PROCEEDINGS
OF THE
California Academy of Sciences
Volume 46
SAN FRANCISCO
PUBLISHED BY THE ACADEMY
1988-1990
PUBLICATIONS COMMITTEE
Tomio Iwamoto, Scientific Editor
Judy Prokupek, Managing Editor
Luis F. Baptista
Thomas F. Daniel
Terrence M. Gosliner
Thomas Moritz
Wojciech J. Pulawski
(US ISSN 0068-547X)
The California Academy of Sciences
Golden Gate Park
San Francisco, California 94 1 1 8
PRINTED IN THE UNITED STATES OF AMERICA
BY ALLEN PRESS, INC., LAWRENCE, KANSAS
CONTENTS OF VOLUME 46
Pages
No. 1. WELSH, HARTWELL H., JR. An ecogeographic analysis of the herpetofauna of
the Sierra San Pedro Martir region, Baja California, with a contribution to the
biogeography of the Baja California herpetofauna. Published December 7,
1988 1-72
No. 2. DANIEL, THOMAS F. Three new species of Hologr aphis (Acanthaceae) from
Mexico. Published December 7, 1988 73-81
No. 3. ANDERSON, M. ERIC, AND ALEX E. PEDEN. The eelpout genus Pachycara (Teleos-
tei: Zoarcidae) in the northeastern Pacific Ocean, with descriptions of two new
species. Published December 7, 1988 83-94
No. 4. UBICK, DARRELL, AND THOMAS S. BRIGGS. The harvestmen family Phalango-
didae. 1 . The new genus Calicina, with notes on Sitalcina (Opiliones: Lama-
tores). Published March 9, 1989 95-136
No. 5. ALMEDA, FRANK. Five new berry-fruited species of tropical American Melas-
tomataceae. Published March 9, 1989 137-150
No. 6. ROBERTS, TYSON R. Systematic revision and description of new species of
suckermouth catfishes (Chiloglanis, Mochokidae) from Cameroun. Published
August 24, 1 989 1 5 1-1 78
No. 7. ROGERS, EDWARD. American dance flies of the Drapetis assimilis species group
(Diptera: Empididae). Published December 20, 1989 179-192
No. 8. GREENE, HARRY W. Ecological, evolutionary, and conservation implications
of feeding biology in Old World cat snakes, genus Boiga (Colubridae). Published
December 20, 1 989 1 93-207
No. 9. ALMEDA, FRANK. New species and taxonomic notes on Mexican and Central
American Melastomataceae. Published December 20, 1989 209-220
No. 10. ANDERSON, M. ERIC. Review of the eelpout genus Pachycara Zugmayer, 1911
(Teleostei: Zoarcidae), with descriptions of six new species. Published Decem-
ber 20, 1989 221-242
No. 1 1 . PARENTI, LYNNE R. A phylogenetic revision of the phallostethid fishes (Ath-
erinomorpha, Phallostethidae). Published December 20, 1989 243-277
No. 12. DANIEL, THOMAS F. New and reconsidered Mexican Acanthaceae. IV. Pub-
lished September 1 1, 1990 279-287
No. 1 3. DANIEL, THOMAS F., AND DIETER C. WASSHAUSEN. Three new species ofJusticia
(Acanthaceae) from Panama. Published September 11, 1990 289-297
No. 14. ALMEDA, FRANK. New species and new combinations in Blakea and Topobea
(Melastomataceae), with an historical perspective on generic limits in the tribe
Blakeeae. Published September 11, 1990 299-326
No. 15. ALMEDA, FRANK. Three new species ofConostegia (Melastomataceae: Micon-
ieae) from southern Central America. Published September 11, 1990 327-335
Index to Volume 46 .. ... 337-346
The Scientific Publications Committee acknowledges with thanks the individuals listed below
who served as manuscript reviewers for this volume.
Paul H. Arnaud, Jr., William C. Burger, Daniel M. Cohen, James C. Cokendolpher, Thomas Croat,
W. G. D'Arcy, L. H. Durkee, Barry E. Hammel, Robert W. Henderson, James Henrickson, Walter
Ivantsoff, Tomio Iwamoto, Vincent F. Lee, Douglas F. Markle, Thomas Morley, Lucinda McDade,
Theodore J. Papenfuss, Walter J. Rainboth, Richard H. Rosenblatt, Jay M. Savage, W. O. Stevens,
Carol Todzia, William J. Turner, Harold K. Voris, Dieter C. Wasshausen, Robert L. Wilbur, Richard
W. Winterbottom, John J. Wurdack.
PROCEEDINGS
OF THE
CALIFORNIA ACADEMY OF SCIENCES
Vol. 46, No. 1, pp. 1-72, 21 figs., 6 tables. December 7, 1988
AN ECOGEOGRAPHIC ANALYSIS OF THE HERPETOFAUNA OF
THE SIERRA SAN PEDRO MARTIR REGION, BAJA CALIFORNIA,
WITH A CONTRIBUTION TO THE BIOGEOGRAPHY OF THE
BAJA CALIFORNIA HERPETOFAUNA
By
Hartwell H. Welsh, Jr.
Pacific Southwest Forest and Range Experiment Station, Forest Service,
U.S. Department of Agriculture, 1700 Bayview Drive,
Arcata, California 95 52 11
ABSTRACT: An ecogeographic analysis of the distributions of the herpetofauna of the Sierra San Pedro
Martir Region of Baja California was undertaken. The Martir Region is part of northern Baja California
from approximately latitude 30°N to 31°N. Over 3,000 locality records from the literature, museums, and field
work were analyzed across a matrix of ecogeographic formations based on regional climates, physiography,
and extant vegetation. Numerical and heuristic methods of biogeographic analysis indicated the 65 species
present occur in seven distinct patterns. Examination of these contemporary patterns within and beyond the
Martir Region led to a hypothesis of five historical patterns of evolutionary development among the Baja
California herpetofauna. This hypothesis indicates that vicariance, dispersal (diffusion), and in situ evolution
are interacting phenomena contributing to contemporary herpetofaunal distributions in peninsular and insular
Baja California and the Pacific Southwest.
Received Sept. 30, 1987. Accepted March 21, 1988.
ences, and concluded that "... continued pro-
INTRODUCTION . . . c ., f
gress in solving the questions of the origins of
There is a growing consensus that ecological biogeographic areas will require biogeographers
factors limit the geographic range of all organ- to clearly distinguish the origins of species from
isms (e.g., Udvardy 1969; MacArthur 1972). Ro- the origins of areas and then develop new meth-
tramel (1973:229) expressed it thusly: ". . . bio- ods of eco-geographic analysis" (Rotramel 1973:
geographic areas originate from and are 230).
maintained by the interactions of organisms with The ecogeographic characteristics of a species'
physical and biotic factors in their environ- distribution can be viewed as the present sum-
ments." He noted a lack of placing proper em- mation of the continuous evolutionary process
phasis on ecological factors in biogeographic of interaction and compromise that occurs be-
analysis, which he attributed to historical influ- tween that species and its environment. A new
adaptation to the environment that promotes
greater survivorship of offspring within an area,
1 Formerly with the U.S. Fish and Wildlife Service, National &
Ecology Center, Dept. of interior, 1 300 Blue Spruce Drive, Ft. or population expansion into new areas, is a new
Collins, Colorado 80524. characteristic of the ecological strategy of that
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
species, and a new chapter in its evolutionary
development. While congruent areas for different
species are theoretically impossible if the areas
are denned strictly in terms of the organisms'
ecological niches, groups of species— both plant
and animal— do coexist and coevolve in loose
congruency in biotic communities. Given that
the ecological parameters of a species are indic-
ative of the historical process through which that
species has evolved, it is logical to assume that
a similarity of ecogeographic constraints among
populations of different species in the same area
may indicate a common history for those species.
This common history may not have entailed di-
rect ecological interactions but nonetheless con-
stitutes a coevolution, where these species' dis-
tributions are similar even when examined on a
scale approaching that of microhabitat. The elu-
cidation of such similar ecogeographic patterns
is the fundamental methodology of historical
biogeography, and their elucidation and inter-
pretation comprises the essence of the discipline.
The remote, mostly inhospitable, and sparsely
populated peninsula of Baja California, pre-
served in a relatively pristine state, is an ideal
natural laboratory for the study of phenomena
that influence and regulate the distributions of
terrestrial vertebrates. The biotic diversity and
general community structure of the flora and fau-
na of the Baja California peninsula was reported
by Nelson (1921). Schmidt (1922) made the first
comprehensive study of the diverse herpeto-
fauna of the Peninsular Region (peninsula and
related islands), describing peninsular distribu-
tions on the basis of Nelson's Faunal Districts.
Savage (1960) presented a classic historical anal-
ysis of the distributions of the peninsular her-
petofauna, adding significantly to our under-
standing of the origins of these distributions and
those of the entire North American herpeto-
fauna. However, Savage's analysis had three
shortcomings: (1) he assumed that the peninsula
had existed as a more or less stable physiographic
unit since the Eocene; (2) he chose not to incor-
porate vital evidence provided by the regional
insular distributions; and (3) despite noting 1 3
different distribution patterns among the pen-
insular herpetofauna, he based his numerical
comparison and subsequent distributional anal-
ysis on only four Herpetofaunal Areas, appar-
ently modeled after Nelson's Faunal Districts.
This resulted in an oversimplification of distri-
bution patterns and relationships, obscuring some
and entirely concealing others. Nelson's (1921)
Faunal Districts and Savage's slightly more re-
fined Herpetofaunal Areas give a general picture
of major herpetofaunal distribution trends on
peninsular Baja California, but under close scru-
tiny, they fail to delineate ecological or geograph-
ic distribution barriers for much of the herpe-
tofauna.
Greater ecological and geographical resolution
is now possible with the extensive taxonomic and
distributional data available from more recent
studies of the regional herpetofauna. Loomis et
al. (1974) compiled an updated checklist and re-
ported the general distributions of the Baja Ca-
lifornia herpetofauna. Linsdale (1932), Tevis
(1944), Murray (1955), Bostic (1971), Murphy
(1975, 1976, 1 983a), Welsh (1976a),Seib( 1980),
and Welsh and Bury (1984) have investigated
community distributions in various parts of the
peninsula. Numerous other authors have con-
tributed to knowledge of distributions and ecol-
ogy for single species of the regional herpeto-
fauna (see species accounts below). However, a
comprehensive ecogeographic study of the entire
Baja California herpetofauna has yet to be con-
ducted. Murphy's (1983a) study is the best at-
tempt to date; using new distributional data from
remote areas of southern Baja California and the
Gulf islands, he defined more accurate southern
peninsular biotic subdivisions for analysis.
Murphy lacked complete and accurate distri-
butional data for some species in the northern
areas; his analysis emphasized the southern pen-
insular patterns. He lumped several northern dis-
tribution patterns together in a "marginal track"
and, I submit, underplayed their overall signif-
icance as evidence of important evolutionary
trends complementing the tectonic events that
shaped the distributions of much of the Baja Ca-
lifornia herpetofauna. The northern peninsular
patterns are important because the north is the
proposed "gate" for species migrating down from
the continent (Savage 1 960), and the majority of
the lineages present in Baja California are rep-
resented there. Its diversity of habitat types,
stemming from its marked elevational gradient,
and from its situation between tropical and tem-
perate latitudes, is the greatest on the peninsula.
This study focuses on the northern area of the
peninsula, specifically on the Sierra San Pedro
Martir Region of north central Baja California
WELSH: BAJA CALIFORNIA HERPETOFAUNA
Sierra San Pedro
Martir Region
Museum of Vertebrate Zoology
University of California
FIGURE 1 . The location of the Sierra San Pedro Martir Region, Baja California Norte, Mexico.
(hereafter referred to as the Martir Region, or
simply the Region). The Martir Region consists
of that area of northern Baja California from
approximately 30°N to 31°N latitude (Fig. 1). It
contains 65 species of reptiles and amphibians.
Four of Nelson's five Faunal Districts and three
of Savage's four Herpetofaunal Areas intersect
within the Region (Fig. 2a). Hastings and Turner
(1965) investigated the climates of Baja Califor-
nia. They correlated their data with the work of
Shreve and Wiggins (1964) on plant distributions
and described six phytogeographic divisions in
Baja California. Three of these vegetation areas
intersect in the Martir Region (Fig. 2b).
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
FIGURE 2. Biotic and climatic subdivisions of the Sierra San Pedro Martir Region, Baja California, Mexico, a) Herpetofaunal
Areas (Savage 1960): (1) Californian; (2) Colorado Desert; (3) Peninsular Desert, (b) Phytogeographic Provinces (after Shreve
and Wiggins [1964] and Hastings and Turner [1965]): (1) San Pedro Martir; (2) Lower Colorado Valley; (3) Vizcaino, (c)
Geographic relationships of the major climatic areas: (1) Pacific Climate; (2) Montane Climate; (3) Gulf Climate; (4) Central
Desert Climate.
My approach was to formulate a base map of
regional habitats with which to compare distri-
butions and seek evidence of patterns across
species. Such evidence was then used to define
potential historical groupings. Independent evi-
dence of congruency among overall distributions
was then sought as corroboration for the group-
ings derived from the regional analysis. Such cor-
roboration I considered as evidence that the re-
gional groupings are each the result of a common
evolutionary history among its members, and
not the result of random processes.
ACKNOWLEDGMENTS
I want to dedicate this work to the memory of
James R. Koplin, a friend and mentor who
showed me the real value of a sense of humor.
The Direction General de la Fauna Silvestre de
Mexico is gratefully acknowledged for providing
me the permits to do this research. My family,
Aida Meling, and the Observatorio Nacional de
Mexico generously provided field logistical sup-
port. C. Olen, A. Szurak, R. Bebout, A. Chakos,
L. Harlow, F. Adams, A. Spolsky, R. Stebbins,
D. Wake, J. Koplin, D. Kitchen, J. Houck, K.
Tollestrup, A. Lind, R. Drewes, and R. B. Bury
provided support, assistance, and encourage-
ment along the way. I am grateful to M. Bogen,
T. Fritts, R. B. Bury, H. M. Smith, R. Etheridge,
T. Papenfuss, K. Tollestrup, H. Greene, M. Ha-
ney, A. Lind, D. Waters, and three anonymous
reviewers for helpful comments on earlier drafts.
I thank C. Houle and R. Nelson for help with
tables and graphics, and L. Mandell-Lewis for
the summary in Spanish. T. Fritts of the San
Diego Natural History Museum, J. Wright of the
Los Angeles County Museum of Natural His-
tory, A. Leviton, R. Drewes, and J. Vindim of
the California Academy of Sciences, and R. Steb-
bins of the Museum of Vertebrate Zoology, Uni-
versity of California, Berkeley, kindly provided
locality records from their respective institu-
tions. I wish to acknowledge the National Ecol-
ogy Center of the U.S. Fish and Wildlife Service,
the Mattole Institute, Humboldt State Univer-
sity, the Museum of Vertebrate Zoology of the
University of California, and the Theodore Roo-
sevelt Memorial Fund for providing support or
facilities that helped bring this project to fruition.
MATERIALS AND METHODS
The following procedures were followed in the
investigations of the distributional patterns of
the herpetofauna of Baja California: (1) deri-
vation of a base map of the ecogeographic for-
mations of the Sierra San Pedro Martir Region
on the basis of climatic data, physiographic fea-
tures, and extant vegetation associations; (2)
WELSH: BAJA CALIFORNIA HERPETOFAUNA
compilation of locality and ecological data for
the 65 herpetofaunal species occurring within the
Martir Region from field investigations, museum
records, and available literature; (3) plotting of
ecogeographic distributions of regional species,
and grouping of species whose distributions are
similar and thus indicative of a possible histor-
ical relationship; (4) assessment of the validity
of such proposed historical groupings by com-
parison of intra-Regional patterns with broader
distribution patterns (in the rest of Baja Califor-
nia and beyond), to seek corroborative evidence
of a common evolutionary history.
Derivation of Regional Ecogeographic
Formations
Heyer (1967:259) described a life zone as "an
ecological altitudinal or latitudinal zone, char-
acterized by specific climate parameters and sec-
ondarily by vegetation." Miller's (1951) "eco-
logic formations" reflected geomorphological
components that sometimes play a greater role
in the distribution of species than do altitudinal
or vegetative parameters. I have incorporated
both these concepts in my division of the Region
into "ecogeographic formations." These forma-
tions were delineated on the basis of both field
data and literature pertaining to physiographic
features, climate, and dominant vegetational as-
sociations (using perennial plant species). Bio-
geographic assemblages generally represent gra-
dients with regard to space, time, and biotic form,
and boundaries between assemblages are there-
fore at best approximate and variable, changing
continuously, and varying somewhat for each
biotic form (Udvardy 1969). Therefore, I have
used ecotones to establish approximate bound-
aries between ecogeographic formations.
Compilation of Species Accounts
The herpetofauna throughout the Martir Re-
gion was surveyed from 22 June to 13 August
1973, from 9 April to 25 June 1974 (except 6
days in late May 1974), for 2 weeks in June 1976,
and 2 weeks in May 1979. Much of the Region
is accessible by road and many areas were reached
by vehicle. In the more remote and inaccessible
central and south central parts of the Region travel
was by foot. A total of 2,187 km by road and
766 km by trail were logged during these inves-
tigations. Approximately 80-90% of the areal ex-
tent of the Martir Region was observed. Eleva-
tion (with a barometric altimeter), physiography,
climate, vegetation, and microenvironmental
conditions were recorded wherever animals
were observed or collected. Distributional, eco-
logical, and behavioral data were recorded on
the 65 species that comprise the herpetofauna
of the Martir Region. Two hundred thirty-one
voucher specimens were collected; now depos-
ited in the Museum of Vertebrate Zoology, Uni-
versity of California, Berkeley. Additional lo-
cality records of captured and released or
observed animals totaled 523. Museum and lit-
erature records used in the study totaled 2,618.
A total of 3,372 locality records were obtained
from 65 collecting sites. The collecting sites are
listed, north to south, in Appendix A, and are
shown in Figure 3. Distributional data from the
following museums were used: California Acad-
emy of Sciences (CAS and SU), San Diego Nat-
ural History Museum (SDNHM), Los Angeles
County Museum of Natural History (LACM),
and the Museum of Vertebrate Zoology, Uni-
versity of California, Berkeley (MVZ). Sources
for literature records are noted in the species ac-
counts. All known localities of museum and lit-
erature records were visited to determine existing
vegetational and physiographic features, and to
note available habitat types. Phylogenetic orga-
nization of families in the species accounts fol-
lows Stebbins (1985). Subspecific nomenclature
was used in all cases where subspecies have been
described; otherwise the currently recognized
species names were used. Common names used
are from Collins et al. (1978) or Stebbins (1985).
Derivation of Regional Distribution Patterns
Two different methods were used to compare
the distributions of regional species. Following
Savage (1960) and Murphy (1983a), numerical
techniques of analysis were used. Peters (1971)
pointed out that although numerical methods in
biogeography are capable of making quantitative
distinction between areas, they cannot distin-
guish whether said areas have a real basis in na-
ture. This problem was addressed by using re-
gional ecogeographic formations as the primary
areas for numerical comparison instead of pre-
established faunal or vegetation provinces or the
equal quadrants technique often used with nu-
merical faunal analyses.
Secondly, a two-phased heuristic method was
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
FIGURE 3. Location of collecting stations in the Sierra San Pedro Martir Region, Baja California Norte, Mexico. Stations
20, 21, 25, 30, 31, 36, 40-43, 51, and 60 have been omitted; these stations are adjacent to the next highest number, e.g., station
20 and 21 are near station 22. Exact locations and elevations of stations are in Appendix A. Topographic intervals equal 610
m (2,000 ft).
used to analyze regional distributions. Phase one
consisted of a comparison of distributions across
a transect of regional ecogeographic formations,
and phase two was a comparison of overall re-
gional distributions as plotted on maps based on
Figures 3, 4, and 5.
My numerical-heuristic strategy of analysis was
similar to the approach of Morafka ( 1 977) except
that he used two numerical techniques, faunal
resemblance coefficients (two formulas) and an
equal quadrants analysis. I used three faunal re-
semblance coefficients and no equal quadrants
analysis. Both the numerical and heuristic meth-
ods are discussed in more detail in the section
on zoogeographic analysis.
Distributions Beyond the Martir Region
Distributional congruency beyond the Martir
Region for those taxa linked ecogeographically
within the Martir Region was considered to be
indicative of possible historical links and a shared
evolutionary history. Evidence for such extra-
regional congruency was sought using indepen-
dent means. The method of establishing and
comparing the combined distributions of related
forms, conspecific or congeneric, or both (indi-
vidual tracks), in order to establish or deny broad
geographical congruency between unrelated
taxa— and thus demonstrate more general dis-
tribution patterns (generalized tracks)— follows
WELSH: BAJA CALIFORNIA HERPETOFAUNA
the principles and method indicated by Croizat
(1964).
RESULTS
THE ENVIRONMENT
The Sierra San Pedro Martir Region refers to
that area of northern Baja California between the
Pacific Ocean, the Gulf of California (Sea of Cor-
tez), and latitudes 3 1°20'N and 30°00'N (Fig. 1),
an area of approximately 1 50 km2. The mountain
range central to the Region, the Sierra San Pedro
Martir, hereafter referred to as the Sierra, is over
3,000 m in elevation, and transects the center of
the Martir Region from north to south. The Col-
orado Desert is east of the Sierra, and foothill
ranges and the San Quintin Plain are to the west.
The Martir Region is at the southern extreme of
the north temperate zone and is influenced by
two diverse climates, the Pacific and Gulf re-
gimes (Hastings and Turner 1965).
Physiographic Features
The topography and geography of the Sierra
San Pedro Martir were described by Nelson
(1921) and Henderson (1960). Woodford and
Harris (1938) described the geology and geo-
morphology of the Region, and Allison (1964)
discussed the topographic relationships of the
Martir Region with the features of surrounding
regions.
Information on the general physiography of
the Martir Region was obtained from topograph-
ic maps (Estados Unidos Mexicanos, Series 501),
flight navigation charts (U.S. Dept. Commerce
CH-22, 1979), and Robinson (1972). Based on
these sources and my field investigations, I sub-
divided the Region into nine primary topograph-
ic elements, each with a unique and relatively
homogeneous physiography (Fig. 4).
The Sierra consists of an elevated, granitic fault
block dated from the mid-Cretaceous (Silver et
al. 1 956). The fault block is a segment of the Baja
California Cordillera, and is considered part of
the Peninsular Range Physiographic Province
(Allison 1964), which extends from the Trans-
verse Ranges of southern California (at the lat-
itude of Los Angeles) south along the length of
Baja California.
The northern extremity of the Sierra is at San
Matias Pass (980 m), which is probably a trans-
verse fault (Allen et al. 1956). The southern ex-
tremity of the Sierra is in the vicinity of Cerro
Matomi (30°24'N, 1 1 5°10'W), south of which the
Sierra gradually decreases in elevation and is
covered by lava-capped mesas. Here, the moun-
tains grade into the North Central Desert. The
area from Cerro Matomi south to 30°N latitude
was designated the North Central Desert To-
pographic Element (Fig. 4).
The eastern slope of the Sierra, designated the
East Scarp Topographic Element (Fig. 4), is a
steep and fairly straight fault scarp rising 1,220-
2,120 m above the desert floor. A series of steep
canyons transect this scarp, and terminate in
piedmont alluvial fans, opening onto Valle de
San Felipe and Valle Chico, two continuous des-
ert valleys paralleling the scarp base at about 600
m elevation. These valleys are part of, and con-
tiguous with, the southern Colorado Desert. The
Colorado Desert extends from California and
Arizona, south along the Gulf coast of Baja Cali-
fornia to the vicinity of Bahia de Los Angeles
(29°00'N, 1 13°15'W). This desert from 31°20'N
latitude south to Bahia de Los Angeles was des-
ignated the South Colorado Desert Topographic
Element (Fig. 4). All topographic elements from
the ridge of the East Scarp, east to the Gulf of
California, comprise the Gulf Slope.
The west slope of the Sierra consists of a series
of lesser fault scarps with intervening steps be-
tween that drop more gradually than the East
Scarp. They range from 915 m to 1,525 m. This
area is designated the West Scarp Topographic
Element. Below the West Scarp Topographic Ele-
ment is a series of foothill ranges that drop grad-
ually to a coastal plain. The foothill ranges and
the coastal plain are designated the Western
Foothills, and the San Quintin Plain Topograph-
ic Elements, respectively (Fig. 4).
The corrugated Main Scarp of the Sierra San
Pedro Martir is 13-16 km wide, 64 km long, and
is bordered and transected by irregular, boulder-
stacked ridges. The Main Scarp contains four
major, and numerous minor, alluvial-filled,
meadow expanses; several of these meadows
contain year-round surface water. The Main Scarp
was subdivided into the Upper Main Scarp To-
pographic Element, consisting of those areas north
of Tasajera Ridge and above 2,300 m (7,550 ft),
and the Lower Main Scarp Topographic Ele-
ment, consisting of those areas south of Tasajera
Ridge and below 2,300 m (Fig. 4). Tasajera Ridge
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
FIGURE 4. Geographic relationships of the major topographic elements of the Sierra San Pedro Martir Region, Baja California
None, Mexico. (1) San Quintin Plain; (2) Western Foothills; (3) West Scarp; (4) Upper Main Scarp; (5) Lower Main Scarp; (6)
East Scarp; (7) South Colorado Desert; (8) North Central Desert; (9) Western Arroyos.
transects the Main Scarp at about midlength, and
runs west to east at 30°57'N latitude.
The Pacific Slope is west of the crest of the
Sierra San Pedro Martir and north of the North
Central Desert Topographic Element (Fig. 4). It
is transected by a system of westerly directed
canyons, carrying partly subterranean runoff from
the Sierra, 72 km west to the Pacific. These ar-
royos were designated the Western Arroyos To-
pographic Element (Fig. 4). In subsequent sec-
tions, the suffix "Topographic Element" has been
omitted, and the capitalized geographic entity is
used to designate these topographic divisions (e.g.,
West Scarp Topographic Element = West Scarp).
Regional Climates
The climate of the Martir Region results from
the interaction of regional topography with two
major climatic regimes that govern weather con-
ditions over most of Baja California. The cli-
matic regimes consist of the relatively cool, moist
Pacific Coastal Regime, produced by the cold
California Current with its prevailing westerly
winds, and the considerably warmer drier Gulf
Regime (Hastings and Turner 1965; Meigs 1966).
Hastings and Turner (1965) postulated that
seasonal differences in variability and amount of
precipitation produce the distinctive vegetation-
WELSH: BAJA CALIFORNIA HERPETOFAUNA
al associations of their phytogeographic prov-
inces (Fig. 2b). Lacking quantitative data, they
lumped the Sierra San Pedro Martir with all of
northwestern Baja California in the San Pedro
Martir Phytogeographic Province. My observa-
tions within this province indicated more com-
plex vegetational relationships. I found this mon-
tane area to have a climate and vegetation
sufficiently unique to warrant separate consid-
eration.
The Sierra has predictable winter and spring
precipitation, often falling as snow in the winter
(J. Alonso, Mexican National Observatory, pers.
comm.). A summer rainfall pattern results from
a thermal low pressure cell to the east of the
Sierra, causing moist air to flow east across the
mountains. Afternoon thunder showers occur al-
most daily from late June through September as
a result of orographic uplift. Data from San Juan
de Dios (elevation 1 ,400 m) in the Sierra Juarez,
1 44 km north of the Sierra San Pedro Martir,
show the highest annual precipitation (33.4 cm)
yet reported for northern Baja California (Has-
tings and Humphrey 1969). The Main Scarp of
the Sierra San Pedro Martir is considerably higher
in elevation than the Sierra Juarez and thus prob-
ably receives the highest annual precipitation in
Baja California.
The precipitation percolates down through the
granitic sand and the soil surface dries rapidly in
the warm summer air. Flash-flooding from un-
usually heavy rainfall occurs occasionally on all
scarps. Lightning regularly accompanies these
storms and is responsible for fires on the forested
plateau. Underbrush is scarce throughout much
of the forest, perhaps due to fires.
The Main Scarp, and the upper reaches of the
West and East scarps, areas influenced by the
high elevation weather system of the Sierra, have
been designated the Montane Climatic Area (Fig.
2c).
Those areas on the west side of the Sierra San
Pedro Martir, encompassing the lower elevations
of the West Scarp, the Western Foothills, and the
Western Arroyos were designated the Pacific Cli-
matic Area (Fig. 2c). This area is characterized
by frequent fog, cool oceanic temperatures, and
relatively predictable winter and spring precip-
itation (Table 1 , Western Foothills).
Those areas on the east slope of the Sierra San
Pedro Martir, and corresponding with the south-
ern portion of the Lower Colorado Valley Phy-
togeographic Province (Table 1), were designated
the Gulf Climatic Area; it encompasses the lower
East Scarp and the South Colorado Desert To-
pographic Elements (Fig. 2c; Table 1). This area
is one of the hottest places in the world (Meigs
1953) and is characterized by a lack of predict-
able rainfall. Shreve (1934) remarked that this
area may be the most sparsely vegetated of any
in North America.
The area designated the Central Desert Cli-
matic Area corresponds with the northern por-
tion of the Vizcaino Phytogeographic Province
(Table 1), and encompasses the North Central
Desert Topographic Element (Fig. 2c; Table 1).
This area has the driest summers of the four
climatic areas. The lack of rainfall in the Central
Desert is mitigated by cool Pacific air and fre-
quent fogs, which provide considerable ground
moisture. Bostic (1971) described the climate of
the Central Desert in greater detail.
The narrow San Quintin Plain Topographic
Element, extending northward to the vicinity of
latitude 30°20'N, was included in the Central
Desert Climatic Area (Fig. 2c). Hastings and
Turner (1965) included this area with north-
western Baja California, placing it in their San
Pedro Martir Phytogeographic Province. Cli-
matic data from this coastal plain were analyzed
and compared with similar data from adjacent
topographic elements to the northeast and south
(Table 1). The climate of the San Quintin Plain
falls between that of the two adjacent areas, but
seasonal precipitation patterns are most similar
to the Central Desert Climatic Area. Floral and
faunal characteristics of the San Quintin Plain
indicate that this area, and the proximate inland
valleys, are within the influence of the Central
Desert Climate (Short and Crossin 1967; Bostic
1968; this study).
References to the four climatic areas of the
Martir Region will use this format: Gulf Climatic
Area = Gulf Climate.
Vegetative Elements
Following the approach of a similar regional
study conducted by Schoenherr ( 1 9 7 6) in the San
Gabriel Mountains of southern California, the
structure and distribution of Martir Region vege-
tation communities were determined. Nelson
(1921) reported on an extensive natural history
survey of Baja California, and described vege-
10
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
TABLE 1 . SUMMARY OF CLIMATIC DATA FROM THE SIERRA SAN PEDRO MARTIR REGION, BAJA CALIFORNIA NORTE, MEXICO;
REPORTED BY PHYTOGEOGRAPHIC PROVINCE AND TOPOGRAPHIC ELEMENT (Fie. 4).
Geographic area
Winter
Spring
Summer
Fall
Annual
Number of
weather
stations'
Phytogeographic province2
Lower Colorado Valley
Vixcaino
San Pedro Martir
Topographic clement3
South Colorado Desert
San Quintin Plain
North Central Desert
Western Foothills
Sierra Scarps4
Mean precipitation (cm)
1.9
5.6
9.2
2.3
8.1
5.5
9.7
0.8
1.0
4.1
0.7
2.8
1.3
3.5
1.4
1.1
1.8
1.3
1.6
0.4
4.0
2.0
2.8
2.7
2.0
2.1
2.1
4.3
Mean temperature (°C)
6.1
10.5
17.8
6.5
13.5
9.5
18.9
(33.4)
13 +
27
Topographic element
South Colorado Desert
15.0
21.8
31.8
24.8
23.2
2
San Quintin Plain
13.2
15.2
19.3
18.0
16.4
4
North Central Desert
13.4
16.3
23.3
19.5
19.1
3
Western Foothills
11.9
16.6
22.7
18.7
17.2
7
Sierra Scarps
(no data)
0
1 See Hastings and Humphrey (1969) for exact locations.
2 Data from Hastings and Turner (1965).
3 Mean precipitation and temperature data were summarized from weather stations within each topographic element; data
are from Hastings and Humphrey (1969).
4 Data from Sierra Juarez; see text.
tational characteristics and listed dominant
species in accordance with Merriam's (1 898) life
zones. Wiggins (1944) surveyed the Sierra San
Pedro Martir and surrounding lowlands, Shreve
(1936) described the complex vegetation west of
the Sierra, and Chambers (1955) described the
flora of a canyon on the East Scarp.
Seven vegetational associations are recognized
within the Martir Region (Fig. 5): Coniferous
Forest, Chaparral, Pinyon-Juniper Woodland,
Creosote Bush Scrub, Coastal Sage Scrub, Cen-
tral Desert Scrub, and Riparian Woodland. These
vegetational associations are modeled after the
communities of Munz and Keck (1949, 1959)
except where noted below. The dominant species
of each association are listed in Table 2.
Coniferous Forest
Coniferous Forest vegetation occurs above
1,830 m on the Sierra Main Scarp, within areas
influenced by the Montane Climate (Fig. 5, 6).
The forest is open and parklike, with sparse un-
dergrowth of low shrubs, forbs, and grasses.
At higher elevations (over 2,400 m), vegeta-
tional elements characteristic of Merriam's
(1898) Canadian and Hudsonian zones (lodge-
pole pine [Pinus murrayana], white fir [Abies
concolor], and quaking aspen [Populus tremu-
loides]) are intermixed with the dominant ele-
ments of Coniferous Forest. This impoverished
boreal vegetation is best developed in deep can-
yons and on north-facing slopes of the Upper
Main Scarp at the north end of the Sierra.
Chaparral
Chaparral covers extensive areas of the Pacific
Slope of the Sierra, between 1,220 m and 2,120
m, and occurs in a narrow belt on the East Scarp
between 1,670 m and 2,120 m (Fig. 5, 7). The
dense and often impenetrable vegetation is com-
monly 2.5-3.0 m in height. Distribution of Chap-
arral on both versants corresponds with the steep
scarp faces where the Montane Climate inter-
grades with the drier and warmer Pacific and
Gulf climates.
Within the Chaparral, and particularly in
shaded canyons and areas with surface and sub-
surface water, are patches of oak woodland. Oak
WELSH: BAJA CALIFORNIA HERPETOFAUNA
11
miles 7
6 km II
Creosote Bush Scrub
Pinyon-Juniper Woodland
Coastal Sage Scrub
Chaparral
Coniferous Forest
Central Desert Scrub
FIGURE 5. Vegetational associations of the Sierra San Pedro Martir Region, Baja California Norte, Mexico. Map is based
on U.S. Dept. Commerce, CH-22 (1979). Distribution of vegetation is based on Shreve and Wiggins (1964), Short and Crossin
(1967), and field investigations. Riparian Woodland Association occurs in all other vegetation types.
woodland in the Sierra San Pedro Martir occurs
primarily as riparian habitat, and is considered
with the Riparian Woodland association.
Pinyon-Juniper Woodland
This vegetational type occurs between 920 m
and 1,830 m on the East Scarp of the Sierra (Fig.
5, 8); this distribution coincides with areas of
higher elevation dominated by the Gulf Climate.
Being high on the east versant of the Sierra, Pin-
yon-Juniper Woodland is probably also influ-
enced to some degree by the Montane Climate.
The trees in the Pinyon-Juniper Woodland vary
from approximately 3.0 to 10.0 m in height, and
are widely dispersed. Open shrub growth occurs
throughout the Woodland. This association ex-
tends around the southern and northern ends of
the range above 920 m and intergrades with
Chaparral.
12
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
TABLE 2. DOMINANT SPECIES OF THE VEGETATIONAL ASSOCIATIONS OF THE SIERRA SAN PEDRO MARTIR REGION, BAJA CA-
LIFORNIA NORTE, MEXICO.
Vegetation association
Dominant species
Coniferous Forest
Coastal Sage Scrub
Pinyon-Juniper Woodland
Central Desert Scrub
Chaparral
Creosote Bush Scrub
Riparian Woodland
(Species composition var-
ies greatly with elevation
and slope; abbreviations
refer to correspondence
with above associations.)
jeffery pine (Pi nus Jeffrey i), sugar pine (P. lambertiana), incense-cedar (Libocedrus decurrens),
rose sage (Salvia pachyphylla), greenleaf manzanita (Arctostaphylos patula), pink-bracted
manzanita (A. pringlei), snow bush (Ceanothus cordulatus), scrub oak (Quercus spp.)
flattop buckwheat (Eriogonumfasciculatum), California sagebrush (Artemisia californica),
white sage (Savia apiana), century plant (Agave shawii), prickly pear (Opuntia littoralis),
coastal cholia (Opuntia prolifera), Ephedra californica
pinyon pine (Finns monophylla and P. quadrifolia), California juniper (Juniperus californica),
Mojave yucca ( Yucca schidigera), barrel cactus (Ferocactus acanthodes), Agave aurea, scrub
oak (Quercus turbinella), sugar bush (Rhus ovata)
datilillo (Yucca valida), elephant tree (Pachycormus discolor), century plant (Agave shawii},
cardon (Pachycereus pringlei), cholia (Opuntia spp.), Franseria chenopodiafolia, Ocotillo
(Fourquiera splendens), Dudleya spp.
bigberry manzanita (Arctostaphylos glauca), Mexican manzanita (A. pungens), chamise (Ade-
nostoma fasciculatum), ceanothus (Ceanothus leucodermis and C. greggi), toyon (Hetero-
meles arbutifolia), red shank (Adenostoma sparifolium), scrub oak (Quercus dumosa), sugar
bush (Rhus ovata)
creosote bush (Larrea tridentata), white bursage (Franseria dumosa), ocotillo (Fourquieria
splendens), mesquite (Prosopis juliflora), agave (Agave spp.), palo verde (Cericidium micro-
phyllum), yucca (Yucca spp.), ironwood (Olneya tesota), cholia (Opuntia spp.), catclaw
(Acacia greggi), cardon (Pachycereus pringlei)
willow (Salix spp.), wild rose (Rosa californica), jeffery pine (Pinus jefferyi), incense-cedar
(Libocedrus decurrens) (CF); Oak (Quercus spp.), coffeeberry (Rhamnus californica) (Chp);
Cottonwood (Populus fremonti) (P-J, CSS); Sycamore (Platamus racemosa), ceanothus (C.
leucodermis and C. greggi (CSS); mesquite (Prosopis juliflora), desert willow (Chilopsis lin-
earis) (CBS and CDS); fan palm (Washingtonia filifera) (CDS)
Creosote Bush Scrub
Creosote Bush Scrub occurs below 1 ,070 m on
the eastern side of the Sierra, in the influence of
the Gulf Climate (Fig. 5, 9). The physiognomy
of this vegetation is predominantly shrub, 0.7-
4.8 m tall, and widely dispersed. Creosote Bush
Scrub vegetation becomes more luxuriant along
the base of the East Scarp, and to a lesser degree
along the immediate Gulf coast of the South Col-
orado Desert. Qualitative observations suggest
more plant species occur in these two marginal
parts of the South Colorado Desert and the shrubs
and cacti attain greater size here than in much
of the central area of the desert. This phenom-
enon may be a result of increased surface or sub-
surface moisture, as Gulf Slope runoff (from the
East Scarp and South Colorado Desert), which
is primarily subterranean, is nearest the surface
in these areas.
Fieldwork in the South Colorado Desert was
restricted primarily to the basal slopes of the
Sierra, and the two proximate desert valleys, Valle
de San Felipe and Valle Chico. The Creosote
Bush Scrub classification is expanded to include
the entire South Colorado Desert, south to lati-
tude 29°N, following Shreve and Wiggins (1 964).
At the north end of the Sierra, Creosote Bush
Scrub extends over San Matias Pass (980 m) into
Valle de Trinidad, where it intergrades with
Coastal Sage Scrub. South of the Sierra San Pedro
Martir (beyond 30°25'N), Creosote Bush Scrub
intergrades with Central Desert Scrub along the
Baja California Cordillera.
Coastal Sage Scrub
The Pacific Slope below 1 ,220 m, from slightly
beyond the north limit of the Martir Region,
south to the vicinity of latitude 30°15'N, is vege-
tationally complex. Shreve (1936) described this
area from Arroyo Santo Tomas to Arroyo So-
corro as being within a 10-mi transition zone
(from north to south) between Chaparral and
Sonoran Desert vegetation. However, the clas-
sification Coastal Sage Scrub was adopted as best
reflecting the life-forms and species composition
that dominate most of this area (Fig. 5, 10). The
WELSH: BAJA CALIFORNIA HERPETOFAUNA
13
FIGURE 6. The Lower Main Scarp Ecogeographic Formation at Rancho Viejo (station 37) in Coniferous Forest adjacent
montane meadow.
plants are mostly "half-shrubs, one to five feet
tall or somewhat woodier and larger, forming a
more open community than chaparral" (Munz
and Keck 1949:97). There are numerous loca-
tions along the lower Pacific Slope of the Region
where elements of desert vegetation dominate.
The most extensive of these areas is the San
Quintin Plain (Fig. 1 1) and the valleys and low
slopes immediately to the east. The Chaparral
occurring on the lower Pacific Slope, with the
exception of scattered stands at high elevations
in the Western Foothills, is mostly dwarfed and
depauperate. The desert scrub and Chaparral
complexes are not the primary associations of
the lower Pacific Slope.
Central Desert Scrub
The scrub vegetation changes noticeably in the
vicinity of latitude 30°15'N on the Pacific Slope.
The Creosote Bush Scrub of the lower East Scarp
and Colorado Desert drainages, the Coastal Sage
Scrub of the Western Foothills, and the Chap-
arral of the southern Sierran slopes interdigitate
with a Sonoran vegetation characterized by sar-
cophyllous forms. Wiggins (1960) described this
area as part of the Central Desert Phytogeograph-
ic Area of the Vizcaino Province. Following the
work of Aschman (1959), Wiggins (1960), and
Bostic (1971), I also considered this area to have
a unique biota. The plant association of this area
is referred to as Central Desert Scrub in order to
locate it geographically and to indicate its close
structural relationship to the aforementioned
scrub vegetational associations (Fig. 5, 12).
This classification is applicable to the vege-
tation of the San Quintin Plain, to the vicinity
of latitude 31°20'N. Central Desert Scrub vege-
tation is denser and more lush on the San Quintin
Plain than in the North Central Desert. Sarco-
phyllous forms dominate in many areas here,
often occurring in impenetrable stands. Bostic
(1971) noted that many of these plants are adapt-
ed structurally to collect moisture from the air
(Yucca valida, Agave spp., Dudleya spp.). The
regular fog and dew along this Plain probably
provide a major source of water for these plants
and explain the rich vegetation here.
Riparian Woodland
This plant association (Fig. 13) is based pri-
marily on similarities of growth form and phys-
ical setting rather than simply on associations of
14
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
FIGURE 7. The West Scarp Ecogeographic Formation in Arroyo San Rafael (station 1 0). The vegetation is primarily Chaparral,
with elements of Pinyon-Juniper Woodland and Riparian Woodland evident on the lower slopes and in the canyon bottom,
respectively. The Upper Main Scarp Ecogeographic Formation is in the background.
plant species. Each of the previously described
vegetational associations contains a riparian ele-
ment, with only a few species common to all
riparian elements. The species vary considerably
relative to elevation, however the presence of
closed overstory, dense understory thickets, and
lush, moist ground vegetation is common to the
Riparian Woodland at all elevations of the Mar-
tir Region. The protective cover and microcli-
matic conditions provided by riparian areas are
important for many of the reptiles and amphib-
ians in the Region.
The most prevalent plant species in riparian
associations are willows (Salix spp.). They com-
pose most of the vegetation in the dense under-
story thickets characteristic of the riparian cor-
ridors, and they are also a major component of
the overstory. In proximity to Creosote Bush
Scrub and Central Desert Scrub, willows are rare
and are generally replaced by mesquite (Prosopis
juliflord) and desert willow (Chilopsis linearis).
Grasses, sedges, and herbs form a dense mat
of streamside meadow throughout riparian as-
sociations. This meadow is most highly devel-
oped within the Coniferous Forest, where exten-
sive areas of wet meadow occur with little or no
overstory. In scrub associations, meadow vege-
tation is reduced to narrow strips at streamside,
rarely more than 1 m wide.
Ecogeographic Formations
On the basis of the physiographic, climatic,
and vegetational data presented above, I have
divided the Martir Region into nine component
ecosystems or ecogeographic formations (Fig. 14).
Boundaries of the nine topographic elements
closely approximate the ecotonal boundaries of
the ecogeographic formations. For practical pur-
poses, these boundaries will subsequently be
treated as synonymous and each ecogeographic
formation will be identified by its topographic
component (Fig. 4). For the zoogeographic anal-
ysis, specimens were recorded first by vegeta-
tional association and secondarily by geographic
locale. Thus, a specimen present in Chaparral on
the Lower Main Scarp was scored in ecotonal
vegetation on the Lower Main Scarp Ecogeo-
graphic Formation.
From west to east and from north to south,
the nine ecogeographic formations of the Sierra
San Pedro Martir Region are: San Quintin Plain
WELSH: BAJA CALIFORNIA HERPETOFAUNA
15
FIGURE 8. The East Scarp Ecogeographic Formation in upper Canon El Cajon (near station 5 1) in Pinyon-Juniper Woodland.
Ecogeographic Formation (northwest) (Fig. 11),
North Central Desert Ecogeographic Formation
(southwest) (Fig. 1 2), Western Foothills Ecogeo-
graphic Formation (Fig. 10), Western Arroyos
Ecogeographic Formation (Fig. 1 3), West Scarp
Ecogeographic Formation (Fig. 7), Upper Main
Scarp Ecogeographic Formation (north) (see Fig.
7), Lower Main Scarp Ecogeographic Formation
(south) (Fig. 6), East Scarp Ecogeographic For-
mation (Fig. 8), and South Colorado Desert Eco-
geographic Formation (Fig. 9).
SPECIES ACCOUNTS
(see Appendix A for a list of collecting stations)
CAUDATA
Plethodontidae
Batrachoseps pacificus major (Camp, 1915),
Garden Slender Salamander
Batrachoseps pacificus ssp.
San Pedro Martir Slender Salamander
Batrachoseps pacificus major is known in
northwestern Baja California south to Arroyo El
Rosario (K. Yanev, pers. comm.). It is known
from Colonett (LACM 3433 1-35) and from near
Santo Tomas (MVZ 55102-12). Other localities
are more common northward.
16
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
FIGURE 9. The South Colorado Desert Ecogeographic Formation near the Junction of Mexico Hwy 3 and 5 in Creosote
Bush Scrub vegetation.
An undescribed subspecies of B. pacificus (see
Yanev 1978) is known locally only from the up-
per scarps of the Sierra San Pedro Martir down
to 1,785 m elevation.
Brame and Murray (1968) speculated that Ba-
trachoseps pacificus major and Batrachoseps pa-
cificus ssp. are sympatric in the Sierra San Pedro
Martir. I found no records of B. p. major in the
Martir Region, east of Santo Tomas, 20 km from
the Pacific coast. Apparently suitable habitat for
Batrachoseps occurs inland along the arroyos of
the Western Foothills, but surface moisture in
these arroyos is seldom permanent, or depend-
able.
I collected a gravid female of Batrachoseps pa-
cificus ssp. on 24 June 1973 at station 25 (MVZ
140701). Two individuals were observed be-
neath a log on moist sand in a clump of willow
on 21 June 1974 at station 36. Other specimens
are known from Arroyo Encantada (station 4 1 ;
CAS 57219-32), Encantada Meadow (SDNHM
4 1 32-36), and La Grulla (SU 1 2943). All of these
localities correspond with riparian woodland
habitat of the coniferous forest. B. pacificus ssp.
also occurs in riparian habitat in the chaparral
on the West Scarp. I collected two specimens
inside rotting logs in oak woodland on 1 1 April
1974 near station 29 at 1,785 m elevation (MVZ
140702-03); a series from La Zanja Creek (SU
12933-42) is from similar habitat.
SALIENTIA
Pelobatidae
Scaphiopus couchii Baird, 1854,
Couch's Spadefoot
Wasserman (1970) indicates that Scaphiopus
couchii is restricted to the Colorado Desert, and
southward along the Gulf Slope to the Cape Re-
gion of southern Baja California. He lists no rec-
ords on the Pacific slope of Baja California north
of the Magdalena Plain. A specimen collected on
Hwy 1, 5.3 km south junction of Hamilton Ranch
Road, Valle de San Quintin (CAS 94808) and
reported by Welsh (1 91 6a) as possibly S. couchii,
is actually a juvenile S. hammondii.
Scaphiopus hammondii Baird, 1854,
Western Spadefoot
This toad is known from several localities on
the Pacific Slope of Baja California north of the
WELSH: BAJA CALIFORNIA HERPETOFAUNA
17
FIGURE 10. The Western Foothills Ecogeographic Formation 12 km west of Rancho San Jose (station 27) in Coastal Sage
Scrub vegetation.
Martir Region. Within the Martir Region, Lins-
dale (1932) reports it from Rancho San Jose (sta-
tion 27) in the Western Foothills; it is known
also from the northern San Quintin Plain at San
Antonio del Mar (SDNHM 42095-96), 32 km
south of San Vicente (LACM 90967-68), along
Mexican Hwy 1, 5.3 km south junction of Ham-
ilton Ranch Road, Valle de San Quintin (CAS
94808), and 10 km south of San Quintin
(SDNHM 41404). These are areas of coastal sage
and central desert scrub.
Bufonidae
Bufo boreas halophilus Baird and Girard, 1 852<2,
California Toad
This toad is abundant in the riparian meadows
and woodlands of the coniferous forest, occur-
ring up to 2,800 m in or near riparian areas of
the entire Pacific Slope; it is uncommon in the
Western Foothills and in coastal areas. It fre-
quents areas near standing or slow-flowing water.
18
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
FIGURE 1 1. The San Quintin Plain Ecogeographic Formation between Mexican Hwy 1 and San Antonio Del Mar (station
9). The vegetation is Central Desert Scrub with elements of Coastal Sage Scrub.
Its southernmost locality is El Rosario (CAS
136714-15).
Adults are primarily nocturnal; diurnal surface
activity in the summer was observed on the Main
Scarp, but mostly during afternoon cloud cover
or showers. Newly metamorphosed juveniles,
common in late spring in the wet meadows of
the Main Scarp, were continuously active, often
in direct sun. Two incidences of cannibalism were
observed among these juveniles.
Adult specimens were collected at stations 3
(MVZ 140712), 26 (MVZ 140704-05), 37 (MVZ
140706-07), and 43 (MVZ 140710); adults were
observed at stations 27, 29, 44, 47, 57, 58, and
62. Juveniles were collected 29 June 1973 at sta-
tion 43 (MVZ 140708-09) and 23 July 1973 at
station 15 (MVZ 14071 1). They were observed
between these dates at stations 12, 15, 27, 39,
and 43. Larvae were observed or collected 25
June 1973 at station 22 (MVZ 140727) and 31
July 1973 at station 15.
Bufo microscaphus calif or nicus (Camp, 1915),
Arroyo Toad
This toad is uncommon on the Lower Main
Scarp in riparian habitats of the coniferous for-
est; 2,300 m was the highest recorded elevation.
WELSH: BAJA CALIFORNIA HERPETOFAUNA
FIGURE 12.
vegetation.
The North Central Desert Ecogeographic Formation at San Juan de Dios (station 64) in Central Desert Scrub
In riparian habitats in the chaparral of the West
Scarp, they are common and also abundant in
riparian habitats in coastal sage scrub at the West
Scarp base. Tevis (1 944) found this species at the
Hamilton Ranch (Rio Santo Domingo) on the
San Quintin Plain; this apparently is the south-
ernmost record in Baja California.
In the San Bernardino Mountains of California
where B. microscaphus is sympatric with B. bo-
reas, the latter species occurs in cooler, moister
habitats (Cunningham 1962). My observations
of the microhabitat affinities of B. boreas and B.
microscaphus in the Martir Region concur with
those reported by Cunningham. Additionally, I
found B. microscaphus only in or near fast-flow-
ing water, whereas B. boreas occurred only in or
near standing or slow-flowing water. It is appar-
ently not known whether these two species of
Bufo deposit their eggs in different habitats or
their larvae show different habitat preferences.
Larvae of both species were found together in
several streams of moderate, steady flow (sta-
tions 15, 37, 43, and 57). A pair of B. micro-
scaphus was observed in amplexus in fast-flow-
ing water at 0905 on 3 May 1974 at station 10.
Adults were active at night; toads were twice
20
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
FIGURE 13. Riparian Woodland vegetation in Arroyo San Rafael at station 14.
observed far from water, in coastal sage scrub
and chaparral. On three occasions, they were ob-
served returning to riparian habitat at dawn.
Adults were collected at stations 14 (MVZ
140719), 43 (MVZ 140713, 140738), 49 (MVZ
140715-18, 1475 12), and 56 (MVZ 1407 14). Ju-
veniles were collected or observed 13 July 1973
at station 57 (MVZ 140739) and 3 August 1973
at station 49. Larvae were observed 1 3 July 1973
at station 43 and 31 July 1973 at station 15.
Bufo punctatus Baird and Girard, 18520,
Red-spotted Toad
I observed this toad in riparian habitats with
flowing water at three localities on the East Scarp
surrounded by rocky creosote bush scrub habitat.
Three specimens are known from the Main Scarp
at La Grulla (SU 12945-47). The species also is
known from the Sierra Juarez (CAS 1 2 1 20 1-09).
Bostic (1971) reported the species in the Central
Desert.
Adults were observed or collected at stations
17, 46 (MVZ 140720, 140747), and 60; juveniles
were collected on 24 April 1974 at station 17
(MVZ 140748). Larvae were observed or col-
lected 14 April 1974 at station 46 (MVZ 140740)
and 1 6 April 1 974 at station 59. Larvae also were
observed on 29 April 1974, 1 km above sta-
tion 4.
Hylidae
Hyla cadaver ina Cope, 1866,
California Treefrog
I found this frog in riparian woodland through-
out the Sierra, but it was rare on the Main Scarp;
the highest elevation recorded was 2,300 m. The
species is common on the West and East scarps
and in the arroyos of the Western Foothills. Adult
males were heard chorusing at station 64, a spring
in the North Central Desert. Duellman (1970)
reported Hyla cadaverina from 32 km east of El
Rosario. South of the Martir Region, it is known
from Bahia de Los Angeles (LACM 1 305, 1 3 1 63-
69), 4.8 km N of Rancho Catavina (LACM
1 3 1 73-95), and from Rancho Santa Ynez (LACM
107923-26, 1 13730-31); these localities are the
southernmost records for the species in Baja Ca-
lifornia.
These treefrogs were usually found among steep
WELSH: BAJA CALIFORNIA HERPETOFAUNA
21
PACIFIC CLIMATE-
MONTANE
CLIMATE
•GULF CLIMATE
CENTRAL DESERT CLIMATE
2800m -
1400m -
San
QuintiYi
Plain
West Foothills
West Arroyos
Upper &
Lower
Main
rarp I
East
Scarp
Colorado Desert
Central
Desert
Scrub
Coastal Sage
Scrub
: — ecotones —
Chaparral
Confer
Forest
Chaparral
Pinyon-Juniper
Creosote Bush
Scrub
0
75 km
100
125
150
FIGURE 14. The ecogeographic formations of the Sierra San Pedro Martir Region, Baja California Norte, Mexico at latitude
3 1°N. An ecogeographic formation is comprised of a topographic element (Fig. 4), a climatic element(s) (Fig. 2c, Table 1), and
a vegetation association (Fig. 5. Table 2). Riparian Woodland Associations are present within all vegetations and form the
primary vegetation in the Western Arroyos. A ninth ecogeographic formation, the North Central Desert, occurs south of latitude
31°N; it is continuous with the San Quintin Plain Ecogeographic Formation and contains similar vegetation.
granitic boulders and cliff facings above fast-
flowing water. Cunningham (1964) observed the
aquatic garter snake, Thamnophis hammondii,
eating this frog. The substrate preference of this
frog may be related to the predatory habits of T.
hammondii, which was twice observed attempt-
ing to climb streamside rock faces in the near
vicinity of perched treefrogs. On two occasions
I observed these frogs on sand, gravel, and aquat-
ic vegetation at streamside. These localities were
isolated riparian sites created by springs on the
lower East Scarp; T. hammondii was not ob-
served at these localities (above stations 4 and
46). Boulders and rock facings were present in
riparian habitat at these localities.
Groups of 5-30 individuals were commonly
observed congregating in direct sunlight in the
mist above cataracts and waterfalls in the after-
noon. Individuals moved back and forth from
shade to sunlight, perhaps for thermoregulation.
Adults were collected at station 35 (MVZ
140746), 41 (MVZ 140721), 49 (MVZ 140729-
31), and 56 (MVZ 140722, 140724-25); they were
observed at or near stations 4, 10, 16, 17, 20,
21, 23, 27, 46, 60, and 64. Juveniles were col-
lected 13 July 1973 at station 56 (MVZ 140723);
larvae were collected on 14 April 1974 at station
46 (MVZ 140741), and were observed through
29 April 1974 at stations 4, 16, 20, 21, 23, and
59.
Hyla regilla hypochondriaca Hallowell, 1854,
Pacific Treefrog
Hyla regilla hypochondriaca ranges to 2,750
m in riparian habitat in the coniferous forest,
22
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
and was found in the riparian woodland in the
Arroyos of the West Scarp and Western Foot-
hills, west to the Pacific coast. Bostic (1971) re-
ported H. r. desert icola (=hypochondriaca) from
two localities in the Central Desert.
All Hyla regilla were observed near standing
and slow-flowing water. Most animals were in
aquatic vegetation at streamside or floating in
the water. H. regilla was rarely observed more
than 1 5 cm above the ground or more than 1 m
from water. This contrasts with H. cadaverina,
which was found in areas of fast water flowing
over granite boulders where the frogs inhabited
rock surfaces often 2-5 m above the water. These
two treefrog species seldom occurred in sym-
patry within the Region as a result of the unique
riparian microhabitats occupied by each; I found
them together at a single locality (station 27). H.
r. hypochondriaca was active diurnally only un-
der conditions of deep shade or cloud cover; most
activity was observed at twilight and at night.
Adults were collected at stations 44 (MVZ
1 1 142-56) and 62 (MVZ 140746), and observed
at stations 1 1, 12, 14, 19, 27, 29, 35, 37, 38, 39,
57, and 62. In excess of several hundred juveniles
were observed 4 July 1973 at station 39. Larvae
were observed 27 July 1973 at station 12 and 1 1
April 1974 at station 29.
Ranidae
Rana aurora draytoni Baird and Girard, 1852a,
California Red-legged Frog
This species was abundant on the Lower Main
Scarp, in riparian woodland, and wet meadow
habitats of the coniferous forest up to an eleva-
tion of 2,200 m. They also were found in Arroyos
of the West Scarp and Western Foothills to the
Pacific coast. Linsdale (1932) reported this species
from San Ramon at the mouth of Rio Santo
Domingo, which is apparently the southernmost
record for the species in Baja California.
I observed Rana aurora basking and feeding
during the day in streamside and aquatic vege-
tation associated with fast, slow, and standing
water. It was not observed in rocky areas with
cascading water. Most of the R. aurora were found
near deep pools which were used for escape cov-
er.
Adult specimens were collected at stations 38
(MVZ 140732) and 39 (MVZ 140733-35); ani-
mals were observed at stations 14, 27, 35, 37,
39, 42, 44, and 49. Larvae with well-developed
legs were collected on 10 August 1973 at station
10 (MVZ 140736).
Rana boylii Baird, 1854,
Foothill Yellow-legged Frog
On several occasions I made extensive search-
es of the Lower Main Scarp near the western edge
at La Grulla (station 39) where Loomis (1965)
reported three specimens of R. boylii. I did not
find this species here or along any other wa-
tersheds in the Sierra. The preferred habitat of
R. boylii— fast- flowing water over gravel or cob-
ble streambeds— is rare in the Martir Region,
and I suggest that R. boylii may be a marginal
species in the Region as a result of habitat lim-
itations. The abundance of R. aurora in montane
riparian habitat may also contribute to the scar-
city ofR. boylii. Stream alterations due to mining
activities and acid rain are other possible expla-
nations for the scarcity or loss of this species from
the Region.
TESTUDINATA
Emydidae
Clemmys marmorata (Baird and Girard, 1852a),
Pacific Pond Turtle
Linsdale (1932) reported Clemmys marmo-
rata from Valle de Trinidad (probably in Rio
San Antonio) and Arroyo San Telmo, on the
Pacific Slope. A record from 3.2 km south of
Valladares (CAS 56884-85) placed the species
in a third arroyo to the south. P. Haneline (pers.
comm.) collected a specimen of C. marmorata
from 30 mi east of El Rosario (Rancho El Metate,
station 65). In May 1974, 1 was informed by the
ranch foreman at this locality that this animal
was brought from a more northern population.
I found C. marmorata in the deep pools of a
perennial stream at Rancho San Antonio on Ar-
royo Santo Domingo, at the base of the West
Scarp (station 49). This site is the southern dis-
tribution limit; incongruously, giant cardon (Pa-
chycreus pringlei) and barrel cactus (Ferocactus
sp.) occur a few meters from streamside. This
turtle is uncommon in the Martir Region, prob-
ably as a result of alteration of riparian habitat
for agricultural purposes.
Seeliger (1 945) reported that the six specimens
she examined from Baja California are not sim-
WELSH: BAJA CALIFORNIA HERPETOFAUNA
23
ilar to either described subspecies in California.
Bury (1970) suggested that an analysis of geo-
graphic variation in this species is needed.
Five specimens were collected at station 49
(MVZ 207759-63). An adult turtle was seen div-
ing in a pool on the West Scarp, at 1,600 m in
Arroyo de San Rafael (station 1 0).
SAURIA
Gekkonidae
Coleonyx switaki (Murphy, 1974),
Barefoot Gecko
Fritts et al. (1982) reported Coleonyx switaki
from eastern San Diego County, California and
from Bahia de Los Angeles south of the Martir
Region. These and previous records (Murphy
1974) and the ecological preferences noted by
Fritts et al. (1982) suggest a contiguous range for
this gecko along the eastern flank of the Penin-
sular Range Mountains, although no specimens
have been reported from the Martir Region.
Coleonyx variegatus variegatus (Baird, 1859),
Desert Banded Gecko
Coleonyx variegatus abbotti Klauber, 1945,
San Diego Banded Gecko
The Desert Banded Gecko, Coleonyx varie-
gatus variegatus, occurs in the Colorado Desert
and south to San Felipe (Klauber 1945). A single
specimen from the vicinity of Bahia de Los An-
geles (CAS 121182) appears to be C. v. penin-
sularis Klauber (A. Leviton, pers. comm.). This
specimen is apparently an unreported northern
range extension for the subspecies on the Gulf
Slope.
Coleonyx variegatus abbotti Klauber, the San
Diego Banded Gecko, occurs on the Pacific Slope
of the Martir Region. This lizard was collected
at Rancho San Jose (SDNHM 24390) and east
of San Telmo (SDNHM 42542) in the Western
Foothills; Bostic (1971) reported two specimens
from the Central Desert.
I collected two specimens of Coleonyx varie-
gatus variegatus in Valle de San Felipe near the
east side of San Matias Pass (MVZ 140742-43).
These specimens resembled C. v. variegatus
(Klauber 1945) except that they had a clear and
distinct nuchal light loop, characteristic of C. v.
abbotti. Klauber ( 1 945) postulated that these races
intergrade across some low passes in the Pen-
insular Range, and the present specimens sup-
port this contention.
Phyllodactylus nocticolus (Dixon, 1964),
Baja California Leaf-toed Gecko
Dixon (1969) indicated that Phyllodactylus
nocticolus ranges along the east scarps of the Pen-
insular Range from California southward down
the length of Baja California. He noted that P.
n. nocticolus occurs through the Martir Region.
Bostic (1971) described a new subspecies, P. n.
sloani, from the Central Desert, thus extending
the species range to the Pacific Slope, south of
the Martir Region. The northernmost locality he
reported was 39 km SE Rosario (29°48'N,
115°33'W).
I collected two specimens of Phyllodactylus
nocticolus at station 65 (MVZ 140744-45) about
25 km NW of the above locality. This site is a
volcanic slope with boulders, in Central Desert
scrub vegetation. The geckos were both under
exfoliating slabs of volcanic rock, 1.3 m above
ground. These two specimens had 21 paraver-
tebral tubercules, within the range of both P. n.
nocticolus and P. n. sloani, but only 28 and 30
longitudinal rows, within the range of P. n. noc-
ticolus and below that of P. n. sloani (Dixon
1964; Bostic 1971).
Iguanidae
Callisaurus draconoides Blainville, 1835,
Zebra-tailed Lizard
Callisaurus draconoides occurs on both Gulf
and Pacific slopes of the Martir Region. It in-
habits all scrub vegetations, and is most abun-
dant in the creosote bush scrub of Valle de San
Felipe and Valle Chico. It ranges through San
Matias Pass into Valle de Trinidad (SDNHM
1 6869-8 1 ), and occurs at Rancho San Jose (Lins-
dale 1932), and 16 km west of station 10 in the
Western Foothills. C. draconoides is uncommon,
with a spotty distribution on the Pacific Slope of
the Region; distribution and abundance increase
southward into the Central Desert (stations 63,
64; Bostic 1971). North of the Region, it occurs
at Punta Banda (LACM 94409) and east of En-
senada(LACM 107136).
This species was found in sandy areas of open
benches and slopes, or on sandy soil around scat-
tered outcrops. Specimens were collected at sta-
24
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
tions 2 (MVZ 140753), 4 (140750-51), and 63
(MVZ 140752); animals were observed at sta-
tions 1, 5, 46, 54, 61, 64, and 65.
Crotaphytus insularis vestigium
Smith and Tanner, 1972,
Baja California Black-collared Lizard
Smith and Tanner (1972) described Crotaphy-
tus insularis vestigium and indicated a Gulf Slope
range for the species through south central Baja
California. It is known from the lower East Scarp
of the Martir Region (SU 17048; Linsdale 1932).
I collected two adult males at stations 49 (MVZ
140754) and 65 (MVZ 140755), on the Pacific
Slope of the Region. These records confirm the
validity of two unreported, and apparently ig-
nored, specimens collected in 1 935 from Rancho
San Jose (station 27; SDNHM 24391-92). Van
Denburgh (1922) reported this species at Trini-
dad, west of San Matias Pass, the northernmost
record on the Pacific Slope. The lizard is now
known from the Martir Region, from the Pacific
Slope of the Central Desert at Mesa de San Carlos
(Bostic 1971), and near Rancho Catavina (LACM
16993, 63176). In Baja California Sur, it occurs
at Arroyo La Purisima, 12.8 km southwest of
Canipole (SU 1 1545); 4.0 km northeast San Jose
de Comondu; at Comondu; at La Purisima; and
9.6 km and 52.8 km north of Canipole (Smith
and Holland 1971). These records indicate a much
greater distribution for this lizard in Baja Cali-
fornia than the Gulf Slope range proposed by
Smith and Tanner (1972, 1974).
Both specimens I collected were basking on
boulders during the midafternoon. The site at
station 65 is a rocky volcanic slope grown to
Central Desert scrub; the site at station 49 is a
granite outcrop in coastal sage scrub.
Dipsosaurus dorsal is dorsal is
Baird and Girard, 1852a,
Desert Iguana
This lizard is known from numerous localities
in the Colorado Desert, south to Bahia de Los
Angeles on the Gulf Slope (Murray 1955). Lins-
dale (1932) reported it from the Vizcaino Desert
near San Ignacio. A second race is recognized
south of the Vizcaino Desert through the Cape
Region (Schmidt 1922). The species is absent
from the Pacific Slope of Baja California north
of latitude 29°N in the Central Desert Region.
I observed Dipsosaurus dorsalis dorsalis along
the base of the East Scarp in Valle de San Felipe
and Valle Chico (stations 5 and 54), and in a
canyon bottom west of the scarp base (station
61). All observations of D. d. dorsalis occurred
in creosote bush scrub on open benches and slopes
with a sandy substrate.
Gambelia wizlizenii copei (Yarrow, 1882),
Cope's Leopard Lizard
Gambelia wizlizenii wizlizenii
(Baird and Girard, 1852*),
Longnose Leopard Lizard
Banta and Tanner (1968) indicated that Gam-
belia wizlizenii copei occupies most of the pen-
insula south of the Martir Region, ranging north
to the vicinity of San Felipe on the Gulf coast,
and that G. w. wizlizenii occurs in the Colorado
Desert of northeastern Baja, from the vicinity of
San Felipe northward. On the Pacific Slope, G.
w. copei ranges on the coast to just north of the
Martir Region and inland north just into the
United States (Banta and Tanner 1968; Mahrdt
1973).
I collected four specimens of Gambelia wiz-
lizenii copei on the Pacific Slope of the Martir
Region. I found two active males during mid-
morning at station 27 (MVZ 140756-57). They
were on sandy soil under shrubs in an area of
scattered granite outcrops, in coastal sage scrub.
A third adult male was captured basking on a
rock in creosote bush scrub at station 2 (MVZ
1 40759). A subadult female was captured emerg-
ing from a burrow at 0815 hr in coastal sage
scrub/oak woodland ecotone at the head of Ar-
royo San Telmo (1,480 m elevation), station 19
(MVZ 140758). I observed one probable G. w.
copei on the lower East Scarp in creosote bush
scrub at station 46.
Petrosaurus mearnsi (Stejneger, 1894),
Banded Rock Lizard
This species has been reported from numerous
East Scarp localities (Van Denburgh 1922; Lins-
dale 1932). It ranges south on the Gulf Slope to
Bahia de Los Angeles (CAS 85338), and occurs
north into California as far as Riverside County.
Petrosaurus is absent from the Pacific Slope north
of the Central Desert region.
I observed Petrosaurus mearnsi only in rocky
WELSH: BAJA CALIFORNIA HERPETOFAUNA
25
areas of both creosote bush scrub and pinyon-
juniper woodland up to 1,220 m. Few individ-
uals were seen on the ground as P. mearnsi seems
to prefer steep rock faces from 1 .5 to 5.0 m above
ground. The angles of slope of perching sites
measured for P. mearnsi were greater than those
of all other species of lizards in the Region, com-
monly exceeding 90°. These lizards were active
all day, with a midmorning peak. The greatest
densities of lizards were at riparian edges in can-
yons of the East Scarp.
Specimens were collected at stations 1 7 (M VZ
140762) and 59 (MVZ 140760-61). I observed
P. mearnsi at stations 4, 16, 20, 21, 23, 46, 48,
59, 60, and 61.
Phrynosoma coronatum (Blainville, 1835),
Coast Horned Lizard
Reeve (1952) indicated that the race Phryno-
soma coronatum schmidti occurs on the Pacific
Slope of the Martir Region, south into the Cen-
tral Desert, and northward to near the United
States border. Bostic (1971) reported the species
from the Central Desert, but he did not indicate
subspecific affinities. P. coronatum appears to be
absent on the Gulf Slope north of latitude 29°N.
I observed lizards in chaparral up to 1 ,400 m
at station 18, and at many localities in coastal
sage scrub. They frequented sand, soil, and rock
rubble substrates. P. coronatum was active in
direct sunlight and on substrates with unusually
high surface temperatures (up to 49°C), when
other lizard species were relatively inactive and
seeking shelter in shade.
Specimens were collected at stations 10 (MVZ
140824), 18 (MVZ 140826), 27 (MVZ 140823,
140825), 28 (MVZ 140821), and 49 (140822).
Phrynosoma mcallii (Hallowell, 1852),
Flat-tailed Horned Lizard
Reeve (1952) reported three specimens of
Phrynosoma mcallii from extreme northeastern
Baja California (Colorado Desert). One record
exists from further south, at the mouth of Gua-
dalupe Canyon, at the foot of the Sierra Juarez
(CAS 1 1 9077). This species has yet to be reported
from the desert immediately east of the Sierra
San Pedro Martir. P. mcallii occurs in the vicin-
ity of sand dunes and sandy flats (Stebbins 1985);
such habitat is present in the Martir Region, but
is not continuous with similar habitat to the north.
Phrynosoma platyrhinos calidiarum (Cope, 1896),
Southern Desert Horned Lizard
Reeve (1952) reported seven records for Phry-
nosoma platyrhinos calidiarum in northeastern
Baja California, four from the South Colorado
Desert in the vicinity of San Felipe. Welsh and
Bury (1984) reported it from three localities in-
land from the Gulf. Other records of note in-
clude: Valle de San Felipe (station 5; SU 18588),
Monies de Media (SDNHM 25252), and 15 km
north of Bahia de San Luis Gonzaga (CAS 91631).
P. platyrhinos may occur south to the vicinity of
Bahia de Los Angeles. Apparently it is absent on
the Pacific Slope of the Martir Region and in the
Central Desert.
Sauromalus obesus obesus (Baird, 1859),
Western Chuckwalla
Gates (1968) first reported Sauromalus obesus
in northern Baja California from specimens col-
lected at six localities on the Gulf Slope. The
northernmost locality was San Matias Pass (sta-
tion 1); the southernmost record was 36 km
southwest of Bahia de San Luis Gonzaga. I col-
lected an adult female at station 1 6 (MVZ 1 40749)
and observed individuals at stations 1 7 and 46
on the east scarp. Apparently this species is ab-
sent on the Pacific Slope of the Region.
Recent records from about 5 km north of Pun-
ta Prieta (MVZ 1 17420 and 1 17467-69) extend
the known range of 5". obesus south into the Cen-
tral Desert. A single record from the Sierra Juarez
at Canon Camillas (SDNHM 43974) fills some
of the gap between Baja California populations
and those known from southern California. This
saxicolous species probably occurs throughout
the east scarps of the Peninsular Range from Cal-
ifornia into the Central Desert of Baja California.
Bostic (1971) reported S. australis from 50 km
south of Punta Prieta in the Central Desert, thus
indicating that a zone of sympatry between S.
australis and S. obesus probably exists in the
Central Desert in the vicinity of Punta Prieta.
Seib (1980) implies that this zone of sympatry
does not exist in the Central Desert, ignoring S.
australis and treating all populations in the Cen-
tral Desert as S. obesus. More work is needed to
clarify relationships among the chuckwallas of
Baja California.
Sauromalus obesus was unusually wary in the
Martir Region. Individuals sought cover in re-
26
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
sponse to movement at distances as great as 70
m. The lizards occurred on high cliff faces and
ledges, and seldom ventured far from crevices
where they sought cover. These large dark lizards
contrast greatly with the pale granite surface of
the Martir Region. The wariness of S. obesus
may allow use of these pale surfaces in an area
with high densities of raptorial birds (Short and
Crossin 1967). I submit that the dark color may
be a therm oregulatory adaptation, which has a
higher selective value than does cryptic color-
ation for predator avoidance. The extreme war-
iness of S. obesus may explain why this species
remained unreported for so long in this area.
The specimen captured 1 2 June 1 974 at station
16 contained six well-developed ova. Scutella-
tion most closely matched the description for the
race Sauromalus obesus obesus (Shaw 1945).
Sceloporus graciosus vandenburghianus
Cope, 1896,
Southern Sagebrush Lizard
This lizard was the most frequently observed
reptile on the Main Scarp of the Sierra. Its dis-
tribution was limited to coniferous forest, coni-
fer-chaparral ecotone, and related riparian edge
habitats of the upper scarps, from 1 ,980 to 2,820
m. Lizards were active from 0500 to 1800 hr on
boulders, logs, and occasionally on the forest floor;
juveniles occurred mostly on the ground. In the
Martir Region, Sceloporus graciosus vanden-
burghianus exhibited frequent climbing behav-
ior. Elsewhere, they are found mostly on the
ground (Stebbins 1966).
The distribution of S. graciosus in the Martir
Region may be influenced by interspecific com-
petition. I expected it to exist in the chaparral of
the East and West scarps of the Region, a seem-
ingly appropriate habitat. In fact, I found it in
stands of open chaparral only above 2,100 m on
the Main Scarp. The closely related 51. occiden-
talis occurs in chaparral on the East, West, and
Main scarps below 2,100 m. This suggests that
51. occidentalis may be limited to lower areas with
higher temperatures, whereas S. graciosus is
found in cooler climes above 2,100 m. The
mechanism of exclusion could be simple aggres-
sive dominance of the larger species (S. occiden-
talis) over the smaller (S. graciosus), driving the
latter away from critical resources at lower ele-
vations (interference competition). Feeding
strategies may also play a role in segregating these
two lizards. In the ecotonal vegetation between
1,970 and 2,100 m where these two lizards co-
exist, limited observations indicated a difference
in their choice of perching and foraging sites.
Adult S. graciosus were most often on elevated
perches and descended to forage on the ground.
S. occidentalis was always observed on the ground
or on low boulders.
Specimens were collected at stations 8 (MVZ
140773-74), 25 (MVZ 140763), 38 (MVZ
140768-69, 147514), 39 (MVZ 140764-67,
147513), 47 (MVZ 140770-71), 55 (MVZ
140772), and 58 (MVZ 140785-86); observa-
tions were made at stations 12, 22, 24, 25, 30,
32, 36, 37, 42, 44, 47, and 55.
Sceloporus magister rufidorsum Yarrow, 1883,
Red-backed Spiny Lizard
Sceloporus magister uniformis
Phelan and Brattstrom, 1955,
Yellowback Spiny Lizard
Phelan and Brattstrom (1955) reported four
subspecies of Sceloporus magister in Baja Cali-
fornia. Hall (1973) indicated that these taxa can
be divided into two distinct karyological groups
indicating at least two, and possibly four, distinct
species. Murphy (1983a), citing Hall and un-
published electrophoretic data, treated the forms
as separate species. I reserve judgement until more
information is available, treating the northern
forms as subspecies. Two forms, one from each
karyological group, occur in the Martir Region:
S. m. uniformis (2n = 26 chromosomes) in the
Colorado Desert south to San Felipe on the Gulf
Slope and S. m. rufidorsum (2n = 30 chromo-
somes) on the Pacific Slope, from just north of
the Martir Region south through the Central
Desert. Subspecific morphological differences are
primarily based on dorsal pattern and coloration
of adult males, with the number of femoral pores
as a secondary distinguishing character (Phelan
and Brattstrom 1955).
I collected three adult males. A male from sta-
tion 2 (MVZ 140799) had a uniform beige dor-
sum and a femoral pore count of 13-13, which
closely matches the description of S. m. unifor-
mis. An adult female from station 2 (MVZ
1 40798) also fits this description. The other males,
collected at station 27 (MVZ 140793-94) were
not similar to the described forms (Phelan and
WELSH: BAJA CALIFORNIA HERPETOFAUNA
27
Brattstrom 1955). These two males had a faint
and diffuse rust color dorsally, and lacked any
pattern or side bars. Femoral pore counts were
17-18 and 16-16, which corresponds to S. m.
rufidorsum (15-20), not to S. m. uniformis (10-
1 5). Other specimens of the S. magister complex
collected were an adult female from station 62
(MVZ 140797), and three juveniles, two males,
and one female from station 64 (MVZ 140795-
96, 1475 1 7). All of these matched the description
for S. m. rufidorsum.
This small sample suggests the possibility of
integration between two forms occurring in the
Martir Region. Similarly, Bostic (1971) exam-
ined 1 3 males from the Central Desert and found
only one specimen with a typical Sceloporus ma-
gister rufidorsum pattern. Further biochemical
work is needed to describe variation in the S.
magister complex in this part of Baja California.
Sceloporus magister was observed in all scrub
vegetation on both the Gulf and Pacific slopes
of the Martir Region, up to 1,220 m, and south
into the Central Desert. It was usually seen in
association with plant forms that afforded pro-
tective cover. Lizard burrows were located at the
base of plant species such as Mojave Yucca (Yuc-
ca schidigera), Agave sp., and Cholla (Opuntia
sp.). Lizards often were observed up to 1.5 m
above ground in a wide variety of woody plants.
I saw S. magister in all regional edaphic types,
including riparian edge, and on all substrates. S.
magister was observed at stations 5, 9, 10, 16,
18, 63, and 65.
Sceloporus occidentalis biseriatus
Hallowell, 1854,
Great Basin Fence Lizard
This lizard reaches its southern limit in the
Martir Region, where it occurs primarily along
the Pacific Slope. It was abundant in the chap-
arral of the West and Upper East scarps, and
uncommon in conifer-chaparral ecotones on the
Lower Main Scarp (stations 43 and 47). This
species was absent above 2, 1 20 m. On the lower
Pacific Slope, in areas of coastal sage scrub, it
was found only in riparian habitat.
In riparian woodland, Sceloporus occidentalis
was highly arboreal, commonly found 1.5 m or
more above ground on trees and fenceposts. Liz-
ards foraged on the ground from these elevated
perches; when startled, they returned to their
perches. In the chaparral and conifer-chaparral
ecotone, S. occidentalis was primarily ground-
dwelling. In the conifer-chaparral ecotone, S. oc-
cidentalis is sympatric with S. graciosus (see
above).
Specimens were collected at stations 1 0 (MVZ
140790), 19 (MVZ 140792, 147518), 27 (MVZ
140791), 29 (MVZ 140788), 35 (MVZ 140789),
43 (MVZ 140781), 47 (MVZ 140782-83), 53
(MVZ 140784), 55 (MVZ 140787, 147515), and
58 (MVZ 140785-86); observations were at sta-
tions 14, 19, 56, and 57.
Sceloporus orcutti Stejneger, 189 3 a,
Granite Spiny Lizard
Sceloporus orcutti occurs on the rocky scarps
of the Martir Region in scrub vegetations, pin-
yon-juniper woodland, and chaparral. It reaches
the conifer-chaparral ecotone on the Lower Main
Scarp (station 39) at 2,080 m. It is most abundant
on the lower East Scarp in rocky creosote bush
scrub. This species is entirely saxicolous, and is
associated with large boulders, outcrops, or rock
faces.
Sceloporus orcutti was extremely wary and dif-
ficult to approach, reacting to my presence at a
distance of about 30-40 m. I observed one lizard
taken by a Harris' Hawk (Parabuteo unicinctus)
and another by a Red-tailed Hawk (Buteo ja-
maicensis). The abundance of large raptorial
birds in the Martir Region, the dark body color
of this spiny lizard, and the pale granite surfaces
frequented by this species may together explain
the extreme wariness displayed by S. orcutti.
Mayhew (1963) mentioned this wariness and
noted a southward increase in its intensity. S1.
orcutti is easily captured by noose at the north
end of its range in the San Jacinto Mountains of
California, but it becomes increasingly difficult
to approach further southward, and is virtually
impossible to noose in the Martir Region. Buteo
populations at the northern extreme of this liz-
ard's range may be less dense, due to human
interference, or there may be a greater choice of
prey animals to select from. Both factors would
result in less intense selective pressure for alert-
ness in S. orcutti.
Sceloporus orcutti was collected at stations 1 7
(MVZ 140776-78), 27 (MVZ 140779, 147516),
46 (MVZ 140775), and 65 (MVZ 140780); ob-
servations were made at stations 2, 4, 10, 1 1, 14,
29, 35, 37, 39, 48, 49, 51, 56, 59, 60, and 62.
28
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
Uma notata notata Baird, 1859,
Desert Fringe-toed Lizard
Schmidt (1922) noted two records for Uma
notata notata from the desert east of the Sierra
Juarez, approximately 160 km north of the Mar-
tir Region. Norris (1958) suggested this are-
nicolous lizard is restricted to contiguous sand
dune habitat, which may explain its apparent
exclusion from the Martir Region, where dune
habitat is not contiguous. On the other hand, the
lack of records for U. notata may be an artifact
of limited collecting.
Urosaurus graciosus graciosus Hallowell, 1854,
Western Brush Lizard
Urosaurus graciosus graciosus is known from
the South Colorado Desert at San Felipe (Lins-
dale 1932); it ranges northward through the des-
erts of California and Arizona. Stebbins (1966)
indicated that U. g. graciosus ranges through much
of northeastern Baja California. I collected two
adult males in creosote bush scrub at the base of
the East Scarp, one in Valle Chico (station 61)
(MVZ 140808), and the other in Valle de San
Felipe near station 1 7 (MVZ 1 40809). Both spec-
imens were 1-2 m above ground in iron wood
trees (Olneya tesotd). These localities extend the
known range of this species slightly south and
east.
Urosaurus' graciosus graciosus is absent from
the East Scarp and the Pacific Slope of the Martir
Region, and is also unknown from the Central
Desert. U. microscutatus and U. lahtelai occur
in these respective areas (see below and Rau and
Loomis 1977); presence of these congeners may
exclude U. g. graciosus.
Urosaurus microscutatus (Van Denburgh, 1894),
Small-scaled Lizard
Urosaurus microscutatus was common on the
rocky scarps of the Martir Region in all scrub
vegetations, chaparral, and pinyon-juniper
woodland; it occurs up to 2, 1 20 m on the East
Scarp, and 1 ,525 m on the West Scarp. The species
is not present on open sandy desert and areas
with few outcrops; it is absent in the South Col-
orado Desert and San Quintin Plain, and un-
common in the Western Foothills. Bostic (1971)
reported it as uncommon in the Central Desert.
This species is arboreal, occurring up to 2.5 m
above ground. It commonly descends from ele-
vated perches to forage on the ground. I found
Urosaurus microscutatus on boulders and in
vegetation, with females and young occasionally
observed on the ground. Greatest densities were
reached in the rocky creosote bush scrub areas
of the East Scarp along the riparian edge.
Specimens were collected at stations 10 (MVZ
140814, 140819), 27 (MVZ 140820), 46 (MVZ
140815-16, 147519), 49 (MVZ 140812-13), 51
(MVZ 14081 1), 53 (MVZ 140810, 147520), and
59 (MVZ 140817-18); observations were made
at stations 4, 11, 17, 19, 23, 48, 60, 61, 62, and
65.
Uta stansburiana elegans (Yarrow, 1883),
California Side-blotched Lizard
Ballinger and Tinkle (1972) indicated that a
single race of the Side-blotched Lizard occurs
throughout peninsular Baja California. This
species is the most common and ubiquitous liz-
ard in the Martir Region at all elevations and in
all vegetations except coniferous forest. On the
Lower Main Scarp it ranges up to 2,120 m in
chaparral.
Uta stansburiana elegans showed no special
edaphic or substrate preference. It is mostly a
ground lizard but occasionally occurs up to 1.5
m above ground on rock outcrops.
Specimens were collected at stations 10 (MVZ
140806), 27 (MVZ 140807), 35 (MVZ 140802),
45 (MVZ 140803), 46 (MVZ 140805), 49 (MVZ
140804, 147522), 54 (MVZ 140800, 147521),
and 56 (MVZ 140801); observations were made
at stations 1, 2, 4, 9, 14, 17, 19, 47, 48, 57, 59,
60, 61, 62, 63, 64, and 65.
Xantusiidae
Xantusia henshawi henshawi Stejneger, 1893&,
Granite Night Lizard
This saxicolous species occurs on all the rocky
scarps of the Sierra San Pedro Martir, up into
the coniferous forest on the Lower Main Scarp
at Arroyo Encantada (2,100 m; CAS 57294-96),
and the conifer-chaparral ecotone at La Grulla
(station 39; Murray 1 955). I collected a specimen
in chaparral at station 19 (MVZ 140828), and
three specimens in coastal sage scrub at station
27 (MVZ 140827-28, 140830). I saw one on the
East Scarp in pinyon-juniper woodland at sta-
tion 5 1 . The species undoubtedly occurs to the
desert floor at the base of the East Scarp (Stebbins
WELSH: BAJA CALIFORNIA HERPETOFAUNA
29
1966; Lee 1975) but its secretive habits and the
ideal protective habitat of the cliffs and canyons
on the East Scarp make collecting difficult. All
animals were found under exfoliating slabs on
large granite boulders. The species is as yet un-
reported in the Central Desert, possibly because
of the marked change in geomorphology that oc-
curs at this latitude. The primarily granitic for-
mations that dominate in the Martir Region and
provide cover for this species give way to vol-
canic formations in the Central Desert.
Xantusia vigilis vigilis Baird, 1858,
Desert Night Lizard
Xantusia vigilis wigginsi Savage, 1952,
Baja California Night Lizard
This lizard occurred most commonly in areas
of open benches and slopes in creosote bush and
Central Desert scrub on both the Gulf and Pacific
slopes. Apparently it is absent in the coastal sage
scrub of the Western Foothills. I collected one
specimen on the rocky lower West Scarp at 1 ,400
m in pinyon-juniper woodland (station 1 0; M VZ
140833).
Savage (1952) indicated that the Martir Re-
gion was the probable area of intergradation be-
tween the races Xantusia vigilis wigginsi of cen-
tral Baja California, and X. v. vigilis of the
southwestern United States. Two adult males
from San Matias Pass (station 1) matched Sav-
age's description of X. v. wigginsi on the basis of
tail pattern, but showed a dorsal pattern found
in both forms. A juvenile from station 1 had
dorsal characteristics of X. v. vigilis. An adult
from station 1 0 completely lacked a pattern, and
did not fit any previous descriptions. A juvenile
from the South Colorado Desert near station 54
had the dorsal pattern of A', v. vigilis and the tail
pattern of X. v. wigginsi. One of each of two
juveniles from station 62 in the north Central
Desert fits closely the descriptions of X. v. wig-
ginsi and X. v. vigilis. A series of specimens at-
tributed to X. v. wigginsi (SDNHM 41340-52),
was collected at San Telmo, a lowland locality
of the Pacific Slope dominated by Central Desert
scrub. The eight specimens I collected show char-
acteristics of both subspecies, and failed to match
either satisfactorily. These findings support Sav-
age's speculation of intergradation in this Re-
gion.
All animals I collected were taken from rubble
and dead fall of Mojave yucca (Agave sp.), barrel
cactus, or Echinocactus sp. Specimens were col-
lected at stations 1 (MVZ 140831, 147523), 2
(MVZ 140836-37), 10 (MVZ 140833), 54 (MVZ
140832), 64 (MVZ 140834-35); observations
were recorded at stations 14 and 65.
Scincidae
Eumeces gilberti rubricaudatus Taylor, 1935,
Western Redtail Skink
Jones (1985) indicated that Eumeces gilberti
rubricaudatus ranges south to the Sierra San Pe-
dro Martir in Baja California, but with an inland
distribution. Rogers and Fitch (1947) reported
E. g. rubricaudatus from the Pacific coast of the
Martir Region at San Antonio del Mar (station
9). A record from La Grulla (station 40; SDNHM
44291) indicates that they range into the conif-
erous forest-chaparral ecotone. I collected three
specimens, from stations 14 (MVZ 140842), 27
(MVZ 140843), and 49 (MVZ 140841) and ob-
served one at station 1 9. All were in deep riparian
woodland; light intensity range 50-85 FC, rela-
tive humidity 33^1% (Welsh 1976a). Jones
(1985) reported this species in the chaparral
vegetation of the Martir Region but failed to cite
specimens, suggesting this may be an assumption
based on habitat associations observed else-
where. Evidence to date suggests that this species
is restricted to the riparian corridors in the Re-
gion, and may occur along the moist coastal cor-
ridor of the San Quintin Plain. A record from
Guadalupe Canyon in the Sierra Juarez (CAS
119087) suggests it may also occur in riparian
vegetation on the East Scarp of the Martir Re-
gion.
Eumeces skiltonianus skiltonianus
(Baird and Girard, 18526),
Western Skink
This lizard is common in the coniferous forest
of the Sierra, occurring up to 2,800 m (station
22). It also occurs on the West Scarp and at lower
elevations on the Pacific Slope; records include
Rancho San Jose (Linsdale 1932) and 6.6 km
east of San Telmo (SDNHM 42741). Coastal rec-
ords for Eumeces skiltonianus in the Martir Re-
gion include: San Antonio del Mar (SDNHM
41990), 10 km southeast Punta Colnett (CAS
57556), San Quintin (CAS 55802), and Colonia
Guerrero (LACM 107935).
30
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
On the lower Pacific Slope, in areas dominated
by scrub vegetation, Eumeces skiltonianus ap-
peared to be restricted to riparian habitat and
the moist coastal corridor. Limited observations
suggest this lizard reaches greatest densities in
areas of riparian edge habitat.
Specimens were collected at stations 37 (MVZ
140838), 39 (MVZ 140839, 147524), and 50
(MVZ 1 40840). Sightings occurred at stations 1 0,
19, 22, and 38.
A site recorded from Catavina in the north
Central Desert has recently come to my attention
(D. Schmoldt, pers. comm. and 35-mm slide).
This would constitute a new southern record for
the subspecies if confirmed by a specimen.
Teiidae
Cnemidophorus hyperythrus schmidti
(Van Denburgh and Slevin, 1921),
Orangethroat Whiptail
Cnemidophorus hyperythrus schmidti occurs
on the Pacific Slope from San Diego County,
California, south through the Central Desert,
where it ranges across to the Gulf coast, and
through the Cape Region (Burt 1931). In the
Martir Region these whiptails were found in scrub
vegetation of the western foothills and north
Central Desert. Bostic (1968) found C. hypery-
thrus on the San Quintin Plain, but he considered
habitat there to be suboptimal; C. labialis was
far more abundant than C. hyperythrus in this
area.
I found Cnemidophorus hyperythrus to be ter-
restrial and diurnal, showing preference for areas
of fine-textured, soft soils with shrub cover. It
was most abundant in riparian/scrub ecotone. I
saw American Kestrels (Falco sparverius) cap-
ture individuals of C. hyperythrus on two occa-
sions. Specimens were collected at stations 27
(MVZ 147525), 49 (MVZ 140844-46), 62 (MVZ
140849-50), and 64 (MVZ 140847^18).
Three subspecies of Cnemidophorus hyp-
erythrus have been described from Baja Califor-
nia: C. h. hyperythrus, C. h. schmidti, and C. h.
beldingi (Linsdale 1932). The northern race, C.
h. beldingi, is generally distinguished by the pres-
ence of two middorsal stripes that are lacking in
the central race, C. h. schmidti (Murray 1955).
Murray noted a difference between races in the
number of supraoculars that were separated by
granules from the frontals, which he considered
a more diagnostic character. Bostic (1971) de-
scribed 15 specimens from the Central Desert,
the presumed range of C. h. schmidti, that match
Murray's characters for C. h. beldingi.
The eight specimens I collected appear to be
of the race C. h. schmidti. Available data (Table
3) indicate only slight differences between the
subspecies. I suggest that there is insufficient dif-
ferentiation to warrant three distinct races. With
most morphological characters showing a north
to south clinal variation, I consider the central
peninsular region to be an area of intergradation
for the ecotypes C. h. beldingi and C. h. hyp-
erythrus. Burt (1931) recognized no subspecies.
Cnemidophorus labialis Stejneger, 1890,
Baja California Whiptail
Cnemidophorus labialis is endemic to north
central Baja California, occurring primarily along
the San Quintin Plain and the Pacific coast of
the Central Desert Region. The northernmost
record for C. labialis is Punta San Jose, about
40 km NW San Vicente (Bostic 1968). It ranges
southward in the Central Desert to the vicinity
of latitude 28°13'N. Bostic (1971) considered
this to be near the southern limit of distribution.
Vegetation in the range of C. labialis on the San
Quintin Plain most closely approximates Central
Desert scrub, with plant density increased and
species composition altered slightly by the fog
and cool, moist air. I did not collect any speci-
mens of C. labialis.
Cnemidophorus tigris multiscutatus Cope, 1892a,
Coastal Whiptail
Cnemidophorus tigris tigris
Baird and Girard, 1852#,
Great Basin Whiptail
Cnemidophorus tigris multiscutatus occurs up
to 1,980 m in the Martir Region, on both sides
of the peninsular divide in all types of scrub vege-
tation, pinyon-juniper woodland, and chaparral.
It ranges to the coast on the Pacific Slope and
south through the Central Desert (Bostic 1971).
On the Gulf Slope, C. t. multiscutatus occurs in
the south Colorado Desert in Valle de San Felipe
and Valle Chico, along the base of the East Scarp.
Murray (1955) collected the Great Basin Whip-
tail (C. t. tigris) from three Gulf coastal localities
in the South Colorado Desert near San Felipe.
I found Cnemidophorus tigris multiscutatus
only at ground level on sand, soil, rock, rubble,
WELSH: BAJA CALIFORNIA HERPETOFAUNA
31
TABLE 3. VARIATION IN MORPHOLOGICAL CHARACTERS AMONG SUBSPECIES OF CNEMIDOPHORUS HYPERYTHRUS. (See footnotes
for sources of published data.)
Character
Subspecies
Martir
specimens
beldingi
schmidti
hyperythrus
Number of granules around midbody
72.8 ± 0.8
75.2 ± 0.6
77.6 ± 0.6
75.0
(66-79) 17'
(66-83) 542
(69-90) 45'
(70-80) 8
Anterior supraoculars separated from
the frontal by granules:
only part of third
53
10" 192
285
0
all of third
0
48" 142
445
8
only part of second
253
3" 162
45
—
all of second
363
152
0
-
Number of middorsal lines:
three
96
0
48"
1
two
1046
24 32
15"
2
one (forked anteriorly)
466
594 562
34
5
1 Data from Walker and Taylor (1968).
2 Data from Bostic (1971).
3 Data from Van Denburgh (1922).
4 Data from Linsdale (1932).
5 Combined data from Murray (1955) and Linsdale (1932).
6 Data from Burt ( 1931).
or leaf litter. I observed it foraging throughout
the day, mostly in the broken shade of the shrub
understory. When startled or chased, these whip-
tails sought shelter in burrows.
Specimens were collected at stations 2 (MVZ
140857), 14 (MVZ 140854), 27 (MVZ 140855),
48 (MVZ 140853), 49 (MVZ 140852, 147526),
56 (MVZ 140851), and 62 (MVZ 140856,
1 47527). These animals resembled Bostic's (1971)
collection of Cnemidophorus tigris multiscutatus
from the Central Desert. C. t. multiscutatus was
observed at stations 4, 5, 10, 17, 18, 19, 35, 46,
48, 49, 52, 57, 59, 60, 61, 62, 63, 64, and 65.
Anguidae
Elgaria multicarinata webbi (Baird, 1858),
San Diego Alligator Lizard
Elgaria multicarinata webbi is known from co-
niferous forest areas on the Lower Main Scarp
at La Grulla (Linsdale 1932), Arroyo Encantada
(CAS 57038^2), El Alcatraz (CAS 57330), and
at station 37. This lizard was found in chaparral
on the West Scarp at stations 10 (MVZ 140860)
and 35 (MVZ 140859) and on the East Scarp at
station 53 (MVZ 1 40858). Western Foothills rec-
ords include: Valladares (Linsdale 1932); Mes-
quite Spring, Valle de Trinidad (SDNHM 1 6864);
and Rancho San Jose (Linsdale 1932). These
Western Foothill records are all from sites con-
taining riparian habitat. The chaparral habitat
occupied by E. m. webbi at higher elevations is
scarce below 1,220 m in the Western Foothills,
and I suspect that its distribution below 1,220
m is localized primarily around the riparian cor-
ridors. Records from Canon del Diablo (SU
17293-95) and Canon El Cajon (Linsdale 1932)
indicate a possible similar riparian distribution
of E. m. webbi on the East Scarp below the chap-
arral association. Coastally, on the San Quintin
Plain, I observed this lizard in Central Desert
scrub at 1.7 km east of San Antonio del Mar
(station 9), and it is also known from Colonia
Guerrero (LACM 75501, 107933). Bostic (1971)
indicated that this race occurs along the moist
coastal belt of the Central Desert Region.
All specimens I collected were in areas of rocky
outcrops, on granitic sand, soil, or rock rubble,
except the individual near station 9, which was
on open sandy ground near vegetation.
Anniellidae
Anniella geronimensis Shaw, 1940,
Baja California Legless Lizard
Anniella geronimensis occurs coastally on the
San Quintin Plain from Colonia Guerrero south
to the vicinity of El Rosario (Shaw 1953; Bezy
et al. 1 977). This species is endemic to the Martir
Region and a single near-shore island. Shaw
32
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
(1953) reported A geronimensis sympatric with
A. pulchra in the vicinity of Colonia Guerrero.
Anniella pulchra Gray, 1852,
Silvery Legless Lizard
Anniella pulchra occurs on the Pacific Slope
within the Martir Region at San Jose (SDNHM
40394), Socorro (SDNHM 8842), east of San
Telmo (SDNHM 42738), San Antonio del Mar
(SDNHM 4900-10), and west of Colonia Guer-
rero (Bezy et al. 1977). A resident told me that
A. pulchra occurred in sandy fields at the bottom
of the arroyo at station 19. This site at 1,450 m
may be the highest record for the species in Baja
California. Stebbins (1954:303) reported A. pul-
chra from 1,950 m in the Giant Forest, Sequoia
National Park in California. Its preference for
moist, fine-textured soils, as reported by Miller
( 1 944), probably limits A. pulchra to arroyo bot-
toms and coastal localities within the Martir Re-
gion; its upward distribution in the Sierra is prob-
ably restricted by the steep, granitic west scarp.
Bury (1983) reported Anniella pulchra from
an arroyo on the West Scarp of the Sierra Juarez,
60 km east southeast of Ensenada. Klauber ( 1 932)
recorded A. pulchra at Canon San Salado, at the
base of the East Scarp of the Sierra Juarez. Bury
(1983) maintained that this locality is probably
Arroyo Salado, 1 9 km south of San Vincente near
the Pacific coast; he indicated that a second rec-
ord from the Gulf Slope near San Felipe is also
probably invalild.
SERPENTES
Leptotyphlopidae
Leptotyphlops humilis cahuilae Klauber, 19316,
Desert Worm Snake
Leptotyphlops humilis humilis
(Baird and Girard, 1853),
Western Worm Snake
Klauber (1940) reported two subspecies of
Leptotyphlops humilis in northern Baja Califor-
nia: L. h. humilis occurs on the Pacific Slope from
San Diego County, California, south into the
Central Desert Region, and L. h. cahuilae occurs
along the Gulf Slope from California into central
Baja California. Klauber also noted a record of
L. h. cahuilae from the Vizcaino Desert.
I collected two specimens of Leptotyphlops hu-
milis on the lower Pacific Slope of the Martir
Region: one was taken at 1950 hr, as it moved
rapidly across a dirt road at 1,100 m in open,
sandy, coastal sage, 6.2 km west of station 28
(MVZ 140861); the other was found at station
27 (MVZ 140862), crushed in a cow track. Both
specimens agreed in dorsal pigmentation and in
number of pigmented scale rows with the de-
scription of L. h. humilis (Klauber 1940). How-
ever, the specimen from station 27 had a dorsal
scale count (285) exceeding that established for
L. h. humilis (range 257-283), and in the range
for L. h. cahuilae (280-305). This suggests a zone
of intergradation on the Pacific Slope of the Mar-
tir Region.
Previously unpublished records of Lepto-
typhlops humilis humilis on the Pacific Slope in-
clude 25 km east of San Telmo (CAS 123717)
and 8.3 km south of Socorro Ranch (SDNHM
19231), San Telmo (SDNHM 42746).
Leptotyphlops humilis cahuilae was first re-
ported from the south Colorado Desert at Punta
San Felipe (Murray 1 955); additional records are
known from 58.5 km north of San Felipe
(SDNHM 44262-63) and 8.3 km north of San
Felipe (CAS 1 36368-69). There are three records
for this subspecies from Bahia de Los Angeles,
at the south end of the Colorado Desert (CAS
103465, LACM 2167, and SDNHM 19998).
Klauber (1940) reported this subspecies as far
south as the Vizcaino Desert.
All records for the races of Leptotyphlops hu-
milis from the Martir Region are from areas of
scrub vegetation.
Boidae
Lichanura trivirgata roseofusca Cope, 1868,
Coastal Rosy Boa
Three forms of the Rosy Boa are now recog-
nized on peninsular Baja California. Klauber
(1933), Lowe (1964), and Stebbins (1966) indi-
cated that there are three subspecies (Lichanura
trivirgata trivirgata, L. t. gracia, and L. t. roseo-
fusca). Klauber (193 la) implied that L. t. gracia
was intermediate between L. t. trivirgata and L.
t. roseofusca on the basis of scale counts and color
description. Gorman (1965) found no evidence
of intergradation between L. t. trivirgata and L.
t. gracia, and recognized two species: L. trivirgata
and L. roseofusca.
Bostic (1971) reported two specimens from the
Central Desert of Baja California that were sim-
WELSH: BAJA CALIFORNIA HERPETOFAUNA
33
ilar to Klauber's (193 la) description of Licha-
nura trivirgata gracia, but with dorsal and ven-
tral scale counts matching L. t. trivirgata. One
of these specimens was collected 167 km north
of San Ignacio in the Vizcaino Desert, previously
the northernmost locality known for L. t. trivir-
gata in Baja California (Gorman 1 965). This form
is known from the Cape Region of Baja Califor-
nia and the west coast of Mexico.
In the Martir Region, the Rosy Boa is known
from the lower Pacific Slope at numerous local-
ities in the Western Foothills and on the San
Quintin Plain, all corresponding to areas of coastal
sage and Central Desert scrub. Specimens from
this area match the description for Lichanura
trivirgata roseofusca, which ranges north into
California along the Pacific coast. Gorman (1 965)
ascribed a single specimen from the South Col-
orado Desert, 25 km south of San Felipe, to L.
roseofusca. Welsh and Bury (1984) reported a
specimen matching Klauber's ( 1 93 1 a} description
for L. t. gracia from 56 km northwest of San
Felipe in the Sierra San Felipe of the South Col-
orado Desert. L. t. gracia occurs in the high des-
erts of California and Arizona between the coast-
al California range of L. t. roseofusca, and the
extreme southern Arizona and Sonora range of
L. t. trivirgata. The northern Mojave Desert range
of L. t. gracia is about 335 km away from the
Central Desert and South Colorado Desert re-
gions of Baja California where Bostic (1971) and
Welsh and Bury (1984) collected L. t. gracia.
In view of this great distance, and the lack of
low desert records for the species north of Baja
California, it seems highly unlikely that a con-
tiguous population of Lichanura trivirgata gracia
exists across this area. The disjunct distribution
of L. t. gracia between populations of L. t. ro-
seofusca and L. t. trivirgata prompted my sug-
gestion (Welsh \916a) that L. t. gracia is a re-
current intermediate form, a thesis more recently
entertained by Yingling (1982).
Colubridae
Arizona elegans eburnata Klauber, \946b,
Desert Glossy Snake
Arizona elegans occidentalis Blanchard, 1 924,
Western Glossy Snake
Arizona elegans pacata Klauber, 1 9466,
Peninsula Glossy Snake
Klauber (19466) reported Arizona elegans oc-
cidentalis at numerous localities along the Pacific
Slope from the Martir Region north into Cali-
fornia. It has been reported as far south as El
Rosario on the Pacific coast (SDNHM 44155).
Seifert (1980) reported two specimens from the
Central Desert south of Laguna Chapala.
A second subspecies, Arizona elegans pacata,
is known from the Central Desert Region. Rec-
ords are known from near Santo Domingo
(Klauber 19466), 43.4 km northwest of San Ig-
nacio (Murray 1955), and 50 km west northwest
of El Arco (Banta and Leviton 1963); more re-
cently collected specimens are at CAS and MVZ
(T. Papenfuss, pers. comm.).
Klauber (19466) reported Arizona elegans
eburnata in the Colorado Desert of California
and Arizona, and implied its presence in north-
eastern Baja California. In June 1974, 1 collected
a juvenile male at 2030 hr, 3.2 km southeast of
Rancho Rosarito (station 62; MVZ 140868). The
site was a sandy arroyo with creosote bush scrub
vegetation. The pattern and scutellation matched
those described for A. e. eburnata (Klauber
1 9466). This specimen, the first for the subspe-
cies in Baja California, extended the range to 80
km southwest of San Felipe, and is also the first
record of the desert glossy snake from the Pacific
Slope. Welsh and Bury (1984) reported six more
specimens of this race from Martir Region lo-
calities northwest of San Felipe. All records of
A. elegans from the Martir Region occurred in
areas of scrub vegetation.
Chilomeniscus cinctus Cope, 1861,
Banded Sand Snake
Chilomeniscus cinctus is known from six lo-
calities on the Pacific Slope of the Martir Region:
Arroyo Seco (SDNHM 48150-52), San Antonio
River, northeast of Arroyo Seco (SDNHM
43378), 14 km east of San Telmo (SDNHM
41325), 20 km east of San Telmo (SDNHM
42737), and Valle de Trinidad (SDNHM 3037 1).
The first five sites are in low-lying canyons that
are immediately east and contiguous with the
San Quintin Plain. These areas are dominated
by Central Desert and coastal sage scrubs. Valle
de Trinidad is at the northwest end of the Sierra
and is dominated by creosote bush scrub that is
contiguous through San Matias Pass into the Col-
orado Desert. The Valle de Trinidad record in-
34
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
dicates that C. cinctus is probably present in the
sandy habitats of the South Colorado Desert.
R. Seib (pers. comm.) reported this species from
two localities in the north Central Desert: Mis-
sion San Fernando 3.3 km west of El Progresso
(MVZ 1 17449) and 5.5 km north of Santa Ines
parador, Mexican Hwy 1 (MVZ 117303). Two
other unpublished records exist for Chilome-
niscus cinctus in this area: 16.7 km south of Ca-
tavina (SDNHM 42054), and 50 km south of El
Marmol (SDNHM 38663). Bostic (1971) re-
ported a specimen from 0.8 km north of San
Javier. Numerous additional records exist for the
Central Desert southward throughout the pen-
insula (R. Seib, pers. comm.), including three
unpublished records from San Ignacio (SDNHM
3828-30) in the Vizcaino Desert.
Chionactis occipitalis annulata (Baird, 1859),
Colorado Desert Shovel-nosed Snake
The subspecies Chionactis occipitalis annulata
is known from sandy alluvial soil in a limited
coastal area north of San Felipe, within 3.2 km
of the Gulf of California along Mexican Hwy 5
(Wake 19660; Cross 1970). Welsh and Bury
(1984) reported four more specimens from this
area and two specimens from 1 1.2 km and 14.2
km west of the Gulf of California. This are-
nicolous snake is not known south of the Martir
Region; the discontinuity of suitable sandy hab-
itat may account for its absence.
Coluber flagellum fuliginosus (Cope, 1895),
Baja California Coachwhip
Coluber flagellum piceus (Cope, 1875),
Red Coachwhip
Wilson (1973) indicated two subspecies of Col-
uber flagellum in the Martir Region: C. / piceus,
ranging south through the Colorado Desert to
the vicinity of Bahia de San Felipe, and C. f.
fuliginosus, occurring at localities on the lower
Pacific Slope, and south through the Central Des-
ert. These areas contain scrub vegetation.
The dark phase of C. f. fuliginosus was ob-
served twice during midmorning in open sandy
coastal scrub at station 27 on the Pacific Slope.
Both snakes were emerging from burrows.
Another dark racer was observed above station
4 on the lower East Scarp but could not be iden-
tified to subspecies.
Wilson (1973) indicated a single record for
Coluber flagellum piceus in the South Colorado
Desert, near San Felipe. Welsh and Bury (1984)
reported seven records from inland localities
north and west of San Felipe. These specimens
were all red phase snakes.
Coluber lateralis lateralis (Hallowell, 1853),
California Striped Racer
Two records exist for Coluber lateralis lateralis
on the East Scarp: Canon el Cajon (Linsdale
1932), and Canon del Diablo (SU 17298). On
the Pacific Slope, this race is known from five
localities: San Antonio Mine (Murray 1955),
Valladares Creek (CAS 56881), Socorro
(SDNHM 10524), and San Jose (SDNHM 5130
and 8856). I collected two specimens on the Pa-
cific Slope: one dead on the road in rocky chap-
arral at 1,320 m near station 29 (MVZ 140863),
and another at station 19 (MVZ 140864). An
individual was observed at 1 140 hr in a hollow
oak trunk in oak woodland; this snake was feed-
ing on a Peromyscus sp. Available records in-
dicate a preference for chaparral and riparian
habitats on the east and west scarps and on the
lower Pacific Slope.
Diadophis punctatus similis Blanchard, 1923,
San Diego Ringneck Snake
Schmidt (1922) recorded Diadophis punctatus
similis from San Matias Pass. This specimen was
probably collected at a spring south of the Pass,
and some 150 m higher, in pinyon-juniper
woodland. This race is known from Rancho San
Jose (SDNHM 36534), where it occurs in ripar-
ian habitat (Aida Meling, pers. comm.), and from
an irrigated slope of Arroyo San Telmo, south
of Colnett (R. Marlow, pers. comm.). These rec-
ords, and habitat preferences of the species in
other areas (Stebbins 1966), indicate it is prob-
ably restricted to the Pacific Slope of the Martir
Region, in riparian woodland corridors below
1,500 m elevation.
Elaphe rosaliae (Mocquard, 1899),
Baja California Rat Snake
Elaphe rosaliae, long known from San Bartolo
and Santa Rosalia in Baja California Sur (Schmidt
1922; see Ottley and Jacobsen 1983 for addi-
tional localities), has been reported more recently
from localities to the north. Hunsaker (1965) re-
WELSH: BAJA CALIFORNIA HERPETOFAUNA
35
ported two specimens at Canon Guadalupe on
the east scarp of the Sierra Juarez, and Stebbins
(1985) reported a specimen from Mt. Spring, Im-
perial Co., California. These records indicate E.
rosaliae ranges in the intervening area on the
eastern slopes of the Peninsular Range through
the Martir Region (Stebbins 1985).
Hypsiglena torquata deserticola Tanner, 1946,
Desert Night Snake
Hypsiglena torquata klauberi Tanner, 1946,
San Diego Night Snake
Tanner ( 1 966a) reviewed available material of
Hypsiglena torquata and proposed that the Pa-
cific Slope distribution for H. t. klauberi in Baja
California was through the Martir Region to the
North Central Desert. Bostic (1971) reported a
specimen of this race from near El Rosario. Tan-
ner ( 1 966fl) suggested that H. t. deserticola occurs
in northeastern Baja along the Gulf Slope. This
was based on five specimens from the Central
Desert Region that showed evidence of intergra-
dation between H. t. klauberi and H. t. desertico-
la. Tanner also indicated that the species occurs
near San Felipe. The above records and known
habitat preferences of this species (Stebbins 1966)
indicate that it probably inhabits all scrub and
xeric woodland areas on both slopes of the Re-
gion.
Lampropeltis getulus californiae
(Blainville, 1835),
California Kingsnake
This snake is known from the East Scarp at
Canon El Cajon (Linsdale 1932). I observed a
juvenile on the East Scarp at station 20. It was
active in shade at 0840 hr on a rocky slope of
pinyon-juniper woodland. This race is also
known on the Gulf Slope at Bahia de San Luis
Gonzaga (CAS 90227), an area of creosote bush
scrub. Lower Pacific Slope records are relatively
numerous, indicating that the snake occurs in
coastal sage and Central Desert scrub south to
El Rosario (SDNHM 1070, 41264). I observed
two specimens on the lower Pacific Slope, at sta-
tion 27; one was in coastal sage scrub, and the
other under a cotton wood tree (Populus sp.) in a
grassy field at the edge of riparian habitat. This
snake has not been reported from the central
South Colorado Desert or from the upper Sierran
scarps.
Schmidt ( 1 922) reported Lampropeltis getulus
yumensis from Volcano Lake in the Colorado
Desert of extreme northeastern Baja California.
Two records from 1 1.6 km east of Cerro Prieto
(32°25'N) are also in this area (MVZ 10761-62).
Lampropeltis zonata agalma
Van Denburgh and Slevin, 1923,
Baja California Mountain Kingsnake
I collected a single Lampropeltis zonata agal-
ma active at 1 635 hr in an area of rocky chaparral
at 1,600 m, 1.7 km west of station 29 (MVZ
140871). This locality was about 10 km down
slope from the nearest coniferous forest habitat.
All other locality records for this snake in the
Martir Region (Zweifel 1952, 1974; SDNHM
46017; and R. Moran, pers. comm.) are from
higher elevations, up to 2,800 m, in areas of
coniferous forest.
Throughout its range, Lampropeltis zonata oc-
curs sympatrically with the wide-ranging L. ge-
tulus, whose range circumscribes the areas oc-
cupied by L. zonata (Stebbins 1966). Storer and
Usinger (1963) indicated that L. getulus is rare
or absent in coniferous forests of the Sierra
Nevada of California, habitat where L. zonata is
found most commonly. L. getulus occurs up to
2,100 m (Stebbins 1966), but is uncommon in
coniferous forest habitats, even where L. zonata
is absent. Perhaps L. getulus is limited by an
elevational factor such as temperature, and L.
zonata replaces it in these higher, cooler areas.
Out of coniferous forest habitats L. zonata is un-
common, but it does occur marginally in chap-
arral throughout its range (Zweifel 1952, 1974;
this study).
Phyllorhynchus decurtatus perkinsi
Klauber, 1935,
Western Leafnose Snake
This snake is known in the Martir Region only
from the South Colorado Desert of the Gulf
Slope; records exist from 21.8 km north of San
Felipe (SDNHM 18917) and 1.6 km north of
San Felipe (SDNHM 38108). Welsh and Bury
(1984) reported four additional specimens from
localities 50 km northwest of San Felipe. The
subspecies P. d. decurtatus occurs from south of
the Martir Region through central Baja Califor-
nia and south into the Cape Region. Specimens
are known from 16 km north of Rancho Cata-
36
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
vina (LACM 20751), 16.2 km south of Punta
Prieta (Bostic 1971), 16.7 km southeast of Mes-
quital (MVZ 50171) in the Central Desert, and
from Bahia de Los Angeles (SU 2 1 800) on the
Gulf coast. R. Seib (pers. comm.) reported 14
additional records from the Central Desert
southward.
Pituophis melanoleucus annectens
Baird and Girard, 1853,
San Diego Gopher Snake
Pituophis melanoleucus bimaris Klauber, 1 946a,
Baja California Gopher Snake
Pituophis melanoleucus affinis Hallowell, 1852,
Sonoran Gopher Snake
Klauber (\946d) reported Pituophis melano-
leucus annectens from numerous localities on the
Pacific Slope of the Martir Region up into co-
niferous forest, and north into California. Mur-
ray (1955) reported this form at 1,400 m in chap-
arral on the West Scarp. It is known along the
coast at San Antonio del Mar (SDNHM 23800
and 32149), and Bostic (1971) reported this race
in the Central Desert, 31.8 km southeast of El
Rosario.
I collected three snakes that all correspond in
pattern and color with Pituophis melanoleucus
annectens. A snake was found crossing a road at
1730 hr, 15.2 km east of station 7 (MVZ 140869);
a second individual was found on the road at
1400 hr, 4.8 km west of station 29 (destroyed in
transit); and a third was found dead on the road
in midafternoon at station 28 (MVZ 140870).
The first locality was in an area of Central Desert/
coastal sage scrub, and the other two sites were
in areas of coastal sage scrub/chaparral ecotone.
The subspecies Pituophis melanoleucus bi-
maris occurs throughout the Central Desert
(Klauber 1 946a), whereas P. m. affinis is known
from two localities in extreme northeastern Baja
(Klauber \946a). The latter form has not yet
been reported in the desert east of the Sierra San
Pedro Martir, but Welsh and Bury (1984) re-
ported a specimen from 9.1 km east of San Ma-
tias Pass that appears to be an intergrade between
P. m. affinis and P. m. annectens.
Klauber (1946«) noted overlap in the ranges
of Pituophis melanoleucus annectens and P. m.
bimaris on the Pacific Slope through the Martir
Region, from El Rosario north to Alamo; he found
no evidence of intergradation in the specimens
he examined from this area.
Rhinocheilus lecontei lecontei
Baird and Girard, 1853,
Western Longnose Snake
Klauber (1941) reported two records from ex-
treme northeastern Baja California. Other rec-
ords from this area include: 51.2 km east of Mex-
icali (LACM 36576), east of Mexicali (SDNHM
19884-85; LACM 20821), and Tajo Canyon
(SDNHM 1 8875). Few records exist for the east-
ern Martir Region; Medica (1975) reported one
specimen from San Felipe, and Welsh and Bury
(1984) reported a specimen from 50 km north
and 12 km west of San Felipe. There are several
records from the Pacific Slope, both north into
California, and within the Martir Region as fol-
lows: 3.2 km and 11.2 km south of Santo Tomas
(LACM 59125, 102711), 8.3 km east of Cabo
Colnett (SDNHM 34003), 3.3 km south of San
Simon (SDNHM 42094), 3.3 km north of San
Simon (SDNHM 42632), and 12.7 km north of
Arroyo Seco (SDNHM 42543). These Pacific
Slope localities are in areas of coastal sage and
Central Desert scrub, whereas the records from
the Gulf Slope correspond with areas of creosote
bush scrub. Records from 1 .7 km west of El Mar-
mol (SDNHM 42439), 24.2 km south of Rancho
Santa Ines (CAS 143725), and San Borja
(SDNHM 42439) confirm this species in the
Central Desert. Lockington (1880) reported this
snake from the vicinity of Magdelena Bay in Baja
California Sur, but Klauber (1941) doubted the
validity of this record. The two recent specimens
from the Central Desert Region, and a record
from Cerralvo Island off the Cape Region (Soule
and Sloan 1 966), give renewed credence to Lock-
ington's record. Medica (1975) stated that this
species will probably be found to inhabit all of
Baja California.
Salvadora hexalepis hexalepis (Cope, 1867),
Desert Patchnose Snake
Salvadora hexalepis klauberi Bogert, 1945,
Baja California Patchnose Snake
Salvadora hexalepis virgultea Bogert, 1939,
Coast Patchnose Snake
A specimen of Salvadora hexalepis was col-
lected at 0707 hr crossing a sandy road in chap-
WELSH: BAJA CALIFORNIA HERPETOFAUNA
37
arral-coastal sage ecotone at 915 m, 5.8 km east
of station 27 (MVZ 140865). A second individ-
ual was found at 0906 hr in chaparral at station
11 (MVZ 140866). Both specimens fit the de-
scription of S. h. virgultea (Bogert 1945). This
subspecies is known from the Pacific Slope of
northern Baja California, from California to the
vicinity of El Rosario. Records in the Martir Re-
gion indicate an affinity to areas of chaparral,
coastal sage, and Central Desert scrub vegeta-
tions.
Bogert (1945) described Salvadora hexalepis
klauberi from the Central Desert, and indicated
a zone of intergradation between S. h. klauberi
and S. h. virgultea in the vicinity of latitude 30°N
(El Rosario). His speculation that S. h. hexalepis
might range into northeastern Baja in the Col-
orado Desert was confirmed by a specimen from
the Gulf Slope of the Martir Region reported by
Welsh and Bury (1984).
Sonora semiannulata Baird and Girard, 1853,
Western Ground Snake
Frost (1983) placed the two banded ground
snakes of Baja California, Sonora bancroftae and
S. mosaueri, in synonymy with the unbanded S.
semiannulata; at that time he concluded that in-
sufficient evidence existed to describe subspe-
cies. Despite Frost's conclusion I think it im-
portant to bear in mind that the banded forms
are superficially different from the unbanded S.
semiannulata, and that their respective distri-
butions are not inconsistent with other unique
elements of the Baja California herpetofauna.
Because of the possible zoogeographic implica-
tions of these populations I have chosen to treat
the banded forms here as subspecies.
Stickel (1943) indicated that S. semiannulata
occurred through the Colorado and Central Des-
ert Regions of Baja California to Santa Rosalia;
however Frost (1983) indicated a lack of records
to confirm the species in the Colorado Desert
east of the Sierra San Pedro Martir.
Klauber ( 1 943) described the banded form So-
nora bancroftae from a single specimen collected
in Arroyo San Telmo, 3.3 km east of San Jorge,
a site on the lower Pacific Slope of the Martir
Region, in an area of Central Desert/coastal sage
scrub ecotone. A second individual is known from
Rancho Buena Vista (SDNHM 36527), 27 km
further east and slightly higher in the same arroyo
in coastal sage scrub.
Stickel (1938) described another banded form,
Sonora mosaueri, from three specimens collect-
ed at Comondu in Baja California Sur. It has
since been reported from the Vizcaino Desert in
the vicinity of Santa Rosalia (SDNHM 44468
and 46808), and in the Central Desert, 37 km
south of El Rosario (SDNHM 46802). Stickel
(1 943) noted that S. mosaueri was closely related
to, and might prove to be a subspecies of 5".
semiannulata. The distribution of the three forms
of Sonora are in disparate, yet contiguous, areas
of desert scrub vegetations.
Tantilla planiceps (Blainville, 1835),
California Blackhead Snake
Cole and Hardy (1981) recognized no subspe-
cies of Tantilla planiceps. Tanner (1966&) re-
ported five specimens from the Pacific Slope of
Baja California, three from north of the Martir
Region, one from 6.4 km south of San Vicente
in the Western Foothills, and one from the Cen-
tral Desert. Bostic (1971) reported a specimen
from the Central Desert, 39 km southeast of El
Rosario. These localities indicate that T. plani-
ceps occurs in coastal sage and Central Desert
scrub vegetations within the Martir Region. A
single record exists from the Gulf Slope at Ar-
royo El Tajo on the east scarp of the Sierra Juarez
(SDNHM 45001), indicating T. planiceps is
probably also present on the Gulf Slope of the
Martir Region.
Thamnophis elegans hueyi
Van Denburgh and Slevin, 1923,
San Pedro Martir Garter Snake
This endemic subspecies of Thamnophis ele-
gans is found only on the Main Scarp of the Sierra
San Pedro Martir (Van Denburgh and Slevin
1923) and occurs in riparian meadow and nearby
moist woodland habitats of the coniferous forest.
Elsewhere, Thamnophis elegans occurs at all
elevations up to 3,000 m, and is associated with
a variety of terrestrial and riparian habitats
(Stebbins 1966). In the Martir Region, T. elegans
is found only in and around marshy meadows
and slow streams above 1,820 m. T. hammondii
is also present, but uncommon in these areas; it
occurs more frequently in and around the fast-
flowing streams in the steep riparian areas below
the Main Scarp where T. elegans is absent. My
observations of foraging of Thamnophis spp. in
38
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
the Martir Region suggest that both species may
be localized to riparian habitats because of the
moisture requirements of their amphibian prey
species. The usually eurytopic T. elegans may be
further limited to marshy habitats of the main
scarp by the presence of T. hammondii. Foraging
strategies may be an important factor in segre-
gating these two species in areas of sympatry such
as the Lower Main Scarp. T. hammondii is no-
ticeably more aquatic, and was seldom observed
out of water. It was observed in pursuit of anuran
larvae and Hyla cadaverina. T. elegans seemed
to prefer moist or damp habitats but was not
observed in the water. It was observed feeding
on toads on marshy ground.
An adult female collected 22 June 1974 (sta-
tion 24) (MVZ 140876) contained eight small
ova. An adult female found under a log at 1916
hr had recently ingested an adult Bufo boreas.
Specimens were collected at stations 12 (MVZ
140875), 25 (MVZ 140872), 37 (MVZ 140873),
and 40 (MVZ 140874); observations were made
at stations 24 and 39.
Thamnophis hammondii (Kennicott, 1 860),
Two-striped Garter Snake
This snake occurs on the East Scarp and the
Pacific Slope of the Region, frequenting only ri-
parian habitat. It occurs to 2, 1 20 m on the Lower
Main Scarp in coniferous forest, on the East Scarp
in creosote bush scrub and pinyon-juniper
woodland, and on the Pacific Slope in arroyos
dominated by chaparral and coastal sage scrub,
ranging west to the San Quintin Plain. A juvenile
was observed at station 65 in an irrigation ditch
that flowed from an apparently permanent spring
in Arroyo Edwardo, an area of Central Desert
scrub. Records from El Rosario (SDNHM 41593;
LACM 1 04069-7 1 ) are the southernmost for this
species.
Thamnophis hammondii was observed under-
water, along streamsides, on streamside rock
faces, and up to 3.4 m above ground in riparian
overstory (usually willow). I twice observed
snakes in apparent pursuit of Hyla cadaverina.
On three occasions, T. hammondii were ob-
served underwater stalking and lunging at anuran
larvae.
Specimens were collected at stations 1 7 (MVZ
140880), 37 (MVZ 140887, 147528), and 49
(MVZ 140878-79). Observations were made at
stations 10, 16, 20, 21, 23, 27, 38, and 65.
Trimorphodon biscutatus vandenburghi
Klauber, 1924,
California Lyre Snake
An adult male, matching Gehlbach's (1971)
description for Trimorphodon biscutatus van-
denburghi, was found dead on a dirt road on the
morning of 22 June 1973, at 1,520 m in rocky
chaparral, 3.5 km west of station 29 (MVZ
140877). This subspecies is known in north-
western Baja California from six other speci-
mens: 1 6 km north of Arroyo Seco on the San
Quintin Plain (Murray 1955), 2.9 km north of
El Rosario de Arriba (LACM 121552), 4.8 km
east of La Rumerosa (SDNHM 121 149), 2.2 km
south of Catavina (CAS 146577), 27 km south
of Santa Ines (MVZ 161561), and 32 km south-
east of Punta Prieta (LACM 126261). Scott and
McDiarmid (1984) showed a lack of records from
the Gulf Slope of northeastern Baja California;
Welsh and Bury (1984) reported the first speci-
men from 34 km east of San Matias Pass in the
South Colorado Desert. Habitat preferences
(Stebbins 1 966) suggest the probable occurrence
of this snake in pinyon-juniper, chaparral, and
all scrub vegetations of the Martir Region. The
absence of contiguous rocky habitat may limit
its distribution in parts of the Colorado Desert.
Gehlbach (1971) indicated that T. b. lyrophanes
of southern and central Baja California is known
as far north as San Ignacio in the Vizcaino Des-
ert.
Vipiridae
Crotalus atrox atrox Baird and Girard, 1853,
Western Diamondback Rattlesnake
This snake ranges into the Colorado Desert of
extreme northeastern Baja California (Schmidt
1 922; Linsdale 1 932), and has been reported from
Central Baja California (Alvarez and Huerta
1 974). It has yet to be found in the Martir Region.
Suitable habitat for this species occurs in the
South Colorado Desert; its apparent absence may
be due to limited collecting in this remote region.
Crotalus cerastes laterorepens Klauber, 1 944,
Colorado Desert Sidewinder
Crotalus cerastes laterorepens is known from
numerous localities in the South Colorado Des-
ert from San Matias Pass to the vicinity of San
Felipe (Linsdale 1932; Murray 1955; Welsh and
WELSH: BAJA CALIFORNIA HERPETOFAUNA
39
Bury 1984), south to San Francisquito on the
Gulf coast (Klauber 1 944). Residents of the Valle
de Trinidad area informed me that it occurs at
that Pacific Slope locality. All regional localities
for this species are areas of creosote bush scrub.
It has not been reported from the Pacific Slope
south of the Martir Region and the Gulf Slope
south of San Francisquito. It may be excluded
from these deserts by the presence of the closely
related form C. enyo.
Crotalus enyofurvus Lowe and Norris, 1954,
Dusky Baja California Rattlesnake
Crotalus enyo furvus was described by Lowe
and Norris (1 954) from four specimens collected
along the San Quintin Plain, north of El Rosario.
Their proposed range for C. e. furvus, based on
vegetative and edaphic affinities, corresponds
with Sonoran habitat of the Pacific coast of the
Martir Region (Short and Crossin 1967), herein
considered as part of the Central Desert scrub
vegetational association. Lowe and Norris's pro-
posed range also corresponds closely to the San
Quintin Plain. A record from 4.6 km north of
Arroyo Seco (SDNHM 46196) is just north of
this proposed range, in Central Desert scrub. Short
and Crossin (1967) indicated that this habitat
extends north to the vicinity of Rio San Antonio
along the base of the foothills and as far as Punta
Banda on the coast. Future work in the area may
reveal that the northern limit of this snake cor-
responds closely with this habitat.
Crotalus enyo enyo is known in more arid hab-
itats as far north as 17.2 km southeast of El Ro-
sario (Bostic 1971), and from numerous southern
localities of the Central Desert, Vizcaino, and
Cape Regions (Lowe and Norris 1954).
Crotalus mitchellii pyrrhus (Cope, 1867),
Southwestern Speckled Rattlesnake
Crotalus mitchellii pyrrhus occurs on the Gulf
and Pacific slopes of the Martir Region. I ob-
served two snakes on the rocky East Scarp at
stations 17 and 21 (MVZ 140884), in creosote
bush scrub and pinyon-juniper, respectively. A
third individual was observed on the lower West
Scarp at station 28 in coastal sage-chaparral eco-
tone. Klauber (1936) indicated the occurrence of
other records from the Martir scarps, Western
Foothills, and northward on both slopes into the
southern U.S. The three snakes I observed were
active during midmorning among boulders. Lit-
erature records and my observations indicate that
C. m. pyrrhus is primarily saxicolous, although
some records exist for other than rocky habitat
(Klauber 1972).
South of the Martir Region on the Gulf Slope,
Crotalus mitchellii pyrrhus is known from Bah ia
de Los Angeles (SDNHM 1 97 1 3-14). Specimens
from 4.8 km north of (LACM 20005) and 28.8
km south of (LACM 20006) Rancho Catavina
are apparently the first records of this species
from the Pacific Slope of the Central Desert. The
subspecies C. m. mitchellii occurs in the Vizcaino
Desert and south into the Cape Region of the
Peninsula.
Crotalus ruber ruber Cope, 18926,
Red Diamondback Rattlesnake
This subspecies occurs from San Diego Coun-
ty, California, south through the Martir Region
and the Central Desert (Klauber 1972). It has
been reported from numerous localities on the
lower Pacific Slope in the Western Foothills and
on the San Quintin Plain (Klauber 1 972). I found
it in the sandy desert at the base of the East Scarp.
Welsh and Bury (1984) reported it from a Gulf
coast locality north of San Felipe.
I observed two specimens at the base of the
East Scarp in creosote bush scrub. One was found
active in late afternoon shade on an open sandy
slope at station 54, the other at station 17, coiled
in the midafternoon shade of an overhanging
rock among talus. I collected one specimen at
night on the Pacific Slope, beside a dirt road, 5
km north of El Burro (station 3) (MVZ 140885),
in an area of chaparral-piny on-juniper ecotone.
Crotalus viridis helleri Meek, 1905,
Southern Pacific Rattlesnake
I found nine records of Crotalus viridis helleri
on the interior lower Pacific Slope of the Martir
Region, all corresponding with areas containing
riparian woodland habitat: La Toya, Valladares
Creek, and Rancho San Jose. Other records from
the Martir and Central Desert regions indicate
that this species has an affinity for coastal hab-
itats: San Antonio del Mar (CAS 57555); West
of San Quintin Bay (SU 12141-42); Camalu
Plains, 8.3 km north of Santo Domingo (SU
1 1574); Punta Camalu (CAS 40140); Playa Ma-
ria Bay (Klauber 1949); and 16.7 km south of
40
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
Jesus Maria (Bostic 1971). In the Sierra San Pe-
dro Martir, I found C. v. helleri up to 2, 1 20 m
on the Lower Main Scarp at stations 37 (MVZ
140822), 38 (MVZ 140883), and 39 (MVZ
147529), in coniferous forest and in conifer-
chaparral ecotone. This snake was most often
encountered in riparian edge habitat in these
areas. I also found a specimen in open chaparral
at 1,300 m. It was coiled in sand on a road at
2008 hr. C. v. helleri has not been reported from
the Gulf Slope of the Martir Region.
This species frequents riparian, coastal, and
upland habitats but is apparently absent from
inland scrub areas of the Martir and Central Des-
ert regions. However, two records are known from
arid situations in the Central Desert: 33.4 km
northwest of Bahia de Los Angeles (Klauber
1949), and 33.4 km southeast of Mezquital
(Murray 1 955). Specimens from the Central Des-
ert are described by Murray (1955) and Bostic
(1971) as light in color. Specimens I found in the
Martir Region were dark, almost black.
Jones (1981) and Ottley and Hunt (1981) re-
ported two records ofCrotalus viridis helleri from
Baja California del Sur; these specimens repre-
sent the southernmost records for the species.
ZOOGEOGRAPHIC ANALYSES
I used both a numerical analysis and a heuristic
comparison of distributions (based on locality
records listed in Species Accounts) across the
matrix of ecogeographic formations (Fig. 1 4) to
determine distributional patterns within the Re-
gion. Regional patterns were then extrapolated
beyond the Region to include the entire distri-
bution of each species in order to test for overall
geographic congruency between species linked by
a common ecogeographic pattern within the Re-
gion (analysis of tracks). Evidence of such overall
distributional congruency was considered indic-
ative of a common evolutionary history among
linked species.
Numerical Analysis
Cheetham and Hazel (1969) reviewed and dis-
cussed the use of numerical methods (Faunal Re-
semblance Factors or FRFs) in biogeography,
noting the inherent advantages, disadvantages,
and assumptions associated with binary coeffi-
cients. They listed some 22 possible formulas for
expressing the relationship of the number of taxa
in common between paired geographic units to
the total number of taxa present. In order to
ameliorate the disadvantages inherent with any
particular FRF, I employed three different FRFs
that tend to complement one another by em-
phasizing different aspects of faunal assemblages.
An FRF that emphasizes similarities can indicate
historical affinities and evolutionary relation-
ships between geographic areas, whereas an FRF
that emphasizes differences can test for sampling
error and the effects of faunas of unequal size,
situations minimized by similarity coefficients.
In addition to using a coefficient of similarity
and a coefficient of difference, I used a third coef-
ficient, one that emphasizes ecological relation-
ships between geographic units— a community
coefficient (Jaccard 1902). Subspecific differ-
ences were ignored for this analysis; distributions
were determined by species.
All available records (see Species Accounts)
were plotted on a schematic, southwest to north-
east, regional transect in the vicinity of latitude
3 1°N (all localities used are between 30°30'N and
3 1°30'N latitude). The ecotones between regional
ecogeographic formations (Fig. 1 4) were used to
delineate the units along the transect used for
numerical comparison. A species was considered
to occur within an ecogeographic formation only
if it ranged across the ecotone; marginal species,
those found only to occur in the ecotone and not
in a particular ecogeographic formation, were not
included. The fauna of each formation was com-
pared with those of all other formations using
the following three FRFs: Savage's (1960) Coef-
ficient of Difference (CD = 1 - C/N2 x 100);
Simpson's (1960) Similarity Coefficient (SC =
C/N, x 100), and Jaccard's (1902: in Morafka
1977) Community Coefficient (CC = [(C/N, +
N2 - C) x 100]), where C = the number of
species in common, N, = the number of species
in the area with the fewest species, and N2 = the
number of species in the area with the most
species.
Faunal resemblance factors are descriptive
rather than statistical indices (Cheetham and Ha-
zel 1969) and thus do not have levels of signif-
icance. I established a "level of significance" for
each coefficient on the basis of usage by previous
workers, and the relative correlation of the dif-
ferent coefficient values derived from the com-
parison of geographic units where N, approaches
or equals N2. My primary emphasis, when eval-
WELSH: BAJA CALIFORNIA HERPETOFAUNA
41
uating FRF values, was to weigh all three values
for each pair of geographic units, treating all units
relative to one another, and to look for obvious
clusters occurring in ranked coefficient values.
Savage (1960) arbitrarily selected CD = 50% or
greater as indicative of a unique faunal grouping
and I follow his usage as a basis. Employing the
50% level of significance with Jaccard's CC and
Simpson's SC indicated to me that they are more
sensitive and less sensitive, respectively, than the
CD for distinguishing unique faunal groupings.
I have therefore partially weighted these coeffi-
cients. Those ecogeographic formation pairs with
at least two of three coefficients meeting the fol-
lowing criteria: CD < 50%, CC > 40%, and SC
> 60%, were considered to be related at a sig-
nificant level.
Thirteen pairs met my criteria for significant
relatedness (Table 4). The relationships between
these 1 3 pairs are depicted phenographically (Fig.
15a-c). Subsequent pairwise analyses between
ecogeographic formations using both average
linkage (UPGMA) and single linkage (nearest
neighbor) methods (SAS 1985) yielded results
nearly identical to those depicted in Figure
15a-c.
Results of the numerical comparison indicate
the presence of four distinct herpetofaunal areas
within the Martir Region, each showing some
degree of overlap with one or more of the ad-
jacent areas (Fig. 1 5a-c). These areas are:
( 1 ) The Upper Main Scarp and the Lower Main
Scarp— these two formations support the same
montane herpetofauna (SC = 100%) with nu-
merous species dropping out with increasing
elevation (CD = 60%). The unequal fauna sizes
between these two formations (7 vs. 20) account
for the high SC and low CC values (CC = 40%).
(2) The Lower Main Scarp, West Scarp, and
the Western Arroyos— these three formations
support a fauna that apparently shares some eco-
logical characteristics and possibly a convergent
history with that of (1) above, but differ in their
adaptation to the warmer and drier habitats of
lower elevations, extending into some relatively
extreme xeric areas where these species occur
along the riparian corridors of the Western Ar-
royos, and similar habitats of the East Scarp.
(3) East Scarp and West Scarp— the high SC
and CC and the low CD values between these
formations indicate strong faunal affinities be-
tween these two rocky xeric woodland areas.
However, each of these two formations dem-
onstrates only slightly less significant, differing
affinities, with proximal formations; a situation
that initially seems somewhat contrary and con-
fusing. The East Scarp shows close faunal links
with the Colorado Desert and the Western Foot-
hills, while the West Scarp shows close affinities
with the Western Arroyos and the Western Foot-
hills (the FRF comparisons of the West Scarp
and the Western Foothills are not depicted phe-
nographically in Fig. 5a-c because they fell just
below my levels of significance: CD = 49%, SC
= 55%, and CC = 36%). I interpret these data to
mean that two distinct overlapping faunas are
present on the East and West scarps, each con-
tinuing with slightly differing distributions onto
adjacent formations.
(4) The Colorado Desert, North Central Des-
ert, San Quintin Plain, and the Western Foot-
hills—all coefficient values between these four
relatively open, scrub desert formations indicate
strong faunal ties.
Further analysis of these numerical compari-
sons follows in conjunction with the results of
the heuristic analysis below.
Heuristic Analysis
The heuristic analysis of distributions of re-
gional species consisted of two parts: ( 1 ) a com-
parison of ecogeographic distributions along a
schematic, southwest to northeast, regional tran-
sect between latitudes 30°30'N and 31°30'N (re-
sults depicted in Table 5); (2) a comparison of
overall regional distributions established by plot-
ting available literature and field records (Species
Accounts) on maps based on Figures 3, 4, and 5
(results illustrated in Fig. 1 6). A species that oc-
curred at two or more localities within a contig-
uous vegetation or physiographic area was as-
sumed to occur throughout the contiguous habitat
where suitable microhabitat existed, but not
above or below the highest or lowest elevation
records. I included locality records from the Sier-
ra Juarez Region to the north and the Central
Desert Region to the south to discern distribu-
tional trends over a wider area.
Three distinct patterns were apparent from the
transect analysis (Table 5): (1) the High Sierran
Fauna (I), with distributions corresponding
closely to those of the Upper and Lower Main
Scarp formations, areas linked by numerical
42
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
WELSH: BAJA CALIFORNIA HERPETOFAUNA
43
TABLE 4. COMPARISON OF REGIONAL HERPETOFAUNAS BY ECOGEOGRAPHIC FORMATION (Fie. 14) USING THREE FAUNAL COEF-
FICIENT FACTORS.' Number of species present in parentheses. C = number of species in common. Formation pairs with two of
three coefficients as follows are considered to be related at a significant level: CD < 50%, SC > 60%, CC ^ 40%.
East Scarp
(26)
Upper Main Scarp
(8)
Lower Main Scarp
(20)
West Scarp
(31)
West Foothills
(29)
C
CD
SC
CC
C
CD
SC
CC
C
CD
SC
CC
C
CD
SC
CC
C
CD
SC
18
49%
69%
41%
0
100%
0%
0%
3
92%
15%
7%
12
66%
38%
22%
22
38%
75%
0
100%
0%
0%
9
66%
45%
24%
20
36%
76%
54%
16
45%
61%
8
60%
100%
40%
6
81%
75%
18%
1
97%
12%
16
49%
80%
45%
5
83%
25%
16
49%
55%
CC
52%
46%
3%
11%
36%
West Arroyos
C
2
9
4
12
16
5
(22)
CD
94%
66%
82%
46%
49%
83%
SC
9%
40%
50%
60%
72%
22%
CC
4%
23%
15%
37%
43%
10%
San Quintin Plain
C
19
13
2
5
15
22
8
(30)
CD
46%
57%
94%
84%
52%
27%
74%
SC
63%
50%
25%
25%
50%
75%
36%
CC
41%
30%
6%
11%
32%
59%
18%
Central Desert
C
26
15
2
6
14
23
6
23
(32)
CD
27%
54%
94%
82%
57%
29%
82%
29%
SC
81%
57%
25%
30%
45%
79%
27%
76%
CC
63%
34%
5%
13%
28%
60%
12%
58%
Colo.
East
Upper
Lower
West
West
West
San
Desert
Scarp
Main
Main
Scarp
Foot-
Ar-
Quintin
Scarp
Scarp
hills
royos
Plain
(35)
(26)
(8)
(20)
(31)
(29)
(22)
(30)
1 CD = Coefficient of Difference (CD = 1 - C/N2 x 100) (Savage 1960); SC = Similarity Coefficient (SC = C/N, x 100)
(Simpson 1960); CC = Community Coefficient [CC = (C/N, + N2 - C) x 100] (Jaccard 1902, in Morafka 1977). N, = number
of species in the area having the fewest species, N2 = the number of species in the area with the most species.
comparison above, and indicating a montane
fauna; (2) the Californian Fauna (II) with distri-
butions corresponding closely to those of the sec-
ond faunal area indicated numerically above, and
encompassing primarily the Lower Main and
West scarps and the Western Arroyos; and (3)
the Colorado Desert Fauna (III). The endemic
Colorado Desert Fauna is apparently too small,
relative to the total number of xerophilic species
present within the South Colorado Desert for-
mation, to have been detected by the FRF anal-
ysis. The considerable overlap in ecogeographic
FIGURE 15. Relationships between herpetofaunas of the ecogeographic formations (Fig. 14) based on (a) Coefficients of
Difference (CD) (Savage 1 960), (b) Coefficients of Similarity (SC) (Simpson 1 960), and (c) Community Coefficients (CC) (Jaccard
1902, in Morafka 1977). Width of bars is directly proportional to degree of relationship, length of solid bars is inversely
proportional to degree of relationship. Formation pairs with CD > 0.50 not included (except the Upper Main Scarp for which
the smallest value is included). Formation pairs with SC < 0.60 not included. Formation pairs with CC < 0.37 not included.
See Table 4 for all pairwise values.
44
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
TABLE 5. ECOGEOGRAPHIC DISTRIBUTION OF THE HERPETOFAUNA ON A SCHEMATIC TRANSECT OF THE SIERRA SAN PEDRO
MARTIR REGION, BAJA CALIFORNIA NORTE, MEXICO, IN THE VICINITY OF LATITUDE 3 1°N. Faunal groupings are based on coincident
regional (Fig. 16) and overall areal (Fig. 11-21) distributions. Solid lines indicate occurrence throughout a formation; X without
lines indicates occurrence restricted within a formation. (X) = occurrence only within Western Arroyos where they transect the
Western Foothills (not considered a part of the foothills fauna). ? ? = expected occurrence. * = resident of Pacific Northwest
Track (see text and Fig. 17). ** = resident of Madrean Track (see text and Fig. 18).
SPECIES
ECOGEOGRAPHIC FORMATION
N S W W W
C Q A F S
Regional Herpetof aunas D P
(CDS) (CDS) (RW) (CSS) (Chp)
L U
M M
S S
(Con) (Con)
E C
S D
(P-J.
CBS) (CBS)
I. High Sierran Scarps Fauna**
Rana boylii
X
X
X
X
X
X
X
X
X
X
X
X )
)
X
{
[ X
[
X
X
:
C
X X
X
X
7
X
hx
Batrachoseps pacificus ssp.
Sceloporus graciosus
Lampropeltis zonata
X
Thamnophis elegans
)
; x
X
X
X
II. Calif ornian Fauna**
Bufo boreas*
Bufo microscaphus
X
X
X X
Hyla cadaverina
X
X
X
(X)
(X)
(X)
X
X
X
X
X
X
X
X J
X 2
X }
C X
C X
C X
C X
X
Hyla regilla
Rana aurora*
Scaphiopus hanunondii
Batrachoseps p. major
X
1C
X
X
X
X
X
X
X
X
X
(X)
(X)
(X)
(X)
Clemmys marmorata*
1C }
Anniella pulchra
Eumeces gilberti
Eumeces skiltonianus
Elgaria multicarinata
X
X
Sceloporus occidentalis
X---
X
X
.-?__.
X
X
X
(X)
X
X
X
X )
Xantusia vigilis
X
)
Crotalus viridis
X
X
X
X
X
X
X
Diadophis punctatus
Coluber lateralis
Xj
?
X
X
X
X
(X)
(X)
(X)
e<
X
X
X
0
. .•>
•>_ _.
X---'
>
i X
Thamnophis hammondii
X
X )
:ones- •
?--x
III. Colorado Desert Fauna
Phrynosoma platyrhinos
Urosaurus graciosus
X
Chionactis occipitalis
X
Crotalus cerastes
X
NCD - North central Desert; SQP - San Quintin Plain; WA - Western Arroyos; WF = Western
Foothills; WS = West Scarp; LMS = Lower Main Scarp; UMS = Upper Main Scarp; ES = East Scarp;
CD » Colorado Desert; see figure 4.
( ) » Dominant vegetation; see table 2.
distribution among the xerophilic species of the
Region made it difficult to distinguish additional
distinct faunal assemblages using the schematic
ecogeographic transect format alone.
However, comparison of overall regional dis-
tributions indicated a total of seven patterns
among the regional herpetofauna (Fig. 1 6). Pat-
terns I, II, and III (Fig. 16) correspond with the
WELSH: BAJA CALIFORNIA HERPETOFAUNA
45
Table 5. Continued.
SPECIES
ECOGEOGRAPHIC FORMATION
N S
C Q
Regional Herpetofaunas D P
(COS) (CDS)
W W W L U E C
A F S H M S D
S S (P-J.
(RW) (CSS) (Chp) (Con) (Con) CBS) (CBS)
IV. Peninsular Range Fauna
A. Pacific Slope Forms:
Anniella geronimensis**
X
K X
X
X '
C
i
f
1C
!
C
?
>
X
;
*
?--
X
X
7
X
Cnemidophorus hyperythrus
Cnemidophorus labialis**
X
X
:
:
[
X
Phrynosoma coronatum
X
}
Crotalus enyo
X
X
X
7
1
3
Sonora s. bancroftae
X
X
X
X
? >
3
x 3
B . Rocky Scarps Forms :
Crotaphytus insularis
Petrosaurus mearnsi
Phyllodactylus nocticolus
X
X
Sauromalus obesus
X
X
Sceloporus orcutti
X
X
Urosaurus microscutatus
7
X
X
X
X
X
X
X
» ---X
X
X
X
Xantusia henshawi**
E
i
C
<L
i
C
C
t
i
C
X
X
X
X
X
X
X
X
X
?
X
X
3
Crotalus mitchellii
X
X- —
C. Flat land Desert Forms:
Callisaurus draconoides
X
X
X
X
?--
(
X
X
X
X
X
X
X
Dipsosaurus dorsalis
Chilomeniscus cinctus
•>
Crotalus ruber
X
3
_x— ?
-X X
Lichanura trivirgata
? X
3
X— -
:
Phyllorhynchus decurtatus
X
X
3
X
V. North American Ubiquitous
Scrubland Fauna
Bufo punctatus
X
-X--?
Schaphiopus couchii
Cnemidophorus tigris
X
Coleonyx variegatus
X J
C 31
X
X
?— X
X
X
X
Gambelia wislizenii
X
;
X
Sceloporus magister
X
;
X
Uta stansburiana
X
3
:
'-X
X
Arizona elegans
X
3
---X--
7
X
>
x,
X
Hypsiglena torquata
X
3
X
Lampropeltis getulus
7
X
X
X---
'--X
X
7
X
X
Leptotyphlops humilis
--X
3
. — ? — .
[
X
— X—
7
X
'--X
?
Coluber flagellum
Pituophis melanoluecus
X
> 1
X
7
X
7
-X— ?
--X-X
X
Rhinocheilus lecontei
X
3
:
l
ec
X— -'
>
Salvadora hexalepis
X
]
X
7
0
Sonora s. semiannulata
7
X-
7
-X
•-X "
. 7
__ -•>
Trimorphodon biscutatus
>
— — -
— X-?
:ones-
_
46
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
FIGURE 16. Distribution patterns of the herpetofauna of
the Sierra San Pedro Martir Region, Baja California Norte,
Mexico. Numerals refer to Table 5. In some cases member-
species occur within, but not across the entire pattern range
depicted: see Table 5 for exact regional distributions of indi-
vidual species. Broken lines indicate fewer species of grouping
indicated occur northward.
three derived above and numerically. The re-
maining species represent two relatively distinct
groupings that occur throughout the xeric scrub-
land formations on both the Pacific and Gulf
slopes (Fig. 16, pattern V), or only within some
of the scrubland formations on one or the other,
or both, slopes (Fig. 16, patterns IVa, IVb, IVc).
Several species subsequently assigned to pattern
IVc (Fig. 16) (Callisaurus draconoides, Chilo-
meniscus cinctus, Crotalus ruber, and Lichanura
trivirgatd) and pattern IVb (Fig. 16) (Crotalus
mitchellii) had regional distributions that were
indistinguishable from members of pattern V (Fig.
1 6) at this stage of analysis. Final assignment of
these species, and those assigned to pattern V,
was not possible until overall distribution pat-
terns were compared (see Analysis of Tracks be-
low).
The remaining three species assemblages (Fig.
16, patterns IVa, IVb, and IVc) are centered pri-
marily along the East or West scarps or both, and
the Western Foothills, corresponding with the
third faunal area derived numerically above (Fig.
15). This third FRF-derived faunal area is the
least clearly defined probably due to the complex
overlapping ecogeographic variations apparent
among the several groups of species within this
faunal area and their overlap with many of the
species of pattern V. Pattern V (Table 5) corre-
sponds closely with the fourth faunal area de-
rived by the FRF analysis and covering the four
scrubland desert formations (Fig. 15). Most
species of pattern IVa (Fig. 1 6; Table 5) occur
only on the Pacific Slope of northern or central
Baja California. Sonora semiannulata bancroftae
and Anniella geronimensis are endemic to the
Martir Region; Crotalus enyo and Cnemidoph-
orus labialis have their northern limits slightly
north of the Martir Region. The remaining two
members of pattern IVa (Phrynosoma corona-
turn and Cnemidophorus hyperythrus) range north
across the U.S. border on the Pacific Slope. All
of these species avoid rocky substrates, preferring
sandy soils. Members of pattern IVc also show
preference for sandy, fine-textured substrates, but
demonstrate wider regional and overall distri-
butions than members of pattern IVa, occurring
throughout the xeric and subxeric scrub habitats
on both the Pacific and Gulf slopes and ranging
extensively beyond the Martir Region (Fig. 16;
Table 5). Pattern IVb species (Fig. 16; Table 5)
are saxicolous species associated with the Pen-
insular Range scarps and bordering rocky scrub-
land habitats. Three of these species (Petrosaurus
mearnsi, Phyllodactylus nocticolus, and Sauro-
malus obesus) occur only along the Gulf Slope
north of 30°N latitude. The remaining pattern
IVb members occur on both slopes of the Pen-
insular Range regionally and northward, except
Crotaphytus insularis, which is absent from the
Pacific Slope north of 32°N latitude. Xantusia
henshawi is absent south of 30°N latitude.
Analysis of Tracks
Lastly, I conducted an analysis of the entire
distribution of each taxon in Baja California us-
ing the Species Accounts and the following lit-
erature: Van Denburgh (1922), Schmidt (1922),
Savage (1960, 1967), Soule and Sloan (1966),
Stebbins (1966, 1985), Bostic (1971), Loomis et
al. (1974), Conant (1975), Murphy (1983a), and
Murphy and Ottley (1983, 1984). This last stage
of the analysis sought to discern the generalized
WELSH: BAJA CALIFORNIA HERPETOFAUNA
47
tracks of the Baja California herpetofauna by
comparing the geographic distributions of con-
sanguineous forms (monophyletic lineages),
seeking spatial congruency among unrelated taxa,
the panbiogeographic method of Croizat ( 1 964;
see also Craw [1983] and Craw and Weston
[1984]). Distributions were examined and com-
pared for each lineage at the subspecific, specific,
and generic levels. In each case of a proposed
generalized track, the number of individual tracks
indicative of the described generalized track ex-
ceeded the minimum three taxon/area state-
ments (three individual tracks) considered as es-
sential to support a hypothesis of a generalized
track or pattern (Craw 1983).
I found evidence of at least five generalized
tracks among the herpetofauna of the Sierra San
Pedro Martir Region: a Northwest Pacific Track,
a Madrean Track, a Colorado Desert Track, a
Peninsular Range Track, and a North American
Desert and Plains Track.
The Pacific Northwest Track (Fig. 1 7)
The 20-plus species on this track are found
along the western Pacific rim from Baja Califor-
nia north through California, Oregon, Washing-
ton, and into Canada, with a few forms occurring
north to Alaska. Many species of this track have
allopatric congeners in the northeastern United
States. This track is characterized by endemic
salamanders, most of which occur north of 35°N
latitude in California; only the genera Ensatina
andAneides occur in Baja California. In southern
California and Baja California, members of the
Pacific Northwest Track (Table 6) occur as dis-
junct populations and are often isolated on
mountain ranges. Clemmys marmorata is the
only reptile of the Martir Region that I believe
is a member of this track. The anurans Bufo
boreas and Rana aurora are included on this
track, but their taxonomic and distributional
patterns do not match the pattern as well as the
other amphibians indicated. This may be due to
their euryecious nature, and they may in fact
have a common zoogeographic history with oth-
er track members. They may also represent
another, more recent overlapping track, as they
appear to be relatively young forms in North
America that may have relatively recently dif-
fused (Pielou 1979) south across cool temperate
regions (Savage 1960). The Martir faunal ele-
ments of the Pacific Northwest Track are mem-
FIGURE 17. Generalized tracks of the herpetofauna of the
Baja California Region: Pacific Northwest Track.
bers of either the High Sierran or Californian
faunas (Table 5; Fig. 16, 1 and II) and have their
southern terminus in the Martir Region.
The Madrean Track (Fig. 1 8)
The Madrean Track lineages exhibit a complex
pattern of distribution across southwestern North
America often involving two or more disjunct
forms, for each lineage, within Baja California
and additional disjunct, closely related forms, to
the north and east in California and Arizona,
with many lineages ranging into southwestern
New Mexico and south into the Sierra Occidental
of Mexico (not all member lineages presently oc-
cur across the full range of the track). These forms
avoid lowland xeric habitats for the most part,
or if they occur in such areas, they have adapted
to microhabitats, and temporal activity pat-
terns that ameliorate against the extremes of low
moisture and high temperature (e.g., Scaphiopus,
Xantusia spp.).
Within Baja California many of these forms
occur in disjunct populations, with some occur-
48
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
TABLE 6. RESIDENT TAXA OF THE GENERALIZED TRACKS OF THE HERPETOFAUNA OF THE BAJA CALIFORNIA REGION. Includes
only those track members that occur within the Baja California Region. ( ) refer to Figures 17-21.
Aneides lugubris
Ensatina eschscholtzii
Bufo boreas
Batrachoseps pacificus major
B. pacificus spp.
Scaphiopus hammondii
Bufo microscaphus
Hyla cadaver ina
H. regilla
Rana boylii
Sceloporus graciosus
S. occidentalis
Xantusia vigilis (2 spp.)
X. henshawi
Phrynosoma mcallii
P. p/atyrhinos
Uma notata
Chrysemys scripta
Bipes biporus
Coleonyx switaki
Phyllodactylus angelensis
P. apricus
P. bugastrolepis
P. homolepidurus nolascoensis
P. nocticolus (9 ssp).
P. partidus
P. santacruzensis
P. tinklei
P. unctus (3 ssp.)
P. xanti
Ca/lisaurus draconoides (6 ssp.)
Crotaphytus insularis (2 ssp.)
Ctenosaura hemilopha
Petrosaurus complex
P. mearnsi
P. repens
P. thalassinus
Phrynosoma coronatum complex
P. cerroense
P. coronatum (3 ssp.)
Bufo punctatus
Scaphiopus couchii
Coleonyx variegatus (6 ssp.)
Uta complex
U. antigua
U. nolascensis
U. pa/meri
U. squamata
U. stellata
U. stansburiana (2 ssp.)
Pacific Northwest Track (17)
Rana aurora
Clemmvs marmorata
Madrean Track (18)
Eumeces gilberti
E. skiltonianus complex
E. skiltonianus
E. lagunensis
Cnemidophorus labialis
Elgaria complex
E. cedrosensis
E. multicarinata
E. paucicarinata
Anniella pulchra
A. geronimensis
Diadophis punctatus
Coluber lateralis complex
C. aurigulus
C. barbouri
C. lateralis
Lampropeltis zonata
Thamnophis hammondii complex
T. hammondii
T. digueti
T. elegans
Crotalus viridis
Colorado Desert Track (19)
Urosaurus graciosus
Chionactis occipitalis
Crotalus cerastes
Peninsular Range Track (20)
Sator complex
S. an gust us
S. grandaevus
Sauromalus complex
5. ater (2 ssp.)
S. australis
S. hispidus
S. klauberi
S. obesus
S. slevini
S. varius
Sceloporus magister complex
5. m. lineatulus
S. m. monserratensis
S. m. rufidorsum
S. m. zosteromus
Sceloporus orcutti complex
S. hunsakeri
S. licki
S. orcutti
Urosaurus complex
U. lahtelai
U. microscutatus
U. nigricaudus
Cnemidophorus deppei complex
C. ceralbensis
C. hyperythrus (3 ssp.)
Lichanura trivirgata (3 spp.)
Chilomeniscus complex
C. cinctus
C. punctissimus
C. savagei
C. stramineus (2 ssp.)
Etaphe rosaliae
Eridiphas slevini (2 ssp.)
Nerodia valida
Phyllorhynchus decurtatus (3 ssp.)
Sonora semiannulata bancroftae
S. s. mosaueri
Tantilla planiceps
Crotalus atrox complex
C. catalinensis
C. exsul
C. ruber (2 ssp.)
C. tortugensis
C. enyo (3 ssp.)
C. mitchellii (4 spp.)
North American Desert and Plains Track (21):
Group A: old vicariated lineages
Leptotyphlops humilis (5 ssp.)
Hypsiglena torquata (7 ssp.)
Lampropeltis getulus complex
L. catalinensis
L. getulus (3 ssp.)
L. nitida
WELSH: BAJA CALIFORNIA HERPETOFAUNA
49
TABLE 6. CONTINUED.
Gainbelia wizlizenii
Sceloporus magister uniformis
Arizona elegans (3 ssp.)
Coluber flagellum (2 ssp.)
Pituophis melanoleucus (3 ssp.)
North American Desert and Plains Track (21):
Group B: post-vicariant lineages
Rhinocheilus leconti
Salvadora hexalepis (3 ssp.)
Sonora s. semiannulata
Trimorphodon biscutatus (2 ssp.)
Crotalus atrox
ring as allopatric species, occupying patches of
relatively mesic habitats scattered sporadically
down the length of the peninsula, with several
lineages present on one or two Gulf or Pacific
islands (e.g., Hyla regilla, Eumeces skiltonianus-
E. lagunensis, Elgaria multicarinata-E. pauci-
carinata, E. cedrosensis, Coluber lateralis-C.
barbouri and C. aurigulus, Thamnophis ham-
mondii-T. digueti, Xantusia vigilis, and Crotalus
viridis). The remaining lineages of this track (Ta-
ble 6) do not, for the most part, occur south of
30°N latitude in Baja California. North of 30°N
latitude into California and northward, and east
across the Mojave Desert through Arizona, New
Mexico, and south into Mexico in the Sierra
Madre Occidental, these taxa demonstrate a sim-
ilar pattern of allopatric distributions with many
of these lineages represented by two or three
species across the scope of the track (e.g., Bufo
microscaphus; Hyla cadaverina, H. regilla-H.
exima; H. arenicolor, Elgaria multicarinata-E.
panamintina-E. kingi; Xantusia v. vigilis-X. v.
arizonae, X. h. bolsonae, etc.; Lampropeltis zo-
nata-L. pyromeland).
The remaining lineages included on this track
(Table 6) are somewhat less clearcut in their pat-
terns of disjunction and taxonomic/geographic
correlation with the above listed taxa, but suf-
ficient aspects of the pattern are apparent in their
distributions to support their inclusion as ele-
ments of the Madrean Track.
Within the Martir Region, elements of the
Madrean Track demonstrate ecogeographic af-
finities coincident with the mesic habitats of the
Gulf and Pacific slopes (Table 5, I, II, and IV).
Colorado Desert Track (Fig. 1 9)
These species have distributions coincident
with the hottest and driest portions of the south-
western deserts, the area where the Mojave, Col-
orado, and Sonoran deserts merge. Martir faunal
elements present on this track occur regionally
only in the Colorado Desert formation (Table 5,
III; Fig. 16, III). Species of the Colorado Desert
Track demonstrate the most limited distribu-
tions and the lowest levels of geographic and
taxonomic diversity within lineages of all the
faunal assemblages described here for Baja Ca-
lifornia. Six species are included; the Colorado
Desert fauna of the Martir Region (Table 5, III),
plus Uma notata and Phrynosoma mcallii.
Peninsular Range Track (Fig. 20)
Lineages of this track have their overall dis-
tributions coincident with all or much of pen-
insular Baja California and the Gulf of California
islands, extending north along the Peninsular
Range to San Gorgonio Pass in southern Cali-
fornia. A few taxa of this track have radiated
northeast into the Mojave Desert (e.g., Crotalus
mitchellii, Lichanura trivirgata, and Dipsosaurus
dorsalis) and southeast into the Sonoran Desert
(e.g., Lichanura trivirgata, Chilomeniscus cinc-
tus, Phyllorhynchus decurtatus, Callisaurus dra-
conoides, Dipsosaurus dorsalis). Several interre-
lated characteristics of the lineages assigned to
this track are: (1) pronounced endemism; (2)
presence of most primitive extant forms; and (3)
greatest levels of geographic and taxonomic di-
versity within lineages, all within insular and
peninsular Baja California. Twenty-six species or
species-complexes meet these criteria (Table 6).
Twenty species-groups of the Peninsular Range
Track occur in the Martir Region (Table 5, IV),
and exhibit three variations in regional distri-
bution (Fig. 16, IVa, IVb, and IVc). Six taxa
occur further south in peninsular and insular Baja
California in the Central and Vizcaino deserts
and the Cape Region (Chrysemys scripta, Bipes
biporus, Ctenosaura hemilopha, Sator spp.,
Eridiphas slevini, and Nerodia valida).
50
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
FIGURE 18. Generalized tracks of the herpetofauna of the
Baja California Region: Madrean Track.
FIGURE 19. Generalized tracks of the herpetofauna of the
Baja California Region: Colorado Desert Track.
North American Desert and Plains Track
(Fig. 21)
Species of this track possess broad distribu-
tions throughout the North American Desert Re-
gion, and into the Great Plains Region. These
species are widespread in Baja California, oc-
curring throughout most insular and peninsular
xeric habitats. The 1 8 taxa of the North Amer-
ican Desert and Plains Track present in the Mar-
tir Region demonstrate broad ecogeographic dis-
tributions throughout the scrub and woodland
habitats of the Region (Table 5; Fig. 16, V).
The lineages of this track can be divided into
two subsets on the bases of distributional vari-
ations on Gulf of California islands, and the
level of divergence they show within the Baja
California Region as a whole: (1) those lineages
that show extensive divergence within the Baja
California Region, including endemic taxa on
some or all of the "old islands" (Soule and Sloan
1966; see also Murphy [1983&] and Murphy and
Ottley [1984]) of the Gulf of California (Table 6,
Desert and Plains Track— group A); and (2) those
lineages that show little divergence throughout
peninsular Baja California and are absent from
the "old islands" (Soule and Sloan 1966) (Table
6, Desert and Plains Track— group B). The taxa
of group A appear to have a shared evolutionary
history within the Baja California Region with
the lineages of the Peninsular Range Track (Ta-
ble 6). The taxa of group B appear to be relatively
recent (post-Pliocene) emigrants in the Baja Ca-
lifornia Region despite sharing contemporary
overall continental distributions with members
of group A.
Geographic diversity and endemism within the
lineages of this track, when considered within
peninsular and insular Baja California, are not
noticeably more extensive than in other parts of
the North American Region (in contrast with
member lineages of the Peninsular Range Track).
On the contrary, with many lineages of this track,
geographic diversity and endemism are greater
beyond, than within, insular and peninsular Baja
California.
WELSH: BAJA CALIFORNIA HERPETOFAUNA
51
FIGURE 20. Generalized tracks of the herpetofauna of the
Baja California Region: Peninsular Range Track. The extent
of this track in southwest Mexico is greater than indicated,
particularly southward.
FIGURE 2 1 . Generalized tracks of the herpetofauna of the
Baja California Region: North American Desert and Plains
Track.
DISCUSSION
Theories on the historical biogeography of the
herpetofauna of the Baja California Region have
advanced on three parallel fronts: (1) with in-
creased knowledge of contemporary herpeto-
faunal distributions and their ecogeographic af-
finities; (2) with new information about past
climates and vegetation assemblages— the geo-
floral literature; and (3) with the advent of a new
paradigm in geology— plate tectonics— depicting
a dynamic rather than static landscape. Nelson
(1921) and Schmidt (1922) compiled the first
syntheses based on early distributional data of
the peninsular biotas. Savage ( 1 960) constructed
paleoecological scenarios using data in the geo-
floral literature, and in conjunction with analyses
of contemporary herpetofaunal distributions,
presented a classic synthesis. Soule and Sloan
( 1 966) and Savage ( 1 967) discussed evolutionary
relationships of the island herpetofaunas of the
Gulf of California and proximate Pacific Ocean
islands, respectively. However, Savage and ear-
lier investigators based their analyses on the as-
sumption of a static or fixed peninsula, requiring
Recent distributions to have been derived pri-
marily from north to south dispersals in the case
of species present on the peninsula, or tempo-
rary land bridges or chance dispersal across water
barriers in the case of species present on islands.
More recently, in conjunction with developing
tectonic theory, Leviton and Tanner (1960),
Robinson (1973), and Smith and Tanner (1974)
examined biogeographic relationships within
single taxa, and indicated that particular species
of the southern Baja California herpetofauna ex-
hibit distributions more consistent with a theory
of vicariant origins resulting from past tectonic
events than with a theory of north to south dis-
persals down a fixed peninsula. Murphy (1975),
following this lead, proposed a "trans-gulfian"
vicariance model, based on the dislodging of pre-
peninsular land masses from further south along
52
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
coastal Mexico during the Miocene, to explain
contemporary distributions of many of the sub-
tropical associated elements of the Cape Region
and associated deep-water islands (Bipes biporus,
Nerodia valida, Ctenosaura hemilopha, Pseud-
emys scripta, Eridiphas sleveni, Sator angustus,
Crotalus catalinensis, and three subspecies of
Leptotyphlops humilis). Welsh (197 '6a, b, 1981)
indicated that distributional data for most of the
Sonoran Desert herpetofauna of Baja California
showed patterns of distribution and differentia-
tion on peninsular and insular land masses in-
dicative of long-term isolation and evolution
consistent with a tectonic vicariant origin. Ele-
ments of the herpetofauna whose distributions
are consistent with this scenario include some or
all members of the following genera within the
Baja California Region: Callisaurus, Crotaphy-
tus, Dipsosaurus, Petrosaurus, Phrynosoma,
Sauromalus, Sceloporus, Ufa, Urosaurus, Cole-
onyx, Phyllodactylus, Cnemidophorus, Lichan-
ura, Leptotyphlops, Chilomeniscus, Hypsiglena,
Lampropeltis, Phyllorhynchus, Sonora, Tantilla,
Trimorphodon, and Crotalus. Murphy (1983a)
developed a comprehensive evolutionary scenar-
io based on geomorphological and paleobotan-
ical evidence and tectonic theory to further elu-
cidate the possible vicariant events responsible
for many contemporary herpetofaunal distribu-
tion patterns. Kim et al. (1976), Wyles and Gor-
man (1978), Murphy and Papenfuss (1979), and
Murphy (1983a) presented electrophoretic data
that, assuming the validity of the "biochemical
evolutionary clock" (Sarich 1977), support the
thesis of a vicariant origin from Miocene tectonic
events to account for distributional patterns
among major components of the contemporary
herpetofauna of Baja California.
Paleogeographic and Paleobotanic History of
the Baja California Region
The following scenario is a shortened version
that follows closely the synthesis developed by
Murphy (19830) but with some modifications as
indicated below. Berggren and Van Couvering
(1974) and Van Couvering (1978) present a geo-
logic time scale that differs from the generally
accepted time scale by placing the Miocene-Plio-
cene boundary at 5.2 MYBP (million years be-
fore present), rather than at 1 1-13 MYBP, and
the Pliocene-Pleistocene boundary at 1.6-2.0
MYBP rather than at 1 MYBP. The Berggren
and Van Couvering time scale will be followed
in subsequent discussion.
Miocene and Earlier
Paleogeography
The "two-fault" hypothesis of Atwater (1970),
which indicates that two major migrating land
masses have been active along Pacific coastal
Mexico since at least the Miocene, is central to
the hypothesis of a tectonic origin for many pres-
ent-day herpetofaunal distributions of Baja
California. Gastil et al. ( 1 972) supported the "two-
fault" theory and indicated further that "Pen-
insular California" was probably separated from
coastal Mexico as several large fragments. These
two land masses are: (1) "Alta California," de-
fined here as consisting of present-day California
from about Santa Barbara to San Francisco and
west of the San Andreas fault (modified from
Gastil et al. 1972), and (2) "Peninsular Califor-
nia," those areas south of the Transverse Ranges
of southern California and west of the Gulf of
California (Gastil et al. 1972). "Alta California"
probably began its traverse 20-25 MYBP when
it was torn from "Peninsular California" (then
south of its present location, and part of coastal
Mexico) from about Bahia San Sebastian Viz-
caino northward. The Pacific islands of Baja Ca-
lifornia north of Vizcaino Bay and the Channel
Islands of California are probably fragments that
resulted from the traverse of "Alta California."
The dating for the separation of "Peninsular
California" from coastal Mexico varies consid-
erably from early Tertiary (60 MYBP; Gastil et
al. 1972) to early Pliocene (5 MYBP; Moore
1 973). Geological evidence (Atwater and Molnar
1973; Gastil and Jensky 1973) suggests that the
southern end of Baja California from about La
Paz south— the Cape fragment— originated from
coastal Mexico near the Jalisco-Nayarit border
in the mid-Miocene (12-14 MYBP). This and
other separate island and island fault blocks have
been simultaneously moving northwestward since
mid-Miocene through Pliocene times. By the close
of the Miocene, "Peninsular California" had mi-
grated 260 km northwest to the region of the
Islas Las Tres Marias (Gastil and Jensky 1973).
Thus it appears that from the mid-Miocene, be-
ginning about 14 MYBP, much of "Peninsular
California" existed as an island archipelago. This
archipelago included much of the present-day
WELSH: BAJA CALIFORNIA HERPETOFAUNA
53
Cape Region and associated deep-water islands
(Santa Catalina, Santa Cruz, San Diego, and Isla
Cerralvo— hereafter referred to collectively with
the proto-Cape Region as the Cape Islands), as
well as upland areas of the present-day Vizcaino
peninsula (possibly including Isla Cedros). Si-
multaneously, the northern Peninsular Ranges
were beginning to uplift and separate as a result
of subduction (Murphy 1983a). Murphy (19830,
fig. 6) indicated that the northern Peninsular
Ranges were not totally separated from western
Mexico at this time. A proto-Gulf of California
was beginning to form about 1 4 M YBP but may
not have become a permanent geographic feature
until around the Miocene-Pliocene boundary, 6-
5 MYBP (Gastil and Jensky 1973.)
Subduction along the fault system continued
to elevate the northern Peninsular Ranges, and
the central Peninsular Ranges began to emerge
as a result of volcanism (Durham and Allison
1 960; Karig and Jensky 1 972). The Sierra Madre
Occidental, the Sierra Madre Oriental, and the
Mexican Plateau were also being elevated as a
result of Miocene tectonic activities (King 1959).
"Thus the only geographically isolated portions
of the extant peninsula in the middle Miocene
were the islands near (and including) the current
Cape area, the Sierra Vizcaino, and several
southern peninsula-associated island localities"
(Murphy 19830). Murphy (19830, fig. 8, and text)
indicated that, by the late Miocene, the Cape
Islands were connected temporarily to more
northern portions of the peninsula, at least until
about the Miocene-Pliocene boundary, when this
connection was apparently inundated.
Paleobotany
Prior to the Eocene, neotropical floral elements
dominated western Mexico; from the Eocene on-
ward, however, declining temperatures and in-
creasing aridity progressively restricted the neo-
tropical-Tertiary geofloral elements southward
(Axelrod 1 979). Oak-pinyon woodlands and arid
tropic-scrub habitats (Madro-Tertiary vegeta-
tion associations) were developing but were re-
stricted to higher elevations of the young devel-
oping mountain regions in the interior. By the
close of the Oligocene the Madro-Tertiary geo-
flora predominated along the mountain chains
and uplands of central Mexico (Axelrod 1975).
The declining temperatures and increasing arid-
ity continued to facilitate the spread of Madro-
Tertiary geoflora in upland areas. Lowland areas
of Mexico west of the Sierra Madre Occidental
were still dominated by a neotropical flora (Ax-
elrod 1975, 1979).
Pliocene
Paleogeography
During the Pliocene, beginning about 5 MYBP,
"Peninsular California" began a more rapid
northwestward movement, from the present re-
gion of the Islas Las Tres Marias to its present
location (Larsen 1972; Atwater and Molner 1973;
Moore 1973). The resurgence of plate interac-
tions and the union of "Peninsular California"
with the Pacific Plate (Atwater and Molner 1 973)
resulted in further uplift of the Peninsular Ranges.
Volcanism continued in the southern Peninsular
Ranges, including the Sierra de la Giganta (Mina
1957). The proto-Gulf of California had com-
pletely formed by 5 MYBP (Karig and Jensky
1972; Gastil et al. 1975) and extended from the
region of the San Gorgonio Pass in southern Cal-
ifornia (Allen 1 957) to its opening into the Pacific
Ocean between the Islas Las Tres Marias and
mainland Mexico (Moore 1973). In the central
and southern peninsula, the Sierra Vizcaino and
possibly the Cape Islands remained isolated from
the northern Peninsular Ranges. During the Plio-
cene the Los Angeles Basin was submerged be-
neath the Pacific Ocean. The combination of the
flooding of the Los Angeles Basin and the for-
mation of the proto-Gulf of California greatly
reduced terrestrial access to the northern Pen-
insula (Durham and Allison 1960); it may have
in fact been totally isolated (Murphy 19830).
Murphy (19830, fig. 9) refers to this probable
restricted or eliminated terrestrial access to the
northern Peninsular Ranges as the San Gorgonio
Barrier.
Pliocene orogenic events brought the Sierra
Madre Occidental, the Sierra Madre Oriental,
and the Mexican Plateau to near their present
elevations (Eardley 1951; King 1959). This re-
sulted in at least a partial barrier to gene exchange
between populations of xerophilic herpetofaunas
in the proto-Chihuahuan and proto-Sonoran xe-
ric habitats on either side at the northern end of
the Sierra Madre Occidental; an area referred to
as the Cochise Filter Barrier (Morafka 1977).
54
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
Paleobotany
The newly formed, moderately high montane
regions initiated the establishment of easterly
rainshadows. Axelrod (1975) indicated that by
the late Pliocene, climates in the middle latitudes
(25-40°N) were essentially equivalent to those of
the present. The deserts as we know them today,
however, did not develop until the last intergla-
cial period (Axelrod 1979). The only paleobo-
tanical data from "Peninsular California" are
from the Mount Eden Beds at the north end of
the Peninsular Range in southern California.
These are characterized by a mixture of Arcto-
Tertiary and Madro-Tertiary geofloral assem-
blages. Eight habitats were represented: (1) des-
ert, (2) arid subtropical scrub, (3) coastal sage,
(4) grassland, (5) chaparral, (6) live oak and wal-
nut woodland, (7) digger pine woodland, and (8)
big-cone pine conifer forest (Axelrod 1937, 1950).
Murphy ( 1 9830) assumes that the desert and arid
subtropical scrub habitats were marginal and that
the area was predominantly mesic in character.
This interpretation is supported by vertebrate
fossils from the same area (Frick 1933). Terres-
trial and aquatic vertebrate fossil remains from
the Pliocene in southern Baja California (Miller
1977, 1980) indicate a warmer climate than the
present one with year-round fresh water, possibly
a subtropical savanna habitat.
Pleistocene
Paleogeography
During this epoch, the most significant process
of change— from the point of view of contem-
porary herpetofaunal distributions— appears to
have been the rise and fall in sea level resulting
from glacial fluctuations that alternately isolated
and then connected many islands in the Gulf of
California to larger peninsular land masses (those
islands separated from the peninsula by an ocean
depth of less than 1 30 m) (Auffenburg and Mil-
stead 1965). Tectonic events continued as well,
separating the Angel de la Guarda island block
(Angel de la Guarda, Partida Norte, San Lorenzo
Norte and Sur) from the peninsula, prior to 1
MYBP (Moore 1 973). Tectonic activity also con-
tinued the elevation of the Peninsular Ranges
(Gastil et al. 1975), completing the Coast Range
Corridor in California (Peabody and Savage
1958), and forcing the waters of the Gulf of Cal-
ifornia to recede to their present position. The
Sierra Vizcaino (Minch et al. 1 976) and the Cape
islands (Mina 1957) were unified with the pen-
insula.
Paleobotany
Geofloral data from Baja California and prox-
imate areas of the southwest from the Pleistocene
indicate a progressive drying and warming trend
with each of the four interglacial periods (Axel-
rod 1966, 1975, 1979; VanDevender and
Spaulding 1979). Axelrod (1966) characterized
the region at the north end of the Peninsular
Ranges during the interglacials as progressing
from subhumid to semiarid, to subdesert, and
finally to desert environments. Increased eleva-
tion of the northern Peninsular Ranges and other
mountain regions of the southwest, created east-
erly rainshadows that promoted increased aridity
and the subsequent expansion of desert habitats
during the Pleistocene (Axelrod 1979). Evidence
suggests that climates in southern Baja California
continued to be relatively mesic during this ep-
och compared with those of northern peninsular
areas. A fossil Pleistocene terrestrial vertebrate
fauna from southern Baja California (Miller 1 977)
indicates both wooded and grassland habitats oc-
curred in the area, indicative of denser vegeta-
tions than those presently in existence. In gen-
eral, excepting the evolving desert plant
communities, evidence indicates that mostly mi-
nor distributional perturbations in vegetation as-
sociations, rather than significant changes in
vegetational components, occurred during the
Pleistocene in southwestern North America (Ax-
elrod 1966, 1975, 1979). Such perturbations were
characterized by shifts in elevation of the vege-
tation communities of Baja California concom-
itant with glacial advances and retreats in more
northern areas.
The Paleobiogeography of the Herpetofauna:
Previous Scenarios
The broad paleobiogeographic scenarios of
Savage (1960) and Murphy (1983«) provide the
most current and comprehensive framework from
which to discuss the evolutionary trends of the
Baja California herpetofauna. Savage (1960) and
Murphy (1983#) each developed a set of perti-
nent data essential to an overall understanding
of the evolutionary mechanisms shaping her-
petofaunal relationships in Baja California. Mur-
WELSH: BAJA CALIFORNIA HERPETOFAUNA
55
phy's paleobiogeographic scenario, based on cur-
rent understanding of plate tectonics, clearly
outdates Savage's hypothesis as to the primary
mechanisms shaping the peninsular herpeto-
faunas. However, Savage developed, in broad
outline, a powerful model to describe evolving
North American herpetofaunal assemblages, the
progenitors of the contemporary peninsular and
insular herpetofaunas. He relied on the geolog-
ical history (King 1958; Durham and Allison
1 960), and more particularly the paleoecological
history, in his reconstruction of Tertiary and
Quaternary environments. This included the
geofloral history as described in the work of Cha-
ney, Axelrod, and MacGinitie (see Murphy
[1983a] or Savage [1960] for complete citations)
and the concept of interdigitating and evolving
geofloras. Savage's inherent assumption was that
herpetofaunas evolve in conjunction with broad
changes in phytogeographic relationships. He de-
nned four contemporary herpetofaunal areas in
Baja California (Savage 1960) (Fig. 2a plus the
Cape Region); these areas closely approximate
the four contemporary phytogeographic prov-
inces of the peninsula as denned by Shreve and
Wiggins (1964; see also this paper Fig. 2b). Sav-
age considered each of his herpetofaunal areas
to be dominated by assemblages of species with
a common evolutionary history. He noted that
a given Recent herpetofauna is composed of ". . .
a compliment of species originating from several
sources. A considerable portion of the species in
each unit, however, originated in or near the area
now occupied by the faunal assemblages. For
these reasons, present-day herpetofaunas are
mixtures of species of diverse historical origins,
although a particular fauna is dominated by de-
rivatives of one of the historical elements de-
nned" (Savage 1960:189).
Murphy (1983a:14) challenged the assump-
tion that particular herpetofaunas evolved in
conjunction with and parallel to particular geo-
floras. His argument is supported by two ex-
amples that illustrate the absence of complete
distributional correlation between groups of her-
petofaunal species and particular contemporary
vegetational associations of Madro-Tertiary
vegetation in Baja California. These arguments
are weak and based on a few rigid and very ex-
acting applications of what was intended only to
be a very general correlation. An exact correla-
tion between a particular vegetational associa-
tion (e.g., the Sonoran Desert flora) and a par-
ticular herpetofauna (e.g., the Colorado Desert
herpetofauna) was never assumed, intended, or
implied in my understanding of Savage's model.
Savage did not rule out the possible ameliorating
effects of physiography or of interspecific com-
petition, for example, in determining distribu-
tion. He used the evolving geofloras as evidence
of existing paleoclimatic conditions and in so
doing made the assumption only that contem-
porary and phylogenetic antecedent herpeto-
faunal assemblages occupy similar climatic re-
gimes. This seems a valid assumption considering
the taxa concerned are terrestrial ectotherms and
have physiological limitations that translate di-
rectly into ecogeographic restrictions (Bartho-
lomew 1958). This assumption also seems well
supported by the fossil record. Murphy's argu-
ments are further weakened by the fact that
throughout his own paleobiogeographic scena-
rio, he employs the same paleobotanical litera-
ture and makes the same general assumption that
geofloras are evidence of paleoclimatic condi-
tions that are indicative of certain historical her-
petofaunal assemblages in a given region during
a given time.
Generalized Tracks and Ecogeographic
Patterns
The strength of Murphy's (1983#) presenta-
tion lies in the excellent paleobiogeographic syn-
thesis he constructs to explain contemporary dis-
tribution patterns, and in the genetic evidence
he presents to support many of his conclusions.
However, I find his herpetofaunal assemblages
("tracks") to have serious shortcomings. Appar-
ently this is a result of his misapplication of the
concept of track (Croizat 1 964), and the concom-
itant lack of distinction between contemporary
distribution patterns and the diverse historical
assemblages from which they derived.
It is important to clarify some terms and con-
cepts critical to subsequent discussion. Murphy
( 1 983#) describes six herpetofaunal areas in pen-
insular Baja California: Californian, Colorado
Desert, Vizcaino Desert North, Coastal Vizcaino
Desert, Peninsular Area (with three subdivi-
sions), and the Cape Area. These herpetofaunal
areas are basically broad phytogeographic units
defined on the basis of major physiographic or
geomorphological transformations and concom-
itant differences in the general botanical com-
56
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
position (based on Nelson 1921; Savage 1960;
Shreve and Wiggins 1964; Bostic 1971; Loomis
et al. 1 974; and Murphy's own fieldwork). They
are comparable in concept to ecogeographic for-
mations (Fig. 14), but lack as fine a resolution
because they each encompass a larger geographic
area containing complex geomorphological for-
mations with numerous vegetational commu-
nities. They are probably best characterized as
ecogeographic provinces.
Murphy (1983a) compiles the herpetofaunal
composition of these ecogeographic provinces and
compares them quantitatively. From this anal-
ysis, he derives what he terms "tracks." He then
states "phylogenetic relationships of the herpe-
tofaunal elements are then superimposed on these
tracks of distribution to suggest hypotheses ex-
plaining the evolutionary relationships of the
herpetofauna of Baja California" (Murphy
1983#:3). The use of the term "track" in this
sense is misleading and confusing. The original
concept of "track" as developed and employed
by Croizat (1964; see also Croizat et al. 1974;
Craw 1983; Craw and Weston 1984) has two
components. The first concept is that of the in-
dividual track; this concept depicts the spatial
(geographical) expression of relatedness among
disparate forms. The geographic distributions of
all related forms of a given taxon (e.g., a species
complex, members of a genus or family) from
disparate geographic areas, when considered to-
gether and depicted spatially, comprise an in-
dividual track. The second concept, that of gen-
eralized or standard track (op. cit.) describes the
phenomenon of congruency in space of more than
one individual track. Both individual and gen-
eralized tracks are spatial expressions of evolu-
tionary time. A generalized track, because of the
spatiotemporal congruency it expresses, indi-
cates historical homology among the individual
tracks, and thus among the life forms represent-
ed. In other words, a generalized track is a spatial
and temporal representation of a group of phy-
logenetically unrelated organisms that shared
proximate space through time, or a common
evolutionary history.
Murphy states (198 3a:l 1): "I re-emphasize that
among the various species constituting each of
these five major generalized tracks of distribu-
tion, some have phylogenetic affinities with trop-
ical Mexico and others with territories north and
east of the peninsula." In my understanding of
Croizat (1964), for unrelated life forms to be
members of the same generalized track, they need
to demonstrate distributional similarities
throughout much or all of their respective ranges.
In other words, one would find consistent spatial
overlap among the phylogenetic lineages of each
of the various member taxa of the generalized
track. From the lack of this spatial overlap among
represented lineages, one would have to conclude
the presence of more than a single generalized
track. Murphy's (1983a) use of the term "track"
to describe herpetofaunal taxa that have present-
day geographic overlap within Baja California,
but lack such spatial congruency among their
respective phylogenetic lineages elsewhere, is in-
consistent with the concept of track.
The common denominator supporting Mur-
phy's faunal groupings is consistent spatial over-
lap among species, geographic or ecologic, or both,
only within parts of Baja California. Lacking the
requisite geographic congruency among repre-
sented lineages beyond the Baja California Re-
gion, one would have to conclude that these
groupings represent contemporary ecogeograph-
ic patterns, not historical patterns or tracks. They
appear to be analogous with the heretofore de-
scribed ecogeographic patterns of the herpeto-
fauna of the Sierra San Pedro Martir Region of
northern Baja California (Fig. 16; Table 5; and
text).
Subsequent use of the term track in this dis-
cussion should be understood in the sense used
by Croizat (1964). When indicating Murphy's
"tracks" quotation marks will be used to distin-
guish his from the generalized tracks that this
investigation indicates exist among the herpe-
tofauna of the Baja California Region. Savage's
( 1 960) historical assemblages appear to be con-
sistent with the temporal aspect of generalized
tracks; these parallels will be indicated below. I
consider his spatial designation of province con-
sistent with the spatial aspects of generalized
tracks, however, like Morafka (1977) I consider
Savage's (1960) provinces to equal superprov-
inces. I follow Morafka (1977) in my use of prov-
ince to indicate a geographical unit with uniform
dominant physiography, vegetation, and fauna
(see Dice 1943). My use of the term track, in the
spatial sense, is synonymous with province or
superprovince as indicated.
The terms track and generalized track, as used
in the temporal sense and assemblage or histor-
WELSH: BAJA CALIFORNIA HERPETOFAUNA
57
ical assemblage, are used interchangeably and
indicate historical biogeographic groupings; they
represent events in evolutionary time. The term
fauna is used to indicate a contemporary group-
ing of forms with similar ecogeographic restric-
tions; a province or track in the spatial sense may
include one or more faunas but only a single
fauna is considered the participant fauna (Mo-
rafka 1977). The terms fauna and track as used
in the spatial sense represent the present, in eco-
logical time; here considered to be subsets of the
historical groupings.
Extant Faunas and Generalized Tracks
Savage (1960) did not list all of those species
or species-groups he considered to have evolved
together as part of a particular historical assem-
blage. But, using his examples of species with
distributions representative of his historical as-
semblages, the list of extant vegetational asso-
ciations occupied by Recent derivatives of his
historical complexes and components, and the
contemporary species composition of Savage's
Herpetofaunal Areas, I assigned each contem-
porary species to a particular historical assem-
blage. The assignment of species and species-
groups, though based on Savage's (1960) work,
is my own interpretation and is thus subject to
errors for which I alone am responsible. Murphy
likewise did not list the member species for each
of his nine possible "tracks," leaving the reader
to decipher from his table 2 (Murphy 1983a)
those species that fit the description for each of
his "minor" and "major generalized tracks." He
was unclear about proper assignment of some
forms (e.g., on page 6, he assigns the Sceloporus
magister complex, minus the nominal species,
to his "Transpeninsular Mesophilic Track," but
on page 32 he discusses this complex as part of
his "Transpeninsular Xerophilic Track"). I have
attempted to interpret Murphy's work and assign
species in accordance with his descriptions; once
again I alone am responsible for any errors.
The following hypothesis of interacting and
overlapping generalized tracks comprising the
contemporary herpetofauna of the Baja Califor-
nia Region is a working one that best integrates
our current knowledge from pertinent fields.
However, it is a hypothesis to be tested and
changed wherever future investigations find it
lacking.
The Generalized Tracks of the
Baja California Herpetofauna
Savage's (1960) and Murphy's (1983a) Cali-
fornian Herpetofaunal Area (Fig. 2a), consisting
of the Pacific Slope of Baja California north of
latitude 30°N, is the most biologically complex
area of the Martir Region, and probably of the
entire peninsula. It is also one of the most poorly
investigated by previous workers. Savage indi-
cated that this area was dominated by an assem-
blage of species with more extensive ranges
northward in coastal and montane California that
shared an evolutionary history associated with
derivatives of the Madro-Tertiary geoflora (the
Californian Component of his Madrean Com-
plex). In addition, he indicated that some ele-
ments of his Sonoran Desert Component of the
Desert and Plains Complex occur within the Cal-
ifornian Area. This study indicates that elements
of four contemporary faunas (High Sierran Scarps,
Californian, Peninsular Range, and North Amer-
ican Ubiquitous Scrubland; Fig. 15, 16; Table
5), representing four different tracks (Pacific
Northwest, Madrean, Peninsular Range, and
North American Desert and Plains: Fig. 17, 18,
20, 21), overlap on the Pacific Slope of Baja Cal-
ifornia north of latitude 30°N.
The Californian Fauna together with the High
Sierran Scarps Fauna (Table 5, I and II) match
closely the species Savage indicated as endemic
to and dominating the Californian Herpetofau-
nal Area. Several species within this area, as not-
ed by Savage, are associated primarily with de-
rivatives of the Arcto-Tertiary geoflora. These
species, the High Sierran Scarps Fauna, are re-
stricted to Arcto-Tertiary floral derivatives in
northern Baja California and southern Califor-
nia. However, outside of Baja California, mem-
bers of this fauna demonstrate broader ecological
niches and occur in conjunction with more me-
sophilic derivatives of the Madro-Tertiary geo-
flora (see Species Accounts; Stebbins 1 966, 1 985).
The variability in niche breadth in different parts
of their range, and the presence of sympatric con-
geners within the Baja California Region, suggest
that the geographic limits of the High Sierran
Scarps Fauna result from competition. Overall
distributions of these species closely match those
of the Californian Fauna, suggesting that these
two faunas have a very similar and perhaps par-
allel evolutionary history in western North
58
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
America. The relatively low Coefficient of Dif-
ference and high Similarity Coefficient and Com-
munity Coefficient values (Table 4; Fig. 1 5) be-
tween areas occupied by these two faunas are
also indicative of a close relationship. With the
exception of three species of the Californian Fau-
na (Rana aurora, Bufo boreas, and Clemmys
marmorata) that are also associated primarily
with Arcto-Tertiary geofloral derivatives, the
species of the High Sierran Scarps Fauna and the
Californian Fauna are considered to be contem-
porary members of the Madrean Track (Fig. 18;
Table 5, 6; and below).
The Pacific Northwest Track
Rana aurora, Bufo boreas, and Clemmys mar-
morata are considered to be members of the Pa-
cific Northwest Track (Fig. 17; Table 5, 6). This
track is synonymous with Savage's Western
American Complex of the Old Northern Ele-
ment. Pacific Northwest Track members have
the most tenuous toehold on the predominantly
xeric Baja California peninsula. Several member
species occur in scattered populations in a few
canyons of extreme northwestern Baja (i.e.,
Aneides lugubris and Ensatina eschscholtzii).
Members of this track probably dispersed into
the Region following completion of the Coast
Range Corridor (Peabody and Savage 1958) dur-
ing the cooler and moister Pleistocene epoch.
Two of these species, Rana aurora and Bufo bo-
reas, appear to have derived from Asian forms,
and as noted by Savage (1960), apparently dis-
persed across the Bering Land Bridge and down
the west coast of North America during the
Pleistocene (Savage's Holarctic Element). It is
reasonable to propose that they constitute a sec-
ond, more recent generalized track, one that
overlaps the older Pacific Northwest Track.
The Madrean Track
The Madrean Track (Fig. 18; Table 6) is con-
sidered synonymous, in terms of process and pat-
tern, with Savage's (1960) Madrean Complex of
the Young Northern Element, but with a con-
siderably different species composition. Savage
distinguished a Californian Component and a
San Lucan Component in his Madrean Complex.
His San Lucan Component includes most of the
species here assigned to the Peninsular Range
Track (Table 6), as discussed below. Included»as
contemporary elements of a San Lucan compo-
nent of my Madrean Track (Table 6) are only
those forms present as sibling species in the Cal-
ifornian and Cape regions (and on some Pacific
coastal islands of Baja California) and absent
from most or all of the Central and Vizcaino
deserts (e.g., Eumeces, Hyla, Elgaria, and Col-
uber "lateralis complex").
Savage (1960) considered his Madrean Com-
plex the oldest assemblage in the Baja California
Region. Such dating probably refers to those taxa
that are here assigned to the Peninsular Range
Track fauna (Table 6). Distribution of popula-
tions on "old" or "deep-water" islands of the
Gulf of California (Soule and Sloan 1966; Mur-
phy 19836), and the concomitant relatively
greater levels of divergence within lineages (see
Species Accounts for taxonomic references), in-
dicate that the members of the Peninsular Range
Track are much older elements on the land masses
of Baja California than are Madrean Track forms.
Biochemical and electrophoretic analyses (Kim
et al. 1976; Wyles and Gorman 1978; Murphy
and Papenfuss 1979; Murphy 19830) indicate a
probable mid-Miocene divergence for several of
the lineages of the Peninsular Range Track.
The Madrean Track fauna, both within and
beyond the Baja California Region, appears to
be a previously more widespread, vicariated (see
Figure 18) assemblage of relatively mesophilic
species-groups. The level of differentiation among
disjunct populations of these forms within the
Baja California Region appears to match Mayr's
(1978) criteria for sibling species or semispecies.
Divergence and endemism within lineages of
Madrean Track elements beyond Baja California
(in California, Arizona, New Mexico, and Mex-
ico) are considerably greater than the level of
sibling species, with most lineages having pop-
ulations along the track that appear to meet the
criteria of superspecies (Mayr 1978). The rela-
tively greater differentiation within lineages of
the Madrean Track beyond, compared to within
Baja California, suggests a long-term historical
relationship among these species-groups prior to
their inhabiting Baja California, and an evolu-
tionary history distinct from, yet probably par-
alleling, the lineages of the Peninsular Range
Track.
Much of existing Baja California appears to
have been inaccessible to southward emigration
of terrestrial vertebrates (see Paleobiogeographic
WELSH: BAJA CALIFORNIA HERPETOFAUNA
History above), possibly from the mid-Miocene
and probably from about the Miocene-Pliocene
boundary until the Pleistocene. If so, after the
mid-Miocene or the Miocene-Pliocene bound-
ary and prior to the Pleistocene, Madrean Track
elements could have reached Baja California only
by being carried along on the migrating land
masses, i.e., by "transgulfian vicariance" (Mur-
phy 1975, 19830). The complete absence of
Madrean Track forms on the "old" or "deep-
water" Gulf islands (Soule and Sloan 1966; Mur-
phy 1983fl, b\ Murphy and Ottley 1983, 1984)
discounts this possibility, although vicariance and
subsequent extinction cannot be entirely ruled
out. An alternative possibility is that some of
these lineages were present on the proto-northern
Peninsular Range land masses during the Mio-
cene and vicariated at a later time (e.g., Miocene-
Pliocene boundary) than did lineages on the Cape
Islands. However, the relatively low levels of di-
vergence between the Baja California popula-
tions of these lineages, their absence from the
"old" or "deep-water" Gulf islands, and the in-
crease in numbers of Madrean Track species
present along a south to north transect in pen-
insular Baja California, strongly suggest a rela-
tively recent (i.e., Pleistocene) diffusion— gradual
movement across hospitable areas over long pe-
riods (Pielou 1979)— southward. The disjunct
distributions (vicariance) of some Madrean Track
elements along the peninsula suggests even more
recent (mid-Pleistocene to Recent) fragmenta-
tion of ranges in conjunction with climatic shifts
toward aridity (Axelrod 1967, 1975, 1979) or
from possible periodic inundation of low-lying
areas by changes in sea level, or both (Flint 1971).
At least two possible scenarios are suggested
by the distributions of the Madrean Track lin-
eages within Baja California. Firstly, the spatial
relationships between the congeners of the Cal-
ifornian and the High Sierran Scarps Faunas (Ta-
ble 5), which I propose to have resulted from
competition, are consistent with a hypothesis of
Pleistocene to Recent periodic diffusions by these
lineages southward into Baja California. Late
Pliocene or Pleistocene formation of a terrestrial
connection between Baja California and south-
ern California land masses and the concomitant
erogenic uplift would have provided southward
emigration routes as well as the upland retreats
consistent with the contemporary distributions
of these faunas. These contemporary distribu-
tions of mesophilic species, and those of more
xerophilic species associated with other tracks
discussed below, indicate that for at least part of
this time, barriers to gene flow existed. Elements
of the Madrean Track probably dispersed in suc-
cessive diffusions across a partial or fluctuating
barrier (San Gorgonio Barrier) during this time.
Secondly, prior to this time, from mid-Miocene
to Pleistocene, it is unclear what the nature of
the terrestrial connection at the northern end of
the extant peninsula was, or even if it existed.
Some elements I have assigned to the Madrean
Track have distributions that suggest they may
have been present on the northern peninsular
land masses at mid-Miocene, when increased
tectonic activity vicariated their ranges, isolating
them from ancestral stocks to the east (e.g., An-
niella geronimensis, Cnemidophorus labialis, and
Xantusia henshawi).
A hypothesis that incorporates both possible
scenarios is preferred. Species like Anniella ge-
ronimensis, Cnemidophorus labialis, and Xan-
tusia henshawi are not good candidates to sup-
port a Pleistocene diffusion hypothesis; their
endemism, narrow niches, and low vagility are
all indicative of species that are evolutionarily
relatively conservative, with their present ranges
probably contracting rather than expanding. They
appear to be the remains of an ancestral fauna,
a proto-Madrean Track fauna that may have vi-
cariated in the Miocene. If this were indeed the
case, these species could arguably be evidence of
another, earlier, generalized track, possibly the
result of the same geomorphological events that
shaped the Peninsular Range track further south
as discussed below. On the other hand, their sym-
patric congeners and the other species of the
Madrean Track, probably made their appearance
in the northern peninsula (and beyond), subse-
quent to Pleistocene climatic changes, sea level
shifts, and related phenomena that promoted an
ingress across the San Gorgonio Barrier. Under
the cool moist conditions of Pleistocene glacial
maxima, the more hydrophilic High Sierran
Scarps Fauna could have undergone diffusion
southward. The presence of Lampropeltis zonata
and Batrachoseps pacificus on Todos Santos Is-
land off the Pacific coast of Baja California sup-
ports the contention that cool, moist conditions
were more widespread in northern Baja Califor-
nia in the past. With glacial retreat and the long-
term drying and warming trend culminating in
60
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
the late Pleistocene to Recent (Axelrod 1967,
1975, 1979), conditions became increasingly fa-
vorable for the westward and southward dis-
persal of the warm-mesophilic Californian Fau-
na. Concomitant with these southward advances
of the Californian Fauna, the High Sierran Scarps
Fauna was increasingly restricted and relegated
to habitats at higher, cooler elevations. With the
advancing aridity, even the Californian Fauna
became restricted; ranges became fragmented and
populations in arid areas relegated to protect mi-
crohabitat, such as their present-day riparian
corridor distributions at lower elevations in the
Martir Region (Table 5; Fig. 16). This scenario
of diffusing and fragmenting faunas is consistent
with the current evidence of geologic and cli-
matic conditions during Pleistocene and Recent
times. It is also consistent with and supportive
of the thesis that sympatry within monophyletic
lineages is evidence of dispersal (Croizat et al.
1974; Platnick and Nelson 1978).
Four interrelated criteria were used to distin-
guish species of the Madrean Track from those
of the Peninsular Range Track (discussed below).
(1) Madrean Track species and species groups,
for the most part, exhibit extensive ranges be-
yond the Baja California Region. Members of
the Peninsular Range Track have their greatest,
and in many cases their total, distribution within
the Baja California Region. (2) In the Baja Ca-
lifornia Region, Madrean Track forms have
comparatively less differentiation within mono-
phyletic lineages compared to lineages of the
Peninsular Range Track, on both peninsular and
insular land masses. (3) Madrean Track forms
are present only on some of the land-bridged
islands (Soule and Sloan 1966; Murphy and Ott-
ley 1983, 1984) of the Gulf of California and on
some of the Pacific coastal islands of Baja Cali-
fornia and California. They are absent from all
of the "old" or "deep-water" islands (Soule and
Sloan 1966; Murphy and Ottley 1983, 1984) of
the Gulf, which are inhabited by numerous en-
demic forms of the Peninsular Range Track lin-
eages and some elements of the North American
Desert and Plains Track (discussed below). (4)
As a general trend, Madrean Track forms appear
to have their closest relationships within mono-
phyletic lineages with forms occurring in Cali-
fornia, Arizona, New Mexico, and the northwest
central uplands of Mexico (e.g., Rana boy Hi
[Zweifel 1955], Lampropeltis zonata [Tanner
1953], Batrachoseps [Wake 19666], Xantusia
[Bezy 1972; Bezy and Sites 1987]). Peninsular
Range Track elements are most closely related
to forms occurring in lowland areas from central
Mexico southward (e.g., Ctenosaura hemilopha
[Bailey 1928; Smith 1972], Eridiphas [Duellman
1958; Leviton and Tanner 1960; Cadle 1984;
Dowling and Jenner 1987], Phyllodactylus [Dix-
on 1964, 1969; Murphy and Papenfuss 1979],
Cnemidophorus hyperythrus and C. ceralbensis
[Lowe et al. 1970; Robinson 1973], Phrynosoma
coronatum [Presh 1969], Sator [Wyles and Gor-
man 1978], Nerodia [Conant 1969], and Bipes
[Kim et al. 1976; Papenfuss 1982]).
Peninsular Range Track
The third contemporary fauna with elements
in Savage's Californian Herpetofaunal Area of
the Martir Region is the Peninsular Range Fauna
(Table 5, IV). Some of these taxa occur also in
the Colorado Desert Herpetofaunal Area (Fig.
2a) and southward, throughout the length of the
peninsula. The Peninsular Range Fauna is part
of the larger Peninsular Range Track (Table 6;
Fig. 20) comprised primarily of forms endemic
to the Baja California Region and associated with
warm-temperate and subtropical floral ele-
ments. Many forms of the Peninsular Range
Track occur only south of the Martir Region in
the Central and Vizcaino deserts, in the Cape
Region, or on islands in the Gulf of California
(see Murphy 1983a; Murphy and Ottley 1983,
1984).
These peninsular and insular endemics are the
oldest lineages present in the Baja California Re-
gion, dating from mid-Miocene (Murphy 1983a).
Savage (1960) indicated the members of this fau-
na were part of two of his southward dispersing
waves: a Sonoran Component of his Desert and
Plains Complex, and a Madrean Complex. How-
ever, the present distributions of these species
and species-groups are more consistent with a
theory of origin in Baja California based on vi-
cariance, each distribution appearing— at least in
part— to result from the migration and fragmen-
tation of peninsular and insular land masses from
coastal Mexico. Murphy (1975) proposed such a
theory of vicariated origin to account for several
of these lineages found in the Cape Region and
on some associated islands; his "trans-gulfian
migration" hypothesis. Welsh (1976<2, b, 1981)
WELSH: BAJA CALIFORNIA HERPETOFAUNA
61
proposed a similar hypothesis to account for the
distributions of most of the Sonoran Desert-as-
sociated lineages of Baja California (see list at
beginning of Discussion), that are considered to
have a shared evolutionary history with the more
clearly vicariated subtropical Cape Island en-
demics of the Peninsular Range Track. Murphy
(1983#) included these Sonoran Desert-associ-
ated lineages and added several more mesophilic
lineages in his revised "trans-gulfian vicariance"
model.
Murphy (1983a) included the mesophilic lin-
eages Batrachoseps, Hyla (=H. regilla progeni-
tor), Eumeces, and Elgaria in his "trans-gulfian
vicariance" model. Exception is taken to the ad-
dition of these lineages, all of which are herein
considered to be part of the Madrean Track. In
adding these mesophilic genera, Murphy ( 1 9 8 3 a)
assumed that the Cape Region or Cape Islands
supported both Neotropical and Madro-Tertiary
geofloral elements during the mid-Miocene when
the area became separated from coastal Mexico:
"In the absence of fossil data, I assume that the
Cape Islands supported a mixture of Neotropi-
cal-Tertiary and Madro-Tertiary geofloras in the
respective tropical and temperate climatic re-
gimes; the temperate regime of the Cape resulting
from altitudinal effects" (Murphy 1983«:16).
Murphy was making the same kind of assump-
tion (inferring presence of historically associated
herpetofauna from presence of geoflora) that he
criticized Savage for making. Despite his appar-
ent contradiction, and "willingness" to join Sav-
age and myself in making this kind of assump-
tion, I think in this particular case, the assumption
is invalid as I shall discuss below.
Axelrod (1979:28) indicated the presence of
both Neotropical geofloral elements (dry tropic
forest and arid tropic scrub) and Madro-Tertiary
geofloral elements (oak-pinyon woodland) along
coastal western Mexico during the Miocene.
However, he stated that ". . . woodland presum-
ably reached southward down a line of volcanos,
linking the flora of the Cape Region with that of
Sonora-Sinaloa-Nayarit-Jalisco, which probably
lived above scattered dry-tropic forest and thorn
forest" (1 979:29). Whereas it seems possible that
volcanism could have provided the requisite up-
land habitat in some parts of coastal western
Mexico as Axelrod indicated, the Cape Islands
land masses are areas of granitic batholith, not
volcanic rock (Mina 1957; Gastil et al. 1975).
Furthermore, there is no evidence that this area
had been uplifted to sufficient height by mid-
Miocene to create the cooler climatic regime req-
uisite for the Madro-Tertiary geoflora.
Even in the event that such vegetational ele-
ments existed at the right time and in the right
place on the geomorphic progenitors of the Cape
Islands, these vegetation communities would
have been present on upland areas as islands of
habitat surrounded by great expanses of Neo-
tropical floral elements (Axelrod 1979:28). As
such, (a) these "islands" would have been far
removed from the northwest central highland and
central plateau areas of Mexico considered to be
the centers of differentiation and evolution of the
Madro-Tertiary geoflora (Axelrod 1975, 1979),
and (b) they would also most probably be highly
depauperate. Therefore, it seems highly unlikely
that these "islands" could contain and support
the mesophilic elements of the ancestral herpe-
tofauna that are thought to have evolved in as-
sociation with the then-developing Madro-Ter-
tiary vegetation.
The possibility that Batrachoseps, Hyla, Eu-
meces, and Elgaria vicariated on the Cape Is-
lands in upland habitats during mid-Miocene
cannot be entirely discounted. However, it seems
highly improbable and less likely than the more
parsimonious explanation that these genera are
elements of the Madrean Track (Fig. 18), whose
origins in the Cape Region are probably, as Sav-
age (1960) suggested, the result of late Pliocene
or Pleistocene diffusions down an existing pen-
insula.
The fossil herpetofaunal evidence that exists
from the Cape Region (Miller 1977) dates from
the late Pliocene and includes Boa, Geochelone,
and Crocodylus, etc., indicating decidedly trop-
ical conditions at least in the vicinity of this low-
land site.
The noninsular forms of the Peninsular Range
Track show three general patterns of distribution
in the Baja California Region: (1) restricted en-
demic lineages— those lineages found only in one
or more of the four southern ecogeographic prov-
inces (Murphy 1983a) of the peninsula or on
associated "deep-water" islands (Soule and Sloan
1966; Murphy 19836), or both; generally south
of latitude 28°N through the Cape Region (e.g.,
Bipes biporus, Chrysemys scripta, Ctenosaura
hemilopha, Sator complex, Nerodia vallida, and
Eridiphas slevini); (2) endemic lineages (note:
62
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
some of these lineages have member-species that
when considered separately, fit the description
in (1) above)— those lineages that range through-
out most or all of the Peninsular Ranges but not
north of the San Gorgonio Barrier (e.g., Cne-
midophorus hyperythrus, Crotaphytus insularis,
Petrosaurus complex, Phyllodactylus complex,
Sceloporus magister complex— those members
with 2n = 30 chromosomes, Sceloporus orcutti
complex, Elaphe rosaliae, Crotalus enyo, and C.
rubef); (3) radiated lineages— those lineages that
have expanded beyond the Peninsular Ranges
into habitats of the Mojave and Sonoran deserts
and north and east (Callisaurus draconoides, Co-
leonyx switaki, Dipsosaurus dorsalis, Phrynoso-
ma coronation, Sauromalus complex,
Chilomeniscus complex, Crotalus mitchellii, Li-
chanura trivirgata, and Phyllorhynchus decur-
tatus).
Murphy (1983a) outlined a plausible paleo-
biogeographic scenario to account for the distri-
butional permutations and divergence within
lineages among the Peninsular Range Track fau-
na. Lineages with member elements distributed
in accordance with each of the above described
geographic patterns consist of sibling species along
the peninsula, with one or two forms in the Cape
Region and another form or forms to the north
across the Isthmus of La Paz (e.g., Petrosaurus,
Urosaurus, Sceloporus orcutti complex, S. ma-
gister complex, Phyllodactylus complex, and
Chilomeniscus complex). Murphy attributed this
species-pairs phenomenon to a temporary or
fluctuating terrestrial corridor between the Cape
Islands (Region) and areas to the north, probably
during the late Miocene or early Pliocene (see
his "Transpeninsular Xerophilic Track"). It
would appear that more than a single vicariating
event, or intermittent gene flow between more
than two areas is involved here, based on the
presence of more than two distinct forms among
several of these lineages on the southern penin-
sula. Murphy (1983a, b) gives detailed account-
ings of the numerous unique patterns among in-
sular distributions of the lineages of the Peninsular
Range Track, and also for those insular-occur-
ring forms I have assigned to the North Amer-
ican Desert and Plains Track (see Table 6 and
below).
Assuming the validity of the hypothesis that
species of the Peninsular Range Track predate—
in Baja California— those species of the Madrean
Track, then future electrophoretic or immuno-
logical distance data should indicate a trend of
greater genetic distance between Peninsular Range
Track members and their southwest Mexican
siblings, than that which would be found between
the Madrean Track forms and their siblings be-
yond the Baja California Region (with the ex-
ception of some of the possible proto-Madrean
Track forms like Cnemidophorus labialis and
Xantusia henshawi). Good (1988) presented al-
lozyme data for the genus Elgaria that indicate
a closer relationship between E. paucicarinata of
the Cape Region and E. panamintina of south-
eastern California, than between E. paucicari-
nata and E. kingi of western Mexico. Following
Savage's (1960) scheme of historical herpeto-
faunas of North America, the lineages of the Pen-
insular Range Track would be called the Pen-
insular Range Complex of the Young Northern
Herpetofauna.
The Colorado Desert Track
Peninsular members of the Colorado Desert
Track (Fig. 19; Table 6) are found only in the
Colorado Desert Herpetofaunal Area (Fig. 2a).
The Colorado Desert was formed during the
Pleistocene; it resulted from the continued ele-
vation of the northern Peninsular Ranges, which
forced the Gulf of California to recede (Atwater
and Molner 1973), and the accumulation of al-
luvial deposits from the Colorado River filling
in the head of the Gulf (Norris 1958). Species of
the Colorado Desert Track have the most re-
stricted distributions of the xerophilic herpeto-
fauna of the Baja California Region; monophy-
letic lineages demonstrate the lowest levels of
interpopulational morphological differentiation.
The limited distributions and minimal intra-
lineal differentiation (see Adest 1977), together
with their endemism, suggest a Pleistocene to
Recent isolation and in situ evolution for these
forms. The presence of congeners of several of
these species among the Peninsular Range Track
fauna (Table 6), suggests that competition may
limit the spread of these species further south
into other desert habitats of Baja California.
This is an incipient track and hence does not
demonstrate the characteristics of spatial and
temporal fragmentation among forms associated
with older tracks (except see the distribution of
the genus Uma; Stebbins 1985). The Colorado
Desert Track becomes apparent when one ex-
WELSH: BAJA CALIFORNIA HERPETOFAUNA
63
amines the pattern of distributions of sibling
forms in the Chihuahuan, Sonoran, and Penin-
sular deserts. In fact, these four desert areas, each
considered here as a generalized track with its
own participant fauna, can also be viewed as
subsets of the more extensive and older North
American Desert and Plains Track (see below).
Morafka (1977:187) hypothesized a faunal
assemblage associated with the Colorado Desert
he called the California Gulf Arch Assemblage.
He considered this assemblage of 10 reptiles
(Dipsosaurus dorsalis, Sauromalus obesus, Cal-
lisaurus draconoides, Lichanura trivirgata, Chi-
lomeniscus cinctus, Chionactis occipitalis, Phyl-
lorhynchus decurtatus, Crotalus cerastes, Crotalus
mitchellii, and Gopherus agassizii) to be a sec-
ondary, more northerly centered ecological unit
of the Sonoran Herpetofauna, a fauna centered
to the south in Sonora and Sinaloa, Mexico. I
agree that the fauna of the Colorado Desert show
strong affinities with the Sonoran Herpetofauna,
but I do not consider it a subset thereof. Despite
the distributional similarities apparent among
these forms, track analysis indicates his is a syn-
thetic assemblage. It appears to me that his Cal-
ifornia Gulf Arch Assemblage is comprised of
Peninsular Range and Sonoran herpetofaunal
elements that have radiated into amenable hab-
itats not otherwise occupied in each of the other
areas, plus Colorado Desert Track forms that
have evolved in situ; in both cases as a result of
post-Pleistocene geomorphological and climatic
changes favoring expansion of extreme desert
habitats.
Adest (1987) analyzed allozyme data for Cal-
lisaurus draconoides from populations in the Cape
Region of Baja California throughout much of
the range including Nevada, and down the west
coast of Mexico to the vicinity of Mazatlan. His
results indicated low genetic diversity through-
out the range, with some minor differences in
populations near Mazatlan and on Isla Ceralvo
in the Gulf of California. His data were incon-
clusive to resolve the question of a Pleistocene
isolation in either a Sonoran or a Cape Region
refugium (Savage 1960) or a California Gulf Arch
refugium (Morafka 1977). However, he con-
cluded that his data did not support a model of
a Pleistocene Cape Region refuge for the species.
Unfortunately he did not include samples from
the population on Isla Angel de La Guarda, a
deep-water island isolated from central penin-
sular Baja California 1 MYBP (Murphy 1 983/7).
Significant differences between this population
and those tested by Adest (1987) would add cre-
dence to a hypothesis of a mid-latitude penin-
sular or insular refuge for Callisaurus draco-
noides.
Following Savage's historical groupings, the
species of the Colorado Desert Track might be
called the Colorado Desert Component of the
Desert and Plains Complex.
The riparian herpetofauna of the Colorado
River delta region (Bufo alvarius, Bufo cognatus,
Bufo woodhousei, Urosaurus ornatus, Thamno-
phis marcianus, and Trionyx spiniferus) were not
included in my ecogeographic analysis because
none of them range south into the Martir Region.
With the exceptions of Bufo alvarius and Uro-
saurus ornatus, which have distributions char-
acteristic of Sonoran Track forms, these species
appear to be post-Pleistocene to Recent radia-
tions from the east. Trionyx spiniferus and Bufo
woodhousei appear to be elements of a western
radiation of the Austriparian Herpetofauna (Sav-
age 1960) centered in the southeastern U.S.
Thamnophis marcianus appears to be a member
of the Chihuahuan Track that has crossed the
Cochise Filter Barrier (Morafka 1977), and Bufo
cognatus appears to be a member of the North
American Desert and Plains Track.
The North American Desert and Plains Track
The North American Desert and Plains Track
(Fig. 21) overlaps four less extensive herpeto-
faunal tracks: the Peninsular Range Track, the
Colorado Desert Track, the Sonoran Desert Track
(encompassing parts of northern Sinaloa, Sono-
ra, and southern Arizona), and the Chihuahuan
Desert Track (including much of Mexico's cen-
tral plateau— see Morafka 1977). Each of these
tracks has an endemic herpetofauna, however
these tracks also share many species and species-
groups. As here employed, the term North Amer-
ican Desert and Plains Track encompasses the
combined areas and fauna common to the Pen-
insular Range Deserts, Colorado Desert, Sono-
ran Desert, and Chihuahuan Desert generalized
tracks, and extends into the high deserts and plains
of the southwest, west, and central North Amer-
ica. This concept of an older (relative to the true
desert tracks listed above), more comprehensive
arid regions track is consistent with Morafka's
(1977) concept of a Miocene-Pliocene "Moha-
64
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
via" Biota, and the Desert and Plains super-
province of Savage (1960).
The fauna of the North American Desert and
Plains Track occurs within all of both Savage's
and Murphy's herpetofaunal areas of Baja Ca-
lifornia. Elements of this ubiquitous fauna ap-
pear to have populated the Baja California Re-
gion by two distinct means:
(1) The majority of these lineages (Table 6,
North American Desert and Plains Track, group
A) apparently accompanied the progenitors of
the lineages of the Peninsular Range Track, hav-
ing vicariated on prepeninsular and insular land
masses as they were separated from coastal Mex-
ico in the mid-Miocene. These species and
species-groups are widely distributed on the "old"
or "deep-water" islands of the Baja California
Region and they all demonstrate relatively high
degrees of morphological differentiation within
monophyletic lineages on the peninsula and as-
sociated islands (Soule and Sloan 1966; Savage
1967; Ballinger and Tinkle 1972; Case 1975;
Murphy \983a, b; Murphy and Ottley 1983,
1984).
(2) Other elements (Table 6, North American
Desert and Plains Track, group B) dispersed
southward along the peninsula after Pleistocene
to Recent climatic changes, which promoted cor-
ridors of xeric habitats at the northern end of the
Peninsular Ranges, allowing for diffusion south-
ward. Evidence for this diffusion hypothesis are
the relatively low levels of morphological differ-
entiation among all peninsular and insular pop-
ulations of each lineage, and their exclusive oc-
currence on land-bridged islands of the Gulf of
California and on some apparently recently iso-
lated land-bridged islands of the Pacific side of
the peninsula (Soule and Sloan 1966; Savage
1967; Ballinger and Tinkle 1972; Case 1975;
Murphy 19836; Murphy and Ottley 1983, 1984).
These recently invading ubiquitous xeric forms
could be considered as a distinct, overlapping
generalized track analogous to the situation with
the earlier Old Northern Element and the more
recent, overlapping Holarctic Element on the Pa-
cific Northwest Track. These more recent lin-
eages might be considered a Holxeric Element
of the Desert and Plains Track.
The hypothesis for the evolution of the North
American Desert and Plains Fauna, proposed by
Savage ( 1 960), incorporates three distinct centers
of evolution of xerophilic herpetofauna of North
America. Dispersal and genetic introgression be-
tween these areas since the late Pleistocene and
possibly during glacial minima has resulted in
the present, relatively continuous distribution of
much of this fauna. The endemic xerophilic her-
petofaunas of the Peninsular Range Track, the
Sonoran Desert Track, and the Chihuahuan Des-
ert Track appear to be comprised of those forms
with a competitive disadvantage in xeric habi-
tats, such as restricted niches, or lower vagility,
or both.
SUMMARY
Theories of the historical biogeography of the
the herpetofauna of the Baja California Region
have developed in conjunction with the growing
geofloral literature and the development of a new
paradigm in geology, that of plate tectonics. Past
syntheses by Schmidt ( 1 922), Savage ( 1 960), and
Murphy (1983a) reflect this evolution of ideas.
Five primary historical patterns or generalized
tracks are proposed to describe the interaction
of historical faunal assemblages that produced
contemporary distributional patterns: the Pacific
Northwest Track, the Madrean Track, the Pen-
insular Range Track, the Colorado Desert Track,
and the North American Desert and Plains Track.
Vicariance of peninsular land masses from west-
ern Mexico as a result of Miocene-Pliocene tec-
tonic activity is the most significant process
shaping extant patterns. In situ evolution during
the Pliocene, Pleistocene, and Recent epochs, as
well as radiations northward from these southern
vicariated areas, and biotic dispersal— both dif-
fusion and secular migration (Pielou 1 979)— from
proximate regions to the north and east played
major secondary roles influencing present dis-
tributional patterns.
RESUMEN
Conduje un analisis ecogeografico de la her-
petofauna de la region de San Pedro Martir en
Baja California Norte, Mexico. Se analizaron mas
de 3,000 archives municipales conseguidos de la
literatura, los museos, y investigaciones en el
campo, por medio de una matriz de formaciones
ecogeograficas basadas en los climas regionales,
la fisiografia, y la vegetation existente. Metodos
numericos y intuitivos de analisis biografico in-
dican que las 65 especies que se encuentran ocu-
rren en siete distintos modelos de distribution.
WELSH: BAJA CALIFORNIA HERPETOFAUNA
65
Se analizan estos modelos contemporaneos en
un contexto geografico mas amplio para encon-
trar indicaciones de sus origenes historicos.
Se han desarrollado las teorias de la biogeo-
grafia historica de la herpetofauna de la region
de Baja California en junto con la creciente li-
teratura geofloral y el desarrollo de un nuevo
paradigmo de la geologia— el de "plate tecto-
nics." Sintesis anteriores por Schmidt (1922),
Savage (1960), y Murphy (1983a) reflejan esta
evolucion de ideas. Se propone cinco patrones
historicos, o "generalized tracks" (Croizat 1964)
para describer la interaction de "faunal assem-
blages" historicos que producen los modelos de
distribucion comtemporaneos: el "Pacific
Northwest Track," el "Madrean Track," el "Pe-
ninsular Range Track," el "Colorado Desert
Track," y el "North American Desert and Plains
Track." La "vicariance" (Croizat et al. 1974) de
tierras peninsulares del oeste de Mexico como
resultado de actividad tectonica es al proceso
mas significante en el porceso de la formation
de los modelos que existen hoy dia. Evolucion
en situ durante las epocas Pliocena, Pleistocena,
y Reciente, ademas de radiaciones hacia el norte
de estas areas "vicariated" del sur, y "biotic dis-
persal" (diffusion and secular migration— Pielou
1979) de las regiones vecinas del norte y el este
tomaron importantes papeles secundarios que
influyeron los modelos de distribucion que esis-
ten ahora.
LITERATURE CITED
ADEST, G. A. 1977. Genetic relationships in the genus Uma
(Iguanidae). Copeia 1977(l):47-52.
. 1987. Genetic differentiation among populations of
the zebratail lizard, Callisaurus draconoides (Sauria: Igua-
nidae). Copeia 1987(4):854-859.
ALLEN, C. R. 1957. San Andreas Fault zone in San Gorgonio
Pass, southern California. Geol. Soc. Am. Bull. 68:315-350.
ALLEN, C. R., L. T. SILVER, AND F. G. STEHLI. 1956. Agua
Blanca Fault— a major transverse structure of northern Baja
California, Mexico. Geol. Soc. Am. Bull. 67(Dec):1664.
ALLISON, E. C. 1964. Geology of areas bordering the Gulf of
California. Pp. 3-29 in Marine geology of the Gulf of Cal-
ifornia. T. H. Van Andel and G. G. Shor, Jr., eds. American
Association of Petroleum Geologists, Memoirs 3.
ALVAREZ, T. AND P. HUERTA. 1974. Nuevo registro de Cro-
talus atrox para la peninsula de Baja California. Rev. Soc.
Mexicana Hist. Nat. 35:1 13-1 15.
ASCHMAN, H. 1959. The Central Desert of Baja California:
demography and ecology. Ibero-Amer. 42. University of
California Press, Berkeley. 282 pp.
ATWATER, T. 1970. Implications of plate tectonics for the
Cenozoic tectonic evolution of western North America. Geol.
Soc. Am. Bull. 81:3513-3535.
ATWATER, T. AND P. MOLNAR. 1973. Relative motion of the
Pacific and North American plates deduced from sea-floor
spreading in the Atlantic, Indian and South Pacific oceans.
Stanford Univ. Publ. Geol. Sci. 11:136-148.
AUFFENBURG, W. AND M. W. MiLSTEAD. 1965. Reptiles in
the Quaternary of North America. Pp. 557-568 in The Qua-
ternary of the United States. H. E. Wright, Jr. and D. G.
Frey. eds. Princeton University Press, Princeton, New Jer-
sey.
AXELROD, D. I. 1937. A Pliocene flora from the Mount Eden
Beds, southern California. Carnegie Inst. Washington Publ.
476:125-153.
. 1950. Further studies of the Mount Eden Flora,
southern California. Carnegie Inst. Washington Publ. 590:
73-117.
. 1966. The Pleistocene Soboba flora of southern Cal-
ifornia. Univ. Calif. Publ. Geol. Sci. 60:1-109.
. 1967. Geologic history of the California insular flora.
Pp. 267-3 1 5 in Proceedings of the symposium on the biology
of the California Islands. R. N. Philbrick, ed. Santa Barbara
Botanic Gardens, Santa Barbara, California.
. 1975. Evolution and biogeography of the Madrean-
Tethyan sclerophyll vegetation. Ann. Mo. Bot. Card. 62(2):
280-334.
. 1979. Age and origin of the Sonoran Desert vege-
tation. Occas. Pap. Calif. Acad. Sci. 132:1-74.
BAILEY, J. W. 1928. A revision of the lizards of the genus
Ctenosaura. Proc. U.S. Natl. Mus. 73(2733): 1-55.
BAIRD, S. F. 1854. Description of new genera and species of
North American frogs. Proc. Acad. Nat. Sci. Phila. 7(2):59-
62.
. 1858. Description of new genera and species of North
American lizards in the museum of the Smithsonian Insti-
tution. Proc. Acad. Nat. Sci. Phila. 7:253-256.
— . 1859. Reptiles of the boundary. In Vol. 2, Pt. 2,
United States and Mexican Boundary Survey, U.S. 34th
Congress, 1st Session, Exec. Doc. 108:1-35.
BAIRD, S. F. AND C. GIRARD. 1852a. Descriptions of new
species of reptiles collected by the U.S. Exploring Expedition
under the command of Capt. Charles Wilkes, U.S.N. Proc.
Acad. Nat. Sci. Phila. 6:174-177.
. 18526. Reptiles in Stanbury, II. Exploration and sur-
vey of the Valley of the Great Salt Lake of Utah. App. C.:
336-353.
. 1853. Catalog of North American reptiles and am-
phibians in the museum of the Smithsonian Institution. Part
I. Serpentes. Smithson. Misc. Collect. 2(5): 1-1 72.
BALLINGER, R. E. AND D. W. TINKLE. 1972. Systematics and
evolution of the genus Uta (Sauria: Iguanidae). Misc. Publ.
Mus. Zool. Univ. Mich. (145):277-279.
BANTA, B. AND A. LEVITON. 1963. Remarks upon Arizona
elcgans pacata. Herpetologica 18(4):277-279.
BANTA, B. AND W. W. TANNER. 1968. The systematics of
Crotaphytus wizlizeni, the leopard lizards (Sauria, Iguani-
dae). Part II. A review of the status of the Baja California
peninsular populations and a description of a new subspecies
from Cedros Island. Great Basin Nat. 28(4): 182-1 94.
BARTHOLOMEW, G. A. 1958. The role of physiology in the
distribution of terrestrial vertebrates. Pp. 81-95 in Zooge-
ography. C. L. Hubbs, ed. Am. Assoc. Advance. Sci., Wash-
ington, D.C.
66
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
BERGGREN, W. A. AND J. A. VAN COUVERING. 1 974. The late
Neogene. Palaeogeogr. Palaeoclimatol. Palaeoecol. 16:1-216.
BEZY, R. L. 1972. Karyotypic variation and evolution of the
lizard family Xantusiidae. Los Angeles Co. Mus. Contr. Sci.
227:1-29.
BEZY, R. L., G. C. GORMAN. Y. L. KIM, AND J. W. WRIGHT.
1977. Chromosomal and genetic divergence in the fossorial
lizards of the family Anniellidae. Syst. Zool. 26(1):57-71.
BEZY, R. L. AND J. W. SITES, JR. 1987. A preliminary study
of allozyme evolution in the lizard family Xantusiidae. Her-
petologica 43(3):280-292.
BLAINVILLE, M. H. D. DE 1835. Description de quelques
especes de reptiles de la California precedee de 1'analyse d'un
systeme general d'erpetologie et d'amphibiologie. Nouv. Ann.
Mus. d'Hist. Nat. 4:232-296.
BLANCHARD, F. N. 1923. Comments on ring-necked snakes
(genus Diadophis), with diagnosis of new forms. Occ. Pap.
Mus. Zool. Univ. Mich. 142:1-9.
. 1924. A new snake of the genus Arizona. Occ. Pap.
Mus. Zool. Univ. Mich. 150:1-5.
BOGERT, C. M. 1939. A study of the genus Salvadora, the
patchnose snakes. Publ. Univ. Calif. Los Angeles Biol. Sci.
1:177-236.
. 1945. Two additional races of the patch-nosed snake,
Salvadora hexalepis. Am. Mus. Nat. Hist. Novit. 1285:1-
14.
BOSTIC, D. L. 1968. Thermal relations, distributions, and
habitat of Cnemidophorus labialis (Sauria: Teiidae). Trans.
San Diego Soc. Nat. Hist. 15(3):21-30.
. 1971. Herpetofauna of the Pacific coast of north
central Baja California, Mexico, with a description of a new
subspecies of Phyllodactylus xanti. Trans. San Diego Soc.
Nat. Hist. 16(10):237-264.
BRAME, A. H., JR. AND K. F. MURRAY. 1968. Three new
slender salamanders (Batrachoseps} with a discussion of re-
lationships and speciation within the genus. Bull. Los An-
geles Co. Mus. Nat. Hist. 4:1-33.
BURT, C. E. 1931. A study of the teiid lizards of the genus
Cnemidophorus with special reference to their phylogenetic
relationships. Bull. U.S. Natl. Mus. 154:1-286.
BURY, R. B. 1970. Clemmys marmorata. Cat. Am. Amphib.
Rept. 100.1-100.3.
. 1983. Geographic distribution. Annie/la nigra ar-
gentea. Herpetol. Rev. 14(3):83-84.
CADLE, J. E. 1984. Molecular systematics of neotropical xe-
nodontine snakes: II. Central American xenodontines. Her-
petologica40(l):21-30.
CAMP, C. L. 1915. Batrachoseps major and Bufo cognatus
californicus, new Amphibia from southern California. Univ.
Calif. Publ. Zool. 12(12):327-334.
CASE, T. J. 1975. Species numbers, density compensation,
and colonizing ability of lizards on islands in the Gulf of
California. Ecology 56(1):3-18.
CHAMBERS, K. L. 1955. A collection of plants from the eastern
flank of the Sierra San Pedro Martir, Baja California. Dudley
Herb. Stanford Univ., Contr. 4(8):323-330.
CHEETHAM, A. H. AND J. E. HAZEL. 1969. Binary (presence-
absence) similarity coefficients. J. Paleontol. 43: 1 1 30-1 1 36.
COLE, C. J. AND L. M. HARDY. 1981. Systematics of North
American colubrid snakes related to Tantilla planiceps
(Blainville). Bull. Am. Mus. Nat. Hist. 17(3): 199-284.
COLLINS, J. T., J. E. HUHEEY, J. L. KNIGHT, AND H. M. SMITH.
1978. Standard common and scientific names for North
American amphibians and reptiles. S.S.A.R. Misc. Publ.
Herpetol. Circ. No. 7, 36 pp.
CONANT, R. 1969. A review of the water snakes of the genus
Natrix in Mexico. Bull. Am. Mus. Nat. Hist. 142(1): 1-140.
. 1975. A field guide to the reptiles and amphibians
of eastern and central North America. Houghton-Mifflin,
Cambridge, Massachusetts. 429 pp.
COPE, E. D. 1861. Contributions to the ophiology of Lower
California, Mexico, and Central America. Proc. Acad. Nat.
Sci. Phila. 13:292-306.
. 1 866. Fourth contribution to the herpetology of trop-
ical America. Proc. Acad. Nat. Sci. Phila. 18:123-132.
. 1867. On Reptilia and Batrachia of the Sonoran
Province of the Nearctic Region. Proc. Acad. Nat. Sci. Phila.
18:300-314.
. 1868. Observations on some specimens of Verte-
brata presented by Wm. M. Gabb, of San Francisco, which
were procured by him in western Nevada and the northern
part of Lower California. Proc. Acad. Nat. Sci. Phila. 20:2.
. 1875. Checklist of North American Batrachia and
Reptilia. Bull. U.S. Natl. Mus. 1:1-104.
. 1892a. A synopsis of the species of the teiid genus
Cnemidophorus. Trans. Am. Philos. Soc. 17(l):27-52.
. 1892fc. A critical review of the characters and vari-
ations of the snakes of North America. Proc. U.S. Natl.
Mus. 14(882):589-694.
. 1895. On some new North American snakes. Am.
Nat. 29:676-680.
. 1 896. On two species of lizards from southern Cal-
ifornia. Am. Nat. 30:676-680.
CRAW, R. C. 1983. Panbiogeography and vicariance cladis-
tics: are they truly different? Syst. Zool. 32(4):43 1-438.
CRAW, R. C. AND P. WESTON. 1984. Panbiogeography: a
progressive research program? Syst. Zool. 33(1): 1-1 3.
CROIZAT, L. 1964. Space, time, form: the biological synthesis.
Published privately by author. Caracas, Venezuela. 88 1 pp.
CROIZAT, L., G. NELSON, AND D. E. ROSEN. 1974. Centers of
origin and related concepts. Syst. Zool. 23(3):265.
CROSS, J. K. 1970. The shovel-nosed snake (Chionactis oc-
cipitalis) in Baja California. Herpetologica 26(1): 134- 140.
CUNNINGHAM, J. D. 1962. Observations on the natural his-
tory of the California Toad, Bufo californicus Camp. Her-
petologica 17(4):255-260.
. 1964. Observations on the ecology of the canyon
treefrog, Hyla californiae. Herpetologica 20(1):55-61.
DICE, L. R. 1943. The biotic provinces of North America.
University of Michigan, Ann Arbor. 79 pp.
DIXON, J. R. 1964. The systematics and distribution of the
lizard genus Phyllodactylus in North and Central America.
New Mexico State Univ. Sci. Bull. 64(1): 1-39.
. 1969. Phyllodactylus xanti. Cat. Am. Amphib. Rept.
79.1-79.2.
DOWLING, H. G. AND J. V. JENNER. 1987. Taxonomy of
American xenodontine snakes. II. The status and relation-
ships of Pseudoleptodeira. Herpetologica 43(2): 190-200.
DUELLMAN, W. E. 1 958. A monographic study of the colubrid
snake genus Leptodeira. Bull. Am. Mus. Nat. Hist. 1 14(1):
1-52.
. 1970. The hylid frogs of Middle America. Monogr.
Mus. Nat. Hist., Univ. Kansas 1:1-753.
DURHAM, J. W. AND E. C. ALLISON. 1960. The geologic his-
tory of Baja California and its marine fauna. Syst. Zool. 9:
47-9 1 .
WELSH: BAJA CALIFORNIA HERPETOFAUNA
67
EARDLEY, A. J. 1951. Structural geology of North America.
Harper, New York. 624 pp.
FLINT, R. F. 1971. Glacial and Quaternary geology. John
Wiley and Sons, Inc., New York. 892 pp.
FRICK, C. 1933. New remains of trilophodont-tetrabelodont
mastodons. Bull. Am. Mus. Nat. Hist. 59:505-652.
FRITTS, T. H., H. L. SMELL, AND R. L. MARTIN. 1982. An-
arbylus switaki Murphy: an addition to the herpetofauna of
the United States with comments on relationships with Co-
leonyx. J. Herpetol. 16(l):39-52.
FROST, D. R. 1 983. Sonora semiannulata. Cat. Am. Amphib.
Rept. 333.1-333.4.
GASTIL, R. AND W. JENSK.Y. 1973. Evidence for strike-slip
displacement beneath the trans-Mexican volcanic belt. Stan-
ford Univ. Publ. Geol. Sci. 1 1:171-180.
GASTIL, R., R. P. PHILLIPS, AND E. C. ALLISON. 1975. Re-
connaissance geology of the state of Baja California. Geol.
Soc. Am. Mem. 14. 170 pp.
GASTIL, R., R. P. PHILLIPS, AND R. RODRIGUEZ-TORRES. 1972.
The reconstruction of Mesozoic California. Twenty-fourth
International Geological Congress 1972(3):2 17-229.
GATES, G. O. 1968. Geographical distribution and character
analysis of the iguanid lizard Sauromalus obesus in Baja
California, Mexico. Herpetologica 24(4):285-288.
GEHLBACH, F. R. 1971. Lyre snakes of the Trimorphodon
biscutatus complex: a taxonomic resume. Herpetologica 27(2):
200-211.
GOOD, D. A. 1988. Allozyme variation and phylogenetic
relationships among the species of Elgaria (Squamata, An-
guidae). Herpetologica 44(2): 154- 162.
GORMAN, G. C. 1965. The distribution of Lichanura trivir-
gata and the status of the species. Herpetologica 21(4):283-
287.
GRAY, J. E. 1852. Descriptions of several new genera of
reptiles, principally from the collection of H. M. S. Herald.
Ann. Mag. Nat. Hist. 2(10):437^40.
HALL, W. P. 1973. Comparative population cytogenetics,
speciation, and evolution in the iguanid lizard genus Sce-
loporus. Ph.D. Dissertation. Harvard University, Cam-
bridge, Massachusetts.
HALLOWELL, E. 1852. Descriptions of new species of reptiles
inhabiting North America. Proc. Acad. Nat. Sci. Phila. 6:
177-182.
. 1853. On some new reptiles from California. Proc.
Acad. Nat. Sci. Phila. 6(7):236-238.
. 1854. Descriptions of new reptiles from California.
Proc. Acad. Nat. Sci. Phila. 7:91-97.
HASTINGS, J. R. AND R. R. HUMPHREY, EDS. 1969. Clima-
tological data and statistics for Baja California. Technical
reports on the meteorology and climatology of arid regions,
no. 18. Univ. Arizona Atmos. Phys., Tucson.
HASTINGS, J. R. AND R. M. TURNER. 1965. Seasonal precip-
itation regimes in Baja California, Mexico. Geografiska An-
naler 47 Ser. A:204-223.
HENDERSON, D. A. 1960. Geography of the Sierra Juarez and
San Pedro Martir, Baja, California, Mexico. Calif. Geogr. 1:
21-28.
HEYER, W. R. 1967. A herpetofaunal study of an ecological
transect through the Cordillera de Tilaran, Costa Rica. Co-
peia 1967(2):259-271.
HUNSAKER, D. 1965. The ratsnake Elaphe rosaliae in north-
ern Baja California. Herpetologica 21:71-72.
JACCARD, P. 1902. Lois de distribution florale dans la zone
alpine. Bull. Soc. Vaudoise Sci. Nat. 38:69-130.
JONES, K. B. 1 985. Eumecesgilberti. Cat. Am. Amphib. Rept.
372.1-372.3.
JONES, L. C. 1981. Geographic distribution: Crotalus viridis
helleri. Herpetol. Rev. 12(2):65.
KARIG, D. E. AND W. JENSKY. 1972. The proto Gulf of Cal-
ifornia. Earth Planet. Sci. Lett. 17:169-174.
KENNICOTT, R. 1860. Descriptions of new species of North
American serpents in the museum of the Smithsonian In-
stitution, Washington. Proc. Acad. Nat. Sci. Phila. 12:328-
338.
KIM. Y. L, G. C. GORMAN, T. J. PAPENFUSS, AND A. K.
ROYCHOUDHURY. 1976. Genetic relationships and genetic
variation in the amphisbaenian genus Bipes. Copeia 1976(1):
120-124.
KING, P. B. 1958. Evolution of modern surface features in
western North America. Am. Assoc. Advance. Sci. Publ. 5 1 :
3-60.
. 1959. The evolution of North America. Princeton
University Press, Princeton, New Jersey. 189 pp.
KLAUBER, L. M. 1924. Notes on the distribution of snakes
in San Diego County, California. Bull. Zool. Soc. San Diego
1:1-23.
. 1931a. A new subspecies of the California boa with
notes on the genus Lichanura. Trans. San Diego Soc. Nat.
Hist. 6(20):305-318.
. 1931/7. Notes on the worm snakes of the southwest
with descriptions of two new subspecies. Trans. San Diego
Soc. Nat. Hist. 6(23):333-352.
. 1932. Notes on the silvery footless lizard, A nniella
pulchra. Copeia 1932(l):4-6.
. 1933. Notes on Lichanura. Copeia 1933(4):2 14-2 15.
. 1935. Phvllorhvnchus, the leaf-nosed snake. Bull.
Zool. Soc. San Diego 12:1-31.
. 1936. Crotalus mitchellii, the speckled rattlesnake.
Trans. San Diego Soc. Nat. Hist. 8(1 9): 149-1 84.
. 1940. The worm snakes of the genus Leptotyphlops
in the United States and northern Mexico. Trans. San Diego
Soc. Nat. Hist. 9(18):87-162.
. 1941. The long-nosed snakes of the genus Rhino-
cheilus. Trans. San Diego Soc. Nat. Hist. 9(29):289-332.
-. 1 943. A new snake of the genus Sonora from Lower
California, Mexico. Trans. San Diego Soc. Nat. Hist. 10(4):
69-70.
. 1944. The sidewinder, Crotalus cerastes, with the
description of a new subspecies. Trans. San Diego Soc. Nat.
Hist. 10(8):91-126.
1945. The geckos of the genus Co/eonyx with de-
scriptions of new subspecies. Trans. San Diego Soc. Nat.
Hist. 10(11): 133-2 16.
-. 1946a. The gopher snakes of Baja California, with
descriptions of new subspecies ofPituophis catenifer. Trans.
San Diego Soc. Nat. Hist. 1 1(1): 1^10.
-. 19466. The glossy snakes, Arizona, with descriptions
of new subspecies. Trans. San Diego Soc. Nat. Hist. 10(17):
311-398.
. 1949. Some new and revised subspecies of rattle-
snakes. Trans. San Diego Soc. Nat. Hist. 1 1(6):61-1 16.
. 1972. Rattlesnakes, their habits, life histories, and
influences on mankind. University of California Press,
Berkeley. 1,533 pp.
LARSON, R. L. 1972. Bathymetry, magnetic anomalies, and
68
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
plate tectonic history of the mouth of the Gulf of California.
Geol. Soc. Am. Bull. 83:3345-3360.
LEE, J.C. 1975. Theautecologyof-Ya/7Ms/a/!m$/zaw/(Sauria:
Xantusiidae). Trans. San Diego Soc. Nat. Hist. 17(19):259-
277.
LEVITON, A. E. AND W. W. TANNER. 1960. The generic al-
location of Hypsiglena slevini Tanner (Serpentes: Colubri-
dae). Occ. Pap. Calif. Acad. Sci. 27:1-7.
LINSDALE, J. M. 1932. Amphibians and reptiles from Lower
California. Univ. Calif. Publ. Zool. 38(6):345-386.
LOCKINGTON, W. N. 1880. List of California reptiles and
batrachia collected by Mr. Dunn and Mr. W. J. Fisher in
1876. Am. Nat. 14:290-296.
LOOMIS, R. B. 1965. The yellow-legged frog, Rana boylei,
from the Sierra San Pedro Martir, Baja California Norte,
Mexico. Herpetologica 21(1):78-80.
LOOMIS, R. B., S. G. BENNETT, S. R. SANBORN, C. H. BARBOUR,
AND H. WEINER. 1974. A handlist of the herpetofauna of
Baja California, Mexico and adjacent islands. California State
University, Long Beach. 1 1 pp.
LOWE, C. H. 1 964. The vertebrates of Arizona. Annotated
check lists. University of Arizona Press, Tucson.
LOWE, C. H. AND K. S. NORRIS. 1954. Analysis of the her-
petofauna of Baja California, Mexico. Trans. San Diego Soc.
Nat. Hist. 12(4):47-64.
LOWE, C. H., J. W. WRIGHT, C. J. COLE, AND R. L. BEZY.
1970. Chromosomes and evolution of the species group of
Cnemidophorus (Reptilia: Teiidae). Syst. Zool. 19(2): 128-
141.
MACARTHUR, R. H. 1972. Geographical ecology: patterns in
the distribution of species. Harper and Row, New York. 269
pp.
MAHRDT, C. R. 1973. Geographic distribution: Crotaphytus
wizlizenii copei. HISS News- J. 1:98.
MAYHEW, W. W. 1963. Biology of the granite spiny lizard,
Sceloporus orcutti. Am. Midi. Nat. 69(2):3 10-327.
MAYR, E. 1978. Origin and history of some terms in system-
atic and evolutionary biology. Syst. Zool. 27(1):88-104.
MEDICA, P. A. 1975. Rhinocheilus. Cat. Am. Amphib. Rept.
175.1-175.3.
MEEK, S. E. 1905. An annotated list of a collection of reptiles
from southern California and northern Lower California.
Field Columb. Mus., Zool. Ser. 7(1): 1-1 9.
MEIGS, P. 1953. World distribution of arid and semi-arid
homoclimates. Pp. 202-210 in Reviews of research on arid
zone hydrology. Arid Zone Prog. 1 . UNESCO, Paris, France.
. 1 966. Geography of coastal deserts. Arid Zone Re-
search 28. UNESCO, Paris, France. 140 pp.
MERRIAM, C. H. 1898. Life zones and crop zones of the
United States. USDA Div. Biol. Surv., Bull. 10.
MILLER, A. H. 1951. An analysis of the distribution of the
birds of California. Univ. Calif. Publ. Zool. 50(6):53 1-643.
MILLER, C. M. 1944. Ecologic relations and adaptations of
the limbless lizards of the genus Anniella. Ecol. Monogr.
14(3):27 1-289.
MILLER, W. E. 1977. Pleistocene terrestrial vertebrates from
southern Baja California. Geol. Soc. Am. Abstract 9:468.
— . 1 980. The late Pliocene Las Tumas Local fauna from
southernmost Baja California. J. Paleontol. 54:762-805.
MINA, V. F. 1957. Bosquejo geologico del Territorio de la
Baja California. Boletin de la Associacion Mexicana de Geo-
logos Petroleros 9:141-269.
MINCH, J. C., R. G. GASTIL, W. FINK, J. ROBINSON, AND A.
H. JAMES. 1976. Geology of the Vizcaino Peninsula. Pp.
136-145 in Aspects of the geologic history of the California
Continental Borderland. D. G. Howell, ed. Pacific Section
of the American Association of Petroleum Geologists, Misc.
Pubs. 24.
MOCQUARD, F. 1 899. Contribution a la faune herpetologique
de la Basse California. Nouv. Arch. Mus. Nat. Hist. 4(1):
297-344.
MOORE, D. G. 1973. Plate-edge deformation and crustal
growth. Gulf of California structural province. Geol. Soc.
Am. Bull. 84:1883-1906.
MORAFKA, D. J. 1977. A biogeographical analysis of the Chi-
huahuan Desert through its herpetofauna. W. Junk B. V.,
The Hague, Netherlands. 3 1 2 pp.
MUNZ, P. A. AND D. KECK. 1949. California plant com-
munities. El Aliso 2(1):87-105.
. 1959. A California flora. University of California
Press, Berkeley, California. 1,681 pp.
MURPHY, R. W. 1974. A new genus and species of eu-
blepharine gecko (Sauria: Gekkonidae) from Baja California.
Mexico. Proc. Calif. Acad. Sci. 40:87-92.
. 1975. Two new blind snakes (Serpentes: Leptotyph-
lopidae) from Baja California, Mexico, with a contribution
to the biogeography of peninsular and insular herpetofauna.
Proc. Calif. Acad. Sci. 40(5):93-107.
. 1976. The evolution of a peninsular and insular her-
petofauna: a drift based alternative hypothesis. M.S. Thesis,
California State University, San Francisco. 7 1 pp.
. 1983a. Paleobiogeography and genetic differentia-
tion of the Baja California herpetofauna. Occ. Pap. Calif.
Acad. Sci. 137:1-48.
. 1983£. The reptiles: origin and evolution. Pp. 130-
158 in Island biogeography in the Sea of Cortez. T. J. Case
and M. L. Cody, eds. University of California Press, Berke-
ley. 508 pp.
MURPHY, R. W. AND J. R. OTTLEY. 1983. A distributional
checklist of the reptiles and amphibians on the islands in
the Sea of Cortez. Appendix 6.1 and 6.2, pp. 429-437 in
Island biogeography in the Sea of Cortez. T. J. Case and M.
L. Cody, eds. University of California Press, Berkeley. 508
pp.
. 1984. Distribution of amphibians and reptiles on
islands in the Gulf of California. Ann. Carnegie Mus. 53:
207-230.
MURPHY, R. W. AND T. J. PAPENFUSS. 1979. Biochemical
relationships, identification, and variation of Phyllodactylus
unctiisand Phyllodactylus paucituberculatus. Biochem. Syst.
Ecol. 8:97-100.
MURRAY, K. F. 1955. Herpetological collections from Baja
California. Herpetologica 1 1(l):33^t8.
NELSON, E. W. 1921. Lower California and its natural re-
sources. Nat. Acad. Sci. 16. First memoir. 194 pp.
NORRIS, K. W. 1958. The evolution and systematics of the
iguanid genus Uma and its relation to the evolution of other
North American desert reptiles. Bull. Am. Mus. Nat. Hist.
114:247-326.
OTTLEY, J. R. AND L. E. HUNT. 1981. Geographic distribu-
tion: Crotalus viridis helleri. Herpetol. Rev. 12(2):65.
OTTLEY, J. R. AND E. E. JACOBSEN. 1983. Pattern and col-
oration of juvenile Elaphe rosaliae, with notes on natural
history. J. Herpetol. 17(2):189-191.
PAPENFUSS, T. J. 1982. The ecology and systematics of the
amphisbaenian genus Bipes. Occ. Pap. Calif. Acad. Sci. 1 36.
42pp.
PEABODY, F. E. AND J. M. SAVAGE. 1958. Evolution of the
WELSH: BAJA CALIFORNIA HERPETOFAUNA
69
Coast Range corridor in California and its effect on the origin
and dispersal of living amphibians and reptiles. Am. Assoc.
Advance. Sci. Publ. 51:159-186.
PETERS, J. A. 1971. A new approach in the analysis of bio-
geographical data. Smithson. Contr. Zool. 107:1-28.
PHELAN, R. L. AND B. H. BRATTSTROM. 1955. Geographical
variation in Sceloporus magister. Herpetologica 1 1(1):1-14.
PIELOU, E. C. 1979. Biogeography. J. Wiley and Sons, New
York. 351 pp.
PLATNICK, N. I. AND G. NELSON. 1 978. A method of analysis
for historical biogeography. Syst. Zool. 27(1):1-16.
PRESH, W. 1969. Evolutionary osteology and relationships
of the horned lizard genus Phrynosoma (family Iguanidae).
Copeia 1969(2):250-275.
RAU, C. S. AND R. B. LOOMIS. 1977. A new species of Uro-
saurus (Reptilia, Lacertilia, Iguanidae) from Baja California,
Mexico. J. Herpetol. ll(l):25-29.
REEVE, W. L. 1952. Taxonomy and distribution of the horned
lizards genus Phrynosoma. Bull. Univ. Kan. Sci. 34(2):817-
960.
ROBINSON, J. W. 1972. Camping and climbing in Baja. La
Siesta Press, Glendale, California. 96 pp.
ROBINSON, M. D. 1973. Chromosomes and systematics of
the Baja California whiptail lizards Cnemidophorus hyp-
erythrus and C. ceralbensis (Reptilia: Teiidae). Syst. Zool.
22(l):30-35.
ROGERS, T. L. AND H. S. FITCH. 1 947. Variation in the skinks
of the Skiltonianus group. Univ. Calif. Publ. Zool. 48:169-
219.
ROTRAMEL, G. L. 1973. The development and application
of the area concept in biogeography. Syst. Zool. 22:227-232.
SARICH, V. M. 1977. Rates, sample sizes and neutrality hy-
pothesis for electrophoresis in evolutionary studies. Nature
265:24-28.
SAS INSTITUTE, INC. 1985. SAS user's guide: statistics, ver-
sion 5 edition. SAS Institute, Gary, North Carolina. 956 pp.
SAVAGE, J. M. 1952. Studies on the lizard family Xantusiidae.
I. The systematic status of the Baja California night lizards
allied to Xantusia vigilis, with the description of a new sub-
species. Am. Midi. Nat. 48(2):467-479.
. 1960. Evolution of a peninsular herpetofauna. Syst.
Zool. 9(3-4): 184-2 12.
. 1967. Evolution of the insular herpetofaunas. Pp.
. 1945. The chuckwalla, genus Sauromalus. Trans.
San Diego Soc. Nat. Hist. 10(15):269-306.
1953. Anniella pulchra and A. geronimensis; sym-
219-227 in Proceedings of the symposium on the biology
of the California islands. Santa Barbara Botanic Gardens,
Santa Barbara, California.
SCHMIDT, K. P. 1922. The amphibians and reptiles of Lower
California and the neighboring islands. Bull. Am. Mus. Nat.
Hist. 46:607-707.
SCHOENHERR, A. A. 1976. The herpetofauna of the San Ga-
briel Mountains, Los Angeles County, California. Special
Publication of the Southwest Herpetological Society, Feb. 1,
1976.95pp.
Scorr, N. J. AND R. W. McDiARMio. 1984. Trimorphodon
biscittatus. Cat. Am. Amphib. Rept. 353.1-353.4.
SEELIGER, L. M. 1945. Variation in the Pacific mud turtle.
Copeia 1945(3): 150-1 59.
SEIB, R. L. 1 980. Baja California: a peninsula for rodents but
not for reptiles. Am. Nat. 1 15(4):6 13-620.
SEIFERT, W. 1980. Geographical distribution: Arizona ele-
gans occipitalis. Herpetol. Rev. 1 1(2):39.
SHAW, C. E. 1940. A new species of legless lizard from San
Geronimo Island, Lower California. Trans. San Diego Soc.
Nat. Hist. 9(24):225-227.
patric species. Herpetologica 8(4): 167-1 70.
SHORT, L. L. AND R. CROSSIN. 1967. Notes on the avifauna
of northwestern Baja California. Trans. San Diego Soc. Nat.
Hist. 14(20):28 1-300.
SHREVE, F. 1934. Vegetation of the northwestern coast of
Mexico. Bull. Torrey Bot. Club 61(7):373-380.
- . 1936. The transition from desert to chaparral in Baja
California. Madrono 3:257-320.
SHREVE, F. AND L. WIGGINS. 1964. Vegetation and flora of
the Sonoran Desert. Stanford University Press, Palo Alto,
California. 1,740pp.
SILVER, L. T., E. G. STEHLI, AND C. R. ALLEN. 1956. Lower
Cretaceous prebatholith rocks of northern Baja California.
Calif. Inst. Tech., Div. Geol. Sci., Contr. No. 799:1-1 1.
SIMPSON, G. G. 1960. Notes on the measurement of faunal
resemblance. Am. Sci. 258:300-31 1.
SMITH, H. M. 1972. The Sonoran subspecies of the lizard
Ctenosaura hemilopha. Great Basin Nat. 32(2): 104-1 1 1.
SMITH, H. M. AND R. L. HOLLAND. 1971. Noteworthy snakes
and lizards from Baja California. J. Herpetol. 5(l-2):56-69.
SMITH, H. M. AND W. W. TANNER. 1 972. Two new subspecies
of Crotaphytus (Sauna, Iguanidae). Great Basin Nat. 32:25-
34.
- . 1974. A taxonomic study of the western collared
lizards, Crotaphytus collaris and Crotaphytus insularis. Brig-
ham Young Univ. Sci. Bull., Biol. Series 19(4): 1-29.
SOULE, M. AND A. J. SLOAN. 1966. Biogeography and dis-
tribution of the reptiles and amphibians on islands in the
Gulf of California, Mexico. Trans. San Diego Soc. Nat. Hist.
STEBBINS, R. C. 1954. Amphibians and reptiles of western
North America. McGraw-Hill Co., New York. 536 pp.
- . 1966. A field guide to western reptiles and amphib-
ians. Houghton Mifflin Co., Boston, Massachusetts. 279 pp.
- . 1985. A field guide to western reptiles and amphib-
ians, 2nd ed., revised. Houghton Mifflin Co., Boston, Mas-
sachusetts. 336 pp.
STEJNEGER, L. 1890. Description of a new lizard from Lower
California. Proc. U.S. Natl. Mus. 12:643-644.
. 1 893a. Annotated list of the reptiles and batrachians
collected by the Death Valley Expedition in 1891, with de-
scription of new species. N. Am. Fauna 7:159-228.
. 1893ft. Diagnosis of a new California lizard. Proc.
U.S. Natl. Mus. 16(944):467.
-. 1 894. Description of Uta mearnsi, a new lizard from
California. Proc. U.S. Natl. Mus. 17:589-591.
STICKEL, W. H. 1938. The snakes of the genus Sonora in the
United States and Lower California. Copeia 1 938(4): 182-
190.
. 1943. The Mexican snakes of the genus Sonora and
Chionactis with notes on the status of other colubrid genera.
Proc. Biol. Soc. Washington 56:109-128.
STORER, T. I. AND R. L. USINGER. 1963. Sierra Nevada nat-
ural history. University of California Press, Berkeley. 374
PP.
TANNER, W. W. 1946. A taxonomic study of the gen us Hyp-
siglena. Great Basin Nat. 5(3-4):25-92.
. 1953. A study of taxonomy and phylogeny of Lam-
propeltis pyromelana Cope. Great Basin Nat. 13(l-2):47-
66.
70
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
. 1966a. The night snakes of Baja California. Trans.
San Diego Soc. Nat. Hist. 14(1 5): 189-1 96.
. 1 966ft. A re-evaluation of the genus Tantilla in the
southwestern United States and northwestern Mexico. Her-
petologica 22(2): 134-1 52.
TAYLOR, E. H. 1935. A taxonomic study of the cosmopolitan
scincoid lizards of the genus Eumeces with an account of
the distribution and relationships of its species. Kans. Univ.
Sci. Bull. 23:1-643.
TEVIS, L. 1944. Herpetological notes from Lower California.
Copeia 1944(1):6-18.
UDVARDY, M. D. F. 1 969. Dynamic zoogeography. Van Nos-
trand Reinhold Co., New York. 445 pp.
U.S. DEPT. COMMERCE. 1979. Operational navigation chart
(ONQCH-22. Aeronautical chart and information center,
St. Louis, Missouri.
VAN COUVERING, J. A. 1 978. Status of late Cenozoic bound-
aries. Geology 6:169.
VAN DENBURGH, J. 1894. Description of three new lizards
from California and Lower California with a note on Phry-
nosoma blainvillii. Proc. Cal. Acad. Sci. 4(2):296-301.
. 1922. The reptiles of western North America. Occ.
Pap. Calif. Acad. Sci. 10:1-1028.
VAN DENBURGH, J. AND J. R. SLEVIN. 1921. Preliminary
diagnosis of more new species of reptiles from islands in the
Gulf of California. Proc. Calif. Acad. Sci. 1 1(4):395-398.
. 1923. Preliminary diagnosis of four new snakes from
Lower California, Mexico. Proc. Calif. Acad. Sci. 1(13):
1-2.
VAN DEVENDER, T. R. AND W. G. SPAULDING. 1979. Devel-
opment of vegetation and climate in the southwestern United
States. Science 204:701-710.
WAKE, D. B. 1 966a. The colubrid snake genus Chionactis in
Baja California, Mexico. Copeia 1966(2):364.
. 1966ft. Comparative osteology and evolution of the
lungless salamanders, family Plethodontidae. Mem. So. Cal-
if. Acad. Sci. 4:1-111.
WALKER, J. M. AND H. L. TAYLOR. 1968. Geographical vari-
ation in the teiid lizard Cnemidophorus hyperythrus. I. The
caeruleus-like subspecies. Am. Midi. Nat. 80(1): 1-27.
WASSERMAN, A. O. 1970. Scaphiopus couchii. Cat. Am. Am-
phib. Rept. 85.1-85.4.
WELSH, H. H., JR. 1976a. Ecogeographic distribution of the
herpetofauna of the Sierra San Pedro Martir Region, Baja
California Norte, Mexico. M.S. Thesis, Humboldt State Uni-
versity, Arcata, California. 169 pp.
. 1976ft. Evolutionary trends and zoogeographic his-
tory of the Baja California herpetofauna. 45 pp. [Unpub-
lished ms, mimeo]
. 1981. Ecogeographic distribution of the northern Baja
California herpetofauna and its relationships with the west-
ern Pacific herpetofauna. Abstract, paper given at American
Society of Ichthylogists and Herpetologists Meeting. Oregon
State University. June 1981.
WELSH, H. H., JR. AND R. B. BURY. 1984. Additions to the
herpetofauna of the south Colorado Desert, Baja California,
with comments on the relationships ofLichanura trivirgata.
Herpetol. Rev 15(2):53-56.
WIGGINS, I. L. 1 944. Notes on the plants of northern Baja
California. Dudley Herb. Stanford Univ. Contr. 3(9):289-
312.
. 1960. The origins and relationships of land flora. In
Symposium: the biogeography of Baja California and ad-
jacent seas. Syst. Zool. 9(1-3): 148-1 65.
WILSON, L. E. 1973. Masticophis flagellum. Cat. Am. Am-
phib. Rept. 145.1-145.4.
WOODFORD, A. O. AND T. F. HARRIS. 1938. Geological re-
connaissance across Sierra San Pedro Martir, Baja Califor-
nia. Bull. Geol. Soc. Am. 9:1297-1336.
WYLES, J. S. AND G. C. GORMAN. 1978. Close relationship
between the lizard genus Sator and Sceloporus utiformis
(Reptilia, Lacertilia, Iguanidae): electrophoretic and im-
munological evidence. J. Herpetol. 12(3):343-350.
YANEV, K. P. 1978. Evolutionary studies of the plethodontid
salamander genus Batrachoseps. Ph.D. Dissertation, Uni-
versity of California, Berkeley.
YARROW, H. C. 1883 (1882). Descriptions of new species of
reptiles and amphibians in the United States National Mu-
seum. Proc. U.S. Natl. Mus. 5(299):438-443.
YINGLING, P. R. 1982. Lichanura, L. trivirgata. Cat. Am.
Amphib. Rept. 294.1-294.2.
ZWEIFEL, R. G. 1952. Pattern variation and evolution of the
mountain king snake, Lampropeltis zonata. Copeia 1952(3):
152-168.
. 1955. The ecology and systematics of the Ranaboylei
species group. Univ. Calif. Publ. Zool. 54:207-292.
. 1974. Lampropeltis zonata. Cat. Am. Amphib. Rept.
174.1-174.4.
WELSH: BAJA CALIFORNIA HERPETOFAUNA 7 1
APPENDIX A. COLLECTING STATIONS IN THE SIERRA SAN PEDRO MARTIR REGION, BAJA CALIFORNIA NORTE, MEXICO.
Station
number
Elevation
Primary
vegetation*
Name and coordinates
ft
m
1
San Matias Pass; 31°19'N, 115°30'W
3,200
980
CBS
2
Vallede Picacho; 31°15'N, 115°35'W
3,400
1,004
CBS
3
Rancho El Burro; 31°14'N, 1 15°36'W
4,450
1,360
P-J
4
Canon Esperanza (mouth); 31°12'N, 1 15°27'W
1,640
500
CBS
5
Valle de San Felipe (center); 31°12'N, 1 15°20'W
1,575
480
CBS
6
Canon Copal (mouth); 3l°09'N, 115°25'W
1,970
600
CBS
7
San Telmo; 31°08'N, 1 16°05'W
400
120
CDS
8
Arroyo Leon (head); 31°08'N, 1 15031'W
7,400
2,260
Conifer
9
San Antonio del Mar; 31°07'N, 116°17'W
75
23
CDS
10
Arroyo de San Rafael, 6 mi E MS Rancho; 31°07'N, 1 15°33'W
5,000
1,525
P-J, Chp
11
Mikes Sky Rancho; 31°06'N, 1 15°38'W
4,150
1,270
Chp
12
Cerro Venado Blanco; 31°06'N, 1 15°29'W
9,325
2,820
Conifer
13
Canon Copal (head); 31°06'N, 1 15°29'W
7,375
2,225
Conifer
14
Arroyo de San Rafael, 2 mi E MS Rancho; 31°05'N, 1 15°37'W
4,350
1,330
Chp
15
Arroyo de San Rafael (head); 31°05'N, 1 15°31'W
7,275
2,210
Conifer
16
Canon del Diablo, 2 mi W mouth; 31°05'N, 1 15°24'W
2,500
760
CBS
17
Canon del Diablo (mouth); 31°05'N, 1 15°22'W
2,175
660
CBS
18
Mikes Sky Rancho, airstrip; 31°04'N, 1 15°38'W
4,600
1,400
Chp
19
Rancho Conception; 3I°04'N, 115°34'W
4,775
1,450
Chp
20
Canon del Diablo, 4 mi W mouth; 31°04'N, 1 15°24'W
2,675
810
P-J/CBS
21
Canon del Diablo, 6 mi W mouth; 31°03'N, 1 15°24'W
3,150
960
P-J/CBS
22
Observatorio Nacional de Mexico; 31°02'N, 1 15°27'W
9,200
2,800
Conifer
23
Canon del Diablo, 8 mi W mouth; 31°02'N, 1 15°24'W
4,000
1,220
P-J
24
Compamento Forestal; 31°01'N, 115°28'W
8,100
2,460
Conifer
25
Yerba Buena Spring; 31°00'N, 1 15°28'W
8,000
2,420
Conifer
26
Vallecitos, SE end; 30°59'N, 115°28'W
8,000
2,420
Conifer
27
Rancho San Jose; 30°58'N, 115°25'W
2,300
640
CSS
28
Socorro; 30°58'N, 115°40'W
4,200
1,280
CSS/Chp
29
Oak Pasture; 30°58'N, 115°36'W
5,900
1,830
Chp
30
Los Llanitos; 30°58'N, 115°26'W
7,900
2,410
Conifer
31
Picacho Spring; 30°58'N, 1 15°25'W
8,200
2,500
Conifer
32
Blue Bottle Mt.; 30°58'N, 115°24'W
9,675
2,950
Conifer
33
Rancho Viejo trail head, La Tasajera ridge; 30°57'N, 1 15°30'W
7,900
2,410
Conifer
34
Llanitos Camp; 30°57'N, 115°26'W
7,700
2,350
Conifer
35
La Jolla; 30°56'N, 115°35'W
5,250
1,600
Chp
36
La Tasajera Ridge, 2 mi NE R. Viejo; 30°56'N, 1 15°28'W
7,000
2,120
Conifer
37
Rancho Viejo; 30°55'N, 1 15°29'W
6,900
2,200
Conifer
38
San Antonio Creek, 1 mi W La Grulla; 30°54'N, 1 15°30'W
6,750
2,060
Conifer
39
La Grulla, W end; 30°54'N, 1 15°29'W
6,800
2,080
Conifer
40
La Grulla, center; 30°54'N, 1 15°27'W
6,900
2,100
Conifer
41
La Encantada Creek; 30°54'N, 115°26'W
7,050
2,150
Conifer
42
Arroyo San Antonio, head; 30°53'N, 1 15°30'W
6,575
2,000
Conifer
43
La Grulla, SE end; 30°53'N, 1 15°26'W
6,950
2,110
Conifer
44
La Encantada; 30°53'N, 1 15°25'W
7,100
2,160
Conifer
45
Arroyo Valladares, Santa Cruz Road; 30°52'N, 1 15°43'W
2,025
620
CSS
46
Canon El Cajon (mouth); 30°52'N, 1 15°14'W
2,125
650
CBS
47
El Alcatraz; 30°51'N, 1 15°26'W
6,800
2,080
Conifer
48
Canon El Cajon, 6 mi W mouth; 30°51'N, 115°16'W
4,000
1,220
P-J/CBS
49
Rancho San Antonio; 30°50'N, 1 15°38'W
1,970
600
CSS
50
El Remudero; 30°50'N, 115°22'W
6,900
2,200
Conifer
51
Canon Barrosa/Canon El Cajon divide; 30°50'N, 1 15°17'W
6,575
2,000
P-J
52
Cowpod Camp, Canon El Cajon; 30°50'N, 1 15°16'W
5,200
1,580
P-J
53
Canon Barrosa (head); 30°49'N, 115°19'W
7,550
2,300
Chp
54
Valle de Chico (center); 30°49'N, 1 15°08'W
1,575
480
CBS
55
Santa Rosa; 30°48'N, 115°2r\V
6,800
2,080
Conifer
56
Mission San Pedro Martir; 30°47'N, 1 15°24'W
5,500
1,675
Chp
57
Santo Tomas; 30°46'N, 1 15°24'W
5,900
1,800
Chp
72
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1
APPENDIX A. CONTINUED.
Elevation
Station
Primary
number Name and coordinates
ft
m
vegetation*
58 Arroyo Alonso (head); 30°46'N, 1 1 5°20' W
6,575
2,000
Conifer
59 Canon Agua Caliente, 9 mi W mouth; 30°41'N, 1 15°16'W
3,125
950
CBS
60 Canon Agua Caliente, 4 mi W mouth; 30°40'N, 1 15°12'W
2,225
680
CBS
6 1 Canon Auga Caliente (mouth); 30°40'N, 1 1 5°09' W
2,200
670
CBS
62 Rancho Rosarito; 30°28'N, 1 1 5°1 8' W
2,750
840
CSS/CBS
63 Rio de Rosarito; 30°2 1'N, 1 1 5°22' W
2,200
670
CDS
64 San Juan de Dios; 30°1 1'N, 11 5°09'W
2,175
660
CDS
65 Rancho El Metate; 30°1 1'N, 1 15°08'W
2,200
670
CDS
* Key to vegetations: Conifer = Coniferous Forest, CSS = Coastal Sage Scrub, CBS
= Creosote Bush Scrub,
CDS = Central
Desert Scrub, Chp = Chaparral, and P-J = Pinyon-Juniper Woodland.
CALIFORNIA ACADEMY OF SCIENCES
Golden Gate Park
San Francisco, California 941 18
PROCEEDINGS
OF THE
CALIFORNIA ACADEMY OF SCIENCES
Vol. 46, No. 2, pp. 73-81, 5 figs. December 7, 1988
THREE NEW SPECIES OF HOLOGRAPHIS (ACANTHACEAE)
FROM MEXICO
By
Thomas F. Daniel
Department of Botany, California Academy of Sciences, Golden Gate Park,
San Francisco, California 94118
ABSTRACT: Three new species of Holographis, H. caput-medusae, If. leticiana, and //. tolantongensis, are
described from Mexico. Diagnostic features of each species are illustrated and the known distribution of the
three species is mapped. A revised key to the entire genus, incorporating data and species that have become
known since the last monograph of Holographis, is presented. Fifteen species are now known for this entirely
Mexican genus.
Received February 16, 1988. Accepted May 11, 1988.
additional collections of previously recognized
Holographis Nees is a small genus of perennial species of Holographis has become available,
herbs and shrubs endemic to Mexico. Since my Some species (e.g., H. argyrea, H. anisophylla,
monograph of the genus (Daniel 1983), in which and H. parayand) are now known from twice as
10 species were recognized, one species (H. pe- many or more collections than they were in 1983.
lorid) has been transferred into Holographis from Fruiting collections have been located for four
Stenandrium Nees (Daniel 1984) and an addi- species (H. argyrea, H. anisophylla, H. peloria,
tional species (H. websteri) has been newly de- and H. pueblensis) whose capsules and seeds were
scribed (Daniel 1986). Continued botanical ex- previously undescribed. The known geographic
ploration in the dry forests of Mexico has revealed distributions of several species (e.g., H. aniso-
the existence of three undescribed and distinctive phylla and H. peloria) have been increased. A
species, bringing the total number of species now key to all species of the genus is presented below
known for the genus to 15. These three new that includes the five taxa recognized in Holo-
species are described below. graphis since the monograph of 1983 and that
Because of the increased botanical activity in incorporates information from other species
Mexico in recent years, a significant number of based on additional collections examined.
KEY TO THE SPECIES OF HOLOGRAPHIS
1 . Leaves opposite (rarely subopposite).
2. Corolla yellow with maroon markings, vertical during an thesis, the upper lip 1-2.5 mm
long; stamens 1.5 mm long with thecae 1-1.3 mm long; style 4 mm long; Balsas basin in
Guerrero. H. argyrea
[73]
74 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 2
2. Corolla white or pinkish, horizontal during anthesis, the upper lip 2.5-5 mm long; stamens
2.8-5.5 mm long with thecae 1.5-2.5 mm long; style 5-7.5 mm long; plants occurring to
the north of Guerrero.
3. Abaxial surface of bracts, bractlets, and calyx glabrous; calyx 4-5 mm long; corolla
white; Tamaulipas H. tamaulipica
3. Abaxial surface of bracts, bractlets, and calyx pubescent; calyx 5-10 mm long; corolla
pinkish; plants occurring to the south and west of Tamaulipas.
4. Corolla glabrous on external surface; calyx lobes aristate at apex; staminode 2 mm
long; Hidalgo H. tolantongensis
4. Corolla pubescent on external surface; calyx lobes acute at apex; staminode 0.5-0.8
mm long; Sonoran Desert region. H. virgata
1 . Leaves whorled, four per node (lower leaves of H. pueblensis sometimes opposite to subop-
posite; rarely opposite on some, but not all, shoots of individuals of H. ehrenbergiand).
5. Leaves orbicular, 0.8-1.3 times longer than wide, the margin spinose toothed; bracts al-
ternate; Chihuahuan Desert region. H. ilicifolia
5. Leaves lanceolate to ovate to elliptic to obovate to oblanceolate, 1.5-6 times longer than
wide, margin entire to undulate; bracts opposite; plants occurring outside of the Chihuahuan
Desert region.
6. Corolla yellow, the upper lip 6-9 mm long; stamens 9-1 1 mm long; seeds glabrous.
7. Inflorescence to 4.5 cm long, many-flowered; bracts 5-1 1 mm long, 1.5-2 mm wide;
calyx 7-10.5 mm long H. pueblensis
1. Inflorescence to 2 cm long, usually reduced to 2 flowers; bracts 1.5-5 mm long, 0.5-
1 mm wide; calyx 3.5-6.5 mm long H. ehrenbergiana
6. Corolla white, pinkish, or reddish (color unknown in H. hintonii), the upper lip 1.5-
5.5 mm long; stamens 1.2-6.5 mm long; seeds pubescent or with hairlike papillae
(glabrous in H. pelorid).
8. Leaves up to 13 mm long; corolla 6.5-7.5 mm long, the lower lip 2.5-3.5 mm long;
stamens 1.2 mm long with thecae 1 mm long; style 1.8-2.5 mm long; Yucatan
Peninsula. H. websteri
8. Leaves up to 105 mm long; corolla 8-17 mm long, the lower lip 4-9 mm long;
stamens 2.5-6.5 mm long with thecae 1.2-2.6 mm long; style 3.5-9.5 mm long;
plants occurring to the west of the Yucatan Peninsula.
9. Bracts 6-1 3 mm long; bractlets 6-10.5 mm long; corolla pubescent with eglandular
trichomes on external surface; seeds glabrous, covered with stout papillae or
tubercules.
1 0. Cauline trichomes to 0. 1 mm long; bracts lanceolate, 0.8-1 .2 mm wide, long-
attenuate and recurved at apex; bractlets and calyx lobes recurved; capsule
puberulent; Guerrero H. caput- medusae
10. Cauline trichomes 0.1-1 mm long; bracts ovate-elliptic to obovate, 2.8-5
mm wide, rounded to acute and erect at apex; bractlets and calyx lobes erect;
capsule glabrous; Durango and Zacatecas H. peloria
9. Bracts 1.5-6 mm long; bractlets 1.5-4 mm long; corolla pubescent with glandular
(sometimes absent in H. parayana and H. pallidd) and eglandular trichomes on
external surface; seeds pubescent with dendroid trichomes.
1 1 . Inflorescence axis pubescent with crooked, interwoven, eglandular trichomes
obscuring the axis; bracts broadly ovate to orbicular, 1.5-2 mm long
H. hintonii
1 1 . Inflorescence axis pubescent with straight to ascendent to recurved to flex-
uose, eglandular (and often glandular) trichomes, these neither crooked, in-
terwoven, nor obscuring the axis; bracts triangular to subulate to lanceolate
to ovate, 2-6 mm long.
1 2. Bracts attenuate to a point but lacking a distinct mucro at apex; upper
DANIEL: NEW SPECIES OF HOLOGRAPHS
75
lip of corolla 3-5.5 mm long; staminode 0.6-1.2 mm long; capsule 1 1-
17 mm long.
13. Cauline trichomes ascendent-appressed, 0.1-0.5 mm long; corolla
horizontal during an thesis; stamens 3-3.2 mm long with thecae 1.2-
1.5 mm long; style 5-6 mm long; seeds 2.5-3 mm long; Chiapas.
H. parayana
13. Cauline trichomes retrorse, 0.05-0.2 mm long; corolla more or less
vertical during anthesis; stamens 5.5-6.5 mm long with thecae 2-
2.6 mm long; style 7-9.5 mm long; seeds 4.5-5.5 mm long; Oaxaca.
H. leticiana
1 2. Bracts mucronate at apex; upper lip of corolla 1.5-3 mm long; staminode
0.1-0.3 mm long; capsule 7-9 mm long.
14. Bractlets 1-1.5 mm wide; corolla 8-10 mm long, the upper lip 1.5-
2 mm long; stamens 2.5-3 mm long; style 3.5-4.5 mm long; capsule
glabrous; Jalisco and Colima H. anisophylla
14. Bractlets 0.8-1 mm wide; corolla 10-12 mm long, the upper lip 2-
3 mm long; stamens 3.5-4 mm long; style 5.3-6 mm long; capsule
pubescent; Sonora and Sinaloa H. pallida
Holographis tolantongensis T. F. Daniel, sp. nov.
(Figures 1, 2)
TYPE. — MEXICO. Hidalgo: Mpio. Cardonal, Barranca de
Tolantongo, lado oeste, 45 km NE of Ixmiquilpan, 10 February
1982, R. Herndndez M. y P. Tenorio L. 6948 (Holotype:
MEXU!; isotype: CAS!).
Suffrutex usque ad 6 dm altus. Caules juniores
pubescentes trichomatibus eglandulosis 0.05-0.7
mm longis. Folia opposita petiolata, lamina lan-
ceolata vel ovata vel elliptica 13-32 mm longa
3.5-14 mm lata 2.3-4.7-plo longior quam latior
margine revoluta. Inflorescentia spicata, brac-
teae lanceolatae 6.5-9 mm longae 1.5-2.5 mm
latae apice acuminatae erectae, bracteolae lanci-
subulatae 5.5-7 mm longae 1-1.5 mm latae apice
subaristatae erectae. Calyx 7.5-10 mm longus,
corolla rosea-purpurea 12-13 mm longa extus
glabra, stamina 4.5 mm longa, stylus 7-7.5 mm
longus. Capsula ignota.
Subshrub to 6 dm tall. Young stems evenly
pubescent with understory of straight to antrorse,
eglandular trichomes 0.05-0.2 mm long and
sparse overstory of straight to flexuose, eglan-
dular trichomes 0.3-0.7 mm long. Leaves op-
posite, petiolate, petioles 2-6 mm long, blades
coriaceous, lanceolate to ovate to elliptic, 13-32
mm long, 3.5-14 mm wide, 2.3-4.7 times longer
than wide, acute at apex, acute at base, adaxial
surface dark green, evenly pubescent with flex-
uose to antrorse-flexuose trichomes to 1 mm long
(strigose), abaxial surface light green, pubescent
along veins and glabrous or nearly so in inter-
costal regions, margin revolute. Inflorescence of
terminal spikes to 3.5 cm long, flowers opposite
at nodes, rachis strigose. Bracts lanceolate, 6.5-
9 mm long, 1.5-2.5 mm wide, acuminate and
erect at apex, abaxial surface strigose. Bractlets
lance-subulate, 5.5-7 mm long, 1-1.5 mm wide,
shorter than bracts, subaristate and erect at apex,
margin often hyaline. Calyx 7.5-10 mm long,
lobes narrowly lanceolate to lance-subulate, un-
equal in length and overlapping one another,
sparsely strigose distally, aristate and erect at apex,
margins often hyaline. Corolla pinkish-purple,
somewhat intermediate between vertical and
horizontal during anthesis, 12-13 mm long, gla-
brous on external surface, tube 7-7.5 mm long,
upper lip 4-4.5 mm long with lobes 2-2.5 mm
long, lower lip 6-7 mm long with obovate lobes
5-6 mm long. Stamens 4.5 mm long, filaments
pubescent, thecae 1.4-1.5 mm long, pollen (ob-
served in Diaphane) prolate, 32-39 nm long (po-
lar diameter), 25-27 nm wide (equatorial di-
ameter), 1.3-1.6 times longer than wide, colpi
indistinct; staminode 2 mm long, pubescent at
apex. Style 7-7.5 mm long, glabrous; stigma
asymmetric, 0.2 mm long; ovary glabrous. Cap-
sule unknown.
DISTRIBUTION AND PHENOLOGY.— Known only
from the Barranca de Tolantongo in central Hi-
dalgo, Mexico (Fig. 3). The plants occur abun-
dantly in rocky soil with Pinus pinceana Gordon,
Quercus opaca Trel., Gochnatia hypoleuca (DC.)
A. Gray, and Juniperus L. at an elevation of
about 2,000 m. Flowering: February.
76
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 2
FIGURE 1 . Holographis tolantongensis (Herndndez M. y Tenorio L. 6948). a, Habit, x 0.5; b, Leaf and section of stem, x 3.3;
c, Portion of inflorescence, x3.5; d, Corolla cut open to show androecium, x3.5; e, Style, x4.6.
Holographis tolantongensis exhibits the fol-
lowing unique combination of characters: op-
posite leaves, bracts with the abaxial surface pu-
bescent, pinkish-purple corollas with the external
surface glabrous, and calyces 7.5-10 mm long.
The only other species of Holographis known
from Hidalgo, H. ehrenbergiana, has mostly
whorled leaves and yellow corollas.
Holographis caput-medusae T. F. Daniel, sp. nov.
(Figures 2, 4)
TYPE. — MEXICO. Guerrero: along road from Petatlan to
Camalotito, 9.8 km SW of Camalotito, 8 March 1987, T. F.
Daniel and B. Bartholomew 4918 (Holotype: CAS!; isotypes:
K!, MEXU!, MICH!, NY!, US!).
Herba perennis usque ad 4 dm alta. Caules
juniores pubescentes trichomatibus inconspicuis
DANIEL: NEW SPECIES OF HOLOGRAPHS
77
FIGURE 2. Scanning electron micrographs of pollen of Holographis. a, H. tolantongensis (Hernandez M. and Tenorio L.
6948), scale equals 6 nm; b, H. caput- medusae (Daniel and Bartholomew 4918), scale equals 6 /*m; c, Holographis leticiana
(Torres C. et al. 637), scale equals 7.5 /mi.
retrorsi eglandulosis usque ad 0.1 mm longis.
Folia verticillata in quoque nodo quatuor sessilia
vel subsessilia, lamina anguste elliptica vel ellip-
tica vel ovato-elliptica vel obovato-elliptica 30-
97 mm longa 10—40 mm lata 2.1-6-plo longior
quam latior margine plana. Inflorescentia spi-
cata, bracteae lanceolatae 6-8 mm longae 0.8-
1.2 mm latae apice longiattenuatae recurvatae,
bracteolae lanceolatae 6-8 mm longae 0.5-1 mm
latae apice longiattenuatae recurvatae. Calyx 8.5-
10 mm longus, corolla albida et rosea 11-14 mm
longa extus pubescens, stamina 3.5 mm longa,
stylus 6.5-7.5 mm longus. Capsula 8-9 mm lon-
ga puberula. Semina 2.2-3 mm longa, 1.8-2 mm
lata.
Perennial herb to 4 dm tall. Young stems in-
conspicuously pubescent with retrorse, eglan-
dular trichomes up to 0. 1 mm long concentrated
in two vertical lines to glabrous. Leaves whorled
(four/node), sessile to subsessile, petioles (if pres-
ent) up to 1 mm long, blades chartaceous, nar-
rowly elliptic to elliptic to ovate-elliptic to ob-
CAPUT-MEDUSAE
LETICIANA
TOLANTONGENSIS
FIGURE 3. Map showing distribution of Holographis caput- medusae, H. leticiana, and H. tolantongensis.
78
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 2
FIGURE 4. Holographys caput-medusae (Daniel and Bartholomew 4918). a, Habit, xQ.5; b, Inflorescence, x 3.4; c, Corolla,
x3.6; d, Corolla cut open to show androecium, x4.3; e, Capsule, x4; f, Seed, x 1 1.5.
ovate-elliptic, 30-97 mm long, 10-40 mm wide,
2. 1-6 times longer than wide, acuminate at apex,
attenuate nearly to or to the node at base, adaxial
surface sparsely pubescent with coarse, antrorse,
eglandular trichomes 0.2-0.5 mm long, abaxial
surface glabrous or nearly so, margin entire to
somewhat undulate, flat. Inflorescence of axillary
spikes to 3 cm long, flowers opposite at nodes,
rachis evenly and densely pubescent with straight,
eglandular trichomes to 0.05 mm long. Bracts
lanceolate, 6-8 mm long, 0.8-1.2 mm wide,
abaxial surface pubescent like rachis, apex long-
attenuate and recurved. Bractlets lanceolate, 6-
8 mm long, 0.5-1 mm wide, approximately equal
DANIEL: NEW SPECIES OF HOLOGRAPHS
79
in length to bracts, pubescent like bracts, apex
long-attenuate and recurved. Calyx 8.5-10 mm
long, lobes equal in length and slightly overlap-
ping one another, pubescent like bracts, long-
attenuate and recurved at apex, margin hyaline
proximally. Corolla whitish with pink lines on
lower lip, horizontal during anthesis, 11-14 mm
long, pubescent on external surface with eglan-
dular trichomes 0.05-0. 1 mm long, tube 5-7 mm
long, upper lip 4.5-5.5 mm long with linear-
elliptic lobes 4.5-5.5 mm long, lower lip 6-7 mm
long with elliptic to obovate lobes 3-4.7 mm
long. Stamens 3.5 mm long, filaments pubescent,
thecae 1.3-1.6 mm long, pollen (observed in Dia-
phane) prolate, 34-42 /um long (polar diameter),
25-27 jum wide (equatorial diameter), 1.4-1.6
times longer than wide, colpi distinct; staminode
0.6 mm long, pubescent at apex. Style 6.5-7.5
mm long, sparsely pubescent; stigma unequally
bilobed, 0.2-0.3 mm long. Capsule 8-9 mm long,
puberulent with trichomes less than 0.05 mm
long. Seeds subelliptical to subrectangular in out-
line, 2.2-3 mm long, 1.8-2 mm wide, surfaces
covered with hairlike papillae to 0. 1 mm long.
DISTRIBUTION AND PHENOLOGY.— Known only
from the seaward slopes of the Sierra Madre Sur
in western Guerrero (Fig. 3) in a region of tropical
dry forest at an elevation of approximately 1 50
m. Flowering and fruiting: March.
Holographis caput- medusae is unique in the
genus by virtue of its relatively long, attenuate-
recurved bracts, bractlets, and calyx lobes. These
give the dense spikes a medusoid appearance.
Holographis caput- medusae is the third species
of the genus to be found in Guerrero. Unlike the
other two, H. argyrea and H. hintonii, it occurs
in the Pacific lowlands. The holotype was col-
lected in a weedy thicket rich in Acanthaceae,
including: Aphelandra scabra (Vahl) Smith, Car-
lowrightia arizonica A. Gray, Didiptera resupi-
nata (Vahl) Juss., Elytraria imbricata (Vahl) Pers.,
Henrya insularis Nees, Odontonema auricula-
turn (Rose) T. Daniel, Ruellia albiflora Fernald,
Tetramerium nervosum Nees, and Tetramerium
tenuissimum Rose.
Holographis leticiana T. F. Daniel, sp. nov.
(Figures 2, 5)
TYPE. -MEXICO. Oaxaca: Distr. Tehuantepec, subida al
Cerro Guiengola por la ladera S donde esta la fabrica de cal,
25 October 1986, M. Leticia Torres C., R. Torres C., y C.
Martinez 637 (Holotype: MEXU!; isotypes: CAS!, and others
to be distributed).
Frutex usque ad 1 m altus. Caules juniores
puberuli trichomatibus eglandulosis 0.05-0.2 mm
longis. Folia verticillata in quoque nodo quatuor
petiolata, lamina ovata vel ovato-elliptica 17-82
mm longa 9-27 mm lata 1.5-3-plo longior
quam latior margine plana. Inflorescentia spi-
cata, bracteae lanceolatae vel lanci-subulatae 2.5-
4 mm longae 0.9-1.1 mm latae apice attenuatae
erectae, bracteolae subulatae 2.3-3.5 mm longae
0.5-0.7 mm latae apice attenuatae erectae. Calyx
4-5 mm longus, corolla rosea 9.5-14.5 mm longa
extus pubescens, staminia 5.5-6.5 mm longa,
stylus 7-9.5 mm longus. Capsula 1 1-17 mm lon-
ga glabra. Semina 5-5.5 mm longa, 3.8-4.5 mm
lata.
Shrub to 2 m tall. Young stems evenly puber-
ulent with retrorse, eglandular trichomes 0.05-
0.2 mm long. Leaves whorled (four/node), pet-
iolate, petioles 2-17 mm long, blades charta-
ceous, ovate to ovate-elliptic, 6-90 mm long, 4-
42 mm wide, 1 .5-3 times longer than wide, acute
to acuminate at apex, acute to attenuate at base,
surfaces pubescent with straight to subflexuose
to antrorse, eglandular trichomes to 0.2 mm long,
margin entire to subundulate, flat. Inflorescence
of axillary spikes to 2.5 cm long, flowers opposite
at nodes, rachis evenly and densely pubescent
with erect to downward pointing, eglandular tri-
chomes 0.1-0.2 mm long. Bracts lanceolate to
lance-subulate, 2-4 mm long, 0.9-1.1 mm wide,
abaxial surface pubescent with erect to subflex-
uose or downward pointing, eglandular tri-
chomes to 0.2 mm long, apex attenuate, erect.
Bractlets subulate, 2-3.5 mm long, 0.5-0.7 mm
wide, shorter than or equal to bracts, pubescent
like bracts, apex attenuate, erect. Calyx 3-5 mm
long, lobes equal in length and slightly overlap-
ping one another, lanceolate, pubescent like
bracts, attenuate and erect at apex. Corolla green-
ish-white with pinkish or maroon stripes on low-
er lip, more or less vertical during anthesis, 9.5-
14.5 mm long, pubescent on external surface with
glandular and eglandular trichomes 0.05-0.2 mm
long, tube 2.5-5 mm long, abruptly expanded
into a throat 2.5-3.5 mm long, upper lip 3-5.5
mm long with triangular lobes 1-1.5 mm long,
lower lip 4.5-7 mm long with obovate lobes 2.5-
5 mm long. Stamens 5.5-6.5 mm long, filaments
pubescent, thecae 2-2.6 mm long, pollen (ob-
served in Diaphane) prolate, 32-39 ^m long (po-
lar diameter), 22-25 /im wide (equatorial di-
ameter), 1.3-1.8 times longer than wide, colpi
distinct; staminode 0.7-1.2 mm long, sparsely
80
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 2
FIGURE 5. Holographs leticiana. a, Habit (Daniel 5386), xQ.5; b, Inflorescence (Daniel 5386), x4; c, Corolla cut open to
show androecium (Torres C. 637), x4; d, Distal portion of style and stigma (Daniel 5386), x 16; e, Capsule (Torres C. 773),
x3.4; f, Seed (Torres C. 773), xg; g, Seminal trichome (Torres C. 773), x42.
pubescent at apex. Style 7-9.5 mm long, very
sparsely pubescent to glabrous; stigma indistinct-
ly bilobed, 0.1 mm long. Capsule 11-18 mm
long, glabrous. Seeds subrectangular to subellip-
tical in outline, 4.5-5.5 mm long, 3.5-4.5 mm
wide, surfaces covered with golden to brown,
dendritic trichomes to 0.4 mm long.
DISTRIBUTION AND PHENOLOGY. —Known only
from the vicinity of the Cerro Guiengola in
southeastern Oaxaca (Fig. 3), where the plants
DANIEL: NEW SPECIES OF HOLOGRAPHS
81
occur in tropical deciduous forest dominated by
legumes and cereoid cacti at elevations from 60
to 600 m. Flowering and fruiting: October-March.
Holographs leticiana can be distinguished from
all other species in the genus by the following
combination of characters: whorled leaves, pink-
ish corollas, and stamens 5.5-6.5 mm long. It is
most similar to H. parayana, a species occurring
approximately 200 km to the east in Chiapas.
The contrasting features of these two species are
summarized in the key above. Plants of this
species collected in November 1987 (Daniel 5386)
were locally common understory shrubs. Flowers
were visited by both bombyliid flies and mega-
chilid bees. However, pollen similar to that of
Holographis was not located on individuals of
either. Other Acanthaceae growing with H. le-
ticiana include Tetramerium oaxacanum T.
Daniel and species of Ruellia L. and Siphono-
glossa Oersted.
PARATYPES.— MEXICO. Oaxaca: along road to Ruinas
Guiengola, 1.3-1.6 km N of Mex. 190 between Jalapa de Mar-
ques and Tehuantepec, T. Daniel 5386 (CAS, DUKE, F, GH,
K, MICH, MEXU, MO, NY, TEX, US); 61.8 km NW of
Tehuantepec, P. Fryxell and E. Lott 3403 (CAS); Distr. Te-
huantepec, Ruinas del Cerro Guiengola, M. L. Torres C. et al.
511 (CAS, MEXU); Distr. Tehuantepec, ladera S del Cerro
Guiengola por la fabrica de cal, M. L. Torres C. et al. 773
(CAS, MEXU); Distr. Tehuantepec, hacia el N a 300 m del
"Mango," por la subida al Ocotal Cerro Guiengola, M. L.
Torres C. et al. 834 (CAS, MEXU).
ACKNOWLEDGMENTS
Funds for field studies were provided by NSF
grants BSR-8304790 and BSR-8609852. I am
most appreciative to the following persons for
their valuable assistance: B. Bartholomew, M.
Baker, and V. Lee (field assistance); M. L. Torres
C., E. Lott, and A. Sanders (making collections
available); C. Sudekum (illustrations); M. A.
Tenorio (SEM operation and photographic as-
sistance); S. Middleton (photography); and F. Al-
meda (selecting loan). I am also grateful to the
curators of the following herbaria for loans or
allowing me to study specimens in their respec-
tive institutions: ASU, CAS, ENCB, F, G, GH,
K, LL, MEXU, MO, P, POM, TEX, and US.
LITERATURE CITED
DANIEL, T. F. 1983. Systematics of Holographis (Acantha-
ceae). J. Arnold Arbor. 64:129-160.
. 1 984. New and reconsidered Mexican Acanthaceae.
Madrono 31:86-92.
. 1986. New and reconsidered Mexican Acanthaceae.
II. Southw. Naturalist 31:169-175.
CALIFORNIA ACADEMY OF SCIENCES
Golden Gate Park
San Francisco, California 941 18
PROCEEDINGS
OF THE
CALIFORNIA ACADEMY OF SCIENCES
Vol. 46, No. 3, pp. 83-94, 6 figs., 2 tables. December 7, 1988
THE EELPOUT GENUS PACHYCARA (TELEOSTEI: ZOARCIDAE)
IN THE NORTHEASTERN PACIFIC OCEAN, WITH
DESCRIPTIONS OF TWO NEW SPECIES
By
M. Eric Anderson
Department of Ichthyology, California Academy of Sciences,
Golden Gate Park, San Francisco, California 94118
and
Alex E. Peden
Vertebrate Unit, Royal British Columbia Museum,
Victoria, British Columbia V8V 1X4, Canada
ABSTRACT: The eelpout genus Pachycara Zugmayer, 1911, is recognized from the North Pacific Ocean for
the first time. Four species are included, two of which— P. gymninium and P. lepinium—aie new. Pachycara
gymninium differs from P. lepinium mainly in having a scaleless nape, the mediolateral lateral line originating
in the pectoral axil versus posterior to the pectoral margin, more numerous precaudal vertebrae, and shorter
pectoral and pelvic fins. The type species of Pachycara— P. obesum—from the abyssal Atlantic, is shown to
be synonymous with the earlier described Maynea bulbiceps Carman, 1899, from the eastern tropical Pacific.
Another species described by Carman (1899)—Phucocoetes suspectus—is placed in Pachycara and redescribed
from three specimens.
Received April 27, 1988. Accepted July 11, 1988.
Anderson (1988) included an Antarctic species,
The eelpout genus Pachycara was erected for first described as Lycodes brachycephalus by
a single specimen taken in the Bay of Biscay at Pappenheim (1912), in Pachycara because it
a depth of 4,780 m by the expeditions of Prince lacked the three apomorphic characters previ-
Albert I of Monaco (Zugmayer 191 la, b). A sec- ously established for Lycodes (Anderson 1984):
ond (juvenile) specimen of the type species, P. 1) possession of cartilaginous laminae along the
obesum, was described from off Virginia by Mar- ventral surface of the dentary bones ("mental
kle and Sedberry (1978). Earlier, Garman (1899) crests" of Andriashev 1 954); 2) loss of the pal-
described Maynea bulbiceps from the eastern atine membrane (oral valve); and 3) loss of head
tropical Pacific (Gulf of Panama). This species pores in adults (except for some paedomorphic
was placed in Pachycara as a putative Pacific species). Anderson (1984) included 11 species in
sister species of P. obesum by Anderson (1984). Pachycara, only three of which had been de-
[83]
84
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 3
FOB.
FIGURE 1. Pachycara bulbiceps (Garman), BCPM 980-99, 525 mm SL, off the Queen Charlotte Islands, British Columbia.
scribed previously, and a revision of the genus
is in progress.
The purpose of this paper is to describe in
advance of the senior author's revision two new
species of Pachycara from the northeastern Pa-
cific Ocean that have been known in collections
for many years. In addition, we redescribe the
type species, P. obesum, herein synonymized with
P. bulbiceps, now known from six North Atlantic
and 32 northeastern Pacific specimens, as well
as the rare P. suspectum, known from three spec-
imens.
METHODS
Measurements were made with an ocular mi-
crometer or dial calipers to the nearest 0.1 mm.
Definitions of characters and methodology fol-
lows Gosztonyi (1977), or, in cases of contradic-
tion, Peden and Anderson (1978) and Anderson
(1982, 1984). All specimens were measured and
radiographed to create Table 1, except 16 of 38
P. gymninium in SIO 67-121. Character state
modifiers are based on those numerically scored
characters of Anderson (1984). Museum abbre-
viations follow Leviton et al. (1985). Measure-
ments are in standard length (SL). Certain pro-
portions expressed in percent head length (HL)
orSL.
KEY TO NORTHEASTERN PACIFIC
PACHYCARA SPECIES
1A. Pelvic fins present __ 2
IB. Pelvic fins absent _
Pachycara bulbiceps (Garman, 1899)
2 A. Number of precaudal vertebrae 21-23;
lateral line with complete, ventral branch
only; head length 1 5.6-1 7.7% SL; anal fin
origin associated with vertebrae 20-22
Pachycara suspectum (Garman, 1899)
2B. Number of precaudal vertebrae 23-31
(usually 25 or more); lateral line with
complete mediolateral and ventral
branches; head length 12.0-15.9% SL;
anal fin origin associated with vertebrae
23-31 3
3A. Scales present on nape; mediolateral
branch of lateral line originating posterior
to pectoral fin margin; pectoral fin length
72.5-89.9% HL; pelvic fin length 11.5-
17.3% HL; number of precaudal verte-
brae 23-26 Pachycara lepinium n. sp.
3B. Scales absent on nape (rarely present in
small patches immediately anterior to
dorsal fin origin); mediolateral branch of
lateral line originating in pectoral axil;
pectoral fin length 63.0-70.8% HL; pelvic
fin length 4.8-1 1.3% HL; number of pre-
caudal vertebrae 28-3 1
_ _ Pachycara gymninium n. sp.
Pachycara bulbiceps (Garman, 1899)
(Figures 1-3)
Maynea bulbiceps Garman, 1899:140, 141, pi. E, fig. 1.
McAllister and Rees 1964:106, 107. Pearcy et al. 1982:387,
399, 400.
Pachycara obesa Zugmayer, 1 9 1 1 a: 1 2; 1 9 1 1 b: 1 34-1 36, pi. VI,
fig. 6. Markle and Sedberry 1978:22-25, fig. 1. Merrett and
Marshall 1981:240. Andriashev 1973:547.
Pachychara obesa Zugmayer (lapsus calami). Fowler 1936:
1056, fig. 47. Andriashev 1986:1 149, text fig.
MATERIAL EXAMINED.— Northeastern Atlantic: MOM, un-
cat. (holotype of Pachycara obesa); Bay of Biscay; 45°27'N,
06°05'W; 4,780 m; PRINCESSE ALICE sta. 1554; 7-9 Sept.
1903. BMNH 1981.6.16:15 (1); off Cap Blanc, West Africa;
20°17.2'N, 21°42.3'W; 4,002-4,007 m; DISCOVERY sta. 9131;
17Nov. 1976.IOS,uncat.(l);offSenegal; 18°08.6'N,20°11.7'W;
3,120 m; DISCOVERY sta. 10884; 12 Aug. 1983. IOS, uncat.
(1); Porcupine Sea Bight; 49°37.1'N, 13°48.5'W; 3,990-3,920
m; DISCOVERY sta. 51803; 26 Sept. 1983. Northwestern
Atlantic: USNM 21561 1 (1); off Virginia; 36°42.0'N, 74°04.5'W;
2,400-2,450 m; EASTWARD cr. 7506, sta. 20; 24-25 July
1975. MCZ 57612 (1); Middle Atlantic Bight; 38°44.0'N,
70°10.0'W; 3,028 m; OCEANUScr. 93, sta. 91 1; 3 1 Mar. 1981.
Gulf of Panama: MCZ 28681 (holotype of Maynea bulbiceps);
06°10'N, 83°06'W; 2,690 m; ALBATROSS sta. 3361; 25 Feb.
1890. UMML 22860 (2); 06°53-49'N, 79°27-28'W; 3,193-
3,201 m; JOHN E. PILLSBURY sta. P-526; 5 May 1967.
UMML 33484 (1);06°47-48'N, 79°13'-80°00'W; 3,04 5-3,218
m; JAMES M. GILLISS sta. GS-3; 14 Jan. 1972. UMML
33485 (2); 06°52'N, 79°28'W; 3,200-3,229 m; JAMES M. GIL-
LISS sta. GS-1; 13-14 Jan. 1972. SIO 81-148 (4); 05°09.8'N,
ANDERSON AND PEDEN: NORTH PACIFIC PACHYCARA
85
FIGURE 2. Pachycara bulbiceps (Garman), diagrammatic of CAS 55588, 338 mm SL, from the Gulf of Panama, showing
extent of lateral lines.
81°41.2'W; 3,900-4,000 m; MELVILLE; 3 Oct. 1981. SIO 85-
33 (4); 05°20.6'N, 81°55.5'W; 3,866 m; DSRV ALVIN dives
1361, 1363; 9-12 Apr. 1984. CAS 55588 (1); recat. from SIO
81-148. Middle American Trench: SIO 73-286 (3); off Costa
Rica; 09°43.8'N, 86°02.0'W; 2,601 m; ALEXANDER AGAS-
SIZ; 19-20 Apr. 1973. SIO 73-268(1); off Guatemala; 13°27.0'N,
91°59.2'W; 3,332 m; ALEXANDER AGASSIZ; 13-14 Apr.
1973. Mexico: SIO 66-35 (1); offGuadalupe Isl.; 29°01.0'N,
118°01.2'W; 3,310 m; ALEXANDER AGASSIZ; 5-6 Apr.
1966. Cascadia Abyssal Plain (off Oregon): OSUO 2559 (2);
45°27.8-27.0'N, 126°1 7.4-2 1.0'W; 2,606 m; YAQUINA sta.
CP-2C, BMT 157; 17 Jan. 1970. OSUO 2560 (2); 46°02.2'N,
1 26°33.5-28.5'W; 2,706 m; YAQUINA sta. CP-2A, BMT 257;
15 Feb. 1971. OSUO 2561 (2); 45°1 8.0-1 7.0'N, 126°26.4-
30.5'W; 2,710 m; YAQUINA sta. CP-2C, BMT 266; 18 Feb.
1971. OSUO 2562 (1); 44°43. l^t6.0'N, 1 27°29.3-28.6' W; 2,8 1 8
m; YAQUINA sta. CP-3E, BMT 3 1 7; 3 Feb. 1 973. CAS 6 1 220
(2); 44°39.7-39.0'N, 1 26°39.6-44.8' W; 2,8 1 6 m; CAYUSE sta.
CP-2E, BMT 253; 30 Sept. 1970. LACM 44316-1 (1); 45°20.6-
20.9'N, 126°35.7-39.7'W; 2,750 m; YAQUINA sta. CP-2C,
BMT 264; 17 Feb. 1971. LACM 44317-1 (1); 46°01.4-05.0'N,
1 27°3 1 .0-32.5' W; 2,835 m; YAQUINA sta. CP-3 A, BMT 322;
12 Mar. 1973. British Columbia: BCPM 980-99 (1); off Queen
Charlotte Isl., W of Tasu Sound; 2,780 m; EASTWARD HO
set 17; 26 Aug. 1979.
DIAGNOSIS.— A species of Pachycara as de-
fined by Anderson (1984) distinguished by the
following combination of characters: pelvic fins
FIGURE 3. Group of Pachycara sp. (presumably P. bulbiceps) near a baitfall in the Gulf of Panama (05°20'N, 8 1°55'W; depth,
3,850 m). Photographed by Craig R. Smith, University of Washington, DSRV ALVIN dive 1 140, 28 Sept. 1981.
86
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 3
TABLE 1 . COUNTS AND MEASUREMENTS OF NORTHEASTERN PACIFIC PACHYCARA SPECIES, INCLUDING ATLANTIC MATERIAL OF
P. BULBICEPS.
P. bulbiceps
(Atlantic)
P. bulbiceps
(Pacific)
P. gymninium
P. lepinium
P. suspectum
Measurements, in percent SL
Head length
12.3-15.0
11.8-14.8
12.0-15.6
13.8-15.9
15.6-17.7
Head width
5.7-8.5
5.5-10.1
5.7-8.2
6.5-11.8
8.0-8.9
Pectoral length
9.9-12.2
8.7-12.9
8.3-10.3
10.2-13.5
9.1-10.1
Predorsal length
17.8-18.7
16.0-19.8
15.9-21.5
15.8-23.2
15.9-18.2
Preanal length
34.9-10.1
35.1-40.0
38.5-44.1
34.3-41.9
35.8-36.0
Body height
6.4-12.7
5.7-11.6
7.4-13.6
8.3-16.9
8.4-9.9
Gill slit length
4.9-5.9
4.1-6.9
3.3-5.2
4.9-9.4
7.0 (in 1)
Caudal fin length
2.4-4.2
1.5-3.6
1.9-5.2
1.5-2.7
3.2 (in 1)
Measurements, in percent HL
Head width
38.8-73.1
43.5-80.5
37.3-64.1
44.5-75.3
49.7-52.7
Upper jaw length
33.9-13.2
37.1-52.7
33.3-52.8
35.9-52.0
38.4-41.4
Pectoral length
67.8-99.5
64.1-99.9
63.0-70.8
72.5-89.9
57.4-58.6
Snout length
19.4-22.7
16.8-26.2
15.8-25.5
17.6-25.9
18.5-22.3
Eye diameter
16.2-26.4
11.5-22.4
13.6-23.3
12.4-18.7
16.8-19.5
Gill slit length
32.9-50.9
31.4-49.8
21.0-43.1
33.9-59.8
39.7 (in 1)
Interorbital width
10.6-17.4
7.0-11.5
9.1-11.3
7.1-13.8
6.4-7.3
Interpupillary width
26.4-38.0
18.2-32.8
26.9-35.7
22.3^t4.4
24.2 (in 1)
Pelvic fin length
absent
absent
4.8-11.3
11.5-17.3
9.7-11.7
Counts
Vertebrae
25-31 +
27-31 +
28-31 +
23-26 +
21-23 +
82-90 =
83-91 =
73-80 =
80-94 =
84-87 =
113-118
112-119
102-109
105-120
105-110
Dorsal fin rays
104-109
105-114
96-103
99-113
100-106
Anal fin rays
86-94
88-97
77-84
85-98
85-89
Caudal fin rays
8-12
10-12
10-12
9-12
9
Pectoral fin rays
16-19
16-18
14-18
15-18
16
Pelvic fin rays
absent
absent
3
3
3
Gill rakers
2-3 +
0-4 +
3-5 +
2-5 +
1-2 +
11-13 =
11-15 =
10-14 =
11-15 =
14-15 =
13-16
11-19
13-19
12-19
16
Vomerine teeth
3-6
2-9
3-8
5-18
3-6
Palatine teeth
2-7
4-10
3-14
7^10
9-12
absent; mediolateral branch of lateral line orig-
inating posterior to pectoral fin margin; scales
absent on nape; vertebrae 25-3 1 + 82-9 1 = 112-
1 19; head length 1 1.8-15.0% SL; anal fin origin
associated with vertebrae 23-30.
DESCRIPTION. — Counts and measurements
presented in Table 1 were compiled from all
known specimens, 82-525 mm SL (no gravid
females). The following description is based on
10 adult males, eight adult females, 10 juveniles
of both sexes, and seven eviscerated specimens.
Head large, ovoid, wider in adults than juveniles.
Smallest specimen (82 mm SL) with most dor-
soventrally depressed head, resulting in more
acute snout than adults. Body short, deep, broad-
er in cross section in adults than juveniles. Body
more slender than similarly sized P. gymninium
and P. lepinium; body height at anal fin origin
of smallest specimen (5.7% SL; Table 1) mea-
suring like slender genus Lycenchelys, however
specimens >160 mm with body height >7.8%
SL. Tail laterally compressed, more so poste-
riorly, tapering gradually to tip. Skin firm, not
gelatinous, but pliable around head, especially
near mouth, cheeks, and nape due to subdermal
lipid layer. Scales minute, cycloid, imbedded,
covering body, tail, and vertical fins to one-half
to three-fourths their height; no scales on head
or nape; scales present on pectoral axil and base,
and on breast to isthmus in largest specimens.
No scales developed in smallest specimen. Eye
circular, larger in proportion to head in smallest
specimens than largest adults. Gill slit short, usu-
ally extending ventrally to lower margin of pec-
ANDERSON AND PEDEN: NORTH PACIFIC PACHYCARA
87
toral base. Very slight, squared-off lobe formed
at dorsal margin of operculum, this best devel-
oped in small juveniles; gill slit of large adults
often without distinct lobe. Single pair of nostrils
at snout tip, nasal tube not reaching upper lip.
Pectoral fin large, its origin at or just below body
midline; posterior margin somewhat wedge-
shaped to almost evenly rounded; lowermost 6-
9 rays shorter and thicker than others, their tips
not at all or slightly exserted; pectoral rays usu-
ally 17-1 8 (Table 2).
Mouth terminal, upper jaw slightly longer in
relation to head in adult males than comparably
sized females; upper jaw 45.0-52.7% HL (X =
48.8%, n = 10) in males over 330 mm SL, 41.8-
44.6% HL (x = 43.1%, n = 5) in females over
330 mm SL. Minute epidermal prickles on lips
and occasionally in patches on snout. Palatine
membrane (oral valve) moderately developed in
adults, not reaching vomer, larger in juveniles,
reaching anterior edge of vomer. Teeth in jaws
and palate small, conical, dentition not sexually
dimorphic. Jaw teeth in single row in smallest
specimen; premaxillae of adults with 2-3 rows
anteriorly, blending into single, posterior row;
dentary with 3-5 rows anteriorly, blending into
single or double posterior row. Vomerine teeth
in irregular patch; palatine teeth in single row.
Cephalic lateralis system reduced, numerical
variation in pores of postorbital and suborbital
canals only. Postorbital pores one, three and four
present on one side, pore one only on the other
in one specimen; pores one and four present in
21 specimens. Two pairs of anterior supraorbital
(nasal) pores, one set mesial to nasal tube, the
other posteromesially. Usually six suborbital
pores, all emanating from ventral ramus of sub-
orbital bone chain, but a seventh pore emanating
from ascending ramus behind eye, just below first
postorbital pore on one side in one specimen.
Eight (not seven as reported by Markle and Sed-
berry 1978) preoperculomandibular pores, four
emanating from dentary, one from anguloartic-
ular, and three from preopercle. Interorbital and
occipital (supratemporal) pores absent. Body lat-
eral line of two branches: mediolateral, origi-
nating at or just posterior to pectoral fin margin,
complete to tail tip (not evident in some mate-
rial), and ventral, originating just posterior to
fourth postorbital pore, coursing gradually across
body to just above anus, then running just above
anal fin to tail tip (Fig. 2).
TABLE 2. PECTORAL FIN RAY COUNTS OF NORTHEASTERN
PACIFIC PACHYCARA SPECIES.
Fin ray number
Species
14
15
16
17
18
19
P. gymninium, n = 30
3
12
11
3
1
0
P. lepinium, n =
30
0
1
3
20
6
0
P. bulbiceps, n =
34
0
0
4
20
9
1*
P. suspectus, n =
3
0
0
3
0
0
0
* USNM 215611, recounted after Markle and Sedberry (1978)
who reported it as 17.
Vertebrae symmetrical, no geographical vari-
ation exhibited numerically. Caudal fin rays vari-
able, with two epural, 4-6 upper hypural, and 3-
5 lower hypural rays. Last anal ray associated
with second preural vertebra, last dorsal ray as-
sociated with third or fourth preural. Dorsal fin
origin associated with vertebrae 4-7, with no free
pterygiophores. All dorsal elements soft rays, first
one not segmented, but bilaterally divided at least
near base. Anal fin origin associated with ante-
penultimate to ultimate precaudal vertebrae, with
3-9 anal fin pterygiophores inserted anterior to
haemal spine of first caudal vertebra.
Gill rakers short, roughly triangular in adults,
more slender and longer in juveniles; adults with
3-4 epibranchial rakers and usually 12-14 on
ceratobranchial; raker denticles absent. Bran-
chiostegal rays six: four articulating with cera-
tohyal and two with epihyal. Two nublike pyloric
caeca, not longer in smallest specimens than larg-
est, as in some zoarcids. Pseudobranch filaments
relatively long, about one eye diameter in length
in adults (less in smallest specimens), numbering
3-7.
Fresh specimens uniformly dark brown, head
and most of pectoral fins darker, but not black.
Pectoral base pale in some preserved specimens,
with irregular gray mottling in IOS DISCOV-
ERY sta. 51803. Dorsal fin of this juvenile and
a large adult (IOS DISCOVERY sta. 10884) with
thin white bands over black background mostly
at mid-body and near tail tip; anal fin with fewer
white bands. Orobranchial chamber pale to dark
brown. Abdomen and eyes dark blue in freshly
preserved specimens.
DISTRIBUTION.— Presently known from both
sides of the North Atlantic at depths of 2,400-
4,780 m, and off the Queen Charlotte Islands,
British Columbia, to the Gulf of Panama, North
Pacific, at depths of 2,601-4,000 m over brown
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 3
and green mud bottoms, where the species may
swarm near food falls or baited traps (Fig. 3).
Pachycara suspectum (Garman, 1899)
(Figure 4)
Phucocoetes suspectus Garman, 1899:137, pi. XXX, fig. 3, 3a.
MATERIAL EXAMINED. — MCZ 28683 (holotype); offAcapul-
co, Mexico; 16°33.0'N,99°52.5'W; 1 ,207 m; ALBATROSS sta.
3418; 11 Apr. 1890. SIO 68-106 (1); Gulf of California;
25°56.6'N, 110°37.5'W; 1,280 m; 22 Jan. 1968. ZMUC
P-761085 (1); Gulf of Panama; 07°22'N, 79°32'W; 915-975
m; GALATHEA sta. 739; 15 May 1952.
DIAGNOSIS.— A species of Pachycara as de-
fined by Anderson (1984) distinguished by the
following combination of characters: pelvic fins
present; lateral line with ventral branch only;
scales absent on nape; vertebrae 21-23 + 84-87
= 105-1 10; head length 15.6-17.7% SL; anal fin
origin associated with vertebrae 20-22.
DESCRIPTION. — Counts and measurements
presented in Table 1 were compiled from all three
specimens (females), 218-280 mm SL; holotype
badly damaged. Head large, ovoid, deepest at
occiput, dorsoventrally depressed. Snout acute,
steeply sloping anteriorly. Body short, deep, ovoid
in cross section. Skin firm, not gelatinous, but
pliable around mouth and cheeks due to
subdermal lipid layer. Scales minute, cycloid,
imbedded, sparse anteriorly, covering body and
tail, absent on unpaired fins, head, nape, and
pectoral base. Eye ovoid, entering dorsal profile
of head when viewed laterally. Gill slit long, ex-
tending ventrally to near pelvic base. Slight,
squared-off lobe formed at dorsal margin of oper-
culum. Single pair of nostrils at snout tip, nasal
tube not reaching upper lip. Pectoral fin large,
its origin well below body midline; posterior
margin wedge-shaped; lowermost 6-7 rays short-
er and thicker than others, their tips very slightly
exserted.
Mouth terminal, upper jaw extending poste-
riorly to vertical through middle of eye. No epi-
dermal prickles on lips or snout. Palatine mem-
brane (oral valve) moderately developed, reaching
anterior margin of vomer. Teeth in jaws and pal-
ate small, conical. Jaw teeth in two rows ante-
riorly, blending into single posterior row in all
three specimens. Vomerine teeth in irregular
patch; palatine teeth in single or double (ZMUC
P-761085) row.
Cephalic lateralis system reduced, numerical
variation not detected (two specimens damaged).
Postorbital pores one, three, and four present in
SIO 68-106. Two pairs of anterior supraorbital
(nasal) pores, one set mesial to nasal tube, the
other dorsally. Seven suborbital bones present
from which emanate six pores from ventral ra-
mus of chain and one from ascending ramus be-
hind eye, just below first postorbital pore. Eight
preoperculomandibular pores, four emanating
from dentary, one from anguloarticular, and three
from preopercle. Interorbital and occipital pores
absent. Body lateral line of single, complete, ven-
tral branch only; lateral line originating just pos-
terior to fourth postorbital pore, coursing steeply
downward on anterior portion of abdomen, then
running just above anal fin to tail tip (Garman
1899, pi. XXX).
Vertebrae symmetrical. Caudal fin with one
epural and four upper and lower hypural rays (in
two). Last anal ray associated with second pre-
ural vertebra, last dorsal ray associated with
fourth preural vertebrae. Dorsal fin origin asso-
ciated with vertebrae 2-3, with no free pteryg-
iophores. All dorsal elements soft rays, first one
not segmented, but bilaterally divided. Anal fin
origin associated with penultimate precaudal
vertebra, with 2-4 anal fin pterygiophores in-
serted anterior to haemal spine of first caudal
vertebra.
Gill rakers short, roughly triangular, ventral-
most blunt, but dorsalmost 5-6 with sharp tips;
raker denticles absent. Branchiostegal rays six:
four articulating with ceratohyal and two with
epihyal. Two nublike pyloric caeca. Pseudo-
branch filaments short, numbering 3-4.
Coloration of fresh specimens not confirmed,
although Garman (1899) reported holotype as
"blackish." Recent material greatly faded. One
gravid female, 218 mm SL (SIO 68-106), with
ova averaging 4.4 mm in diameter.
DISTRIBUTION.— Currently known from the
Gulf of California south to the Gulf of Panama
at depths of 9 1 5-1 ,280 m over green and brown
mud bottoms.
Pachycara gymninium n. sp.
(Figure 5)
Lycodes sp. Hubbs et al., 1979:14.
Lycenchelys "E." Pearcy et al., 1982:387.
HOLOTYPE. -USNM 280121 (female, 422 mm SL); British
Columbia, W of Tasu Sound, Queen Charlotte Islands;
52°38.0'N, 132°05.8'W; trap, 2,744 m; TALAPUS set 18; 4
Feb. 1980.
ANDERSON AND PEDEN: NORTH PACIFIC PACHYCARA
89
FIGURE 4. Pachycara suspectum (Garman), SIO 68-106,
218 mm SL, from the Gulf of California.
PARATYPES.— British Columbia: BCPM 980-121 (1); same
as holotype. BCPM 979-11365 (2); W of Tasu Sound; trap,
2,744 m; EASTWARD HO set 5; 16 Aug. 1979. BCPM 979-
1 1 369 (6); W of Tasu Sound; trap, 2,889 m; EASTWARD HO
set 12; 22 Aug. 1979. BCPM 980-100 (1); W of Tasu Sound;
trap, 2,926 m; EASTWARD HO set 18; Aug. 1979. USNM
221 1 13 (1); recat. from BCPM 979-1 1369. NMC 86-0444 (1);
W of Tasu Sound; trap, 2,743-2,926 m; EASTWARD HO set
11; 21 Aug. 1979. Cascadia Abyssal Plain: CAS 60524 (1);
45C18. 7-18. 5'N, 126°34.4-28.6'W; trawl, 2,750 m; YA-
QUINA sta. CP-2C, BMT 265; 18 Feb. 1971. CAS 60525 (1);
45-21. 5-21. 7'N, 127°33.0-38.5'W; trawl, 2,800 m; YA-
QUINA sta. CP-3C, BMT 280; 1 8 May 1971. California: OSUO
2558 (1); Mendocino Ridge; 40°45.8'N, 127°31.6'W; trap, 3,225
m; USGS S. P. LEE sta. L2-86-5T; 20-21 July 1986. SIO 67-
121 (36); SW of Cordell Bank; 37°57.4'N, 123°40.5'W; trap,
2,707-3,2 19m; THOMAS WASHINGTON; 16-1 7 June 1967.
Mexico: SIO 65-69 (3); off Guadalupe Isl., 4 mi E of southern
tip; trap, 1,829 m; 4-5 Mar. 1965. SIO 68-120 (6); Gulf of
California; 25°35.4'N, 109°46.0'W; trap, over 1,728 m;
THOMAS WASHINGTON; 24-25 Jan. 1968. SIO 70-12 (3);
offBaja California Norte; 31°10.0-06.9'N, 1 18°52.6-53.4'W;
trap, 2,524 m; MELVILLE; 16-17 Dec. 1969.
ADDITIONAL MATERIAL. — SIO 66-19 (2); off Guadalupe Isl.,
Mexico; 29°07.8'N, 118'12.5'W; trap, 2,378 m; ALEX-
ANDER AGASSIZ; 1-2 Apr. 1966.
DIAGNOSIS.— A species of Pachycara as de-
fined by Anderson (1984) distinguished by the
following combination of characters: pelvic fins
present, their lengths 4.8-1 1.3% HL; mediolat-
eral branch of lateral line originating in pectoral
axil just posterior to vertical through pectoral
base; scales absent on nape, or, if present, not
extending anterior to line connecting anterodor-
sal edges of gill slits; pectoral fin length 63.0-
70.8% HL; vertebrae 28-31 + 73-80= 102-109;
head length 12.0-15.6% SL; anal fin origin as-
sociated with vertebrae 27-3 1 .
DESCRIPTION. — Counts and measurements
presented in Table 1 were compiled from all
known specimens, 76-422 mm SL (including one
gravid female). The following description is based
on 10 adult males, nine adult females, 42 juve-
niles of both sexes, and two eviscerated adults.
Head large, ovoid, wider in adults than juveniles.
Body short, deep, broader in cross section in
adults than juveniles. Tail laterally compressed,
more so posteriorly, tapering gradually to tip.
Skin firm, not gelatinous, but pliable around head,
especially near mouth, cheeks and nape, owing
to subdermal lipid layer. Scales minute, cycloid,
imbedded, covering body, tail and vertical fins
to about half their height; usually no scales on
nape, but when present, restricted to small patches
immediately anterior to dorsal fin origin, or,
rarely, extending anteriorly to line connecting
anterodorsal edges of gill slits. Scales usually ab-
sent on pectoral base, when present, usually in
irregular, ovoid patch not extending from dorsal
to ventral margin of fin and not extending onto
fin more than 1 5% of its length. Scales on right
cheek in small patch between preoperculoman-
dibular pores seven and eight in one specimen
(SIO 67-121, 235 mm SL female), scales other-
wise absent on head. No scales developed in
smallest specimen, 76 mm SL. Eye circular, larg-
er in proportion to head in small specimens than
largest adults. Gill slit somewhat shortened, usu-
ally extending ventrally to just above ventral
margin of pectoral base, but restricted to mid-
height of pectoral base in a few specimens, and
extending to below ventral margin of pectoral
base in a few others. Slight, rounded or squared-
TL.B.
FIGURE 5. Pachycara gymninium n. sp., USNM 280121 (holotype, 422 mm SL), off the Queen Charlotte Islands, British
Columbia.
90
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 3
FIGURE 6. Pachycara lepinium n. sp., USNM 280120 (holotype, 465 mm SL), off the Queen Charlotte Islands, British
Columbia.
off, posteriorly-directed lobe formed at dorsal
margin of operculum, its dorsalmost margin usu-
ally straight, but deeply concave in one speci-
men. Single pair of nostrils at snout tip, nasal
tube very short, usually not reaching upper lip
(reaching upper lip only in few juveniles less than
220 mm SL). Pectoral fin large, its origin just
below body midline; posterior margin nearly
evenly rounded (except in few injured speci-
mens); ventralmost 3-5 rays shorter and thicker
than others, their tips not at all, or just slightly
exserted; pectoral rays usually 15-16 (Table 2).
Mouth terminal, upper jaw slightly longer in
relation to head in adult males than comparably
sized adult females; upper jaw 40.5-52.8% HL
(je = 43.9%, n = 12) in males over 280 mm SL,
37.7-41.8% HL (Jc = 39.5%, n = 1 1) in females
over 295 mm SL. Minute, epidermal prickles on
anterior portions of lips and chin, these often
nearly rubbed off in large specimens. Palatine
membrane (oral valve) moderately developed,
almost reaching anterior margin of vomer in most
specimens. Teeth in jaws and palate small, con-
ical, dentition not sexually dimorphic. Jaw teeth
in single row anteriorly in smallest specimens,
in 4-5 anterior rows in adults, blending into sin-
gle posterior row in all. Vomerine teeth in irreg-
ular patch or arched row; palatine teeth in single
row.
Cephalic lateralis pore system reduced, nu-
merical variation in pores of postorbital and sub-
orbital canals only. Postorbital pores one, three,
and four present on both sides in one specimen;
pores one and four only present in 1 8; pores one
and four on one side, pores one, three, and four
on the other in nine specimens. Two pairs of
nasal pores, one set mesial to nasal tube, the
other posteromesially. Usually six suborbital
pores, all emanating from ventral ramus of sub-
orbital bone chain, but a seventh pore emanating
from ascending ramus behind eye just below first
postorbital pore in seven specimens, and, in one,
from just above sixth pore on cheek. Eight pre-
operculomandibular pores, four emanating from
dentary, one from anguloarticular, and three from
preopercle. Interorbital and occipital (supratem-
poral) pores absent. Body lateral line of two
branches: mediolateral, originating in pectoral
axil on vertical through pectoral base, complete
to tail tip, and ventral, originating just posterior
to fourth postorbital pore, coursing gradually
across body to just above anus, then running to
tail tip above anal fin base.
Vertebrae symmetrical, no geographical vari-
ation exhibited numerically. Caudal fin rays vari-
able, with one or two epural, 4-6 upper hypural,
and 4-5 lower hypural rays. Last anal ray asso-
ciated with second preural vertebra, last dorsal
ray associated with fourth preural vertebra. Dor-
sal fin origin associated with vertebrae 3-5, with
no free pterygiophores. All dorsal elements soft
rays, first one not segmented, but bilaterally di-
vided at least near base. Anal fin origin associ-
ated with antepenultimate to ultimate precaudal
vertebrae, with 3-5 pterygiophores inserted an-
terior to haemal spine of first caudal vertebra.
Gill rakers short, roughly triangular in adults,
more slender, but not pointed, in smallest spec-
imens; adults with 3-5 epibranchial rakers and
usually 1 1-1 3 on ceratobranchial; raker denticles
absent. Branchiostegal rays six: four articulating
with ceratohyal and two with epihyal. Two nub-
like pyloric caeca, not appreciably longer in
smallest specimens than largest. Pseudobranch
filaments relatively long, about one eye diameter
in length (or more in smallest specimens), num-
bering 3-5.
Fresh specimens uniformly dark brown, head
and pectoral fin darker, almost black. Lining of
orobranchial chamber dark brown or black. Mar-
gins of vertical fins and peritoneum black.
One gravid female known (CAS 60524, 350
mm SL), with 80 ova 5.2-7.2 mm in diameter
(X = 5.8 mm). Stomach contents of several spec-
ANDERSON AND PEDEN: NORTH PACIFIC PACHYCARA
91
imens consisted of amphipods, isopods, poly-
chaetes, and trap bait.
DISTRIBUTION.— Off the Queen Charlotte Is-
lands, British Columbia, south to ofFGuadalupe
Island, Mexico, and in the Gulf of California in
1 ,829-3,2 1 9 m over brown and green mud bot-
toms.
ETYMOLOGY.— From the Greek yv^vba (naked)
and Iviov (nape) in reference to the species' scale-
less head.
Pachycara lepinium n. sp.
(Figure 6)
Lycodes sp. Hubbs et al., 1979:14.
Lycenchelys "D." Pearcy et ah, 1982:387.
HOLOTYPE.-USNM 280120 (male, 465 mm SL); British
Columbia, W of Tasu Sound, Queen Charlotte Islands;
52°38.0'N, 1 32°05.8'W; trap, 2,744 m; TALAPUS set 8; 4 Feb.
1980.
PARATYPES.- British Columbia: BCPM 980-121 (1); same
as holotype. BCPM 980-98 (1); W of Tasu Sound; trap, 2,889
m; EASTWARD HO set 6; Aug. 1979. NMC 86-0445 (1); W
of Tasu Sound; trap, 2,744 m; EASTWARD HO set 5; 16
Aug. 1979. USNM 221 1 14 (1); W of Tasu Sound; trap, 2,889
m; EASTWARD HO set 12; 22 Aug. 1979. Cascadia Abyssal
Plain: CAS 60526 (1); 44°23.0-25.4' N, 125°34.0-35.0'W; trawl,
2,878m; YAQUINA sta. CP- 1 F, BMT 4 1 0; 4 Aug. 1 974. CAS
60527 (1); 45°46.9^5.7'N, 126°35.5'W; trawl, 2,665 m; YA-
QUINA sta. CP-2A, BMT 259; 15 Feb. 1971. CAS 60528 (1);
44°41. 7-43. 6'N, 125°33.5-32.5'W; trawl, 2,825 m; YA-
QUINA sta. CP-1E, BMT 120; 6 Oct. 1969. CAS 60529 (1);
45°53.1-51.8'N, 126°39.0-42.5'W; trawl, 2,713 m; YA-
QUINA sta. CP-2A, BMT 154; 16 Jan. 1970. CAS 60530 (1);
44°29.6-45.2'N, 126°29.0-35.9'W; trawl, 2,818 m; YA-
QUINA sta. CP-2D, OTB 90; 23 Oct. 1965. OSUO 2563 (1);
45°09.3-10.5'N, 125°38.3-38.0'W; trawl, 2,669 m; YA-
QUINA sta. CP- 1C, BMT 93; 15-16July 1 969. LACM 443 15-1
(1); 44°39.7-39.0'N, 126°39.6-44.8'W; trawl, 2,816 m; CA-
YUSE sta. CP-2E, BMT 253; 30 Sept. 1970. California: SIO
66-50 (1); off Cape Mendocino; 40°34.6-34.7'N, 125°51.4-
51.5'W; trap, 2,970 m; ALEXANDER AGASSIZ; 21-22 May
1966. SIO 60-475 (1); off Farallon Isl.; trawl, ca. 1,800 m; 10
Nov. 1960. SIO 59-288 (1); San Clemente Basin; 32°39.0'N,
1 18°09.3'W; trap, 2,027 m; 23-24 July 1959. SIO 59-289 (3);
San Clemente Basin; 32°39.0'N, 118°09.3'W; trap, 2,023 m;
23-24 July 1959. SIO 65-452 (4); San Clemente Basin;
32°38.7'N, 1 18°08.4'W; trap, 2,008 m; ALEXANDER AGAS-
SIZ; 25 Sept. 1965. LACM 321 14-2 (1); San Clemente Basin;
32°38.0-37.5'N, 1 18°09.0-06.4'W; trap, 1,829-1,939 m; VE-
LERO IV sta. 12521; 10 Dec. 1968. Mexico: SIO 59-365 (5);
off Cabo Colnet; 31°02.7'N, 116°59.0'W; trap, 2,140 m; 27-
28 Oct. 1959.
ADDITIONAL MATERIAL. -SIO 59-364 (1); off Cabo Colnet,
Mexico; 31°00.5'N, 1 18°06.0'W; trap, 1,728 m; 27 Oct. 1959.
SIO 59-366 (1); off Cabo Colnet, Mexico; 31°02.7'N,
116°59.3'W; trap, 2,140 m; 27-28 Oct. 1959.
DIAGNOSIS.— A species of Pachycara as de-
nned by Anderson (1984) distinguished by the
following combination of characters: pelvic fins
present, their length 1 1.5-17.3% HL; mediolat-
eral branch of lateral line originating posterior
to pectoral fin margin; scales present on nape,
extending to interorbital region; pectoral fin length
72. 5-89.9% HL; vertebrae 23-26 + 80-94= 105-
120; head length 13.8-15.9% SL; anal fin origin
associated with vertebrae 23-26.
DESCRIPTION. — Counts and measurements
presented in Table 1 were compiled from all
known specimens, 221-597 mm SL (including
one gravid female). The following description is
based on six adult males, eight adult females,
and 1 6 juveniles of both sexes. Head large, ovoid,
wider in adults than juveniles. Body relatively
short, deep, broader in cross section in adults
than juveniles. Tail laterally compressed, more
so posteriorly, tapering gradually to tip. Skin firm,
not gelatinous, but somewhat pliable around
head, especially near mouth and cheeks, due to
subdermal lipid layer. Scales minute, cycloid,
imbedded, covering body, tail, and vertical fins
almost to their margins; scales on nape extend
anteriorly to vertical one eye diameter anterior
to dorsalmost edge of gill opening in smallest
specimen, to interorbital region in adults. Scales
present in pectoral axil and on base, extending
on outer surface of fin to half or more its length;
scales present on cheeks. Eye circular, relatively
larger in proportion to head in smallest speci-
mens than largest. Gill slit long, extending ven-
trally almost to pelvic fin origin; slight, rounded,
posteriorly-directed lobe formed at dorsal mar-
gin of operculum. Single pair of nostrils at snout
tip, nasal tube very short, not reaching upper lip.
Pectoral fin large, its origin just below body mid-
line; posterior margin almost evenly rounded;
lowermost 5-6 rays shorter and thicker than oth-
ers, their tips somewhat excised; pectoral rays
usually 17-18 (Table 2).
Mouth subterminal in largest adults, terminal
in young; upper jaw longer in relation to head in
adult males than comparably sized adult females;
upper jaw 48.8-52.0% HL (X = 50.3, n = 5) in
males over 380 mm SL, 41.7-46.4% HL (X =
44.8%, n = 7) in females over 380 mm SL. Mi-
nute, epidermal prickles on anterior portions of
lips and chin. Palatine membrane (oral valve)
well developed, reaching anterior edge of vomer.
Teeth in jaws and palate small, conical; dentition
not sexually dimorphic. Jaw teeth in two irreg-
ular rows anteriorly in smallest specimens, in 4-
5 anterior rows in adults, blending into single
posterior row in all. Vomerine teeth in irregular
92
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 3
patch; palatine teeth in one or two (largest spec-
imens) irregular rows.
Cephalic lateralis pore system reduced, nu-
merical variation present only in pores of post-
orbital canal. Postorbital pores one, three, and
four present, emanating from frontal, pterotic
and lateral extrascapular, respectively; pores one
and four only in two specimens; pores one and
four on one side, one, three, and four on the other
in two other specimens. Two pairs of nasal pores,
one set mesial to nasal tube, the other set pos-
teromesially. Six suborbital pores, all emanating
from ventral ramus of suborbital bone chain.
Eight preoperculomandibular pores, four ema-
nating from dentary, one from anguloarticular,
and three from preopercle. Interorbital and oc-
cipital (supratemporal) pores absent. Body lat-
eral line of two branches: mediolateral, origi-
nating at or just posterior to pectoral fin margin,
complete to tail tip, and ventral, originating just
posterior to fourth postorbital pore, coursing
gradually across body to just above anus, then
running just above anal fin to tail tip.
Vertebrae symmetrical. Individuals from ba-
sins of southern California and northern Baja
California with significantly more vertebrae than
those from the north (specimens from British
Columbia to Cape Mendocino with 105-110
vertebrae [n = 13, * = 107.2, SD = 1.58], spec-
imens from the Farallon Islands to Baja Cali-
fornia with 1 1 6-1 20 vertebrae [n = 1 6, x = 1 1 7.2,
SD = 0.49]). Caudal fin rays variable, with two
epural, 3-5 upper hypural and 3-5 lower hypural
rays. Last anal ray associated with second preural
vertebra, last dorsal ray associated with third or
fourth preural vertebrae. Dorsal fin origin asso-
ciated with vertebrae 4-5, with no free pteryg-
iophores. All dorsal fin elements soft rays, first
1-2 not segmented, but bilaterally divided at least
near base. Anal fin origin associated with ante-
penultimate to ultimate precaudal vertebra, with
2-5 pterygiophores inserted anterior to haemal
spine of first caudal vertebra.
Gill rakers short, roughly triangular, relatively
longer on dorsal portion of ceratobranchial; larg-
est adults with four epibranchial rakers and 1 3-
15 on ceratobranchial; raker denticles absent.
Branchiostegal rays six: four articulating with
ceratohyal and two with epihyal. Two nublike
pyloric caeca, not appreciably longer in smallest
specimens. Pseudobranch filaments relatively
long, about one eye diameter in length in both
adults and juveniles, numbering 2-5.
Fresh specimens uniformly dark brown, head
and pectoral fin darker, almost black. Lining of
orobranchial chamber dark brown or black. Mar-
gins of vertical fins and peritoneum black.
One gravid female, 412 mm SL (SIO 59-364,
ovary somewhat damaged), had ova averaging
4.3 mm in diameter. Stomach contents of several
specimens included polychaetes, bivalves, and
amphipods.
DISTRIBUTION.— Off the Queen Charlotte Is-
lands, British Columbia, south to ofTGuadalupe
Island, Mexico, at depths of 1 ,728-2,970 m over
brown and green mud bottoms. Often taken in
traps and trawls with P. gymninium, but this
species not yet known from the Gulf of Califor-
nia.
ETYMOLOGY.— From the Greek XCTTIS (scale) and
Iviov (nape) in reference to the species' scaly head.
DISCUSSION
The distribution of P. bulbiceps may at first
glance appear anomalous, in that the species is
widely distributed in the North Atlantic, but oth-
erwise restricted to the northeastern Pacific con-
tinental rise and adjacent abyss. Anderson (1988)
provided evidence that the pan-American, trop-
ical, abyssal fish fauna was sundered by tectonic
uplifting in the region of the Panamanian isth-
mus at least by the mid-Miocene. The present
Caribbean deep-water ichthyofauna is very dif-
ferent from that of the eastern tropical Pacific
(Garman 1899; Anderson et al. 1986). Thus, the
Pacific and Atlantic stocks of P. bulbiceps would
have had ample time to speciate if there were no
South Atlantic-South Pacific populations linking
the northern groups. Because abyssal trawl sam-
ples from the southwestern Atlantic and south-
eastern Pacific are generally lacking for most re-
gions, isolation of the northern populations
cannot be evaluated. Based on the paleogeog-
raphy of the American tropics cited by Anderson
(1988), the Atlantic and Pacific populations
probably are not isolated from each other, and
future abyssal collections off Chile, Argentina,
Brazil, and probably southern Africa should pro-
duce specimens of P. bulbiceps. Among the char-
acters presented in Table 1 , as well as traditional
data analysis used by us before (Peden and An-
derson 1978; Peden 1981; Anderson 1982), in-
cluding features of the axial skeleton, gill arches,
dentition, squamation, and lateral lines, no nu-
merical or morphological differences were de-
ANDERSON AND PEDEN: NORTH PACIFIC PACHYCARA
93
tected between the Atlantic and Pacific popula-
tions, thus isolation seems improbable. This is
in contrast to North Atlantic and North Pacific
populations of the grenadier Coryphaenoides ar-
matus, which exhibited minor electrophoretic but
significant biometric differences, resulting in
Wilson and Waples' (1984) placement of these
forms as separate subspecies. Similarly, Small
(1981) found specific differences in the cosmo-
politan bathyal genus Antimora and recognized
an endemic North Pacific species (A. microlepis),
while a second species (A. rostratd) was shown
to be circumglobal but excluded from the North
Pacific.
The geographic variation noted for P. lepi-
nium, but not found in P. gymninium, is difficult
to explain, but may relate to differing environ-
mental conditions that affect vertebral number
in the basins of the Southern California Bight.
Still, no variation was noted in the eight indi-
viduals of P. gymninium from this area, or in P.
bulbiceps along its entire range. If the distribu-
tions of these species are broader than present
samples reflect (and they probably are), signifi-
cant geographic variation may be found when
adequate material is analyzed along the ranges
of all species, as was found for the eelpouts Ly-
codapus mandibularis (Peden 1979) and Lycodes
diapterus (Peden 1981). The lack of variation
noted for P. gymninium may be due to either the
present poor sample size or is a valid specific
character, as found in another eelpout genus,
Gymnelus (Anderson 1982).
ACKNOWLEDGMENTS
For assistance with specimen loans and data
management we thank Karsten Hartel, Susan
Jewett, Shelly Johnson, Robert Lavenberg, Nigel
Merrett, William Pearcy, Richard Rosenblatt,
Jeffrey Seigel, and David Stein. Ray Scarsbrook
collected the British Columbia material, includ-
ing holotypes of both new species and generously
donated them to the British Columbia Provincial
Museum. Patricia Drukker-Brammall rendered
Figures 1, 5, and 6. Craig R. Smith generously
provided Figure 3.
LITERATURE CITED
ANDERSON, M. E. 1982. Revision of the fish genera Gymnelus
Reinhardt and Gymnelopsis Soldatov (Zoarcidae), with two
new species and comparative osteology of Gymnelus viridis.
Natl. Mus. Nat. Sci., Publ. Zool. (17): 1-76.
. 1 984. On the anatomy and phytogeny of the Zoar-
cidae (Teleostei: Perciformes). Ph.D. Dissertation, College
of William and Mary, Williamsburg, Virginia. 254 pp.
. 1988. Studies on the Zoarcidae (Teleostei: Perci-
formes) of the southern hemisphere. I. The Antarctic and
subantarctic regions. Antarct. Res. Sen, Amer. Geophys.
Union, Biol. Antarct. Seas 19 (in press).
ANDERSON, M. E., R. E. CRABTREE, H. J. CARTER, K. J. SULAK,
AND M. D. RICHARDSON. 1986. Distribution of demersal
fishes of the Caribbean Sea found below 2,000 meters. Bull.
Mar. Sci. 37(3):794-807.
ANDRIASHEV, A. P. 1 954. Fishes of the northern seas of the
USSR. Zool. Inst, Acad. Sci., Contr. Fauna USSR 53:1-
566.
. 1973. Zoarcidae. In Checklist of the fishes of the
northeastern Atlantic and of the Mediterranean. J.-C. Hu-
reau and T. Monod, eds. Unesco, Paris, 1:540-547.
. 1 986. Zoarcidae (including Lycodidae). In Fishes of
the north-eastern Atlantic and the Mediterranean. P. J. P.
Whitehead, M.-L. Bauchot, J.-C. Hureau, J. Nielsen, and E.
Tortonese, eds. Unesco, Paris, 3:1 130-1 150.
FOWLER, H. W. 1936. The marine fishes of West Africa. Bull.
Amer. Mus. Nat. Hist. 70(2):607-1493.
GARMAN, S. 1899. Reports on an exploration off the west
coasts of Mexico, Central and South America, and off the
Galapagos Islands, in charge of Alexander Agassiz, by the
U.S. Fish Commission steamer "Albatross" during 1891,
Lieut.-Commander Z. L. Tanner, U.S.N., commanding.
XXVI. The fishes. Mem. Mus. Comp. Zool. Harv. Coll. 24:
1-431.
GOSZTONYI, A. E. 1977. Results of the research cruises of
FRV "Walther Herwig" to South America. XLVIII. Revi-
sion of the South American Zoarcidae (Osteichthyes: Blen-
nioidei), with the description of three new genera and five
new species. Arch. FischWiss. 27(3): 19 1-249.
HUBBS, C. L., W. I. FOLLETT, AND L. J. DEMPSTER. 1 979. List
of the fishes of California. Occ. Papers Calif. Acad. Sci. (133):
1-51.
LEVITON, A. E., R. H. GIBBS, JR., E. HEAL, AND C. E. DAWSON.
1985. Standards in herpetology and ichthyology: Part I.
Standard symbolic codes for institutional resource collec-
tions in herpetology and ichthyology. Copeia 1985(3):802-
832.
MARKLE, D. F. AND G. R. SEDBERRY. 1978. A second spec-
imen of the deep-sea fish, Pachycara obesa, with a discussion
of its classification and a checklist of other Zoarcidae off
Virginia. Copeia 1978(l):22-25.
MCALLISTER, D. E. AND E. I. S. REES. 1964. A revision of
the eelpout genus Melanostigma with a new genus and with
comments on Maynea. Bull. Natl. Mus. Canada 199:85-
109.
MERRETT, N. R. AND N. B. MARSHALL. 1981. Observations
on the ecology of deep-sea bottom-living fishes collected off
northwest Africa (08°-27°). Prog. Oceanogr. 9:185-244.
PAPPENHEIM, P. 1912. Die Fische der Deutschen Siidpolar-
Expedition 1901-1903. I. Die Fische der Antarktis und
Subantarktis. Deutsche Sudpolar Exped. 13:160-182.
PEARCY, W. G., D. L. STEIN, AND R. S. CARNEY. 1982. The
deep-sea benthic fish fauna of the northeastern Pacific Ocean
on Cascadia and Tufts Abyssal Plains and adjoining conti-
nental slopes. Biol. Oceanogr. l(4):375-428.
PEDEN, A. E. 1979. Meristic variation of Lycodapus man-
dibularis (Pisces: Zoarcidae) and oceanic upwelling on the
west coast of North America. J. Fish. Res. Bd. Canada 36(1):
69-76.
. 1981. Meristic variation of four fish species exhib-
94
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 3
iting lowest median counts in Georgia Strait, British Colum-
bia. Can. J. Zool. 59(4):679-683.
PEDEN, A. E. AND M. E. ANDERSON. 1978. A systematic re-
view of the fish genus Lycodapus (Zoarcidae) with descrip-
tions of two new species. Can. J. Zool. 56(9): 1925-1 961.
SMALL, G. J. 1981. A review of the bathyal fish genus An-
timora (Moridae: Gadiformes). Proc. Calif. Acad. Sci. 42(13):
341-348.
WILSON, R. R., JR. AND R. S. WAPLES. 1984. Electrophoretic
and biometric variability in the abyssal grenadier Cory-
phaenoides armatus of the western North Atlantic, eastern
South Pacific and eastern North Pacific oceans. Mar. Biol.
80:227-237.
ZUGMAYER, E. 191 la. Diagnoses des poissons nouveaux
provenant des campagnes du yacht "Princesse-Alice" ( 1 90 1
a 1910). Bol. Inst. Oceanogr., Monaco (193):1-14.
. 1 9 1 1 b. Poissons provenant des campagnes du yacht
"Princesse-Alice." Result. Camp. Sci. Prince Albert I 35:1-
174.
CALIFORNIA ACADEMY OF SCIENCES
Golden Gate Park
San Francisco, California 94 1 1 8
PROCEEDINGS
OF THE
CALIFORNIA ACADEMY OF SCIENCES
Vol. 46, No. 4, pp. 95-136, 22 figs., 4 tables. March 9, 1989
THE HARVESTMEN FAMILY PHALANGODIDAE.
1. THE NEW GENUS CALICINA, WITH NOTES ON SITALCINA
(OPILIONES: LANIATORES)
By
Darrell Ubick
Department of Entomology, California Academy of Sciences, Golden Gate Park,
San Francisco, California 94118-4599
and
Thomas S. Briggs
Department of Entomology, California Academy of Sciences, Golden Gate Park,
San Francisco, California 94118-4599
ABSTRACT: The Californian phalangodid genus Sitalcina Banks is shown to be polyphyletic, and is restricted
to include only those species closely related to the type species Sitalces californica Banks. A new genus,
Calicina, described to accommodate most of the remaining species, is defined by the presence of a tarsal spur
on the male palpus, penis glans with apically directed stylus, ovipositor with double row of apical setae, and
ovipositor cuticle with microspines. At least the first of these characters is a synapomorphy.
Of the 25 species of Calicina recognized, 18 are transferred from Sitalcina: breva Briggs (elevated from
subspecies), cloughensis Briggs and Horn, digita Briggs and Horn, ensata Briggs, kaweahensis Briggs and
Horn, keenea Briggs, macula Briggs, mariposa Briggs, minor Briggs and Horn, morroensis Briggs, palapraeputia
Briggs, piedra Briggs, polina Briggs, sequoia Briggs and Horn, serpentinea Briggs and Horn, sierra Briggs and
Horn, topanga Briggs, and yosemitensis Briggs. The remaining species are new: arida, basalta, conifera,
diminua, dimorphica, galena, and mesaensis.
Based on genitalic characters, primarily glans structure, four species groups and nine subgroups are rec-
ognized. Their relationships are hypothesized using cladistic analysis and vicariance biogeography. The
distribution of Calicina is strongly concordant with that of the slender salamander, Batrachoseps, and with
the presumed distributions of exotic terranes.
Interspecific variation in the somatic morphology of Calicina appears to be the result of paedomorphosis
and troglobism. These modifications, present in most species, are probably adaptations to xeric environments
and, possibly, competition with other phalangodids.
Received October 7, 1987. Accepted May 11, 1988
„ „ Phylogeny 1 2 1
TABLE OF CONTENTS „. , ~~
Biogeography 125
Abstract 95 Natural History 130
Introduction 96 Ecology 131
Materials and Methods 96 Acknowledgments 135
Taxonomy 9 8 Literature Cited 1 3 5
[95]
96
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
INTRODUCTION
A major problem in the systematics of har-
vestmen is the apparent scarcity of reliable ge-
neric characters. This stems largely from the
classic studies of Roewer (e.g., 1 923) which over-
emphasized the importance of easily visible so-
matic characters, such as variation in the tarsal
count, in denning genera. This practice has re-
sulted in a confusing proliferation of small and
monotypic genera. For example, the 1 7 known
species of the eastern Nearctic Phalangodidae are
currently assigned to eight genera (Goodnight and
Goodnight 1942, 1967).
In sharp contrast to this general trend is the
treatment of the phalangodids of California: only
two genera are recognized for an unusually rich
fauna of 39 described species. One of these,
Banksula Roewer, includes strictly cavernico-
lous species having uniformly high tarsal counts
(4-6-5-6) and the other, Sitalcina Banks, con-
tains epigean species with lower counts (3-5-5-5
or less). Recent studies (Briggs 1974; Briggs and
Ubick 1981) have demonstrated that Banksula
is well denned by both somatic and genitalic
characters: the palpal femur is armed with dorsal
spines and the penis has a bifurcate ventral plate.
On the other hand, Briggs (1968) found no de-
rived character uniting the 29 species of Sital-
cina. He even commented on the heterogeneity
of the genus, listing interspecific differences in
size, tarsal count, presence or absence of eyes,
and type of sexually dimorphic structures. He
tentatively arranged the species into three groups:
(1) those with sexually dimorphic trochanters,
(2) those with sexually dimorphic palps, and (3)
those lacking sexual dimorphisms. However, in
the absence of additional characters and because
of the overall somatic similarity of the species,
he did not question the validity of the genus.
Our present investigations of the genitalic
morphology of Sitalcina indicate that the genus
is indeed polypliyletic. The strongest evidence
comes from the morphology of the penis where,
on the basis of the glans structure, two groups of
species are evident. The first is characterized by
a folded, caudally directed glans that swings open
180° along a dorsoapical arc during expansion
(Fig. Ib). The second is characterized by an api-
cally directed glans that expands by a rectilinear,
telescoping movement out of the shaft (Fig. 6f,
g). Such striking genitalic differences are com-
parable to those used to distinguish families
(Martens 1986).
It is now evident that Sitalcina must be re-
stricted to include only those species closely re-
lated to the type species, Sitalces californica
Banks. This monophyletic group, corresponding
to group "1" of Briggs (1968), is defined by the
characters listed in Table 1. In addition to the
folding glans, all Sitalcina males have trochan-
teral spurs on leg IV (Fig. la). The ovipositor of
Sitalcina is distinctive in various features, of
which the imbricate cuticle sculpturing is pos-
sibly synapomorphic.
The nine species that now comprise Sitalcina
(along with three others still unplaced that ap-
parently belong to other genera) will be treated
in a future paper. The purpose of this study is to
describe the new genus represented by the re-
maining species, for which we propose the name
Calicina. The species ofCalicina are revised and
their morphological and spatial relationships hy-
pothesized using cladistics and vicariance bio-
geography. Analyses of natural history and ecol-
ogy are provided.
MATERIALS AND METHODS
The 800 specimens examined during the course
of this study were collected almost exclusively
by Briggs and associates over the past two de-
cades. Specimens from early collections were
originally stored in Oudemans' Fluid, which gave
excellent initial preservation (Briggs 1968:3).
However, it is now evident that prolonged stor-
age in this solution (about 10 years or more)
discolors the specimens, results in deterioration
of the internal tissues, and makes genitalic ex-
traction difficult. By comparison, specimens pre-
served in 75% to 85% ethanol for much longer
periods do not show signs of such damage.
The genitalia can usually be extracted by sim-
ply opening the operculum and squeezing the
abdomen with forceps. When this fails it is nec-
essary to push out the genitalia carefully with a
probe inserted into the anal region or, as a last
resort, to dissect it out of the body cavity. Males
often require additional treatment with KOH in
order to fully expand the glans, which normally
lies retracted within the shaft. For this, entire
specimens are soaked in cold 1 0% KOH solution
for about two minutes and then transferred to a
water bath where, aided by gentle squeezing of
the abdomen, maximum expansion can occur.
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
97
TABLE 1 . COMPARATIVE MORPHOLOGY OF SITALCINA AND CALICINA.
Sitalcina
Calicina
Total length
Eyes
Tarsal count
1.50-2.40 mm
present
3-5-5-5
0.77-1.67 mm
present or absent
3-5-5-5 to 3-3-4-4
Palp
Patella
Tarsus ($)
3 spine-bearing tubercles
unmodified
2 spine-bearing tubercles
with spur (usually)
Leg IV
Trochanter (6)
with spur
unmodified
Penis
Glans
folding
telescoping
Ovipositor1
Shape
Cuticle
bent
imbricate2
straight
with microspines
Apical setae
Shape
Length
Number
Arrangement
apically hooked3
long— cross at center
12-20
not grouped4
evenly bent
shorter
12, rarely 14 or 16
in triads
Subapical setae
Number
none
1-3 pairs, rarely none
1 Species of Sitalcina examined: californica, borregoensis, madera, scapula, and sura.
2 In S. californica.
3 Except S. scapula.
4 Except 5. sura.
Unless otherwise indicated, the specimens ex-
amined during this study, including all primary
types, are deposited at the California Academy
of Sciences. Additional specimens are from the
collections of the American Museum of Natural
History (AMNH) and the University of Califor-
nia at Berkeley (UCB).
Complete species descriptions are given for new
species; additional descriptions for those treated
by Briggs (1968). All specimen measurements are
in millimeters.
ABBREVATIONS FOR COLLECTORS
AGO -Andy G. Grubbs
AK- Albert Kurz
AKSJ- Albert K. S. Jung
AL— Albert Lee
BL— Benton Leong
BM— Barbara Martin
CF-C. Fox
DB-D. Broussard
OCR— D. Craig Rudolph
DR- David Rentz
DU-Darrell Ubick
EVI— E. Vanlngen
GAM-G. A. Marsh
GL— Galen Leung
GT- Gordon Tang
JG— Joseph Gee
JRH-J. R. Heifer
JSB-J. S. Buckett
KH- Kevin Horn
KK-K. Kay
LMS-L. M. Smith
MG— Michael Gardner
MM— M. McEachern
MW— Michael Wong
NB-N. Boice
RL— Robert Lem
ROS— Robert O. Schuster
SK-S. Kubota
SO -Stefan Ong
SW- Steve Winterath
TO— Toshiro Ohsumi
TSB— Thomas S. Briggs
VFL— Vincent F. Lee
WCR- Warren C. Rauscher
WES -Warren E. Savary
WI- Wilton I vie
98
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
FIGURE 1. Sitalcina californica. (a) Trochanter IV of male, lateral view (180x). (b) Expanded glans penis, sublateral view
(240 x ). Arrow indicates movement of glans during expansion, (c) Ovipositor, lateral view ( 1 80 x ). (d) Ovipositor, close-up of
lateral surface showing imbricate cuticle (l,800x). (e) Ovipositor, apical view (300 x).
TAXONOMY
Family PHALANGODIDAE Simon
Subfamily PHALANGODINAE Simon
Calicina Ubick and Briggs, new genus
DIAGNOSIS.— Males of Calicina are unique
among the Nearctic phalangodid genera in hav-
ing a telescoping penis glans (as opposed to a
folding one). Additionally, most males of Cali-
cina have a dorsal spur on the palpal tarsus, which
is not known in other phalangodids. Females are
distinct from those of Sitalcina, and possibly oth-
er genera, in having ovipositors with a double
setal fringe and a cuticle with microspines. Most
species of Calicina may be distinguished from
other phalangodid genera by their reduced tarsal
counts of less than 3-5-5-5.
TYPE SPECIES. — Sitalcina mariposa Briggs,
1968.
ETYMOLOGY.— The generic name is a contrac-
tion of California Sitalcina and is feminine in
gender.
DESCRIPTION.— Body length 0.77-1.67. Color
ranges from pale yellow to orange to reddish-
brown. Body cuticle minutely and densely tu-
berculate, with scales, setae, and nipplelike tu-
bercles. Scute with prominent ocular tubercle at
anterior margin; varying in size, shape, and de-
gree of rugosity (see Briggs 1968, fig. 45-57, 59,
62, 63, 65, 66). Eyes well developed (Fig. 3a) or
reduced, lacking retina and occasionally cornea
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
99
FIGURE 2. Calicina mariposa, lateral view of male (left appendages omitted). AT = anterior tubercles, OT = ocular tubercle,
TS = tarsal spur of palpus, GO = genital operculum, P = penis, legs numbered.
(=lens) (Fig. 3b). Carapace with two rows of an-
terior tubercles numbering from 0 to 4 or rarely
to 8 pairs (Fig. 2, 3a, b). Ozopore well developed,
on anteriolateral margin, with distinct posterior
channel. Tergites I-V fused, VI-VIII free, with
median, transverse row of tubercles interspersed
with short setae. Venter textured similarly as scute
but more densely setose; seven free sternites with
median, transverse rows of tubercles. Genital
operculum between coxae IV, apically truncate
in males, rounded in females. Coxae I and II with
endites.
Leg sculpturing consisting of minutely tuber-
culate scales; calcanei and tarsi smooth. Leg II
longest, leg length formula (longest to shortest)
II,IV,I,III. Tarsal counts 3-5-5-5, 3-4-4-5, or 3-4-
4-4; rarely 3-5-4-5 or 3-3-4-4. Tarsal claws single
on legs I and II; paired on legs III and IV. Ju-
veniles with hind claws on onychium, arolium
between paired claws. Palpi and chelicerae dis-
tally setose, with smooth cuticle of appressed
scales. Palpi robust, ventrally armed with spine-
bearing tubercles (femur with three ectobasally
and one mesodistally; patella with 1 pair; tibia
and tarsus with 2 pairs each) (Fig. 3c). Male pal-
pal tarsus with a well developed dorsal spur (Fig.
4a), with vestigial spur (Fig. 4b), or without spur;
occasionally with enlarged mesal spine-bearing
tubercle (Fig. 4c, d).
Penis without muscles, composed of basal sac,
shaft, and apical glans. Basal sac inflatable. Shaft
(=truncus) cylindrical, sclerotized; length 3.5-7
x width; apicoventral part (= ventral plate)
rounded or attenuated, set with setae. Glans tele-
scoping, with basal segment and stylus, rarely
with additional (middle) segment (palapraepu-
tia). Collar lobes (=1Titillator of Martens 1986)
on basal segment (mariposa and digita groups),
on middle segment (palapraeputia), or appar-
ently absent (serpentinea group). Stylus apical
and variable in shape. Accessory structures either
thin, scalelike parastyli (=7Konduktor of Martens
1986) (digita group) or robust, lobed dorsal pro-
cess (serpentinea group).
Ovipositor short, straight, and cylindrical when
expanded; length 1.5-2.0 x width; divided into
basal and distal segments. Lateral and/or apical
surface covered with minute, pointed tubercles
(microspines). Apex with 2 or 4 lobes, occasion-
ally with dorsal folds; with setal fringe of 1 2 api-
cal setae (10 in diminua) grouped in triads and
1-4 pairs of subapical setae (most species) or with
14 apical setae only (palapraeputia and serpen-
tinea subgroup); occasionally with a pair of apical
teeth (Fig. 5).
DISTRIBUTION AND SPECIES.— The genus Cali-
cina is found only in the central Sierra Nevada
and the Coast Ranges of California. The 25 known
species represent four species groups and nine
subgroups.
Key to the Calicina Species Groups
and Subgroups
Based on Characters of the Male Genitalia
1 . Basal glans segment with a pair of collar
lobes (Fig. 16a-d) 2
- Basal glans segment without collar lobes
(Fig. 1 6e-h) 7
2. Parastyli absent (Fig. 6, lOa-d) 3
- Parastyli present (Fig. 7-9, lOe-h) (digita
group) 4
3. Collar lobes simple (Fig. 6)
mariposa group and subgroup
- Collar lobes bilobed (Fig. lOa-d)
arida subgroup, in part (doughensis)
4. Parastyli ventral (Fig. 7) digita subgroup
- Parastyli lateral or dorsal 5
5. Parastyli lateral, broad and ornate (Fig. 8)
kaweahensis subgroup
100
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
FIGURE 3. Calicina, morphology, (a) C. mariposa, male paratopotype, anterior part of scute showing ocular tubercle with
large cornea, a row of paracular anterior tubercles (AT), and the ozopore (O) (230 x ). (b) C. serpentinea, female, same view
showing reduced number of anterior tubercles and degenerate cornea (C) (230 x ). (c) C. mariposa, male paratopotype, chelicerae
and palpi showing tarsal spurs (TS) (80 x ).
- Parastyli dorsal, narrow 6
6. Parastyli separate; stylus straight; ventral
plate acuminate (Fig. 9) topanga subgroup
- Parastyli fused; stylus sinuous; ventral plate
unmodified (Fig. lOe-h)
arida subgroup, in part (aridd)
Middle segment of glans completely sur-
rounding stylus; collar lobes hornlike (Fig.
1 1) palapraeputia group and subgroup
Middle segment of glans forming the dorsal
process that is distinct from stylus; lobes
variable (Fig. 1 6g, h) (serpentinea group) 8
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
101
FIGURE 4. Calicina, male palpal tarsi, (a) C. mariposa, showing large tarsal spur (310x). (b) C. mesaensis, showing reduced
tarsal spur (3 10 x ). (c) C. minor, showing enlarged mesobasal spine-bearing tubercle (X) (470 x ). (d) C. ensata, showing enlarged
mesoapical spine-bearing tubercle (X) (310x).
102
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
FIGURE 5. Calicina, ovipositors, (a) C. mariposa, lateral view (270 x). (b) C. mariposa, close-up of lateral surface showing
microspines (2,700x). (c) C. mariposa, apical view (390 x). (d) C. topanga, lateral view showing subapical setae (S) (230 x). (e)
C. palapraeputia, ventral view showing apical tooth (T) (170x). (f) C. palapraeputia, apical view showing microspines and
anterior margin of genital operculum (230 x). (g) C. serpentinea, lateral view showing apical tooth (310x). (h) C. serpentinea,
apical view showing microspines (390 x).
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
103
104
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CAL1CINA
105
8. Dorsal process with apical bifurcation (Fig.
1 2) minor subgroup
- Dorsal process with apical lobes 9
9. Dorsal process with 5 apical lobes; ventral
plate with at most 4 pairs of lateral setae
(Fig. 13, 14) serpent inea subgroup
- Dorsal process with at most 4 apical lobes;
ventral plate with 5 to 8 pairs of lateral
setae (Fig. 1 5) sequoia subgroup
THE MARIPOSA SPECIES GROUP
(The mariposa Subgroup)
DIAGNOSIS.— Members of the mariposa group
may be recognized by their simple genitalia. The
glans consists of a stylus and a basal segment
bearing collar lobes and lacks additional sclerites.
The ovipositor contains the full complement of
12 apical and 3 pairs of subapical setae. The
dorsal grooves on the collar lobes are synapo-
morphic for the group. The three species be-
longing to this group are all summer-active forest
dwellers.
DISTRIBUTION.— The mariposa group is found
only in the central Sierra Nevada in Yosemite
National Park at elevations of 1,200-2,000 m.
Calicina mariposa (Briggs, 1968),
new combination
(Fig. 2, 3a, c, 4a, 5a-c, 6a, b, 16a, b)
Sitalcina mariposa Briggs, 1968:31.
DIAGNOSIS.— This species differs from other
members of the group in having a high tarsal
count of 3-5-5-5; males lack an apical notch on
the ventral plate of the penis.
ADDITIONAL DESCRIPTION.— Carapace with 6-
8 pairs of anterior tubercles. Penis large; ventral
plate apically entire and with 8 pairs of marginal
setae; collar lobes pointed and pocketlike; stylus
long and sinuous. Ovipositor with 1 2 apical setae
and 3 pairs of lateral, subapical setae; micro-
spines on apical and lateral surfaces.
TYPE SPECIMENS.— Male holotype, female allotype, and three
paratypes (one male, two females), under logs in a giant sequoia
grove at Mariposa Big Trees, Yosemite National Park, Mari-
posa County, 28 July 1967, TSB and KH.
SPECIMENS EXAMINED.— Four males, 12 females.
NEW RECORDS.- MARIPOSA COUNTY: Mariposa Big
Trees, 26 July 1984 (TSB, SK, GT); TUOLUMNE COUNTY:
Yosemite National Park: Tioga Pass Road, June 1970 (TSB);
9.2 km E Crane Flat Junction, 27 July 1984 (TSB, SK, GT);
Wawona Camp, 17 Sep. 1941 (WI; AMNH collection).
NATURAL HISTORY. —Known from sequoia and
yellow pine forests; collected from June to Sep-
tember.
Calicina yosemitensis (Briggs, 1968),
new combination
(Fig. 6e-g)
Sitalcina yosemitensis Briggs, 1968:31.
DIAGNOSIS.— Males of this species have a glans
with a uniquely modified stylus tip; females may
be separated from other group members in hav-
ing an ovipositor with only 1 pair of subapical
setae.
ADDITIONAL DESCRIPTION.— Carapace with 4-
8 pairs of anterior tubercles. Penis medium-sized;
ventral plate with apical notch and 7 pairs of
lateral setae; collar lobes small, pointed, apically
grooved; stylus tip with lateral expansions. Ovi-
positor as in mariposa but with only 1 pair of
subapical setae.
TYPE SPECIMENS. — Male holotype, female allotype, and three
paratypes (two males, one female), under fallen bark in conif-
erous forest 5.0 mi (8.0 km) E Crane Flat Ranger Station,
Yosemite National Park, Tuolumne County, 28 July 1967,
TSB and AL.
SPECIMENS EXAMINED.— Five males, three females.
NEW RECORD. -TUOLUMNE COUNTY: Yosemite Na-
tional Park: 3.8 mi (6.1 km) NE Crane Flat Ranger Station,
21 July 1968 (TSB, MW, GL).
Calicina conifera Ubick and Briggs,
new species
(Fig. 6c, d)
DIAGNOSIS.— This species differs from others
in the mariposa group by its small size, pale col-
oration, and absence of retina.
ETYMOLOGY.— The specific name refers to the
coniferous biome in which this species lives.
DESCRIPTION.— Color yellowish orange. Car-
FIOURE 6. Male genitalia of the mariposa group, (a, b) C. mariposa (Crane Flat Junction), ventral and lateral views, (c, d)
C. conifera (holotype), ventral and lateral views, (e-g) C. yosemitensis. (e, f) Dorsal and lateral views showing retracted glans
(paratopotype). (g) Dorsolateral view showing expanded glans, arrow indicates movement of glans during expansion (3.8 mi NE
Crane Flat). Note: Unless otherwise indicated, the scale bar for these and all subsequent genitalic drawings equals 0.30 mm,
and all drawings are of fully expanded glandes.
106
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
apace with 3 or 4 pairs of anterior tubercles.
Retina absent, lens present. Tarsal count 3-4-4-5.
Male: Total body length, 1.20. Scute length,
0.80. Scute width, 0.88. Eye tubercle length, 0. 1 2.
Eye tubercle width, 0.16. Leg II length, 2.13.
Palpal tarsus with dorsal spur. Penis small; ven-
tral plate apically notched and with 7 pairs of
marginal setae; collar lobes small, rounded and
apically grooved; stylus uniformly tapering and
sinuous.
Female: Total body length, 1 .00. Scute length,
0.63. Scute width, 0.7 1 . Eye tubercle length, 0. 1 3.
Eye tubercle width, 0. 1 1 . Ovipositor as in mar-
iposa but with two pairs of subapical setae.
TYPE SPECIMENS.— Male holotype and two female para types,
under fallen bark in a mixed coniferous forest at 1.8 km E
Crane Flat Junction, Tuolumne County, 29 July 1984, TSB,
SK, and GT.
SPECIMENS EXAMINED.— Only the type series.
THE DIGITA SPECIES GROUP
DIAGNOSIS.— The parastyli on the glans penis
is the presumed synapomorphy for this group.
The digita Subgroup
DIAGNOSIS.— The digita subgroup is distin-
guished from all others by the unique male gen-
italia: the ventral plate is arrow-shaped and the
glans bears a pair of slender, ventral parastyli.
Males have a large dorsal spur on the palpal tar-
sus.
DISTRIBUTION.— Middle elevation Sierra Ne-
vada from Amador south to Tulare counties.
Calicina digita (Briggs and Horn, 1967),
new combination
(Fig. 7a-d, 16c, d)
Sitalcina digitus Briggs and Horn, 1967:51. Briggs, 1968:20.
DIAGNOSIS. — This species is easily distin-
guished from the other in its subgroup by a higher
tarsal count of 3-5-5-5.
ADDITIONAL DESCRIPTION.— Carapace with 6
or 7 pairs of anterior tubercles. Glans with a pair
of short, scalelike parastyli and long, sinuous sty-
lus. Ventral plate with 7 pairs of setae. Ovipositor
similar to that ofmariposa but with lower density
of microspines and with 2 or 3 pairs of subapical
setae.
TYPE SPECIMENS.— Male holotype, female allotype, and 20
paratypes (12 males, eight females), under fallen bark at 2.2
mi (3.5 km) S Giant Forest, Sequoia National Park, Tulare
County, 15 May 1966, TSB and KH.
SPECIMENS EXAMINED.— Twenty-seven males, 22 females.
NEW RECORDS. -MARIPOSA COUNTY: 3 mi (5 km) E
Briceburg, 6 Feb. 1 954 (ROS; AMNH collection); Merced Riv-
er, N Fork near Main Fork, 3 1 Mar. 1 983 (TSB, SO). FRESNO
COUNTY: Saturday Cave, near Boyden Cave, 15 May 1966
(TSB); Big Creek, 7 Aug. 1956 (ROS; UCB collection). TU-
LARE COUNTY: Sequoia National Park: 0.9 mi (1.4 km) S
Giant Forest Village, 17 Aug. 1984 (TSB, VFL, DU); 4 mi (6.4
km) N Hospital Campground, 13 Feb. 1967 (JSB, MG).
NATURAL HISTORY.— Found in a variety of
biomes, including forests (giant sequoia-yellow
pine forest, oak woodland) and grassland, and
collected throughout the year. There is one col-
lection from a cave.
REMARKS.— The disjunct populations do not
appear to differ in reproductive or somatic char-
acters.
Calicina sierra (Briggs and Horn, 1967),
new combination
(Fig. 7e-g)
Sitalcina sierra Briggs and Horn, 1967:49.
Sitalcina sierra sierra Briggs and Horn. Briggs, 1968:20.
DIAGNOSIS. — Calicina sierra, a paedomorphic
species, is most readily distinguished from digita
by its low tarsal count of 3-4-4-4.
ADDITIONAL DESCRIPTION.— Carapace with 3
or 4 pairs of anterior tubercles. Penis similar to,
but smaller than, that of digita. Ovipositor as in
mariposa.
JUVENILES.— One of the two juveniles is an ear-
ly instar (from Scat Cave) having a tarsal count
of 1-1-2-2, the other is a penultimate (McLean's
Cave area, 2 Nov. 1975) just prior to moulting.
Both instars lack anterior tubercles. The late in-
star has several robust spines on each of the free
tergites; the early instar has only one pair of such
spines restricted to tergite VII. The eyes of both
instars are well developed, as in adults.
TYPE SPECIMENS.— Male holotype, female allotype, and five
paratypes (three males, two females), 6.2 mi (10.0 km) N Mer-
cer Caverns, Sheep Ranch Road, Calaveras County, 26 Mar.
1966, TSB and KH.
SPECIMENS EXAMINED. — Forty-three males, 30 females, two
juveniles.
NEW RECORDS.-CALAVERAS COUNTY: Music Hall Cave,
7.2 km ESE Angel's Camp, 18 May 1977 (AGG, SW, MM,
NB); Carlow's Bat Cavern, 7 km SE Angel's Camp, 25 May
1977 (DB, SW, AGG); Scat Cave, 4 mi (6.4 km) W Columbia,
26 Mar. 1979 (BM, SW); unnamed cave, Peruvian Gulch, 2
mi (3.2 km) NW Columbia, 10 May 1980 (DU); Speleogen
Cave, 7 km SE Angel's Camp, 20 Apr. 1980 (TSB, DU).
TUOLUMNE COUNTY: Lower van Trump Mine, Fox Gulch,
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
107
FIGURE 7. Male genitalia of the digita subgroup, (a-d) C. digita. (a, b) Lateral and subventral views showing retracted glans
(Merced River), (c, d) Lateral view of expanded glans and dorsal view of left collar lobe (Giant Forest), (e-g) C. sierra (McLean's
Cave area), ventral, dorsal, and lateral views.
23 Sep. 1975 (TSB, RL); Porcupine Skull Cave, 2 mi (3.2 km)
N Columbia, 11 Feb. 1979 (OCR, SW); 2.5 mi (4 km) NW
Columbia, 2 Feb. 1979 (OCR, SW); S Fork Stanislaus River
at Marble Quarry Road, 5 Mar. 1981 (DU); 2.5 mi (4 km) N
Columbia, McLean's Cave area, 2 Nov. 1975 (TSB, RL), 2
Apr. 1979 (OCR, BM, SW).
DISTRIBUTION.— Known from Amador, Ca-
laveras, and Tuolumne counties.
NATURAL HISTORY.— Found in digger pine-oak
forests during the winter and spring months (No-
vember-April). C. sierra is also known from
caves, where it is active throughout the year,
although it has never been collected in associa-
tion with the sympatric, cavernicolous species of
Banksula. In an extensive survey of the Cala-
veras Formation caves, Banksula was recorded
from more than 30 caves and C. sierra from six
different ones (Briggs and Ubick 1981).
The kaweahensis Subgroup
DIAGNOSIS.— Males of the kaweahensis sub-
group are unique in having a penis with ornate,
lateral parastyli.
108
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
FIGURE 8. Male genitalia of the kaweahensis subgroup, (a-d) C. galena, (a) Sublateral view showing retracted glans (holotype).
(b-d) Lateral, dorsal, and ventral views (paratopotype). (e-g) C. kaweahensis (Rocky Hill), dorsal, ventral, and lateral views.
DISTRIBUTION. — Known only from Tulare
County.
Calicina kaweahensis (Briggs and Horn, 1966),
new combination
(Fig. 8e-g)
Sitalcina kaweahensis Briggs and Horn, 1966:266. Briggs, 1968:
22.
DIAGNOSIS. — This species may be distin-
guished by the combination of large size (body
length greater than 1 .4 mm) and a tarsal count
of 3-5-4-5. Males have a glans with uniquely
modified parastyli.
ADDITIONAL DESCRIPTION.— Carapace with 4
pairs of anterior tubercles. Penis with collar lobes
in basal position; parastyli dorsally fused and
apically attenuated; stylus straight and basally
enlarged; ventral plate with 8 pairs of marginal
setae. Ovipositor as in mariposa but with 2 pairs
of subapical setae and with a pair of large apical
teeth; microspines on entire distal segment but
sparse on the ventral surface.
TYPE SPECIMENS. — Female holotype, male allotype, and two
paratypes (damaged), under granitic rocks at 0.3 mi (0.5 km)
N Junction of Woodlake-Lemoncove Road on road to Badger,
Tulare County, 26 Nov. 1965, TSB and VFL.
SPECIMENS EXAMINED.— Seventeen males, 11 females, four
specimens of undetermined sex (damaged).
NEW RECORDS. -TULARE COUNTY: 0.3 mi (0.5 km) N
junction Woodlake-Lemoncove Road, 18 Dec. 1965 (KH);
near Lake Kaweah Offices on Hwy 198, 14 May 1966 (TSB);
near Kaweah Dam on Hwy 198, 26 Nov. 1965 (VFL); 2.1 mi
(3.4 km) E Rocky Hill Town, 19 Dec. 1966 (TSB, KH); 1.4
mi (2.3 km) E Rocky Hill, 22 Jan. 1967 (TSB, KH, AKSJ);
Yokohl Valley, near Lindsay, Feb. 1971 (TSB); Three Rivers,
Nov. 1967 (TSB).
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
109
NATURAL HISTORY.— Known from oak wood-
land biomes and collected throughout the winter
and spring months (November-May).
Calicina galena Ubick and Briggs,
new species
(Fig. 8a-d)
DIAGNOSIS.— Males of this species differ from
all others by the ornately scrolled parastyli and
the unique transverse rows of setae on the ventral
plate.
ETYMOLOGY.— The specific name refers to the
type locality and is a name in apposition.
DESCRIPTION.— Carapace with 1 pair of ante-
rior tubercles. Eyes present. Tarsal count 3-5-5-5
or 3-5-4-5.
Male: Total body length, 1.17. Scute length,
0.79. Scute width, 0.72. Eye tubercle length, 0. 1 7.
Eye tubercle width, 0.12. Leg II length, 1.94.
Color yellowish orange. Palpal tarsus without
dorsal spur. Penis with ventral plate bearing 6
pairs of setae arranged in two transverse rows;
collar lobes reduced to small flaps; parastyli in-
tricately scrolled; stylus short and spinelike.
Female: Total body length: 1.01. Scute length,
0.69. Scute width, 0.78. Eye tubercle length, 0. 1 3.
Eye tubercle width, 0.12. Color orange. Ovipos-
itor as in mariposa.
VARIATION.— Some specimens from the John-
sondale localities have the basal tarsomeres of
tarsus III fused, giving a tarsal count of 3-5-4-5.
JUVENILES.— The single known juvenile is very
pale, has a tarsal count of 3-5-4-5, and has well
developed eyes. As in adults, one pair of anterior
tubercles is present, and the abdominal tergites
lack robust setae.
TYPE SPECIMENS.— Male holotype and 11 paratypes (two
males, eight females, one juvenile), Galena Creek, Mountain
Home State Forest, Tulare County, 4 Sep. 1967, TSB. Ten
paratypes (seven males, three females), 2.25 mi (3.6 km) W
Johnsondale, Tulare County, 1,600 m, 2 July 1988, TSB and
DU. Nineteen paratypes (12 males, seven females), Double
Bunk Meadows, 3.5 air mi (5.6 km) SW Johnsondale, Tulare
County, 1,850 m, 3 July 1988, TSB and DU.
SPECIMENS EXAMINED.— Only the type series.
NATURAL HISTORY.— The original series was
collected beneath granite rocks in a giant sequoia
forest. The recent collections are from mixed co-
niferous forests; specimens were collected be-
neath and within decaying logs, large slabs of
fallen bark, and dense mats of duff. The popu-
lation at 3.6 km W Johnsondale is sympatric
with a species of Sitalcina.
The topanga Subgroup
DIAGNOSIS.— This subgroup is distinguished
from all others by several unique male genital
characters: apex of ventral plate attenuated and
armed with two pairs of strong setae; stylus
straight and elongated, somewhat flattened;
parastyli short, hornlike in shape and dorsally
situated; and collar lobes small and covered with
minute tubercles.
DISTRIBUTION.— Members of this subgroup
occur in the southern Coast Ranges and Te-
hachapi Mountains, the southernmost distribu-
tion of the genus.
Calicina topanga (Briggs, 1968),
new combination
(Fig. 5d, 9c, d)
Sitalcina topanga Briggs, 1968:25.
DIAGNOSIS. — This species may be distin-
guished from others in the subgroup by its high
tarsal count of 3-5-5-5.
ADDITIONAL DESCRIPTION.— Carapace with 2
or 3 pairs of anterior tubercles. Penis with ventral
plate attenuated and bearing 7 pairs of setae, of
which four apical ones are enlarged; collar lobes
earlike lateral projections with microspines;
parastyli dorsal, short and hornlike; stylus long
and straight, somewhat flattened. Ovipositor as
in mariposa but with 2-4 pairs of subapical setae
(one specimen with 9 setae).
TYPE SPECIMENS.— Male holotype, female allotype, and four
paratypes (two males, two females), under sandstone in a dense
oak forest at 3.3 mi (5.3 km) N Topanga Beach in Topanga
Canyon, Los Angeles County, 7 Apr. 1966, TSB, KH, and
VFL.
SPECIMENS EXAMINED.— Thirty-eight males, 32 females, one
specimen of undetermined sex.
NEW RECORDS. -SANTA BARBARA COUNTY: Santa Ynez
Mountains, Cold Springs, 19 Dec. 1967 (TSB, AKSJ). Santa
Cruz Island: canyon N UCSB Field Station, 1 9 Dec. 1 967 (TSB,
KH, AKSJ), 21 Dec. 1967 (TSB); 0.1 mi (0.2 km) SW UCSB
Field Station, 20 Dec. 1967 (AKSJ); Canada del Puerto, Pris-
oners Harbor, 21 Dec. 1967 (KH, AKSJ); Raven's Wood Can-
yon, 22 Dec. 1967 (TSB); Cristi Pines, 19-22 Dec. 1967 (TSB,
KH, AKSJ).
NATURAL HISTORY.— Known from both closed
and open forests (broadleaf evergreen, digger pine-
oak, and oak woodland), and collected during
the winter and spring months (December-April).
110
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
FIGURE 9. Male genitalia of the topanga subgroup, (a, b) C. morroensis (Madonna Inn), ventral and subdorsal views, (c, d)
C. topanga (Santa Cruz Island), subventral and lateral views, (e-g) C. keenea (holotype), lateral, ventral, and dorsal views.
REMARKS: The mainland and island popula-
tions of topanga appear similar in both genitalic
and somatic characters.
Calicina keenea (Briggs, 1968),
new combination
(Fig. 9e, f)
Sitalcina keenea Briggs, 1968:25.
DIAGNOSIS.— This species is distinguished from
others in its group by its tarsal count, 3-4-4-5.
ADDITIONAL DESCRIPTION.— Carapace with 2
or 3 pairs of anterior tubercles. Penis similar to
but smaller than that of topanga; ventral plate
with 6 pairs of lateral setae. Ovipositor as in
mariposa but with 4 pairs of subapical setae and
with a lower density of microspines.
TYPE SPECIMENS. — Male holotype and female allotype, under
granite in moist oak grassland at 3.2 mi (5.1 km) NW Keene,
Kern County, 19 Mar. 1967, VFL.
SPECIMENS EXAMINED.— Only the type series.
Calicina morroensis (Briggs, 1 968),
new combination
(Fig. 9a, b)
Sitalcina morroensis Briggs, 1968:26.
DIAGNOSIS.— This paedomorphic species dif-
fers from others in the group by the combination
of low tarsal count of 3-4-4-4 and degenerate eyes
(lacking retina and sometimes cornea).
ADDITIONAL DESCRIPTION.— Carapace with 1
pair of anterior tubercles. Penis typical for the
group; ventral plate with strong apical setae and
6 pairs of lateral setae; stylus broad and flat,
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
111
FIGURE 10. Male genitalia of the arida subgroup, (a-d) C. cloughensis. (a-c) Lateral, ventral, and dorsal views (topotype).
(d) Dorsal view of slide-mounted specimen showing retracted glans (holotype). (e-h) C. arida. (e-g) Lateral, dorsal, and ventral
views (holotype). (h) Sublateral view showing retracted glans (paratopotype).
parastyli very small. Ovipositor similar to that TYPE SPECIMENS. -Male holotype and female allotype, under
K- . ; •,-,-/• j j serpentine in marshy grassland at 0.8 mi (1.3 km) N Morro
of manposa but with 3 pairs of strongly reduced Bay San Luis Obispo County 2 Mar 1 967> VFL.
Subapical setae and with Sparse distribution of SPECIMENS EXAMINED. -Twenty-nine males, 25 females.
microspines. NEW RECORDS. -SAN LUIS OBISPO COUNTY: San Luis
112
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
Obispo, behind Madonna Inn, 22 Mar. 1970 (TSB, GL); 2.4
mi (3.9 km) NW San Simeon in serpentine grassland, 13 Feb.
1987 (TSB, VFL).
The arida Subgroup
DIAGNOSIS.— Members of this subgroup are
distinguished from all other species by their
unique glans penis: the stylus is strongly sinuous
and the collar lobes are greatly enlarged. Males
also lack spurs on their palpal tarsi but have
enlarged mesobasal, spine-bearing tubercles.
DISTRIBUTION.— Known from a single locality
in San Benito County and a cave in Tulare Coun-
ty.
Calicina arida Ubick and Briggs,
new species
(Fig. 10e-h)
DIAGNOSIS. — Males of this paedomorphic
species are unique in the genus in having the
parastyli represented by a single rod.
ETYMOLOGY. —The specific name is a Latin ad-
jective meaning dry, and refers to the climatic
conditions in the vicinity of the type locality.
DESCRIPTION.— Color yellowish orange. Car-
apace with 1 pair of anterior tubercles. Retina
absent, cornea present. Tarsal count 3-4-4-4.
Male: Total body length, 0.80. Scute length,
0.57. Scute width, 0.56. Eye tubercle length, 0.10.
Eye tubercle width, 0.13. Leg II length, 1.65.
Palpal tarsus without dorsal spur but with en-
larged mesobasal spine-bearing tubercle. Penis
relatively stout; ventral plate with slight apical
attenuation and 6 pairs of setae; collar lobes
greatly enlarged and apically directed; parastyli
represented by a single, dorsal rod; stylus strong-
ly sinuous and dorsally directed.
Female: Total body length, 0.73. Scute length,
0.50. Scute width, 0.53. Eye tubercle length, 0.06.
Eye tubercle width, 0.08. Ovipositor with 1 2 api-
cal setae, with 3 pairs of subapical setae and with
a pair of short apical teeth; microspines sparsely
distributed apically and laterally.
TYPE SPECIMENS. — Male holotype and 18 paratypes (nine
males, nine females), under serpentine rocks in an oak gully
at Panoche Road, 12.7 mi (20.4 km) SE Paicines, San Benito
County, 1 Dec. 1984, TSB, VFL, and DU. One male parato-
potype, 1 Dec. 1971, KH. Two paratypes (male, female) de-
posited at AMNH collection.
SPECIMENS EXAMINED. — Eleven males, 10 females.
Calicina cloughensis (Briggs and Horn, 1967),
new combination
(Fig. lOa-d)
Sitalcina cloughensis Briggs and Horn, 1967:52. Briggs, 1968:
23.
DIAGNOSIS.— This species is the only troglo-
bitic Calicina and is recognized by its pale col-
oration, complete loss of eyes (cornea as well as
retina), and a high tarsal count (3-5-5-5). Fur-
thermore, males of cloughensis have bilobed col-
lar lobes and a unique stylus bearing a median
enlargement.
ADDITIONAL DESCRIPTION.— Carapace with 1
pair of anterior tubercles. Penis with ventral plate
bearing 6 pairs of lateral setae; collar lobes di-
vided into apical and basal portions; stylus long
and sinuous, with crescent-shaped enlargement
at midlength, and with apical extension. Ovi-
positor with 1 2 apical setae (nine in one speci-
men) and 2 or 3 pairs of subapical setae; mi-
crospines sparsely distributed on lateral surface.
JUVENILES.— Two of the paratypes are early in-
star juveniles having tarsal counts of 1 -2-2-2 and
completely lacking eyes. Both lack anterior tu-
bercles but have robust spines on the free tergites
(adults have normal setae).
TYPE SPECIMENS. — Male holotype, female allotype, and six
paratypes (male, three females, two juveniles), in Ladder Room
of Clough Cave at South Fork Ranger Station 1,070 m (3,500
ft), Sequoia National Park, Tulare County, 14 May 1966, TSB,
VFL, and KH.
SPECIMENS EXAMINED. — Four males, nine females, two ju-
veniles.
NEW RECORDS. -TULARE COUNTY: Clough Cave, 1 8 Apr.
1979 (SW, EVI), 31 Aug. 1986 (TSB, KH, AKSJ).
REMARKS.— The stylus of the holotype male
lacks the apical extension found in other speci-
mens.
THE PALAPRAEPUTIA SPECIES GROUP
(The palapraeputia Subgroup)
DIAGNOSIS.— The single species representing
this group has unique male genitalia: glans of
penis three segmented with middle segment
bearing a pair of hornlike collar lobes.
DISTRIBUTION.— Found only in Fresno Coun-
ty.
Calicina palapraeputia (Briggs, 1968),
new combination
(Fig. 5e, f, 11, 16e, f)
Sitalcina palapraeputia Briggs, 1968:29.
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
113
DIAGNOSIS.— Same as for group.
ADDITIONAL DESCRIPTION.— Carapace with 2
or 3 pairs of anterior tubercles. Genital opercu-
lum with concave anterior margin. Penis with
ventral plate truncate, apically elongated and
armed with 7 pairs of setae; glans with two col-
lars; collar lobes raptorial; stylus stout. Ovipos-
itor with short distal segment, with 7 pairs of
apical setae, and with a pair of large apical teeth;
microspines present only at apex.
TYPE SPECIMENS. — Male holotype, female allotype, and six
paratypes (three males, three females), under serpentine rocks
in open grassland at 7 mi (1 1 km) NE Piedra, Fresno County,
21 Jan. 1967, TSB and KH.
SPECIMENS EXAMINED.— Fifteen males, 10 females.
NEW RECORDS. -FRESNO COUNTY: 10.2 mi (16.4 km)
SW Trimmer, 27 Jan. 1968 (KH, JG); 8.7 mi (14 km) SW
Trimmer, 27 Jan. 1968 (TSB, KH); 5.3 mi (8.5 km) SW Trim-
mer, 27 Jan. 1968 (TSB).
NATURAL HISTORY.— Found beneath serpen-
tine and granite rocks in oak woodland and grass-
land biomes.
THE SERPENTINEA SPECIES GROUP
DIAGNOSIS.— Males of this group have a dorsal
process on the glans penis. Females differ from
those in the digita and mariposa groups in having
an ovipositor with only apical microspines.
The minor Subgroup
DIAGNOSIS.— The morphology of the penis,
dorsal process with ventrally directed basal lobes
but no apical lobes, is diagnostic for the subgroup.
DISTRIBUTION.— Known only from San Mateo
County.
FIGURE 1 1 . Male genitalia of C. palapraeputia. (a, b) Lat-
eral and dorsal views showing partially expanded glans (5.3
mi SW Trimmer), (c) Ventral view showing fully expanded
glans (10.2 mi SW Trimmer).
Calicina minor (Briggs and Horn, 1 966),
new combination
(Fig. 4c, 12)
Sitalcina minor Briggs and Horn, 1966:263. Briggs, 1968:27.
DIAGNOSIS.— This paedomorphic species is
distinguished from all others by its unique glans
penis: basal segment with transverse rows of small
tubercles and dorsal process lacking apical lobes
but with prominent basal lobes.
ADDITIONAL DESCRIPTION.— Carapace with at
most 1 pair of anterior tubercles. Male palpal
tarsus with enlarged mesobasal spine-bearing tu-
bercle instead of spur. Penis with ventral plate
apically produced and bearing 7 pairs of setae;
dorsal process apically bifurcate, with ventrally
directed basal lobes; stylus straight and tapering,
basally thickened. Ovipositor with 6 pairs of api-
cal and 3 pairs of subapical setae; microspines
restricted to apex.
TYPE SPECIMENS. — Female holotype, male allotype, and four
paratypes (one male, three females), under serpentine rocks on
grassland hillside at 0.75 mi (1.2 km) N Crystal Springs Dam
on County Road No. 14, San Mateo County, 23 Jan. 1966,
TSB and KH.
SPECIMENS EXAMINED.— Twenty males, 23 females.
NEW RECORDS. -SAN MATEO COUNTY: Edgewood Park,
serpentine grassland on N half of park, 2 Jan. 1 983 (TSB, VFL,
DU); 6 Apr. 1986 (TSB, TO); 2 Jan. 1987 (TSB, VFL, DU);
24 Dec. 1987 (TSB, VFL, WES, DU).
114
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
FIGURE 12. Male genitalia of C. minor (Edgewood Park), (a) Ventral view of penis shaft, (b-d) Ventral, dorsal, and lateral
views.
DISTRIBUTION. — This species is presently
known only from Edgewood Park. No specimens
have been collected at the type locality (despite
repeated attempts) since the completion of the
adjacent highway. Also, the series of collections
from Santa Clara County listed by Briggs and
Horn (1966) represent an undescribed species be-
longing to a related genus.
NATURAL HISTORY.— All collections are from
serpentine grassland.
The serpentined Subgroup
DIAGNOSIS. —This subgroup, the largest in Cal-
icina, includes those species whose males have
a glans with a ventral stylus and a dorsal process
bearing five apical lobes. The palpal spur, when
present, is small or vestigial.
DISTRIBUTION.— The subgroup has a disjunct
distribution with two populations in the lower
Sierra Nevada foothills and a third along the
central and northern Coast Ranges.
Key to Males of the Species of the
serpentined Subgroup
1 . Glans with a pair of basal lobes (Fig. 1 3a-
c, 18) piedrd
- Glans without basal lobes (Fig. 1 3d-m, 1 4)
2
2. Stylus longer than dorsal process (Fig. 1 3d-
m) 3
- Stylus shorter than dorsal process (Fig. 1 4)
5
3. Tarsal count 3-4-4-5 serpentined
- Tarsal count 3-4-4-4 4
4. Dorsal process with lateral lobes as long as
wide (Fig. 1 3d-f ) ensata
- Dorsal process with lateral lobes longer than
wide (Fig. 1 31, m) polind
5. Body with dark maculations macula
- Body concolorous 6
6. Dorsal process with lateral lobes as long as
wide (Fig. 14a-f) _ 7
- Dorsal process with lateral lobes longer than
wide (Fig. 14J-1) 8
FIGURE 13. Male genitalia of the serpentined subgroup. 1. (a-c) C. piedra (holotype), lateral, ventral, and dorsal views, (d-
f) C. ensata (Tollhouse), (d, e) Ventrolateral and dorsal views, (f ) Dorsolateral view of stylus tip. (g-k) C. serpentinea. (g, h)
Dorsal and ventral views (Monte Vista), (i-k) Ventrolateral, dorsal, and ventral views (San Felipe Road). (I, m) C. polina
(Novato), dorsolateral and ventral views. (Scale bar equals 0.25 mm, except for "f" where it equals 0.10 mm.)
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
115
116
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
7. Median lobe of dorsal process apically
notched (Fig. 14d-f) basalta
- Median lobe of dorsal process apically
pointed (Fig. 14a-c) breva
8. Tarsal count 3-4-4-5 dimorphica
- Tarsal count 3-4-4-4 mesaensis
Calicina piedra (Briggs, 1968),
new combination
(Fig. 13a-c)
Sitaldna piedra Briggs, 1968:24.
DIAGNOSIS.— This species, the largest of the
serpentinea group, is easily distinguished by the
structure of the male genitalia: dorsal process of
glans subequal in length to stylus, with a pair of
basoventral lobes, and with small apical lobes.
ADDITIONAL DESCRIPTION.— Carapace with 2
or 3 pairs of anterior tubercles. Penis with ventral
plate apically attenuated and with 5 pairs of lat-
eral setae; stylus flat, broad, and sinuous. Ovi-
positor as in minor but without subapical setae.
TYPE SPECIMENS. — Male holotype, female allotype, and fe-
male paratype, under rocks in an oak grassland at 1 .6 mi (2.6
km) SW Piedra, Fresno County, 21 Jan. 1967, TSB, KH, and
AJ.
SPECIMENS EXAMINED.— Only the type series.
Calicina serpentinea (Briggs and Horn, 1 966),
new combination
(Fig. 3b, 5g, h, 13g-k, 16g, h)
Sitaldna serpentinea Briggs and Horn, 1966:268. Briggs, 1968:
26.
DIAGNOSIS.— This species differs from others
in its group by the combination of a tarsal count
of 3-4-4-5 and structure of the male genitalia
(stylus longer than the dorsal process and median
lobe narrow and elongated).
ADDITIONAL DESCRIPTION.— Carapace with 2
or 3 pairs of anterior tubercles. Eyes variable,
cornea present or absent. Tarsal spur of male
palpi may be absent or extremely small. Penis
with ventral plate bearing 7 pairs of setae; dorsal
process with narrow median lobe (with or with-
out an apical notch) and rounded lateral lobes.
Ovipositor with apical microspines, with 6-8
pairs of apical setae, and with one pair of apical
teeth.
VARIATION. — Calicina serpentinea is poly-
morphic with regard to several characters. In-
dividuals from the three westernmost localities
(dense forest biomes) lack corneas and tarsal
spurs, which are present in the remaining pop-
ulations (grassland and oak woodland biomes).
The female from Monte Vista lacks the paired
apical teeth on the ovipositor. In individuals from
Santa Clara and Contra Costa counties, the me-
dian lobe is apically entire (Fig. 1 3g, h), in others
it is apically notched (Fig. 1 3i-k).
JUVENILES.— The two known juveniles, from
Bob's Secret Cave, are middle instars having tar-
sal counts of 2-2-3-3. They lack anterior tuber-
cles. As in adults from this locality, the eyes are
completely reduced (lacking both retina and cor-
nea) and the free tergites lack enlarged spines.
TYPE SPECIMENS. — Female holotype, male allotype, and six
paratypes (two males, four females), under serpentine in marshy
grassland along Silver Creek Road, 5 mi (8 km) SE Tully Road,
San Jose, Santa Clara County, 27 Feb. 1966, TSB and CF.
SPECIMENS EXAMINED. — Fifty-one males, 38 females, two ju-
veniles, seven specimens of undetermined sex.
NEW RECORDS. -SANTA CLARA COUNTY: Silver Creek
Road, 3.6 mi (5.8 km) from San Felipe Road, 2 Jan. 1983
(TSB, VFL, DU); Silver Creek at San Felipe Reservoir, 27
Nov. 1966 (TSB, AKSJ); Monte Vista, near Permanente Quar-
ry, 24 Mar. 1 968 (TSB); Metcalf Road, 1 .8 mi (2.9 km) N Hwy
101,2 Jan. 1983 (TSB, VFL, DU). SANTA CRUZ COUNTY:
Cave Gulch, Bob's Secret Cave, 6 May 1984 (TSB, VFL, DU);
Cave Gulch, 1 8 June 1 984 (TSB, DU). SAN BENITO COUN-
TY: Lone Tree Road, 8.6 mi (13.8 km) W Fairview Road, 27
Dec. 1984 (TSB, VFL, DU).
DISTRIBUTION.— Known from the southern Bay
Area from San Mateo and Contra Costa to San
Benito counties.
NATURAL HISTORY. — This wide-ranging
species occurs in a variety of habitats; most com-
monly beneath serpentine in grassland and oak
woodland biomes, but also from redwood,
broadleaf evergreen, and digger pine-oak asso-
ciations. There is also one collection from a cave.
Most specimens were taken from under serpen-
tine, but some were also collected beneath lime-
stone and sandstone rocks, and fallen bark. While
more commonly collected during the rainy sea-
son, this species is apparently active throughout
the year in mesic refugia.
Calicina polina (Briggs, 1968),
new combination
(Fig. 131,m)
Sitaldna polina Briggs, 1968:30.
DIAGNOSIS.— A unique dorsal process with a
cordate median lobe and narrow lateral lobes
distinguishes males of this paedomorphic species
from all other species.
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
117
ADDITIONAL DESCRIPTION.— Carapace with 1
or 2 pairs of anterior tubercles. Penis with stylus
longer than dorsal process; ventral plate with 6
pairs of setae. Ovipositor with 7 pairs of apical
setae; microspines apical.
TYPE SPECIMENS.— Male holotype, female allotype, and three
paratypes (two males, one female), under serpentine rocks in
a moist grassland at El Polin Spring, Presidio, San Francisco
County, 11 Dec. 1966, TSB.
SPECIMENS EXAMINED.— Forty-six males, 31 females, one
specimen of undetermined sex.
NEW RECORDS. -SAN FRANCISCO COUNTY: Presidio
Golf Course, 9 Nov. 1975 (TSB). MARIN COUNTY: 1 mi
(1.6 km) NW Novato, 2 Apr. 1967 (TSB); Marin Dell Ranch,
near Novato, 2 Jan. 1972 (TSB, AK); Novato, San Marin
Drive, 7 Mar. 1982 (TSB, DU), 14 Mar. 1982 (DU); Novato,
W of Burdell Mountain, 100 m (350 ft), 2 Jan. 1986 (TSB,
DU); Novato, SW of Burdell Mountain, 170 m (550 ft), 2 Jan.
1986 (TSB, DU). SONOMA COUNTY: 2.5 mi (4 km) E Shell-
ville, 2 Apr. 1967 (TSB); ALAMEDA COUNTY: Cull Canyon
Road, 1 km N Crow Canyon Road, 22 Jan. 1984 (TSB, DU);
Cull Canyon Road, 1.5 mi (2.4 km) N Crow Canyon Road,
22 Jan. 1984 (TSB, DU); Woolsey Canyon, Berkeley, 21 Dec.
1 983 (TSB, VFL, DU); Redwood Road, 3. 1 mi (5 km) N Castro
Valley Road, 22 Jan. 1984 (TSB, DU); Chabot Reservoir at
Lake Chabot Road, 6 Apr. 1982 (TSB, DU). CONTRA COS-
TA COUNTY: Wildcat Canyon Park near Hill Road, 30 Jan.
1984 (TSB), 14 July 1984 (TSB).
DISTRIBUTION.— This species is found in the
northern part of the San Francisco Bay Region,
from Sonoma south to San Francisco and Ala-
meda counties.
NATURAL HISTORY. —The collections ofpolina
west of San Francisco Bay (Sonoma, Marin, and
San Francisco counties) are from serpentine
grassland, whereas those from the East Bay are
from beneath sandstone in oak forests.
Calicina ensata (Briggs, 1968),
new combination
(Fig. 4d, 13d-f)
Sitalcina ensata Briggs, 1968:21.
DIAGNOSIS. — Males of this paedomorphic
species may be distinguished from others in the
subgroup in having a stylus which is longer than
the dorsal process and has an apical prong.
ADDITIONAL DESCRIPTION.— Carapace with 2
or 3 pairs of anterior tubercles. Penis with ventral
plate bearing 7 pairs of lateral setae; dorsal pro-
cess with large lateral lobes; stylus long, tapering.
Ovipositor with 12 apical setae and 3 pairs of
subapical setae; microspines apical.
TYPE SPECIMENS. — Male holotype, female allotype, and 10
paratypes (eight males, two females), under rhyolite in an oak
woodland at 2.0 mi (3.2 km) NW Trimmer, Fresno County,
16 Apr. 1967, TSB.
SPECIMENS EXAMINED.— Sixteen males, seven females.
NEW RECORD.— FRESNO COUNTY: W entrance to Watts
Valley, 28 Jan. 1968 (TSB).
NATURAL HISTORY.— This species is known
from oak woodland and has been collected be-
neath granite, rhyolite, and serpentine rocks.
Calicina macula (Briggs, 1968),
new combination
(Fig. 14g-i)
Sitalcina macula Briggs, 1968:23.
DIAGNOSIS.— This is the only Calicina with a
dark marbled color pattern. Males of this pae-
domorphic species may also be distinguished by
the proportions of the dorsal process lobes and
in having enlarged mesoapical spine-bearing tu-
bercles on male palpal tarsi.
ADDITIONAL DESCRIPTION.— Carapace with 2
pairs of anterior tubercles. Penis typical for the
serpentinea group; dorsal process with notched
median lobe; ventral plate with 7 pairs of lateral
setae. Ovipositor with 1 2 apical and 3 pairs of
subapical setae; microspines apical.
TYPE SPECIMENS.— Male holotype, female allotype, and nine
paratypes (eight males, one female), under serpentine rocks in
an oak grassland at 9 mi (14.5 km) SE Academy, Fresno Coun-
ty, 16 Apr. 1967, TSB.
SPECIMENS EXAMINED.— Only the type series.
Calicina dimorphica Ubick and Briggs,
new species
(Fig. 14m, n)
DIAGNOSIS.— Males of this species are distin-
guished from all others in the group in having a
dorsal process with a diamond-shaped central
lobe and elongated lateral lobes. Males lack the
palpal spur but have an enlarged mesoapical
spine-bearing tubercle on the palpal tarsus.
ETYMOLOGY.— The specific name refers to the
sexually dimorphic palpal spination character-
istic of this species.
DESCRIPTION.— Color yellowish orange. Car-
apace with 3 pairs of anterior tubercles. Eyes well
developed. Tarsal count 3-4-4-5.
Male: Total body length, 1.22. Scute length,
0.86. Scute width, 0.90. Eye tubercle length, 0. 1 7.
Eye tubercle width, 0.18. Leg II length, 2.07.
Palpal tarsus without dorsal spur but with en-
larged mesoapical spine-bearing tubercle. Penis
118
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
h
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
119
with ventral plate bearing 7 pairs of setae, of
which 2 distal pairs are enlarged. Dorsal process
with long, tapering lateral lobes and diamond-
shaped median lobe.
Female: Total body length, 1 .00. Scute length,
0.63. Scute width, 0.60. Eye tubercle length, 0. 1 3.
Eye tubercle width, 0.12. Palpal tarsus without
enlarged spine-bearing tubercle. Ovipositor with
apical microspines, 12 apical setae, and 1 pair
of subapical setae.
TYPE SPECIMENS.— Male holotype and three paratypes (two
males, one female) under granite in an oak grassland at the NE
entrance to Watts Valley, Fresno County, 28 Jan. 1968, TSB.
SPECIMENS EXAMINED.— Only the type series.
Calicina mesaensis Ubick and Briggs,
new species
(Fig. 4b, 14J-1)
DIAGNOSIS. — Males of this paedomorphic
species differ from all others in having a dorsal
process with a quadrate median lobe and elon-
gate lateral lobes.
ETYMOLOGY. —The specific name is a reference
to the type locality, Table Mountain.
DESCRIPTION.— Color yellow-orange. Cara-
pace with 2 pairs of anterior tubercles. Eyes well
developed. Tarsal count 3-4-4-4.
Male: Total body length, 1.14. Scute length,
0.87. Scute width, 0.82. Eye tubercle length, 0. 1 8.
Eye tubercle width, 0.17. Leg II length, 2.02.
Palpal tarsus with greatly reduced dorsal spur
(Fig. 4b). Penis with 7 pairs of setae on ventral
plate. Dorsal process with quadrate median lobe
and tapering lateral lobes. Stylus with apical
prong.
Female: Total body length, 1.13. Scute length,
0.78. Scute width, 0.71. Eye tubercle length, 0.15.
Eye tubercle width, 0. 1 3. Ovipositor with apical
microspines and 7 pairs of apical setae.
TYPE SPECIMENS.— Male holotype and 25 paratypes (1 1 males,
14 females), under basalt rocks in an oak grassland at Table
Mountain, 2.8 mi (4.5 km) N Millerton Lake Road on Sky
Harbor Road, Fresno County, 31 Mar. 1985, TSB, TO, WCR,
and DU. Two paratypes (male, female) deposited at the AMNH
collection.
SPECIMENS EXAMINED.— Only the type series.
Calicina breva (Briggs, 1968),
new combination
(Fig. 14a-c)
Sitalcina sierra breva Briggs, 1968:21.
DIAGNOSIS. — Males of this paedomorphic
species may be distinguished by the structure of
the dorsal process: median lobe pentagonal, sub-
equal to the laterals, and ventral plate with only
5 pairs of setae.
ADDITIONAL DESCRIPTION.— Carapace with 1
or 2 pairs of anterior tubercles. Penis with ventral
plate bearing 5 pairs of small setae; dorsal process
with three subequal, cloverlike lobes; stylus short
and straight. Ovipositor with apical microspines
and 7 pairs of apical setae.
TYPE SPECIMENS.— Male holotype, female allotype, and five
paratypes (three males, two females), under basalt rocks in a
grassland at 1 mi (1 .6 km) S Knight's Ferry, Stanislaus County,
11 Apr. 1967, TSB.
SPECIMENS EXAMINED.— Only the type series.
Calicina basalta Ubick and Briggs,
new species
(Fig. I4d-f)
DIAGNOSIS.— This paedomorphic species is
most closely related to breva from which it may
be distinguished by the notched median lobe of
the dorsal process of the penis.
ETYMOLOGY.— The specific name refers to the
microhabitat of this species.
DESCRIPTION.— Color yellow. Carapace with 1
or 2 pairs of anterior tubercles. Retina occasion-
ally absent. Tarsal count 3-4-4-4.
Male: Total body length, 1.08. Scute length,
0.70. Scute width, 0.77. Eye tubercle length, 0. 1 3.
Eye tubercle width, 0.15. Leg II length, 2.07.
Palpal tarsus without dorsal spur. Penis similar
to that of breva except that dorsal lobe has notched
median lobe.
Female: Total body length, 0.86. Scute length,
0.60. Scute width, 0.63. Eye tubercle length, 0. 1 2.
Eye tubercle width, 0. 1 2. Ovipositor with apical
microspines and 7 pairs of apical setae.
FIGURE 14. Male genitalia of the serpentinea subgroup. 2. (a-c) C. breva (paratopotype), lateral, dorsal, and ventral views,
(d-f) C. basalta (holotype), lateral, dorsal, and ventral views, (g-i) C. macula (holotype), sublateral, dorsal, and ventral views
showing retracted glans. (j-1) C. mesaensis (holotype), lateral, dorsal and ventral views, (m, n) C. dimorphica (paratopotype),
lateral and ventral views.
120
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
FIGURE 1 5. Male genitalia of the sequoia subgroup, (a-e) C. diminua (paratopotype). (a, b) Dorsal and lateral views, (c-e)
Ventral, dorsal, and lateral views, (f-p) C. sequoia, (f) Ventral view of penis, truncus (Little River), (g-j) Lateral to ventral views
(Little River), (k, 1) Dorsal and ventral views (Pine Ridge Road), (m) Lateral view showing retracted glans (Mendocino). (n-p)
Lateral, dorsal, and ventral views (Usal Creek). (Scale bar equals 0.25 mm, except for "c-e" where it equals 0.10 mm).
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
121
TYPE SPECIMENS. — Male holotype and five paratypes (three
males, two females), under basalt rocks in a grassland area 8
mi (12.9 km) SW Jamestown, Tuolumne County, 10 Feb. 1968,
TSB. Six paratypes (three males, three females) taken in similar
conditions at 1 3 mi (2 1 km) E Oakdale, Stanislaus County, 1 0
Feb. 1968, TSB and BL.
SPECIMENS EXAMINED.— Seven males, five females.
The sequoia Subgroup
DIAGNOSIS.— The presence of at least 5 pairs
of regularly spaced lateral setae on the ventral
plate and the extremely complex dorsal process
are diagnostic for this subgroup.
DISTRIBUTION.— Known from the northern
Coast Ranges of California.
Calicina diminua Ubick and Briggs,
new species
(Fig. 15a-e)
DIAGNOSIS.— This paedomorphic species, the
smallest Calicina, is most easily distinguished by
the elaborate structure of the glans penis.
ETYMOLOGY. —The specific name is a reference
to the small size of this species.
DESCRIPTION.— Color yellow. Carapace with-
out anterior tubercles. Eyes, both retina and lens,
absent. Tarsal count 3-4-4-4.
Male: Total body length, 0.77. Scute length,
0.49. Scute width, 0.53. Eye tubercle length, 0.08.
Eye tubercle width, 0.09. Leg II length 1 .4 1 . Pal-
pal tarsus with dorsal spur. Penis small; ventral
plate with 5 pairs of lateral setae and two triads
of ventral setae; dorsal process ornate; stylus short
and spinelike.
Female: Total body length, 0.60. Scute length,
0.46. Scute width, 0.46. Eye tubercle length, 0.06.
Eye tubercle width, 0.07. Ovipositor with only
10 apical setae and without subapical setae or
microspines.
TYPE SPECIMENS. — Male holotype and 11 paratypes (four
males, seven females), under serpentine on a grassland hillside
at Novato, San Marin Drive, Marin County, 2 Jan. 1986, TSB
and DU. Five paratypes, 0.5 km SW of type locality, 7 Dec.
1985 (male) and 28 Apr. 1968 (three males, one female), TSB.
Two paratypes (male, female) deposited at the AMNH collec-
tion.
SPECIMENS EXAMINED.— Nine males, eight females.
Calicina sequoia (Briggs and Horn, 1966),
new combination
(Fig. 15f-p)
Sitalcina sequoia Briggs and Horn, 1966:267. Briggs, 1968:28.
DIAGNOSIS.— In addition to their distinctive
male genitalia (dorsal process large with 1 or 2
pairs of lobes and ventral plate with 6 to 8 pairs
of lateral setae) these small phalangodids have
the most reduced tarsal count of any Calicina
(3-4-4-4 or less), but unlike most other paedo-
morphic species, have well developed eyes.
ADDITIONAL DESCRIPTION.— Tarsal count usu-
ally 3-3-4-4, occasionally 3-4-4-4. Carapace with
1 pair of anterior tubercles. Penis with ventral
plate bearing 6-8 pairs of lateral setae and 2 or
3 pairs of ventral setae; dorsal process with two
pairs of lobes (the basal pair sometimes missing)
and an apical portion capable of hyperextension;
stylus ventral, spinelike. Ovipositor with apical
microspines, 6 pairs of apical setae, 2 pairs of
subapical setae, and 1 pair of apical teeth.
VARIATION.— The five specimens from the
southernmost localities (Pine Ridge Road and
Talmage) have the higher tarsal count of 3-4-4-
4 and lack dorsobasal lobes on the dorsal process
(Fig. 1 5k, 1) but, otherwise, do not differ in gen-
italic features.
JUVENILES.— The only known juvenile, from
Mendocino, is an early instar having a tarsal count
of 1-1-2-2. As in adults, the eyes are well de-
veloped. However, the anterior tubercles are ab-
sent and the free tergites are armed with robust
spines which are absent in adults.
TYPE SPECIMENS. — Female holotype, male allotype, and six
paratypes, under rocks and logs in a Douglas fir forest at 2.3
mi (3.7 km) S Piercy, Mendocino County, 13 Mar. 1966, TSB
andKH.
SPECIMENS EXAMINED.— Twenty- two males, 24 females, one
juvenile, seven specimens of undetermined sex.
NEW RECORDS. -MENDOCINO COUNTY: 2.3 mi (3.7 km)
5 Piercy, 17-18 June 1966 (KH, VFL); Mendocino, 4 May
1963 (OCR); 5 mi (8 km) S Usal Creek, 17 Apr. 1976 (TSB);
Pine Ridge Road, 1.6 mi (2.6 km) S Low Gap Road, W of
Ukiah, 10 Feb. 1985 (TSB); Casper, 3 Aug. 1957 (JRH, GAM;
AMNH collection); Talmage, 29 July 1959 (LMS, ROS; UCB
collection).
NATURAL HISTORY.— This species is known
from redwood and Douglas fir forests and has
been collected beneath both logs and rocks. Six
samples contained individuals of "Sitalcina"
cockerelli Goodnight and Goodnight.
PHYLOGENY
MONOPHYLY OF CALICINA
Calicina is believed to be monophyletic on the
basis of its unique palpal tarsus and possible syn-
apomorphies in the ovipositor and glans mor-
phologies. Among the Nearctic phalangodids,
only the species of Calicina have a telescoping
122
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
FIGURE 16. Calicina, glans types (a, c, e, g = lateral views;
b, d, f, h = dorsal views), (a, b) C. mariposa, ST = stylus, CL
= collar lobe, (c, d) C. digita, PS = parastylus. (e, f) C. pala-
praeputia, MS = middle segment, (g, h) C. serpentined, DP =
dorsal process.
glans (Fig. 1 6). Representatives of all other nom-
inal genera were examined and found to have a
folding glans. At this time it is not possible to
determine which of these character states is de-
rived. Outgroup comparison cannot be used since
the sister group of these (presumably closely re-
lated) genera is not known. It may be argued that
the folding glans, because of its widespread oc-
currence, is plesiomorphic. However, the tele-
scoping glans is functionally simpler, whereas the
folding glans is usually associated with other de-
rived genital characters, such as the bifurcate
ventral plate ofBanksula Roewer, Texella Good-
night and Goodnight, and all Appalachian gen-
era.
The tarsal spur is not known in any other pha-
langodid genus and is, clearly, a synapomorphy
for Calicina (Fig. 17, component 1). However,
several species that we include in Calicina lack
the spur, which we interpret to be the result of
a character reversal. The tarsal spur is most
strongly developed in the two least derived taxa
of Calicina (the mariposa group and the digita
subgroup) but is reduced in size or absent in most
of the remaining species. Of the 1 1 species that
completely lack the spur, most are closely related
(based on genitalic characters) to species having
distinct, but small, spurs (the seven species be-
longing to the kaweahensis, serpentinea, and se-
quoia subgroups). Likewise, on the basis of their
glans structure, the two species in the arida
subgroup and the monotypic minor subgroup are
clearly representatives of the digita and serpen-
tinea species groups, respectively. This leaves only
one unassociated species, palapraeputia, that
lacks the spur. However, despite the unique glans
structure of this species, its generic placement is
not questioned because, in addition to having a
telescoping glans, its ovipositor closely resembles
that of some Calicina species in microspine dis-
tribution and setal arrangement (compare Fig.
5e, f and 5g, h).
The ovipositor of Calicina differs from that of
Sitalcina in several characters (Table 1). Of these,
microspines may be derived because, with few
exceptions, they do not occur in other Califor-
nian phalangodid species. Also, two setal series
(apical and subapical) present in most Calicina
species have not been observed in Sitalcina nor
in other Californian phalangodid genera. Further
investigation will be necessary to determine the
polarities of these character states and their uni-
versality.
SISTER GROUP
There are two possible sister groups of Cali-
cina. The first consists of the genera which have
a telescoping glans. Of the European phalango-
dids, apparently only Ptychosoma Soerensen has
this type of glans (see figures in Brignoli 1968).
Two additional species with telescoping glans
have recently been described in Scotolemon Lu-
cas (espanoli Rambla, 1973 and balearicus Ram-
bla, 1977) but, based on the published genital
illustrations, appear to be congeneric with Pty-
chosoma. Although similar to Calicina in general
body structure, Ptychosoma differs in having a
greater number of palpal spine-bearing tubercles,
in having a strongly modified ventral plate, and
in lacking sexually dimorphic structures.
A telescoping glans is likewise found in some
Southeast Asian genera. Buparellus Roewer and
Bupares Thorell have a glans structure quite sim-
ilar to some Calicina species, but they differ in
having more strongly armed palpi (see figures in
Suzuki 1985). The Japanese Parabeloniscus Su-
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
123
TABLE 2. LIST OF CHARACTERS AND THEIR PRESUMED POLARITIES.
Character
Plesiomorphic
Apomorphic
1.
male palpal tarsus
unmodified
with spur
2.
parastylus
absent
present
3.
stylus shape
sinuous
straight
4.
parastylus insertion
ventral
(a) lateral, (b) dorsal
5.
microspine distribution
distal half of ovipositor
restricted to apex
6.
glans segments
two
three
7.
dorsal process (DP)
absent
present
8.
DP apex
bifurcate
lobate
9.
basoventral lobes of DP
present
absent
10.
median lobe length
shorter than paramedian lobes
longer than paramedian lobes
11.
stylus length
subequal to DP
(a) longer than DP
(b) shorter than DP
12.
lateral lobe shape
length = width
length > width
zuki shares with Calicina a similar glans and
ventral plate spination but has increased palpal
spination and a sexually dimorphic third tergite
(elongated in males) (see figures in Suzuki 1 973).
Although placed in the Phalangodinae, these gen-
era also lack the characteristic eye tubercle and
may belong elsewhere.
Finally, the genitalia of Assamiidae and Bian-
tidae strongly resemble those of some Calicina
species (see figures in Martens 1977, 1986).
Members of these families, however, have highly
modified palpi (segments are elongated and at-
tenuated and spination is reduced) and generally
lack well defined eye tubercles.
The relationship of Calicina to the above taxa
depends on the phylogenetic significance of the
telescoping glans. If this glans structure is de-
rived, then these and additional taxa may ac-
tually be closely related. On the other hand, if
the telescoping state is plesiomorphic, then dif-
ferences in somatic characters assume greater
importance. In this case, the somatically similar
Californian phalangodids become the second
group of potential candidates for the Calicina
sister group. Of these, the most likely possibility
is a recently discovered (and undescribed) phal-
angodid genus that has a glans that unfolds and
telescopes during expansion. This relationship
will be explored in a future study.
INTRAGENERIC RELATIONSHIPS
The proposed relationships among the Cali-
cina species are based primarily on the male gen-
italia. The characters used, along with their pre-
sumed polarities, are given in Table 2 and the
resulting cladograms are presented in Figures 1 7
and 18.
As described earlier, four types of glans occur
in Calicina. These appear to define the major
lineages that are here recognized as species groups.
Each group is defined by unique structures or
character states, interpreted as autapomorphies:
the mariposa group by dorsally grooved collar
lobes, the digita group by parastyli, the pala-
praeputia group by an additional glans segment,
and the serpentinea group by a dorsal process
(Fig. 16).
The mariposa group contains species with the
simplest glans morphology (Fig. 6, 16a, b). This
glans contains two segments: the basal bears a
pair of collar lobes and the apical consists of a
sinuous stylus. This structural simplicity is as-
sumed to be plesiomorphic. The presence of ple-
siomorphic somatic characters in this group is at
least consistent with this interpretation. First,
Calicina mariposa has a tarsal count of 3-5-5-5.
With the exception of a few species in the digita
group, all other Calicina species have lower
counts. High tarsal counts, 3-5-5-5 or higher,
occur in virtually all other Holarctic phalango-
dids. Second, the males of all species in the mar-
iposa group have well developed tarsal spurs.
This state, as argued previously, is presumed to
be plesiomorphic within the genus.
The glans in the digita group is similar to that
in the mariposa group, but contains an additional
pair of sclerites, the parastyli (PS) (Fig. 1 6c, d,
component 2). The digita subgroup has ventral,
scalelike parastyli and a sagittate ventral plate
(Fig. 7). It is the most generalized element of the
group because it retains the high tarsal count,
124
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
Calicina subgroups
mar dig kaw top ari pal min ser seq
A(R)
an
R
A(R)
R
Ab
P
P
P
Calicina serpentinea subgroup
pie dim mes mac bre bas ens pol ser
A(R) A(R)
P P
A(R) A
9.
10.
11.
12.
P = plesiomorphy
A = apomorphy
a,b = alternate apomorphic states
R = character reversal
(R) = reversal in some species
C. piedra
FIGURE 17. Cladogram with data matrix showing the re-
lationships of Calicina subgroups using characters given in
Table 2. Abbreviations: mar = mariposa, dig = digita, kaw =
kaweahensis, top = topanga, ari = arida, pal = palapraeputia,
min = minor, ser = serpentinea, and seq = sequoia.
well developed tarsal spurs, and a mariposa-like
sinuous stylus. The stylus of the kaweahensis and
topanga subgroups is straight, and considered to
be derived (component 3). In the former, the PS
are broad, ornate, and laterally inserted (Fig. 8,
component 4a). In the topanga subgroup the PS
are rodlike and inserted dorsolaterally to the sty-
lus (Fig. 9). The two species of the arida subgroup
have an unusual glans morphology for their group
(Fig. 10). In arida the PS are represented by a
single dorsal rod, whereas in doughensis they are
totally absent. Despite the absence of PS, dough-
ensis is undoubtedly a member of this subgroup
because of its close relationship with arida. (See
the diagnosis of the subgroup for synapomor-
phies.) Using the digita subgroup as an outgroup,
the dorsal position of the PS in the arida and
topanga subgroups is synapomorphic (compo-
nent 4b).
Calicina palapraeputia, the sole representative
FIGURE 18. Cladogram with data matrix showing the re-
lationships of species in the serpentinea subgroup using char-
acters given in Table 2. The drawing is the glans, ventral view,
of C. piedra. Abbreviations: pie = piedra, dim = dimorphica,
mes = mesaensis, mac = macula, bre = breva, bas = basalta,
ens = ensata, pol = polina, ser = serpentinea, DP = dorsal
process, ML = median lobe, PML = paramedian lobe, LL =
lateral lobe, BL = basal lobe, and ST = stylus.
of its group, differs from all other Calicina species
in having a three-segmented glans (Fig. 11, 1 6e,
f ). If the stylus and basal segment of this glans
are homologous to comparable parts of the mar-
iposa glans, the middle segment is left unasso-
ciated. One possibility is that this segment rep-
resents a highly modified PS that enveloped and
fused around the stylus, as appears to occur in
kaweahensis (Fig. 8e-g). Alternatively, the mid-
dle segment could have developed from similarly
modified collar lobes from an arida-like ancestor
(Fig. lOe-h). Either of these mechanisms would
probably be accompanied by the fusion of these
structures (PS or collar lobes) on the middle seg-
ment, but no evidence of this has been detected.
Furthermore, both interpretations require the in-
dependent acquisition of the middle segment
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
125
lobes in palapraeputia. A more parsimonious al-
ternative (and the one we accept) is that the lobes
in palapraeputia are actually collar lobes, ho-
mologous to those found in the mariposa and
iigita groups. In this case the palapraeputia glans
:ould be derived from a mariposa type through
the development of an additional collar on the
3asal segment of the latter.
In the serpentinea group the glans has a de-
*ived, lobe-bearing structure, the dorsal process
'Fig. 1 6g, h, component 7). In most species (the
serpentinea subgroup) the dorsal process (DP)
sears five apical lobes (Fig. 13, 14). The remain-
ng three species all have unique DP morphol-
Dgies. Calicina minor, the sole representative of
ts subgroup, has the simplest DP bearing a pair
:>f dorsobasal lobes and an apical bifurcation in-
stead of lobes (Fig. 1 2). Of the members of the
sequoia subgroup, the DP of diminua is divided
.nto two lateral, extremely ornate, hemispheres;
whereas in sequoia it contains a pair of basal and
ipical lobes in addition to a complex apical re-
>ion (Fig. 1 5). Despite fundamental differences
n the glans structure, the last two species seem
:o be related as they share a unique ventral plate
spination. The presence of apical lobes on the
DP appears to be a synapomorphy uniting all
species of the serpentinea group except minor
component 8).
The species of the serpentinea subgroup all have
•ather similar glans morphologies, with the ex-
option of piedra. The piedra glans differs from
;he others in four characters: (1) DP with basal
obes, (2) paramedian lobes longer than median
obe, (3) lateral lobes wider than long, and (4)
stylus subequal to DP in length (Fig. 1 3a-c, 1 8).
Because basal lobes are found in the other
subgroups, their presence is a synapomorphy for
;he entire serpentinea group, but a plesiomorphy
it the subgroup level. The implication that piedra
,s the most generalized species in the subgroup
s compatible with its relatively higher tarsal count
3-4-4-5). The remaining species in this subgroup
ire united in lacking basal lobes (component 9)
ind in having relatively larger median lobes
[component 10). They are subdivided into two
groups based on stylus length. Using piedra as
the outgroup, stylus lengths either longer (com-
ponent 1 la) or shorter (component 1 Ib) than the
DP appear to be synapomorphic. Finally, elon-
gated lateral lobes (component 12) are synapo-
morphic for some species.
Because of the serpentinea group's unique glans
morphology, its relationship to the other groups
would appear remote. However, its ovipositor
structure implies a close relationship with pa-
lapraeputia. In both groups the distribution of
microspines is restricted to the apical surface of
the ovipositor which, using mariposa as the out-
group, is synapomorphic (component 5). This
relationship suggests that the glans morphologies
in the two groups are likewise closely related;
namely, that the dorsal process (DP) is homol-
ogous to the palapraeputia middle glans segment
(component 6). The DP could conceivably have
developed from the middle segment by the en-
largement of the stylus opening along the ven-
troapical surface. In fact, the basal lobes of the
DP, evident in members of all subgroups, may
well be vestiges of the ventral surface of the mid-
dle segment.
Two alternative explanations for the origin of
the DP are possible. First, the DP might repre-
sent highly derived parastyli. In kaweahensis, for
example, the parastyli are greatly enlarged, par-
tially envelop the stylus, and are fused dorsally
(Fig. 8e-g). Further enlargement could result in
a DP. However, this interpretation requires the
concomitant loss of collar lobes in the serpenti-
nea group. Second, the DP might be homologous
to collar lobes. In arida the collar lobes are large
and apically produced (Fig. lOe-h) and could
conceivably form a DP through fusion and ad-
ditional enlargement. However, this interpreta-
tion requires a character reversal of parastyli in
the serpentinea group. Furthermore, both of these
explanations imply a sister group relationship
between the serpentinea group and either ka-
weahensis or arida which is not supported by any
other known character.
BlOGEOGRAPHY
As is to be expected of cryptozoic organisms
having low dispersal potential, the species of Cal-
icina are allopatric or parapatric (see Fig. 19).
The four known instances of sympatry, between
palapraeputia and ensata, diminua and polina,
digita and kaweahensis, and digita and clough-
ensis, are all between members of different groups
or subgroups. Furthermore, these cases of sym-
patry may not actually involve interspecific in-
teraction. In the last case, for example, clough-
ensis is troglobitic and digita epigean. Also,
though diminua and polina live in close prox-
126
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
sequoia-
diminu
polina
minor-
sierra
serpentinea
arida
morroensis
topanga
\
conif era
yosemitensis
mariposa
\
dimorphica
palapraeputia
digita
cloughensis
galena
FIGURE 19. Map of central California showing the distribution of species of Calicina. A black dot indicates the localities of
paedomorphic species, a circle those of non-paedomorphics. Overlapping symbols denote sympatry.
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
127
FIGURE 20. Distribution of the subgroups ofCalicina and a cladogram of their relationships. A dotted line connects disjunct
elements and a dashed line indicates the presumed boundaries of the Sierran exotic terranes (from Hendrickson 1 986). Abbre-
viations: K = kaweahensis, M = minor, and P = palapraeputia.
128
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
dimorphica
O ensata
"7" piedra
FIGURE 2 1 . Map of central California showing the distribution of the species of the serpentined subgroup with a cladogram
of their relationships. The 'x' indicates the presumed time of the subgroup's disjunction by the Central Valley. Abbreviation:
m = macula.
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
129
imity, they have never been collected in the same
rock outcrops.
An area cladogram of the groups and subgroups
is given in Figure 20. Two gradients are apparent.
An altitudinal gradient exists between the mar-
iposa, digita, and serpentined groups, which oc-
cupy progressively lower elevations. A latitudi-
nal gradient is also evident: the digita group occurs
primarily along the southern mountains and
coast, whereas the serpentinea group occupies the
central and northern parts of the state.
The distribution pattern of the subgroups of
Calicina is compatible with an origin through
vicariance. The initial speciation events are like-
ly to have taken place in the Sierra. The Sierran
component of Calicina contains representatives
of all four groups, including the most generalized
members (mariposa group, digita subgroup, and
piedrd). The initial speciations may have resulted
from two barriers (given the basal trichotomy)
located between the ranges of the mariposa, dig-
ita, and serpentinea groups.
It is worth noting that these presumed barriers
closely correspond to the proposed boundaries
of exotic terranes given by Hendrickson (1986)
and indicated in Figure 20. The coastal Calicina
distributions also correspond closely to the maps
of exotic terranes. The species topanga and mor-
roensis each occupy separate terranes, and the
coastal distribution of the serpentinea group (with
the exception of the two westernmost serpentinea
localities) is located within the Franciscan com-
plex of terranes.
A few taxa exhibit disjunct distributions that,
in conjunction with our knowledge of geologic
history, may be used for estimating their mini-
mum ages. Thepolina distribution surrounds the
San Francisco Bay, suggesting that the species
predates the formation of the bay. Similarly, to-
panga occurs on Santa Cruz Island as well as the
adjacent mainland. If the currently disjunct dis-
tribution resulted from the formation of the San-
ta Barbara Channel, then the species must be
rather old. Recent estimates for a land connec-
tion to these islands are much older (although
unspecified) than the previous model of a Pleis-
tocene land bridge (Wenner and Johnson 1980).
It may be more plausible, however, that the San-
ta Cruz Island population represents a more re-
cent colonization from the mainland. Rafting is
a commonly evoked mechanism (for example,
by Yanev (1982) for slender salamanders) and a
good possibility for the forest-dwelling topanga.
The Central Valley is currently an obvious and
impenetrable barrier for Calicina (owing to the
absence of favorable mesic habitats) and divides
the distributions of two subgroups, arida and
serpentinea. The Sierran uplift, which formed the
Central Valley, is thought to have begun in the
early Pleistocene. A barrier could eventually have
resulted from the drying effects of glacial retreats,
as proposed for the populations of Ortholasma
levipes Shear and Gruber (1983:12). Our inter-
pretation of the serpentinea subgroup relation-
ships indicates that the Central Valley could have
been responsible for the disjunction of ensata
from serpentinea-polina (indicated by an "x" in
Fig. 21).
Another method of estimating the time of cla-
dogenic events in Calicina is to use the findings
for another organism from the same ancestral
biota. A good example is the slender salamander,
Batrachoseps. It appears very likely that Batra-
choseps and Calicina experienced a comparable
evolutionary history for three reasons: their sim-
ilar habitat preferences, congruent distributions,
and compatible phylogenies.
The ecological requirements of Batrachoseps,
as described by Yanev (1980) are remarkably
similar to those of Calicina. Both genera include
species restricted to forests or oak woodlands.
Both are specialized for subterranean life, appear
at the surface only during favorable periods, and
are able to survive in small isolates of suitable
habitat. Studies on Batrachoseps indicate that
individuals are extremely sedentary and have a
very low dispersal potential, as is probably the
case for Calicina.
The phylogeny of Batrachoseps consists of three
main lineages. The basal branch, currently rep-
resented by two disjunct isolates, is thought to
have formerly ranged along the eastern Sierra
Nevada, close to that of the present distribution
of the mariposa group. The second branch is
represented by B. attenuatus, which has a coastal
component that is almost identical to that of the
serpentinea group and a Sierran component in-
tersecting the distribution of breva and basalta.
The third branch contains the remaining four
species whose combined distribution is very sim-
ilar to that of the digita group.
On the basis of electrophoretic and immuno-
logical studies, Yanev (1980) proposed an origin
for Batrachoseps in the early Eocene (ca. 50 mya).
However, Hendrickson (1986) criticized these
values as being far too recent and suggested that
130
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
TABLE 3. FIELD DATA OF CALICINA.
Taxa
N'
Biome
Habitat
Season2
mariposa gp
10
sequoia-yellow pine
snags, logs
summer
digita
9
sequoia-yellow pine, oak wood, grass
logs, gra, sla
all year
digita (cave)
sierra
sierra (cave)
1
6
8
sequoia-yellow pine
digger pine-oak
digger pine-oak
cave
lim, sch, sla
caves
spr3
win-spr
all year
galena
kaweah
3
12
sequoia-yellow pine
oak wood
gra, logs
gra
summer
win-spr
topanga
9
broadleaf evergreen, digger pine-oak,
oak wood
logs, bas, snd, bre
win-spr3
kee
mor
1
4
oak wood
grass
gra
srp
win-spr
win-spr
clough
arida
3
2
oak wood
oak wood
cave
srp
srp-sum3
win-spr
pala
5
oak wood
gra, srp
win-spr
minor
4
grass
srp
win-spr
serp sbgp"
serp
serp (cave)
29
12
1
oak wood, grass
redwood-broadleaf evergreen, digger
pine-oak, oak wood, grass
redwood
spr, bas, gra, snd, rhy
logs, srp, lim, snd
cave
win-spr
all year
spr3
diminua
sequoia
3
11
grass
redwood-Douglas fir
srp
logs, rocks
win-spr
all year
Abbreviations: N = number of collection samples; gp = group; sbgp = subgroup; kaweah = kaweahensis; kee = keenea; mor
= morroensis; clough = cloughensis\ pala = palapraeputia\ serp = serpentinea; wood = woodland; grass = grassland; bas = basalt;
bre = breccia; gra = granite; lim = limestone; rhy = rhyolite; sch = schist; sla = slate; snd = sandstone; srp = serpentine; spr
= spring; sum = summer; win = winter.
' Roughly 3/4 of all samples include biome and habitat data.
2 Time of the year when adults are active.
3 Probably active all year.
4 Excluding serpentinea.
the isolation of Batrachoseps must have taken
place between the late Jurassic and the early Cre-
taceous. This conclusion was based on the cor-
relation between cladistic branching and geologic
events, especially the history of exotic terranes.
The geologic history of the Californian ter-
ranes is still largely unresolved. Most of the ter-
ranes are believed to have distant southern
origins, presumably in the vicinity of southern
North America. Their northward displacement
as isolated units is thought to have commenced
in the mid-Mesozoic. Subsequent accretions to
the North American Plate range from late Ju-
rassic to Cretaceous for the Sierran terranes, to
early Eocene for the southern coastal terranes,
and to late Miocene for the western elements of
the Franciscan complex. If the present distri-
bution pattern of Calicina is the result of isola-
tion on terranes, then a possible scenario is that
the original breakup of the Sierran terranes iso-
lated the ancestral populations of Calicina, which
eventually evolved into distinct species. Follow-
ing accretion to their present position, some dis-
persal from the terranes would be necessary to
account for the present distribution, especially
for the south Sierran members of the serpentinea
group. The later arrival of the coastal terranes
suggests widespread dispersal from the Sierran
populations.
NATURAL HISTORY
Information on the natural history of Calicina
comes from our field observations and the data
associated with some 1 30 collection samples. This
information, summarized in Table 3, is dis-
cussed below.
HABITAT.— Species of Calicina occur in mesic
habitats (in conditions of total darkness and high
humidity) but are apparently absent from those
areas that are periodically inundated or situated
in saturated soils. Most species are found exclu-
sively beneath medium to large-sized rocks that
are in contact with the soil and undisturbed. Oth-
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
131
er species also occur beneath decomposing logs.
One group of species (the mariposa group) is
found exclusively under logs and, more com-
monly, beneath and among fragments of fallen
bark that surround large snags. Four species have
been recorded from caves. One of these (clough-
ensis) is a troglobite, the others are troglophiles
with predominantly surface populations. These
habitat preferences of Calicina are similar to those
previously recorded for all Californian phalan-
godids (Briggs 1968) except that Calicina, unlike
some species of Sitalcina, has never been col-
lected in leaf litter.
Species of Calicina are associated with many
rock types, the most frequent being serpentine.
Of the 64 collection samples that indicate specific
rock associations, 28 are from serpentine. Other
frequently recorded rocks are granite (13 sam-
ples), basalt (seven), sandstone (seven), and lime-
stone (four, excluding cave collections). The Coast
Range species have been recorded primarily from
serpentine and sandstone, the Sierran ones from
granite and basalt.
BIOME.— The species of Calicina are known
from a variety of biomes (see also Fig. 22). Some
are apparently restricted to dense forests. Mem-
bers of the mariposa group and galena are known
only from high elevations (1,200-2,000 m) in
primary stands of giant sequoia-yellow pine as-
sociation. Calicina sequoia has been collected only
in redwood-Douglas fir association. Three ad-
ditional species have been collected in dense for-
ests. Calicina digita is recorded from giant se-
quoia-yellow pine, topanga and serpentinea from
broadleaf evergreen, the latter also in association
with redwood. However, these three species are
also known from more open forests (digger pine-
oak and oak woodland), as well as from grass-
land. All remaining species are known only from
oak woodland and grassland biomes.
ACTIVITY.— The period of adult surface activ-
ity varies between the species. Three patterns are
evident. (1) Species active during the rainy sea-
son (winter to spring). These species (14) live in
grassland and oak- woodland biomes (one species
from digger pine-oak). (2) Species active during
the summer. Species of the mariposa group and
galena, collected in the summer months, are
known only from giant sequoia-yellow pine for-
ests. (3) Species active throughout the year. The
residents of dense, low-elevation forests (sequoia
and forest populations of serpentinea, digita, and
possibly topanga) have a year-long adult activity
period. Likewise, the cavernicolous populations
of sierra have been collected throughout the year.
It is expected that the other cavernicolous Cal-
icina also have year-round activity.
LIFE CYCLE.— Almost all individuals of Cali-
cina encountered in the field and in collections
are adults. Despite repeated efforts to collect ju-
veniles, only three subadults are known from
epigean habitats (five additional juveniles are
known from caves). This absence of juveniles at
the surface suggests that development in Calicina
takes place within the soil. The appearance of
adults of species from drier environments co-
incides with the onset of the rainy season. Nor-
mally, only one or a few individuals are present
on each rock or log undersurface. However, in
several species we have observed aggregations of
1 0 or more individuals, often containing at least
one mating pair.
COHABITANTS.— The habitat of Calicina is
densely populated by a rich assortment of or-
ganisms. The most commonly observed insects
are campodeid diplurans and collembolans. Ob-
servations suggest that collembolans are the like-
ly prey ofBanksula (Briggs and Ubick 1 98 1) and
probably also of Calicina.
Several laniatorid opilionids live sympatri-
cally with Calicina. Zuma Goodnight and Good-
night is known from dense coniferous forests at
higher elevations of the Sierra Nevada. Sitalcina
is found in forests of both coastal and Sierran
regions. "Sitalcina" cockerelli Goodnight and
Goodnight lives in the northern coastal conif-
erous forests. Banksula is restricted to caves of
the central Sierra foothills, but has never been
collected sympatrically with Calicina.
In the central Coast Range and Sierran foothill
regions, numerous spiders are repeatedly col-
lected with Calicina. Some of the more common
(and distinctive) are leptonetids (Archoleptone-
td), telemids (Usofild), oonopids (Orchestina and
Scaphielld), amaurobioids (Blabomma, Calym-
maria, Pimus, and Titiotus) and many gnapho-
sids. Interestingly, these taxa, with the exception
of the oonopids, gnaphosids, and Pimus, all have
local cavernicolous representatives.
CONSERVATION.— Several Calicina species are
known only from single localities and may be
facing extinction as a result of habitat destruc-
tion. An extreme example is arida whose known
range consists of several square meters in a small
ravine, surrounded by extremely xeric or dis-
turbed environments. Several additional species
132
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
TABLE 4. THE RELATIONSHIP BETWEEN PAEDOMORPHIC AND
TROGLOBITIC CHARACTERS (TARSAL COUNT AND EYE Loss) AND
XERIC ENVIRONMENTS (OAK-GRASSLAND BIOMES) IN CALICINA.
Calicina species
Tarsal
count
Eyesight
Biome
mariposa gp.
mariposa sbgp.
mariposa
3-5-5-5
yosemitensis
3-4-4-5
conifera
3-4-4-5
blind
digita gp.
digita sbgp.
digita
3-5-5-5
xeric
(most)
sierra
3-4-4-4
xeric
kaweahensis sbgp.
kaweahensis
3-5-4-5
xeric
galena
3-5-5-5
(3-5-4-5)
topanga sbgp.
topanga
3-5-5-5
blind
(some)
keenea
3-4-4-5
xeric
morroensis
3-4-4-4
blind
xeric
arida sbgp.
cloughensis
3-5-5-5
blind
arida
3-4-4-4
blind
xeric
palapraeputia gp.
palapraeputia sbgp.
palapraeputia
3-4-4-5
xeric
serpentinea gp.
minor sbgp.
minor
3-4-4-4
blind
xeric
serpentinea sbgp.
piedra
3-4-4-5
xeric
ensata
3-4-4-4
xeric
serpentinea
3-4-4-5
blind
xeric
(most)
polina
3-4-4-4
blind
xeric
macula
3-4-4-4
blind
xeric
(some)
mesaensis
3-4-4-4
xeric
dimorphica
3-4-4-5
xeric
breva
3-4-4-4
blind
xeric
basalta
3-4-4-4
blind
xeric
(some)
sequoia sbgp.
diminua
3-4-4-4
blind
xeric
sequoia
3-3-4-4
(3-4-4-4)
are known from single localities: basalta, clough-
ensis, conifera, diminua, dimorphica, keenea,
macula, mesaensis, minor, and piedra. Conser-
vation programs will be necessary if these or-
ganisms and their habitats are to be protected.
ECOLOGY
Calicina exhibits considerable size-related in-
terspecific variation. Larger species have rela-
tively higher tarsal counts, darker pigmentation,
greater number of anterior tubercles, larger tarsal
spurs (of male palpi), and well developed eyes.
These structures are reduced in size or absent in
their smaller relatives. The possibility that one
of these character clusters is synapomorphic is
incompatible with our cladistic analysis, as nei-
ther group of species represents a clade. How-
ever, there appears to be a correlation between
reduced size and xeric biomes (see Table 4 and
Fig. 22). This section will explore the hypothesis
that the character transformations result from
two phenomena, paedomorphosis and troglob-
ism, and represent adaptations to xeric condi-
tions.
PAEDOMORPHOSIS.— The direction of the so-
matic character transformation can be inferred
by comparison with the ancestral state. As ar-
gued in the phylogeny section, the probable out-
group for the remaining species of Calicina
is mariposa, the species with the simplest
(=plesiomorphic) glans morphology. The forest-
dwelling mariposa is a large, strongly pigmented
species, which has a high tarsal count (3-5-5-5),
the greatest number of anterior tubercles (seven
to eight pairs), a large tarsal spur, and well de-
veloped eyes. That the character states associated
with large size are plesiomorphic is further sup-
ported by the condition in other Phalangodidae.
With the exception of troglobitic species, vir-
tually all remaining Nearctic phalangodids are
large, strongly pigmented forest dwellers with high
tarsal counts and well developed eyes. It is, thus,
reasonable to suppose that the reduction in size
and structure is derived.
Ontogenetic transformations were ascertained
indirectly by examining the juvenile morpholo-
gy. A total of eight specimens, representing five
species, were examined (attempts to collect ad-
ditional juveniles were not successful). All ju-
veniles are white, lacking any of the orange pig-
mentation found in adults, all lack tarsal spurs,
and all, with the exception of one penultimate
instar (galena), lack anterior tubercles. Based on
relative size, these specimens represent early (four
individuals), middle (two), and penultimate in-
stars (two). The only observable correlation with
growth, other than size, is an increase in tarsal
count. The earliest instars have tarsal counts of
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALICINA
133
4.0-1
3.0-
O
: Pi
2.5-
2.0-
1.5
TC < 3-4-U-U
si
O
• € O
br
pa
O
• O ka
TC> 3-4-U-5
:. dn
me
C /
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
TOTAL BODY LENGTH (MM)
FIGURE 22. A plot showing the distribution of paedomorphic and troglobitic characters in Calicina. The Total Body Length
and Leg II Length values are from the measurements of male holotypes (male allotypes of digita, kaweahensis, minor, and
serpentined). A circle indicates species having well developed eyes, a black dot denotes blind species, and a mixed symbol
indicates the species with partial retinal loss. The vertical line separates those species with tarsal counts of 3-4-4-4 or less
(paedomorphics) from those with higher counts. The forest-dwelling species are encircled by a solid line, those from oak woodland
and grassland by a dotted line, and those living in both densely forested and grassland biomes by a dashed line. Abbreviations:
a = arida, ba = basalta, br = breva, cl = cloughensis, co = conifera, dg = digita, dn = diminua, dr = dimorphica, e = ensata,
g = galena, ka = kaweahensis, ke = keenea, me = macula, me = mesaensis, mi = minor, mo = morroensis, mr = mariposa,
pa = palapraeputia, pi = piedra, po = polina, si = sierra, sq = sequoia, sr = serpentinea, t = topanga, and y = yosemitensis.
134
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
1 - 1 -2-2 (sierra and sequoia) and 1 -2-2-2 (dough-
ensis); later instars (serpentined) have 2-2-3-3;
and the oldest (galena), 3-5-4-5. Eyes are present
in species with eyed adults (sierra, galena, and
sequoia), absent in the others (doughensis and
serpentined). Despite the small sample size, it
seems probable that these observations apply to
all Calidna, given the generally conservative na-
ture of juvenile characters.
Using mariposa as the outgroup, the character
states found in the small species of Calidna are
derived. However, a comparison with their pre-
sumed ontogeny indicates that they are plesio-
morphic. The inescapable conclusion is that the
small species are derived through the retention
of juvenile characters; namely, that they are pae-
domorphic. The other possible alternative is that
the juvenile states are in reality character re-
versals; i.e., that penultimate instars have adult
somatic morphology. However, this latter pos-
sibility is clearly less parsimonious and not sup-
ported by the morphology of the two penulti-
mates examined.
Paedomorphosis is a widespread phenomenon
in Calidna. With the exception of three species
(mariposa, digita, and topangd), all species have
some degree of paedomorphic modification.
However, only the 1 2 species having the lowest
tarsal counts (3-4-4-4; 3-3-4-4 in sequoia) and
the high correlation to other characters (see Fig.
22) are here considered to be significantly pae-
domorphic. Given our phylogenetic interpreta-
tions, paedomorphic species are present in all
subgroups except mariposa, kaweahensis, and
palapraeputia, and must have evolved indepen-
dently at least nine times. Most of the paedo-
morphics (nine) belong to the serpentinea group;
the remainder belong to the digita group, where
they occupy the periphery of the group's distri-
bution (see Fig. 1 9).
Paedomorphosis has been well documented in
Opiliones. In her review, Rambla (1980) listed
many examples of the phenomenon in Caddidae,
Ischyropsalidae, Nemastomatidae, and Penta-
nychidae. Briggs (1986), in addition to pointing
out the presence of paedomorphic Calidna (as
Sitaldnd) species, included Triaenonychidae
among the examples.
TROGLOBISM.— Although only one species of
Calidna (doughensis) is an obligate cavernicole,
troglobitic characters are found in many species
of Calidna, including most paedomorphics. The
most obvious character is eye loss. Eight of the
paedomorphic species lack retinae, along with
three non-paedomorphics (excluding doughen-
sis). However, eye loss does not appear to be a
paedomorphic character, as juveniles of eyed
species have well developed eyes (for example,
the early instars of sierra and sequoia). Addi-
tional character states of paedomorphics, such
as pale coloration and small body size, could also
be troglobitic adaptations. Another common
troglobitic character is increased appendage
length. Preliminary results (Fig. 22) suggest that
the blind species of Calidna have slightly longer
leg II lengths than comparable-sized eyed ones.
SELECTIVE PRESSURES.— In his exhaustive study
of heterochrony, Gould (1977) differentiates two
distinct phenomena included in paedomorpho-
sis: progenesis and neoteny. Progenesis evolves
(through r selection) in harsh, unstable, density-
independent situations that favor rapid matu-
ration. This is achieved by shortening the life
cycle through premature maturation, and results
in adults juvenilized in both size and shape. Neo-
teny, on the other hand, evolves (through K se-
lection) in stable, density-dependent situations
that favor maintenance. Here the selection is not
for early maturation but for adaptive larval char-
acters. Adult neotenics are, therefore, juvenilized
in shape but not size.
The small size of paedomorphic Calidna
strongly suggests progenesis. As mentioned ear-
lier, all paedomorphic species (except sequoia)
reside exclusively in oak woodland and grassland
biomes, whereas non-paedomorphics occur pre-
dominantly in dense forests. The former biomes
are much more xeric and experience severe dry
seasons, which may favor both the small size and
shorter life cycle of progenetics. The presence of
troglobitic characters in paedomorphic Calidna
indicates a prolonged subterranean existence,
which would clearly favor a reduction in size.
Also, a shorter life cycle would be an advantage
in harsh environments having short growing sea-
sons. This was postulated by Shear (1975) for
another progenetic harvestman, Caddo pepper-
ella Shear.
A somewhat weaker argument may be made
that the paedomorphosis in Calidna is the result
of neoteny. First, it is theoretically possible
(though perhaps less parsimonious) that the re-
duction in size and juvenilization are indepen-
dently derived; the former resulting from pro-
portioned dwarfism, the latter from neoteny.
Second, the presence of troglobitic characters in
paedomorphic Calidna suggests cavelike selec-
tive pressures. Studies cited by Culver (1982)
UBICK AND BRIGGS: NEW PHALANGODID GENUS, CALIC1NA
135
indicated that certain cave salamanders become
neotenic in response to a low food supply. If the
subterranean habitat of paedomorphic Calicina
is cavelike in these respects, such as constant
conditions and resource scarcity, then K selec-
tion and neoteny emerge as possibilities.
Neoteny is also a possibility for the three larg-
est species of Calicina: kaweahensis, palaprae-
putia, and piedra. In contrast to the forest-dwell-
ing habits of other large species, these species
inhabit grassland biomes. They exhibit some re-
duction in the tarsal count, most pronounced in
the latter two (which also lack tarsal spurs on the
male palpi). Large size in combination with ju-
venile structures suggests neoteny.
Two situations suggest a relationship between
competition and paedomorphosis. First, Cali-
cina sequoia is unique among the paedomorphic
species. On the basis of its low tarsal count (most
populations have 3-3-4-4), it is the most pae-
domorphic species. However, it has well devel-
oped eyes, lives in dense forests, and has a year-
round activity period. Paedomorphosis here does
not seem to be a response to xeric environment.
Interestingly, this species is fully sympatric with
another phalangodid, " Sitalcina" cockerelli. The
two species not only occupy the same biomes,
but share identical habitats (we have six samples
containing both species). They differ most strik-
ingly in size: cockerelli has a body length two to
three times that of sequoia. It seems conceivable
that paedomorphosis could have evolved in se-
quoia as a means of reducing competition for
similar prey.
Second, competition may also be involved in
the evolution of paedomorphosis in other Cali-
cina species. The Coast Range species, from po-
lina south to morroensis, are closely sympatric
with species of Sitalcina. However, Sitalcina
species are non-paedomorphic forest dwellers,
whereas the coastal Calicina species are predom-
inantly paedomorphics from oak woodland and
grassland biomes (except for some populations
of serpentined). Since the original biomes for
Calicina were most probably dense forests, it is
likely that the coastal Calicina species also re-
sided in forests. Perhaps Calicina was excluded
from the more favorable environments by Si-
talcina.
ACKNOWLEDGMENTS
For their assistance with both field and labo-
ratory work we thank Vincent F. Lee, Kevin Horn,
Albert K. S. Jung, along with several other for-
mer Galileo High School students. Specimens of
Calicina and related phalangodids were kindly
loaned by Norman I. Platnick, James C. Co-
kendolpher, and William A. Shear. Special thanks
go to Mary Ann Tenorio for producing the scan-
ning electron micrographs; she and Leo Andres
gave invaluable assistance in artistic matters.
William A. Shear kindly made available numer-
ous detailed drawings of related phalangodids,
and Lynne R. Parenti provided an essential ref-
erence and additional information on biogeog-
raphy. Finally, we thank James C. Cokendol-
pher, Daphne G. Fautin, Willis J. Gertsch, David
H. Kavanaugh, Vincent F. Lee, Norman I. Plat-
nick, Wojciech J. Pulawski, William A. Shear,
and others for critically reading our manuscript
and offering many useful suggestions.
LITERATURE CITED
BRIGGS, T. S. 1968. Phalangids of the laniatorid genus Si-
talcina (Phalangodidae: Opiliones). Proc. Calif. Acad. Sci.
(ser. 4)36:1-32.
. 1974. Phalangodidae from caves in the Sierra Ne-
vada (California) with a redescription of the type genus
(Opiliones: Phalangodidae). Occas. Pap. Calif. Acad. Sci.
108:1-15.
. 1986. Neotenic morphology in Pacific Coast opil-
ionids. Proc. Ninth International Congress of Arachnology,
Panama, p. 303. (Abstr.)
BRIGGS, T. S. AND K. HOM. 1966. Five new species of Pha-
langodidae from California (Opiliones). Pan-Pac. Entomol.
42:262-269.
. 1967. New Phalangodidae from the Sierra Nevada
Mountains (Opiliones). Pan-Pac. Entomol. 43:48-52.
BRIGGS, T. S. AND D. UBICK. 1981. Studies on cave har-
vestmen of the central Sierra Nevada with descriptions of
new species ofBanksula. Proc. Calif. Acad. Sci. 42:3 1 5-322.
BRIGNOLI, P. M. 1968. Note su Sironidae, Phalangodidae e
Trogulidae italiani, cavernicoli ed endogei (Opiliones). Fragm.
Entomol. 5:259-293.
CULVER, D. C. 1982. Cave life. Evolution and ecology. Har-
vard University Press, Massachusetts. 185 pp.
GOODNIGHT, C. J. AND M. L. GOODNIGHT. 1942. New Pha-
langodidae (Phalangida) from the United States. Amer. Mus.
Novit. 1188:1-18.
. 1967. Opilionids from Texas caves (Opiliones, Pha-
langodidae). Amer. Mus. Novit. 2301:1-8.
GOULD, S. J. 1 977. Ontogeny and phylogeny. Harvard Uni-
versity Press, Massachusetts. 501 pp.
HENDRICKSON, D. A. 1986. Congruence of bolitoglossine bio-
geography and phylogeny with geologic history: paleotrans-
port on displaced suspect terranes? Cladistics 2:1 13-129.
MARTENS, J. 1977. Opiliones aus dem Nepal-Himalaya. III.
Oncopodidae, Phalangodidae, Assamiidae (Arachnida).
Senckenb. Biol. (1976) 57:295-340.
. 1986. Die Grossgliederung der Opiliones und die
Evolution der Ordnung (Arachnida). Proc. Tenth Interna-
tional Congress Arachnol., Jaca, 1986, 1:289-310.
RAMBLA, M. 1973. ContribucionalconocimientodelosOpil-
136
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 4
iones de la fauna iberica. Estudio de los subordenes Lani-
atores y Palpatores (pars.). Summary of Ph.D. Thesis. Uni-
versity of Barcelona. 2 1 pp.
. 1977. Un nuevo Scotolemon cavernicola de la isla
de Mallorca (Arachnida, Opiliones, Phalangodidae). Speleon
23:7-13.
1980. Neoteny in Opiliones. Proc. Eighth Interna-
tional Congress Arachnol., Vienna, 1980:489^92.
ROEWER, C.-F. 1923. Die Weberknechte der Erde. Verlag
von Gustav Fischer, Jena. IV, 1116 pp.
SHEAR, W. A. 1975. The opilionid family Caddidae in North
America, with notes on species from other regions (Opil-
iones, Palpatores, Caddoidea). J. Arachnol. (1974) 2:65-88.
SHEAR, W. A. AND J. GRUBER. 1983. The opilionid subfamily
Ortholasmatinae (Opiliones, Troguloidea, Nemastomati-
dae). Amer. Mus. Novit. 2757:1-65.
SUZUKI, S. 1973. Opiliones from the South-west Islands, Ja-
pan. J. Sci. Hiroshima Univ., Ser. B, Div. 1 (Zool.) 24:205-
279.
. 1985. A synopsis of the Opiliones of Thailand
(Arachnida). I. Cyphophthalmi and Laniatores. Steenstrupia
11:69-110.
WENNER, A. M. AND D. L. JOHNSON. 1980. Land vertebrates
on the California Channel Islands: sweepstakes or bridges?
Pp. 497-530 in The California islands: proceedings of a mul-
tidisciplinary symposium. D. Power, ed. Santa Barbara Mu-
seum of Natural History, Santa Barbara.
YANEV, K. P. 1980. Biogeography and distribution of three
parapatric salamander species in coastal and borderland Cal-
ifornia. Pp. 531-550 in The California islands: proceedings
of a multidisciplinary symposium. D. Power, ed. Santa Bar-
bara Museum of Natural History, Santa Barbara.
CALIFORNIA ACADEMY OF SCIENCES
Golden Gate Park
San Francisco, California 94 1 1 8
PROCEEDINGS
OF THE
CALIFORNIA ACADEMY OF SCIENCES
Vol. 46, No. 5, pp. 137-150, 6 figs.
March 9, 1989
FIVE NEW BERRY-FRUITED SPECIES OF
TROPICAL AMERICAN MELASTOMATACEAE
By
Frank Almeda
Department of Botany, California Academy of Sciences, Golden Gate Park,
San Francisco, California 94118-9961
ABSTRACT: Five species of berry-fruited Melastomataceae are described as new: Blakea fuchsioides from
western Panama; Clidemia hammelii from Costa Rica, Panama, and western Colombia; Miconia calocoma
and M. dissitiflora from Costa Rica; and Tococa croatii from eastern Panama. Predaceous mites of the genus
Ololaelaps are recorded as residents in the foliar pocket domatia of C. hammelii. This is the first report of
an association between a mite and a species of Melastomataceae. Discussions, distributional notes, and
diagnostic illustrations are provided for each of the species.
Received June 22, 1988. Accepted September 14, 1988.
INTRODUCTION
Ongoing field work in little-explored areas of
Central America and the study of numerous col-
lections sent for identification continue to yield
many new taxa among the berry-fruited genera
of Melastomataceae. In this paper I describe new
species in Blakea, Clidemia, Miconia, and To-
coca. The taxonomic notes and discussions in-
cluded here emphasize the diagnostic characters
that make each of the new species unusual or
anomalous additions to their respective genera.
Blakea fuchsioides Almeda, sp. nov.
Figure 1
TYPE: — PANAMA. Chiriqui: La Fortuna area, ca. 7 mi N
of Los Planes de Hornito in forest along small draw, elev. 3,600
ft (1,097 m), 26 Aug. 1983, Hammel & Kress 13473 (holotype:
CAS!; isotype: DUKE).
Frutex epiphyticus. Ramuli sicut petioli folia subtus inflo-
rescentia hypanthiaque modice pilis castaneis laevibus (0.5-)
l-1.5(-2.5) mm longis armati. Petioli 0.8-2.4 cm longi, lam-
ina 3-8. 9 x 2-4.7cmelliptico-ovatavelovataapiceacuminata
basi rotundata vel subcordata, 5-nervata, membranacea vel
subcoriacea et integra vel obscure dentata. Flores 6-meri in
quoque nodo superiore singuli vel bini, pedicellis 3.2-1 1.5 cm
longis, bracteae elliptico-lanceolatae; bracteae exteriores 3-4.5
x 1 cm acutae liberae; bracteae interiores 2.2-3.5 x 0.9-1.2
cm acutae liberae. Hypanthium (ad torum) 5-7 mm longum
extus strigillosum; calycis tubus 5 mm longus, lobis 4-6 mm
longis. Petala 19-23 x 14-15 mm elliptico-ovata apicerotun-
dato. Filamenta 5-7 mm longa; antherae 2-3.5 x 2 mm ob-
longae lateraliter non cohaerentes apicaliter biporosae; con-
nectivum dorsaliter ca. 0.5 mm supra thecarum basim
inconspicue tuberculatum. Stylus 1.8-3.2 cm; ovarium 6-lo-
culare omnino inferum apice glabro (cono et collo non evo-
luto).
Pendent viny epiphytes or hemiepiphytes ad-
hering to the bark of host trees by adventitious
roots. Older cauline internodes glabrate, terete,
becoming striate or cracked in age. Distal in-
dument of smooth rusty brown hairs (0.5-)1-
1.5(-2.5) mm long. Leaves of a pair slightly un-
equal in size; petioles 0.8-2.4 cm long; blades
membranaceous to subcoriaceous, 3-8.9 cm long
and 2-4.7 cm wide, elliptic-ovate to ovate, acu-
minate at apex, broadly rounded to subcordate
at base, margin typically entire but varying to
[137]
138
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 5
FIGURE 1 . Blakea fuchsioides Almeda. A, habit, x ca. %; B, representative leaf (lower surface), x ca. %; C, enlargement of
lower leaf surface showing pubescence and elevated median nerve; D, representative flower with petals and stamens removed,
x%; E, outer floral bract (adaxial surface), x2; F, petal, x4; G, stamens, lateral view (left) and ventral view (right), xca. 15;
H, seeds, xca. 15. (A-F from the holotype; G from Hammel 2261.)
remotely dentate, 5 -nerved below, the outermost
pair of subparallel primaries often inconspicuous
and concealed by the revolute margins on drying,
the transverse secondaries essentially parallel on
the upper surface and diverging from the ele-
vated primaries at nearly right angles, moder-
ately to densely strigillose to velutinous above
with widely to antrorsely spreading smooth hairs,
(0. 5-) 1-1. 5 mm long, but often varying to gla-
brate at maturity with the pubescence persisting
ALMEDA: FIVE NEW MELASTOMATACEAE
139
to varying degrees in the furrows created by the
impressed primary nerves, moderately to dense-
ly hirsute below with smooth rusty brown hairs
1-2.5 mm long. Flowers 6-merous, pendent, sol-
itary, or rarely geminate in the axils of distal
branches; pedicels 3.2-11.5 cm long, densely
covered with minutely roughened brown spread-
ing hairs 1 mm long. Floral bracts red, foliaceous,
sessile, and entire, moderately to densely covered
with roughened spreading hairs 0.5-1 mm long
on both surfaces; outer bracts 3-5 -nerved, free
to the base, 3-4.5 x 1 cm, lanceolate, apex acute,
base truncate to rounded; inner bracts 7-14-
nerved, free to the base, 2.2-3.5 x 0.9-1.2 cm,
lanceolate, apex acute, base truncate to rounded.
Hypanthia (at anthesis) campanulate, 5-7 mm
long to the torus (vascular ring), moderately to
densely strigillose with obscurely roughened hairs
0.5 mm long. Calyx tube ca. 5 mm long, flaring
and flangelike; calyx lobes copiously pubescent
with minutely roughened brown hairs 0.25-0.5
mm long, deltoid basally between sinuses but
abruptly tapered into linear-oblong, mostly en-
tire segments 4-6 mm long and 1 mm wide. Pet-
als glabrous, pink (fide McPherson 7733) or
magenta (fide Hammel & Kress 13473), elliptic-
ovate, 19-23 mm long, 14-15 mm wide, con-
nivent and somewhat concave adaxially with
slightly involute margins when expanded, apex
rounded. Stamens 12, isomorphic, free and
forming a pendent circle around the style; fila-
ments complanate, 5-7 mm long; anthers linear-
oblong to ovoid-oblong, truncate at apex, 2-3.5
mm long, 2 mm wide, each anther tipped with
two confluent pores; connective slightly thick-
ened dorsally and dilated basally about 0.5 mm
above base of anther thecae into a bluntly round-
ed deflexed caudiform appendage about 0.5 mm
long. Ovary inferior, 6-celled, glabrous at apex
but not expanded into a cone or collar. Style
straight, glabrous, 1.8-3.2 cm long, typically
exserted beyond the petals and sometimes ex-
ceeding the subtending pair of floral bracts; stig-
ma truncate to slightly rounded. Berry globose,
about 1 cm long and 1 cm in diameter. Seeds
numerous, clavate to narrowly pyriform or
bluntly deltoid, 1(-1.5) mm long, beige with a
smooth glossy testa and a prominent lateral raphe.
ADDITIONAL SPECIMENS EXAMINED. — PANAMA. Bocas del
Toro: along trail on divide separating Chiriqui and Bocas del
Toro, ca. 08°45'N, 82°15'W, 22 Oct. 1985, McPherson 7199
(CAS); Fortuna Dam region along continental divide trail,
08°45'04"N, 82°15'W, 23 Dec. 1986, McPherson & Aranda
10062 (CAS). Bocas del Toro/Chiriqui Border: above Fortuna
Dam, ca. 08°45'N, 82°15'W, 3 Dec. 1985, McPherson 7733
(CAS). Chiriqui: road to Fortuna Dam site N of Gualaca, 22.7
mi [36.8 km] beyond bridge over Rio Esti, 22 Nov. 1979,
Antonio 2770 (CAS); La Fortuna hydroelectric project, on ridge
behind camp, 22 Mar. 1978, Hammel 2261 (MO); trail west
from Fortuna Dam camp to La Fortuna, 08°43'N, 82°14'W,
23 Mar. 1985, Hampshire & Whitefoord 903 (BM); 2 km S of
Fortuna Lake, trail east, 08°43'N, 82°14'W, 24 Mar. 1985,
Hampshire & Whitefoord 912 (BM); campamento Bijao en
Fortuna, 20 Mar. 1976, Mendoza et al. 249 (PMA, US); entre
alto de Guayabo y la linea divisoria continental, 25 Sep. 1976,
Correa et al. 2837 (PMA, US).
DISTRIBUTION.— Known only from the wet
cloud forests of the Fortuna Dam region in west-
ern Panama along the Chiriqui/Bocas del Toro
border at 1,097-1,400 m.
PHENOLOGY.— Flowering specimens have been
collected from August through December; the
only available fruiting collection was made in
March.
Labels on collections of this species describe
it as a climbing or pendent epiphyte. I have not
studied this species in the field but its climbing
habit, adventitious roots, and pendent inflores-
cences lead me to suspect that it will ultimately
be shown to be a secondary hemiepiphyte like
Miconia arboricola Almeda (Almeda 1 984). Sec-
ondary hemiepiphytes are vine-like plants that
germinate terrestrially, ascend nearby trees by
adventitious roots, and later become epiphytic
by losing root contact with the ground (Putz and
Holbrook 1986).
In addition to its copious cover of brown
smooth hairs on distal branchlets, petioles, lower
leaf surfaces, and floral bracts, B. fuchsioides is
distinguished by an extraordinary combination
of floral characters. As emphasized by the spe-
cific epithet, the flowers of this species are rem-
iniscent of Fuchsia (Onagraceae) in posture and
coloration. In B. fuchsioides ihe flowers are borne
on flexible pendent pedicels exposing floral or-
gans in a way that favors hovering pollinators.
The connivent petals form a broad tube-like con-
formation, and the floral parts are vividly col-
ored. The hypanthia, calyx lobes, and decussate
floral bracts that closely subtend the flowers are
bright red, whereas the petals are reportedly pink
or magenta.
If future field observations confirm that flow-
ers of this species produce nectar, we can predict
that this species is probably bird pollinated be-
140
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 5
cause it exhibits so many of the features asso-
ciated with ornithophily (Percival 1965; Faegri
and van der Fiji 1971; Proctor and Yeo 1973).
This species and other undescribed Central
American species currently under study (Alme-
da, in prep.) exhibit anther characters that weak-
en the tenuous distinction between Blakea and
the closely related segregate genus Topobea. The
traditional differences between these genera are
based on anther shape and the relative size and
proximity of the apical pores. In Blakea, the an-
thers are oval, oblong, or elliptic and blunt or
broadly rounded at the summit with two typi-
cally well-separated (often minute) apical pores.
In Topobea, the anthers are linear-oblong to lan-
ceolate or subulate with the dorsally inclined api-
cal pores approximate and often confluent. An-
ther size and shape in B. fuchsioides agree with
the syndrome typically found in Blakea. The
broad confluent pores, however, dictate assign-
ment of this species to Topobea. The truncate
position of the anther pores is characteristic of
neither genus and underscores the difficulty of
assigning unusual new species to either of the
genera as they are currently denned. Until a
counterargument can be advanced for the con-
tinual recognition of Topobea based on some-
thing other than arbitrary characters, it seems
appropriate to follow Baillon (1879) and Mac-
bride ( 1 94 1 ) in submerging Topobea in the older
and much larger genus, Blakea.
Because B. fuchsioides is one of the most un-
usual species yet to be described in Blakea, it is
not surprising that its relationships are obscure.
A peculiar feature which has not been reported
in other species of Blakea or Topobea is the del-
icate, veil-like membrane that covers the broad
anther pores prior to anthesis. These veil-like
coverings may well function to preclude exces-
sive intrafloral pollen dispersion prior to petal
expansion on the pendent flowers.
Clidemia hammelii Almeda, sp. nov.
Figure 2
TYPE. -COSTA RICA. Heredia: Finca La Selva, OTS [Or-
ganization for Tropical Studies] Field Station on Rio Puerto
Viejo just E of its junction with Rio Sarapiqui, elev. about 100
m. Slopes along Q. El Salto, 2,900 m S, 2 Sep. 1980, Hammel
9682 (holotype: CAS!; isotype: DUKE).
Frutex vel arbuscula 1-5 m. Ramuli teretes primum sicut
folia novella inflorescentia sparse vel modice setosi pilis 2-3.5
mm longis et modice vel sparse stellulato-furfuracei demum
glabrati. Petioli 1-4 cm longi; lamina 1 1.5-29 x 6.1-14.5 cm
elliptica apice acuminata basi acuta vel cuneata, 5-7-nervata
vel 5-7-plinervata, nervi in axillis acarodomatiis instructi. In-
florescentia primum terminalis demum lateralis pauciramosa
plus minusve deflexa; flores 5-meri, pedicellis (ad anthesim)
1-2.5 mm longis, bracteolis 1.5-2.5 mm longis subulatis per-
sistentibus. Hypanthium (ad torum) 1 .5-2.5 mm longum; calyx
primum in cono apiculato clausus demum in lobos irregulares
persistentes ruptus, dentibus exterioribus 0.5-1 mm eminen-
tibus. Petala 3-3.5 x 1-1.5 mm oblonga-rotundato glabra.
Stamina isomorphica vel paulo anisomorphica glabra; fila-
menta 1.5-2 mm longa; antherarum thecae 1-1.25 x 0.5 vel
1.25-1.75 x 0.5 mm oblongae, poro paulo dorsaliter inclinato;
connectivum nee prolongatum nee appendiculatum. Stylus 5-
6 mm glaber in ovarii collo 0.5 mm immersus; ovarium
5-loculare omnino inferum apice modice glanduloso-setuloso.
Shrubs or small trees 1-5 m tall. Internodes
terete, glabrate at maturity; distal branchlets,
vegetative buds, and inflorescences beset with a
sparse to moderate covering of smooth, spread-
ing hairs (2-3.5 mm) that is typically intermixed
with inconspicuous, early deciduous, asperous-
headed hairs and underlain by a moderate to
dense understory of stellulate-furfuraceous or
short asperous-headed hairs. Leaves of a pair
typically somewhat unequal in size; petioles 1-
4 cm long; blades membranaceous, 1 1.5-29 cm
long and 6.1-14.5 cm wide, elliptic, apex long-
acuminate, base acute to cuneate, margin entire
to inconspicuously crenulate, 5-7-nerved or 5-
7-plinerved abaxially with pocket domatia typ-
ically formed in the angle between the median
nerve and each of the two proximal lateral nerves,
moderately strigose to subhirsute above with
smooth hairs mostly 1-2 mm long, moderately
hirsute below with a mixture of smooth hairs (1-
2.5 mm long) and minute glandular hairs essen-
tially restricted to the elevated primary and higher
order venation. Inflorescence a pseudolateral
modified dichasium, 3-6 cm long, divaricately
branched from the base; bracts of the rachis nodes
paired, narrowly lanceolate to subulate, 3-3.5
mm long, 1 mm wide, sparingly stellulate-fur-
furaceous to glabrate; bracteoles sessile, persis-
tent, paired but sometimes fused into an incon-
spicuous nodal collar, narrowly lanceolate to
subulate, 1.5-2.5 mm long, up to 0.5 mm wide,
essentially glabrous and entire but terminating
in a solitary hair. Pedicels 1-2.5 mm long, spar-
ingly to moderately stellulate-furfuraceous and
basally encircled by a deciduous, tufted ring of
barbellate or asperous-headed hairs. Hypanthia
(at anthesis) campanulate, 1.5-2.5 mm long to
the torus (vascular ring), moderately to sparsely
covered with spreading smooth hairs 0.5-2 mm
ALMEDA: FIVE NEW MELASTOMATACEAE
141
FIGURE 2. Clidemia hammelii Almeda. A, habit, x%; B, representative flower with stamens removed (left) and floral bud
showing rupturing calyx (right), x 9; C, petal, x 10; D, larger (antepetalous) stamens, dorsal view (left) and 3/4 lateral view (right),
x ca. 1 0; E, smaller (antesepalous) stamens, % lateral view (left) and dorsal view (right), x ca. 1 2; F, mature berry, x 3; G, seeds,
x 16. (A from Folsom 5864; B-E from de Nevers 3962; F, G from Hammel 12668.)
long and a sparse understory of sessile stellulate-
furfuraceous hairs. Calyx closed in bud and
crowned by an apiculum 0.5 mm long but rup-
turing irregularly at anthesis into 3-5 hyaline,
persistent, rounded lobes 0.5 mm long; calyx teeth
5, subulate, 0.5-1 mm long. Petals 5, glabrous
and reflexed, white or reportedly pale pink, ob-
long, rounded and often bluntly erose apically,
otherwise entire, 3-3.5 mm long, 1-1.5 mm wide.
Stamens 10, isomorphic or alternately subiso-
morphic with larger stamens inserted on gla-
brous torus opposite petals and smaller stamens
inserted opposite calyx lobes; filaments glabrous,
complanate and constricted distally, 1.5-2 mm
long; anthers 1-1.5 (-1.75) mm long and 0.5 mm
wide, yellow, linear-oblong, deeply channeled
142
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 5
ventrally between thecae, truncate to slightly
emarginate distally with a circular, somewhat
dorsally inclined pore; connective simple, slight-
ly thickened dorsally but not dilated or prolonged
below thecae. Ovary inferior, 5 -celled, fluted and
distended apically into a glandular-puberulent,
lobulate, stylar collar mostly 0.5 mm high that
becomes depressed and bowl-like on maturing
berries. Style straight, glabrous, 5-6 mm long;
stigma truncate. Berry reportedly dark purple at
maturity, globose, 6-7 mm long and 5-6 mm in
diameter. Seeds numerous, galeiform to deltoid,
0.5 mm long, white or beige, irregularly angulate
with a densely papillate testa and a lateral flat-
tened or somewhat convex raphe.
ADDITIONAL SPECIMENS EXAMINED.— COSTA RICA. Car-
tago: Valle Escondido, 30 Mar. 1966, Schnell 627 (US). He-
redia: Finca La Selva, OTS Field Station on Rio Puerto Viejo,
just E of its junction with Rio Sarapiqui, 21 Jul. 1981, Hammel
11026 (DUKE); forest N of road to Puerto Viejo at Chilamate,
22 May 1 982, Hammel 12423 (CAS, CR); hills near Chilamate
along road between Puerto Viejo and La Virgen, 30 May 1 982,
Hammel 12668 (CAS); Chilamate de Sarapiqui, S end of Cer-
ros Sardinal, N side of Rio Sarapiqui, 10°27.5'N, 84°4'W, 3
Jul. 1985, Hammel & Grayum 14103 (CAS); Finca El Bejuco,
S base of Cerros Sardinal, Chilamate de Sarapiqui, 10°27'N,
84°04'W, 2 Jun. 1985, Grayum & Jacobs 5355 (CAS); between
Rio Peje and Rio Sardinalito, Atlantic slope of Volcan Barva,
10°18.5'N, 84°04'W, 8 Apr. 1986, Grayum 6910 (CAS-2
sheets); Tirimbina, 9 Jun. 1971, Proctor 32244 (MO). PAN-
AMA. Bocas del Toro: between Fortuna and Chiriqui Grande,
8.5 mi [13.7 km] N of bridge over the Fortuna Lake, 08°46'N,
82°14'W, 10 Mar. 1985, Croat 60151 (CAS); Fortuna to Chi-
riqui, ca. 3 km N of continental divide, 08°48'N, 82°1 1'W, 12
Feb. 1986, Hammel & McPherson 14466 (CAS); Fortuna Dam
region, near Chiriqui Grande, 08°45'N,82°1 5' W, 18 Jan. 1986,
McPherson 807 9 (CAS); 8.5 road-miles [13.7 km] from bridge
near Fortuna Dam on road towards Chiriqui Grande, 08°50'N,
82°15'W, 10 Mar. 1985, McPherson 6760 (CAS); road to Chi-
riqui Grande, 6 km N of continental divide, 08°50'N, 82°07' W,
2 Aug. 1984, Todzia et al. 2558 (CAS); Fortuna Dam area to
Chiriqui Grande, 2 Aug. 1984, D'Arcy et al. 15996 (CAS);
above Chiriqui Grande ca. 2 mi E on a side road ca. 10 road-
miles below continental divide, 08°55'N, 82°10'W, 11 Mar.
1988, Almeda et al. 6082 (CAS, MO, PMA, US). Code: 2 mi
N of Cerro Pilon, 16 Mar. 1973, Liesner 726 (CAS, MO);
between La Junta and Limon, 5 hour walk N of Alto Calvario,
1 1 Oct. 1977, Folsom 5864 (CAS); Caribbean side of divide
at El Cope, 08°45'N, 80°35'W, 4 Feb. 1983, Hamilton & Da-
vidse 2776 (CAS); road from La Pintada to Coclesito, 08°45'N,
80°30'W,7 Feb. 1983, Hamilton &Davidse 28 15 (CAS). Colon:
9 mi [14.6 km] S of Portobello, 17 Jul. 1970, Croat 11372 (F,
US); Rio Iguanita and inland to 2 km, 7 Feb. 1981, D'Arcy
14615 (CAS). Darien: vicinity of gold mining camp of T. Kit-
tredge on headwaters of Rio Tuquesa ca. 2 air km from con-
tinental divide, 26 Aug. 1974, Croat 27210 (F); forested ridges
and valleys at Mamey, 6 Mar. 1982, Whitefoord & Eddy 374
(BM, MO). Panama: 13 km N of El Llano on road to Carti,
13 Feb. 1973, Busey 374 (US); El Llano-Carti road, 14.8 km
N of Panamanian Hwy., 27 Jan. 1977, Folsom et al. 1479
(MO). San Bias: Nusagandi, trail from camp NW to a quebrada,
09°19'N, 78°15'W, 31 Jul. 1984, de Nevers & de Leon 3595
(CAS); Nusagandi, El Llano-Carti road, 09°19'N, 78°15'W, 10
Aug. 1984, de Nevers & Gonzalez 3655 (CAS); Llano-Carti
road, 19 km from Interamerican Hwy., 09°19'N, 78°55'W, 4
Sep. 1 984, de Nevers & Herman 3831 (CAS); Llano-Carti road,
20 km from Interamerican Hwy., 09°19'N, 78°55'W, 30 Sep.
1 984, de Nevers et al. 3962 (CAS). COLOMBIA. Choco: forest
NW of Alto Curiche, 20 May 1967, Duke & Idrobo 11240
(US).
DISTRIBUTION.— Currently known from river
banks, shaded sites, and light gaps in rain forests
from the Puerto Viejo region of northeastern
Costa Rica south through Panama to an area
northwest of Alto Curiche in western Colombia,
from sea level to 900 m.
PHENOLOGY.— Flowering and fruiting occur
sporadically throughout the year.
Although variable in foliar size and the degree
to which the elevated primaries diverge from the
median nerve on lower foliar surfaces, C. ham-
melii is readily denned by the following set of
characters: a divaricately branched pseudolateral
inflorescence; an irregularly rupturing apiculate
calyx; a fluted, glandular-puberulent stylar col-
lar; and seeds that are galeiform to angularly del-
toid in outline with a densely papillate testa. There
also appears to be some intrafloral variation in
staminal size among the three flowering collec-
tions available for study. A collection from Pan-
ama (de Nevers et al. 3962) has stamens that are
alternately somewhat unequal in size (see Fig. 2),
whereas Hammel 9682 from Costa Rica and
Todzia et al. 2558 from Panama have stamens
that are clearly isomorphic. This slight variation
may prove to be taxonomically inconsequential
but warrants further observation as additional
flowering collections are made.
Although C. hammelii appears to have no close
allies among described congeners, it bears a su-
perficial vegetative resemblance to C. reitziana
Cogn. & Gleason and C. costaricensis Cogn. &
Gleason. Both differ from C. hammelii in having
an adaxially pubescent torus, regularly devel-
oped (nonrupturing) calyx lobes, and smooth
seeds.
A distinctive, but not always conspicuous, fea-
ture of C. hammelii is the occurrence of small
funnel-shaped pockets or chambers in the angles
between the median nerve and the base of each
proximal lateral nerve (see Fig. 3). The function
of these structures in any kind of symbiotic re-
lationship is unclear. Of the numerous collec-
tions cited above, label information on only one
ALMEDA: FIVE NEW MELASTOMATACEAE
143
B
FIGURE 3. Clidemia hammelii Almeda. A, representative leaf (abaxial surface), xca. '/»; B, enlargement of lower leaf surface
showing pocket domatia, xca. 5; C, the mite, Ololaelaps sp., dorsal view (above) and ventral view (below), xca. 20. (A, B from
Almeda et al. 6082.)
from Panama (de Never s & Herman 3831) makes
note of "ants in the leaves." During recent field
work I examined several individuals of another
Panamanian population (Almeda et al. 6082) and
found mites of the genus Ololaelaps (Gamasida,
family Laelapidae) in the domatia of at least three
individuals (see Fig. 3). No other insects or ar-
thropods, however, were encountered in the 33
domatia examined. According to G. W. Krantz
(in litt., 16 June 1988) a domatium is an odd
habitat for Ololaelaps because most are de-
scribed from moss and litter substrates, or from
the nests of small mammals. Because mites of
this genus are thought to be predaceous, Krantz
suggests that the mites may feed on resident
nematodes, or they may simply use the domatia
as shelters.
In his review of plant domatia, Jacobs (1966)
concludes that there is no evidence to demon-
strate a biological relationship between domatia-
bearing plants and mites. Benson (1985), how-
ever, notes that scavenging mites are fairly
common opportunistic invaders of plant cavi-
ties. He suggests that pocket domatia, often re-
ferred to as acarodomatia, probably function as
feeding sites for sucking homopterans (i.e., the
coccids or scale insects, and the membracids or
treehoppers). This ant-coccid linkage is intrigu-
ing when one considers that homopterans can
attract ants with their honeydew secretions and
gain protection from other predaceous insects.
Charles Turner (pers. comm., 26 May 1988), who
is currently working on mite ecology, proposes
a facultative mutualistic interaction to explain
the plant-mite association. He suggests that the
pocket domatia may serve as egg nurseries and
help to shelter the mites from predators. The
mites, in turn, benefit the plant by feeding on
fungal spores and the eggs of predaceous insects.
The plant-mite relationship thus remains un-
solved, requiring more sophisticated field-ori-
ented study than it has received so far.
This species is named for Barry E. Hammel,
a student of neotropical Clusiaceae whose col-
144
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 5
lecting efforts in Costa Rica and Panama have
added many new and interesting plants to the
flora of Mesoamerica.
Miconia calocoma Almeda, sp. nov.
Figure 4
TYPE. -COSTA RICA. Heredia: Finca La Selva, OTS Field
Station on Rio Puerto Viejo just E of its junction with Rio
Sarapiqui, southeast corner, elev. 100m, 17 Apr. \9&l,Folsom
9776 (holotype: CAS!; isotype: DUKE).
Frutex vel arbuscula 3-5 m. Ramuli teretes sicut petioli
foliorum subtus venae primariae inflorescentia hypanthiaque
modice pilis stellatis induti. Petioli 4-10 mm longi; lamina
(4.2-)8.5-19.5 x (2.3-)5-8.5 cm elliptica vel ovato-elliptica
apice acuminata basi rotundato-obtusa, 5(-7)-plinervata,
membranacea et undulata vel undulato-dentata. Inflorescentia
3-7 cm longa multiflora; flores 4-meri, pedicellis (ad anthesim)
0.5 mm longis, bracteolis 2-3 mm longis anguste oblongis
persistentibus. Hypanthium (ad torum) 2 mm longum; calyx
primum in cono apiculato clausus demum in lobos regulares
persistentes ruptus, dentibus exterioribus 1-1.5 mm eminen-
tibus. Petala 3-4 x 1.5-2.5 mm obovato-oblongaglabra. Stam-
ina isomorphica glabra; filamenta 1.5-2 mm longa; anthera-
rum thecae 1-1.5 x 0.5 mm angustae oblongae, poro paulo
ventraliter inclinato; connectivum nee prolongatum nee ap-
pendiculatum. Stylus 5-10 mm glaber; ovarium 4-loculare et
'A inferum glaber.
Shrubs or small trees 3-5 m tall. Older cauline
internodes terete and essentially glabrous. Distal
branchlets, petioles, and inflorescences densely
covered with a mixture of rusty brown sessile-
stellate and stipitate-stellate hairs. Leaves of a
pair typically unequal in size; petioles 4-10 mm
long; blades membranaceous, (4.2-)8.5-19.5 cm
long and (2.3-)5-8.5 cm wide, elliptic to elliptic-
ovate, apex acuminate, base typically obtuse to
broadly rounded but sometimes varying to
slightly oblique at base and then somewhat de-
current on petiole, margin undulate to undulate-
dentate; 5(-7)-plinerved with the inner pairs of
subparallel primary nerves elevated and diverg-
ing from median nerve in opposite, subopposite,
or irregularly alternate fashion at successive points
above the blade base, sparingly stellate pubescent
to glabrous above at maturity, copiously stellate
pubescent on elevated primary nerves below with
a sparser hair covering on prominulous network
of transverse secondary and higher order veins.
Inflorescence terminal but sometimes appearing
pseudolateral because of elongation of axillary
shoots, paniculiform with ultimate branchlets
terminating in multiflowered, congested glomer-
ules; bracts of rachis nodes paired, linear-oblong,
2-5 mm long, 0.5-1 mm wide, glabrous above
and copiously stellate below; bracteoles persis-
tent, 2-3 per pedicel, sessile, linear-oblong, at-
tenuate apically, 0.5-1.5 mm long, 0.25-0.5 mm
wide, margin entire with a solitary apical hair,
glabrous above and sparingly to moderately stel-
late below. Pedicels copiously stellate pubescent,
0.5 mm long but inconspicuous and concealed
by congested glomerules. Hypanthia (at anthesis)
campanulate, 2 mm long to the torus (vascular
ring), densely stellate pubescent throughout. Ca-
lyx closed in bud and crowned by an apiculum
0.25 mm long but rupturing at anthesis into 2-
4 persistent deltoid to semicircular hyaline lobes
mostly 1 mm long and 1-1.5 mm wide basally;
exterior calyx teeth 4, linear-oblong, 1-1.5 x 0.5
mm, copiously stellate and widely spreading to
recurved on fruiting hypanthia. Petals 4, gla-
brous, white, obovate to oblong-obovate, widely
spreading to reflexed, rounded apically, 3-4 mm
long and 1.5-2.5 mm wide distally. Stamens 8,
isomorphic; filaments glabrous, complanate,
constricted and incurved distally, 1.5-2 mm long;
anthers 1-1.5 mm long and 0.5 mm wide, yellow,
linear-oblong, laterally compressed and deeply
channeled ventrally between thecae, truncate to
slightly emarginate distally with a somewhat
ventrally inclined terminal pore; connective
thickened dorsally but not dilated or prolonged
below the filament insertion. Ovary ca. % infe-
rior, 4-celled, and glabrous. Style erect, often
somewhat curved distally, glabrous, 5-10 mm
long and conspicuously overtopping the sur-
rounding stamens; stigma truncate to capitellate.
Berry reportedly blue-black to purple at matu-
rity, globose, 6-8 mm long and 6-8 mm in di-
ameter. Seeds numerous, obovoid to pyriform,
1 mm long, white or beige, somewhat angulate
with a densely tuberculate testa and a lateral flat-
tened raphe on the opposing face.
ADDITIONAL SPECIMENS EXAMINED.— COSTA RICA. Here-
dia: S of Puerto Viejo, 2 km S of Magsasay Penal Colony, W
of road, 5 Feb. 1983, Garwood et al. 1125 (BM, MO); Finca
La Selva, OTS Field Station on Rio Puerto Viejo, just E of its
junction with Rio Sarapiqui, 3,000 m line, South Boundary
along Central Trail, 2 May 1981, Folsom 9990 (CAS); Finca
La Selva, OTS Field Station, South Boundary, 1 ,900 m E, 6
Aug. 1 980, Hammel 9452 (DUKE); Finca La Selva, OTS Field
Station, South Boundary, 1,600 m E, 27 Sep. 1980, Hammel
9872 (DUKE); Finca La Selva, OTS Field Station, South
Boundary on slope just W of Q. Esquina, 1 5 Jul. 1 982, Hammel
& Trainer 13181 (CAS). Limon: Cerro Coronel, E of Laguna
Danto, 10°41'N, 83°38'W, 16-23 Jan. 1986, Stevens 23730
(MO).
DISTRIBUTION.— A little-collected rain forest
species currently known from northeastern Costa
ALMEDA: FIVE NEW MELASTOMATACEAE
145
FIGURE 4. Miconia calocoma Almeda. A, habit, x %; B, representative leaf (lower surface), x %; C, enlargement of lower
leaf surface showing stellate pubescence; D, representative flower, x!2; E, petal, xlO; F, stamens, % lateral view (left) and
ventral view (right), x 10; G, mature berry, x4; H, seeds, x 15. (A-F from the holotype; G, H from Hammel 13181.)
Rica in an area extending from Cerro Coronel
just south of Barra del Colorado southwest to the
OTS La Selva Field Station and vicinity at 20-
200m.
PHENOLOGY.— Flowering and fruiting speci-
mens have been collected from January through
May and July through September, respectively.
Miconia calocoma is distinguished by its rusty
stellate pubescence, 4-merous flowers, irregular-
ly rupturing apiculate calyx, unappendaged an-
thers, 4-celled ovary, and densely tuberculate,
obovoid to pyriform seeds.
Placement of this species into Cogniaux's
(1891) sectional classification depends on wheth-
er one chooses to emphasize characters of the
calyx or stamens. In having oblong unappen-
daged stamens with ventrally inclined apical
pores, M. calocoma would appear to belong to
146
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 5
section Amblyarrhena. Within this section, how-
ever, no particular species or group of species
can be singled out as close relatives. The irreg-
ularly rupturing apiculate calyx is characteristic
of section Laceraria, but M. calocoma lacks the
dorsally appendiculate anthers that are typical of
that section. Within this latter section, M. cal-
ocoma is most similar to M. rupticalyx Wurdack
of Venezuela and M. wagneri Macbride of Bo-
livia and Peru. These two allopatric taxa share
several diagnostic features with M. calocoma in-
cluding the stellate pubescence, rupturing apic-
ulate calyx, and 4-merous flowers. Characters that
consistently separate M. rupticalyx from M. cal-
ocoma include the larger (12-28 x 9-19 cm)
ovate, 5-7-nerved leaves with rounded serrulate
margins, reniform bracteoles, dorsally inclined
anther pores, and 2(-3)-celled ovary. Miconia
wagneri is readily separated from M. calocoma
by many of these same features but it differs most
notably in having larger (15-25 x 8-12 cm), 5-
7-nerved leaves that are entire or remotely cal-
lose-denticulate and anther connectives that are
dilated at the filament insertion into dorso-basal
spurs.
The epithet for this species is derived from the
Greek word kalos, beautiful, and the Latin word
coma (home in Greek), hair tuft, in reference to
the attractive covering of rusty stellate hairs on
branchlets, leaves, and inflorescences.
Miconia dissitiflora Almeda, sp. nov.
Figure 5
TYPE.— COSTA RICA. Puntarenas: above Golfito along road
to television tower, elev. 50-500 m, 16 Jul. 1977, Almeda et
al. 3093 (holotype: CAS!; isotypes: BM!, BR!, CR!, DUKE!,
F!, MO!, NY!, US!).
Frutex 1-3 m. Ramuli teretes sicut foliorum subtus venae
primariae inflorescentia hypanthiaque sparse vel modice glan-
duloso-furfuracei. Folia subsessilia vel petiolis l-5(-9) mm
longis; lamina 6.5-22 x 3.1-10cm,ellipticavelelliptico-ovata
apice acuminata vel attenuata basi paulo cordata vel rotundata,
5(-7)-plinervata, membranacea et undulata vel subintegra. In-
florescentia 7-20.5 cm longa multiflora; flores 5-meri, pedi-
cellis (ad anthesim) 1.5-2.5 mm longis; bracteolis 0.5 mm
longis subulatis. Hypanthium (ad torum) 1.5 mm longum; ca-
lyx primum in cono apiculato clausus demum in lobos irre-
gulares persistentes ruptus, dentibus exterioribus 0.5 mm em-
inentibus. Petala 3.5 x 1.5-2 mm oblongo-lanceolata glabra.
Stamina paulo anisomorphica glabra; filamenta 1.5-2 mm lon-
ga; antherarum thecae 2 x 0.5 vel 1.5 x 0.5 mm angustae
oblongae, poro paulo dorsaliter inclinato; connectivum nee
prolongatum nee appendiculatum. Stylus 2.5-3 mm glaber;
ovarium 5-loculare omnino inferum apice modice puberulo.
Sparingly branched shrubs 1-3 m tall. Older
cauline internodes terete and glabrous. Distal
branchlets and inflorescences moderately to
sparingly stellate-furfuraceous. Leaves of a pair
somewhat unequal in size, typically subsessile
and clasping or sometimes with petioles l-5(-9)
mm long; blades membranaceous to subcoria-
ceous at maturity, 6.5-22 cm long and 3.1-10
cm wide, elliptic to elliptic-ovate, apex acumi-
nate to attenuate, base rounded to subcordate
but sometimes varying to slightly oblique, mar-
gin bluntly undulate-dentate to subentire, 5(-7)-
plinerved below with inner pairs of subparallel
primary nerves elevated and diverging from me-
dian nerve in opposite fashion at successive points
above blade base, glabrous above, glabrous to
sparingly stellate on and adjacent to elevated pri-
maries below. Inflorescence a terminal, divari-
cately branched paniculiform dichasium 7-20.5
x 7-30 cm; bracts of the rachis nodes paired,
lance-triangular, 0.5-4 mm long, 0.5-1 mm wide,
essentially glabrous throughout or sparingly stel-
late pubescent below; bracteoles paired, gla-
brous, sessile and evidently fused into a short
nodal collar usually evident as an elevated in-
terpetiolar ridge, lance-triangular to subulate, 0.5
mm long, 0.5 mm wide at base, margin entire.
Pedicels sparingly stellate-furfuraceous to gla-
brous, 1.5-2.5 mm long. Hypanthia (at an thesis)
campanulate, 1.5 mm long to torus (vascular ring),
glabrous to sparingly stellulate puberulent ba-
sally. Calyx closed in bud and crowned by an
apiculum ca. 0.5 mm long but rupturing irreg-
ularly at anthesis into 2-5 persistent hyaline lobes
1-1.5 mm long; exterior calyx teeth 5, subulate,
0.5 mm long, glabrous, erect to antrorsely
spreading. Petals 5, glabrous, white, oblong-lan-
ceolate, acute to retuse apically, 3.5 mm long and
1.5-2 mm wide, margin entire. Stamens 10, al-
ternately unequal with the larger stamens in-
serted on the torus opposite the petals and the
smaller ones inserted opposite the calyx lobes;
filaments glabrous, complanate, somewhat con-
stricted and geniculate distally, 1.5-2 mm long;
anthers 1.5-2 mm long and 0.5 mm wide, yellow,
linear-oblong, truncate distally with a somewhat
dorsally inclined terminal pore; connective
slightly thickened dorsally but not dilated or pro-
longed below the filament insertion. Ovary in-
ferior, 5 -celled, minutely and caducously puber-
ulent at the summit. Style straight, glabrous, 2.5-
3 mm long; stigma truncate. Berry globose, 2.5-
3.5 mm long and ca. 2.5 mm in diameter. Seeds
numerous, white, galeiform, 0.5 mm long with
ALMEDA: FIVE NEW MELASTOMATACEAE
147
H
FIGURE 5. Miconia dissitiflora Almeda. A, habit, x%; B, enlargement of inflorescence node showing bracts and stellate-
furfuraceous pubescence; C, floral bud showing rupturing calyx and apiculum, x 1 2; D, flower with petals and stamens removed,
x 1 2; E, petal, x 9; F, larger (antepetalous) stamens, dorsal view (left) and lateral view (right), x 6; G, smaller (antesepalous)
stamens, lateral view (left) and dorsal view (right), xca. 6; H, mature berry, xca. 10; I, seeds, xca. 18. (A-G from the holotype;
H, I from Liesner 2013.)
a densely muricate or verrucose testa on the con-
vex side and a lateral raphe on the opposing face.
ADDITIONAL SPECIMENS EXAMINED.— COSTA RICA. Pun-
tarenas: main ridge and NE slopes of Fila de Cal, between San
Vito and Ciudad Neily, 08°41'N, 82°56.5'W, 13 Sep. 1985,
Graytim et al. 6034 (CAS); NE slopes of Fila de Cal between
San Vito and Ciudad Neily, 08°41'N, 82°56.5'W, 12Jul. 1985,
Hammel & Grayum 14187 (CAS); Rincon de Osa, Rio Agua
Buena to ca. 4 km above it, 10 Feb. 1974, Liesner 20 13 (CAS);
Corcovado National Park, trail from base of hills to Los Chiles,
9 Jul. 1977, Liesner 3071 (CAS, CR); trail from radio tower
on ridge above Golfito, opposite docks, 08°39'N, 83°10'W, 25
Jul. 1977, Webster 22045 (CAS).
DISTRIBUTION.— Known only from the Golfo
Dulce region of southern Costa Rica in an area
extending from Los Chiles in Corcovado Na-
148
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 5
tional Park eastward to Fila de Cal between San
Vito and Ciudad Nelly at 20-620 m.
PHENOLOGY.— Flowering specimens have been
collected in July and September; the only fruiting
specimen was collected in July.
Miconia dissitiflora is a well-defined species,
readily distinguished by its sparse stellate-
furfuraceous indument, divaricately branched
inflorescence, irregularly rupturing calyx, unap-
pendaged anisomorphic stamens, and oblong-
lanceolate petals. It is also unusual and evidently
unique among New World berry-fruited mela-
stomes in having the larger stamens inserted on
the torus opposite the petals. Among neotropical
melastomes this condition is known only in the
capsular-fruited genus Monochaetum.
The problem of placing M. dissitiflora into a
section parallels the ambiguous situation de-
scribed above for M. calocoma. An emphasis on
staminal characters would dictate placement into
section Amblyarrhena, whereas characters of the
calyx suggest a relationship with the species of
section Laceraria. The new species bears no strong
resemblance to any of the described taxa cur-
rently placed in section Amblyarrhena. A search
for morphologically similar species in section
Laceraria has only served to highlight the am-
biguities involved when assessing relationships
based on different character sets. Miconia cen-
trodesma of section Laceraria shares an irregu-
larly rupturing calyx and a sparse stellate-fur-
furaceous indument with M. dissitiflora, but it
differs in a number of diagnostic reproductive
characters. It has 4-merous flowers, isomorphic
stamens, dorsally appendiculate anthers, a 2-3-
celled ovary, and narrow angular-oblong tuber-
culate seeds. The seeds of M. centrodesma are
unusual in having a dilated testa at the distal end
that forms a foot or spur composed of enlarged
cells that collapse on drying. This "foot," which
can be observed only in fresh or hydrated ma-
terial, imparts an overall shape reminiscent of a
miniature boomerang. Thus, the seeds of M cen-
trodesma differ dramatically from the muricate
or verrucose galeiform seeds of M. dissitiflora.
These considerable differences suggest that the
relationship between M. centrodesma and M.
dissitiflora is not a particularly close one. Con-
sequently, M. dissitiflora is probably best re-
garded as an isolated species not readily placed
within the sectional classification proposed by
Cogniaux(1891).
The epithet chosen for this species is derived
from the Latin word dissitus, remote or lying
apart, referring to the well-spaced position of
flowers on the divaricately branched inflores-
cence. It is this character, together with the small
apiculate buds and plinerved leaves, that give M.
dissitiflora an aspect reminiscent of Conostegia
cinnamomea (Beurl.) Wurdack. The latter differs
in having isomorphic stamens, a calyptriform,
circumscissile calyx, and oblong, deltoid seeds
that are somewhat angulate on the convex face.
Tococa croatii Almeda, sp. nov.
Figure 6
TYPE.— PANAMA. Darien: Serrania de Pirre, 9-10 km due
N of Alto de Nique, elev. 1,520-1,560 m, 27 Jul. 1976, Croat
37873 (holotype: CAS!; isotype: MO!).
Frutex 0.5-15 m. Ramuli primum rotundato-quadrangulati
demum teretes sicut foliorum venae primariae et secundariae
sparse vel modice setosi pilis laevibus laxis paulo retrorsis 0.5-
2mmlongis. Petioli2.5-12.5cmlongi;lamina 12-19.5 x 8.3-
1 5.3 cm ovata apice acuminata basi rotundata vel cordata, 7(-
9)-nervata, membranacea et crenulata ciliata. Inflorescentia
10-18 cm long multiflora; flores 5-meri, pedicellis (ad anthe-
sim) 1-4 mm longis, bracteolis 1 mm longis subulatis persis-
tentibus. Hypanthium (ad torum) 2-2.5 mm longum; calyx
primum in cono apiculato clausus demum in lobos regulares
persistentes ruptus, dentibus exterioribus 0.25-0.5 mm emi-
nentibus. Petala 5-6 x 3.5-5 mm obovata glabra. Stamina
isomorphica glabra; filamenta 2.5-3.5 mm longa; antherarum
thecae 1.5-2 x 1 mm oblongae inter se cohaerentes poro 0.5
mm diam.; connectivum dorsaliter ad basim dente ca. 1-1.5
mm longo descendenti armatum. Stylus 5.5-8.5 mm; ovarium
5-loculare omnino inferum apice sparsissime pilis 0.5-1 mm
longis glandulosis armato.
Shrubs 0.5-1.5 m tall. Distal branchlets
subquadrangular, sparsely to moderately beset
with a caducous cover of straight spreading and/
or somewhat crisped hairs mostly 1-2 mm long.
Leaves of a pair equal to slightly unequal in size;
petioles 2.5-12.5 cm long; blades thin and mem-
branaceous, 12-19.5 cm long and 8.3-15.3 cm
wide, broadly ovate to cordate, apex acuminate,
base broadly rounded to cordate, margin ciliate-
crenulate, 7(-9)-nerved below with all primary
nerves diverging from a common point at base
of blade, sparingly to moderately setulose above
with smooth appressed to somewhat spreading
hairs mostly 0.5-2 mm long, moderately fine se-
tulose on and between primary nerves below with
tardily caducous, simple, and glandular hairs 0.5-
1 mm long. Inflorescence a terminal multiflow-
ered panicle 10-18 cm long with subverticillate
branching at secondary nodes; rachis moderately
to copiously covered with spreading glandular
hairs (0.5-1 mm long) and a ground layer of
sessile glandular hairs; bracts of rachis nodes
paired, linear-oblong, 2.5-3 mm long, 0.5 mm
ALMEDA: FIVE NEW MELASTOMATACEAE
149
FIGURE 6. Tococa croatii Almeda. A, habit, x'/i; B, flower with petals and staminal ring removed, x4; C, petal, x5; D,
staminal ring as seen from above (left) and lateral view (right), xca. 4; E, seeds, xca. 25. (A-D from the holotype; E from
Gentry & Mori 13817.)
wide, sparsely glandular setulose below inter-
mixed with or replaced by an inconspicuous
ground layer of sessile glandular hairs; bracteoles
paired, sessile, and persistent, subulate to seti-
form, 1 mm long, essentially glabrous or with a
sparse cover of sessile glandular hairs. Pedicels
1-4 mm long, beset with minute glandular hairs
up to 0.5 mm long. Hypanthia (at anthesis) cam-
panulate, 2-2.5 mm long to torus (vascular ring),
moderately covered with spreading smooth glan-
dular hairs 0.5-1 mm long intermixed with or
replaced by a ground layer of sessile dot-like
glandular hairs. Calyx closed in bud, crowned by
an apiculum 0.5 mm long but rupturing regularly
at anthesis into 5 triangular hyaline lobes 2x2
mm; exterior calyx teeth 5, fused to calyx lobes
for much of their length with the free subulate
portions projecting 0.25-0.5 mm from the distal
portion of each calyx lobe. Petals 5, glabrous,
white (fide Croat 37873), pink (fide Hartman
12458), or bluish-purple (fide Hartman 12404),
obovate, rounded apically, 5-6 mm long and 3.5-
5 mm wide, margin entire. Stamens 10, iso-
morphic; filaments glabrous, complanate, 2.5-
3.5 mm long; anthers 1.5-2 mm long, 1 mm
wide, laterally compressed and coherent in a ring,
apex truncate to somewhat emarginate on ven-
tral side; connective thickened dorsally and pro-
longed at base into a retrorse triangular appen-
dage 1-1.5 mm long. Ovary inferior, 5 -celled,
sparingly glandular-setose at summit surround-
ing style base. Style straight, minutely papillate
distally, 5.5-8.5 mm long and conspicuously
overtopping surrounding stamen ring; stigma
150
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 5
truncate to capitellate. Berry reportedly blue at
maturity, globose, 5-6 mm long, 5-6 mm in di-
ameter. Seeds numerous, ovoid, white, 0.5 mm
long with a shallowly scrobiculate testa.
ADDITIONAL SPECIMENS EXAMINED. — PANAMA. Darien:
Cerro Mali, vicinity of base camp, 1,400 m, near Colombian
border, 23 Jan. 1975, Gentry & Mori 13817 (F, US); SW ridge
leading to Alturas de Nique on border with Colombia, 1 , 1 00-
1,200 m, 30 Dec. 1980, Hartman 12404 (CAS); Alturas de
Nique and ridge leading SW, 1,250-1,500 m, 31 Dec. 1980,
Hartman 12458 (CAS).
DISTRIBUTION.— A little-collected species of
eastern Panama currently known from Cerro Mali
in the Serrania del Darien and the region of Al-
turas de Nique in the Serrania de Pirre in south-
ern Darien province near the Panama/Colombia
frontier at 1,100-1,560 m.
PHENOLOGY.— Flowering specimens have been
collected in July and December; the only known
fruiting specimen was collected in January.
Tococa croatii is denned by its caducous cover
of straight or crisped cauline hairs (1-2 mm long),
regularly rupturing hyaline calyx, and shallowly
scrobiculate, ovoid seeds. The most extraordi-
nary feature of this species is the morphology of
the anther thecae. These are distally truncate to
emarginate with a comparatively broad terminal
pore, laterally coherent in a ring, and beset dor-
sally with deflexed triangular appendages. Among
described species of Tococa, T. croatii shares its
habit, inflorescence architecture, staminal de-
tails, and fruit morphology only with T. sym-
phyandra (Triana) Cogn. of Colombia and Ec-
uador. The consistent differences between these
species involve foliage characters and pubes-
cence details of the distal internodes, hypanthia
and ovaries. Tococa symphyandra has long sub-
retrorse-setose hairs (0.5-1.5 cm long) on young
branchlets and petioles, didymous formicaria
about 1 cm long that are immersed in the blade
base, and essentially glabrous hypanthia and
ovaries. The derived staminal characters that
unite T. croatii and T. symphyandra as a dis-
tinctive species pair are also the features that
make them anomalous members of the genus.
These staminal characters would also make them
equally anomalous if included in Miconia or any
other genus of the Miconieae. Consequently, the
precise relationships and best generic disposition
for these two species must await detailed mono-
graphic studies of both Miconia and Tococa.
This species is named for Thomas B. Croat, a
student of neotropical Araceae who has contrib-
uted significantly to our knowledge of the neo-
tropical flora through his many fine collections
and publications.
ACKNOWLEDGMENTS
The research presented here was supported, in
part, by U.S. National Science Foundation Grants
DEB 76-83040, DEB 78-25620, and BSR
8614880 (Flora Mesoamerica). I am grateful to
the following individuals who contributed to this
study by assisting me in the field or collecting
specimens at my request: Greg de Nevers, Thom-
as F. Daniel, Barry E. Hammel, Gordon Mc-
Pherson, and Robert L. Wilbur. I thank the Mu-
seo Nacional de Costa Rica, the Organization for
Tropical Studies, and the Missouri Botanical
Garden for logistical support. I also thank John
J. Wurdack for comments on the manuscript; G.
W. Krantz (Dept. of Entomology, Oregon State
University) for identifying the mites; Charles
Turner (U.S. Department of Agriculture-ARS,
Albany, California) for information on plant-
mite associations; and the curators and staffs of
the following herbaria who either made their fa-
cilities available for on-site study or provided
loans and/or gifts of important collections: BM,
BR, C, CR, DUKE, F, G, K, MEXU, MO, NY,
P, PMA, US. The illustrations are the work
of Terry Bell (Fig. 1, 3-6) and Ellen del Valle
(Fig. 3).
LITERATURE CITED
ALMEDA, F. 1 984. New and noteworthy additions to the Me-
lastomataceae of Panama. Proc. Calif. Acad. Sci. 43(17):
269-282.
BAILLON, H. E. 1879. Melastomacees. Hist. PI. 7:1-65. (En-
glish transl. 7:1-65. 1881.)
BENSON, W. W. 1985. Amazon ant-plants. Pp. 239-266 in
Amazonia. G. T. Prance and T. E. Lovejoy, eds. Pergamon
Press, Oxford, England.
COGNIAUX, C. A. 1891. Melastomaceae. Pp. 1-1256 in
Monographiae phanerogamarum, Vol. 7. A. de Candolle and
C. de Candolle, eds. G. Masson, Paris.
FAEGRI, K. AND L. VAN DER PUL. 1971. The principles of
pollination ecology, 2nd ed. Pergamon Press, Oxford, En-
gland.
JACOBS, M. 1966. On domatia— the viewpoints and some
facts. I, II, III. Proc. Koninkl. Nederl. Akad. Wetensch. C69:
275-316.
MACBRIDE, J. F. 1 94 1 . Melastomataceae. In Flora of Peru.
Field Mus. Nat. Hist., Bot. Sen 13:249-521.
PERCIVAL, M. 1965. Floral biology. Pergamon Press, Oxford,
England.
PROCTOR, M. AND P. YEO. 1973. The pollination of flowers.
Williams Collins Sons and Co. Ltd., London, England.
PUTZ, F. E. ANDN. M. HOLBROOK. 1986. Notes on the natural
history of hemiepiphytes. Selbyana 9(l):61-69.
PROCEEDINGS
OF THE
CALIFORNIA ACADEMY OF SCIENCES
Vol. 46, No. 6, pp. 151-178, 13 figs., 2 tables. August 24, 1989
SYSTEMATIC REVISION AND DESCRIPTION OF NEW
SPECIES OF SUCKERMOUTH CATFISHES
(CHILOGLANIS, MOCHOKIDAE)
FROM CAMEROUN
By
Tyson R. Roberts
California Academy of Sciences, San Francisco, California 94118
ABSTRACT: Ten species of Chiloglanis occur in Cameroun: C. batesii Boulenger, 1904 (all or almost all river
basins); C. benuensis Daget and Stauch, 1963 (Niger); C. cameronensis Boulenger, 1904 (Lokunje, Kribi,
Kienke, Ntem); C, disneyi Trewavas, 1974 (Mungo and Cross); C. harbinger new species (Lokunje); C. niger
new species (Niger); C. polypogon new species (Cross); C. reticulatus new species (Congo); C. sanagaensis
new species (Sanaga); and C. voltae Daget and Stauch, 1963 (Niger).
Chiloglanis disneyi, C. harbinger, C. niger, C. polypogon, and C. sanagaensis are known only from Cameroun;
C. benuensis occurs in the River Benue in Cameroun and Nigeria. Chiloglanis reticulatus also occurs in the
Congo basin in northwestern Zaire. Chiloglanis batesii (including its synonym C. micropogon Poll, 1952)
occurs in suitable habitats throughout the Niger and Congo basins, and is the only species of Chiloglanis
known from the Chad basin. Chiloglanis cameronensis is the only species known in the Ogooue basin (Gabon).
Chiloglanis voltae is also known from the Benue in Nigeria and the Volta basin (Ghana and Burkina Faso).
In all of the major river systems of Cameroun two or even three species of Chiloglanis occur sympatrically.
Sympatric species usually differ strongly in sucker morphology, number and length of barbels, dentition,
coloration, adult size, and other characteristics. Secondary sexual dimorphism is marked in some species,
apparently absent in others.
Received September 22, 1987. Accepted January 9, 1989.
uensis Daget and Stauch, 1963; C. cameronensis
INTRODUCTION „ . innj, ^ j- • T im/i /~
Boulenger, 1904; C. disneyi Trewavas, 1974; C.
Chiloglanis belongs to the endemic African micropogon Poll, 1952; and C. voltae Daget and
catfish family Mochokidae. Adults range in size Stauch, 1963. For a review of the literature on
from about 18 to 100 mm standard length. The species of Chiloglanis and their distribution see
sucker or oral disc is formed by the enlarged Daget etal. (1986). Two recent papers have dealt
upper and lower lips. The maxillary barbels lie with Chiloglanis from parts of Cameroun: the
free at the sides of the upper lip but the man- Camerounian portion of the Benue (Niger basin)
dibular or mental (chin) barbels are largely in- by Daget and Stauch (1963), and the Mungo and
corporated into the lower lip. Meme, small coastal streams in western Cam-
A total of 34 species of Chiloglanis has been eroun, by Trewavas (1974). The results of these
recognized, of which six have been reported from studies and my interpretations of them are in-
Cameroun: C. batesii Boulenger, 1 904; C. ben- corporated in the present paper.
[151]
152
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 6
Most species of Chiloglanis are represented in
museum collections by only a few specimens,
often in poor condition. In 1980 I travelled ex-
tensively in Cameroun, collecting in the Niger,
Ndian, Cross, Sanaga, and Congo basins. Ad-
ditional collections were made in Cameroun in
1987-88. Specimens of Chiloglanis were col-
lected readily with a nylon flyscreen pushnet in
rocky substrate, gravel bars, and sand riffles. This
resulted in discovery of four undescribed species.
An additional new species, from the Lokunje ba-
sin, was discovered in the fish collection of the
California Academy of Sciences. I visited the
Lokunje in December 1987 but was unable to
collect additional specimens. Five of the six
species previously reported from Cameroun are
recognized as valid. After study of over 40 lots
and nearly 2,000 specimens referable to C.
batesii and C. micropogon, including type spec-
imens, I regard C. micropogon as a junior syn-
onym of C. batesii. Chiloglanis batesii is perhaps
the most widely distributed species of Chilo-
glanis. It occurs in suitable habitat in all or al-
most all river basins of Cameroun, throughout
the Niger and Zaire basins, and is the only species
of the genus in the Chad basin; it is absent west
of the Niger basin. Chiloglanis disneyi, hitherto
known only from the Mungo, is reported from
the Cross basin. The new species described from
Cameroun are: C. niger (Niger basin); C. poly-
pogon (Cross); C. sanagaensis (Sanaga); C. har-
binger (Lokunje); and C. reticulatus (Congo). The
first four are known only from Cameroun. The
fifth is also present in the Congo basin in Zaire.
Little is known about the biology of Chilo-
glanis or of any other mochokids. At least some
species of Chiloglanis have surprisingly large eggs.
Ovarian eggs of 1.7 and 1.9 mm were reported
in C. batesii of 28 and 35.5 mm, respectively, by
Trewavas (1974:363). A 37.2 mm specimen from
Kivu has ovarian eggs of 2.0 mm. The total num-
ber of eggs in this specimen is at least a dozen,
but no more than 20 (all but one or two have
been left in situ). The holotype of C. reticulatus
is a gravid female with about 40 eggs to 1 .8 mm.
In C. batesii and in C. reticulatus, ripe ovaries
contain several size classes of eggs. Sexually ma-
ture males of all species have lobate testes, which
in very ripe individuals may cause abdominal
expansion comparable to that caused by enlarged
ovaries of gravid females. In several species males
have an elongate genital papilla, the functional
significance of which is unknown. Males of a few
species have the humeral process of the pectoral
girdle greatly enlarged, but the significance of this
is also unknown. Chiloglanis feed mainly on in-
sects and attached algae. The species differ strik-
ingly in dentition and other morphological fea-
tures related to feeding, but these differences have
not been correlated with particular foods.
STUDY MATERIAL
Material examined is deposited in the follow-
ing institutions (standard symbolic codes from
Leviton et al. 1985): BMNH, British Museum of
Natural History; CAS and CAS-SU, California
Academy of Sciences; FMNH, Field Museum of
Natural History (Chicago); MCZ, Museum of
Comparative Zoology; MHNG, Museum d'His-
toire Naturelle, Geneve; MNHN, Museum Na-
tional d'Histoire Naturelle; MRAC, Musee Roy-
al de FAfrique Centrale (Tervuren); UMMZ,
Museum of Zoology, University of Michigan; and
USNM, Smithsonian Institution.
Camerounian and non-Camerounian material
of species of Chiloglanis found in Cameroun is
listed in the systematic account of each species.
The abbreviation "R.", always placed in front
of the proper name, stands for River, riviere, Rio
(Spanish), and rio (Portuguese). I have adopted
this practice when writing on African freshwater
fishes because the same river is often shared by
two or more English-, French-, Spanish-, or Por-
tuguese-speaking countries; authors in French
usually refer to any river, regardless of what
country it is in, as "riviere," while authors in
English use "River"; and although in English the
word "River" may precede or succeed the proper
name, in the other languages the word for river
always precedes the proper name.
The following comparative material of non-
Camerounian species of Chiloglanis and of the
suckermouthed mochokid genera Atopochilus and
Euchilichthys has been examined:
Chiloglanis brevibarbis Boulenger, 1902: MHNG 2000.68-82,
15: 29.3-57.7 mm, Kenya, R. Athi at Kabaa.
Chiloglanis congicus Boulenger, 1920: CAS(SU) 66706, 51.6
mm, Congo basin, probably Stanley Pool; CAS 44887, 2:
58.0-59.5 mm, aquarium specimens; CAS 60812, 60.7 mm,
Congo basin, R. Inkisi below Chute de Sanga.
Chiloglanis deckenii Peters, 1868: MHNG 1443.79-83, 5: 25.8-
33.4 mm, Kenya, R. Kerio near Tambach.
Chiloglanis neumanni Boulenger, 1911: UMMZ 199975,
199988, and 200183, 19: 19.5-40.8 mm, Zambia, Luapula-
Congo basin.
ROBERTS: CAMEROUNIAN SUCKERMOUTH CATFISHES
153
Chiloglanis niloticusBoulenger, 1900:MNHN 1907-230,39.4
mm, R. Nile at Arko, Sudan (syntype).
Chiloglanis occidental Pe\\egrin, 1933: UMMZ 182016, 24.0
mm, tributary of R. Senegal; CAS(SU) 62926, 56: 19.9-40.5
mm, Ghana, R. Ankwao; CAS(SU) 62927, 4: 30.7^6.4 mm,
Ghana, Opon Valley; CAS(SU) 62928, 56.7 mm, Ghana, R.
Ahiri near Wasa Akropong; MCZ 48063, 25.6 mm, Ghana,
Pra basin, Twifu Prasu; V. G. F. Smith collection, uncat.,
26.4 mm, Nigeria, Niger basin, R. Lere, Jos Plateau; BMNH
1935.5.29:52, 18.8 mm, Nigeria, Niger basin, R. Kaduna.
Chiloglanis polyodon Norman, 1932: BMNH 1 90 1 . 1 2.26, Sier-
ra Leone, headwaters of R. Bagbwe (holotype).
Chiloglanis sardinhai Ladiges and Voelker, 1961: USNM
265443, 5: 28.1-37.1 mm, Angola, Benguela, Caibambo.
Atopochilus savorgnani Sauvage, 1878: CAS(SU) 15715, Cam-
eroun, R. Ntem, Nyabessan.
Atopochilus sp. undet. a: CAS(SU) 64793, 98.2 mm, Congo
basin, Stanley Pool.
Atopochilus sp. undet. b: CAS 60815, 31.5 mm, Congo basin,
R. Lufu.
Euchilichthys dybowskii (Vai\\anl, 1892): MNHN 1892-81, 34.0
mm, Congo basin, Oubanghi (syntype); MCZ 48356, 5: 40.8-
63.2 mm, Congo basin, R. Mbomou at Gozobangui.
Euchilichthys royauxi Boulenger, 1902: CAS(SU) 64612, 2:
89.6-175 mm, Congo basin, rapids of lower Congo at Kin-
suka.
Characters Useful in Distinguishing
Species of Chiloglanis
Apart from Trewavas (1974), almost no au-
thors have given adequate discussion or defini-
tions of the characters used to define species of
Chiloglanis. Some easily observed features of the
sucker and dentition have not been utilized, and
discussion of variation with age or sex is gen-
erally lacking. Observation of previously unob-
served or unreported features, such as tubercles
and taste buds, is facilitated by scanning electron
microscopy. Some useful characters, none pre-
viously utilized, can be observed in radiographs.
SUCKER AND BARBELS.— All species of Chi-
loglanis have a well-developed sucker or oral disc
(Fig. 1), made of the greatly expanded and united
upper and lower lips. The lower lip incorporates
much or almost all of the mental (mandibular or
chin) barbels. The most obvious differences in
the sucker of various species involves the mental
barbels. In some species they project consider-
ably from the posterior border of the lower lip,
in others they project moderately, and in still
others slightly or not at all. Some species have
only two prominent mental barbels on each side
of the lower lip. These correspond to the primary
or primitive pair of mental barbels found in al-
most all groups of catfishes. In other species there
are up to five auxiliary mental barbels on each
side of the lower lip. In most species with aux-
iliary mental barbels, these occur inside the outer
primary mental barbel, but one species (C. poly-
pogori) usually has an auxiliary mental barbel
external to the outer primary mental barbel.
Camerounian species of Chiloglanis with mental
barbels projecting considerably from the lower
lip (i.e., with "large" mental barbels) are C. ben-
uensis, C. polypogon, C. reticulatus, and C. vol-
tae. Chiloglanis reticulatus and C. voltae have
only primary mental barbels, while C. benuensis
and C. polypogon have well-developed auxiliary
mental barbels. Chiloglanis voltae is unique in
having very elongate outer mental barbels and
very short inner ones. Species with mental bar-
bels projecting very little ("minute" mental bar-
bels) are C. batesii, C. disneyi, C. harbinger, and
C. niger. Chiloglanis cameronensis and C. sa-
nagaensis have moderately projecting mental
barbels.
Other obvious differences between species in-
volve size and shape of the sucker. In some species
it is transversely oval (i.e., broader than long)
and relatively (proportionately) small compared
to the size of the fish, e.g., in C. benuensis, C.
polypogon, and C. reticulatus. In others it is rel-
atively round, e.g., C. disneyi, C. harbinger, and
C. niger. The last two species have perhaps the
largest suckers of any species of Chiloglanis. The
papillae on the sucker also differ among species.
In C. benuensis, C. polypogon, and C. reticulatus
they are especially large, particularly on the me-
dian part of the lower lip. In C. voltae the papillae
are greatly reduced or absent. In C. sanagaensis
they are extremely numerous, evenly distributed
and of nearly the same size over virtually the
entire sucker. In all species the two halves of the
lower lip are separated distally by a median cleft.
In most species the cleft is very deep, so that
most of the median papillate area of the lower
lip is also cleft, but in C. niger the cleft is shal-
lower and the median papillate area undivided.
The boundary between upper and lower lips is
relatively well defined in some species, e.g., C.
benuensis, C. polypogon, and C. reticulatus, but
is indistinguishable in others, e.g., C. niger. The
upper lip of C. disneyi extends posteriorly under
the lower and is broadly fused to it, but retains
a free posterior margin where it has joined the
lower lip.
Each half of the lower lip has two medullary
cartilages. Difficult to see in some species or spec-
154
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 6
FIGURE 1 . Suckers of Camerounian species of Chiloglanis. a, C. reticulatus, 31.3 mm paratype (Congo basin); b, C. polypogon,
25.0 mm paratype (Cross basin); c, C. sanagaensis, 28.6 mm paratype (Sanaga basin); d, C. batesii, 21.7 mm (Sanaga basin); e,
C. disneyi, 23.0 mm (Cross basin);/ C. niger, 27.8 mm paratype (Niger basin).
imens, in others they are large, naturally reddish
or purplish, and readily observed with trans-
mitted light. The distal portion of each medullar
cartilage typically is trifurcate; occasionally a car-
tilage may have four branches (see Trewavas
1 974:360-36 1 , fig. 5a). The branches of the med-
ullary cartilages are longest in species with large
barbels. In C. reticulatus and C. voltae, which
have no auxiliary mental barbels, the medullary
cartilages are nevertheless tri- or quadrifurcate.
ROBERTS: CAMEROUNIAN SUCKERMOUTH CATFISHES
155
The large central branch of each medullary car-
tilage lies in the middle of a primary mental bar-
bel and extends almost to its tip. The lateral
branches, much shorter than the central one, ex-
tend along the lateral margins of the portions of
the disc projecting posteriorly between the bases
of the barbels. In C. polypogon, in which the
portions of the disc between the primary mental
barbels are greatly prolonged as auxiliary mental
barbels, a lateral branch of the medullary carti-
lage lies in the center of each auxiliary barbel.
DENTITION.— The dentition of Chiloglanis, like
that of most mochokids, is highly specialized,
and its description has given rise to much con-
fusion. As in Synodontis, the mochokid genus
with the largest number of species, the teeth are
S-shaped. An exception is provided by C. voltae,
in which the teeth are almost straight. The mor-
phology of the teeth has been described and il-
lustrated for many species of Synodontis (Poll
1971), but this has not been done for Chiloglanis.
The premaxillary (upper jaw) teeth are of two
sizes, large and small. The large teeth are ar-
ranged in two oval patches, one on each pre-
maxillary, separated by a toothless median gap.
The teeth in these patches form somewhat irreg-
ular rows that nevertheless can be counted; the
number varies from two to eight. Most species
have three to five rows, C. batesii and C. disneyi
three or four, C. harbinger to seven, C. niger to
seven or eight; C. voltae only two. In addition to
the oval patches of large teeth, the upper jaw
bears numerous small teeth forming a shallowly
V-shaped series across the angle of the upper jaw
where the flat external surface bearing the oval
patches curves sharply upwards toward the roof
of the mouth. Some of these teeth are firmly
implanted on the premaxillary bones, while oth-
ers lie more or less loose in the membrane be-
tween the premaxillary bones. The innermost or
most posterior row of teeth in this series may be
somewhat longer and more slender, forming a
row of ciliiform teeth (sensu Trewavas 1974:359
et seq.). This tooth row, while it varies in prom-
inence and degree of differentiation from the rest
of the small premaxillary teeth, is present in most
species of Chiloglanis. According to Trewavas
(1974:358-359) the disposition of small pre-
maxillary teeth differs in C. batesii and C. mi-
cropogon: "in C. batesii the attached and free
teeth compose a uniform group, but in C. mi-
cropogon the median teeth are still smaller and
sections show that the membrane in which they
lie forms a minute pouch opening backwards,
the teeth at the edge of the pouch forming a
transverse series across the midline. Examined
with a dissecting microscope, these latter appear
as a row of minute ciliiform teeth on a small
membranous curtain." This lengthy quote pro-
vides the best account of a peculiar feature of the
dentition of Chiloglanis, and purportedly a basis
for distinguishing C. batesii and C. micropogon.
My observations suggest that a complete row of
ciliiform teeth is characteristic of most species
of Chiloglanis except C. voltae. In many freshly
preserved specimens identified by me as C.
batesii, the ciliiform teeth are highly visible, and
the membranous pouch described by Trewavas
is fully evident. In other specimens, including
some collected by Bates at Efulen (type locality
of C. batesii), they appear at first glance to be
absent. However, in at least some of these spec-
imens the delicate membrane bearing the cili-
iform teeth is shrunken and collapsed against the
roof of the mouth, so that the teeth no longer
project normally downwards. In this connection
it may be noted that a small specimen from the
R. Lobi in Cameroun, originally identified as C.
batesii by Boulenger, was re-identified as C. mi-
cropogon by Trewavas (1974:361), and that part
of the type series of C. batesii was collected in
tributaries of the Lobi. It may also be noted that
in another species, C. brevibarbis, Trewavas
( 1 974:364) reported ciliiform teeth present in only
one specimen among several she examined.
The mandibular or lower jaw teeth, arrayed in
a relatively small area near the symphysis of the
lower jaw in all species of Chiloglanis except C.
voltae, usually comprise an internal row of func-
tional teeth and an external row (rarely two or
three external rows) of replacement teeth (Fig.
2). In most species the teeth in the replacement
row erupt nearly simultaneously, so that de-
pending upon the stage in the replacement cycle,
teeth of the lower jaw are either in only a single
row (the functional row) or in two rows (func-
tional and replacement). This simple situation
may be complicated when the row of functional
teeth has been shed or only partially shed, and
only part of the replacement tooth row has moved
into functional position. In some species, espe-
cially those with numerous teeth such as C. dis-
neyi, C. harbinger, and C. niger, the replacement
process tends to be very regular, and it seems
156
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 6
FIGURE 2. Mandibular teeth of Camerounian species of Chiloglanis. a, C. reticulatus, 31.3 mm paratype; b, C. polypogon,
25.0 mm paratype; c, C. sanagaemis, 28.6 mm paratype; d, C. batesii, 21.7 mm; e, C. disneyi, 23.0 mm;/ C. niger, 27.8 mm
paratype.
that an old row of functional teeth is shed nearly replaced by corresponding teeth from the re-
simultaneously or as a unit, to be replaced at placement row before the more lateral functional
once by a complete row of replacement teeth. In teeth are shed. The sequence of tooth replace-
others, e.g., C. polypogon and C. congicus, it seems ment has not been investigated in detail,
that those near the symphysis fall out first, to be In Chiloglanis generally, allowing for individ-
ROBERTS: CAMEROUNIAN SUCKERMOUTH CATFISHES
157
ual variation due to loss or injury of teeth and
irregularities in replacement, there are basically
two rows of mandibular teeth, an internal func-
tional row and an external replacement row; the
number of teeth of the left and right side of the
lower jaw is usually equal. The number of teeth
in the replacement row usually corresponds to
that in the functional row. There is some increase
in number of teeth with growth, but throughout
the size range of specimens examined for this
report, the ontogenetic increase does not amount
to more than four to six functional mandibular
teeth even in species with the highest tooth counts.
In numerous specimens of C. batesii, the func-
tional tooth row has 6 teeth on the left side and
6 on the right, and the replacement tooth row
has the same count. Thus, the basic number of
mandibular teeth in this species usually is 6 + 6.
Although some specimens of C. batesii have few-
er than 6 + 6 (usually 5 + 5 or 4+4) I have not
observed any with more than 6 + 6.
Species of Chiloglanis previously reported to
have numerous mandibular teeth are C. congi-
cus, C. disneyi, C. lamottei Daget, 1948, C. mar-
lieri Poll, 1952, and C. polyodon. These were
discussed in connection with the description of
C. disneyi by Trewavas (1974:365-367). Chi-
loglanis congicus has mandibular teeth tightly
bunched together ("en bouquet"). The teeth near
the symphysis (middle of the bouquet) have their
crowns projecting more than those at the sides.
Trewavas gave the mandibular tooth count as 1 8
but did not explain how the count was made. In
four specimens of C. congicus I find the total
number of mandibular teeth is 12, 16, 16, and
1 9. In the first three specimens there are basically
two rows of teeth (one functional, one replace-
ment) with 4+4 or 5 + 5 teeth in each row. The
fourth specimen has two or three rows of teeth,
one functional and one or two replacement, with
the rows deranged and some teeth missing. On
each side of any of the rows the maximum num-
ber of teeth is 4. Chiloglanis lamottei, known
only from the holotype from the R. Cavally, has
a single row of 1 4 mandibular teeth (Daget 1 948),
presumably 7 + 7. Chiloglanis polyodon, known
only from the holotype from the R. Bagbwe, was
originally described as having mandibular teeth
"arranged in two parallel rows, each consisting
of 1 5 or 16 teeth, those of the inner row larger
and more curved than those of the outer (Nor-
man 1932:1 84). Some 40 years later the holotype
was re-examined and found to have 1 5 inner and
14 outer mandibular teeth bunched together"
TABLE 1 . Mandibular tooth counts in species of Chilogla-
nis.
C. polypogon
C. reticulatus
C. congicus
C. sanagaensis
C benuensis
C. cameronensis
C. batesii
C. lamottei
C. polyodon
C. disneyi
C. niger
C. marlieri
C. harbinger
3 + 3 or 4+4
3 + 3 to 5 + 5?
4+4 (occasionally 5 + 5?)
4+4 or 5 + 5 (rarely 6+6)
3 + 3 to 6+6
4+4 to 8 + 8
4+4 to 6 + 6
7 + 7
7 + 7 to 8 + 8
8 + 8 to 10+10
8+8 to 10+10
13+ 13. (or 14+14?)
13+13to 15+15
(Trewavas 1974:367). Both accounts are cited
here because of the possibility that in the inter-
vening period a tooth was removed from each
row, possibly by Norman himself. The basic
mandibular tooth count for C. polyodon presum-
ably is 7 + 7 or 8 + 8. Chiloglanis marlieri is known
only from the holotype collected in the R. Nda-
kirwa, a tributary of the R. Lowa, west of Lake
Kivu. As originally illustrated this specimen has
one complete mandibular tooth row of 1 5 + 1 5
teeth, plus an incomplete internal row with 7
teeth on the extreme left side, 8 on the extreme
right, and none in the large intervening space
(Poll 1952:227, fig. 2). Trewavas (1974) re-ex-
amined this specimen and reported the mandib-
ular teeth as "one row of 26, a widely interrupted
row of 6 + 7 and 2 incompletely erupted teeth at
the midline." Thus the basic tooth count pre-
sumably is 13 + 13 or 14+14. Mandibular tooth
counts for various species of Chiloglanis are pre-
sented in Table 1 .
TASTE BUDS.— Taste buds (Figs. 3, 4) were re-
ported (not figured) on the papillae or "bosses of
thickened epithelium" on the oral disc of C.
batesii and C. micropogon (regarded here as a
synonym of C. batesii) by Trewavas (1974). I
have observed taste buds with scanning electron
microscopy in one or two specimens of five
species from Cameroun, and failed to find them
in specimens of a sixth (Figs. 3, 4). The following
remarks are presented as generalizations on the
taste buds of these six species, but the limited
number of specimens on which they are based
should be kept in mind. The buds are largest and
most abundant in C. polypogon and C. reticu-
latus. In these species they are distributed fairly
uniformly on the papillae and barbels of the low-
er lip. In C. reticulatus the buds are about 50 /*m
diameter, and the sensory field (with numerous
158
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 6
W^i -• V '•". '
FIGURE 3. Taste buds in Camerounian species of Chiloglanis. a, b, C. reticulatus, 31.3 mm; c, d, C. polypogon, 25.0 mm; e,
f, C. sanagaensis, 28.6 mm. Horizontal field widths: a, c, e (top row) = 450 nm\ b, d, /(bottom row) = 45 /*m.
minute pores for the sensillae) of each bud is
about 1 8 nm diameter. The sensory field projects
from the summit or apex of the taste bud (Fig.
36). The density of the taste buds, calculated
from Figure 3b, is about 240 per square mm; this
is about the same as that reported for taste buds
in the mouth and pharynx of the common Eur-
asian loach Cobitis taenia Linnaeus, 1758 by Ja-
kubowski (1983:850). The taste buds and sen-
sory fields of C. polypogon are of about the same
size, but the sensory field tends to be sunken into
the middle of the bud rather than elevated (Fig.
3d). A similar condition is reported for the ex-
ternal taste buds of Cobitis taenia by Jakubowski
(1983:851, fig. ID) but has not been observed in
other Chiloglanis. Chiloglanis disneyi has rela-
tively large, numerous buds on the lower lip. The
buds are up to 22 ^m diameter and the sensory
fields to 7 or 8 /j.m diameter (Fig. 4d). Slightly
larger buds occur in C. batesii (Fig. 46). Extensive
viewing of the sucker of C. niger with SEM re-
vealed no large or well formed taste buds. Only
a few small structures tentatively identified as
taste buds were observed on papillae on the lower
lip (Fig. 4f). The largest bud is 1 2 nm diameter
with a sensory field of about 8 /urn diameter. The
sensory field has fewer pores than observed in
other species of Chiloglanis. Taste buds appar-
ently are absent in C. sanagaensis. I could not
find any on the sucker of a 28.6 mm specimen.
Although the oral papillae of this specimen have
raised areas similar to those where taste buds
occur in other species, their summit is occupied
only by densely convoluted microvillar ridges,
without any sensillar pores. The structures su-
perficially most similar to taste buds observed
in C. sanagaensis are illustrated in Figure 3e, f).
The species of Chiloglanis with the largest and
most numerous taste buds, C. polypogon and C.
reticulatus, have suckers with large barbels and
large papillae. Taste buds of intermediate size
and abundance occur in C. batesii and C. disneyi,
ROBERTS: CAMEROUNIAN SUCKERMOUTH CATFISHES
159
FIGURE 4. Taste buds in Camerounian species of Chiloglanis (cont.). a, b, C. batesii, 21.7 mm; c, d, C. disneyi, 23.0 mm; e,
f, C. niger, 27.8 mm. Horizontal field widths: a, c, e (top row) = 450 nm; b, ^./(bottom row) = 45 nm.
species with greatly reduced barbels and rela-
tively small papillae. Chiloglanis niger, with a
very large sucker, minute barbels, and moder-
ately large papillae, has taste buds few in number
and poorly developed. Chiloglanis sanagaensis,
with moderately elongate barbels and numerous
well-developed papillae, apparently lacks taste
buds. Finally, it should be noted that taste buds
are not limited to the sucker in Chiloglanis. Rel-
atively large buds occur on the dorsal surface of
the head in C. reticulatus (Fig. 5e, f).
TUBERCLES.— Tubercles, composed of keratin-
ized and sometimes also greatly enlarged super-
ficial epidermal cells, occur on the head and body
of all or almost all species of Chiloglanis. In con-
trast to the situation in most cyprinoid fishes,
the tuberculation apparently is not sexually di-
morphic. Among Camerounian species it is
particularly well developed in C. reticulatus, C.
polypogon, and C. batesii. Although less evident
because they are fewer and further apart, well-de-
veloped tubercles are present also in C. disneyi,
C. niger, and C. sanagaensis. The Camerounian
(and Zairean) species with the most striking tu-
berculation is C. reticulatus, in which large spec-
imens of both sexes have long tuberculated ridges
forming a network or reticulum covering the dor-
sal and lateral surfaces of the head and much of
the body. The ridges are especially evident on
the head. They are up to 10 cells wide; the crest
of a ridge is regularly tuberculate, each tubercle
consisting of about a dozen enlarged epidermal
cells (Fig. 5d). The surface of the non-tuberculate
polygonal cells forming the ridge is not smooth,
the center of each cell being occupied by a low-
lying roughened protuberance evidently equiv-
alent to an unculus (for review of unculi see Rob-
erts 1982). The epidermal surfaces between the
ridges, in contrast, consist of slightly smaller, non-
unculiferous (and therefore relatively smooth-
surfaced) polygonal cells, widely scattered taste
buds, and even more widely scattered pores. The
160
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 6
FIGURE 5. Epidermal features on head of Chiloglanis reticulatus. a, dorsal view of head of 31.3 mm paratype, with elongate
tuberculate ridges well developed (but not connecting up to form a network as in larger specimens); b-e, dorsal surface of head
posterior to eye of 41.6 mm holotype, with well-developed network of tuberculated ridges and isolated taste buds;/ enlarged
view of taste bud in e. Horizontal field widths: a, 7.5 mm; b, 1 mm; c, 750 ^m; d, 360 ^m; e, 1 10 /j.m; and/ 1 1 /am.
pores, two of which are seen in Figure 5b, c, lie
in shallow depressions about 50 nm diameter
and are themselves about 20 pm diameter. The
taste buds, elevated above the general epidermal
surface, are also about 20 /um diameter (Fig. 5/).
The gross aspects of all of these features (but not
their cellular nature) can be observed under the
dissecting microscope. Although no other species
of Chiloglanis observed by me has tuberculated
ridges forming such an extensive and well-de-
fined network as C. reticulatus, more or less elon-
gate tuberculated ridges are present on some larg-
er specimens of C. batesii, C. occidentalis, and
C. polypogon.
Numerous round or oval tubercles, comprising
some 60 to 80 cells and about 40 MHI diameter,
occur on the dorsal surface of the head of C.
polypogon (Fig. 6a, b). Similar tubercles were
observed in C. batesii from Ngudi (Fig. 6e). Some
larger C. batesii have tuberculated ridges, but
these were not observed with SEM. In C. poly-
pogon the cells comprising a tubercle are uncu-
liferous, whereas in the specimens of C. batesii
observed they are not. In a 28.6 mm C. sana-
gaensis, prominent tubercles were observed on
the dorsal surface of the snout. These have a
round base 30 to 35 nm diameter, and a rounded
or pointed apex to 20 to 30 nm high. They appear
to have a perimeteral dehision ring (Fig. 6d). The
head of this specimen exhibits many shallow cra-
ter-like depressions, tentatively identified as scars
from tubercles that had been shed. Of the species
of Chiloglanis examined with SEM, the weakest
tuberculation is in C. niger. Tubercles on its head
are widely separated, up to 40 /urn diameter, but
not elevated, consisting of only a single layer of
thin, apparently individually dehiscent polygo-
nal epidermal cells, with irregular projections
ROBERTS: CAMEROUNIAN SUCKERMOUTH CATFISHES
161
FIGURE 6. Tubercles on head of Camerounian species of Chiloglanis. a, b, C. polypogon, 25.0 mm; c, d, C. sanagaensis, 28.6
mm; e, C. batesii, 25.7 mm (R. Ngudi);/ C. niger, 27.8 mm. Horizontal field widths: a, 250 nm; b, c, 115 nm; d, 105 nm; e,
450 /urn; and/ 105 nm.
(unculi?) arising from their centers or cell mar-
gins (Fig. 6/).
FIN RAYS.— Counts of fin rays are of limited
usefulness in distinguishing species of Chilogla-
nis. Most useful are dorsal and pectoral fin ray
counts. Most Camerounian species usually have
5 soft dorsal fin rays, but C. batesii, C. niger, and
C. voltae usually have 6. However, occasional
specimens of the former species have only 5 and
occasional specimens of the latter species have
6. Chiloglanis harbinger has 4 or 5. Many species
have 7 or 8 soft pectoral fin rays; no species has
been observed with fewer than 7. Chiloglanis
disneyi has 8 or 9, C. harbinger and C. niger 9
or 10.
In many specimens the number of anal fin rays
is difficult to determine. I have not obtained ad-
equate data to evaluate anal fin ray counts for
distinguishing species. Camerounian species have
a total of 7 to 12 anal fin rays, but much of the
variation is in the number of anterior simple
rays, which range from 2 to 5. Especially in spec-
imens with more simple rays, the anteriormost
ones tend to be very small, sometimes impos-
sible to count accurately in whole specimens or
even in radiographs. The number of branched
anal fin rays is usually 7 (range 5 to 8). All species
of Chiloglanis examined have 7 pelvic fin rays
and 8/9 principal caudal fin rays.
COLORATION.— The basic elements of color
pattern— lightly and darkly pigmented areas—
are remarkably constant in many species of Chi-
loglanis. This can be appreciated readily by com-
paring the photographs of various Camerounian
species in this paper. The main feature of the
color pattern is three broad dark vertical bands
on the dorsolateral body surface. Although the
outline of the bands is irregular, closer exami-
nation reveals the same or similar irregularities
in the bands of quite different species. This is
especially true for the very irregularly shaped
second band, lying between the adipose and anal
162
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 6
fins. This band has an anteroventral projection
where its ventral portion is more or less exten-
sively broken or invaded by a depigmented area
arising between the base of the pelvic and anal
fins and extending dorsoposteriorly. The posi-
tion and shape of this pale area, and the resulting
configuration of the dark vertical band between
the adipose and anal fins, is very evident in the
photos of C. reticulata, C. polypogon, C. sana-
gaensis, and C. disneyi. It is also visible in the
photos of C. niger and C. batesii, although in
these specimens the basic pattern is obscured by
their overall duskiness. The barred color pattern
on the caudal peduncle and caudal fin is similar
in many species. Some species, including C. poly-
pogon and C. reticulatus, have small, round, pale
spots scattered on the dark vertical bands of the
body. At first glance these spots appear irregular,
but their number and position is similar in var-
ious species. The peculiar conservatism of the
irregularly banded or barred color pattern of
Chiloglanis is paralleled by a number of other
groups of catfishes, notably the Southeast Asian
bagrid genus Leiocassis.
Most species of Chiloglanis are dull and cryp-
tically or disruptively colored in life. Live Chi-
loglanis niger are uniformly dark, almost black.
This is perhaps also true in C. harbinger for which
coloration has been observed only in preserved
specimens. The most brilliantly colored species
probably is C. disneyi, which has well defined
bright yellow and black bands. Sexual dichro-
matism is unknown.
A few species have distinctive color features.
Nearly all species have dark vertical bars on the
dorsolateral part of the body that end more or
less abruptly at about the level of the paired fins.
In C. harbinger the large anterior bar extends
partially onto the abdomen, the melanophores
becoming fewer and wider apart towards the ven-
tral midline. In C. cameronensis this bar often
continues without diminished intensity right
across the abdomen, which is therefore black.
This is especially so in samples of this species
from the Ntem and Ogooue basins. In most spec-
imens of C. niloticus and its close relative (or
conspecific) C. occidentalis, the upper and lower
caudal fin lobes each bear a dark horizontal stripe.
Although C. occidentalis is widely distributed in
western Africa, including the Niger basin, it has
not been found in Cameroun. None of the species
of Chiloglanis known from Cameroun have cau-
dal fin lobes with horizontal stripes. The dorsal
surface of the snout of C. batesii has a triradiate
pattern. This consists of a dark or dusky median
mark at the snout tip separated by pale inter-
spaces from an anterolateral dark or dusky mark
on each side of the snout. This pattern, observed
in specimens from throughout the extensive range
of C. batesii, is absent or poorly developed in
other species of Chiloglanis. Chiloglanis reticu-
latus, occurring sympatrically with C. batesii in
the Congo basin, often has a biradiate pattern on
the snout. This is due to a dark or dusky an-
terolateral mark on each side of the snout as in
C. batesii, but the large intervening area includ-
ing the snout tip is pale.
SEXUAL DIMORPHISM. — Testes of maturing
Chiloglanis have numerous lobes. In extremely
ripe males the lobes may be so enlarged as to
cause abdominal expansion comparable to that
of gravid females. In many species males have
an elongate genital papilla. Secondary sexual di-
morphism occurs in several species. Often this
involves enlargement of the anal fin, and some-
times also enlargement and modification of the
shape of the caudal fin. In males of C. macrop-
terus Poll and Stewart, 1975, from the R. Luongo
in Zambia (Congo basin), the anal and caudal
fins are both enlarged; the caudal fin, moderately
forked in females, is triangular in large males,
with the distal part of the triangle formed by
elongation of the middle caudal fin rays. Very
similar sexual dimorphism of the anal and caudal
fins occurs in a West African species, C. normani
from the R. Cavally, Cote d'lvoire. The only
Camerounian species in which sexual dimor-
phism of the caudal fin has been observed is C.
batesii. In this species the caudal fin is very deep-
ly forked, with the upper lobe considerably en-
larged and pointed or even filamentous in mature
males. Such modification of the upper caudal fin
lobe of male Chiloglanis has been reported pre-
viously only in C. micropogon, here considered
conspecific with C. batesii. In addition to the
enlarged and pointed upper caudal fin lobe, ma-
ture males of C. batesii have lobate testes, an
elongate genital papillae, enlarged anal fin, and
one other remarkable feature.
Radiographs of Chiloglanis from Cameroun
and other countries reveals several species with
a striking sexual dimorphism hitherto unre-
ported. In these species males have the humeral
process of the pectoral fin girdle more or less
ROBERTS: CAMEROUNIAN SUCKERMOUTH CATFISHES
163
TABLE 2. Vertebral counts in species of the suckermouth
mochokid genera Chiloglanis, Atopochilus, and Euchilichthys.
TABLE 2. Continued.
Abdom-
Abdom-
inal
Caudal
Total
inal Caudal Total
C. batesii
A. sp. undet. a
Mungo
15
16-17
3 1(2), 32(2)
Stanley Pool 19 16 35(1)
Cross
15
16-17
31(1), 32(1)
A. sp. undet. b
Ja
14
17-18
3 1(3), 32(1)
Luala
14-15
16-17
30(2), 3 1(7)
Lufu 21 16 37(1)
Lufu
14-15
15-17
30(4), 3 1(5), 32(1)
E. dybowskii
Inkisi
14-15
16-17
3 1(5), 32(5)
Oubanghi 19 13 32(1)
Kwango
15-16
16
3 1(1), 32(5)
Mbomou 18-20 12-14 31(2), 32(2)
Mbomou
14-15
16-17
31(2)
Kivu
15
17-18
32(1), 33(1)
E. royauxi
C. brevibarbis
Kinsuka 23-24 17-18 40(1), 42(1)
Athi
15-17
18-20
34(3), 35(10)
C. cameronensis
greatly enlarged. The enlargement is manifested
Lokunje
15
15-16
30(1), 31(1)
externally as a large, skin-covered convex boss
Mvila
Ntem
14-15
15
16-17
16
31(4)
31(2)
just above the pectoral fin. It is most pronounced
Diala
15
16-17
•* * v**/
3 1(2), 32(1)
in C. batesii, but also occurs in C. occidentalis
Ivindo
14-15
16
30(2), 3 1(2)
and C. polypogon. It apparently does not occur
C. congicus
in C. brevibarbis, C. cameronensis, C. congicus,
Congo
16-17
15-16
32(3), 33(1)
C. disneyi, C. niger, C. reticulatus, C. sanagaen-
C. deckenii
sis, or C. voltae, for all of which radiographs as
Kerio
16-18
15-17
33(4) 34(2)
well as specimens have been examined.
"»*"*\r/J •/~\* V
VERTEBRAL COUNTS.— Vertebral counts, until
C. disneyi
recently seldom employed in systematic studies
Mungo
Cross
15
15-16
15-16
15-17
30(1), 31(1)
31(6) 32(3)
of catfishes, can be very useful in distinguishing
•* * \U/J */A*\-//
species. Smaller or less elongate species of cat-
C. harbinger
fishes often have fewer vertebrae than their larger
Lokunje
15-16
15-16
30(1), 3 1(9)
or more elongate close relatives. Thus, the sisorid
C. niger
catfish Bagarius bagarius Hamilton-Buchanan,
Menchum
15
16
3KD
1822, is a small species superficially similar to
C. occidentalis
the very large species Bagarius yarrelli Sykes,
Ankwao
14-15
16-18
3 1(1), 32(3)
1841, but differs from it in having only 38-42
Pra
14-15
16-18
3 1(4), 32(2)
(mode 39) vertebrae instead of 40-45 (mode 43)
C. polyodon
(Roberts 1 983:437). Vertebral counts are of com-
Bagbwe
15
16
3KD
parable utility in distinguishing species of the
C. polypogon
bagrid genera Leiocassis and Mystus and of the
Cross
14-15
16-18
30(1), 32(10), 32(8)
pangasiid or schilbeid genus Pangasius (Roberts,
C. reticulatus
in press). Vertebral counts are determined readi-
Congo
14-15
16-19
3 1(5), 32(4), 33(5)
ly from radiographs of Chiloglanis. In some cat-
fish groups, due to variation in the length and
C. sanagaensis
Sanaga
14-16
16-17
31(1), 32(4), 33(1)
partial or total fusion of the anteriormost (We-
berian) vertebral centra with each other, and oc-
C. sardinhai
casional fusion of additional more-posterior ver-
Caimbambo
14-15
15-17
30(1), 3 1(4)
tebrae with them, it is difficult to obtain reliable
C. voltae
counts of abdominal vertebrae. In Chiloglanis,
Oti
13-14
16-17
29(2), 30(5), 31(2)
however, the four Weberian vertebrae have cen-
Lakdo
13-14
16-18
30(3), 3 1(1)
tra of about the same size that remain separate;
A. savorgnani
thus, they often may be counted directly from
Ntem
17-18
14-15
32(5), 33(3)
radiographs. Furthermore, the fourth vertebra is
164
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 6
readily identifiable by the large ossa suspensoria
and the fifth vertebra by a large but slender pair
of normally shaped ribs. In practice I commence
a count of the vertebral column with the fifth
vertebra and proceed backwards. Abdominal
vertebrae are defined as all those lying anterior
to the anal fin and its pterygiophores. If the an-
teriormost anal fin pterygiophore approximates
or even contacts the hemal spine of a vertebra
but fails to lie in front of it, the vertebra is count-
ed as abdominal. The hypural complex is count-
ed as one vertebra. This method gives repeatable
counts for most catfishes and many other fish
groups. Vertebral counts from a radiographic
survey of Chiloglanis and other mochokid genera
with sucker mouths are presented in Table 2.
The counts are of little use in distinguishing
most species of Chiloglanis, all of which are more
or less short bodied and none very large. Many
species have around 30 or 31 to 33 vertebrae.
The smallest species, C. voltae, has 29-3 1 . This
is close to the lowest number of vertebrae found
in other groups of catfishes and in ostariophysan
fishes generally. Somewhat higher counts (33-35)
occur in species from East Africa including C.
deckeniiand C. brevibarbis. Of particular interest
are the vertebral counts of the genera Atopochilus
and Euchilichthys, in which the species generally
are far larger (to 600 mm) than Chiloglanis and
have up to 42 vertebrae. In addition there is a
shift in the ratio of abdominal to caudal verte-
brae. In Chiloglanis the ratio is close to unity,
n:n or n:n+ 1 or n+2, while in Atopochilus and
Euchilichthys the abdominal vertebrae are more
numerous than the caudal vertebrae, so the ratio
is n:n— 2 to n— 7. This increase in abdominal
vertebrae appears related to an elongate abdo-
men accommodating an extremely convoluted
gut.
SYSTEMATIC ACCOUNT
Chiloglanis Peters, 1868
Chiloglanis Peters, 1868: 599 (type species Chi-
loglanis deckenii Peters, 1868, by monotypy).
Chiloglanis (gender masculine) differs from all
other genera of the endemic African catfish fam-
ily Mochokidae, except Atopochilus Sauvage,
1878 and Euchilichthys Boulenger, 1 900, in hav-
ing upper and lower lips greatly expanded and
united to form a sucker. Some species of Syno-
dontis have large lips superficially resembling
those of Chiloglanis but are readily distinguish-
able because they have mental barbels free from
sucker instead of incorporated into it. In Ato-
pochilus and Euchilichthys the mandibular teeth
are very numerous and spread across the entire
free margin of the lower jaw, whereas in all species
of Chiloglanis except C. voltae, the mandibular
teeth are relatively few and restricted to a rela-
tively small area near the symphysis of the lower
jaw. In Chiloglanis the pectoral fin spine is in-
variably smooth, never serrate, while in Atopo-
chilus and Euchilichthys it bears few to many
serrae. Euchilichthys and Atopochilus are very
poorly represented in Cameroun. Euchilichthys
dybowskii is present in the R. Ngoko (lower R.
Ja, Congo basin) and Atopochilus savorgnani in
the R. Ntem.
KEY TO CHILOGLANIS OF CAMEROUN
1A. Mental barbels moderately to very long
(Fig. 1 a-c) 2
IB. Mental barbels short, sometimes absent
(Fig. \d- /) 7
2 A. No auxiliary mental barbels (Fig. la) 3
2B. From 2 to 5 well developed auxiliary
mental barbels on each side of lower lip
(Fig. 1 b) 4
3A. Sucker with well developed papillae; large
teeth on premaxilla in several rows; man-
dibular teeth well developed; adults to
over 40 mm (Congo basin) ... C. reticulatus
3B. Sucker without papillae; large teeth on
premaxilla in two rows; mandibular teeth
in two rows; adults to 27 mm (Niger ba-
sin) C. voltae
4A. Mental barbels very large; sucker trans-
versely oval, middle of lower lip with
greatly enlarged papillae (Fig. la, b) 5
4B. Mental barbels moderately large; sucker
round, or longer than broad, entire sur-
face covered with papillae of similar size
(Fig. 1 c) 6
5A. Accessory mental barbels no more than
3 on each side of lower lip; mandibular
teeth 4+4 to 7 + 7 (Niger basin)
C. benuensis
5B. Accessory mental barbels from 3 to 5 on
each side of lower lip; mandibular teeth
3 + 3 or 4+4 (Cross basin) C. polypogon
6A. Body relatively elongate; abdomen not
greatly expanded, more or less exten-
sively darkly pigmented; mandibular teeth
ROBERTS: CAMEROUNIAN SUCKERMOUTH CATFISHES
165
4 +4 to 8 + 8 (coastal basins east of Sanaga
basin) C. cameronensis
6B. Body stubby; abdomen considerably ex-
panded, pale; mandibular teeth usually
4+4 or 5 + 5, rarely 6+6 (Sanaga basin)
C. sanagaensis
7 A. Mandibular teeth (except in very small
specimens) 8 + 8 or more; sexual dimor-
phism absent or weak 8
7B. Mandibular teeth 3 + 3 to 6 + 6; sexual di-
morphism marked, larger males with en-
larged anal fin, enlarged and pointed up-
per lobe of caudal fin, and greatly enlarged
and expanded humeral process of pec-
toral girdle (in all or nearly all river ba-
sins) C. batesii
8 A. Sucker relatively huge; large premaxillary
teeth in seven or eight rows; top and sides
of body extensively covered with mela-
nophores, so it appears almost uniformly
dark; pectoral fin soft rays 9 or 1 0; caudal
peduncle moderately to very slender 9
8B. Sucker relatively small; large premaxil-
lary teeth in three or four rows; top and
sides of body with very sharply contrast-
ing dark vertical bands alternating with
pale areas; pectoral fin soft rays 8 or 9;
caudal peduncle relatively deep
C. disneyi
9 A. Mandibular teeth 8+8 to 10+10; dorsal
fin soft rays 6; adipose fin short (Niger
basin) C. niger
9B. Mandibular teeth 13+13 to 15 + 15; dor-
sal fin soft rays 4 or 5; adipose fin long
(Lokunje basin) C. harbinger
Chiloglanis reticulatus new species
(Figs, la, 2, 3a, b, 5, 7)
HOLOTYPE.— CAS 60786, 41.6 mm gravid female, Camer-
oun, Congo basin, R. Mwamedjwel, a very small stream 2-3
km W of Yokadouma, Tyson R. Roberts, 15 Jan. 1980.
PARATYPES.— CAS (SU) 47471, 36.3 mm, Cameroun, Congo
basin, R. Momjepom, Yokadouma, A. I. Good, 20 Mar. 1936;
CAS 60787, 3 1 .0 mm, Zaire, Congo basin, R. Lubazi, tributary
of R. Luala, 3 km S of Kibunzi, Tyson R. Roberts and Donald
J.Stewart, 17 Aug. 1973; CAS 60788, 10: 19.1-38.1 mm, Zaire,
Congo basin, R. Lufu, 2 km NW of Lufu village, Tyson R.
Roberts and Donald J. Stewart, 2 Aug. 1973 CAS 60789, 43:
9.4-37.4 mm, Zaire, Congo basin, tributary of R. Luala 26 km
N of Kibunzi, Tyson R. Roberts and Donald J. Stewart, 17
and 24 Aug. 1973.
Chiloglanis reticulatus is a relatively small
species, largest specimen 41.6 mm standard
length; sucker relatively small, transversely oval,
with very large barbels; auxiliary mental barbels
absent; large teeth of premaxillary in 2 to 4 rows;
mandibular teeth in 1 or 2 rows, very irregularly
arranged, basic count from about 2 + 2 to 5 + 5;
eyes large; dorsal fin soft rays usually 5 (rarely
6); pectoral fin soft rays 7 or 8; adipose fin long,
its origin well in advance of a vertical line through
anal fin origin; caudal peduncle moderately deep;
caudal fin moderately forked, lower lobe longer
than upper. Larger specimens of both sexes with
well-developed tuberculated epidermal ridges on
head and body, characteristically forming a dense
network or reticulum on dorsal surface of head.
The largest male, 37.4 mm, from tributary of R.
Luala, has anal fin and lower lobe of caudal fin
greatly enlarged; humeral process of pectoral spine
not enlarged. Its genital papilla is elongate, with
a peculiarly enlarged and swollen tip, so that it
resembles somewhat the proboscis of an ele-
phant. Sexual dimorphism absent or slight in
other specimens examined. Holotype in life with
chocolate-brown dark bands and cream-colored
interspaces.
ETYMOLOGY.— The name reticulatus (Latin,
adjective) refers to the reticulum or network of
tuberculated epidermal ridges.
DISTRIBUTION.— Known only from the north-
western portion of the Congo basin in Cameroun
and Zaire.
Chiloglanis voltae Daget and Stauch, 1963
Chiloglanis voltae Daget and Stauch, 1963:99, fig. 4 (type lo-
cality Bougouri Ba au pont de Nabere, Volta basin; also
reported from R. Benoue a Lakdo, Cameroun); Roman 1 966:
1 54 (Volta Noire pres de Boromo).
MATERIAL EXAMINED.— CAMEROUN: Niger basin: MNHN
1962-1275,53:1 4.6-1 7.6 mm, R. Benoue a Lakdo (paratypes);
NIGERIA: Niger basin: CAS 64128, 33: 16.9-27.0 mm, R. Be-
nue 22 km E of Makurdi; GHANA: Volta basin: MNHN 1984-
551, 102: 15.1-26.3 mm, R. Oti a Sabari; MNHN 1984-522,
2: 22.9-27.2 mm, R. Wawa a Yadzo. BURKINA FASO: Volta
basin: MNHN 1962-1280, 24.2 mm, Bougouri Ba au pont de
Nabere (holotype); MNHN 1961-591, 24: 14.6-16.9 mm, same
collection as holotype (paratypes).
Chiloglanis voltae is the smallest known species
of Chiloglanis, largest specimen 27.2 mm stan-
dard length. Sucker moderately large, non-pa-
pillose; outer mental barbel elongate, inner short;
no auxiliary mental barbels; teeth conical, not
strongly S-shaped; mandibular teeth in 1 or 2
irregular rows, not concentrated near symphysis
but spread across entire mouth opening; eye large;
dorsal fin soft rays usually 6; pectoral fin soft rays
166
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 6
FIGURE 7. Chiloglanis reticulatus, holotype, gravid female 41.6 mm (Congo basin).
7 or 8; caudal peduncle moderately deep; caudal
fin moderately forked; caudal fin lobes rounded,
lower slightly longer than upper. Body with three
dark or dusky vertical bars separated by pale
areas; margin of dark bars defined by concentra-
tion of melanophores not observed in other
species of Chiloglanis; caudal fin without well
developed bars or other marks. Sexual dimor-
phism unknown. No observations available on
gonads of mature individuals. Total vertebrae
29-3 1 , fewer than in any other species of Chi-
loglanis for which data are available.
DISTRIBUTION.— Known only from the R. Be-
nue (Niger basin) of Cameroun and Nigeria and
the Volta basin. In the R. Benue near Makurdi,
I collected C. voltae on sandy bottom, far from
ROBERTS: CAMEROUNIAN SUCKERMOUTH CATFISHES
167
FIGURE 8. Chiloglanis polypogon, holotype, immature female 41.5 mm (Cross basin).
any other kind of habitat, in sympatry with C.
benuensis.
Chiloglanis benuensis Daget and Stauch, 1963
Chiloglanis benuensis Daget and Stauch, 1963:98, fig. 3 (type
locality R. Benoue £ Lakdo, Cameroun).
MATERIAL EXAMINED.— CAMEROUN: Niger basin: MNHN
1962-1273, 38.4 mm, R. Benoue a Lakdo (holotype); MNHN
1962-1274, 40: 18.9-34.8 mm, collected with holotype (para-
types); NIGERIA: Niger basin: CAS 64127, 20: 20.3-35.0 mm,
R. Benue 22 km E of Makurdi.
Chiloglanis benuensis (largest specimen 38.4
mm) has relatively small, transversely oval suck-
168
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 6
er and very large barbels; up to three auxiliary
mental barbels on each side of lower lip, but no
auxiliary barbel lying external to outer primary
mental barbel; eye large; mandibular teeth 3 + 3
to 5 + 5; soft dorsal fin rays 5 or 6; pectoral fin
soft rays 8; adipose fin long, its origin far anterior
to a vertical line through anal fin origin; caudal
peduncle moderately slender; caudal fin mod-
erately forked, lower lobe somewhat longer than
upper; sexual dimorphism unknown.
DISTRIBUTION. — Chiloglanis benuensis is
known only from the R. Benue in Cameroun and
Nigeria. In the Benue near Makurdi it was col-
lected sympatrically with C. voltae on a sandy
bottom in shallow, swift-flowing water.
Chiloglanis polypogon new species
(Figs. \b, 2b, 3c, d, 6a, b, 8)
HOLOTYPE.— CAS 60790, 41.5 mm immature female, Cam-
eroun, Cross basin, high gradient streams along road from
Mamfe to Bamenda, 88-94 km SW of Bamenda, Tyson R.
Roberts, 17 Mar. 1980.
PARATYPES.-CAS 60791, 1 1: 20.2-39.7 mm, collected with
holotype; CAS 60792, 11: 23.0-26.0 mm, Cameroun, Cross
basin, riffles of R. Badi where it flows into R. Cross and main-
stream of R. Cross near Mamfe, Tyson R. Roberts, 1-5 Mar.
1980; CAS 60793, 7: 46.4-54.9 mm, Cameroun, Cross basin,
small stream on road from Mamfe to Bamenda, 98 km from
Bamenda and about 24 km from Mamfe, Tyson R. Roberts,
17 Mar. 1980; MRAC 76-32-P-2213-218, 3: 35.4-38.9 mm,
Cross basin, 1 4 miles [23 km] W of Mamfe towards Eyomojok,
H. Stenholt Clausen and Jorgen J. Scheel, 8 Mar. 1966.
Chiloglanis polypogon is a moderately large
species, largest specimen 54.9 mm; sucker rela-
tively small, transversely oval, with large pa-
pillae and very large barbels; 3-5 auxiliary men-
tal barbels present on each side of lower lip;
majority of specimens with an auxiliary mental
barbel external to outer primary mental barbel
on one or both sides of lower lip (not observed
in any other species of Chiloglanis); large pre-
maxillary teeth in 3 or 4 irregular rows; man-
dibular teeth large, 3 + 3 or 4+4; dorsal fin soft
rays usually 5 (5 in 24 specimens, 6 in 5); pectoral
fin soft rays 7 or 8; adipose fin long, its origin
well in advance of a vertical through anal fin
origin; caudal peduncle moderately slender; cau-
dal fin moderately forked, lower lobe longer than
upper; larger specimens with longitudinally ori-
ented tuberculated epidermal ridges on dorsal
surface of head; sexual dimorphism absent or
slight, males without enlarged humeral process.
In life pale areas between dark vertical bars and
entire ventral surface pinkish or pale salmon,
unlike any other species observed by me. Bars
on body intermediate in distinctness between
those of C. disneyi (very distinct) and C. batesii
(poorly defined), species with which it occurs
sympatrically in Cross basin. Body with pale
round spots observed in C. reticulatus and some
other species but not in C. batesii and C. ca-
meronensis.
Pupil of eye with flattened upper margin, con-
tractile; contraction involves downward pro-
gression of a silvery membrane or "curtain" from
ventral margin of exposed portion of eyeball.
During maximum contraction, pupil is dorsally
directed, almost out of horizontal line of vision.
A contractile pupil has not been observed in oth-
er species of Chiloglanis.
ETYMOLOGY.— The name polypogon (noun in
apposition, from the Greek "poly," multiple, and
"pogon," beard) refers to the numerous mental
barbels.
DISTRIBUTION.— Known only from the Cam-
erounian portion of the Cross basin.
Chiloglanis cameronensis Boulenger, 1 904
Chiloglanis cameronensis Boulenger 1904:18 (type locality
Efulen, S. Cameroon); Boulenger 1911:482, fig. 360; Matthes
1964:118 (R. Kululu, tributary of R. Tshuapa near Ikela,
Cuvette Centrale of Congo basin, Zaire; identification doubt-
ful, specimen unavailable for examination); Gery 1965:379
(swamps of R. Ivindo near Makokou, Ogooue basin, Gabon;
Roman 1971:134, fig. 58 (Ntem basin, Rio Muni).
MATERIAL EXAMINED. -CAMEROUN: BMNH 1904.7.1:90-96,
7: 32.5-41.2 mm, Efulen, Bates collection (syntypes); BMNH
1 906.5.28:68-69, 45.3 mm, Efulen (Bates collection); CAS(SU)
15448, 42.0 mm, Efulen, R. Chenge, Kribi basin; CAS(SU)
47468, 25.5 mm, R. Mfiande or R. Seng, Ebolowa, Ntem or
Campo basin; CAS(SU) 1 5450, 36.3 mm, R. Mfiande, Ebo-
lowa, Ntem basin; CAS(SU) 15451, 4: 35.7^0.2 mm, "Ebo-
lowa (Mejap II 12 mi W) Minya'a R. trib. of Mvila R."; CAS(SU)
47470, 2: 25.6-34.2 mm, Bikui, Lolodorf, Lokunje basin;
MHNG 757.69, 33.3 mm, Efulen; MNHN 1988-1 163, 3: 34.3-
42.9 mm, R. Tchengue on road from Kribi to Ebolowa, about
80 km W of Ebolowa, Kienke basin. Rio MUNI: MRAC 173147-
148, 28.7 mm, R. Guoro, tributary of R. Ntem. GABON: MRAC
73-02-P-191 1-921, 5: 23.5-33.7 mm, Ogooue basin, Makou-
kou; MHNG 2169.56-58, 4: 18.9-26.5 mm, Ogooue basin,
marigots apres Ybiegn, Makoukou; MRAC 80-27-P-89-95, 4:
19.6-28.2 mm, R. Diala.
Chiloglanis cameronensis is a medium-sized
species, largest specimen 45.3 mm; sucker very
large, longer than broad (not transversely oval),
extensively covered with moderately large pa-
pillae; barbels short, but well developed, usually
papillose; up to 3 auxiliary mental barbels on
each side of lower lip; large premaxillary teeth
ROBERTS: CAMEROUNIAN SUCKERMOUTH CATFISHES
169
in 3 or 4 rows; mandibular teeth large, usually
about 4+4 or 5 + 5, but some specimens with as
many as 7 + 7 or 8 + 8; eyes large; dorsal fin soft
rays usually 5; pectoral fin soft rays 7 or 8; adi-
pose fin long, its origin considerably in advance
of a vertical through anal fin origin; caudal pe-
duncle moderately to very slender (most slender
in specimens from Ogooue); caudal fin not deeply
forked, lobes rounded or even truncate, lower
slightly longer than upper. Dark bands and pale
interspaces on body fairly distinct. Abdomen
more or less darkly pigmented, almost black in
some specimens, due to melanophores contin-
uous with those on anterior vertical bar of body
extending across abdomen. This abdominal pig-
mentation, absent or poorly developed in other
species, highly variable: in some older and poorly
preserved specimens, scarcely evident; in some
specimens from Cameroun (including Efulen),
on only a relatively small area between the pelvic
fin bases; in specimens from Ntem and Ogooue,
entire abdomen darkened.
DISTRIBUTION.— Coastal basins of SE Camer-
oun and Rio Muni, including Lokunje, Kribi,
Kienke, and Ntem basins; Ogooue basin (Ga-
bon). Only species of Chiloglanis known from
Ogooue.
Chiloglanis sanagaensis new species
(Figs. \c, 2c, 3e,f,6c, d, 9)
HOLOTYPE.— CAS 60794, 26.8 mm male, Cameroun, Sanaga
basin, gravel bars and riffles in lower 1-2 km of R. Nchit where
it flows into R. Mbam (site of summer palace of Sultan of
Foumban), Tyson R. Roberts, 23 Mar. 1980.
PARATYPES.— CAMEROUN: CAS 60795, 25: 18.6-28.6 mm,
collected with holotype; MNHN 1988-1164, 11: 31.4-39.9
mm, Sanaga basin, R. Makenene at bridge on road to Ndimeki
( 1 00 m from Bafia-Bafoussam route), about 1 1 0 km S of Ba-
foussam, Tyson R. Roberts, 1 Jan. 1988; MNHN 1988-1 165,
12: 21.1^12.8 mm, Sanaga basin, R. Mana 23 km N of Bafia
on route from Bafia to Bafoussam, Tyson R. Roberts, 1 Jan.
1988.
Chiloglanis sanagaensis is a medium-sized
species, largest specimen 42.8 mm; sucker mod-
erately large, round, with numerous papillae of
uniform size evenly distributed over almost its
entire surface; barbels short but well developed,
each side of lower lip with 3 auxiliary mental
barbels; large premaxillary teeth in 3 or 4 rows;
mandibular teeth usually 4+4 or 5 + 5, rarely
6+6; eyes moderately large; dorsal fin soft rays
5 or 6; pectoral fin soft rays 7 or 8; adipose fin
low and very long, its origin far in advance of a
vertical through anal fin origin; caudal peduncle
slender; caudal fin almost truncate, very slightly
emarginate or forked, lobes slightly rounded,
lower a bit longer than upper. Secondary sexual
dimorphism not observed.
This species characterized by short, stocky or
stubby body and swollen abdomen. Swelling ev-
ident in virtually all specimens, seemingly due
mainly to exceptionally long and highly coiled
intestine. A specimen of 27.0 mm has intestine
just over 30 mm long, more than double relative
length of intestine observed in C. batesii and sev-
eral other species.
ETYMOLOGY.— The name sanagaensis, adjec-
tive, refers to the Sanaga basin, source of all
known specimens of this species.
DISTRIBUTION.— Known only from the Sanaga
basin.
Chiloglanis batesii Boulenger, 1 904
(Figs. Id, Id, 4a, b, 6e, 10)
Chiloglanis batesii Boulenger, 1 904: 1 9 (type locality "Efulen"
and "streams tributary of the Lobi River, 1 5 or 20 miles
S.W. of Efulen" [South Cameroon]); Boulenger 1911:485,
fig. 363 (Efulen, R. Lobi, R. Ja); Nichols and Griscom 1917:
720 (lower Congo; Faradje; no description); Trewavas 1974
(morphology, comparison with C. micropogori).
Chiloglanis micropogon Poll, 1952:228, figs. 3, 4 (type locality
R. Nzokwe, affluent de la R. Ulindi, Territ. Kabare; and R.
Tshinganda, Bunyakiri, affluent de la R. Luhoho, West Kivu);
Daget 1954:307, fig. 116 (upper Niger); Poll 1967:255, fig.
119 (R. Luachimo, Angola); Trewavas 1974:361, figs. 5, 6
(Mungo, Meme, and Lobi basins, Cameroun; R. Kaduna,
Niger basin, Nigeria).
MATERIAL EXAMINED.— CAMEROUN: BMNH 1904.7.127,28.2
mm, "15 or 20 miles S.W. of Efulen in streams trib. to R.
Lobi" (syntype, a male with elongate genital papilla, elongate
upper caudal fin lobe, and enormously expanded humeral pro-
cess); BMNH 1904.7.1.97-100, 4: 27.8-33.2 mm, Efulen, Bates
collection (syntypes); BMNH 1904.10.26:53-62, 1 1: 27.5-34.0
mm, R. Ja; BMNH 1909.4.29.97, 26.1 mm, R. Lobi; CAS
60796, 2: 36.7-39.3 mm, high-gradient streams along Mamfe-
Bamenda road, 16-22 km N of Numba and 88-94 km SW of
Bamenda, Cross basin; CAS 60797, 22.2 mm, riffles in mouth
of R. Badi where it flows into R. Cross at Mamfe; CAS 60798,
6: 19.2-26.7 mm, R. Nchit where it flows into R. Mbam,
Sanaga basin; BMNH 1973.5.14:362-363, 2: 19.8-20.0 mm,
R. Mungo at Baduma; BMNH 1973.5.14:347-354, 3: 13.7-
24.5 mm, R. Blackwater, Mungo basin; BMNH 1973.5.14:
355-359, 2: 22.0-26.3 mm, R. Dilolo at Bolo, Mungo basin;
BMNH 1973.5.14:364-370,2: 31.9-35.8 mm, R. Wowe, Mungo
basin; CAS(SU) 15447, 9: 29.0-33.3 mm, Efulen, R. Chenge,
Kribi basin; MNHN 1988-1155, 3: 26.2-28.7 mm, Kienke
basin, R. Tchengue on road from Kribi to Ebolowa. GUINEA:
Niger basin: MNHN 1986-750, 85: 18.5-34.6 mm, R. Milo.
MALI: Niger basin: MNHN 1961-590, 14: 30. 1-35.2 mm, Oey-
anko, near Bamako. NIGERIA: Niger basin: MNHN 1988-11 54,
4: 21.7-37.8 mm, R. Owena 25 km NE of Ondo on road to
170
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 6
FIGURE 9. Chiloglanis sanagaensis, holotype, male 26.8 mm (Sanaga basin).
Akure; MNHN 1988-1153, 8: 16.3-20.0 mm, R. Osse near
Uyere, on road from Benin City to Akure; BMNH 1 978. 1 2. 1 3.7-
8, 2: 26.5-27.3 mm, Gurara Falls near Abuja; MNHN 1977-
14, 3: 24.1-37.6 mm, R. Lere, Gindiri, Jos Plateau; MNHN
1977- 1 5, 3: 27. 1-40.0 mm, R. Shemankar, tributary of R. Gu,
Jos Plateau; Victor. G. Smith collection, uncat., 34.9 mm, R.
Assob, Gengera Kibo, Jos Plateau. REPUBLIQUE CENTRAFRI-
CAINE: Chad basin: MNHN 1988-1 156, 8: 21.5-27.5 mm, R.
Doukouma about 65 km W of Mbres; MNHN 1988-1157,
131: 19.3-31.8 mm, R. Ouaham 20 km upriver from Batan-
gafo; MNHN 1988-1158,2:31 .3-36. 1 mm, R. Koumbala be-
low Chutes de Matakil; MNHN 1988-1 159, 5: 22.2-25.7 mm,
R. Gounda (near Chad border). Congo basin: MNHN 1988-
1 1 6 1 , 20: 1 8.2-3 1 .6 mm, R. Mbomou near Zemio; MCZ 48360,
ROBERTS: CAMEROUNIAN SUCKERMOUTH CATFISHES
171
FIGURE 10. Chiloglanis batesii, 36.7 mm (Sanaga basin).
R. Mbomou at rapides de Gozobangui; MNHN 1988-1160,
41: 17.6-29.6 mm, R. Chinko at Rafai; MNHN 1988-1162,
4: 31.3-36.0 mm, small tributary of R. Euwou about 30 km
NW of Alindao. ZAIRE: Congo basin: CAS 60799, 26.4 mm,
R. Yambi below Chute de Bidi, 2 1 km E of Sundi-Mamba;
CAS 60800, 60: 16.5-27.2 mm, R. Ngudi at Chute Mayeux,
23 km NE of Sundi-Lutete; CAS 60801, 139: 13.7-30.4 mm,
R. Lufu, Songoloko district; CAS 60802, 2: 22.7-25.7 mm, R.
Lubazi, tributary to R. Luala near Kibunzi; CAS 60803, 35:
1 9.2-29.3 mm, tributary to R. Luala near Kibunzi; CAS 60804,
25.8 mm, R. Kwilu near Moerbeke; 60805, 29.6 mm, R. Sen-
gezi, tributary to R. Inkisi-Lukusu 87 km W of Kinshasa; CAS
60807, 2: 23.9-26.4 mm, R. Konzi, tributary to R. Kwango,
at bridge om road from Kenge to Masi-Manimba; FMNH
62918, 22.2 mm, R. Dungu, Nagero; MRAC 91480-81, 2:
33.1-37.2 mm, R. Tshinganda, tributary of R. Luhoho, Bun-
172
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 6
yakiri, Kivu (paratypes of C. micropogori); MNHN 1987-1607,
32.0 mm, R. Loya near Mafifi.
Chiloglanis batesii is a small species, largest
known specimen 40 mm standard length; sucker
relatively large, round or longer than broad rath-
er than transversely oval; mental barbels greatly
reduced; mandibular teeth 6 + 6 or fewer (often
5 + 5, occasionally 4 + 4); eyes large; dorsal fin soft
rays usually 6 (sometimes 5); pectoral soft rays
7 or 8; adipose fin high and posteriorly rounded,
but relatively short, its origin only slightly in
advance of a vertical through anal fin origin; cau-
dal fin deeply forked; caudal peduncle very slen-
der; penultimate 5-7 vertebrae, in juveniles as
well as adults of both sexes, with distal ends of
neural and hemal spines flared (anteroposteriorly
expanded). Sexual dimorphism exceptionally
pronounced, subadult as well as adult males usu-
ally with enlarged anal fin, enlarged and pointed
or filamentous upper caudal fin lobe, and greatly
enlarged humeral spine. Most Chiloglanis with
6+6 or fewer mandibular teeth have relatively
small suckers and long barbels. No other species
has a more slender caudal peduncle, or males
with such an elongate upper caudal fin and en-
larged humeral spine at such small size. Neural
and hemal spines in caudal peduncle slender in
all or almost all other species of Chiloglanis ex-
amined except some larger specimens of C. oc-
cidentalis.
Coloration in live specimens is recorded in my
fieldnotes for two Camerounian localities, high-
gradient tributaries of the R. Cross and the R.
Nchit, Sanaga basin. Two large Cross specimens
had body with dusky, poorly defined bands, al-
most mottled. Pale areas on body (more distinct-
ly outlined than dusky bands) pale yellowish or
dull olivaceous yellow. Caudal fin with dusky
vertical bar across base of lobes, light areas pale
yellow. Dusky markings on anal fin present but
less distinct than in C. disneyi collected at same
locality. Eye with small oval black pupil, iris dull
gray-green. Six specimens from the Nchit (Sa-
naga) had pale areas on body grayish or oliva-
ceous or pale yellow; alternating dark and pale
areas on snout forming in all six specimens a
triradiate pattern not evident in specimens of C.
sanagaensis collected at same locality.
DISTRIBUTION. — Chiloglanis batesii probably
is the species of Chiloglanis with the most ex-
tensive distribution. It is found throughout the
Niger and Congo basins, in the headwaters of the
Chad basin in Republique Centrafricaine, and in
almost every river basin in Cameroun except the
Ndian, Nyong, and Ogooue. It occurs sympatri-
cally with six of the nine other species of Chi-
loglanis found in Cameroun.
Chiloglanis disneyi Trewavas, 1974
(Figs. \e, 2e,4c,d, 11)
Chiloglanis disneyi Trewavas, 1974:365, figs. 6, 8 (type locality
Wowe, Mungo basin, Cameroun).
MATERIAL EXAMINED.— CAMEROUN: BMNH 1973:5.14.342,
34.6 mm, Mungo basin, R. Wowe, Disney collection (holo-
type); BMNH 1973.5.14:343-344, 2: 23.5-23.8 mm, R.
Dilolo, Mungo basin; BMNH 1973.5.1:345, 22.5 mm, R.
Blackwater, Mungo basin; CAS 60808, 20: 22.4-31.4 mm,
high-gradient streams of R. Cross along road from Mamfe to
fBamenda.
Chiloglanis disneyi apparently is a small
species, largest specimen only 35 mm (Trewavas
1974:362); sucker moderately large, round; pa-
pillae of upper lip with distinctive frond-like
margins (Trewavas 1974:365-366, fig. 8); bar-
bels very small; no auxiliary mental barbels; large
premaxillary teeth in 3 or 4 rows; mandibular
teeth usually in 2 rows, each with 8 + 8 to 10+10
teeth; eye small; dorsal fin soft rays usually 5
(specimens from Cross basin have 5 in 18, 6 in
2); pectoral fin soft rays 8 or 9; caudal peduncle
deep; caudal fin slightly forked, lobes rounded,
lower distinctly larger than upper. Secondary
sexual dimorphism apparently absent, but larg-
est specimens available lack well-developed go-
nads.
Live specimens from Cross basin have body
with three well-defined, almost black, vertical
bands; dorsal surface of head to dorsal fin origin
or first dark band on body yellowish orange, or-
angish brown, or brownish in different individ-
uals; pale bands on body, dorsal, adipose, and
caudal fins bright yellow or lemon yellow in all
individuals; pectoral, pelvic, and anal fins pale
yellow; abdomen off-white or cream; pectoral fin
with faint transverse dusky band near its middle;
adipose and anal fin base with black spots an-
teriorly; caudal fin with four large round yellow
spots surrounded by sharply defined black marks;
eye black. Alternating dark and pale areas on
head and body more clearly demarcated in C.
disneyi than in any other species of Chiloglanis
observed by me in Cameroun.
DISTRIBUTION.— Known only from Mungo and
Cross basins in Cameroun.
ROBERTS: CAMEROUNIAN SUCKERMOUTH CATFISHES
173
FIGURE 1 1 . Chiloglanis disneyi, 3 1 .4 mm (Cross basin).
COMMENT.— I have compared the specimens
collected in the Cross with two of Trewavas's
specimens collected in the Mungo, and find they
agree in most respects. The sucker is relatively
larger in the Cross specimens; the smaller suckers
of the Mungo specimens may be due to shrinkage
in preservative. The color pattern is barely dis-
cernible in the Mungo specimens, which are in
poor condition.
Chiloglanis niger new species
(Figs. lf,2f,4e,f,6f, 12)
HOLOTYPE.— CAS 60809, 34.4 mm, Cameroun, Niger basin,
R. Menchum below high waterfall near Befang, Bamenda high-
lands, Tyson R. Roberts, 10 Mar. 1980.
174
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 6
FIGURE 12. Chiloglanis niger, holotype, 34.4 mm (Niger basin).
PARATYPES.-CAS 60810, 7: 21.7-35.3 mm, collected with
holotype.
Chiloglanis niger (largest known specimen 35.3
mm) has a huge sucker, rounded and somewhat
longer than broad; barbels very small; no aux-
iliary mental barbels; large premaxillary teeth in
7 or 8 rows; mandibular teeth usually in 2 rows;
basic mandibular tooth count 8 + 8 to 10+10 in
largest specimens (6 + 6 and 7 + 7 in smallest); eye
very small; dorsal fin soft rays 6; pectoral fin soft
rays 9 or 10; adipose fin short, its origin only
slightly in advance of a vertical through anal fin
origin; caudal peduncle slender; caudal fin mod-
erately forked, lobes rounded, equal. No sexual
ROBERTS: CAMEROUNIAN SUCKERMOUTH CATFISHES
175
FIGURE 13. Chiloglanis harbinger, holotype, immature female 43.5 mm (Lokunje basin).
dimorphism observed (specimens probably im-
mature).
Color of live specimens (collected on a black
coarse sand or gravel bottom) uniformly dark,
almost black, on entire dorsolateral surface of
head and body; abdomen pale. Alternating dark
bars and pale intervening areas on body, so typ-
ical of species of Chiloglanis, but not noted in
live fish, faintly discernible preserved specimens.
ETYMOLOGY.— The name niger (Latin, black,
noun in apposition) refers to the Niger basin as
well as to the color in life.
DISTRIBUTION. — Known only from type local-
ity in R. Menchum, Niger basin, Bamenda high-
lands of Cameroun.
Chiloglanis harbinger new species
(Fig. 13)
HOLOTYPE.— CAS 6081 1, 43.5 mm immature female, Cam-
eroun, R. Lokunje near Bipindi, A. I. Good, 17 Feb. 1936.
PARATYPES.-CAS(SU) 47469, 9: 34.5-44.0 mm, collected
with holotype.
Chiloglanis harbinger (largest specimen 44.0
mm) has a huge sucker, longer than broad; mi-
nute barbels; no auxiliary mental barbels; teeth
very numerous; large premaxillary teeth in up to
7 rows; mandibular teeth usually in 2 rows, basic
mandibular tooth count about 13+13 to 15+15
(teeth of type specimens damaged by string passed
through mouth and gills); eye very small; dorsal
fin soft rays usually 5 (4 in two specimens); pec-
176
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 6
toral fin soft rays 9 or 10; adipose fin long and
low, its origin well in advance of a vertical through
anal fin origin; caudal peduncle moderately deep;
caudal fin moderately forked, lobes rounded,
nearly equal. Males with a moderately elongate
genital papilla; no sexual dimorphism observed
in size of anal fin, shape of caudal fin, or humeral
process of pectoral girdle.
Color in life unknown. Preserved specimens
with numerous large melanophores more or less
uniformly distributed over entire dorsolateral
surface of body; vertical dark bars and interven-
ing pale areas on body (present in most species
of Chiloglanis) not evident; melanophores ex-
tending from sides of body onto sides of abdo-
men but rapidly becoming sparser toward middle
of abdomen (hence very different from uniformly
distributed melanophores on abdomen in some
specimens of C. cameronensis); a dark vertical
bar on base of caudal fin as well as one in its
middle (very dark area near tip of caudal fin in
lateral photograph of holotype an artifact due to
shadow from tip of caudal fin, which is broken
and bent toward camera).
ETYMOLOGY.— The name harbinger (Middle
English, noun in apposition) is an allusion to the
large number of teeth in the suckermouth mo-
chokid genera Atopochilus and Euchilichthys, a
condition presaged or foreshadowed in this
species of Chiloglanis.
DISTRIBUTION.— Known only from the Lo-
kunje basin in southeastern Cameroun.
ACKNOWLEDGMENTS
My interest in Chiloglanis was kindled when
I visited the late botanist and ichthyologist F. R.
Irvine at his home in Achimota shortly after my
arrival in Ghana in 1961 and was given some
specimens collected by Ing Zwilling in Nigeria.
Permission for fieldwork in Cameroun was
granted by the Ministry for Scientific and Tech-
nological Research of the Republic of Cameroun.
Fieldwork in 1980 was supported by the Fonds
Leopold III pour la Conservation et la Explo-
ration de la Nature and in 1988 by the National
Geographic Society. Fieldwork in Zaire in 1973
was sponsored by the Office National de la Re-
cherche et du Developpement of the Republic of
Zaire and supported by the National Geographic
Society. Most of the photographs are by Al Cole-
man of the Museum of Comparative Zoology,
Harvard. Scanning electron micrographs are by
Mary Ann Tenorio of the SEM lab, California
Academy of Sciences. Radiographs were pre-
pared by David Catania in the Department of
Ichthyology, California Academy of Sciences, and
Jeanine Abel, Laboratoire d'Ichtyologie, Mu-
seum National d'Histoire Naturelle (Paris). Prof.
Michal Jakubowski of the Department of Com-
parative Anatomy of Jagiellonian University
(Cracov, Poland) kindly provided information
concerning taste buds. For arranging loans of
specimens I am obliged to Susan Jewett, Karsten
Hartel, Douglas Nelson, Alwynne Wheeler,
Volker Mahnert, Dirk Thys van den Auden-
aerde, and Victor G. F. Smith. The paper was
written in the Department of Ichthyology of the
California Academy of Sciences and in the La-
boratoire d'Ichtyologie of the Museum National
d'Histoire Naturelle.
LITERATURE CITED
BOULENGER, G. A. 1900. On some little known African si-
luroid fishes of the subfamily Doradinae. Ann. Mag. Nat.
Hist., ser. 7, 6:520-529.
. 1902. On the fishes collected by Mr. S. L. Hinde in
the Kenya district, with descriptions of four new species.
Proc. Zool. Soc. London 2:221-224.
-. 1904. Descriptions of new West-African fishes. Ann.
Mag. Nat. Hist., ser. 7, 14:16-20.
. 1911. Catalogue of the freshwater fishes of Africa. 2.
Brit. Mus. Nat. Hist.
. 1920. Poissons recueillis au Congo Beige par 1'ex-
pedition du Dr. C. Christy. Ann. Mus. Congo Beige, ser. 1 ,
2:1-38, pis. 23-25.
DAGET, J. 1948. La collection des poissons d'eau douce de
1'IFAN. Cat. Inst. Fr. Afr. Noire (Dakar). 59 pp.
. 1954. Les poissons du Niger superieur. Mem. Inst.
Fr. Afr. Noire 36. 391 pp.
DAGET, J., J.-P. GOSSE, AND D. F. E. THYS VAN DEN AUDEN-
AERDE, EDS. 1986. Check-list of the freshwater fishes of
Africa. 2. ISNB (Brussels), MRAC (Tervuren), ORSTOM
(Paris).
DAGET, J. AND A. ILTIS. 1965. Poissons de Cote d'lvoire (eaux
douces et saumatres). Mem. Inst. Fr. Afr. Noire 74. 385 pp.
DAGET, J. AND A. STAUCH. 1963. Poissons de la partie ca-
merounaise de bassin de la Benoue. Mem. Inst. Fr. Afr.
Noire. 68:85-107.
GERY, J. 1965. Poissons du bassin de 1'Ivindo. Biol. Gab. 1:
375-393.
JAKUBOWSKI, M. 1983. New details of the ultrastructure (TEM,
SEM) of taste buds in fishes. Zeitschr. mikrosk.-anat. Forsch.
(Leipzig) 97:849-862.
LADIGES, W. AND J. VOELKER. 1961. Untersuchungen iiber
die Fische in gebirgsgewassern des Wasserscheidenhoch-
lands in Angola. Mitt. Hamb. Zool. Mus. Inst. 59:1 17-140.
LEVITON, A. E., R. H. GIBBS, Jr., E. HEAL, AND C. E. DAWSON.
1985. Standards in herpetology and ichthyology. Part I.
Standard symbolic codes for institutional resource collec-
tions in herpetology and ichthyology. Copeia 1985(3):802-
832.
ROBERTS: CAMEROUNIAN SUCKERMOUTH CATFISHES
177
MATTHES, H. 1 964. Les poissons du lac Tumba et de la region
d'Ikela. Ann. Mus. Roy. Afr. Centrale (Tervuren), ser. oct.,
sci. zool. 126. 206 pp., 6 pis.
NICHOLS, J. T. AND L. GRISCOM. 1917. Freshwater fishes of
the Congo basin. Bull. Amer. Mus. Nat. Hist. 37:653-756.
NORMAN, J. R. 1932. A collection of fishes from Sierra Leone.
Ann. Mag. Nat. Hist., ser. 10, 10:180-185.
PELLEGRIN, J. 1 933. Voyage du Ch. Alluaud et P. A. Chappuis
en Afrique occidental francaise (dec. 1930-mars 1930). 4.
Poissons. Arch. Hydrobiol. (Stuttgart) 26:101-120.
PETERS, W. C. T. 1868. Ueber eine von dem Baron Carl von
der Decken entdeckte neue Gattung von Welsen, Chiloglanis
Deckenii, und einige andere Susswasserfische aus Ostafrika.
Monatsber. Akad. Wiss. Berlin 1868:598-602, pi. 2.
POLL, M. 1952. Poissons des rivieres de la region des lacs
Tanganyika et Kivu recueillis par G. Marlier. Rev. Zool.
Bot. Afr. 46:221-236.
. 1967. Contribution a la faune ichthyologique de
1'Angola. Diamang Publ. Cult. (Lisbon) 75. 381 pp.
. 1971. Revision des Synodontis africains (famille
Mochocidae). Ann. Mus. Roy. Afr. Centr., ser. oct., sci. zool.
191.497pp., 13 pis.
POLL, M. AND D. J. STEWART. 1975. Un Mochocidae et un
Kneriidae nouveaux de la riviere Luango (Zambia), affluent
du bassin du Congo (Pisces). Rev. Zool. Bot. Afr. 89:151-
158.
ROBERTS, T. R. 1982. Unculi (horny projections arising from
single cells), an adaptive feature of the epidermis of ostar-
iophysan fishes. Zool. Scripta 11:55-76.
. 1983. Revision of the south and southeast Asian
sisorid catfish genus Bagarius, with description of a new
species from the Mekong. Copeia 1983(2):435-445.
. In press. The freshwater fishes of western Borneo (Ka-
limantan Barat, Indonesia). Mem. Calif. Acad. Sci. 14.
ROMAN, B. 1966. Les poissons des hauts-bassins de la Volta.
Ann. Mus. Roy. Afr. centr., ser. oct., sci. zool. 150. 191 pp.
. 1971. Peces de Rio Muni Guinea Ecuatorial (Aguas
dulces y salobres). Fund. La Salle Cien. Nat. (Barcelona).
295 pp.
SAUVAGE, H. E. 1878. Notice sur la faune ichthyologique de
1'Ogooue. Bull. Soc. Philomath. Paris, ser. 3, 7:90-103.
SMITH, V. G. F. 1983. Fish from fast-flowing rivers and streams
on the Jos Plateau. Nigerian Field 47:201-21 1.
TREWAVAS, E. T. 1974. The freshwater fishes of Rivers Mun-
go and Meme and Lakes Kotto, Mboandong and Soden,
West Cameroon. Bull. Brit. Mus. Nat. Hist., Zool. 26:331-
412, 5 pis.
CALIFORNIA ACADEMY OF SCIENCES
Golden Gate Park
San Francisco, California 94118
PROCEEDINGS
OF THE
CALIFORNIA ACADEMY OF SCIENCES
Vol. 46, No. 7, pp. 179-192, 36 figs.
December 20, 1989
AMERICAN DANCE FLIES OF THE
DRAPETIS ASSIMILIS SPECIES GROUP
(DIPTERA: EMPIDIDAE)
By
Edward Rogers
Naval Medical Research Institute, Detachment Lima, Peru,
APO Miami, Florida 34031-0008
ABSTRACT: Eight American species are newly recognized as belonging to the Drapetis assimilis species group,
including five new species described herein: Drapetis arnaudi n. sp. from Florida; D. cerina n. sp. from New
Mexico; D. destituta n. sp. from Baja California; Z). Solaris n. sp. from Mexico (Distrito Federal); and D.
tonilosa n. sp. from Baja California. Lectotypes are designated for D. latipennis Melander, 1902 and D.
discalis Melander, 1918. Drapetis discalis Melander is removed from Crossopalpus Bigot and redescribed.
Drapetis infumata Melander is redescribed. A key to American species is provided.
Received November 10, 1988. Accepted August 24, 1989.
INTRODUCTION
Flies of the genus Drapetis are tiny cursorial
predators that are seldom observed or collected.
Members of the D. assimilis species group have
been found on tree trunks and in flowers, and
reared from litter in hollow trees; their other hab-
its are unknown.
The D. assimilis group was first alluded to by
Collin (1961) in discussing characters shared by
the Palearctic species, Drapetis assimilis Fallen,
D. simulans Collin, and D. arcuata Loew. Sub-
sequently, Kovalev (1972) named and further
characterized the group, basing it upon the same
three species and D. ingrica Kovalev. This clas-
sification was followed by Chvala (1975) in a
treatment of Scandinavian Tachydromiinae.
The last revision of the American Drapetis
(Melander 1918) recognized no infrasubgeneric
categories, although three members of the D. as-
similis group were among the 1 1 valid species
then known. At that time, distinctions among
Drapetis Meigen, Crossopalpus Bigot (=Eudra-
petis Melander), and Elaphropeza Macquart were
poorly understood, with the result that for the
last 70 years, the American D. assimilis group
species have been divided between Drapetis and
Crossopalpus. The present paper corrects this un-
natural arrangement by uniting the group and
placing it within Drapetis, s. str.
None of the D. assimilis group species have
Holarctic distributions, and despite study, none
are yet known from the Neotropical Region.
METHODS
Specimens from the following collections were
studied. Abbreviations given in parentheses are
used in the text to denote depositories: (AC) Col-
lection of the author; (AMNH) American Mu-
seum of Natural History; (CAS) California Acad-
emy of Sciences; (CIS) University of California
[179]
180
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 7
at Berkeley; (CNC) Canadian National Collec-
tion; (INHS) Illinois Natural History Survey;
(MCZC) Museum of Comparative Zoology;
(PHA) Paul H. Arnaud, Jr., Collection; (UCD)
University of California at Davis; (UCR) Uni-
versity of California at Riverside; (IUCM) Iowa
State University; (UK) University of Kansas;
(USNM) United States National Museum of
Natural History; (WSU) Washington State Uni-
versity. In listings of materials examined, all col-
lection data preceding a specimen apply to that
specimen, unless data that follow are in conflict.
Type specimens of all American species were
examined and compared with original descrip-
tions of types and with specimens and descrip-
tions of this study. Descriptions of new species
have been based solely on the holotype, with
variation among individuals discussed in suc-
ceeding paragraphs.
Male genitalia have been drawn in ventrolat-
eral and dorsolateral views, using a camera lu-
cida; wings have been traced from microscope
slide projections. Terminology of genitalia fol-
lows Griffiths (1972).
Drapetis Meigen
Drapetis Meigen, 1822:91. Type species: Drapetis exilis Mei-
gen, 1 822 (by monotypy).
A diagnosis of this genus is given elsewhere
(Rogers 1983). Drapetis is not to be confused
with its sister genera, Crossopalpus Bigot and
Elaphropeza Macquart. Drapetis discalis Melan-
der was originally described in Eudrapetis Me-
lander, a junior synonym of Crossopalpus. It, and
all species treated here, belong to Drapetis Mei-
gen.
Drapetis assimilis Group
DESCRIPTION.— Males: Head round in frontal
view; palpus light to dark brown; antenna uni-
formly brown, the second segment bearing a dis-
tal circlet of setae with one ventral seta longer
than all others in the circlet, arista long in relation
to the short third segment. Thorax scantily to-
mentose; when present, propleural tomentum not
reaching anterior spiracle; pteropleuron pol-
ished; scutellar tomentum invading posterior edge
of mesonotum between calli. Wing lacking mac-
ulations, veins R4+5 and M converging or sub-
parallel distally; second basal (M) cell long, dis-
tally extending to below the junction of vein R,
with the costa; rm crossvein attaching just prox-
imad of center of M cell; me crossvein perpen-
dicular to Cu or forming a slightly obtuse prox-
imal angle at junction with Cu. Legs long; anterior
femur tomentose ventrally; center femur lacking
corrugation patterns, ventral surface bearing one
or more longitudinal rows of long spines; pos-
terior femur lacking strong anteapical spines on
posterior surface; anterior and center tibiae to-
mentose along flexor surfaces; center tibia lack-
ing a strong apical tooth; posterior tibia lacking
true extensor bristles, apical lappet short, blunt,
unarmed. Abdomen black, tergites lacking in-
volutions on corners; sclerotization of tergite I
interrupted near dorsal midline, other tergites
entire; tergites IV and V laterally bearing flat-
tened spines; sternites I, II, III, VI, VII, and VIII
entire; sternite IV divided through center; scler-
otization of segment VIII narrow and of uniform
width dorsally; aedeagus short, concealed within
periandrium. Females: As in males, except all
abdominal sternites entire.
The D. assimilis group is distinguished from
all others by the unique combination of a long
projecting seta on the venter of the second an-
tennal segment, a long basal cell, and division of
the male abdominal sternite IV. The antennal
seta and basal cell have been used before to sep-
arate the D. assimilis group from other Drapetis
in keys (Kovalev 1972); the importance of the
divided sternite IV has been overlooked.
With the exception of D. latipennis Melander,
all of the American D. assimilis group species
have widely polished orbits, and the tomentum
of the prothorax fails to reach the antero ventral
lateral corner of the humerus. The center tibia
bears two parallel rows of short spines along its
flexor length in all American species except D.
torulosa n. sp. Sternite V is divided in a few
species, as noted in their respective descriptions.
None of the D. assimilis species have the long
exserted aedeagus, divided sternite HI, wing vein
anomalies, nor tibial deformations that appear
in various species allied to D. divergens Loew.
Species related to D. dividua Melander have ex-
tensor spines on the posterior tibiae and a pe-
dunculate structure articulating with the left peri-
andrium, both absent in the D. assimilis group.
Members of the D. naica group (Rogers 1983)
have an apical thorn on the flexor surface of the
center tibia, widely divergent radial and medial
veins, a short M cell, and oval head shape. Rel-
atives of D. exilis Meigen lack the long ventral
ROGERS: AMERICAN DRAPETIS
181
1
0.1
mm
FIGURES 1-5. Third antennal segments of paratype males. Figure 1. Drapetis arnaudi n. sp. Figure 2. Drapetis destitute, n.
sp. Figure 3. Drapetis cerina n. sp. Figure 4. Drapetis Solaris n. sp. Figure 5. Drapetis torulosa n. sp.
seta on the second antennal segment, present in
the D. assimilis and D. naica groups, and fre-
quently have corrugations on the center femur.
KEY TO SPECIES OF THE D. ASSIMILIS GROUP
This key is designed for provisional identifi-
cation of unmacerated males. Identifications
should be confirmed by examination of cleared
genitalia. Females are seldom identifiable.
(Males)
1. Anterodorsal surface of center femur bulg-
ing near proximal quarter, Figure 34; center
tibia lacking flexor spines; genitalia, Figures
35, 36 torulosa n. sp.
- Anterodorsal surface of center femur shaped
normally, Figure 31; center tibia bearing
flexor spines 2
2. Posterior tibia bearing a group of long,
downcurved setae about distal third of ex-
tensor surface, Figure 28; genitalia, Figures
29, 30 latipennis Melander
- Posterior tibia bearing short setae of nearly
uniform length along extensor surface 3
3. Setae of anterior surface of center femur
directed ventrally, Figure 3 1 ; genitalia, Fig-
ures 32, 33 Solaris n. sp.
- Setae of anterior surface of center femur all
directed distally 4
4. Halter yellow; one pair of vertical bris-
tles. 5
- Halter brown; two or three pairs of verti-
cals 6
5. Ventral spines of center femur arranged in
a single row; propleuron partially tomen-
tose; genitalia, Figures 21-23
discalis Melander
- Ventral spines of center femur arranged in
two or more rows; propleuron completely
polished; genitalia, Figures 15, 16
arnaudi n. sp.
6. Gena obscured by eye in profile; anterior
surface of center femur thickly whitish yel-
low tomentose; two pairs of vertical bris-
tles; genitalia, Figures 17, 18 cerina n. sp.
- Gena narrowly visible beneath eye in pro-
file; anterior surface of center femur pol-
ished; three pairs of verticals 7
7. Bearing a tubercle or a blunt raised projec-
tion near posteroventral proximal third of
posterior femur; propleuron usually par-
tially tomentose; abdominal tergites VI and
VII partially polished; genitalia, Figures 24-
27 infumata Melander
- Lacking a tubercle or prominence on pos-
teroventral surface of posterior femur; pro-
pleuron completely polished; abdominal
tergites VI and VII completely tomentose;
genitalia, Figures 1 9, 20 destituta n. sp.
Drapetis arnaudi n. sp.
(Figs. 1, 10, 15, 16)
DIAGNOSIS.— One pair of vertical bristles; pro-
pleuron completely polished; halter yellow; male
genitalia as in Figures 15, 16.
DESCRIPTION.— Male. Length 1.1 mm. Head
dark brown; one pair of verticals; frons width at
ventral end one-fifth length of third antennal seg-
ment; face linear, slightly narrowed toward cen-
ter; gena entirely visible in profile; third antennal
segment shaped as in Figure 1 , arista five times
its length. Thorax dark brown; propleuron com-
pletely polished. Wing 1.3 mm long, neuration
as in Figure 10. Legs and coxae bright yellow,
ultimate tarsal articles of each leg dark brown.
Abdomen laterally bearing tiny flattened spines
on segments IV and V.
TYPES. -Holotype: male (CAS 13448), USA:
182
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 7
14
FIGURES 6-14. Wing venation of males. Figure 6. Drapetis cerina n. sp., paratype. Figure 7. Drapetis discalis, Brewster Co.,
Texas. Figure 8. Drapetis torulosa n. sp., paratype. Figure 9. Drapetis latipennis, Orange Co., California. Figure 10. Drapetis
arnaudi n. sp., paratype. Figure 11. Drapetis Solaris n. sp., paratype. Figure 12. Drapetis discalis, Riverside Co., California.
Figure 13. Drapetis destituta n. sp., paratype. Figure 14. Drapetis infumata, Chelan Co., Washington.
ROGERS: AMERICAN DRAPETIS
183
18
FIGURES 15-18. Genitalia of male paratypes. Figures 15 (ventrolateral view) and 16 (dorsolateral view). Drapetis arnaudi
n. sp. Figures 1 7 (ventrolateral view) and 1 8 (dorsolateral view). Drapetis cerina n. sp.
184
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 7
labelled, "FLA: Highlands Co., Archbold Biol.
Sta., 7.X. 1964, P. H. Arnaud, Jr." Paratypes,
same locality: 2 males, 1 0.X. 1 964 (CAS and AC),
1 male, 7.X. 1964 (CAS), 1 male, 12.X.1964(AQ.
REMARKS.— The paratypes have three rows of
ventral spines on the center femur, merging dis-
tally into two rows, not visible on the holotype.
The posterior femur bears a single row of poorly
developed ventral setae.
This species is named for its collector, Dr. Paul
H. Arnaud, Jr.
Drapetis cerina n. sp.
(Figs. 3,6, 17, 18)
DIAGNOSIS.— Two pairs of vertical bristles; gena
obscured by eye in profile; halter brown; anterior
surface of center femur whitish yellow tomen-
tose; male genitalia as in Figures 17, 18.
DESCRIPTION. — Male. Length 1.8 mm. Head
black; two pairs of verticals, outer pair short;
frons width at ventral end one-seventh length of
third antennal segment; face linear; gena ob-
scured by eye in profile; third antennal segment
shaped as in Figure 3, arista six times its length.
Thorax black; one weakly developed humeral
bristle; propleural tomentum present, narrowly
failing to reach anteroventral lateral corner of
humerus; halter dark brown. Wing 2.0 mm long,
neuration as in Figure 6. Legs and coxae yellow,
with a brownish cast in coxae and tarsi; anterior
surface of center femur completely coated with
translucent, waxy, whitish yellow tomentum,
ventral surface bearing brown spines set in two
irregular rows, spines longest near centers of rows;
flexor spines of center tibia very poorly devel-
oped; posterior femur with one row of ventral
setae and four long anteroventral distal bristles
near knee. Abdomen laterally bearing erect flat-
tened spines on tergites IV and V.
TYPES. -Holotype: male (USNM 762713),
USA: labelled, "Las Vegas HS, 14.5, NM, H. S.
Barber Collector." Paratype: 1 male, USA: New
Mexico: Santa Fe Co., Santa Fe, VII (USNM).
REMARKS.— The anterior surface of the center
femur looks much as though it were smeared
with wax, producing greasy reflections. The ad-
jective cerina refers to this waxy quality.
Drapetis destituta n. sp.
(Figs. 2, 13, 19,20)
DIAGNOSIS.— Three pairs of strong vertical
bristles; propleuron completely polished; halter
dark brown; abdominal tergites VI and VII com-
pletely tomentose; male genitalia as in Figures
19,20.
DESCRIPTION.— Male. Length 14 mm. Head
black; three verticals; frons width at ventral end
one-sixth length of third antennal segment; face
linear; gena entirely visible in profile; third an-
tennal segment shaped as in Figure 2, arista four
times its length. Thorax dark brown; propleuron
completely polished; halter dark brown. Wing
1 .6 mm long, neuration as in Figure 1 3. Legs and
coxae yellow, except tarsi and posterior knees
brown; center femur bearing two rows of strong,
dark brown ventral spines, merging distally into
one row at about half length, continuing along
entire femur; posterior femur bearing two rows
of long yellow setae along entire length, one row
anteroventral, the other posteroventral. Abdo-
men laterally bearing long flattened spines on
tergites IV and V; tergites VI and VII completely
tomentose.
TYPES. -Holotype: male (CAS 13451), MEX-
ICO: labelled, "MEXICO: Baja California: Agua
Caliente (San Carlos), 18.5 km east of Manea-
dero, 6. VII. 1 973, Paul H. Arnaud, Jr." Paratype:
1 male, same data (PHA).
REMARKS.— Maceration of the paratype re-
veals that abdominal sternite V is divided. Rec-
ognition of this species relies primarily on
genitalic differences; it lacks other unique char-
acteristics.
Drapetis discalis (Melander)
(Figs. 7, 12, 21-23)
Eudrapetis discalis Melander, 1918:198.
Drapetis (Eudrapetis) discalis; Melander, 1928:310.
Drapetis (Crossopalpus) discalis; Melander, 1965:477.
DIAGNOSIS. —One pair of vertical bristles; male
genitalia as in Figures 21-23.
REDESCRIPTION. — The original description will
suffice to recognize D. discalis with the following
corrections based on the lectotype: body shining
brown; halter yellowish-clear; second, third, and
fourth sections of costa proportioned 2.4:2.8:8. 1 ;
rm crossvein at five-twelfths of the length of sec-
ond basal cell; outer two sections of vein M,+2
proportioned 1:2.
TYPES.— Melander labelled his Washington
specimen "type," and his California specimen
"paratype," but neglected to publish a designa-
tion of the holotype. Lectotype male (USNM),
here designated, USA: labelled, "Wawawai Wash
ROGERS: AMERICAN DRAPETIS
185
22. VI." Paralectotype male (USNM), USA: Cal-
ifornia: Santa Clara Co., Palo Alto.
OTHER MATERIAL EXAMINED.— USA: California: Contra
Costa Co., 1 male, Orinda Village, 18. VI. 1970, E. I. Schlinger
(AC); Monterey Co., 3 males, 25.IX.1934, A. L. Melander
(USNM), 1 male, 8 km SW Greenfield, 3.V.1975, P. A. Rude
(AC), 1 male, Soledad, 20. V.I 956, H. R. Moffitt (UCD); Or-
ange Co., 1 male, 14. VI. 1929, P. W. Oman (UK); Riverside
Co., 1 male, 8 km S of Sage, 16.IV. 1965, C. A. Toschi (CIS),
2 males, Whitewater Cyn., 1 1 .IV. 1 965, J. T. Doyen (CIS); San
Bernardino Co., 1 male, Loma Linda, 13.V.1953, R. E. Ryck-
man (USNM), 1 male, Upper Santa Ana River, 28. IX. 1953,
A. L. Melander (USNM), 1 male, Mountain Home Cyn.,
27.IX.1955, A. L. Melander (USNM), 1 male, Verdemont,
I.V.I 946, A. L. Melander (USNM); San Diego Co., 1 male,
4.8 km WSW of Escondido, 23.111.1975, P. A. Rude (AC);
Santa Barbara Co., 1 male, Santa Barbara, 16. VI. 1951, K. W.
Tucker (UCD); Santa Clara Co., 2 males, Los Gatos, off Stacia
Street, Quercus trunk, 30.V/23.VII.1977, E. J. Rogers (AC), 5
males, Stanford, 19.XI.1952, P. H. Arnaud (PHA); Shasta Co,
1 male, McArthur, 12.X.1952, E. I. Schlinger (UCD); Ventura
Co., 1 male, Saticoy, 5.V.1924, S. E. Banders (CAS). Idaho:
Nez Perce Co., 1 male, Lewiston Hill, 3 I.V.I 924, A. L. Me-
lander (USNM), 1 male, 30. VI. 1 975, E. J. Rogers (AC). Texas:
Brewster Co., 2 males, Panther Junction, Big Bend Nat'l. Park,
1,067 m, 14.V.1959, J. F. McAlpine (CNC); Gillespie Co., 1
male, Pedernales River, 4.IV.1955, W. W. Wirth (USNM);
San Patricio Co., 1 male, Padres Id. near Port Aransas,
23.111.1965, J. G. Chillcott (CNC). Utah: Cache Co., 1 male,
Blacksmith Fork Cyn., 1 1. VIII. 1975, G. F. Knowlton (UCD).
Washington: Asotin Co., 1 male, Fields' Spring State Park,
31. VII. 1971, W. J. Turner (WSU), 1 male, 27.4 km S of An-
atone, 594 m, fls. Lonicera, 15.IV. 1977, W. J. Turner (WSU);
Whitman Co., 3 males, Yakawawa Cyn., 1 1.3 km NW of Col-
ton, 762 m, 25/27. VI. 1977, W. J. Turner (WSU), 1 male, Big
Almota Cyn., Almota, 25.VII.1973, D. Corredor and S. Ber-
kenkamp (WSU). MEXICO: Baja California: 1 male, 9 km E
of Hamilton Ranch Arr. Santo Domingo, 23.IV. 1963, H. B.
Leech and P. H. Arnaud (CAS).
REMARKS.— The gena is usually narrowly vis-
ible in profile (not in specimens from Brewster
Co., Texas). The ventral half of the propleuron
is usually tomentose (completely polished in all
Texas specimens). Most individuals show a sin-
gle, poorly distinguishable humeral bristle. The
halter may be yellow or (rarely) brown. Ventral
spines of the center femur are arranged in a single
row, and the posterior femur bears a ventral row
of setae. Abdominal sternite V is divided.
Distinguishing between D. discalis and D. in-
fumata Melander is difficult. Cerci of the two
species vary widely in shape (Figs. 21, 22, 24,
25), and Pacific Coast D. discalis occasionally
have brown halters, or infumated wings, or have
the me cross vein located proximad of the junc-
tion of vein R, with the costa, all as in D. infu-
mata. Difference in the number of vertical bris-
tles appears to be the best character for separation.
mm
FIGURES 1 9, 20. Genitalia of paratype male Drapetis des-
tituta n. sp. Figure 19. Ventrolatral view. Figure 20. Dorso-
lateral view.
186
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 7
21
23
FIGURES 21-23. Genitalia of male Drapetis discalis. Figure 21 (ventrolateral view). Brewster Co., Texas. Figure 22 (ventro-
lateral view). Cerci, Monterey Co., California. Figure 23 (dorsolateral view). Santa Clara Co., California.
FIGURES 24-27. Genitalia of male Drapetis infumata. Figure 24 (ventrolateral view). Cerci, Chelan Co., Washington. Figure
25 (ventrolateral view). Monmouth Co., New Jersey. Figure 26 (dorsolateral view). Chelan Co., Washington. Figure 27 (dor-
solateral view). Monmouth Co., New Jersey.
ROGERS: AMERICAN DRAPETIS
187
25
27
0.1 mm
188
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 7
Eudrapetis Melander ( 1 9 1 8) is an obvious syn-
onym of Crossopalpus Bigot (1857). Crossopal-
pus have produced genae, a single pair of ocellar
bristles, and lack mesopleural setae. Like other
Drapetis, D. discalis has short genae, two pairs
of ocellars, and hairy mesopleurae.
Drapetis infumata Melander
(Figs. 14, 24-27)
Drapetis latipennis Melander, 1 902:209 (in part, male paratype
from Milwaukee, Wisconsin).
Drapetis infumata Melander, 1918:194.
Drapetis naica Melander, 1918:195 (in part, male paratype
from Avon, Idaho).
DIAGNOSIS.— Three pairs of strong vertical
bristles; a gentle prominence or pointed tubercle
near posteroventral proximal third of posterior
femur; male genitalia as in Figures 24-27.
REDESCRIPTION.— Males. Length 1.2 to 1.5 mm.
Head dark brown; three pairs of verticals; frons
width at ventral end one-third to one-fifth length
of third antennal segment. Thorax black to brown;
one weak humeral bristle; propleural tomentum
present, not attaining anteroventral lateral cor-
ner of humerus nor anterior spiracle; halter dark
brown. Wing 1.6 to 1.9 mm long, occasionally
infumated along veins, me crossvein located
proximad to junction of R, with costa, Figure
14. Legs and coxae dark yellow to dark brown,
anterior femur and coxa and proximal areas of
other femora lightest colored; center femur bear-
ing a single row of ventral spines; posterior femur
bearing one row of posteroventral setae along
distal half, one row of short to long setae along
entire ventral length, and a posteroventral tu-
bercle or node near proximal third. Abdomen
laterally bearing tiny flattened spines on tergites
IV and V, closely appressed to body; abdominal
sternite V usually divided; genitalia as in Figures
24-27.
TYPES— Holotype (D. infumata): female
(USNM), CANADA: labelled, "Nelson, British
Columbia." Paratype (D. infumata) female
(USNM), USA: Priest Lake, Bonner Co., Idaho
(may be one of two females from Priest Lake,
collected 1. VIII. 19 16, but lacking a type label).
Paratype (of D. latipennis) male (AMNH 729),
USA: Milwaukee, Wisconsin, 23.VI.1895, W. M.
Wheeler. Paratype (of D. naica) male, USA:
Avon, Idaho (USNM).
OTHER MATERIAL EXAMINED. — USA: California: El Dorado
Co., 1 male, Blodgett Forest, 21 km E Georgetown,
12. VIII. 1975, P. A. Rude(AC);HumboldtCo., 1 male, Willow
Creek, 12. VIII. 1948, W. W. Wirth (USNM); Madera Co., 1
male, SE slope of Green Mtn., 2,316 m, 20. VIII. 1971, H. B.
Leech (CAS); Mono Co., 1 male, 8 km N Bridgeport, Huntoon
Forest Camp, 2,073 m, 22. VIII. 1966, P. H. Arnaud (AC);
Siskiyou Co., 1 male, Poker Flat, 1,536 m, 13.VIII.1966, H.
B. Leech (CAS); Tuolumne Co., 1 male, Pinecrest, 1 1 .VIII. 1 948,
P. H. Arnaud (PHA). Georgia: Dade Co., 1 male, Cloudland
Cyn. State Park, 8.V.1952, O. Peck (CNC). Idaho: Kootenai
Co., 1 male, Carlin Bay, Lk. Coeur d'Alene, 16 km N of Har-
rison, 640 m, 14/20. VII. 1981, W. J. Turner (AC), 1 male, 217
23. VIII. 1977 (WSU), 1 male, 24/28. VII. 1977 (WSU); Latah
Co., 42 males, Lost Creek, 19.3 km ENE of Potlatch, 823 m,
5. VIII. 1979, W. J. Turner (WSU), 6 males (AC), 1 male,
1. VII. 1980 (WSU), 1 male, 7. VII. 1980 (WSU), 1 male (AC),
7 males, 23. VII. 1980 (WSU), 1 male (AC), 2 males, Styrchnine
Creek, 24.1 km ENE Potlatch, 884 m, 1/3.VH.1980 (WSU),
1 male, 7/9. VII. 1980 (WSU), 1 male, Physocarpus, 1. VII. 1980
(WSU), 4 males, 1 1 .3 km NNE of Moscow, 823 m, 24. VII. 1980,
W. J. Turner (WSU), 5 males, Big Meadow Rec. Area, 1 1.3
km N of Troy, 914 m, 31. VII. 1979, W. J. Turner (WSU), 3
males (AC), 1 male, L. Sand Creek nr. Bonami Creek, 25.7
km E of Potlatch, 884 m, 9. VIII. 1979, W. J. Turner (WSU).
Iowa: Boone Co., 1 male, Ledges State Park, 23.VI.1961, J. L.
Laffoon (UICM). Michigan: Midland Co., 1 male, 5. VII. 1951,
R. R. Dreisbach (USNM). New Jersey: Monmouth Co., 1 male,
Long Branch, 1 l.VL, C. W. Johnson (MCZC). Oregon: Baker
Co., 3 males, Up. Goose Creek, 54.7 km SE of Union, 1,268
m, 13/1 9. VII. 1975, E. J. Davis (WSU), 1 male (AC), 1 male,
Low. Goose Creek, 57.9 km SE of Union, 1,219 m, 13/.
19.VII.1975, E. J. Davis (WSU), 1 male, Velvet Creek, 22.1
km SE of Union, 1,439 m, 13/1 9. VII. 1975, E. J. Davis (AC);
Grant Co., 1 male, 40 km N Mt. Vernon, 24.VII.1974, P. H.
Arnaud (CAS); Union Co., 1 male, Low. Lick Creek, 41.8 km
SE of Union, 1,305 m, 2 1/23. VII. 1977, E. J. Davis (WSU).
Virginia: Montgomery Co., 1 male, Blacksburg, 640 m,
28. VI. 1962, J. G. Chillcott (AC); Roanoke Co., 1 male, Mt.
Roanoke, 1 8. V.I 965, J. G. Chillcott (AC). Washington: Asotin
Co., 1 male, 6.4 km S of Anatone, 1,097 m, 12. VIII. 1980, W.
Turner (WSU), 1 male, Fields' Spring State Park, 1,067-1,219
m, 30. VI. 1975, W. J. Turner (WSU); Chelan Co., 1 male,
Lucerne, 29. VIII. 19 19, A. L. Melander (USNM). CANADA:
Manitoba: 1 male, Ninette, "Betula glandulosa, Populus bal-
samifera associate," 15. VII. 1958, J. G. Chillcott (CNC). On-
tario: Ottawa, 1 male, 26. VI. 1 958, J.G. Chillcott (AC), 1 male,
9.VI.1962, J. R. Vockeroth (CNC), 1 male, Maynooth,
22.VI.1953, J. F. McAlpine (CNC). Quebec: 1 male, Old Chel-
sea, Summit King Mtn., 351 m, 25. VI. 1962, J. R. Vockeroth
(CNC), 1 male, 9. VIII. 1961 (CNC). Saskatchewan: 1 male,
Uranium City, 59°34'N 108°36'W, 18. VI. 1962, J. G. Chillcott
(CNC), 1 male, Wallwort, 52°33'N 104°03'W, 20. VII. 1 942, J.
D. Ritchie (UK).
REMARKS.— Development of the ventral setae
and tubercle of the posterior femur shows con-
siderable variation. Specimens from the Atlantic
Coast have long setae, a distinct nipple-like tu-
bercle, and genitalia confirming substantially to
Figures 25 and 27. Specimens collected between
Ontario and California have short setae, and more
often a gentle prominence rather than a tubercle,
with genitalia resembling Figures 24 and 26. Cal-
ROGERS: AMERICAN DRAPET1S
189
28
29
FIGURES 28-30. Drapetis latipennis. Figure 28. Posterior view of male hind leg, Douglas Co., Kansas. Figures 29 (ventrolateral
view) and 30 (dorsolateral view). Male genitalia, Orange Co., California.
ifornia material often shows well-developed tu-
bercles and genitalia similar to Atlantic Coast
specimens. Abdominal tergites VI and VII are
mostly polished in specimens from Lost Creek,
Idaho. Specimens from Velvet Creek, Oregon,
and Coeur d'Alene Lake, Idaho, lack all pro-
pleural tomentum. Sternite V is undivided in
specimens from Big Meadow Creek, Idaho.
190
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 7
31
32
mm
FIGURES 3 1-33. Drapetis Solaris n. sp., paratype male. Figure 3 1 . Anterior view of femur of center leg. Figures 32 (ventrolateral
view) and 33 (dorsolateral view). Genitalia.
Individuals from the Pacific Coast have so far
been collected only in mountainous regions,
where wing infumation occurs only sporadically
and is coincident with generalized melanism.
Drapetis latipennis Melander
(Figs. 9, 28-30)
Drapetis latipennis Melander, 1902:209.
DIAGNOSIS.— Three strong pairs of verticals;
gena obscured by eye in profile; propleural to-
mentum profuse, extending to anteroventral lat-
eral corner of humerus; posterior femur bearing
five to seven long, sturdy anteroventral bristles
along distal quarter; posterior tibia bearing a
group of long, down turned setae near distal third
of extensor surface; genitalia as in Figures 29, 30.
TYPES. — Lectotype (here designated): male
(USNM), USA: labelled, "Lawrence Kans."
Paralectotypes, same locality: 1 male (USNM),
3 females (AMNH 730). A male from Milwau-
kee, Wisconsin (AMNH 729) included in the
original type series is Drapetis infumata.
OTHER MATERIAL EXAMINED.— USA: California: Orange Co.,
2 males, Irvine Ranch nr. Bonita Creek., "on Platanus,"
24. V.I 963, E. I. Schlinger (UCR). Illinois: Champaign Co., 1
ROGERS: AMERICAN DRAPETIS
191
0.1 mm
FIGURES 34-36. Drapetis torulosa n. sp., paratype male. Figure 34. Anterior view of center leg. Tomentose area indicated
by stippling. Figures 35 (ventrolateral view) and 36 (dorsolateral view). Genitalia.
male, Urbana, 29.VI.1915 (IHNS). Kansas: Douglas Co., 1
male, Atherton (USNM). New York: Ulster Co., 1 male, Cher-
rytown, 8/1 8.VIII. 1971, P. and B. Wygodzinsky (AMNH).
REMARKS.— Flattened spines of abdominal
tergites IV and V are long, thick and erect. The
peculiar long setae on the posterior tibia are pres-
ent in both sexes, but are shorter in females;
Drapetis assimilis Meigen has less developed long
setae in the same location.
Drapetis Solaris n. sp.
(Figures 4, 11,31-33)
DIAGNOSIS.— Center femur bearing ventrally
directed setae across center of anterior surface,
Figure 31; genitalia as in Figures 32, 33.
DESCRIPTION.— Male. Length 1.6 mm. Head
black; three pairs of verticals, outer two short;
frons width at ventral end one-sixth length of
third antennal segment; face linear; gena entirely
visible in profile; third antennal segment shaped
as in Figure 4, arista seven times its length. Tho-
rax black, except posterior edge of mesopleuron
and dorsal edge of sternopleuron dark brown;
humerus bearing an inconspicuous bristle; pro-
pleural tomentum limited to ventral margin; hal-
ter light brown. Wing 1.6 mm long, neuration as
in Figure 1 1 . Legs and coxae chiefly dark brown,
slightly lighter at knees, on tarsal articles and
anterior tibia, and in ventral half of anterior coxa;
center femur bearing a single row of ventral spines
along entire femoral length, spines longer proxi-
192
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 7
mally, anterior surface of femur bearing ventrally
directed setae and an unusually sturdy antero-
ventral distal bristle; center tibia very slightly
bowed; a row of very long ventral setae along
entire length of posterior femur. Abdomen lat-
erally bearing long, flattened spines on tergites
IV and V.
TYPES. -Holotype: male (CAS 13455), MEX-
ICO: labelled, "MEXICO: Mexico: Teotihuacan
Pyramid to the Sun. 27.XII.1970, P. H. and M.
Arnaud/Coll. at flowers Cassia tomentosa L.f.
Fam. Leguminosae Det. D. Breedlove." Para-
types: 3 males, same data (CAS).
REMARKS.— The face of one paratype is con-
stricted ventrally, and another has no discernible
outer vertical. The species is named in reference
to its type locality.
Drapetis torulosa n. sp.
(Figs. 5, 8, 34-36)
DIAGNOSIS.— Three pairs of strong verticals;
anterodorsal surface of center femur with a to-
mentose callosity erupting at proximal quarter
and gradually subsiding distally, Figure 34, pos-
teroventral surface bearing one row of bristles,
bristles lengthening distally; center tibia lacking
flexor spines; genitalia as in Figures 35, 36.
DESCRIPTION. — Male. Length 1.4 mm. Head
black; three pairs of verticals; frons width at ven-
tral end one-eighth length of third antennal seg-
ment; face strongly constricted at ventral end;
gena entirely visible in profile; third antennal
segment shaped as in Figure 5, arista five times
its length. Thorax dark brown; propleuron com-
pletely polished, three weak pale setae along ver-
tical suture; halter dark yellow. Wing 1.6 mm
long, neuration as in Figure 8. Legs and coxae
yellow, except tarsi yellowish brown; center fe-
mur with one very strong, nearly erect antero-
ventral distal bristle near knee, and strong yellow
posteroventral bristles in a single row, bristles
lengthening distally along proximal three-quar-
ters of femur; anterodorsal surface of center fe-
mur thickly tomentose, with a long swelling be-
ginning at proximal quarter and distally subsiding
across remaining length of femur; center tibia
lacking flexor spines; posterior femur lacking
ventral setae, bearing a rounded basal lump on
posteroventral surface just distad of trochanter.
Abdomen laterally bearing tiny flattened spines
on tergites IV and V, closely appressed to body.
TYPES. -Holotype: male (CAS 13459), MEX-
ICO: labelled, "MEXICO: Baja California: Agua
Caliente (San Carlos), 18.5 km east of Maneade-
ro, 6. VII. 1973, Paul H. Arnaud, Jr." Paratypes,
same data, 1 male (AC), 2 males (PHA).
REMARKS.— Abdominal sternite V of a macer-
ated paratype is divided through the center. This
species is named in reference to the peculiar
modification of the center femur.
ACKNOWLEDGMENTS
Assistance of the following persons is grate-
fully recognized: for loans of specimens: P. H.
Arnaud, Jr., G. W. Byers, S. I. Frommer, L. V.
Knutson, the late J. L. Laffoon, A. F. Newton,
J. A. Powell, R. O. Schuster, H. J. Teskey, F. C.
Thompson, W. J. Turner, the late P. W. Wygod-
zinsky; for loan of equipment: H. V. Daly; for
help in manuscript preparation: C. C. Magowan,
M. B. McAdow, and E. I. Schlinger; for funding:
B. A. Czuba and the late A. P. Philips.
LITERATURE CITED
BIGOT, J. F. M. 1857. Essai d'une classification generate et
synoptique de 1'ordre des Insectes Dipt6res. Ann. Soc. Ento-
mol.Fr. 26:551-564.
CHVALA, M. 1975. The Tachydromiinae (Dipt. Empididae)
of Fennoscandia and Denmark, Vol. 3. Scandinavian Sci-
ence Press, Klampenborg. 336 pp.
COLLIN, J. E. 1961. Empididae. British flies, Vol. VI. Cam-
bridge University Press, Cambridge. 782 pp.
GRIFFITHS, G. C. D. 1972. The phylogenetic classification of
Diptera Cyclorrhapha, with special reference to the structure
of the male postabdomen. Dr. W. Junk, The Hague. 340 pp.
KOVALEV, V. G. 1 972. Diptera of the genera Drapetis Mg.
and Crossopalpus Bigot (Empididae) of the European part
of the USSR. Entomol. Obozr. 61:173-196.
MEIGEN, J. W. 1822. Systematische Beschreibung der be-
kannten europaischen zweifliigeligen Insekten, Vol. III.
Hamm. 416 pp.
MELANDER, A. L. 1 902. A monograph of the North American
Empididae. Part I. Trans. Am. Entomol. Soc. 28:195-367,
pis. 5-9.
. 1918. The dipterous genus Drapetis Meigen (family
Empididae). Ann. Entomol. Soc. Am. 11:183-221.
-. 1928. Diptera. Fam. Empididae. Genera Insectorum
185:1-434.
. 1965. Family Empididae (Empidae, Hybotidae). Pp.
446-48 1 in A catalog of the Diptera of America north of
Mexico. A. Stone, C. W. Sabrosky, W. W. Wirth, R. H.
Foote, and J. R. Coulson, eds. U.S.D.A., Agric. Res. Serv.,
Agric. Handb. 276.
ROGERS, E. 1983. The Neotropical species of Drapetis Meigen
(Diptera: Empididae). Syst. Entomol. 8:431-452.
PROCEEDINGS
OF THE
CALIFORNIA ACADEMY OF SCIENCES
Vol. 46, No. 8, pp. 193-207, 2 figs. December 20, 1989
ECOLOGICAL, EVOLUTIONARY, AND CONSERVATION
IMPLICATIONS OF FEEDING BIOLOGY IN OLD WORLD
CAT SNAKES, GENUS BOIGA (COLUBRIDAE)
By
Harry W. Greene*
Museum of Vertebrate Zoology,
University of California, Berkeley, California 94720
* Research Associate, Department of Herpetology,
Fellow, California Academy of Sciences,
Golden Gate Park, San Francisco, California 94118
ABSTRACT: Cat snakes forage actively and sometimes consume multiple prey items at a site. Some species
use an unusual style of constriction and venom to immobilize prey, which is typically swallowed head-first.
Approximately 300 prey items from 21 species demonstrate intra- and interindividual, ontogenetic, and geo-
graphic dietary variation. Most species of Boiga are small to moderately long snakes that eat mainly lizards,
or huge snakes that eat lizards as juveniles and mainly mammals and/or birds as adults. Probably predation
on birds by introduced B. irregularis on Guam is a retained primitive feeding pattern, whereas that species'
diet of rodents on New Guinea and euryphagy by B. dendrophila throughout its range are correlated with
local ecological factors. With few exceptions (e.g., a mouse-deer in B. dendrophila), prey/predator mass ratios
are not large compared to other snakes. Morphological and ecological considerations suggest that, like B.
irregularis, several other species of cat snakes could have catastrophic impact as introduced predators on
island bird and bat faunas.
Received February 16, 1989. Accepted July 6, 1989.
evolutionary, and conservation implications of
The Old World genus Boiga includes approx- that variation.
imately 25 species of moderate to large, rear- The feeding biology of cat snakes is of special
fanged colubrids (Leviton 1970; Kroon 1973; interest for at least three reasons:
Rasmussen 1979), often called cat or tree snakes (1) Boiga is morphologically and ecologically
because of their nocturnal, arboreal habits and diverse, and thus offers interesting material for
vertical pupils. This report surveys published in- broader evolutionary considerations. For ex-
formation on feeding in cat snakes, and analyzes ample, although rear-fanged snakes play prom-
additional data based on the examination of mu- inent roles in discussions of venom evolution
seum specimens. It addresses the nature of di- (e.g., Smith and Bellairs 1947; Kardong 1979,
etary and behavioral variation within and among 1 980; Savitzky 1 980), an understanding of that
2 1 species of Boiga, and discusses ecological, problem is hampered by a lack of information
[193]
194
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 8
on the feeding biology and phylogenetic rela-
tionships of informative taxa (Cadle 1983).
(2) Cat snakes might be important arboreal
predators on vertebrates in many Old World
tropical ecosystems, and three to five species of
Boiga sometimes occur sympatrically (e.g., Tay-
lor 1922a; Lim 1955; Lloyd et al. 1968; Kroon
1973; Inger and Colwell 1977). Autecological
analyses provide data for subsequent commu-
nity-level research, information that could not
be obtained for snakes at a single site during a
short-term study (Greene 1988a).
(3) Predation by Boiga irregularis was a key
factor in the recent, precipitous decline of bird
populations on the Pacific island of Guam, where
that species was inadvertently introduced in the
1940s (Savidge 1984, 1987, 1988; Conry 1988;
Fritts 1988). Island populations of fruit bats, al-
ready widely threatened by direct human per-
secution, might be at risk as well (Wiles 1988).
Data on the natural diets of cat snakes could
prove useful in managing this problem and pre-
venting additional catastrophes.
METHODS
Stomach contents were examined through
ventral slits in specimens in the Academy of Nat-
ural Sciences of Philadelphia (ANSP); American
Museum of Natural History, New York (AMNH);
California Academy of Sciences, San Francisco
(CAS); Field Museum of Natural History, Chi-
cago (FMNH); Museum of Comparative Zool-
ogy, Harvard University, Cambridge (MCZ);
Museum of Southwestern Biology, University of
New Mexico, Albuquerque (MSB); Museum of
Vertebrate Zoology, University of California,
Berkeley (MVZ); and National Museum of Nat-
ural History, Washington (USNM). I avoided
especially soft, brittle, or otherwise fragile spec-
imens; all others for each species were examined
to preclude a bias for large prey caused by open-
ing only snakes with externally palpable food
items. My results also incorporate data on 10
New Guinea specimens of Boiga irregularis that
contained prey (R. Shine, in litt.) in the Austra-
lian Museum (AM) and Queensland Museum
(QM), and specific published information on
feeding. I took care to account for redundancy
among literature records and museum specimens
(e.g., Schmidt 1927 and Pope 1935; Smith 1916
and 1930), and sometimes obtained information
not mentioned previously.
Direction of ingestion, identity, and linear di-
mensions of prey items were recorded whenever
possible. Intact prey items (or a reference spec-
imen of comparable size) and the predators were
weighed after blotting and draining them briefly
on paper towels. Abbreviations refer to snout-
vent length (SVL), total length (TL), maximum
total length (MTL), and prey/predator mass ratio
(MR). All bird or lizard eggs in a stomach were
counted as one item, because they might repre-
sent a single location event and because some-
times the exact number of items was impossible
to determine.
SPECIES ACCOUNTS
Boiga angulata (Peters 1861)
Philippine Islands, MTL 1.4 m (Leviton 1970).
Four snakes contained five agamid lizards ( 1 Ca-
lotessp.,3Dracosp.,x= 1.3/snake; Griffin 1910;
CAS 154175, CAS-SU 24179; FMNH 61620).
MRs for two Draco sp. in snakes with SVLs of
646 and 1,079 mm were 0.39 and 0.12, respec-
tively. Two Draco sp. and an unidentified agamid
were swallowed head-first.
Boiga blandingi (Hallowell 1857),
Blanding's Tree Snake
Africa, MTL 2.7 m (Pitman 1974; Goodman
1985). Fifteen specimens contained at least 30
prey items (x = 2.0/snake): a 1.8 m TL snake
contained 1 agamid lizard (Agama agamd) and
ca. 8 bats (Eptesicus tenuipennis, Vespertilioni-
dae; Tadarida sp., Molossidae); 5 each contained
1 bat; 1 contained bats and 1 lizard; 1 contained
a chameleonid lizard; 1 contained 4 birds; 1 con-
tained a sunbird (Nectarinafamosa); 4 each con-
tained 1 bird, 2 swallowed head-first; and 1 con-
tained 3 bats, swallowed tail-first, MRs = 0.03/
each (Laurent 1956; Woodward 1960; Barry
1961; Cozens 1961; Dunger 1961; Menzies 1961;
Pitman 1962; Hedges 1983; FMNH 179319-20;
MCZ 54007). Three adults in one tree each con-
tained young weaver birds, Plesiositagra cucul-
latus, and others contained fledgling birds and
immature rats (Jones 1961).
Boiga ceylonensis (Giinther 1858),
Sri Lankan Cat Snake
India and Sri Lanka, MTL 1 .6 m (Smith 1 943).
Forty-two prey items included 8 frogs (1 Rana
GREENE: OLD WORLD CAT SNAKES
195
limnocharis, Ranidae; 5 Philautus sp., Rhaco-
phoridae), 21 lizards (1 Gekkonidae; 2 Calotes
sp., 3 C. versicolor, 1 Charasia dor sails, 3 Salea
horsfieldi, Agamidae), 3 birds (1 fledgling), and
2 mice (Wall 1919 [including snakes referred to
as B. nuchalis], 1923, 1924;Hutton 1949; Taylor
1 953; FMNH 1 3 1 380, 1 67006). Two lizards were
swallowed head-first.
Boiga cyanea (Dumeril and Bibron 1854),
Green Cat Snake
India and Southeast Asia, MTL 2.3 m (Smith
1943). Seven snakes contained 1 snake (a female
B. cyanea, 1,350 mm TL, in a 1,260 mm TL
male) and 6 birds (1 sparrow [Passer sp., Pas-
seridae], 1 probable wagtail [Motacilla sp., Mo-
tacillidae, in a snake of 821 mm SVL], 1 week-
old domestic chick [Gallus gallus, Phasianidae];
Smith and Kloss 1915; Saint Girons 1972; Frith
1977; MCZ 58260). One bird had been swal-
lowed head-first.
Boiga cynodon (Boie 1827),
Dog-toothed Cat Snake
Southeast Asia and Indo- Australian Archipel-
ago; MTL 2.8 m (Smith 1 943; Saint Girons 1 972;
Kroon 1973; Tweedie 1983; FMNH 131809).
Nineteen snakes contained 20 items (x = 1.05/
snake), including 3 lizards (Draco sp., D. volans,
Agamidae; Gekko gecko, Gekkonidae), 1 2 birds,
4 or more bird eggs, and 1 rodent (Flower 1899;
Auffenberg 1980; ANSP 26421; CAS 8533,
125173; FMNH 53454-6, 71628, 131809,
131811, 150883, 158652-3, 168006, 183749).
One lizard, 1 rodent, and 5 birds were swallowed
head-first. A small snake (SVL = 360 mm) had
eaten a Draco sp., and a juvenile (SVL = 640
mm) contained a relatively large bird (MR =
0.44). An intact bird's egg, maximum diameter
51 mm, was swallowed by a snake with head
length of 52.2 mm.
Boiga dendrophila (Boie 1827),
Mangrove Snake
Southeast Asia and nearby archipelagos, MTL
2.8 m (Burger 1975; Tweedie 1983). Fifty-four
prey items in 52 snakes (Jc = 1 .04/snake) included
1 frog, 15 lizards (9 scincids [2 Lamprolepis
smaragdinum]; 2 lacertids, Tachydromus sexlin-
eatus; 3 agamids [2 Calotes sp., 1 C. cristatelus];
1 varanid, Varanus salvator); 1 snakes (1 B. cyn-
odon, 2 Chrysopelea ornata, 1 Psammodynastes
pulverulentus, Colubridae; 1 Calloselasma rho-
dostoma, Viperidae); 17 mammals (2 bats [1
Eonycteris spelea, Pteropodidae], 1 mouse-deer
[Tragulus javanicus, Tragulidae], 14 rodents
[probably mostly Muridae, including Rattus sp.]);
and 13 birds (Flower 1899; Smith 1916, 1930;
Taylor 1922b; Brongersma 1934; Harrison 1962;
Leviton 1970; Lim 1973; Minton and Dunson
1978; Smith 1985; ANSP 264 17-8; CAS 12362-
3, 15313; CAS-SU 28525-6; FMNH 53460,
53462, 53466, 68912, 71630, 96609, 120393,
131808, 150879, 158679, 178602, 230061).
Juvenile Boiga dendrophila eat mainly ecto-
therms (lizards and rarely snakes), whereas adults
eat mainly birds and/or mammals (5/6 times vs.
25/34 items, respectively, using Brongersma's
[1934] sample and museum records; P = 0.01,
Fisher Exact Test). Sample sizes are insufficient
for statistical comparisons, but I detect no evi-
dence for major geographic variation in the diet
of adults: frequencies of lizards, snakes, birds,
and mammals, respectively, are 1, 1, 5, and 3
for Borneo; 5, 2, 2, and 5 for the Philippine Is-
lands (all of the mammals were in Palawan man-
grove snakes); 2, 1,3, and 0 for Sulawesi; and 1,
3, 0, and 2 for mainland Asia. The largest snake
containing a lizard and the smallest containing
an endotherm had SVLs of 1 , 1 1 2 mm and 840
mm, respectively. Four lizards, 3 snakes, 3 birds,
4 rodents, and 1 adult mouse-deer were swal-
lowed head-first, the latter by a 1.7 m TL snake.
MRs for 2 lizards and a bird were 0.10-0.18 (x
= 0.14).
Boiga dightoni (Boulenger 1894)
India, MTL 1.3 m (Smith 1943). One adult
(SVL = 915 mm) contained an agamid lizard,
Calotes versicolor, swallowed head-first, MR =
0.24 (FMNH 217699 [Inger et al. 1984]).
Boiga drapiezii (Boie 1827),
White-spotted Cat Snake
Southeast Asia, MTL 2.1 m (Tweedie 1954,
1983). Eight prey items included 5 lizards (1
Gekkonidae; 1 Draco sp., 1 Gonocephalus sp.,
Agamidae), 1 rhacophorid frog, and 2 birds (Lim
1956; FMNH 131804, 138594, 196805-6). Both
agamids and the frog were swallowed head-first.
MRs for the gecko and Draco were both 0.25.
196
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 8
Boiga forsteni (Dumeril and Bibron 1854)
Sri Lanka and India, MTL 2.3 m (de Silva
1 980). Stomachs contained birds and a bat (Wall
1921).
Boiga gokool (Gray 1834),
Eastern Gamma Cat Snake
Asia and India, MTL 1.2m (Smith 1 943). Two
snakes contained a mouse and an agamid lizard,
the latter swallowed head-first (Wall 1910; MCZ
58261).
Boiga irregularis (Merrem 1790),
Brown Tree Snake
New Guinea, northern Australia, and nearby
archipelagos; MTL 2.3 m (Fritts 1988). Fifty-
nine snakes contained 65 items (1-3 per stom-
ach, x = 1.10), including 2 frogs (1 Rana sp.,
Ranidae), 19 lizards (1 agamid, Gonocephalus
modestus; 10 scincids [2 Emoia cyanura, 3 Lam-
prolepis smaragdinum], 8 gekkonids [1 Gehyra
sp., 2 G. mutilata, 1 Gekko vittatus]), 13 birds
(1 passerine, 1 owlet-nightjar [Aegotheles ben-
nettii, Aegothelidae], 2 starlings [Aplonis sp.,
Sturnidae], 1 quail [Coturnix chinensis, Phasian-
idae]), and 30 mammals (1 marsupial bandicoot,
Peramelidae; 1 bat; 1 shrew, Croddura sp.; 17
rodents (5 juveniles, 1 adult Rattus exulans, 1
R. rattus, 1 Mus musculus, all murids; Rooij 1917;
Kopstein 1926; Hediger 1933; Pendleton 1947;
McDowell 1984; AM 4553, 11825, 12443, 12542,
14858, 86923, 87463, 87466; AMNH 40233,
42339, 42371, 59083, 62031-2, 65513, 73962,
85727-8, 85731, 95553, 95607, 101080, 101083;
CAS 49910, 94027, 103388, 113600, 121224;
FMNH 14076, 41980; QM 4381, 13.1280;
USNM 6197, 61918, 61920, 159971, 195594,
213431, 215938, 215942, 215947, 216007,
216009, 237136, 237634). Stomachs with mul-
tiple prey included 1 with 2 E. cyanura, 1 with
2 G. mutilata, 1 with 2 young Aplonis sp., and 1
with 3 nestling rodents (apparently a single
species). Three "bird" records consisted of 1 egg,
several eggs, and 1 bird plus 2-5 eggs.
Forty-eight measurable individuals containing
prey had SVLs of 292-1,710 mm. A 285 mm
SVL specimen had an obvious umbilical scar,
and six females containing enlarged ova or shelled
oviductal eggs had SVLs of 840-1,305 mm. On
those bases, I divided snakes containing prey into
"juveniles" (equal to or less than 750 mm SVL)
and "adults" (greater than 750 mm SVL). Ju-
veniles take significantly more ectothermic prey
(snake, frogs, lizards) vs. endothermic prey (birds,
mammals), whereas adults show the opposite bias
(14:3 vs. 6:32, respectively; P < 0.01, x2 test).
The largest snake containing a lizard was 1 ,000
mm SVL, the smallest containing an endotherm
was 646 mm SVL, and the shift from emphasis
on one prey type to the other occurs at ca. 800-
900 mm SVL (Fig. 1). Larger snakes ate larger
prey than did smaller snakes (Fig. 2). Prey/pred-
ator mass ratios ranged from 0.004-0.241 (x =
0.106, n = 13), and the largest item was a 32 g
starling (Aplonis sp.) in a 152 g Boiga irregularis
from the Bismark Archipelago.
The 65 prey items are from throughout the
species' range except for Australia, where the diet
of this species is under study by R. Shine. There
is significant geographic variation in diet (P <
0.01, x2 test) if 38 records for endothermic prey
are divided into those from mainland New
Guinea vs. those from smaller islands (all others
except Sulawesi, including the Solomon, Bis-
marck, and Molucca groups). Snakes from New
Guinea ate mammals more frequently than they
ate birds (19:3), in agreement with summaries of
unspecified numbers of prey (Room 1 974; Parker
1982) and general comments about diet on the
tags of AMNH 74506 and 82334. The single diet
record from Sulawesi, also a "large island," was
of a mammal. Snakes from "smaller islands"
consumed birds and their eggs more frequently
than they ate mammals (10:6).
Direction of ingestion was determined for 36
prey items. Twenty-eight were swallowed head-
first, including 2 frogs, 8 lizards, 1 1 birds, 1 bat,
1 snake, and 5 rodents. Seven were swallowed
tail-first, including 2 lizards, 1 bird, and 4 ro-
dents. One lizard was swallowed bent doubled.
Savidge (1 988) analyzed 494 prey items in 353
Boiga irregularis from an introduced population
on Guam. As in natural populations, those snakes
ate primarily lizards as juveniles and birds and/
or mammals as adults. Her mean of 1.4 items/
snake is higher than I observed, perhaps because
she counted eggs as individual items in a stom-
ach. Wiles (1988) reported three juvenile fruit
bats in the stomach of a 2.5 m TL snake on
Guam, and indirect evidence that these snakes
were raiding bat roosts.
GREENE: OLD WORLD CAT SNAKES
197
Boiga jaspidea (Dumeril and Bibron 1854),
Jasper Cat Snake
Southeast Asia, MTL 1.5 m (Taylor 1965;
Tweedie 1983). Stomachs contained 1 agamid
lizard (Calotes sp.), 1 bird, and 1 tree-mouse
(Chiropodomys gliroides, Muridae) in a 1,213 mm
TL snake (Lim 1956, 1967).
Boiga kraepelini Stejneger (1902)
East Asia and nearby islands, MTL 1.6 m
(Kuntz 1963). Four snakes contained 2 agamid
lizards (swallowed head-first), 2 birds, and 2 bird
eggs, ready to hatch (Pope 1935; CAS 18891-2;
FMNH 24986).
Boiga multifasciata (Blyth 1860),
Himalayan Cat Snake
Himalayan region, MTL 1.06 m (Smith 1943).
One contained an agamid lizard, Japalura var-
iegata(Wa\\ 1909).
Boiga multomaculata (Boie 1827),
Large-spotted Cat Snake
East Asia, Java, Sumatra, and Borneo; MTL
1.2 m (Smith 1943). Stomachs contained 6 birds
(1 Motadlla sp., Motacillidae) and 8 agamid liz-
ards (Calotes sp., C. versicolor); three lizards (1
with MR = 0.58) and a bird were swallowed
head-first (Wall 1926a, b; Schmidt 1927; Pope
1935; FMNH 26451, 105704, 178666; MSB
40769).
Boiga nigriceps (Gunther 1863),
Dark-headed Cat Snake
South Asia and nearby archipelagos; MTL 1.7
m (Rooij 1917). One snake contained a bird,
swallowed head-first (FMNH 128150).
Boiga ocellata Kroon (1973)
South Asia; MTL 1.7 m (Kroon 1973). Two
adults contained a bird and a bird egg (FMNH
191998-9).
Boiga ochracea (Gunther 1868),
Tawny Cat Snake
Asia, MTL 1.3m (Smith 1 943). Ten prey items
included 8 agamid lizards (2 Japalura variegata,
3 Calotes sp.), 1 fledgling bird, and 1 mass of
bird eggs; three of the agamids were swallowed
head-first (Wall 1909, 1926a, as B. hexagonatus;
CAS 12365, 95252).
Boiga pulverulenta (Fischer 1856),
Powdered Tree Snake
Africa, MTL 1.3 m (Pitman 1974). Ten snakes
contained 13 prey items (x = 1.3/snake): 3 mice
(2 juvenile Praomys sp., Muridae, in 1 snake), 2
agamid lizards, and 8 birds (Werner 1909; Pit-
man 1974; FMNH 4017, 59006-7; LACM
38688, 46328; MCZ 29354, 29357; USNM
223775). Two lizards and 6 birds were swallowed
head-first; 2 snakes containing lizards were 330
and 850 mm SVL, whereas 5 containing birds
were ca. 700-780 mm SVL. MRs were 0.12 for
each lizard and 0. 1 5 for each of 4 birds in 2
snakes (x — 1 .4). One record of a bat in this
species (Woodward 1960) is based on a misi-
dentified B. blandingi (see above).
Boiga schultzei Taylor (1923)
Philippine Islands; MTL 1.4 m (Taylor 1923).
A juvenile, SVL 310 mm, had swallowed 2 gek-
konid lizards, Gehyra mutilata, head first (CAS
62153).
Boiga trigonata (Schneider in Bechstein 1 802),
Gamma Cat Snake
Southern Asia, India, Sri Lanka; MTL 1.2 m
(Smith 1 943). ". . . the commonest contents were
birds, chickens and eggs" (Lindberg 1932). Four-
teen snakes contained 1 5 prey (x = 1 .07 items/
snake), including 1 1 lizards (Acanthodactylus sp.,
Lacertidae; at least 4 Calotes sp., 2 C. versicolor,
Agamidae; 1 scincid; 1 set of 3 lizard eggs) and
3 birds (1 Saxicoloides fulica, Muscicapidae); two
lizards and 1 bird were swallowed head-first (Wall
1907a, 1919, 1921; Minton 1966; Sharma and
Vazirani 1977; CAS 13686; FMNH 131958,
166998, 171761). Ants and grasshoppers in the
stomach of a small Indian specimen (Sunder-
singh 1960) were probably remnants from the
stomach contents of a prey lizard.
DISCUSSION
INDIVIDUAL AND GEOGRAPHIC DIETARY
VARIATION. — Intraspecific variation in snake
diets can include intra- and interindividual, on-
198
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 8
togenetic, seasonal, and geographic components
(Arnold 1977; Greene 1984; Seib 1985; Mushin-
sky 1987; Macias Garcia and Drummond 1988;
Savidge 1988). At one extreme, Thamnophis ele-
gans (Colubridae) shows extensive geographic
variation, and litter mates have genetic predis-
positions for alternative prey types (Arnold 1977);
conversely, the diet of Micrurus fulvius (Elapi-
dae) varies only slightly at all sizes and through-
out the year over a large geographic range (Greene
1984).
In studies of museum specimens, evidence for
intraindividual dietary variation comes from
multiple prey types in single stomachs, as in a
Boiga ceylonensis that contained a bird and two
mice, and two B. blandingi that each contained
bats and a lizard. All such records for B. angu-
lata, B. cynodon, B. dendrophila, B. irregularis,
B. kraepelini, B. ochracea, B. pulverulenta, B.
schultzei, and B. trigonata apparently are mul-
tiples of one prey species, as are several records
for B. blandingi. Of 353 B. irregularis on Guam
that contained prey, 67 had more than one item
of the same "prey type" (e.g., lizards and/or liz-
ard eggs, birds and/or bird eggs); 14 had lizards
and birds or mammals, as expected for small
adults; and only five (<6% of stomachs with
multiple items) contained birds and/or their eggs
and mammals (Savidge 1988). These data sug-
gest that individual cat snakes rarely eat birds
and mammals, at least within a short time pe-
riod. Some species, however, often consume more
than one of the same prey type (and even same
prey species), and perhaps this is especially true
for B. blandingi. Surely the clumped distribution
of roosting bats and nestling birds facilitates mul-
tiple ingestion.
There is significant geographic dietary varia-
tion in Boiga irregularis within its natural range,
and microgeographic variation in predation on
rodents vs. birds in urban vs. forest habitats by
introduced snakes on Guam (data from Savidge
[1988, table 1], x2 test, P < 0.01). The data for
other species of Boiga are inadequate to distin-
guish geographic from individual and ontoge-
netic variation.
Perhaps diet is plastic at some localities, or
throughout all or part of the range of some species
of Boiga. If so, interspecific and geographic pat-
terns could result from individual responses to
local prey characteristics and/or availability,
rather than from dietary evolution. The infre-
quent occurrence of multiple prey types in in-
dividual stomachs of several species of Boiga is
consistent with stable prey preferences, whereas
there evidently has been a rapid local response
by B. irregularis to shifting prey densities on
Guam (Savidge 1988, and see above; Green well
et al. [1984] described a possible example in-
volving Thamnophis sirtalis and birds on an is-
land, and Chiszar [1989] discussed the behav-
ioral basis for prey switching in B. irregularis).
Genetic differentiation and phenotypic plasticity
are not mutually exclusive, in that response to
local conditions could override genealogically
constrained prey preferences, and a thorough as-
sessment of these possibilities for cat snakes will
require studies of the ontogeny and control of
feeding behavior in several populations.
PREY/PREDATOR SIZE RELATIONSHIPS AND
ONTOGENETIC VARIATION.— As in other snakes
(Greene 1984; Seib 1984, 1985), larger individ-
uals of Boiga eat larger prey, and also continue
to eat relatively small prey (Fig. 2). Most MRs
for eight species of Boiga are not especially large
compared with those of other snakes (0.004-0.58,
x = 0.16, n = 32; cf. Voris and Moffett 1981;
Greene 1983, 1984, 1989; Seib 1984, 1985;Jayne
et al. 1988). However, a 1.7 m B. dendrophila
would weigh no more than 1 kg (pers. obs.) and
lesser mouse-deer weigh 0.7-2.0 kg (Lekagul and
McNeely 1977), so the latter is a relatively large
item. A bandicoot in B. irregularis and prey
snakes longer than the B. cyanea and B. dendro-
phila that contained them (Smith and Kloss 1915;
Smith 1916) likewise were surely relatively heavy;
a captive B. irregularis consumed a domestic ro-
dent with MR = 0.60 (Chiszar 1989). Records
for B. irregularis and B. pulverulenta demon-
strate that a cat snake can ingest at least a third
of its mass if it eats more than one item in a nest
or roost.
Data from its natural range and the introduced
Guam population agree that Boiga irregularis
changes from eating mainly lizards to eating
mainly birds and/or mammals at a SVL of ca.
80-100 cm. Similar shifts evidently occur in B.
dendrophila and B. pulverulenta, and seem likely
for other large species of Boiga (Fig. 1). Perhaps
that pattern reflects simply an ability of larger
snakes to swallow larger prey, since slopes and
intercepts for endothermic vs. ectothermic prey
are similar (analysis of covariance for data in Fig.
2, P > 0.05). However, large snakes eat endo-
GREENE: OLD WORLD CAT SNAKES
199
V)
I«H
o
CD
DC
CD
12H
• Endothermic prey
U Ectothermic prey
Boiga irregularis
N = 47
1
o
o
0)
DC
I 1 1 r—
20 30 40 50 60 70 80 90 100110120130140150160170180190200210
B 8-
SVL (cm)
6-
4H
CD
.O
• Endothermic prey
ED Ectothermic prey
(17 species)
N = 69
1 1.. 1. 1
i 1 1 1 1 1 1 1— r
20 30 40 50 60 70 80 90 100110120130140150160170180190200210
SVL (cm)
FIGURE 1 . Records for ectothermic and endothermic prey with respect to predator snout-vent length in cat snakes, genus
Boiga. For this comparison, multiple items of one prey type in a single stomach were counted once.
therms that are smaller than the ectotherms tak-
en by smaller snakes (Fig. 2), implying that larger
snakes fail to encounter small ectotherms or have
undergone a real change in prey preference.
FEEDING BEHAVIOR.— Most Boiga prey mainly
on diurnal, arboreal lizards (especially agamids
of the genus Calotes) and diurnal birds. Studies
on free-living and captive B. irregularis suggest
that these nocturnal hunters must search for rath-
er than ambush inactive prey (Fritts et al. 1987;
Savidge 1988; Chiszar 1989; Chiszar etal. 1985,
1 988). Jones (in Pitman 1958) saw three B. blan-
dingi "pushing their heads into one nest after the
other and taking the young [weaver birds]. ..."
However, a B. dendrophila constricting an Eo-
nycteris spelea in a tree after dark probably am-
bushed the bat while it was feeding on or ap-
proaching fruit, as that species typically roosts
in caves (Harrison 1962).
Some Boiga cynodon, B. dendrophila, and B.
trigonata constrict prey with the tail as well as
hold and presumably envenomate an animal with
their teeth (Wall 1907b, 1921; Barach 1952;
Murphy 1977; however, Lim et al. [1960] stated
200
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 8
2.0-
O.
D)
0.5-
0
A
A
0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Log Predator Mass
FIGURE 2. Relationship between log predator mass and log
prey mass for Boiga irregularis (triangles, n = 8) and eight
other species of Boiga (circles, n = 13). Hollow symbols in-
dicate ectothermic prey; solid symbols indicate endothermic
prey. For this comparison, multiple items of one prey type in
a single stomach were counted once.
that B. cynodon does not constrict prey). Boiga
multomaculata constricts lizards, but the precise
method is unknown (Campden-Main 1 970). Boi-
ga blandingi and B. ceylonensis do not constrict
prey, whereas B. forsteni and B. irregularis vary
in this regard (Groves 1973; Murphy 1977; de
Silva 1980; Parker 1982; Shine and Schwaner
1985; Fritts 1988). Young (1983) observed that
B. blandingi controlled prey with loops of the
body, but did not constrict; that B. cyanea over-
powered live mice with constriction "in a man-
ner similar to that of a rat snake or a king snake";
that B. cynodon did not constrict, but controlled
live prey with loose coils of the anterior body;
and that B. dendrophila was "an efficient and
powerful constrictor . . . upon warm blooded prey,
but it rarely employed this behavior ... on
snakes." Captive B. irregularis swallow small prey
immediately and apply constricting coils to rel-
atively heavy prey (Chiszar 1989).
Additional inter- and intraspecific compari-
sons will be required to elucidate the variation
in and evolution of prey-immobilization behav-
ior in Boiga. Constriction with the tail is rare
among snakes (Murphy 1977; Greene 1977) and
absent in the outgroup to Boiga (see below), sug-
gesting three possibilities: (1) It is a derived state
for Boiga, either as yet unobserved or secondarily
lost in most species; (2) some other derived at-
tribute of Boiga (e.g., arboreality, long tails, mul-
tiple constriction of nestlings) favors this type of
constriction and it arose repeatedly in the genus;
or (3) constriction with the tail is a derived at-
tribute for a monophyletic subset of Boiga, a
behavioral synapomorphy for those species.
The venom of B. blandingi is neurotoxic and
can kill a large rodent or a small monkey in less
than an hour (Wakeman 1966; Groves 1973;
Levinson et al. 1976), and at least some other
species have venom that is slightly or moderately
toxic to humans (e.g., Deraniyagala 1955; Whit-
aker 1970; Burger 1975; Minton and Dunson
1978; Cox 1988; Fritts 1988). A captive adult B.
dendrophila (TL ca. 1.8 m) had difficulty sub-
duing and swallowing adult laboratory rats (Ba-
rach 1952), but those animals can be powerful
opponents for captive snakes (pers. obs.).
Head-first prey ingestion is the norm in nature
for Boiga irregularis (28/36 items, P < 0.01, x2
test), as it is for many other snakes (e.g., Greene
1976, 1983, 1984, 1989; Ananjeva and Orlov
1982; Voris and Voris 1983; Seib 1984, 1985).
Of 6 1 cases for which direction of ingestion was
determined for 1 6 other species of Boiga, 58 were
swallowed head-first; 3 bats were eaten tail-first.
EVOLUTIONARY IMPLICATIONS.— An evolu-
tionary consideration of feeding necessitates the
categorization of dietary variation into alterna-
tive states and the identification of appropriate
outgroups, so that the distribution of states among
populations, taxa, and outgroups can be used to
specify transformations during the history of that
clade (see Arnold 1981; Kluge 1987; Losos and
Greene 1988). A phylogenetic consideration of
feeding in Boiga and its relatives, although based
on incomplete data, has implications that justify
the following preliminary analysis.
The precise relationships of Boiga to other col-
ubrids are unknown; Rasmussen (1979, 1985)
placed it with the African genera Crotaphopeltis,
Dipsadoboa, and Telescopus, so these are used
here as a collective outgroup. Species of Crota-
phopeltis are terrestrial and feed on toads. Species
of Dipsadoboa are arboreal and feed on frogs and
lizards. The Eurasian species of Telescopus evi-
dently are terrestrial, whereas the African T. dhara
and T. semiannulatus are arboreal; all eat lizards
and occasionally birds and bats. Species in all
three genera are nocturnal and reach MTLs of
ca. 1 m (Corkhill 1932; Pitman 1974; Broadley
1983).
Most species of Boiga are small to moderately
long snakes that eat lizards throughout their lives,
or large snakes that eat lizards as juveniles and
endotherms as adults. Division of the second of
GREENE: OLD WORLD CAT SNAKES
201
those qualitative groups into four subgroups
yields a total of five dietary character states:
Diet-1 consists primarily of lizards throughout
life, with an occasional bird or mammal. Boiga
ceylonensis possesses this diet, as do perhaps B.
angulata, B. dightoni, B. gokool, and B. ocracea.
These species have MTLs of 1.2-1.6 m.
Diet-2 consists of lizards in juveniles, mainly
of birds in adults. Boiga cynodon, some popu-
lations of B. irregularis, and B. multomaculata
exhibit this diet, as do perhaps B. drapiezii, B.
forsteni, B. kraepeleni, B. nigriceps, B. ocellata,
and B. trigonata. These species have MTLs of
1.2-2.8 m.
Diet-3 consists of lizards in juveniles, mainly
of rodents in adults. Only Boiga irregularis (MTL
= 2.3 m) on New Guinea exhibits this state.
Diet-4 consists of lizards in juveniles, mainly
of mammals and birds in adults. Boiga blandingi
and perhaps B. cyanea, B. jaspidea, and B. pul-
verulentus exhibit this state, and have MTLs of
1.3-2.8 m. However, sample sizes are quite small
except for the first species, and in no case can
the possibility of geographic variation for diet-2
and diet-3 be excluded. Also, bats are probably
functionally equivalent to birds in shape and ar-
boreal roosting behavior (Cundall and Greene,
unpubl.; many other snakes that eat bats also eat
birds, Schaetti 1 984), so that B. blandingi per-
haps should be assigned diet-2.
Diet- 5 consists of lizards in juveniles, mainly
of snakes, birds, and mammals in adults. Stom-
ach contents and observations on captives (Young
1983) suggest that Boiga dendrophila is more eu-
ryphagic than other species in the genus.
Ingroup and outgroup comparisons indicate
that Boiga is primitively nocturnal and perhaps
derivatively arboreal. Diet-1 and relatively small
MTL are found in a few species and in the out-
group, but lacking knowledge of intrageneric re-
lationships, we cannot determine if their pres-
ence in Boiga is primitive or a reversal. Diet-2
and large size are widespread in species of Boiga
and absent in the outgroup (with the possible
exception of diet-2 in some Telescopus), but we
cannot be sure if those characteristics are derived
for the genus or within Boiga, perhaps more than
once.
Diet-2 is probably primitive for adult Boiga
irregularis, based on widespread occurrence in
the genus, suggesting that diet-3 is derived within
that species, but the reasons for intraspecific geo-
graphic variation are not known. Birds are rarely
eaten by B. irregularis in New Guinea, although
the island has a rich avifauna (Beehler et al. 1 986).
Several species of New Guinea birds exhibit nest-
ing attributes that are possibly antipredator spe-
cializations (J. M. Diamond, University of Cal-
ifornia, Los Angeles, in litt.), but perhaps so do
birds on other islands and the mainland that are
commonly eaten by other species of Boiga. A
niche shift in response to competitive pressure
seems unlikely, because no other New Guinea
snakes are known to feed heavily on birds
(McDowell 1969, 1972, 1975, 1984; Shine 1980a;
Parker 1982; Malnate and Underwood 1988).
Perhaps New Guinea arboreal rodents are so
much more diverse and numerous than on near-
by archipelagos (Laurie and Hill 1954; Menzies
and Dennis 1979) that they provide "better" food
per unit effort or are simply encountered more
often than birds.
Based on ingroup and outgroup comparisons,
diet-4 and diet-5 are probably derived within
Boiga. Euryphagy evidently reaches a maximum
for the genus in B. dendrophila and is correlated
with a derived preference for riparian habitats
(Taylor 1965; Frith 1977; Minton and Dunson
1 978; Dring 1 979; Cranbrook and Furtado 1 988;
this proclivity is not reported for other species
of Boiga or the outgroup genera). Some theoret-
ical foraging models predict that habitat spe-
cialization results in broader diets and use of
suboptimal prey (e.g., MacArthur and Pianka
1966; Krebs et al. 1983). The relevance of those
models for cat snakes could be explored with
captive animals, in part, by testing the presump-
tion that alternative prey types vary in profitabil-
ity (see Godley 1980).
My findings on feeding biology in Boiga un-
derscore some cautionary lessons with regard to
the evolution of rear-fanged snakes:
(1) Ontogenetic and/or geographic variation
can be sufficiently large that a proportional rep-
resentation of overall diet for a species will ob-
scure functional, biogeographic, and phyloge-
netic correlates: the overall sample for Boiga
irregularis from throughout its natural range con-
sists of approximately 32% lizards, 21% birds,
and 44% mammals, but most individuals feed
entirely on lizards as juveniles and on either
mainly rodents or mainly birds as adults.
(2) The preceding analysis assumes that diets
really are attributes of populations and taxa, and
202
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 8
that they have evolved conservatively (i.e., with
minimal homoplasy). High rates of reversals and
convergence can only be detected by detailed
phylogenetic analysis, so there is as yet no basis
for postulating this alternative for Boiga. The
other possibility, that individual dietary plastic-
ity accounts fully for geographic and systematic
patterns, is discussed above (see INDIVIDUAL AND
GEOGRAPHIC DIETARY VARIATION).
(3) There is no evidence that the venom ap-
paratus of Boiga is functionally associated with
feeding on prey that are especially heavy or oth-
erwise costly, although rare ingestion of large prey
remains a possibility (see B. dendrophila, above,
and Greene 1986a).
(4) Rear fangs arose at the level of Boiga or,
more likely, a more inclusive group (Rasmussen
1 979, 1 985). Whatever the ecological role of ven-
om injection (if any) in extant cat snakes, its
possible adaptive origin (if any) in the clade con-
taining Boiga cannot be identified until addi-
tional information on suprageneric relationships
and natural history are available (see Cadle 1983;
Greene 1986a; Schaefer and Lauder 1986). To
the extent those limitations apply to other clades,
conclusions about the adaptive significance of
rear-fanged dentition are premature.
ECOMORPHOLOGICAL SUITES AND CONVERGENT
ANALOGUES.— Qualitative ecomorphological an-
alogs have been identified among certain geo-
graphically and phylogenetically distant snakes
(e.g., Shine 1980a, b, c; Henderson and Binder
1980), and suggestive parallels can be drawn be-
tween species of Boiga and certain other taxa.
Some resemblance to Afro-Asian species of Te-
lescopus might be either convergent or homol-
ogous (Rasmussen 1979, 1985); the two genera
are largely allopatric. Rear-fanged colubrids of
the Malagasy genus Lycodryas (Rax worthy 1988)
and neotropical genus Imantodes (especially the
I. cenchoa species group, Myers 1982) are ar-
boreal, nocturnal, and resemble small species of
Boiga in general body form (e.g., large head and
eyes, slender neck) and diet. The New World,
rear-fanged genus Trimorphodon is said to be
related to Boiga (Dowling et al. 1983) or to sev-
eral genera of New World colubrines (Minton
and Salanitro 1972; Cadle 1988; R. Lawson, CAS,
pers. comm.). Both species of Trimorphodon are
arboreal, nocturnal, and feed largely on sleeping,
diurnal iguanid lizards; only certain populations
of T. biscutatus occasionally include birds and
bats in the diet (Greene, unpubl. data). A non-
venomous, diurnal neotropical colubrid, Pseustes
poecilonotus, resembles B. dendrophila in size,
diet, and defensive behavior; however, even ju-
venile P. poecilonotus eat small birds (Greene
1979, and unpubl. data). I detect no particular
convergent resemblance, beyond arboreality, ir-
rascibility (Greene 1988b), and nocturnality, be-
tween Boiga and Australian elapids of the genus
Hoplocephalus (Shine 1983).
CAT SNAKES AS INTRODUCED PREDATORS.—
This study provides potentially useful insights
with regard to the threat that introduced cat
snakes pose for island prey faunas:
(1) Heavy predation on birds by Boiga irre-
gularis on Guam probably reflects the primitive
diet for this species (see above), rather than a
feeding shift in response to special local condi-
tions. It is not surprising that this snake makes
heavy inroads on island bird populations, es-
pecially if acceptable alternative prey (e.g., ro-
dents, see above) are unavailable or rare, and/or
if the native avifauna has evolved behavioral and
demographic attributes in the absence of pred-
ators (see Jackson [1974] for a bird with behav-
ioral defenses against nest predation by snakes).
The effects of an introduced snake on native is-
land prey populations might be all the more rapid
in the absence of natural predators on snakes
(Greene 1988a).
(2) Neonate Boiga irregularis are evidently too
small to eat endothermic prey, and starvation of
juveniles is probably a major source of natural
mortality in some snake populations (e.g., Fitch
1 949). Potential control measures for introduced
Boiga might include manipulating the availabil-
ity or characteristics (e.g., artificial toxicity) of
lizard prey.
(3) Boiga irregularis is a moderately large snake
with a venom delivery system and constricting
behavior that could permit it to subdue a wide
range of prey sizes and types. Like several other
species of Boiga, it possesses a relatively wide
skull and long dentaries (Marx and Rabb 1972),
derived characteristics that are likely associated
with increased gape (Rasmussen 1979; Arnold
1983), thereby facilitating a diet of birds, eggs,
and bats (Cundall and Greene, unpubl. data).
Although not so highly specialized as to be un-
able to forage on the ground, evidently all species
of Boiga climb well and are nocturnal.
Each of the above attributes of cat snakes might
GREENE: OLD WORLD CAT SNAKES
203
enhance their ability to feed on sleeping diurnal
birds and their eggs, or on roosting and foraging
bats. For those reasons and/or because of known
predation on birds and bats, all moderate to large
Boiga are potentially catastrophic invaders of
isolated island faunas. Species for special concern
include B. blandingi, B. cyanea, B. cynodon, B.
dendrophila, B. drapeizii, B. forsteni, B. irregu-
laris, B. jaspidea, B. kraepelini, B. nigriceps, B.
ocellata, B. pulverulenta, B. schnitzel, and B. tri-
gonata. There is a clear potential for additional
introductions: recently individual B. irregularis
have reached several Pacific islands (Fritts 1 988),
at least one B. dendrophila landed on Taiwan in
a shipment of lumber (A. H. T. Yu, pers. comm.),
and several species of large Boiga are commonly
available in the pet trade (pers. obs.).
EPILOGUE
I have never seen a Boiga in the field, and this
research was made possible by the efforts of nu-
merous collectors, curators, authors of anec-
dotes, and sponsors of natural history publica-
tions. Few of those individuals intended their
specimens for studies on the evolution of feeding,
and even fewer imagined that their labors would
have implications for an urgent conservation
problem. Nevertheless, starting from scratch to
learn something comparable about the broader
context for avian predation by introduced cat
snakes on Guam would have required many years
of work. Therein surely lie some lessons (see also
Greene 1986b; Greene and Losos 1988).
ACKNOWLEDGMENTS
I thank P. Alberch, J. E. Cadle, R. C. Drewes,
R. F. Inger, A. E. Leviton, H. Marx, R. W.
McDiarmid, C. W. Myers, H. L. Snell, and J. W.
Wright for access to specimens; B. Beehler, W.
C. Brown, R. I. Crombie, J. M. Diamond, T. H.
Fritts, M. LeCroy, W. Z. Lidicker, C. Luke, E.
D. Pierson, R. Shine, A. H. T. Yu, R. Zusi, and
R. G. Zweifel for identifying prey items, supply-
ing data, or other help; A. E. Leviton for patiently
encouraging my interests in Asiatic snakes; J. E.
Cadle for critical comments on the manuscript;
and the U.S. Fish and Wildlife Service, National
Science Foundation (BSR 83-00346), and Field
Museum of Natural History (Karl P. Schmidt
Fund) for financial support.
LITERATURE CITED
ANANJEVA, N. B. AND N. L. ORLOV. 1982. Feeding behaviour
of snakes. Vertebrata Hungarica 21:25-31.
ARNOLD, E. N. 1981. Estimating phytogenies at low taxo-
nomic levels. Z. Zool. Syst. Evolutionforsch. 19:1-35.
ARNOLD, S. J. 1977. Polymorphism and geographic variation
in the feeding behavior of the garter snake Thamnophis ele-
gans. Science 197:676-678.
. 1983. Morphology, performance and fitness. Amer.
Zool. 23:347-361.
AUFFENBERG, W. 1980. The herpetofauna of Komodo, with
notes on adjacent areas. Bull. Florida State Mus. 25:39-156.
BARACH, J. P. 1952. Some observations on the mangrove
snake in captivity. Herpetologica 8:102-103.
BARRY, D. H. 1961. Correspondence: bat-eating snakes. Ni-
ger. Field 26:42-43.
BECHSTEIN, J. M. 1 802. Herrn De la Cepede's Naturgeschich-
te der Amphibien oder der eyerlegenden vierfussigen Thiere
und der Schlangen. Eine Fortsetzung von Buffon's Natur-
geschichte aus dem Franzosischen ubersetzt und mil An-
merkungen und Zusatsen versehen, Vol. 4. Industrie Comp-
toir, Weimar.
BEEHLER, B. M., T. K. PRATT, AND D. A. ZIMMERMAN. 1986.
Birds of New Guinea. Princeton Univ. Press, Princeton, New
Jersey.
BLYTH, E. 1860. Report of curator, zoological department.
J. Asiatic Soc. Bengal 29:87-1 15.
BOIE, F. 1827. Bemerkungen tiber der Familien und Gattun-
gen der Ophidier. Isis von Oken 20:506-566.
BOULENGER, G. A. 1894. Description of a new snake found
in Travancore by Mr. S. Dighton, Pirmaad. J. Bombay Nat.
Hist. Soc. 8:528.
BROADLEY, D. G. 1983. FitzSimons' snakes of southern Af-
rica. Delta Books, Johannesburg, South Africa.
BRONGERSMA, L. D. 1 934. Contributions to Indo-Australian
herpetology. Zool. Meded. Rijksmus. Nat. Hist. 17:161-251.
BURGER, W. L. 1975. A case of mild envenomation by the
mangrove snake, Boiga dendrophila. Snake 6:99-100.
CADLE, J. E. 1983. Problems and approaches in the inter-
pretation of the evolutionary history of venomous snakes.
Mem. Inst. Butantan 46:255-274.
. 1988. Phylogenetic relationships among advanced
snakes: a molecular perspective. Univ. California Publ. Zool.
119:1-77.
CAMPDEN-MAIN, S. M. 1970. A field guide to the snakes of
South Vietnam. Smithsonian Inst., Washington, D.C.
CHISZAR, D. A. 1989. Behavior of the brown tree snake (Boi-
ga irregularis): a study in applied comparative psychology.
In press in Contemporary issues in comparative psychology.
D. A. Dewsbury, ed. Sinauer Associates, Sunderland, Mas-
sachusetts.
CHISZAR, D., D. CARRILLO, P. RAND, AND J. CHISZAR. 1985.
Nocturnal activity in captive brown tree snakes (Boiga ir-
regularis). Bull. Maryland Herpetol. Soc. 21:115-118.
CHISZAR, D., K. KANDLER, R. LEE, AND H. M. SMITH. 1988.
Stimulus control of predatory attack in the brown tree snake
(Boiga irregularis). 2. Use of chemical cues during foraging.
Amphibia-Reptilia 9:77-88.
CONRY, P. J. 1 988. High nest predation by brown tree snakes
on Guam. Condor 90:478-482.
CORKHILL, N. L. 1932. The snakes of Iraq. J. Bombay Nat.
Hist. Soc. 35:552-572.
204
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 8
Cox, M. J. 1988. Serious effects from the bite of the red cat-
eye snake, Boiga nigriceps. Bull. Chicago Herp. Soc. 23:162.
COZENS, B. 1 96 1 . Correspondence: bat-eating snakes. Niger.
Field 26:44.
CRANBROOK, E. AND J. I. FURTADO. 1988. Freshwaters. Pp.
225-250 in Key environments: Malaysia. E. Cranbrook, ed.
Pergamon Press, Oxford.
DERANIYAGALA, P. E. P. 1955. A colored atlas of some ver-
tebrates from Ceylon, Vol. 3. Serpentoid Reptilia. Govt.
Press, Colombo.
DE SILVA, P. H. D. H. 1980. Snake fauna of Sri Lanka, with
special reference to skull, dentition and venom in snakes.
Dept. Govt. Printing, Colombo, Sri Lanka.
DOWLING, H. G., R. HIGHTON, G. C. MAHA, AND L. R. MAXSON.
1983. Biochemical evaluation of colubrid snake phylogeny.
J.Zool. 201:309-329.
DRING, J. C. M. 1979. Amphibians and reptiles from north-
ern Trengganu, Malaysia, with descriptions of two new geck-
os: CnemaspisanA Cyrtodactylus. Bull. Brit. Mus. (Nat. Hist.)
34:181-241.
DUMERIL, A. M. C. AND G. BiBRON. 1854. Erpetologie gene-
rale ou histoire naturelle complete des reptiles. Paris.
DUNGER, G. T. 1961. Correspondence: bird-eating snakes.
Niger. Field 26:43.
FISCHER, J. G. 1856. Neue Schlangen des Hamburgischen
Naturhistorischen Museums. Abh. Naturwiss. Ver. Ham-
burg 3:81.
FITCH, H. S. 1949. Study of snake populations in central
California. Amer. Midi. Nat. 41:1-150.
FLOWER, S. S. 1 899. Notes on a collection of reptiles made
in the Malay Peninsula and Siam from November 1896 to
September 1898, with a list of species recorded from those
countries. Proc. Zool. Soc. London 1899:600-696.
FRITH, C. B. 1977. A survey of the snakes of Phulet Island
and the adjacent mainland areas of peninsular Thailand.
Nat. Hist. Bull. Siam Soc. 26:263-316.
FRITTS, T. H. 1988. The brown tree snake, Boiga irregularis,
a threat to Pacific islands. U.S. Fish Wild. Serv. Biol. Rept.
88:1-36.
FRITTS, T. H., N. J. SCOTT, JR., AND J. A. SAVIDGE. 1987.
Activity of the arboreal brown tree snake (Boiga irregularis)
on Guam determined by electrical outages. Snake 19:5 1-58.
GODLEY, J. S. 1980. Foraging ecology of the striped swamp
snake, Regina alleni, in southern Florida. Ecol. Monogr. 50:
411-436.
GOODMAN, J. D. 1985. Two record size Blanding's tree snakes
from Uganda. East Afr. Nat. Hist. Soc. Bull. 1985:56-57.
GRAY, J. E. 1834. Illustrations of Indian zoology chiefly se-
lected from the collection of Maj.-Gen. Hardwicke, Vol. 2.
London.
GREENE, H. W. 1976. Scale overlap, a directional sign stim-
ulus for prey ingestion by ophiophagous snakes. Z. Tierpsy-
chol. 41:113-120.
. 1977. Phylogeny, convergence, and snake behavior.
Ph.D. Dissertation, Univ. Tennessee, Knoxville.
-. 1979. Behavioral convergence in the defensive dis-
. 1 986a. Diet and arboreality in the emerald monitor,
Varanus prasinus, with comments on the study of adapta-
tion. Fieldiana, Zool. (new ser.) 31:1-12.
-. 1 986b. Natural history and evolutionary biology. Pp.
plays of snakes. Experientia 35:747-748.
-. 1983. Dietary correlates of the origin and radiation
99-108 in Predator-prey relationships: perspectives and ap-
proaches from the study of lower vertebrates. M. E. Feder
and G. V. Lauder, eds. Univ. Chicago Press, Chicago.
. 1988a. Species richness in tropical predators. Pp.
259-280 in Tropical rainforests: diversity and conservation.
F. Almeda and C. M. Pringle, eds. Cal. Acad. Sci., San Fran-
cisco.
. 1988b. Antipredator mechanisms in reptiles. Pp. 1-
152 in Biology of the Reptilia, Vol. 16. Ecology B. Defense
and life history. C. Gans and R. B. Huey, eds. Alan R. Liss,
Inc., New York.
. 1989. Defensive behavior and feeding biology of the
Asian mock viper, Psammodynastes pulverulentus (Colu-
bridae), a specialized predator on scincid lizards. Chin. Herp.
Res. 2:21-32.
GREENE, H. W. AND J. B. Losos. 1988. Systematics, natural
history, and conservation. BioScience 38:458-462.
GREENWELL, M. G., M. HALL, AND O. J. SEXTON. 1984. Phe-
notypic basis for a feeding change in an insular population
of garter snakes. Devi. Psychobiol. 17:457-463.
GRIFFIN, L. E. 1910. A list of snakes from the island of Polillo,
P. I., with descriptions of a new genus and two new species.
Philippine J. Sci. 5:211-215.
GROVES, F. 1973. Reproduction and venom in Blanding's
tree snake. Int. Zoo Yrbk. 13:106-108.
GUNTHER, A. 1858. Catalogue of colubrine snakes in the
collection of the British Museum. London.
. 1863. Third account of new species of snakes in the
collection of the British Museum. Ann. Mag. Nat. Hist., Ser.
3, 12:348-365.
. 1868. Sixth account of new species of snakes in the
of snakes. Amer. Zool. 23:431-441.
. 1984. Feeding behavior and diet of the eastern coral
snake, Micrurus fulvius. Spec. Publ. Mus. Nat. Hist. Univ.
Kansas 10:147-162.
collection of the British Museum. Ann. Mag. Nat. Hist., Ser.
4, 1:1-8.
HALLOWELL, E. 1857. Notice of a collection of reptiles from
the Gaboon Country, West Africa, recently presented to the
Academy of Natural Sciences of Philadelphia, by Dr. Henry
A. Ford. Proc. Acad. Nat. Sci. Philadelphia 1857:48-73.
HARRISON, T. 1962. Cat snake kills fruit bat. Malay. Nat. J.
16:153-154.
HEDGES, N. G. 1983. Reptiles and amphibians of East Africa.
Kenya Literature Bureau, Nairobi.
HEDIGER, H. 1933. Uber die von Herrn Dr. A. Buehler auf
der Admiralitats-Gruppe und einigen benachbarten Inseln
gesammelten Reptilien und Amphibien. Verh. Naturforsch.
Ges. Basil 44: 1-25.
HENDERSON, R. W. AND M. H. BINDER. 1 980. The ecology
and behavior of vine snakes (Ahaetulla, Oxybelis, Thelotor-
nis, Uromacef): a review. Contr. Biol. Geol. Milwaukee Publ.
Mus. 37:1-38.
HUTTON, A. F. 1949. Notes on the snakes and mammals of
the High Wavy Mountains, Madura District, S. India. J.
Bombay Nat. Hist. Soc. 48:454-460.
INGER, R. F. AND R. K. COLWELL. 1977. Organization of
contiguous communities of amphibians and reptiles in Thai-
land. Ecol. Monogr. 47:229-253.
INGER, R. F., H. B. SHAFFER, M. KOSHY, AND R. BAKDE. 1 984.
A report on a collection of amphibians and reptiles from the
Ponmudi, kerala, South India. J. Bombay Nat. Hist. Soc.
81:551-570.
GREENE: OLD WORLD CAT SNAKES
205
JACKSON, J. A. 1974. Gray rat snakes versus red-cockaded
woodpeckers: predator-prey adaptations. Auk 91:342-347.
JAYNE, B. C., H. K. VORIS, AND K. B. HEANG. 1988. Diet,
feeding behavior, growth, and numbers of a population of
Cerberus rhynchops (Serpentes: Homalopsinae) in Malaysia.
Fieldiana, Zool. (new ser.) 50:1-15.
JONES, T. 1961. Notes on bat eating snakes. Niger. Field 26:
69-70.
KARDONG, K. V. 1979. "Protovipers" and the evolution of
snake fangs. Evolution 33:433-443.
. 1980. Evolutionary patterns in advanced snakes.
Amer. Zool. 20:269-282.
KLUGE, A. G. 1987. Cladistic relationships in the Gekko-
noidea (Squamata, Sauria). Misc. Publ. Mus. Zool. Univ.
Michigan 173:1-54.
KOPSTEIN, F. 1926. Reptilien von den Molukken und den
benachbarten Inseln. Zool. Mededel. 9:71-1 12.
KREBS, J. R., D. W. STEPHENS, AND W. J. SUTHERLAND. 1983.
Perspectives in optimal foraging. Pp. 165-216 in Perspec-
tives in ornithology: essays presented for the centennial of
the American Ornithologists' Union. A. H. Brush and G. A.
Clark, Jr., eds. Cambridge Univ. Press, Cambridge.
KROON, C. 1973. A new colubrid snake (Boiga) from south-
eastern Asia. Copeia 1973:580-586.
KUNTZ,R. E. 1963. Snakes of Taiwan. U.S. Naval Res. Unit
No. 2, Taipei.
LAURENT, R. F. 1956. Contribution a Pherpetologie de la
region des Grands Lacs de I'Afrique centrale. Ann. Mus.
Roy. Congo Belg. Ser. 8, 48:1-390.
LAURIE, E. M. O. AND J. E. HILL. 1954. List of land mammals
of New Guinea, Celebes and adjacent islands, 1758-1952.
Brit. Mus. (Nat. Hist.), London.
LEKAGUL, B. AND J. A. MCNEELY. 1977. Mammals of Thai-
land. Kurusapha Ladprao Press, Bangkok.
LEVINSON, S. R., M. H. EVANS, AND F. GROVES. 1976. A
neurotoxic component of the venom from Blanding's tree
snake (Boiga blandingi). Toxicon 14:307-312.
LEVITON, A. E. 1970. Contributions to a review of Philippine
snakes, XI. The snakes of the genus Boiga. Philippine J. Sci.
97:291-314.
LIM, B. L. 1955. Snakes collected near Kuala Lampur. Malay.
Nat. J. 9:122-125.
. 1956. The natural food of some Malayan snakes.
Malay. Nat. J. 10:139-143.
-. 1967. Snakes collected in Gunong Brinchang, Cam-
eron Highlands, Pahang. Malay. Nat. J. 20:121-127.
. 1973. Unusual predators of the mouse-deer. Malay.
Nat. J. 26:170.
LIM, B. L., J. P. J. WATERS, AND A. R. BEN OMAR. 1960.
Observations on some captive snakes. Malay. Nat. J. 14:
181-187.
LINDBERG, K. 1932. Snakes on the Barsi Light Railway (Dec-
can). J. Bombay Nat. Hist. Soc. 35:690-697.
LLOYD, M., R. F. INGER, AND F. W. KING. 1968. On the
diversity of reptile and amphibian species in a Bornean rain
forest. Amer. Nat. 102:497-515.
Losos, J. B. AND H. W. GREENE. 1988. Ecological and evo-
lutionary implications of diet in monitor lizards. Biol. J.
Linn. Soc. 35:379^07.
MACARTHUR, R. H. AND E. R. PiANKA. 1 966. On the optimal
use of a patchy environment. Amer. Nat. 100:603-609.
MACIAS GARCIA, C. AND H. DRUMMOND. 1988. Seasonal and
ontogenetic variation in the diet of the Mexican garter snake,
Thamnophis eques, in Lake Tecomulco, Hidalgo. J. Her-
petol. 22:129-134.
MALNATE, E. V. AND G. UNDERWOOD. 1988. Australasian
natricine snakes of the genus Tropidonophis. Proc. Acad.
Nat. Sci. Philadelphia 140:59-201.
MARX, H. AND G. B. RABB. 1972. Phyletic analysis of fifty
characters of advanced snakes. Fieldiana, Zool. 63:1-321.
MCDOWELL, S. B. 1 969. Toxicocalamus, a New Guinea genus
of snakes of the family Elapidae. J. Zool. 159:443-511.
. 1972. The species of Stegonotus (Serpentes, Colu-
bridae) in Papua New Guinea. Zool. Meded. 47:6-26.
-. 1975. A catalogue of the snakes of New Guinea and
the Solomons, with special reference to those in the Bernice
P. Bishop Museum. Part II. Anilioidea and Pythoninae. J.
Herp. 9:1-79.
. 1984. Results of the Archbold expeditions. No. 1 12.
The snakes of the Huon Peninsula, Papua New Guinea.
Amer. Mus. Novitates 2775:1-28.
MENZIES, J. I. 1961. Correspondence: bat-eating snakes. Ni-
ger. Field 26:44.
MENZIES, J. I. AND E. DENNIS. 1 979. Handbook of New Guinea
rodents. Wau Ecol. Inst. Handbk. (7).
MERREM, B. 1 790. Bietrage zur Naturgeschichte. Priv. Publ.,
Duisburg and Lemgo.
MINTON, S. A. 1966. A contribution to the herpetology of
West Pakistan. Bull. Amer. Mus. Nat. Hist. 134:27-184.
MINTON, S. A. AND W. A. DUNSON. 1978. Observations on
the Palawan mangrove snake, Boiga dendrophila multidncta
(Reptilia, Serpentes, Colubridae). J. Herp. 12:107-108.
MINTON, S. A. AND S. K. SALANITRO. 1972. Serological re-
lationships among some colubrid snakes. Copeia 1 972:246-
252.
MURPHY, J. B. 1977. An unusual method of immobilizing
prey by the dog-tooth cat snake, Boiga cynodon. Copeia
1977:182-184.
MUSHINSKY, H. R. 1987. Foraging ecology. Pp. 302-334 in
Snakes: ecology and evolutionary biology. R. A. Seigel, J. T.
Collins, and S. S. Novak, eds. Macmillan Publ. Co., New
York.
MYERS, C. W. 1 982. Blunt-headed vine snakes (Imantodes)
in Panama, including a new species and other revisionary
notes. Amer. Mus. Novitates 2738:1-50.
PARKER, F. 1 982. The snakes of Western Province. Div.
Wildlife, Papua New Guinea.
PENDLETON, R. C. 1947. A snake "den" tree on Guadalcanal
Island. Herpetologica 3:189-190.
PETERS, W. 1861. Eine zweite Ubersicht der von Hrn. F.
Jagor auf Malacca, Java, Borneo und den Philippinen gesam-
melten unde dem Kgl. Zoologischen Museum uberstandten
Schlangen. Monatsb. Akad. Wiss. Berlin 1861:683-691.
PITMAN, C. R. S. 1958. Snake and lizard predators on birds.
Bull. Brit. Ornithol. Club 78:82-86.
. 1 962. More snake and lizard predators of birds. Bull.
Brit. Ornithol. Club 82:45-55.
. 1974. A guide to the snakes of Uganda. Wheldon
and Welsey, Ltd., Codicote.
POPE, C. H. 1935. The reptiles of China. Amer. Mus. Nat.
Hist., New York.
RASMUSSEN, J. B. 1979. An intergeneric analysis of some
boigine snakes— Bogert's ( 1 940) Group XIII and XIV (Boigi-
nae, Serpentes). Dansk. Naturhist. For. Vidensk. Meddel.
141:97-155.
. 1985. A re-evaluation of the systematics of the Af-
206
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 8
rican rear-fanged snakes of Bogert's Groups XIII-XVI, in-
cluding a discussion of some evolutionary trends within
Caenophidia. Pp. 531-548 in Proceedings of the Interna-
tional Symposium on African vertebrates: systematics, phy-
logeny and evolutionary ecology. K.-L. Schuchmann, ed.
Zool. Forsch. Mus. Alexander Koenig, Bonn.
RAXWORTHY, C. J. 1988. Reptiles, rainforest and conserva-
tion in Madagascar. Biol. Conserv. 43:181-211.
Rootr, N. 1917. The reptiles of the Indo-Australian Archi-
pelago. E. J. Brill Publ., Leiden.
ROOM, P. M. 1974. Lizards and snakes from the Northern
District of Papua, New Guinea. Brit. J. Herpetol. 5:38-46.
SAINT GIRONS, H. 1972. Les serpents du Cambodge. Mem.
Mus. Nat. Hist. Natur., nouv. ser., 74:1-170.
SAVIDOE, J. A. 1 984. Guam: paradise lost for wildlife. Biol.
Conserv. 30:305-317.
. 1987. Extinction of an island forest avifauna by an
introduced snake. Ecology 68:660-668.
. 1988. Food habits of Boiga irregularis, an introduced
predator on Guam. J. Herp. 22:275-282.
SAVITZKY, A. H. 1980. The role of venom delivery strategies
in snake evolution. Evolution 34:1 194-1204.
SCHAEFER, S. A. AND G. V. LAUDER. 1 986. Historical trans-
formation of functional design: evolutionary morphology of
feeding mechanisms in loricarioid catfishes. Syst. Zool. 35:
489-508.
SCHAETTI, B. 1 984. Fledermause als Nahrung von Schlangen.
Bonn. Zool. Beitr. 35:335-342.
SCHMIDT, K. P. 1927. The reptiles of Hainan. Bull. Amer.
Mus. Nat. Hist. 54:395^65.
SEIB, R. L. 1 984. Prey use in three syntopic neotropical racers.
J. Herpetol. 18:412-420.
. 1985. Euryphagy in a tropical snake, Coniophanes
fissidens. Biotropica 17:57-64.
SHARMA, R. C. AND T. G. VAZIRANI. 1977. Food and feeding
habits of some reptiles of Rajasthan. Rec. Zool. Survey India
73:77-93.
SHINE, R. 1 980a. Ecology of the Australian death adder Acan-
thophis antarcticus (Elapidae): evidence for convergence with
the Viperidae. Herpetologica 36:281-289.
. 1 980b. Reproduction, feeding and growth in the Aus-
tralian burrowing snake Vermicella annulata. J. Herpetol.
14:71-77.
. 1 980c. Ecology of eastern Australian whipsnakes of
the genus Demansia. J. Herpetol. 14:381-389.
. 1983. Arboreality in snakes: ecology of the Australian
elapid genus Hoplocephalus. Copeia 1983:198-205.
SHINE, R. AND T. SCHWANER. 1 985. Prey constriction be ven-
omous snakes: a review, and new data on Australian species.
Copeia 1985:1067-1071.
SMITH, B. E. 1985. A systematic survey of the squamate
reptiles of eastern Mindanao, Philippine Islands, with notes
on ecomorphology. M.S. Thesis, Louisiana State Univ. 78
pp.
SMITH, M. A. 1916. On a collection of reptiles and batra-
chians from peninsular Siam. J. Nat. Hist. Soc. Siam 2: MS-
HI.
. 1930. The Reptilia and Amphibia of the Malay Pen-
insula. Bull. Raffles Mus. 3:1-149.
. 1943. The fauna of British India, Ceylon, and Burma,
including the whole of the Indo-Chinese sub-region, Reptilia
and Amphibia, Vol. III. Serpentes. Taylor and Francis, Lon-
don.
SMITH, M. AND A. o'A. BELLAIRS. 1947. The head glands of
snakes, with remarks on the evolution of the parotid glands
and teeth of the Opisthoglypha. Zool. J. Linn. Soc. 41:351-
370.
SMITH, M. A. AND C. B. KLOSS. 1915. On reptiles and ba-
trachians from the coast and islands of south-east Siam. J.
Nat. Hist. Soc. Siam 1:237-249.
STEJNEGER, L. 1 902. A new opisthoglyp snake from Formosa.
Proc. Biol. Soc. Washington 15:1 15-1 17.
SUNDERSINGH, P. P. H. 1 960. Reptilian fauna of Pilani with
a note on the food contents. Part I— Ophidia. Proc. Rajas-
than Acad. Sci. 7:52-59.
TAYLOR, E. H. 1922a. Herpetological fauna of Mount Maki-
ling. Philippine Agricul. 11:127-139.
. 1922b. The snakes of the Philippine Islands. Phil-
ippine Bur. Sci. Publ. 16:1-312.
. 1923. Additions to the herpetological fauna of the
Philippine Islands, III. Philippine J. Sci. 22:515-557.
. 1953. Report on a collection of Ceylonese serpents.
Univ. Kansas Sci. Bull. 35:1615-1624.
. 1 965. The serpents of Thailand and adjacent waters.
Univ. Kansas Sci. Bull. 45:609-1096.
TWEEDIE, M. W. F. 1954. Notes on Malayan reptiles No. 3.
Bull. Raffles Mus. 25:107-1 17.
. 1983. The snakes of Malaya, 3rd ed. Singapore Nat.
Printers, Singapore.
VORIS, H. K. AND M. W. MOFFETT. 1981. Size and proportion
relationship between the beaked sea snake and its prey. Bio-
tropica 13:15-19.
VORIS, H. K. AND H. H. VORIS. 1983. Feeding strategies in
marine snakes: an analysis of evolutionary, morphological,
behavioral and ecological relationships. Amer. Zool. 23:41 1-
425.
WAKEMAN, B. N. 1966. Uganda's poisonous snakes: further
observations on feeding habits. Uganda J. 30:101-103.
WALL, F. 1907a. Notes on snakes collected in Fyzabad. J.
Bombay Nat. Hist. Soc. 18:101-129.
. 1907b. A popular treatise on the common Indian
snakes. Part 7. Echis carinatus, Boiga trigonatus. J. Bombay
Nat. Hist. Soc. 18:525-554.
. 1909. Notes on snakes from the neighbourhood of
Darjeeling. J. Bombay Nat. Hist. Soc. 19:337-357.
. 1910. Notes on snakes collected in upper Assam.
Part II. J. Bombay Nat. Hist. Soc. 19:825-845.
. 1919. Notes on a collection of snakes made in the
Nilgiri Hills and the adjacent Wynaad. J. Bombay Nat. Hist.
Soc. 26:552-584.
1921. Ophidia taprobanica, or the snakes of Ceylon.
Govt. Printer, Colombo.
. 1923. Notes on snakes collected on Annasigalla Es-
tate from August, 1920, to December, 1921. Spolia Zeylan-
ica 12:252-270.
. 1 924. Notes on Ceylon snakes collected by Mr. W.
W. A. Phillips. Spolia Zeylanica 13:71-88.
. 1926a. Notes on snakes collected in Burma in 1924.
J. Bombay Nat. Hist. Soc. 30:805-821.
. 1926b. Snakes collected in Burma in 1925. J. Bom-
bay Nat. Hist. Soc. 31:558-566.
WERNER, F. 1 909. Uber neue oder seltene Reptilien des Na-
turhistorischen Museums in Hamburg. Mitt. Naturhist. Mus.
Hamburg 26:205-247.
GREENE: OLD WORLD CAT SNAKES 207
WHITAKER, R. 1970. Slight reaction from bites of the rear- WOODWARD, S. F. 1960. A bat-eating snake. Niger. Field 25:
fanged snakes Boiga ceylonensis (Gunther) and Dryophis na- 1 72-1 74.
5M/M5 (Lacepede). J. Bombay Nat. Hist. Soc. 67:113. YOUNG, R. A. 1983. Feeding behavior of captive Boiga. Bull.
WILES, G. J. 1988. The status of fruit bats on Guam. Pac. Chicago Herp. Soc. 18:89-91.
Sci. 41:148-157.
CALIFORNIA ACADEMY OF SCIENCES
Golden Gate Park
San Francisco, California 94 1 1 8
PROCEEDINGS
OF THE
CALIFORNIA ACADEMY OF SCIENCES
Vol. 46, No. 9, pp. 209-220, 4 figs., 1 table.
December 20, 1989
NEW SPECIES AND TAXONOMIC NOTES ON MEXICAN AND
CENTRAL AMERICAN MELASTOMATACEAE
By
Frank Almeda
Department of Botany, California Academy of Sciences, Golden Gate Park,
San Francisco, California 94118-4599
ABSTRACT: Descriptions, diagnostic illustrations, and discussions are presented for four new species of
Miconia: M. grayumii from Costa Rica and Nicaragua; M. ibarrae from Mexico; M. iteophylla and M.peltata
from Panama. A reappraisal of the floral morphology ofLeandra consimilis necessitates its transfer to Miconia
and the adoption of a new name, M. ligulata. A chromosome number of n = 11 is reported for Pilocosta
oerstedii subsp. campanensis for the first time. To reflect the uniqueness and relationships of subsp. campanensis
it is elevated to specific status based on a reconsideration of morphological, chromosomal, and ecogeographic
attributes.
Received June 9, 1989. Accepted August 22, 1989.
INTRODUCTION
Recent field work in Central America has gen-
erated a rich reservoir of specimens upon which
to base a treatment of the Melastomataceae for
Flora Mesoamericana. In the course of studying
this material I continue to detect many new taxa,
significant range extensions, nomenclatural
problems in need of adjustment, and other in-
formation that enhances our understanding of
previously described taxa. This paper records
some of these discoveries and observations to
make the names available for use in other re-
gional checklists and floristic treatments of the
family.
Miconia grayumii Almeda, sp. nov.
(Fig. 1)
TYPE. -COSTA RICA. Heredia: Finca La Selva, Field Sta-
tion of the Organization for Tropical Studies on the Rio Puerto
Viejo E of its junction with the Rio Sarapiqui, elev. ca. 100
m, 22 Oct. 1982, McDowell 576 (holotype: CAS!; isotype:
DUKE).
Frutex vel arbuscula 1-5 m. Ramuli obscure quadrangulati
demum teretes sicut petioli foliorum subtus venae primariae
inflorescentiaque modice granuloso-furfuracei. Petioli (2-)6-
21 mm longi; lamina 5.5-12.7 x 1.2-4.3 cm anguste elliptica
apice anguste gradatimque acuminate basi late acuta vel ob-
tusa, 3-plinervata, membranacea et undulato-denticulata vel
integra. Inflorescentia 2.5-5 cm longa multiflora; flores 5-meri,
pedicellis (ad anthesim) 2-2.5 mm longis, bracteolis 0.5 x 0.25
mm persistentibus. Hypanthium (ad torum) 1 mm longum
sparsiuscule caduceque granuloso-f urfuraceum; lobis interior-
ibus 0.25 x 0.5 late deltoideis, dentibus exterioribus 0.25-0.5
mm eminentibus. Petala 1.5-2 x 0.5 mm obovato-oblonga
intus minute papillosa extus glabra. Stamina isomorphica gla-
bra; filamenta 1-1.5 mm longa; antherarum thecae 0.5-0.75
x 0.25 mm anguste oblongae, poro paulo dorsaliter inclinato;
connectivum nee prolongatum nee appendiculatum. Stylus 2.5-
3 mm glaber; ovarium 5-loculare et % inferum glabro.
Shrub or small trees 1-5 m tall. Older cauline
internodes subquadrate to terete, sparsely to
moderately ferrugineous scurfy-pulverulent to
glabrate. Distal branchlets, petioles, elevated pri-
[209]
210
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 9
FIGURE 1. Miconia grayumii Almeda. A, habit, x ca. %; B, representative leaf (lower surface), x 1; C, flower (at anthesis)
with all petals and stamens removed, x 1 5; D, petal (adaxial surface), x ca. 22; E, stamens, dorsal view (left) and lateral view
(right), x30; F, mature berry, x ca. 5; G, seeds, x ca. 15. (A from Folsom 8914; B from Smith 152; C-F from the holotype;
G from Hammel 12011.)
mary leaf veins below, and inflorescences mod-
erately scurfy-pulverulent. Leaves of a pair
somewhat unequal in size; petioles (2-)6-2 1 mm
long; blades membranaceous, 5.5-12.7 cm long
and 1.2-4.3 cm wide, narrowly elliptic, apex at-
tenuate to gradually acuminate, base broadly
acute to obtuse and typically inaequilateral but
not decurrent on the petiole, margin undulate-
denticulate to entire, 3-plinerved, the innermost
pair of elevated primaries diverging from the me-
dian nerve in opposite or irregularly alternate
fashion at successive points above the blade base,
ALMEDA: MEXICAN AND CENTRAL AMERICAN MELASTOMATACEAE
211
the transverse secondaries prominulous and
spaced 3-5 mm apart at the widest portion of
the blade, glabrous above at maturity, sparingly
scurfy-pulverulent and/or obscurely glandular-
pulverulent to glabrate below. Inflorescence a
paniculiform dichasium 2.5-5 cm long divari-
cately branched from the base, terminal but
sometimes becoming pseudolateral because of
axillary bud elongation; bracts of rachis nodes
persistent, lance-triangular, 0.5-3 mm long, 0.25
mm wide, glabrous above, sparsely scurfy-pul-
verulent to glabrate below; bracteoles sessile and
persistent, paired and fused into a short incon-
spicuous nodal collar or elevated interpetiolar
ridge, lance-triangular to subulate, 0.5 mm long,
0.25 mm wide, glabrous above and sparingly
scurfy-pulverulent to glabrous below, margin en-
tire. Pedicels 2-2.5 mm long, sparingly and de-
ciduously scurfy-pulverulent to glabrate. Hypan-
thium (at anthesis) globose, 1 mm long to the
torus, sparsely and deciduously scurfy-pulveru-
lent. Calyx lobes 5, rounded-deltoid, hyaline and
glabrous, 0.25 mm long, ca. 0.5 mm wide basally
between sinuses, the margins entire; exterior ca-
lyx teeth 5, subulate, 0.25-0.5 mm long, exceed-
ing and obscuring the lobes, essentially glabrous
throughout; torus glabrous adaxially. Petals 5,
obscurely papillose adaxially, otherwise gla-
brous, white, obovate-oblong, rounded apically,
1.5-2 mm long, 0.5 mm wide distally. Stamens
10, isomorphic; filaments glabrous, complanate,
1-1.5 mm long; anthers 0.5-0.75 mm long, 0.25
mm wide, yellow, oblong, rounded to truncate
at the apex with a retuse to somewhat dorsally
inclined terminal pore; connective thickened
dorsally but not prolonged or appendaged at the
filament insertion. Ovary (at anthesis) ca. % in-
ferior, 5-celled, globose, glabrous apically. Style
erect and conspicuously curved distally, gla-
brous, 2.5-3 mm long; stigma punctiform to
truncate. Berry blue-black or purple at maturity,
globose, 3-5 mm long and 3-5 mm in diameter.
Seeds angulate-pyramidate, 0.5-1 mm long, beige,
smooth with obscure but distinct verruculose an-
gles on the convex face.
PHENOLOGY.— Flowering and fruiting occur
sporadically from February through November.
DISTRIBUTION.— Rain forests and stream banks
in the Caribbean lowlands of Costa Rica and
Nicaragua at 40-180 m.
ADDITIONAL SPECIMENS EXAMINED.— COSTA RICA. Here-
dia: Finca La Selva, OTS Field Station on Rio Puerto Viejo,
trail at perimeter of successional plots, 13 Feb. 1981, Folsom
8914 (CAS); Finca La Selva, OTS Field Station, West Bound-
ary, 800-1,200 m line, 19 Feb. 1981, Folsom 9041 (CAS); W
of San Jose on Rio Sarapiqui, across Rio Sardinal at edge of
LomasSardinal, 4 Feb. 1983, Garwoodetal. 1070 (EM); Finca
La Selva, Loop Trail, near SW Trail, 15 Jul. 1979, Grayum
1834 (CAS); Finca La Selva, East Boundary Trail, 2,300 m
line, 19 Jul. 1979, Grayum & Sperry 1854 (CAS); Finca La
Selva, forested knoll E of Boundary Trail on way to Q. Ar-
boleda swamp, 14 Mar. 1980, Hammel8076 (CAS); Finca La
Selva, bank along Rio Puerto Viejo, 550 S x 1,500 E m grid,
23 Jun. 1980, Hammel 9103 (CAS); Finca La Selva, near Loop
Trail, 100 m S, 28 Aug. 1980, Hammel 9603 (CAS); Finca La
Selva, on ridge in plot HI, 4 May 1982, Hammel 12011 (CAS):
Finca La Selva, West Boundary, 1,200 m S, 28 May 1982,
Hammel 12576 (CAS); Finca La Selva, El Sura Trail, 1,200
m line, 25 Sep. 1982, McDowell 199 (CAS); Finca La Selva,
around 1,850 m E, 1,950 m S, 1 Sep. 1981, Smith 152 (CAS).
Limon: Cerro Coronel, E of Rio Zapote, ca. 1.5 km from Rio
Colorado, 10°40'N, 83°40'W, 13 Mar. 1987, Stevens etal. 24795
(CAS). NICARAGUA. Zelaya: Nueva Guinea, Colonia Yo-
laina, 13 Aug. 1982, Araquistain 3069 (CAS); Cano Costa Ri-
quita, 1.8 km SW of Colonia Naciones Unidas, 11°43'N,
84°18'W, 6-7 Nov. 1977, Stevens 4971 (CAS).
The first collections ofAficonia grayumii, made
just over a decade ago, were initially referred to
M. brenesii Standley because of similarities in
foliar morphology, indument type, and the di-
varicately branched dichasia. Critical study of
the improved material now available necessi-
tates the recognition of two species based on con-
sistent morphological and ecological differences.
All of the perceived differences between M. gra-
yumii and M. brenesii involve details of the flow-
ers and fruits. The latter has broadly deltoid calyx
teeth that do not greatly exceed or obscure the
calyx lobes, petals that are smooth on both sur-
faces, and anthers that are cuneate and broadest
distally with flaring dorsally inclined pores. This
broad pore and its thick connective that is pro-
longed for a short distance (0.25 mm) below the
thecae give the anthers of M brenesii an unusual
dorsally arcuate conformation when dry. Some
of the most readily observed differences between
M. grayumii and M. brenesii involve stylar and
seed characters. In M. brenesii the straight style
is terminated by a capitellate stigma and its seeds,
although pyramidate like M. grayumii, have
completely smooth polished angles. There are
also pronounced ecogeographic differences be-
tween these two species. In Costa Rica, M. bre-
nesii is largely confined to a more westerly dis-
tribution at higher elevations (900-1,600 m)
whereas M. grayumii occurs in the eastern low-
lands (40-180 m) that extend northward to Nic-
aragua.
212
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 9
Despite the many vegetative similarities be-
tween M. brenesii and M. grayumii, the latter
appears to be closest to M. ligulata Almeda (also
discussed below). Petal, stamen, and seed mor-
phology are identical in these two species. How-
ever, in M. ligulata the indument consists of pi-
noid hairs, the leaves are larger (12-28 x 4-10
cm) and decurrent on the petiole, the torus is
puberulent adaxially, the petals are glabrous
throughout, and the style is straight.
This species is named for Michael Grayum
who first collected this species at the La Selva
Field Station in 1979. His subsequent field work
in many other areas of Costa Rica has added
numerous interesting plants to the flora of the
region.
Miconia ibarrae Almeda, sp. nov.
(Fig. 2)
TYPE.— MEXICO. Veracruz: road to Lazaro Cardenas, 4 km
W of Estacion de Biologia Tropical Los Tuxtlas, 95°04' and
95°09'W, 18°36'N, municipio San Andres Tuxtla, elev. 400 m,
3 Jun. 1986, Sinica 738 (holotype: CAS!; isotype: MEXU).
Arbor parva 4-9 m. Ramuli quadrangulati demum teretes
sicut foliorum subtus venae primarieae inflorescentia hypan-
thiaque modice vel sparse stellulato-furfuracei demum glabra-
ti. Petioli 0.8-2.1 cm longi; lamina 5-11.5 x 2.5^t.l cm el-
liptica vel elliptico-lanceolata apice acuminata basi acuta vel
obtusa, 3(-5)-plinervata, membranacea et undulato-denticu-
lata. Panicula 5-10 cm longa multiflora; flores 5-meri, pedi-
cellis (ad anthesim) 0.5 mm longis, bracteolis 0.25-0.5 x 0.25
mm persistentibus. Hypanthium (ad torum) 2-2.5 mm Ion-
gum; calyx 0.5 mm longus truncatus vel vix undulatus, den-
tibus exterioribus ca. 0.25 mm eminentibus. Petala 2.5-3 x
1 .5 mm obovato-oblonga, extus et intus dense papillosa. Stam-
ina isomorphica glabra; filamenta 2 mm longa; antherarum
thecae 2-2.5 x 0.5 mm oblongae, poro ventraliter inclinato;
connectivum nee prolongatum nee appendiculatum. Stylus 3.5-
5 mm glaber; ovarium 5-6-loculare omnino inferum apice
glabro.
Trees 4-9 m tall. Uppermost branchlets, veg-
etative buds, and inflorescences sparsely to mod-
erately stellate-furfuraceous, the distal branchlet
nodes bearing prominent interpetiolar lines or
ridges. Leaves of a pair slightly unequal in size;
petioles 0.8-2.1 cm long; blades membrana-
ceous, 5-1 1.5 cm long and 2.5-4.1 cm wide, el-
liptic to elliptic-lanceolate, apex acuminate, base
acute to obtuse, margin undulate-denticulate, 3(-
5)-plinerved, the innermost pair of elevated pri-
maries diverging from the median nerve in op-
posite or alternate fashion above the blade base,
above glabrous at maturity, below essentially gla-
brous or sparingly and deciduously stellate-fur-
furaceous on the elevated primary nerves. Inflo-
rescence a terminal multiflowered panicle 5-10
cm long; bracts of the rachis nodes persistent,
narrowly lance-triangular, 1-1.5 mm long, 0.25
mm wide at the base, sparingly stellulate-furfura-
ceous to glabrate; bracteoles sessile and persis-
tent, subulate, 0.25-0.5 mm long, 0.25 mm wide,
sparingly stellulate-furfuraceous, margin entire.
Pedicels 0.5 mm long but lengthening to 1 mm
in fruit, sparingly stellate. Hypanthia (at anthe-
sis) campanulate, 2-2.5 mm long to the torus,
sparingly and deciduously stellulate-furfura-
ceous to glabrous. Calyx tube ca. 0.5 mm long,
truncate or obscurely undulate with 5 inconspic-
ous and early deciduous triangular calyx teeth
ca. 0.25 mm long that barely project beyond ca-
lyx rim; torus glabrous adaxially. Petals 5, con-
spicuously papillose on both surfaces, white, ob-
ovate-oblong, rounded to irregularly undulate
apically, 2.5-3 mm long, 1.5 mm wide distally.
Stamens 10, isomorphic; filaments glabrous,
complanate, 2 mm long; anthers 2-2.5 mm long,
ca. 0.5 mm wide, erect to slightly incurved api-
cally, yellow, linear-oblong, laterally compressed
and deeply channeled ventrally between thecae,
rounded at the apex with a ventrally inclined
terminal pore; connective thickened dorsally but
not prolonged or appendaged at the filament in-
sertion. Ovary (at anthesis) completely inferior,
5-6 -celled, globose to depressed-conic, the apex
fluted and glabrous but becoming rounded in
fruit. Style somewhat declinate and incurved dis-
tally, glabrous, 3.5-5 mm long; stigma puncti-
form to truncate. Berry globose, greenish-yellow
when plump and fleshy, 4-5 mm long and 5-5.5
mm in diameter. Seeds irregularly pyramidate,
1.5 mm long, beige, smooth and vernicose with
a rounded or bluntly angulate convex face.
PHENOLOGY.— Flowering and fruiting speci-
mens have been collected from April through
July and June through August, respectively.
DISTRIBUTION.— Rain forests of the munici-
pios of Catemaco and San Andres Tuxtla in Ve-
racruz, Mexico at 100-500 m.
ADDITIONAL SPECIMENS EXAMINED. — MEXICO. Veracruz:
Municipio Catemaco, 6 km SW of Sontecomapan, 1 2 May
1 972, Beaman 5953 (F, MEXU); Estacion de Biologia Tropical
Los Tuxtlas, Cerro Vigia, 21 May 1981, Gentry & Lott 32208
(CAS, MEXU, MO, US); Municipio San Andres Tuxtla, La-
guna Escondida, 2 km NE of Estacion de Biologia Tropical
Los Tuxtlas, 22 Jul. 1983, Ibarra 754 (CAS); Estacion de Bio-
logia Tropical Los Tuxtlas, 95°04' and 95°09'W, 18°34' and
18°36'N, Lot 67, 24 Jun. 1984, Ibarra & Cedilla 1792 (CAS,
TEX), Estacion de Biologia Tropical Los Tuxtlas, 95°04' and
95°09'W, 1S°34' and 18°36'N, Lot 67, 21 Aug. 1984, Ibarra
ALMEDA: MEXICAN AND CENTRAL AMERICAN MELASTOMATACEAE
213
FIGURE 2. Miconia ibarrae Almeda. A, habit, x %; B, representative leaf (lower surface), x %; C, representative flower with
petals and stamens removed (left) and floral bud (right), x ca. 7; D, petal (adaxial surface), x 9; E, stamens, lateral view (left)
and ventral view (right), x 8; F, mature berry, x 6; G, seeds, x ca. 6. (A-E from the holotype; F from Ibarra 754; G from Ibarra
& Sinica 1931.)
& Sinica 1931 (CAS); Laguna Escondida, 2.5 km NE of Es-
tacion de Biologia Tropical Los Tuxtlas, 95°04' and 95°09'W,
18°34' and 18°36'N, 23 Apr. 1985, Ibarra at al. 2407 (CAS).
This restricted species is defined by its sparse
cover of white stellate-furfuraceous hairs on the
lower leaf surfaces, prominent interpetiolar lines
or ridges on distal branchlets, undulate-dentic-
ulate leaves, truncate to obscurely undulate ca-
lyx, papillose petals, and irregularly pyramidate
seeds that are smooth and vernicose on the
rounded or bluntly angulate convex face. In leaf
shape, calyx development, stamen morphology,
214
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 9
and petal details, Miconia ibarrae is most similar
to M. fulvostellata L. O. Williams, a more wide-
spread species that also occurs in the Los Tuxtlas
region (Ibarra and Sinica 1987). In M. fulvostel-
lata, the branchlets, petioles, lower leaf surfaces,
inflorescences, and hypanthia (at anthesis) are
densely covered with a ferrugineous stellate in-
dument that conceals the surfaces of these struc-
tures. A superficial comparison of these two
species may give the impression that M. fulvo-
stellata is a densely pubescent extreme of M.
ibarrae. However, M. fulvostellata also differs
from the new species by a suite of other diag-
nostic characters: its style is straight, its stigma
is capitate, and its pyramidate seeds have pus-
ticulate angles on the convex face and a raphe
on the concave side invested with a soft hyaline
tissue that collapses irregularly when dry.
This species is named for Guillermo Ibarra M.,
coauthor of a useful checklist of the plants of the
Tropical Biological Station at Los Tuxtlas, Ve-
racruz. It was through his generosity that a fine
set of flowering and fruiting material was made
available for this study. In the published check-
list, M. ibarrae is listed as Miconia sp. (Ibarra
and Sinica 1987).
Miconia iteophylla Almeda, sp. nov.
(Fig. 3)
TYPE.— PANAMA. Code: along Rio San Juan below its
junction with Rio Tife, elev. 1,200 ft (366 m), 11 Jun. 1978,
Hammel 3393 (holotype: CAS!; isotype: MO).
Frutex 0.5-1.5 m. Ramuli primum obscure quadrangulati
demum teretes sicut laminarum subtus venae primariae pe-
tiolique modice granuloso-furfuracei. Petioli 2-1 1 mm longi;
lamina 4.5-9 x 0.6-1.7 cm anguste elliptica apice acuminata
basi acuta vel acuminata, 3-plinervata, membranacea et in-
tegra. Panicula 2-7 cm longa multiflora; flores 5-meri, pedi-
cellis (ad anthesim) 1-2 mm longis, bracteolis 0.5 x 0.25 mm
persistentibus. Hypanthium (ad torum) 1-1.5 mm longum sparse
caduceque granuloso-furfuraceum; calycis tubus 0.25 mm lon-
gus, lobis subtruncatus vel vix undulatus, dentibus exterioribus
subulatis 0.25 mm eminentibus. Petala glabra 2-2.5 x 0.75
mm oblongo-elliptica apice rotundato. Stamina isomorphica
glabra; filamenta 1-1.5 mm longa; antherarum thecae 1.25-
1.5 x 0.25 mm angustae oblongae, poro paulo dorsaliter in-
clinato; connectivum nee prolongatum nee appendiculatum.
Stylus 3.5 mm glaber; ovarium 5-loculare et % inferum apicem
versus sparse granuloso-puberulum.
Shrubs 0.5-1.5 m tall. Uppermost branchlets
subquadrate to terete, moderately covered with
a brown scurfy-pulverulent indument, the distal
branchlet nodes bearing interpetiolar lines or
ridges. Leaves of a pair somewhat unequal in
size; petioles 2-1 1 mm long; blades membrana-
ceous, 4.5-9 cm long and 0.6-1.7 cm wide, nar-
rowly elliptic, apex acuminate, base acute to acu-
minate, margin entire, 3-plinerved, the inner pair
of elevated primaries diverging from the median
nerve in opposite or subopposite fashion, the
transverse secondaries prominulous and spaced
l-2(-3) mm apart at the widest portion of the
blade, sparingly and deciduously scurfy-pulver-
ulent above when young, deciduously scurfy-pul-
verulent on the elevated primaries and higher
order venation below. Inflorescence a terminal
multiflowered panicle 2-7 cm long typically
branching 0.6-3 cm above the node initiating the
inflorescence, the rachis beset with brown scurfy
hairs; bracts of the rachis nodes sessile and per-
sistent, 1.5-3.5 mm long, 0.25-0.5 mm wide,
glabrous above, sparsely and deciduously scurfy-
pulverulent below; bracteoles sessile and persis-
tent, paired and fused laterally into a short nodal
collar or elevated ridge, linear-oblong to narrow-
ly and bluntly triangular, 0.5 mm long, ca. 0.25
mm wide, essentially glabrous. Pedicels 1-2 mm
long, moderately to densely scurfy-pulverulent.
Hypanthia (at anthesis) campanulate, 1-1.5 mm
long to the torus, deciduously scurfy-pulverulent.
Calyx tube 0.25 mm long, subtruncate, hyaline
and entire, the lobes barely discernible as 5 de-
pressed triangular undulations; exterior calyx
teeth 5, subulate, 0.25 mm long, exceeding and
obscuring the lobes, glabrous throughout; torus
fimbrillate-puberulent adaxially. Petals 5, gla-
brous, white, elliptic-oblong, rounded apically,
2-2.5 mm long, 0.75 mm wide. Stamens 10, iso-
morphic; filaments glabrous, complanate, 1-1.5
mm long; anthers 1.25-1.5 mm long, 0.25 mm
wide, yellow, linear-oblong, truncate, rounded or
emarginate at the apex with a somewhat dorsally
inclined terminal pore; connective thickened
dorsally but not prolonged or appendaged at the
filament insertion. Ovary (at anthesis) % inferior,
5 -celled, globose, apex fluted and sparingly glan-
dular puberulent but becoming rounded and gla-
brous in fruit. Style straight, glabrous, 3.5 mm
long; stigma truncate to capitellate. Berry glo-
bose, red turning purple-black when mature, 2-
4 mm long and 3-4 mm in diameter. Seeds ir-
regularly angulate-pyramidate, 0.5 mm long, rus-
ty brown, smooth, and vernicose with polished
angles on the convex face.
PHENOLOGY.— Flowering specimens have been
collected in June and July; fruiting collections
ALMEDA: MEXICAN AND CENTRAL AMERICAN MELASTOMATACEAE
215
B
FIGURE 3. Miconia iteophylla Almeda. A, habit, x ca. 1; B, representative leaf (lower surface), x ca. 1; C, flower (at anthesis)
with all petals and stamens removed, x 10; D, petal (adaxial surface), x 19; E, stamens, dorsal view (left) and lateral view (right),
x ca. 14; F, mature berry, x ca. 7; G, seeds, x20. (A-E from Hammel 3330; F from the holotype; G from Antonio 3669.)
have been made in February, June, and Decem-
ber.
DISTRIBUTION. —Rain forests on the Caribbean
slope of Panama in Bocas del Toro and Code
provinces where it typically grows near moving
water at 200-450 m.
ADDITIONAL SPECIMENS EXAMINED.— PANAMA. Bocas del
Toro: upper Rio San Pedro, 13 Jul. 1979, Gordon 59Db (MO).
Code: trail from Cano Susio to Cerro Tife on the Atlantic
slope, 3 Feb. 1980, Antonio 3669 (CAS); Atlantic slope NW
of El Cope, along Rio San Juan near fork with Rio Tife, 9 Jun.
1978, Hammel 3330 (CAS); Caribbean side of divide at El
Cope, 3 Feb. 1983, Hamilton & Davidse 2628 (CAS); between
216
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 9
Cano Susio and waterfall at base of Cerro Tife, 13 Dec. 1980,
Sytsma et al. 2541 (CAS).
Miconia iteophylla is distinguished by its brown
scurfy-pulverulent indument, narrow elliptic
leaves, linear-oblong petals, and pyramidate seeds
with smooth polished angles on the convex face.
It is related to a small group of species that in-
cludes M. ligulata and M. peltata, both of which
are also treated herein. These three species share
a scurfy-pulverulent indument, linear-oblong
petals, a deciduously glandular puberulent ovary
apex, and a torus that is puberulent adaxially.
The brown scurfy indument is superficially sim-
ilar in all of these taxa. In M. iteophylla, however,
it consists of branlike scales unlike the pinoid
hairs of M ligulata and M. peltata. The seeds of
M. iteophylla and M. peltata are identical in shape
and sculptural details which suggests that these
two species may be more closely related to one
another than either is to M. ligulata. In general
aspect M. iteophylla resembles M. ligulata, which
is readily differentiated by its larger (12-2 8 x 4-
10 cm) 5-plinerved leaf blades that are decurrent
on the petiole, and pyramidate seeds with ver-
ruculose angles on the convex face.
The epithet for this species is derived from the
Greek words itea, willow, and phyllon, leaf, in
reference to the narrow willowlike leaf blades.
This species is referred to locally as "ugua" (fide
Gordon 59Db, MO) by the Guaymi, the largest
Indian group in Panama (Gordon 1982).
Miconia ligulata Almeda, nom. nov.
Leandra consimilis Gleason, Ann. Missouri Bot. Gard. 45:268.
1958.
TYPE. -PANAMA. Las Minas, Allen 2702 (holotype: NY).
Non Miconia consimilis Pilger, Verh. Bot. Ver. Brand. 47: 1 70.
1 905. This latter entity is a taxonomic synonym ofGraffenrieda
gracilis (Triana) L. O. Williams (Williams 1 963; Wurdack 1 970).
In the protologue, Gleason described the petals
of this species as oblong-lanceolate. This is un-
derstandable in view of the fact that the distal
petal margins tend to become involute and su-
perficially appear acute when pressed and dried.
Traditionally berry-fruited species in the Mico-
nieae with terminal inflorescences and acute pet-
als have been placed in Leandra. The petals of
Leandra consimilis are actually linear-oblong and
rounded to obtuse apically like several described
species in Miconia. To rectify its taxonomic
placement, Leandra consimilis is here trans-
ferred to Miconia. This requires the adoption of
a new name because the epithet consimilis is al-
ready pre-empted in Miconia. When Gleason de-
scribed this species it was known from only the
Panamanian holotype. Wurdack (1978) reported
it from Costa Rica south to Venezuela and I have
recently recorded that its range extends north-
ward to Nicaragua (Almeda, in press). An ex-
panded description is provided here to reflect the
range of variation exhibited by the species
throughout its known range.
Shrubs or small trees (1.5-)4-9 m tall. Up-
permost branches rounded-quadrate to quadri-
sulcate, the young branchlets, petioles, primary
leaf veins below, and inflorescences moderately
to densely covered with a brown scurfy indument
of pinoid hairs. Leaves of a pair equal to some-
what unequal in size; petioles 0.7-2.5 cm long;
blades membranaceous, 1 2-28 cm long and 4-
1 0 cm wide, elliptic, apex attenuate to long-acu-
minate, base gradually tapering and decurrent on
the petiole, margin entire to undulate, 5-pli-
nerved, the innermost pair of elevated primaries
diverging from the median nerve in opposite or
alternate fashion, the transverse secondaries
prominulous and spaced 3-7 mm apart at the
widest portion of the blade, sparsely scurfy-pul-
verulent to glabrate above, moderately and de-
ciduously scurfy-pulverulent on the secondary
and higher order veins below. Inflorescence a
terminal multiflowered panicle 4-1 1.5 cm long
typically branching (0.5-)1. 5-3.5 cm above the
node initiating the inflorescence; bracts of the
rachis nodes persistent, 1-5 mm long, 0.25-1
mm wide, glabrous above, moderately covered
with scurfy-pulverulent or short-stalked pinoid
hairs below; bracteoles sessile and persistent,
paired and fused laterally to form a nodal ridge
or shallow collar, narrowly triangular, 0.5 mm
long, 0.25-0.5 mm wide, glabrous above and
sparingly scurfy-pulverulent to glabrate below.
Pedicels 0.5-1 mm long with an indument like
that of the bracteoles. Hypanthia (at anthesis)
campanulate, 1-1.5 mm long to the torus, de-
ciduously scurfy-pulverulent to stellulate-furfu-
raceous. Calyx tube 0.25 mm long, subtruncate,
hyaline, and entire or undulate with 5 rounded-
triangular lobes 0.25 mm long; exterior calyx teeth
5, subulate, 0.25 mm long, equaling or exceeding
and obscuring the lobes, glabrous throughout;
torus sparsely fimbrillate-puberulent or glandu-
lar puberulent adaxially. Petals 5, glabrous, white,
ALMEDA: MEXICAN AND CENTRAL AMERICAN MELASTOMATACEAE
217
linear-oblong, obtuse to rounded apically, 2-3.5
mm long, 0.75-1 mm wide. Stamens 10, iso-
morphic; filaments glabrous, complanate, 1 mm
long; anthers 0.75-1 .25 mm long, 0.25 mm wide,
yellow, linear-oblong, truncate to slightly emar-
ginate at the apex with a somewhat dorsally in-
clined terminal pore; connective thickened dor-
sally but not prolonged or appendaged at the
filament insertion. Ovary (at anthesis) % to %
inferior, 5 -celled, globose, apex lobed or fluted
and sparsely glandular puberulent but becoming
rounded and glabrous in fruit. Style straight, gla-
brous, 3-4 mm long; stigma truncate but not
conspicuously dilated. Berry globose, pink but
turning blue-purple when mature, 2-4 mm long
and 3-4 mm in diameter. Seeds angular-pyrami-
date, ca. 0.5 mm long, brown, smooth with ver-
ruculose angles on the convex face.
PHENOLOGY. — Flowering sporadically from
January through October, fruiting specimens have
been collected from August through January.
DISTRIBUTION.— Widespread in wet evergreen
forests on the Caribbean slope of Central Amer-
ica from eastern Nicaragua (Zelaya) south and
east through Costa Rica and Panama to north-
western Venezuela (Zulia) from sea level to 1 , 1 00
m.
REPRESENTATIVE SPECIMENS EXAMINED.— NICARAGUA.
Zelaya: SW flank of Cerro Hormiguero, 1 8 Apr. 1979, Grijalva
440 (CAS). COSTA RICA. Alajuela: between San Lorenzo and
Los Angeles de San Ramon, above Rio San Lorenzo, 10°14'N,
84°32'W, 20 Sep. 1978, Burger & Antonio 11176 (CAS). Here-
dia: road between Puerto Viejo and La Virgen, near Chilamate,
30 May 1 982, Hammel 12670 (CAS); Finca La Selva, the OTS
Field Station on the Rio Puerto Viejo just E of its junction
with the Rio Sarapiqui, 1 Aug. 1980, Hammel 9403 (CAS).
Limon: 2 airline km SSE of Islas Buena Vista in the Rio Col-
orado, 83°40'W, 10°40'N, 13-14Sep. 1 986, Davidse & Herrera
31064 (CAS); Hacienda Tapezco-Hda. La Suerte, 29 air km
W of Tortuguero, 10°30'N, 83°47'W, 22 Aug. 1979, Davidson
& Donahue 8665 (CAS); Cerro Coronel, E of Rio Zapote, along
and above new road within 1 km of Rio Colorado, 10°40'N,
83°40'W, 13-14 Sep. 1986, Stevens & Montiel 24356 (CAS).
PANAMA. Code: La Mesa above El Valle de Anton, N of
Cerro Gaital, 8°37'N, 80°06'W, 26 Jul. 1984, de Nevers et al.
3524 (CAS); foothills and summit of Cerro Caracoral, near La
Mesa and N of El Valle de Anton, 10 Sep. 1981, Knapp 1092
(CAS); Atlantic slope of the Continental Divide near sawmill
above El Cope, 8°40'N, 80°36'W, 1 3 Feb 1982, Knapp & Dres-
sier 3 506 (CAS); Rio Guanche, 9°30'N, 79°39'W, 15 Oct. 1980,
Sytsma 1666 (CAS). Comarca de San Bias: El Llano-Carti
Road, Continental Divide, 9°19'N, 78°55'W, 25 Aug. 1984, de
Nevers 3760 (CAS). Darien: Cerro Pirre valley between Pirre
and next most southerly peak, 10-20 Jul. 1977, Folsom 4429
(CAS). Panama: top of Cerro Campana, 8 Sep. 1982, D'Arcy
15092 (CAS); Pipeline road, N of Gamboa, 9°10'N, 79°45'W,
4 Aug. 1984, de Nevers & Hews 3635 (CAS). VENEZUELA.
Zulia: Cano Helena, Sierra Perija, Delascio & Benkowski 3197
(US.)
Miconia ligulata belongs to the group of three
species discussed above under M. iteophylla.
Among its close allies it is unique in having wide
leaf blades (4-10 cm) that are gradually tapered
at the base and decurrent on the petiole. The
verruculose angles on the convex seed face con-
stitute another distinctive but less conspicuous
feature of M. ligulata. The new name for this
species is derived from ligula, Latin for little
tongue or strap, in reference to the elongate strap-
like petals.
Miconia peltata Almeda, sp. nov.
(Fig. 4)
TYPE.— PANAMA. Panama: near Cerro Jefe, along road to-
wards Alto Pacora, forested slopes ca. 850 m, ca. 9°15'N,
79°30'W, 27 Dec. 1985, McPherson 7882 (holotype: CAS!;
isotype: MO).
Frutex vel arbor parva 3-6 m. Ramuli primum paulo com-
pressi demum teretes sicut petioli foliorum subtus venae pri-
mariae inflorescentiaque pilis pinoideis ca. 0.25 mm longis
plus minusve caducis dense induti. Petioli 1.3-5 cm longi;
lamina peltata 6. 7-14.5 x 3.3-8.3cmovatavelelliptico-ovata
apice acuminata basi rotundata vel subcordata, 5-7-nervata,
subcoriacea et undulato-denticulata vel subintegra. Inflores-
centia 3-4 cm longa multiflora, ramis oppositis divaricatis;
flores 5-meri, pedicellis (ad anthesim) 1 mm longum, pilis
plerumque breviuscule stipitato-stellatis vel stellulato-pinoide-
is praeditum. Calycis tubus 0.25 mm longus, lobis ca. 0.25
mm altis late triangularibus, dentibus exterioribus triangulari-
bus 0.25 mm eminentibus. Petala glabra 2.5 x 0.5 mm ob-
longa-elliptica apice rotundato vel rotundato-obtuso. Stamina
isomorphica glabra; filamenta 1 mm longa; antherarum thecae
0.75-1 x 0.25 mm anguste oblongae, poro paulo dorsaliter
inclinato; connectivum nee prolongatum nee appendiculatum.
Stylus 1.5 mm glaber; ovarium 5-loculare et % inferum apice
sparsiuscule glanduloso-puberulo.
Shrubs or small trees 3-6 m tall. Uppermost
branchlets compressed and two-sided but be-
coming rounded with age. Distal branchlets, pet-
ioles, primary leaf veins below, and inflores-
cences densely and somewhat deciduously
covered with a rusty brown indument of pinoid
hairs. Leaves of a pair somewhat unequal in size,
peltate with the petiole attached 5-8 mm from
the basal edge, petioles 1.3-5 cm long; blades
subcoriaceous, 6.7-14.5 cm long and 3.3-8.3 cm
wide, ovate to elliptic-ovate, apex acuminate, base
broadly rounded to subcordate, margin incon-
spicuously undulate-denticulate to subentire, 5-
7-nerved abaxially, the transverse secondaries
prominulous and 2-3 mm apart at the widest
218
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 9
FIGURE 4. Miconia peltata Almeda. A, habit, x ca. %; B, representative leaf (lower surface), x ca. %; C, petal (adaxial
surface), x 1 2; D, stamens, dorsal view (left) and lateral view (right), x ca. 1 7; E, mature berry and bracteoles, x 6; F, enlargement
of pinoid hairs on fruiting pedicel, x ca. 60; G, seeds, x 16. (A, B, E, F from Tyson 3440; C, D, G from the holotype.)
portion of the blade, glabrous above, the sec-
ondaries and higher order veins below sparingly
beset with spreading pinoid hairs and an incon-
spicuous appressed glandular puberulence. Inflo-
rescence a multiflowered paniculiform dicha-
sium 3-4 cm long divaricately branched from
the base; bracts of the rachis nodes sessile and
persistent, 1.5-4 mm long, 0.5 mm wide, sparse-
ly beset with pinoid hairs; bracteoles sessile and
persistent, subulate, 0.25-0.5 mm long, mostly
less than 0.25 mm wide, essentially glabrous
throughout. Pedicels 1 mm long. Hypanthia (at
anthesis) campanulate to subglobose, 1.5 mm
long to the torus, moderately to sparsely covered
ALMEDA: MEXICAN AND CENTRAL AMERICAN MELASTOMATACEAE
219
TABLE 1 . Comparison of chromosome numbers, pollen size, and ecogeographic ranges in Pilocosta.
Haploid
number
Pollen diameter
Ecogeographic range
P. campanensis
P. erythrophylla
P. nana
P. oerstedii
n = 11
n = 7, 14
n = 33
n= 18
16.5-1 8.8 urn
14.1-16.5 /tm
25.5-30.6 /tm
17.6-2 1.2 Mm
Central Panama at 300-1,000 m
Central Costa Rica at 1,480-2,000 m
Costa Rica through central Panama to Colombia (Magdalena)
and Ecuador (Pichincha) at 450-1,800 m
Central Costa Rica to western Panama at 900-2,000 m
with stipitate-stellate or short pinoid hairs. Calyx
tube 0.25 mm long, subtruncate, hyaline and en-
tire, with 5 triangular lobes ca. 0.25 mm long;
exterior calyx teeth 5, bluntly triangular, 0.25
mm long, glabrous throughout; torus inconspic-
uously nmbrillate puberulent adaxially. Petals 5,
glabrous, reddish-pink, oblong-elliptic, rounded
to obtuse apically, 2.5 mm long, 0.5 mm wide.
Stamens 10, isomorphic; filaments glabrous,
complanate, 1 mm long; anthers 0.75-1 mm long,
0.25 mm wide, pale yellow, oblong, rounded to
subtruncate at the apex with a somewhat dorsally
inclined terminal pore; connective thickened
dorsally but not prolonged or appendaged at the
filament insertion. Ovary (at anthesis) % inferior,
5-celled, globose, apex fluted and sparingly glan-
dular puberulent but rounded and glabrous in
fruit. Style straight, glabrous, 1.5 mm long; stig-
ma truncate. Berry globose, purple when mature,
2-2.5 mm long and 3-4 mm in diameter. Seeds
irregularly angulate-pyramidate, 0.5 mm long,
brown, smooth with polished angles on the con-
vex face.
PHENOLOGY.— Flowering and fruiting speci-
mens have been collected in October, December,
and February.
DISTRIBUTION.— Endemic to Cerro Jefe in cen-
tral Panama at 850-1,000 m.
ADDITIONAL SPECIMENS EXAMINED.— PANAMA. Panama:
Cerro Jefe, 7 Oct. 1986, Aranda & Valdespino 185 (CAS, MO,
PMA); E slope of Cerro Jefe, 8 Feb. 1966, Tyson 3440 (MO-
2 sheets).
Miconia peltata also belongs to the group of
three species discussed above under M. iteo-
phylla. Because of its ovate to elliptic-ovate pel-
tate leaves it is recognizable even in sterile con-
dition. In addition to its unusual leaves, M. peltata
differs from closely allied species in having red-
dish-pink petals and an inflorescence that is di-
varicately branched from the base. The epithet
peltata is derived from the Greek word, pelte,
small shield. This refers to the attachment of the
petioles to the lower foliar surface away from the
margin of the blade.
Pilocosta campanensis (Almeda & Whiffin)
Almeda, stat. nov.
Pilocosta oerstedii (Triana) Almeda & Whiffin subsp. campa-
nensis Almeda & Whiffin, Syst. Hot. 5(3):306. 1980 [1981].
TYPE.— PANAMA. Panama: Cerro Campana, 10 Dec. 1967,
Lewis et al. 3069 (holotype: MO!; isotypes: COL, DUKE!, F!,
K!, NY!, UC!).
When Trevor Whiffin and I described this as
a subspecies of Pilocosta oerstedii we based our
decision on similarities in habit and androecial
details (Almeda and Whiffin 1981). We also en-
tertained the possibility of describing it as a sub-
species of P. nana because of similarities in the
kind and distribution of hypanthial pubescence.
At the time it was the only taxon in the genus
for which we had no chromosome information.
Based on information then available we recog-
nized two subspecies under P. oerstedii because
of their non-overlapping geographic distribution
and elevational displacement. A recent study of
chromosome numbers in neotropical Melasto-
mataceae (Almeda and Chuang, in prep.) reveals
that this taxon has a haploid number of n = 11
(count based on field collected buds of Almeda
et al. 5880, CAS). Based on morphological, cy-
tological, and ecogeographic data, it is clear that
Pilocosta oerstedii subsp. campanensis has all the
attributes that characterize species in this genus
(Table 1). I therefore elevate it to species status.
An alternative disposition would be to make it
a subspecies of P. nana. This, however, would
make no sense in view of their pronounced mor-
phological and cytological differences and the fact
that I have found individuals of these two taxa
growing next to one another near La Mesa in
Code Province, Panama.
220
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 9
A chromosome number of n = 1 1 in P. cam-
panensis reconfirms our previous assertion re-
garding the complex cytological evolution within
this small natural genus. It also sheds new light
on the relationship of P. nana and provides some
clues on the evolutionary polarity of character
states judged to be of taxonomic value in Pilo-
costa. With n = 33, P. nana appears to be a
hexaploid derivative of P. campanensis. The
polyploid gain in P. nana has evidently been
accompanied by the derivation of an herbaceous
habit and subisomorphic androecium from the
woody habit and dimorphic androecium of P.
campanensis.
ACKNOWLEDGMENTS
Fieldwork for this study was supported, in part,
by U.S. National Science Foundation Grant BSR
8614880 (Flora Mesoamericana) and the In-
House Research Fund of the California Academy
of Sciences. For special assistance during the
course of this study I thank Gerrit Davidse, Barry
E. Hammel, Guillermo Ibarra M., and Gordon
McPherson. I am also grateful to the Missouri
Botanical Garden and the Smithsonian Tropical
Research Institute for logistical support in the
field; Carolyn Rendu for the illustrations; Tsan-
lang and Fei-mei Chuang for assistance with
chromosome counts; and the curators and staffs
of the following herbaria (acronyms fide Holm-
gren et al. 1981) for providing special loans, gifts,
and/or assistance during study visits: BM, BR,
CR, DUKE, F, K, MA, MEXU, MO, NY, P,
PMA, US.
LITERATURE CITED
ALMEDA, F. In press. Melastomataceae. In Flora de Nica-
ragua. W. D. Stevens, ed. Missouri Botanical Garden, St.
Louis.
ALMEDA, F. AND T. WHIFFIN. 1981. Pilocosta, a new genus
of tropical American Melastomataceae. Syst. Bot. 5(3):294-
311.
GORDON, B. L. 1982. A Panama forest and shore. The Box-
wood Press, Pacific Grove, California.
HOLMGREN P. K., W. KEUKEN, AND E. K. SCHOFIELD. 1981.
Index Herbariorum. Part 1, 7th ed. The herbaria of the world.
RegnumVeg. 106:l^t52.
IBARRA M., G. AND S. SINICA C. 1987. Listados floristicos
de Mexico. VII. Estacion de Biologia Tropical Los Tuxtlas,
Veracruz: 1-51. Universidad Nacional Autonoma de Mex-
ico, Mexico.
WILLIAMS, L. O. 1 963. Tropical American plants V. Fieldiana
Bot. 29(10):545-586.
WURDACK.J. J. 1970. Certamen Melastomataceis XV. Phy-
tologia 20:369-389.
. 1978. Suplemento a las Melastomaceas de Venezue-
la. Acta Bot. Venez. 13:125-172.
CALIFORNIA ACADEMY OF SCIENCES
Golden Gate Park
San Francisco, California 94 1 1 8
PROCEEDINGS
OF THE
CALIFORNIA ACADEMY OF SCIENCES
Vol. 46, No. 10, pp. 221-242, 14 figs. December 20, 1989
REVIEW OF THE EELPOUT GENUS
PACHYCARA ZUGMAYER, 1911 (TELEOSTEI: ZOARCIDAE),
WITH DESCRIPTIONS OF SIX NEW SPECIES
By
M. Eric Anderson
Department of Ichthyology, California Academy of Sciences, Golden Gate Park,
San Francisco, California 94118
ABSTRACT: The eelpout genus Pachycara Zugmayer, 1911 is revised to include 12 species, 6 of which are
reported here as new: P. shcherbachevi from the Indian Ocean; P. sulkai and P. crossacanthum from the
Atlantic; and P. mesoporum, P. pammelas, and P. rimae from the Pacific. Pachycara is a primitive lycodine
eelpout genus characterized by its large body size, small head pores, lack of mental crests, and low para sphenoid
wing. It is most closely related to Lycodes Reinhardt, 1832, and Thermarces Rosenblatt and Cohen, 1986.
The species of Pachycara are characterized by their lateral line patterns, variably present or absent pelvic
fins, counts of the axial skeleton, gill raker morphology, head pore patterns, and branchiostegal ray number.
The genus is distributed in the eastern Pacific, Antarctic, Atlantic and Indian oceans from the upper continental
slope to the abyss (200-4,780 m).
Received February 21, 1989. Accepted July 1, 1989.
, Considerable taxonomic confusion exists in the
literature on eelpouts, especially for the rare deep-
The eelpout genus Pachycara was erected by sea species. As was found earlier (Anderson
Zugmayer (191 la, b) for a species known from 1982a), comparison of recently collected speci-
a single specimen taken in the abyssal North At- mens of Pachycara with older, published mate-
lantic. Anderson (1988a) considered the Antarc- rial was difficult because of a lack of, or poor
tic genus Austrolycichthys Regan, 1 9 1 3 a junior state of, types and the use of non-diagnostic char-
synonym of Pachycara, broadening the diagnosis acters by previous authors. As a contribution to
chiefly by rejecting the presence or absence of the clarification of the systematics of Pachycara,
pelvic fins as a generic character in eelpouts (see I provide descriptions and/or diagnoses for all
also Anderson 1988b). Anderson and Peden species, list synonymies, provide a key, distri-
(1988) reported the genus from the North Pacific butional maps and illustrations to facilitate iden-
for the first time, further broadening its limits, tification, and compare Pachycara to its putative
and described two new species. They cited my sister genera.
unpublished dissertation (Anderson 1984) as a Prior to the postwar rejuvenation of deep-sea
source of information on the genus. This paper exploration, specimens here referred to Pachy-
is an update of that study, revised to include 1 2 cam were few in collections, and, in fact, con-
species, six of which are new. sisted only of the types of P. bulbiceps (Garman,
[221]
222
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 10
1 899) (=P. obesa Zugmayer, 1 9 1 1 ), P. suspectum
(Garman, 1 899), P. crassiceps (Roule, 1916), and
about 60 specimens of P. brachycephalum (Pap-
penheim, 1912). The more important recent col-
lections of deep-sea fishes that resulted in the
capture of Pachycara specimens are briefly dis-
cussed below.
Specimens from the Pacific Ocean were first
taken during the round-the-world expedition of
the Danish ship GALATHEA which collected
the rare P. suspectum in the Gulf of Panama in
1952. This expedition had in the previous year
collected the holotype of P. shcherbachevi in the
Bay of Bengal. The bulk of the North Pacific
species were taken by free vehicle set line traps
by personnel of the Scripps Institution of Ocean-
ography off Mexico and southern California, by
bottom trawls operated from ships of Oregon
State University, and by Canadian experimental
sablefish (Anoplopoma fimbrid) traps off the
Queen Charlotte Islands (summarized in An-
derson and Peden 1988). Specimens from the
southeastern Pacific have come from operations
of DSV ALVIN in the Galapagos Rift Zone (Co-
hen and Haedrich 1983), Chilean fisheries re-
connaissance, supplied to the author by Ismael
Kong, University of Antofagasta, and deep-sea
bottom trawls by the ship ANTON BRUUN
during the National Science Foundation's South-
eastern Pacific Biological Oceanographic Pro-
gram (SEPBOP). Explorations in the Atlantic
Ocean began with the capture of the type species,
P. obesa (=P. bulbiceps) by the expeditions of
Prince Albert I of Monaco in the Bay of Biscay
(Zugmayer 191 la, b). Subsequently, other east-
ern Atlantic material has come from expeditions
of the Institute of Oceanographic Sciences, U.K.
(Merrett and Marshall 1981), the BENGUELA
VII expedition of the Institute de Investigaciones
Pesqueras de Barcelona (now Institute de Cien-
cias del Mar), fish trapping cruises of the Office
de la Recherche Scientifique et Technique Outre-
Mer, Dakar, Senegal, exploratory fisheries cruis-
es of the USSR (Golovan' 1978), and similar
cruises by the South African ship AFRICANA
around the Cape of Good Hope. Specimens of
Pachycara from the northwestern Atlantic are
relatively few and were taken by bottom trawls
in the Middle Atlantic Bight (Musick 1979; Hae-
drich et al. 1980), the Caribbean Sea (Anderson
et al. 1986), by bottom traps around Puerto Rico,
supplied to the author by Dannie A. Hensley,
University of Puerto Rico, and from operations
of DSV ALVIN in the Gulf of Mexico (Paull et
al. 1984).
METHODS AND MATERIALS
Measurements were made with dial calipers or
ocular micrometer to the nearest 0.1 mm. Os-
teological observations were made on cleared and
stained specimens (Dingerkus and Uhler 1977)
and drawings made with the aid of a camera
lucida. Definitions of characters and measure-
ments follow those of Anderson (1982a, 1984).
Institutional abbreviations are as listed in Lev-
iton et al. (1985), or as emended by Leviton and
Gibbs(1988).
This review is based on 249 specimens, 76-
597 mm standard length (SL), housed in 21 mu-
seum collections in 12 countries (listed by ab-
breviation in each account; measurements of
specimens in SL). All type specimens were ex-
amined except the holotype of Lycenchelys cras-
siceps Roule, 1916, thought to be lost. A lecto-
type of Lycodes brachycephalus Pappenheim,
1912, is selected. Full accounts and illustrations
are provided for all species except those recently
reviewed (P. brachycephalum [see Anderson
1988a], P. bulbiceps, P. gymninium, P. lepinium,
and P. suspectum [see Anderson and Peden
1988]), but expanded diagnoses for these species
are provided which include other, important
characters, for ease of reference under one cover.
The osteological section of the description of the
genus is intentionally brief, and the reader should
consult Anderson (1982a) for a more thorough
account of a generalized zoarcid (which includes
Pachycara}.
Species accounts are presented alphabetically
by ocean basin in the following order: Atlantic,
Indian, Pacific, and Antarctic.
Abbreviations for bone names used in the text
figures are as follows:
act— actinosts
ang— anguloarticular
boc — basioccipital
br r— branchiostegal ray
cerhy— ceratohyal
cl— cleithrum
cor— coracoid
den— dentary
ect— ectopterygoid
epihy— epihyal
ANDERSON: REVIEW OF EELPOUT GENUS PACHYCARA
223
epo— epioccipital (epiotic)
exo— exoccipital
fr— frontal
hyom — hyomandibula
hyphy — hypohyals
inop — interopercle
inthy— interhyal
lat eth— lateral ethmoid
mes — mesethmoid
met— metapterygoid
op— opercle
pal— palatine
par— parietal
pas— parasphenoid
pel — postcleithrum
pel— pelvic (basipterygium)
pmx — premaxilla
pop— preopercle
ptem — posttemporal
pto— pterotic
pts — pterosphenoid
quad— quadrate
scap— scapula
sop— subopercle
sph— sphenotic
supcl — supracleithrum
urhy— urohyal
Pachycara Zugmayer, 1911
Pachycara Zugmayer, 191 la: 12 (original description; expand-
ed in Zugmayer, 1 9 1 1 b: 1 34-1 36, pi. VI, fig. 6. Type species:
Pachycara obesa Zugmayer, 1911 (=Maynea bulbiceps Gar-
man, 1 899), by monotypy.
Austrolycichthys Regan, 1 9 1 3:244-245, fig. 2 (type species: Ly-
codes brachycephalus Pappenheim, 1 9 1 2, by subsequent des-
ignation of Jordan, 1920:550).
Pachycarichthys Whitley, 1931:334 (improper emendation).
Pachychara (lapsus calami). Fowler, 1936:1056, fig. 437; An-
driashev, 1986:170, text fig.
DIAGNOSIS. — Body robust, tail short; skin firm;
mental crests (cartilaginous ridges on dentary)
absent; parasphenoid wing not extending above
mid-height of trigeminofacialis foramen; pala-
topterygoid (palatal) arch well developed, with
ectopterygoid and mesopterygoid broadly artic-
ulating with quadrate; cephalic lateralis pores few,
small, rounded; precaudal vertebrae 21-32; sub-
orbital bones 6-8; pelvic fins present or absent,
their presence individually variable in two
species; scales, body lateral line, vomerine and
palatine teeth, pseudobranch (except in one
species), and pyloric caeca present.
DESCRIPTION.— Head large, ovoid, snout gen-
erally bluntly rounded, never dorsoventrally de-
pressed in adults; young juveniles with some-
what depressed heads, but snout not as long and
tapering as lycenchelyine (Anderson 1984) gen-
era. Largest males with wider head than in fe-
males owing to greater development of adductor
mandibulae (in those species with sample size
large enough to test statistically). Body elongate,
robust, subcircular in cross-section, its greatest
depth midway between pectoral base and anus;
body height at anal fin origin 7.7-16.9% SL in
specimens over 1 60 mm SL. Tail high, increas-
ingly laterally compressed posteriad. Skin firm,
not gelatinous, covering all fins. Scales minute,
cycloid, imbedded, covering body usually in-
cluding abdomen (except in smallest specimens
and P. rimae), bases of vertical fins, pectoral base
and pectoral axil. Complete mediolateral and
ventral (Andriashev 1954, fig. 142) lateral line(s)
developed as single row of superficial neuromasts
("free lateralis organs"; Springer and Freihofer
1976), their origins variable and usually species
specific. Eye large and ovoid in young, often rel-
atively small and rounded in adults, often en-
tering dorsal profile of head. Nostril single, tu-
bular, at snout tip, nasal tube not reaching upper
lip except in some young. Pectoral fin large, of
14-19 rays, middle rays longest in large speci-
mens, lowermost rays thickened, slightly exsert-
ed at tips. Pectoral base low on body, its origin
just below midline, insertion on abdomen. Gill
slit extending ventrally to lower margin of pec-
toral base or below. Ovary single. Pseudobranch
filaments 2-7 (absent in P. rimae). Two nublike
pyloric caeca, usually better defined in young.
Stomach and intestines pale. Peritoneum black.
Reproductive mode unknown, probably ovipa-
rous.
Mouth terminal or subterminal, lips fleshy but
not thickened. Upper lip free across snout, lower
lip adnate at dentary symphysis. Oral valve (pal-
atal membrane) moderately developed, usually
reaching anterior margin of vomer, but often
constricted at sides except in very small speci-
mens.
Cephalic lateralis pores small, rounded, rela-
tively few. Six (usually; eight in P. crassiceps)
suborbital bones from which arise 5-7 pores. One
to four postorbital pores present, their number,
like those of suborbitals, individually variable
within a species. Two nasal (anterior supraor-
224
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 10
hyom
pmx
pop
quad
den
ang
FIGURE 1. Jaws, suspensorium, and opercular bones ofPachycara bulbiceps. BMNH 1981.6.16:15, 332 mm SL; right lateral
view.
bital) pores present at snout tip arising from fo-
ramina in anterior and posterior ends of nasal
bone. Interorbital pore(s) absent. Occipital pores
absent except in P. mesoporum and in two of six
specimens of P. crossacanthum. Preopercular and
mandibular canals joined; preoperculomandibu-
lar pores and bone foramina system in plesio-
morphic state for family (and no variation not-
ed), with four pores emanating from dentary, one
from anguloarticular, and three from preopercle
(Fig. 1). Head speckled with small, white super-
ficial neuromasts, particularly on nape and
cheeks; no pattern to these discernible.
Neurocranium well ossified, somewhat box-
like (Fig. 2). Lateral ethmoid broad, not strongly
decurved, but deeply sculptured at surface for
attachment of palatine ligaments. Anterior, cup-
shaped section of mesethmoid broad, deep. Para-
sphenoid wing low, not reaching mid-height of
trigeminofacialis foramen, broadly articulating
with frontal and pterosphenoid. Pterosphenoid
almost completely forming anterior border of tri-
geminofacialis foramen. Frontals unfused, with
deep, open channel anteriorly passing supraor-
bital lateralis canal. Sphenotic and parietal sep-
arated by frontal and pterotic. Lateral protuber-
ance for attachment of levator arcus palatini
muscle a weak ridge. Parietals separated from
midline by supraoccipital, with no channel
through posterior margin bracing supratemporal
commissure (which is incomplete or absent in
all but two species). Supraoccipital large, anterior
ramus extending well under frontals, mesial crest
low; supraoccipital narrowly contacting exoccip-
ital posteriorly. Prootic and sphenotic articulat-
ing dorsally forming hyomandibular condyle.
Hyomandibular foramen well separated from
condyle (Fig. 2A). Prootic also articulating dor-
epo
lot eth
exo
epo
exo
FIGURE 2. Neurocranium ofPachycara gymninium, BCPM
980-100, 281 mm SL: (A) left lateral view; (B) dorsal view.
ANDERSON: REVIEW OF EELPOUT GENUS PACHYCARA
225
inthy
epihy
plem
hyphy
urhy
FIGURE 3. Right hyoid bar ofPachycara gymninium, BCPM
980-100, 281 mm SL; lateral view.
sally with pterotic and intercalar, which form a
ridge extending onto exoccipital. Single pair of
small, barrel-like lateral extrascapulars present.
Foramina passing ninth (glossopharyngeal) and
tenth (vagus) cranial nerves located on low, lat-
eral ridge well away from margins of exoccipital.
Teeth present on jaws (absent on maxilla),
vomer, palatine and pharyngeal bones. Palato-
pterygoid (palatal) arch well developed, ecto-
pterygoid and mesopterygoid broadly overlapping
anterior and dorsal margins of quadrate (Fig. 1).
Hyomandibular posterior ramus not elongated.
Opercle (posteriorly) and subopercle poorly os-
sified. Metapterygoid relatively large and thick.
Most or all of ceratohyal-epihyal juncture
smooth, space between bones cartilage-filled;
bone weakly interdigitating dorsally in large
specimens of one observed species (Fig. 3). Hy-
pohyals separated by wide, cartilage-filled space.
Interhyal ventral surface with concave fossa; dor-
sal tip not greatly sculptured. Six branchiostegal
rays present (except in P. rimae), with four ar-
ticulating with ceratohyal and two with epihyal.
Urohyal large and broad, with well developed
fossae dorsally for attachment of sternohyoideus
muscle.
Ceratobranchial five (lower pharyngeals) den-
tigerous, unfused. Three pairs of ossified infra-
pharyngobranchials (upper pharyngeals), asso-
ciated with gill arches 2-4. Three ossified
basibranchials; fourth an irregular cartilaginous
pad. Uncinate processes on epibranchials three
and four only. Gill rakers present on epibran-
chials, ceratobranchials, and usually one or two
on hypobranchials, numbering 10-21 on first
arch.
Posttemporal ventral ramus weak or absent
(Fig. 4). Supracleithrum with thin, poorly ossi-
fied posterior lamina; median ridge well devel-
oped. Scapular foramen enclosed by bone; scap-
FIGURE 4. Left pectoral girdle ofPachycara bulbiceps, CAS
6 1 220, 445 mm SL; lateral view. Pelvic bones greatly reduced
in size in this species.
ula with well developed posterior strut. Four
actinosts present bearing all pectoral rays but
dorsalmost 2-3 (borne by scapular strut). Pelvic
fins absent (P. bulbiceps), present (10 species), or
variably present and absent (P. mesoporum and
P. sulaki), with three (rarely two) soft rays. Sin-
gle, needlelike postcleithrum present. Cleithrum
with deep dorsal and ventral grooves for muscle
attachments.
Vertebrae symmetrical. Epipleural ribs on ver-
tebrae 1-21 or 22. Pleural ribs on third to ulti-
mate precaudal vertebrae. Dorsal fin origin as-
sociated with vertebrae 2-8, with no free dorsal
pterygiophores. First, or first and second, dorsal
element a "flexible spine" (unsegmented and un-
branched, but bilaterally divided at base); oc-
casionally branched at tip, but this may be a
preservation artifact. Last dorsal ray associated
with second through fifth (usually fourth) preural
vertebrae. Anal fin origin associated with ante-
penultimate to ultimate precaudal vertebrae, with
1-9 anal pterygiophores (all bearing soft rays)
226
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 10
inserted anterior to haemal spine of first caudal
vertebrae. Last anal ray associated with second
(rarely third) preural vertebrae. Caudal fin with
1-3 epural and 3-6 upper and lower hypural rays,
totaling 8-12 rays.
COMPARATIVE REMARKS.— Recently, Arnulf et
al. (1987) elaborated on the osteology and rela-
tionships of Thermarces Rosenblatt and Cohen,
1986, thought to be derived from Pachycara by
the latter authors. Thermarces was diagnosed by
Rosenblatt and Cohen, in part, on the basis of
four characters shown by Arnulf et al. (1987) to
have different states than those originally de-
scribed (those of original description given first):
(1) one or two suborbital ossifications vs. six; (2)
pectoral radials unossified vs. ossified; (3) two
pairs of infrapharyngobranchials vs. three; and
(4) postcleithrum absent vs. present. The speci-
men of Thermarces cerberus (SIO 82-46) studied
by Rosenblatt and Cohen was poorly ossified and
I have observed, on subsequent restaining, a
postcleithrum, six suborbital bones, four pecto-
ral radials, and a cartilaginous third pharyngeal
tooth plate. These structures, although bony,
picked up very little or no alizarin-red stain,
however. The skeletons of many deep-sea zoar-
cids are relatively unossified (Anderson and
Hubbs 1981) and endochondral bone may not
fully ossify in individuals of some species living
in possibly environmentally stressful habitats.
Stresses to normal calcium metabolism in the
hydrothermal vent habitat of Thermarces may
include high dissolved sulphide and metal ion
levels (J. M. Edmonds, Massachusetts Institute
of Technology, pers. comm.), making alizarin
staining difficult (Anderson 1 984). This seems to
have been the case with the specimen cleared
and stained by Rosenblatt and Cohen (1986) for
illustration.
On the basis of the detailed osteology pre-
sented by Arnulf et al. (1987) and my studies,
Thermarces does appear to be derived from, and
very close to, Pachycara, although Arnulf et al.
suggested Thermarces might be placed in a new
subfamily someday. However, Thermarces is
separable from Pachycara by only four charac-
ters: ( 1 ) pelvic bone absent; (2) scales absent; (3)
lateral line absent; (4) flesh gelatinous. Although
Arnulf et al. presented a detailed osteological de-
scription of Thermarces, their discussion on re-
lationships and characters is flawed. For exam-
ple, their table 1 is a very limited assemblage of
essentially non-diagnostic characters of higher
groups, some of which are redundant. Their
subfamily "Brotulinae" refers to the Parabro-
tulidae (Nielsen 1973, 1986; Anderson 1986).
Subfamily Neozoarcinae was placed in Stichaei-
dae by Anderson (1984). Melanostigma was
shown not to possess a basisphenoid bone by
Anderson and Hubbs (1981). The attribution and
interpretation of many characters of zoarcids by
Arnulf et al. are mostly incorrect and a detailed
discussion of these and an osteological analysis
will be presented elsewhere.
KEY TO SPECIES OF PACHYCARA
1A. Branchiostegal rays six; pseudobranch
present 2
IB. Branchiostegal rays four or five; pseu-
dobranch absent .. Pachycara rimae n. sp.
Galapagos Rift Zone
2A. Occipital (supratemporal) pores absent
(rarely present in P. crossacanthum; pre-
caudal vertebrae 26-28 3
2B. Single, mesial occipital pore present;
precaudal vertebrae 24-26
Pachycara mesoporum n. sp.
Peru and Chile
3A. Lateral line with mediolateral and ven-
tral branches*; precaudal vertebrae 23-
32; dorsal fin origin associated with ver-
tebrae 3-8; anal fin origin associated with
vertebrae 23-3 1 4
3B. Lateral line of ventral branch only; pre-
caudal vertebrae 21-23; dorsal fin origin
associated with vertebrae 2-3; anal fin
origin associated with vertebrae 20-22
........ Pachycara suspectum (Garman, 1899)
Eastern Tropical Pacific
4A. Dorsal fin origin associated with verte-
brae 3-7; precaudal vertebrae 23-30;
head length 1 1.8-19.9% SL 5
4B. Dorsal fin origin associated with verte-
bra 8; precaudal vertebrae 32; head
length 1 1.4% SL
Pachycara shcherbachevi n. sp.
Indian Ocean
5A. Gill rakers on lower limb (ceratobran-
chial) of first arch blunt, triangular 6
* Lateral line condition unknown in P. shcherbachevi n. sp.
ANDERSON: REVIEW OF EELPOUT GENUS PACHYCARA
227
80
60
40
20
0
20
40
60
FIGURE 5. Distribution of selected Pachycara species. Circles: P. bulbiceps; squares: P. crassiceps; hexagons: P. crossacanthum;
triangles: P. sulaki; star: P. shcherbachevi.
5B. Gill rakers on lower limb of first arch
bi- or tricuspid
Pachycara crossacanthum n. sp.
Eastern Tropical Atlantic
6A. Origin of ventral branch of lateral line
just posterior to last postorbital pore .. 7
6B. Origin of ventral branch of lateral line
on or behind vertical through posterior
third of pectoral fin
Pachycara sulaki n. sp.
Western Tropical Atlantic
7A. Pelvic fins present 8
7B. Pelvic fins absent
Pachycara bulbiceps (Garman, 1899)
NE Pacific and N Atlantic
8A. Suborbital pores along ventral ramus
(beneath eye) six 9
8B. Suborbital pores beneath eye five
Pachycara brachycephalum
(Pappenheim, 1912)
Antarctica
9A. Origin of mediolateral branch of lateral
line on or behind vertical through pec-
toral axil 1 0
9B. Origin of mediolateral branch of lateral
line on nape, just posterodorsal to last
postorbital pore
_ Pachycara pammelas n. sp.
Chile
1 OA. Precaudal vertebrae 27-3 1 origin of me-
diolateral branch of lateral line anterior
to vertical through posterior third of
pectoral fin 1 1
10B. Precaudal vertebrae 23-26; origin of
mediolateral lateral line at or posterior
to vertical through posterior margin of
pectoral fin Pachycara lepinium
Anderson and Peden, 1988
Northeastern Pacific
11 A. Total vertebrae 102-109; caudal verte-
brae 73-80; D 96-103; A 77-84; pec-
toral fin length 63.0-70.8% HL
Pachycara gymninium
Anderson and Peden, 1988
Northeastern Pacific
1 IB. Total vertebrae 1 1 1-118; caudal verte-
brae 84-88; D 105-1 12; A 86-92; pec-
toral fin length 81.6-97.0% HL
Pachycara crassiceps (Roule, 1916)
Eastern Atlantic
Pachycara bulbiceps (Garman, 1899)
(Figs. 1, 4, 5)
Maynea bulbiceps Garman, 1899: 140- 141, pi. E, fig. 1 (original
description. Type locality: Gulf of Panama). McAllister and
Rees, 1964:106, 107. Pearcy et al., 1982:387, 399, 400.
Pachycara obesa Zugmayer, 191 la:12; 191 lb:134-136, pi. VI,
fig. 6 (type locality: Bay of Biscay). Markle and Sedberry,
228
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 10
1978:22-25, fig. 1. Merrett and Marshall, 1981:240. An-
driashev, 1 973:547. Haedrich and Merrett, 1988:1335, 1338
(as obesum).
-Paehychara obesa Zugmayer (lapsus calami). Fowler, 1936:
1056, fig. 437. Andriashev, 1986:1 149, text fig.
Pachycara bulbiceps (Garman). Anderson and Peden, 1988:
84-88, figs. 1-3.
MATERIAL EXAMINED.— See Anderson and Peden (1988:84-
85).
DIAGNOSIS.— Pelvic fins absent; mediolateral
lateral line originating posterior to pectoral fin
margin; scales absent on nape; vertebrae 25-31
+ 82-91 = 112-119; D 104-114; A 86-97; P,
16-19; dorsal fin origin associated with vertebrae
4-7; gill rakers 0-4 + 11-15=1 1-19; branchios-
tegal rays six; suborbital pores 6-7; postorbital
pores 1-3; head length 1 1.8-15.0% SL; pectoral
fin length 8.7-1 2.9% SL.
DISTRIBUTION.— Both sides of the North At-
lantic at depths of 2,400-4,780 m and in the
northeastern Pacific from the Queen Charlotte
Islands, Canada, to the Gulf of Panama at depths
of 2,60 1-4,000 m (Fig. 5).
Pachycara crassiceps (Roule, 1916)
(Figs. 5, 6)
Lycenchelys crassiceps Roule, 1916:16 (original description.
Type locality: Bay of Biscay; based on two syntypes appar-
ently lost).
Lycodes macrops (non Gunther 1880). Vaillant, 1888:306-
309.
Lycodes mucosus (non Richardson 1855). Vaillant, 1888:31 1-
312.
Lycenchelys crassiceps. Roule, 1919:64-66, pi. V, fig. 1. Mer-
rett and Marshall, 1981:240.
Lycodes sp. Golovan', 1974:289.
Lycodes atlanticus (non Jensen 1 902). Golovan', 1978:226.
Lycodes crassiceps. Andriashev, 1986:1 137, text fig.
MATERIAL EXAMINED. -BMNH 1 98 1 .6.23:3 (448 mm); Por-
cupine Sea Bight, SW of Ireland (51°14.7'N, 13°16.3'W);
CHALLENGER sta. 50509; trawl, 1,490-1,523 m; 3 June 1979.
ZIN 45347 (415 mm); off Cabo Barbas, Western (Spanish)
Sahara (22°20'N, 1 7°30'W); ZVEZDA KRYMA sta. 260; trawl,
1,540-1,620 m; 16 Feb. 1973. BMNH 1981.6.23:1, 2 (432-
518 mm); off Cap Blanc, Mauritania (20°09.1'N, 18°08.8'W);
DISCOVERY sta. 9 1 33-7; trawl, 2, 1 30-2, 1 9 1 m; 26 Nov. 1 976.
CAS 55587 (337 mm); off Senegal (13°12.5'N, 17°43.4'W); L.
AMARO; trap, 1,300 m; 10 Feb. 1983. ZIN 45348 (509 mm);
off Cape Fria, Namibia (18°26'S, 14°06'E); EVRIKA; trawl,
1,100-1,125 m; 22 Dec. 1975. HPB 168/1985 (513 mm); off
Namibia (23°05'S, 12°41'E); BENGUELA VII sta. 2; trawl,
652 m; 9 Aug. 1984. PEM 13784 (395 mm); off Cape of Good
Hope, South Africa (33°40.0'S, 1 7°25.8'E); AFRICANA cruise
060, sta. A6999-01-04B; trawl, 923 m; 5 Mar. 1988. SAM
uncat. (396 mm) off Cape of Good Hope (34°54.9'S, 1 8°1 2. 1 'E);
AFRICANA cruise 060, sta. A7038-16-04B; trawl, 917 m; 14
Mar. 1988. SAM 31601 (189 mm); off Cape of Good Hope
(33°37.5'S, 17°24.5'E); AFRICANA cruise 060, sta. A7002-
06-03B; trawl, 668 m; 8 Mar. 1988. RUSI 28210 (380 mm);
off Cape of Good Hope (34°25.6'S, 17°41.2'E); AFRICANA
sta. A4310; trawl, 760 m; 7 July 1986.
COUNTS AND MEASUREMENTS.— Vertebrae 27-
30 + 84-90 = 111-118; D 105-112; A 86-94;
C 10-12; P, 17-19; P2 3; vomerine teeth 6-8;
palatine teeth 8-17; gill rakers 3 + 12-13; bran-
chiostegal rays six; pseudobranch filaments 4-5.
Following measurements in percent SL: head
length 12.6-14.5; head width 8.0-10.3; pectoral
fin length 10.7-13.3; predorsal length 17.3-21.6;
preanal length 38. 1-43.7; body height 10.7-13.6;
gill slit length 5.5-7.4; caudal fin length 2.3-4.6.
Following measurements in percent HL: head
width 59.4-70.7; upper jaw length 38.7-45.4;
pectoral fin length 81.6-97.0; snout length 17.4-
23.4; eye diameter 17.3-20.8; gill slit length 40.4-
50.7; interorbital width 8.5-11.4; interpupillary
width 25.4-33.3; pelvic fin length 9.9-18.6 Pec-
toral base/length ratio 32.8-44.3.
DIAGNOSIS. — Pelvic fins present; mediolateral
lateral line originating above pectoral fin just
posterior to pectoral base; scales present on nape;
vertebrae 27-30 + 84-90 =111-118; gill rakers
15-21; pectoral fin length 81.6-97.0% HL.
DESCRIPTION.— Head ovoid, dorsal profile of
snout more steeply sloping in large adults than
juveniles. Scales extending anteriorly onto nape
to about two eye diameters behind posterior
margin of eye in large specimens; scales present
on abdomen to isthmus, pectoral base and axil,
and extending onto pectoral fin up to about half
its length; scales present on unpaired fins almost
to their margin. Eye circular, entering dorsal pro-
file of head in large specimens. Gill slit extending
to ventral margin of pectoral base. Lobe at dorsal
margin of gill slit not formed in some large spec-
imens; weak and rounded in juveniles. Pectoral
fin origin at or slightly below body midline, in-
sertion on abdomen; posterior margin of fin
wedge-shaped (cf. Lycodes), pectoral rays 4-8
longest; ventralmost three or four rays thickened,
tips slightly exserted.
Mouth subterminal, slightly oblique, upper jaw
extending posteriorly to middle of eye. No epi-
dermal prickles evident on snout or lips. Oral
valve reaching anterior edge of vomer and co-
alesced with lateral margin of plate opposite
vomer. Jaw teeth small, conical; dentary with 4-
5 irregular rows anteriorly, blending into single
posterior row; premaxilla with 3-4 anterior rows,
blending into single posterior row. Vomerine teeth
in irregular patch, longer and sharper in juveniles
ANDERSON: REVIEW OF EELPOUT GENUS PACHYCARA
229
FIGURE 6.
pelvic fins.
Pachycara crassiceps (Roule), BMNH 1 98 1 .6.23: 1-2, 518m SL, from off Cap Blanc, Mauritania. Arrow indicates
than adults. Palatine teeth not retrorse, in double
row at least anteriorly, in all.
Cephalic lateralis system with two or three
postorbital pores (pores one and four, arising from
frontal and lateral extrascapular, or pores one,
three, and four). Two pairs of anterior supraor-
bital (nasal) pores, one set mesial to nasal tube,
the other posteriorly. Six or seven suborbital
pores, six arising from ventral ramus of bone
chain and one (absent in three specimens; absent
on one side only in one specimen) from ascend-
ing ramus behind eye just ventral to first post-
orbital pore. Eight preoperculomandibular pores,
four arising from dentary, one from anguloartic-
ular, and three from preopercle. Interorbital and
occipital pores absent. Body lateral line system
with mediolateral branch originating above pec-
toral fin about one eye diameter posterior to pec-
toral base; ventral branch originating just pos-
terior to fourth postorbital pore; both branches
complete to tail tip. Free superficial neuromasts
(about 8-10) forming dorsal "lateral line" orig-
inating above pectoral base on horizontal through
middle of eye and extending posteriorly to ver-
tical through posterior margin of pectoral fin or
slightly anteriorly.
Dorsal fin origin associated with vertebrae five
or six. Anal fin origin associated with ultimate
precaudal vertebrae (numbers 25-30), with 3-5
anal pterygiophores inserted anterior to haemal
spine of first caudal vertebra. Last dorsal ray
associated with fourth preural vertebra, last anal
ray associated with second preural vertebra. Cau-
dal fin with two epural, 4-5 upper and 4-5 lower
230
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 10
FIGURE
•^^^^^^^•^•••^^^••^^
7. Pachycara crossacanthum n. sp., CAS 62408, paratype, 227 mm SL, from off Senegal.
hypural rays. Gill rakers, small, dorsalmost on
lower limb (ceratobranchial) triangular, sharply
pointed, ventralmost blunt; ventralmost four
rakers in ZIN 45347 with tricuspid tips. Bran-
chiostegal rays six. Pseudobranch filaments long,
simple.
Color uniformly black (adults) or dark choc-
olate brown (juveniles), in recently preserved
material. Eye blue. Abdomen with bluish tinge
in juveniles, black in adults. Lining of orobran-
chial cavity black, lateral regions of palate pale.
DISTRIBUTION.— Known in eastern Atlantic
from SW of Ireland to off South Africa on upper
and lower continental slopes at depths of 652-
2,191 m(Fig. 5).
REMARKS.— This species was originally de-
scribed from two syntypes, neither of which were
found. However, a neotype is not selected here
as the syntypes have not been sufficiently searched
for, and diagnostic features (counts of the axial
skeleton from x-radiography) published in Roule
(1919) permit the distinction of P. crassiceps from
its close congener, P. crossacanthum (below).
Pectoral fin ray counts in Roule (1919) are, how-
ever, undoubtedly erroneous.
Pachycara crossacanthum n. sp.
(Figs. 5, 7, 8)
HOLOTYPE.— CAS 55586 (female, 370 mm); off Senegal
(12°58.0'N, 17°41.8'W); L. AMARO sta. Casamance; fish trap,
900 m; B. Seret, 8 Feb. 1983.
PARATYPES.-MNHN 1988-1170 (3; 227-378 mm); col-
lected with holotype. CAS 62408 (female, 227 mm); offSenegal
(15'47.8'N, 17°07.1'W); L. AMARO, sta. 6; fish trap, 900 m;
D. Gaertner, 29 Mar. 1984. MNHN 1988-1169 (female, 353
mm); offPointe Banda, Gabon (04°00'S, 10°08'E); NIZERY
sta. 14-76; fish trap, 1,050 m; A. Crosnier, 20 May 1976.
COUNTS AND MEASUREMENTS.— Vertebrae 26-
28 + 75-79 = 101-107; D 97-103; A 77-83; C
10-12 P, 17-18; P2 3; vomerine teeth 5-14; pal-
atine teeth 6-16; gill rakers 3 + 12-13; branchi-
ostegal rays six; pseudobranch filaments 4-5.
Following measurements in percent SL: head
length 14.8-16.9; head width 8.6-9.5; pectoral
fin length 9.2-10.5; predorsal length 17.6-19.7;
preanal length 41.3-44.2; body height 9.6-1 1.8;
gill slit length 5.8-6.7; caudal fin length 3.4-5.0.
Following measurements in percent HL: head
width 52.7-64.5; upper jaw length 35.5-45.6;
pectoral fin length 54.6-66.0; snout length 19.1-
23.2; eye diameter 1 5.2-20.2; gill slit length 34.9-
41.8; interorbital width 6.9-10.3; interpupillary
width 21.1-25.3; pelvic fin length 12.2-18.0.
Pectoral base/length ratio 38.9-50.9.
DIAGNOSIS.— Pelvic fins present; dorsal gill
rakers on first arch bi- or tricuspid at tips (adults
and subadults); scales absent on nape and ab-
domen; postorbital pores four; vertebrae 26-28
+ 75-79 = 101-107; origin of mediolateral lat-
eral line near dorsal margin of pectoral base; oc-
cipital pores present or absent.
DESCRIPTION.— Head ovoid, snout not bluntly
rounded (as some species), but steep anteriorly.
Scales extending anteriorly on body in wedge-
shaped pattern to vertical from half pectoral fin
mm
FIGURE 8. Outer surface of right first gill arch of holotype
of Pachycara crossacanthum n. sp., showing furcated gill rak-
ers; gill lamellae not shown.
ANDERSON: REVIEW OF EELPOUT GENUS PACHYCARA
231
FIGURE 9. Pachycara sulaki n. sp., USNM 29281 1, holotype, 189 mm SL, from off Puerto Rico.
length posterior to pectoral margin (usually in
smallest specimens) to vertical through middle
of pectoral; scales absent on abdomen, nape, pec-
toral axil and base. Eye ovoid, entering dorsal
profile of head. Gill slit short, not reaching ven-
tral margin of pectoral base. Lobe at dorsal mar-
gin of gill slit rounded and weakly developed in
three specimens (including holotype), squared-
off and well denned in others. Pectoral fin origin
well below body midline, insertion on abdomen;
posterior margin of fin evenly rounded, middle
rays longest; ventralmost four or five rays thick-
ened, tips slightly exserted.
Mouth subterminal, oblique, upper jaw ex-
tending posteriorly to middle of eye, or slightly
in advance. No epidermal prickles evident on
snout or lips. Oral valve weak, not reaching an-
terior edge of vomer, and not coalesced with lat-
eral margin of palate opposite vomer. Jaw teeth
in large specimens long, sharp, retrorse ante-
riorly; dentary with 4-5 irregular rows anteriorly,
blending into single posterior row; premaxilla
with two (males) or three (females) anterior rows,
blending into single posterior row. Vomerine teeth
in irregular patch; palatine teeth not retrorse, in
double row in large females, in single row in all
others.
Cephalic lateralis system with four postorbital
pores arising from frontal (pore one), pterotic
(pore two), between pterotic and lateral extra-
scapular (pore three) and just posterior to lateral
extrascapular (pore four). Two pairs of anterior
supraorbital (nasal) pores, one set mesial to nasal
tube, the other posteriorly. Seven suborbital
pores, six arising from ventral ramus of bone
chain and one from ascending ramus behind eye
just ventral to first postorbital pore. Eight pre-
operculomandibular pores, four arising from
dentary, one from anguloarticular and three from
preopercle. Interorbital and occipital pores ab-
sent, except in CAS 62408 which has three mi-
nute occipital pores and MNHN 1988-11 70, 227
mm SL, which has a single, right lateral occipital
pore. Body lateral line with mediolateral branch
originating in pectoral axil near dorsal margin of
pectoral base; ventral branch originating just
posterior to fourth postorbital pore; both branch-
es complete to tail tip.
Dorsal fin origin associated with vertebra four.
Anal fin origin associated with ultimate precau-
dal vertebra (numbers 26-28), with three anal
pterygiophores inserted anterior to haemal spine
of first caudal vertebra. Last dorsal ray associated
with fourth preural vertebra, last anal ray asso-
ciated with second preural vertebra. Caudal fin
with 1-2 epural, 4-5 upper and 4-6 lower hypural
rays.
Gill rakers on upper limb (epibranchial) of first
arch long, sharply pointed in young, bi- or tri-
furcate in large specimens; rakers on lower limb
(ceratobranchial) tubular, those of large speci-
mens with bi-or tricuspid tips (Fig. 8); lowermost
1-4 rakers of some specimens simple, pointed.
Branchiostegal rays six. Pseudobranch filaments
4-5, tips furcate in large specimens.
Color uniformly dark brown, margins of dor-
sal, anal and pectoral fins black; eye and abdo-
men bluish. Peritoneum and inner palate black.
Lips and lining of lateral sides of palate white.
DISTRIBUTION.— Known off western tropical
Africa from Senegal to Gabon on the upper con-
tinental slope at depths of 900-1,050 m (Fig. 5).
ETYMOLOGY. —From the Greek updaaos (fringe)
and 'ciKCivQa (thorn) in reference to the species'
furcate gill rakers.
Pachycara sulaki n. sp.
(Figs. 5, 9)
Zoarcid fish. Paull et al., 1984:965, 966.
Pachycara sp. Anderson et al., 1986:800.
HOLOTYPE.— USNM 2928 1 1 (male, 1 89 mm); Mona Passage
off Puerto Rico (1 8°28.7'N, 67°20.6'W); PEZMAR, Deep Trap-
ping Project, sta. 36; medium fish trap, 2,000 m; D. A. Hensley,
12-13 Feb. 1986.
PARATYPES.— USNM 233627 (female, 161 mm); Venezuela
Basin, Caribbean Sea (13°31.3-25.7'N, 64°45.1-40.4'W); USNS
BARTLETT sta. E092; trawl, 3,510-3,469 m; M. D. Richard-
son and party, 28 Nov. 1981. SIO 87-33 (female, 133 mm);
232
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 10
Florida Escarpment, Gulf of Mexico (26°05'N, 84°54'W); DSV
ALVIN dive 1771; Floe Sucker, 3,303 m; C. Paull and party,
7 Nov. 1986.
COUNTS AND MEASUREMENTS.— Vertebrae 27-
30 + 80-85 = 107-1 13; D 99-107; A 81-89; C
9-1 1; P, 16-17; P2 0 or 3; vomerine teeth 4-8;
palatine teeth 4-7; gill rakers 2-3 + 9-13 = 11-
16; branchiostegal rays six; pseudobranch fila-
ments 3-4. Following measurements in percent
SL: head length 14.8-17.7; head width 7.3-9.0;
pectoral fin length 9.8-10.6; predorsal length
20.8-23.5; preanal length 39.0-41 .7; body height
8.3-9.5; gill slit length 4.8-6.5; caudal fin length
2. 1-3.8. Following measurements in percent HL:
head width 47.8-56.1; upper jaw length 36.4-
50.0; pectoral fin length 55.5-68.8; snout length
1 6.4-20. 1 ; eye diameter 1 7.8-23.9; gill slit length
28.0-42.1; interorbital width 6.4-10.5; interpu-
pillary width 23.8-32.8; pelvic fin length 5.7-
13.9 (in two). Pectoral base/length ratio 38.9-
40.0.
DIAGNOSIS.— Pelvic fins present or absent; or-
igin of mediolateral lateral line just posterior to
last (fourth) postorbital pore; origin of ventral
lateral line on or behind vertical through pos-
terior third of pectoral fin; suborbital pores 5-6;
dorsal fin origin associated with vertebrae 6-7;
vertebrae 27-30 + 80-85 = 107-1 1 1; predorsal
length 20.8-23.5% SL.
DESCRIPTION.— Head ovoid, long, dorsal pro-
file evenly tapering; snout bluntly rounded. Scales
extending anteriorly on body in wedge-shaped
pattern to vertical through middle of pectoral fin
(all three specimens juveniles; adults probably
more densely scaled); scales absent in these ju-
veniles on dorsal fin anteriorly (present on pos-
terior third), abdomen, head, nape, pectoral base
and axil. Eye circular, entering dorsal profile of
head. Gill slit reaching ventral margin of pectoral
base or just above it. Lobe at dorsal margin of
gill slit weak, rounded. Pectoral fin origin just
below body midline, insertion on abdomen; pos-
terior margin of fin evenly rounded, middle rays
longest; ventralmost five rays thickened, tips
slightly exserted. Pelvic fins present except in
SIO 87-33; of three soft rays each in other two
specimens.
Mouth subterminal, oblique, upper jaw ex-
tending posteriorly to middle of eye in females
(paratypes), or its rear margin in male holotype.
No epidermal prickles evident on snout or lips.
Oral valve well developed in these small speci-
mens, overlapping anterior edge of vomer and
not coalesced with lateral margins of plate op-
posite vomer. Jaw teeth relatively long and sharp
in male, smaller and conical in females; outer
jaw teeth retrorse; both jaws (all specimens) with
two irregular rows of teeth blending into single,
posterior row. Vomerine teeth in irregular patch,
enlarged in male; palatine teeth not retrorse, in
single row.
Cephalic lateralis system with two postorbital
pores arising from frontal (pore one) and lateral
extrascapulars (pore four). Two pairs of anterior
supraorbital (nasal) pores, one set mesial to nasal
tube, the other posteriorly. Five (USNM 233627)
or six suborbital pores, all arising from ventral
ramus of bone chain. Eight preoperculomandib-
ular pores, four arising from dentary, one from
anguloarticular and three from preopercle. In-
terorbital and occipital pores absent. Superficial
neuromasts prominent on head an anterior por-
tion of body, with series on preopercle, behind
eye running along occiput in converging row,
around rictus of mouth, and snout between nos-
trils. Body lateral line with mediolateral branch
originating just behind posteriormost (number
four) postorbital pore; this branch undulating
across body somewhat, then straightening above
anus; ventral branch originating on or slightly
behind (holotype) posterior third of pectoral fin;
both branches complete to tail tip.
Dorsal fin origin associated with vertebrae 6-
7. Anal fin origin associated with penultimate or
ultimate precaudal vertebrae (numbers 26-30),
with 3-5 anal pterygiophores inserted anterior to
haemal spine of first caudal vertebra. Last dorsal
ray associated with fourth preural vertebra, last
anal ray associated with second preural vertebra.
Caudal fin with two epural, 3-5 upper hypural
and four lower hypural rays.
Gill rakers on upper limb (epibranchial) and
dorsalmost rakers on lower limb (ceratobran-
chial) narrow and sharply pointed; ventralmost
ceratobranchial rakers blunt, triangular. Bran-
chiostegal rays six. Pseudobranch filaments long,
numbering 3-4.
Live color notes or photographs of all three
specimens available, each one quite different.
Holotype uniformly dark chocolate brown, as
most other congeners, unpaired fins black, eye
blue. Paratype USNM 233627 from the Carib-
bean Sea with lustrous dark blue body, typical
of many deep-sea fishes such as halosaurs and
ANDERSON: REVIEW OF EELPOUT GENUS PACHYCARA
233
alepocephalids (Markle 1980, fig. 1), head and
pectoral fins black. Paratype SIO 87-33 relatively
blanched, body uniformly sandy gray, occiput,
nape and pectoral axil occupied by a wide, white
band, eye and abdomen blue, unpaired fins trans-
lucent.
DISTRIBUTION.— Known only from three spec-
imens taken in abyssal waters of the tropical
northwestern Atlantic at depths of 2,000-3,510
m (Fig. 5).
ETYMOLOGY.— Named after Kenneth J. Sulak,
intrepid explorer of the great murky depths of
the Atlantic Ocean, to honor his many contri-
butions to knowledge of the deep-sea fishes of
this and adjacent regions.
Pachycara shcherbachevi n. sp.
(Figs. 5, 10)
HOLOTYPE.-ZMUC P-761147 (female, 236 mm SL); Bay
of Bengal (15°54.0'N, 90°17.0'E); GALATHEA sta. 314; trawl
(HOT/4800), 2,600 m; 1930 hr; 3 May 1951.
COUNTS AND MEASUREMENTS.— Vertebrae 32
+ 90 = 122; D 112; A 94; C 12; P, 17; P2 3;
vomerine teeth two; palatine teeth 4/4; gill rakers
2+14; branchiostegal rays six; pseudobranch
filaments three. Following measurements in per-
cent SL: head length 1 1.4; head width 7.7; pec-
toral fin length 1 0.0; predorsal length 16.6; preanal
length not measurable; body height 7.7; caudal
fin length 3.0. Following measurements in per-
cent HL: head width 67.7; upper jaw length 39.0;
pectoral fin length 87.7; snout length 16.0; eye
diameter 23.8; interorbital width 13.8; interpu-
pillary width 28.6; pelvic fin length 8.9. Pectoral
base/length ratio 36.0.
DIAGNOSIS.— Pelvic fins present; vertebrae 32
+ 90 = 122; dorsal fin origin associated with
vertebra eight; head length 1 1 .4% SL.
DESCRIPTION. — Specimen trawl damaged—
completely skinned, body torn before anus; scales,
lateral line, and most head pores not evident.
Head ovoid, small; its length shorter in relation
to SL than any Pachycara specimen, irregardless
of size. Eye large, ovoid, entering dorsal profile
of head when viewed laterally. Gill opening dam-
aged, tissue covering branchiostegal membranes
torn, but adhering on left side in places, and, with
skin along pectoral base, gill opening appearing
to extend ventrally to just above lower pectoral
base. Dorsal margin of gill slit with short, ante-
riorly directed emargination. Pectoral fin origin
well below body midline, insertion on abdomen;
posterior margin of fin (from reconstruction uti-
lizing both fins) evenly rounded, middle rays
longest; ventralmost four rays thickened.
Mouth subterminal, upper jaw (premaxilla and
maxilla with some flesh) posteriorly relatively
high, somewhat rectangular. Oral valve reaching
anterior edge of vomer and coalesced with lateral
margin (left) of palate opposite vomer. Jaw teeth
small, conical, retrorse anteriorly; dentary with
14 teeth in two irregular rows anteriorly, blend-
ing into single, posterior row; premaxillae with
1 3 (right) or 1 1 (left) teeth, also in two irregular
rows anteriorly, blending into single, posterior
row. Two small vomerine teeth. Four palatine
teeth on both bones, in short row.
Cephalic lateralis pores evident from connec-
tive tissue tubules and impressions in exposed
subdermal lipid layer, not all pores remaining.
Presumably two pairs of anterior supraorbital
pores mesial to nasal tube. Six suborbital pores,
all arising from ventral ramus of bone chain un-
der eye. Eight preoperculomandibular pores, four
arising from dentary, one from anguloarticular,
and three from preopercle. Interorbital and oc-
cipital pores presumably absent. Postorbital pores
one (behind eye) and four (anterodorsal upper
end of gill slit) observed, others (pore two and
three) may or may not be present in the species.
Dorsal fin origin associated with vertebra eight;
posteriormost origin in genus, except for P. ri-
mae, n. sp., which also has dorsal origin with
eighth vertebra. Anal fin origin associated with
penultimate precaudal vertebra (number 31), with
four anal pterygiophores inserted anterior to hae-
mal spine of first caudal vertebra. Last dorsal ray
associated with fourth preural vertebra; last anal
ray associated with second preural vertebra. Cau-
dal fin with two epural, five upper hypural, and
five lower hypural rays.
Gill rakers small, sharply pointed and pyrami-
dal dorsally; ventralmost five much smaller than
dorsalmost, but distinct. Branchiostegal rays six.
Pseudobranch filaments three, very small.
Color unknown, but bits of black skin adhere
to head, body, and fins. Perhaps all black in life.
Peritoneum black. Lining of orobranchial cham-
ber dark brown in alcohol, probably black in life
also.
DISTRIBUTION.— Known from a single speci-
men from the abyssal Bay of Bengal, northwest
of the Andaman Islands, in 2,600 m (Fig. 5).
234
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 10
FIGURE 10.
trawl).
Pachycara shcherbachevi n. sp., ZMUC P-761 147, holotype, 236 mm SL, Bay of Bengal (specimen damaged by
ETYMOLOGY.— Named after Yuri Nikolaevich
Shcherbachev, P. P. Shirshov Institute of Ocean-
ology, Academy of Sciences, USSR, friend and
colleague, in honor of his pioneering contribu-
tions to knowledge of the deep-sea fishes of the
Indian Ocean.
Pachycara gymninium Anderson and Peden, 1988
(Fig. 11)
Lycodes sp. Hubbs et al., 1979:14.
Lycenchelys "E." Pearcy et al., 1982:387.
Pachycara gymninium Anderson and Peden, 1988:88-91,
fig. 5.
ANDERSON: REVIEW OF EELPOUT GENUS PACHYCARA
235
MATERIAL EXAMINED. — See Anderson and Peden (1988:88-
89).
DIAGNOSIS. — Pelvic fins present, their length
4.8-11.3% HL; mediolateral lateral line origi-
nating in pectoral axil just posterior to vertical
through pectoral base; scales absent on nape, or,
if present, not extending anterior to line con-
necting anterodorsal edges of gill slits; vertebrae
28-31 + 73-80 = 102-109; D 96-103; A 77-
84; P, 14-18; dorsal fin origin associated with
vertebrae 3-5; anal fin origin associated with ver-
tebrae 27-31; gill rakers 3-5 + 10-14 = 13-19;
branchiostegal rays six; suborbital pores 6-7;
postorbital pores 2-3; head length 12.0-15.6%
SL; pectoral fin length 63.0-70.8% HL.
DISTRIBUTION.— Eastern North Pacific off the
Queen Charlotte Islands, Canada, south to the
Gulf of California at depths of 1,829-3,225 m
(Fig. 11).
Pachycara lepinium Anderson and Peden, 1988
(Fig 11)
Lycodes sp. Hubbs et al., 1979:14.
Lycenchelys "D." Pearcy et al., 1982:387.
Pachycara lepinium Anderson and Peden, 1988:91-92, fig. 6.
MATERIAL EXAMINED.— See Anderson and Peden (1988:91).
DIAGNOSIS.— Pelvic fins present, their length
11.5-17.3% HL; mediolateral lateral line origi-
nating posterior to pectoral fin margin; scales
present on nape; vertebrae 23-26 + 80-94 =
105-120; D 99-113; A 85-98; P, 15-18; dorsal
fin origin associated with vertebrae 4-5; anal fin
origin associated with vertebrae 23-26; gill rak-
ers 2-5 + 11-15 = 12-19; branchiostegal rays
six; suborbital pores six; postorbital pores 2-3;
head length 13.8-15.9% SL; pectoral fin length
72.5-89.9% HL.
DISTRIBUTION.— Eastern North Pacific off the
Queen Charlotte Islands, Canada, south to off
Cabo Colnett, Baja California Norte, Mexico, at
depths of 1,728-2,907 m (Fig. 1 1).
Pachycara suspectum (Garman, 1899)
(Fig. 11)
Phucocoetes suspectus Garman, 1 899: 1 37, pi. XXX, figs. 3, 3a.
Pachycara suspectum (Garman). Anderson and Peden, 1988:
88, fig. 4.
MATERIAL EXAMINED.— See Anderson and Peden (1988:88).
DIAGNOSIS.— Pelvic fins present; lateral line of
ventral branch only; scales absent on nape; ver-
tebrae 21-23 + 84-87 = 105-110; D 100-106;
100U
80
60C
40C
FIGURE 1 1 . Distribution of eastern Pacific species of Pa-
chycara. Closed circles: P. gymninium; squares: P. lepinium;
hexagon: P. rimae, triangles: P. mesoporum; stars: P. suspec-
tum; open circles: P. pammelas.
A 85-89; P, 16; dorsal fin origin associated with
vertebrae 2-3; anal fin origin associated with ver-
tebrae 20-22; gill rakers 1-2+14-15 = 16; bran-
chiostegal rays six; suborbital pores seven; post-
orbital pores three; head length 15.6-17.7% SL;
pectoral fin length 57.4-58.6% HL.
DISTRIBUTION.— Eastern North Pacific in the
Gulf of California south to the Gulf of Panama
at depths of 915-1,280 m (Fig. 1 1).
Pachycara rimae n. sp.
(Figs. 11, 12)
Zoarcidae (partim). Cohen and Haedrich, 1983:376.
Pachycara (?). Cohen et al., 1985:229.
HoLOTYPE.-LACM 44699-1 (male, 403 mm SL); vie. Ga-
lapagos Rift hydrothermal vents (00°47.8'N, 86°09.9'W); DSV
ALVIN; baited trap, 2,500 m; J. Corliss, 16 Mar. 1977.
236
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 10
FIGURE 12. Pachycara rimae n. sp., LACM 44699-1, holotype, 403 mm SL, Galapagos Rift Zone. Illustration of right side
shown in left lateral view (left side damaged).
COUNTS AND MEASUREMENTS.— Vertebrae 26
+ 67 = 93; D 86; A 70; C 9; P, 1 5; P2 2; vomerine
teeth three; palatine teeth 8-9; gill rakers 3 + 7;
branchiostegal rays four or five; pseudobranch
filaments absent. Following measurements in
percent SL: head length 13.7; head width 10.1;
pectoral fin length 9.3; predorsal length 21.6;
preanal length 44.2; body height 14.1; gill slit
length 6.0; caudal fin length 1 .7. Following mea-
surements in percent HL: head width 74.2; upper
jaw length 54.3; pectoral fin length 68.1; snout
length 18.9; eye diameter 13.0; gill slit length
43.7; interorbital width 12.7; interpupillary width
42.1; pelvic fin length 4.2. Pectoral base/length
ratio 42.7.
DIAGNOSIS.— Pelvic fins nublike, of two soft
rays; vertebrae 26 + 67 = 93; D 86; A 70; lateral
line of mediolateral branch only; dorsal fin origin
associated with vertebra eight; pseudobranch ab-
sent; branchiostegal rays four or five; suborbital
pores five; gill rakers 10.
DESCRIPTION.— Head deep, rounded, some-
what shorter than similarly sized congeners. Snout
steeply sloping anteriorly. Left side of body and
tail damaged, but skin pliable around head and
dorsum owing to subdermal lipid layer. Scales
extending anteriorly on body in wedge-shape
pattern to vertical just anterior to posterior mar-
gin of pectoral fin; scales absent on head, nape,
dorsum, pectoral base, abdomen, and unpaired
fins. Eye small, circular, entering dorsal profile
of head when viewed laterally. Gill slit short,
extending ventrally to opposite pectoral ray 1 2.
Lobe at dorsal margin of gill slit weakly devel-
oped, gill slit not continued anteriorly. Pectoral
fin origin well below body midline; left pectoral
fin badly damaged in capture, right fin deformed,
apparently injured in life, with tips of dorsalmost
four rays missing (Fig. 1 2); six ventralmost pec-
toral rays thickened, slightly exserted at tips.
Mouth terminal, oblique, upper jaw extending
posteriorly to middle of eye or beyond (snout
region twisted to one side). No epidermal prick-
les evident on snout or lips. Oral valve just reach-
ing anterior margin of vomer and coalesced with
lateral margins of palate opposite vomer. Teeth
small, conical, sharp. Jaw teeth in double row
anteriorly, blending into single, posterior row.
Vomerine teeth three, in an arc; palatine teeth
in single row.
Cephalic lateralis system with two postorbital
pores arising from frontal (pore one) and lateral
extrascapular (pore four). Two pairs of anterior
supraorbital (nasal) pores, one set anteromesial
to nasal tube, the other posteromesially. Five
suborbital pores, all arising from ventral ramus
of suborbital bone chain. Eight preoperculoman-
dibular pores, four arising from dentary, one from
anguloarticular, and three from preopercle. In-
terorbital and occipital (supratemporal) pores
absent. Body lateral line mediolateral, originat-
ing posterior to rear margin of pectoral fin, com-
plete to tail tip; no ventral branch.
Dorsal fin origin associated with vertebra eight.
Anal fin origin associated with ultimate precau-
dal vertebra (number 26), with two anal pteryg-
iophores inserted anterior to haemal spine of first
caudal vertebra. Last dorsal ray associated with
fourth preural vertebra, last anal ray associated
with second preural vertebra. Caudal fin with
two epural, four upper hypural, and four lower
hypural rays.
Gill rakers very short, triangular, lowermost
rakers on first arch mere nubs; raker denticles
absent. Branchiostegal rays four on right side (two
articulating with ceratohyal, two with epihyal)
and five on left side (three articulating with cer-
atohyal, two with epihyal). Pseudobranch absent.
Color uniformly light brown (Cohen et al. 1985:
229); eye and abdomen bluish. Peritoneum and
lining of orobranchial chamber dark brown in
alcohol, probably black in life. Head and right
ANDERSON: REVIEW OF EELPOUT GENUS PACHYCARA
237
FIGURE 13. Pachycara mesoporum n. sp., CAS 62406, holotype, 485 mm SL, from off Antofagasta, Chile.
pectoral fin with scattered whitish blotches and
streaks.
DISTRIBUTION.— Known only from the holo-
type taken in the abyssal southeastern Pacific in
2,500 m.
ETYMOLOGY.— From the Latin "rima" (fis-
sure). A genitive noun in apposition, alluding to
the species' capture at the Galapagos Rift Zone
(Cohen and Haedrich 1983; Cohen et al. 1985).
Pachycara mesoporum n. sp.
(Figs. 11, 13)
HOLOTYPE.— CAS 62406 (male, 485 mm); off Antofagasta,
Chile; Chilean fisheries vessel; trawl, depth uncertain (ca. 300
? m); I. Kong Urbina, 18 Apr. 1986.
PARATYPES.-LACM 437 1 7- 1 (5; 1 70-2 1 7 mm); offTrujillo,
Peru (08°26'S, 80°36.5'W); ANTON BRUUN sta. 650E; free
vehicle set line trap, 1,830 m; L. Knapp and party, 8-9 June
1966. ZIN 48394 (male, 189 mm); off Laguna Grande, Peru
( 1 4°44'S, 76°1 2' W); DMITRY MENDELEEV cr. 20, sta. 1 654;
Galathea trawl, 1,495-1,430 m; N. V. Parin, 2335-0120 hr,
1 9-20 Mar. 1978. MNHNC 6653 (female, 3 1 8 mm); off Arica,
Chile (18°26'S, 70°37'W); TIBERIADES; trawl, 760 m; I. Kong
Urbina, 17 Feb. 1981.
COUNTS AND MEASUREMENTS.— Vertebrae 24-
26 + 78-87 = 103-1 13; D 96-108; A 80-90; C
9-10; P, 15-18; P2 absent or three; vomerine
teeth 3-12; palatine teeth 2-11; gill rakers 2-3
+ 1 1-16 = 13-18; branchiostegal rays six; pseu-
dobranch filaments 3-4. Following measure-
ments in percent SL: head length 12.6-15.9; head
width 6.1-8.2; pectoral fin length 8.8-12.1; pre-
dorsal length 15.1-18.1; preanal length 33.6-38.1;
body height 7.8-9.5; gill slit length 4.1-6.9; cau-
dal fin length 2.1-3.2. Following measurements
in percent HL: head width 44.6-52.5; upper jaw
length 34.7-51.7; pectoral fin length 57.1-81.1;
snout length 16.0-21.2; eye diameter 17.2-22.9;
gill slit length 28.1-44.3; interorbital width 5.9-
9.0; interpupillary width 24.0-29.6; pelvic fin
length 8.0-8.8 (in two of eight). Pectoral base/
length ratio 32.8^6.3.
DIAGNOSIS.— Pelvic fins present or absent; sin-
gle, mesial occipital pore present: postorbital
pores 3-4; mediolateral lateral line originating
slightly anterior to vertical through posterior
margin of pectoral fin; ventral lateral line origi-
nating just posterior to last postorbital pore; ver-
tebrae 24-26 + 78-87 = 103-113.
DESCRIPTION.— Following description based on
one adult male (holotype), one adult female
(MNHNC 6653), three juvenile males, and three
juvenile females. Head ovoid in juveniles and
adult female, more triangular in adult male. Male
with dorsoventrally depressed head and straight-
er snout than others; adult female and juveniles
with bluntly rounded snouts. Scales extending
anteriorly to vertical through dorsal fin origin,
or slightly anterior to it; scales absent on nape
(rare or absent anterior to line connecting upper
edge of gill slit and dorsal fin origin). Scales pres-
ent on unpaired fins (almost to their margin pos-
teriorly), abdomen to isthmus, in pectoral axil
and on pectoral base except adult female and one
juvenile male. Eye ovoid, more rounded in adult
male, entering dorsal profile of head except in
adult female; eye diameter 17.2% HL in adult
male, 20.2-22.9% HL in others. Gill slit short,
not reaching ventral margin of pectoral base ex-
cept in adult male, in which it reaches slightly
beyond lowest pectoral ray; gill slit length 42.2%
HL in male, 28.1-34.0% HL in juveniles, and
44.3% HL in adult female owing to its anterior
extension (below). Lobe at dorsal margin of gill
slit small, squared-off in holotype and some ju-
veniles; dorsal margin not extending anteriorly
in some juveniles; operculum broadly rounded
and extending anteriorly almost one eye diam-
eter in adult female. Pectoral fin origin on body
midline in smallest juveniles, well below midline
in adults, insertion on abdomen; posterior mar-
gin of fin usually evenly rounded, middle rays
longest, however, female with left fin distinctly
wedge-shaped and right fin injured, with tips of
upper and middle rays missing (fin deeply
notched). Ventralmost four or five pectoral rays
thickened, tips slightly exserted.
238
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 10
Mouth subterminal, oblique, snout and upper
jaw more produced in adult male than others;
upper jaw extending posteriorly to rear margin
of eye in adult male (upper jaw length 5 1 .7% HL),
or to middle of eye in others (37. 1-42. 1% HL in
adult female and juveniles). No epidermal pric-
kles evident on snout or lips. Oral valve weak,
not reaching anterior edge of vomer in adults,
but not coalesced with lateral margin of palate
opposite vomer in any specimen. Jaw teeth in
adult male long, caniniform, slightly retrorse (Fig.
1 3); dentary with three irregular rows anteriorly,
blending into single, posterior row; premaxilla
with two irregular anterior rows, blending into
single, posterior row; outer row on premaxilla
about five times as large as inner row. Jaw teeth
in adult female and juveniles small, blunt, con-
ical; dentary with four irregular rows anteriorly
(juveniles with 2-3 rows), blending into single,
posterior row; premaxilla with three irregular an-
terior rows (juveniles with two), blending into
single, posterior row. Vomerine teeth in irregular
patch or gradual arc (holotype). Vomerine and
palatine teeth retrorse and caniniform in adult
male, small, blunt and conical in others; palatine
teeth in irregular, double row in adult female, in
single row in others.
Cephalic lateralis system with three postor-
bital pores arising from frontal (pore one), be-
tween pterotic and lateral extrascapular (pore
three), and just posterior to lateral extrascapular
(pore four) except in adult female in which pore
two is present on both sides, arising from pterot-
ic. Two pairs of anterior supraorbital (nasal)
pores, one set mesial to nasal tube, the other
posteriorly (posteromesially in adult male). Usu-
ally seven suborbital pores, with six arising from
ventral ramus of bone chain and one from as-
cending ramus behind eye. Exceptions to this are:
(1) LACM 43717-1, 192 mm, which has five
pores arising from ventral ramus and one from
ascending ramus on right side, and five from
ventral ramus and two from ascending ramus on
left side; and (2) LACM 437 1 7- 1 , 1 72 mm, which
has suborbital pores two and four oddly doubled
(pore at these positions are situated one atop the
other) and no third pore, giving seven pores aris-
ing from ventral ramus and one from ascending
ramus. Eight preoperculomanibular pores, four
arising from dentary, one from anguloarticular
and three from preopercle. Interorbital pores ab-
sent. Occipital (supratemporal) canal with single,
mesial pore in all. Body lateral line with medio-
lateral branch originating on vertical about one
eye diameter anterior to posterior margin of pec-
toral fin; mediolateral branch not visible in great-
ly faded juveniles in LACM 43717-1; ventral
branch originating just posterior to fourth post-
orbital pore, gradually descending across abdo-
men to above anal fin base; both branches com-
plete to tail tip.
Dorsal fin origin associated with vertebrae 4-
5. Anal fin origin associated with ultimate pre-
caudal vertebra (numbers 24-26), with 3-4 anal
pterygiophores inserted anterior to haemal spine
of first caudal vertebra. Last dorsal ray associated
with second, fourth, or fifth preural vertebrae,
last anal ray associated with second preural ver-
tebra. Caudal fin with 1-2 epural, four upper
hypural, and 4-5 lower hypural rays. Pelvic fins
absent except in MNHNC 6653 and ZIN 48394
which have small, fleshy fins of three rays each.
Gill rakers on upper limb (epibranchial) of first
arch moderately long, sharply pointed; ventral-
most rakers on lower limb (ceratobranchial) of
adult female and juveniles small, but sharply
pointed, those of adult male blunt, with squared-
off tips. Branchiostegal rays six. Pseudobranch
filaments long, numbering 3—4.
Recently preserved specimens (collected 1978-
1986) uniformly dark brown in alcohol, probably
black in life. Eye blue. Peritoneum, lining of gill
cavity and inner palate black. Outer palate and
lips pale.
DISTRIBUTION.— Known from eight specimens
taken off western South America from the upper
slope to the edge of the Peru-Chile Trench, re-
liably known from depths of 760-1,830 m (Fig.
11). Capture depth of holotype unknown; from
fisheries vessel.
ETYMOLOGY.— From the Greek n'taos (middle)
and iropos (hole, pore) in reference to the unique
condition of the species' occipital canal with its
single, mesial pore.
REMARKS.— This is the second species of
Pachycara known in which pelvic fins are present
or absent (the other is P. sulaki) and the fourth
zoarcid reported with this condition (others are
Crossostomus chilensis [see Gosztonyi 1 977] and
Lycenchelys monstrosa [see Anderson 1982b]).
The two specimens with pelvic fins (ZIN 48394
and MNHNC 6653) also exhibit slightly elevated
vertebral counts (111 and 113 vs. 103-106 for
others), higher gill raker counts (16 and 18 vs.
ANDERSON: REVIEW OF EELPOUT GENUS PACHYCARA
239
FIGURE 14. Pachycara pammelas n. sp., CAS 62407, holotype, 336 mm SL, from off Punta Lobos, Chile.
13-15 for others), and higher pectoral ray counts
(16 and 18 vs. 15-16 for others). It was tempting
to consider the presence of two closely related
species with mesial occipital pores in the sample
of eight specimens. However, this unique con-
dition, plus the identical lateral lines, squama-
tion, and state of dentition, makes it apparent
that the noted anatomical differences cannot be
considered above the range of individual vari-
ation found in either other Pachycara species
represented by larger sample sizes, or other zoar-
cids.
Pachycara pammelas n. sp.
(Figs. 11, 14)
HOLOTYPE.— CAS 62407 (female, 336 mm); off Punta Lobos,
Chile (20°50'S, 70°27'W); Chilean fisheries vessel; trawl, 610
m; I. Kong Urbina, 21 Feb. 1981.
PARATYPES.-SIO 72-183 (2; 96-192 mm); off Arica, Chile
(18°44.7^t0.2'S, 70-40.7-35. 1'W); South Tow Exped., 1,097-
l,152m;R. L. Wisner and party, 1 152-1 330 hr, 7 May 1972.
LACM 43732-1 (2; 152-217 mm); off Coquimbo, Chile
(30°06.0'S, 71°42.6'W); ANTON BRUUN sta. 708; free ve-
hicle set line trap, 1 ,000 m; L. W. Knapp and party, 1 4 Aug.
1966. LACM 43726-1 (2; 226-231 mm); off Valpariso, Chile
(33°39.0'S, 72°09.5'W); ANTON BRUUN sta. LWK 66-41;
trawl, 1,170-1,480 m; L. W. Knapp and party, 10 Aug. 1966.
COUNTS AND MEASUREMENTS.— Vertebrae 23-
26 + 81-88 = 105-1 14; D 99-109; A 83-92; C
10-1 1; P, 15-18; P2 3; vomerine teeth 3-9; pal-
atine teeth 2-14; gill rakers 1-3 + 12-17 = 14-
20; branchiostegal rays six; pseudobranch fila-
ments 2-5. Following measurements in percent
SL: head length 12.8-15.6; head width 5.7-9.0;
pectoral fin length 7.1-9.2 predorsal length 1 5.0-
18.4; preanal length 34.0-37.0; body height 6.3-
10.5; gill slit length 4.0-6.4; caudal fin length 1 .6-
2.7. Following measurements in percent HL: head
width 39.9-67.4; upper jaw length 34.1-52.7;
pectoral fin length 50.8-64.2; snout length 15.6-
22.2; eye diameter 19.5-30.4; gill slit length 25.7-
50.2; interorbital width 6.7-8.9; interpupillary
width 23.7-33.2; pelvic fin length 7.4-13.3. Pec-
toral base/length ratio 33.0-51.4.
DIAGNOSIS.— Pelvic fins present; origin of me-
diolateral lateral line immediately posterodorsal
last postorbital pore; origin of ventral lateral line
immediately posterior to last postorbital pore;
vertebrae 23-26 + 81-88 = 105-1 14; no scales
on nape.
DESCRIPTION.— Following description based on
one adult female (holotype), two juvenile males,
and four juvenile females. Head ovoid, snout
bluntly rounded in females giving head more
rounded appearance than males. Males (1 52 and
1 92 mm) with head elongate, dorsoventrally de-
pressed, snout long, gently tapering. Scales ex-
tending anteriorly on body to nape where present
in patch extending in front of dorsal fin origin in
holotype (absent there in smaller paratypes);
scales present on unpaired fins (except 96 mm
juvenile), abdomen, pectoral axil and base (latter
in holotype only). Eye ovoid, entering dorsal pro-
file of head. Gill slit short, extending just to lower
margin of pectoral base or slightly above. Lobe
at dorsal margin of gill slit weak, gill slit not
extending anteriorly in largest male; lobe round-
ed or squared-off. Pectoral fin origin well below
body midline in larger specimens (closer to mid-
line in smallest two), insertion on abdomen; pos-
terior margin of fin evenly rounded, middle rays
longest; ventralmost four or five rays thickened,
tips very slightly exserted.
Mouth subterminal, oblique, snout more pro-
duced in males than females; upper jaw extend-
ing posteriorly to middle of eye in females, to
just beyond its posterior margin in largest male.
No epidermal prickles evident on snout or lips.
Oral valve well developed, overhanging anterior
margin of vomer in holotype, coalesced with sides
of palate posterior to vomer in all. Jaw teeth
small, blunt, conical; dentary with four irregular
rows anteriorly in holotype, blending into single,
posterior row consisting of 2-4 teeth; dentary
teeth in two to four (largest two paratypes, LACM
43726- 1) irregular anterior rows in others, blend-
ing into single, posterior row. Premaxilla with
two anterior rows of teeth, blending into single,
240
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 10
posterior row except in smallest specimen (SIO
72-183, 96 mm) which has single row of nine
teeth. Vomerine teeth in patch or arc (small spec-
imens); palatine teeth blunt, not retrorse, in dou-
ble row in holotype, single row in others.
Cephalic lateralis system with three postor-
bital pores arising from frontal (pore one), be-
tween pterotic and lateral extrascapular (pore
three), and posterior to lateral extrascapular (pore
four). Largest male (SIO 72-183, 192 mm) with
pores one and four only. Two pairs of anterior
supraorbital (nasal) pores, one set mesial to nasal
tube, the other posteriorly. Seven suborbital
pores, six arising from ventral ramus of bone
chain and one from ascending ramus behind eye
just ventral to first postorbital pore; eight sub-
orbital pores on right side only in LACM 43732- 1 ,
152 mm). Eight preoperculomandibular pores,
four arising from dentary, one from anguloartic-
ular, and three from preopercle. Interorbital and
occipital pores absent. Body lateral line difficult
to detect in faded paratypes, clear in holotype;
mediolateral branch originating about half an eye
diameter or less posterodorsal to last (number
four) postorbital pore; ventral branch originating
immediately posterior to postorbital pore four;
both branches complete to tail tip.
Dorsal fin origin associated with vertebra 4-
5. Anal fin origin associated with ultimate or
penultimate precaudal vertebrae (numbers 23-
26), with 2-4 anal pterygiophores inserted an-
terior to haemal spine of first caudal vertebra.
Last dorsal ray associated with fourth preural
vertebra, last anal ray associated with second
preural vertebra. Caudal fin with two epural, 4-
5 upper hypural and four lower hypural rays.
Gill rakers on upper limb (epibranchial) of first
arch simple, sharply pointed; rakers on lower
limb (ceratobranchial) laterally compressed, with
squared-off tips; six lower limb rakers in holo-
type with shallowly bicuspid tips. Branchiostegal
rays six. Pseudobranch filaments 2-5, minute in
smallest specimen.
Holotype uniformly black (paratypes faded),
eye blue. Peritoneum, gill cavity and inner palate
black. Outer palate and inner surface of lips pale.
Adult female with ripe ova measuring 3.9-4.2
mm in diameter (not counted).
DISTRIBUTION. — Known from seven speci-
mens captured along the edge of the Peru-Chile
Trench off Chile at depths of 610-1,480 m (Fig.
11).
ETYMOLOGY.— From the Greek iran (an inten-
sifier meaning "all") and ntXas (black), a noun
in apposition, alluding to the coloration of the
species.
Pachycara brachycephalum (Pappenheim, 1912)
Lycodes brachycephalus Pappenheim, 1912:179-180, pi. X,
fig. 3.
Austrolycichthys brachycephalus (Pappenheim). Regan, 1913:
244-245, fig. 2; Waite, 1 9 1 6: 1 5, pi. I, text fig. 2, 2A; Norman,
1938:81, 83, 84.
Pachycara brachycephalus (Pappenheim). Anderson, 1988a:
74-77, figs. 16-19.
MATERIAL EXAMINED.— See Anderson (1988a:74-75).
LECTOTYPE (herein designated).— ZMB 18929 (male, 169
mm); Wilhelm Land, Antarctica; GAUSS winter station; fish
traps, 385 m; Deutsche Sudpolar Exped., 1902.
DIAGNOSIS.— Pelvic fins present; mediolateral
lateral line originating just posterior to pectoral
fin margin; ventral lateral line originating above
dorsal edge of gill slit; scales absent on nape;
vertebrae 23-27 + 71-86 = 95-1 13; D 88-107;
A 72-90; P 16-18; dorsal fin origin associated
with vertebrae 3-7; gill rakers 0-3 + 9-13 = 9-
16; branchiostegal rays six; suborbital pores 5 +
1 (5 + 0 in one); postorbital pores 1-2; head
length 14.0-19.9% SL; pectoral fin length 10.0-
12.7%SL.
DISTRIBUTION.— Coastal Antarctica in inner
basins and on outer slope at depths of 200-
1,810m.
DISCUSSION
So few specimens of Pachycara are known for
most species that it is difficult to consider the
evolutionary history of the genus. Phylogenet-
ically, Pachycara seems to have arisen as a deep-
slope sister taxon or precursor of Lycodes, dif-
ferentiated by its reduced parasphenoid wing
(apomorphy), retention of the oral valve and ce-
phalic pores, and lack of developed mental crests
(plesiomorphies; see Anderson 1984). Ther-
marces is derived with respect to Pachycara in
four reductive characters: ( 1 ) absence of pelvic
bones; (2) absence of scales; (3) absence of lateral
lines; and (4) gelatinous flesh.
The present distributions of Pachycara species
indicate two main historical biogeographic tracks:
(1) North Atlantic-eastern Pacific; and (2) west-
ern South America-Antarctica (P. brachycepha-
ANDERSON: REVIEW OF EELPOUT GENUS PACHYCARA
241
/wm). With only one specimen known, the origin
and distribution of P. shcherbachevi (Indian
Ocean) remains a mystery, but it appears closer
to P. bulbiceps than the two West African species
in its high vertebral count, small head, and re-
trograde dorsal fin origin. Recently, I reviewed
the mid-Miocene tectonic history of the Amer-
ican tropics as it relates to the distribution of
deep-sea bottom fishes (Anderson 1988a), and
Pachycara in particular (Anderson and Peden
1988). I concluded that the closing of the Pan-
amanian seaway to deep-demersal fishes about
7-10 million years ago did not sunder Atlantic
and Pacific stocks of P. bulbiceps (if there were
any), and that the present two populations pos-
sibly exchange genes through as yet unsampled
South Pacific and South Atlantic populations
(Anderson and Peden 1988). This type of distri-
bution (Fig. 5) invites speculation on the origin
and relationships of P. sulaki, presently known
only from the western tropical Atlantic. This
species expresses states of three characters that
link it with southeastern Pacific forms (P. me-
soporum and P. rimae) that may have phyloge-
netic utility: (1) possession of only five suborbital
pores along the ventral ramus in some individ-
uals; (2) dorsal fin origin retrograde, associated
with vertebrae six to seven (P. sulaki) or eight
(P. rimae); and (3) absence of pelvic fins in some
individuals (P. sulaki and P. mesoporuni). How-
ever, a phylogenetic analysis of Pachycara is pre-
cluded by uncertainty about the extent of distri-
bution and the variability of many character
states, particularly of the axial skeleton, denti-
tion, cephalic pores, and the occasional lack of
pelvic fins. This, in turn, produces a cautionary
appraisal of the specific status of the present sam-
ples of P. bulbiceps (Pacific and Atlantic popu-
lations) and P. mesoporum (range of variation)
that only future sampling may clarify.
ACKNOWLEDGMENTS
For specimen donations or loans, photo-
graphs, and data compilation I am grateful for
the help of C. Allue, A. P. Andriashev, M.-L.
Bauchot, D. M. Cohen, A. Cornejo C., J. F. Gras-
sle, K. E. Hartel, D. A. Hensley, P. A. Hulley,
S. Jewett, S. R. Johnson, I. Kong U., C. Karrer,
R. J. Lavenberg, D. F. Markle, R. Melendez, N.
R. Merrett, A. V. Neyelov, J. G. Nielsen, W.G.
Pearcy, A. E. Peden, B. Ranchod, M. D. Rich-
ardson, C. R. Robins, R. H. Rosenblatt, J. A.
Seigel, B. Seret, C. R. Smith, D. L. Stein, H. J.
Walker, and R. R. Wilson. W. I. Follett provided
financial assistance for part of this study. K. Klitz
rendered the specimen illustrations (Figs. 7, 9,
12-14).
LITERATURE CITED
ANDERSON, M. E. 1982a. Revision of the fish genera Gym-
nelus Reinhardt and Gymnelopsis Soldatov (Zoarcidae), with
two new species and comparative osteology of Gymnelus
viridis. Natl. Mus. Nat. Sci., Publ. Zool. (17): 1-76.
. 1982b. Biological results of the University of Miami
Deep-Sea Expeditions. 1 36. A new eelpout (Teleostei: Zoar-
cidae) from the eastern tropical Pacific Ocean. Bull. Mar.
Sci. 32(1):207-212.
-. 1 984. On the anatomy and phylogeny of the Zoar-
cidae (Teleostei: Perciformes). Ph.D. Dissertation, College
of William and Mary, Williamsburg, Virginia. 254 pp.
-. 1 986. Family no. 95: Parabrotulidae. P. 343 in Smiths'
sea fishes. M. M. Smith and P. C. Heemstra, eds. MacMillan
South Africa, Ltd.
. 1988a. Studies on the Zoarcidae of the southern
hemisphere. I. The Antarctic and subantarctic regions. Ant.
Res. Ser. 47:59-113.
. 1 988b. A new genus of California eelpout (Teleostei:
Zoarcidae) based on Maynea californica Starks and Mann,
1911. Proc. Calif. Acad. Sci 45(5):89-96.
ANDERSON, M. E., R. E. CRABTREE, H. J. CARTER, K. J. SULAK,
AND M. D. RICHARDSON. 1986. Distribution of demersal
fishes of the Caribbean Sea found below 2,000 meters. Bull.
Mar. Sci. 37(3):794-807.
ANDERSON, M. E. AND C. L. HUBBS. 1981. Redescription and
osteology of the northeastern Pacific fish Derepodichthys
alepidotus (Zoarcidae). Copeia 1 98 1(2):34 1-352.
ANDERSON, M. E. AND A. E. PEDEN. 1988. The eelpout genus
Pachycara (Teleostei: Zoarcicdae) in the northeastern Pacific
Ocean, with descriptions of two new species. Proc. Calif.
Acad. Sci. 46(3):83-94.
ANDRIASHEV, A. P. 1954. Fishes of the northern seas of the
USSR. Zool. Inst, Acad. Sci., Contr. Fauna USSR 53:1-
566.
. 1973. Zoarcidae. In Check-list of the fishes of the
north-eastern Atlantic and of the Meditrranean. J.-C. Hu-
reau and T. Monod, eds. Unesco, Paris. 1:540-547.
-. 1986. Zoarcidae (including Lycodidae). In Fishes of
the northeastern Atlantic and the Mediterranean. P. J. P.
Whitehead, M.-L. Bauchot, J.-C. Hureau, J. Nielsen, and E.
Tortonese, eds. Unesco, Paris. 3:1 130-1 150.
ARNULF, I., F. J. MEUNIER, AND P. GEISTDOERFER. 1987. Os-
teologiede Thermarces cerberus Rosenblatt and Cohen, 1986,
Zoarcidae des sources hydrothermales du Pacifique Est, sui-
vie d'une discussion sur sa classification. Cybium 1 1 (2): 141-
158.
COHEN, D. M. AND R. L. HAEDRICH. 1983. The fish fauna of
the Galapagos thermal vent region. Deep-Sea Res. 30(4A):
371-379.
COHEN, D. M., R. H. ROSENBLATT, AND R. L. HAEDRICH. 1985.
242
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 10
Identity of thermal vent fishes in the eastern Pacific: an
interim report. Biol. Soc. Wash., Bull. 6:229-230.
DINGERKUS, G. AND L. UHLER. 1977. Enzyme clearing of
alcian blue stained whole small vertebrates for demonstra-
tion of cartilage. Stain Tech. 52(4):229-232.
FOWLER, H. W. 1 936. The marine fishes of west Africa. Bull.
Amer. Mus. Nat. Hist. 70(2):607-1493.
CARMAN, S. 1 899. Reports of an exploration off the west
coasts of Mexico, Central and South America, and off the
Galapagos Islands, in charge of Alexander Agassiz, by the
ALBATROSS during 1891. XXVI. The fishes. Mems. Mus.
Comp. Zool. 24:1-431.
GOLOVAN', G. G. 1 974. Preliminary data on the composition
and distribution of the bathyal ichthyofauna (in the Cap
Blanc area). Oceanol., Akad. Nauk SSSR 14(2):288-290.
. 1978. Composition and distribution of the ichthy-
ofauna of the continental slope of northwestern Africa. Tru-
dy Inst. Oceanol. 1 1 1:195-258.
GoszTONYi, A. E. 1977. Results of the research cruises of
FRV "Walther Herwig" to South America. XLVIII. Revi-
sion of the South American Zoarcidae (Osteichthyes, Blen-
nioidei), with the description of three new genera and five
new species. Arch. FischWiss. 27(3): 19 1-249.
GUNTHER, A. 1 880. Report on the shore fishes procured dur-
ing the voyage of H.M.S. "Challenger" in the years 1873-
1876. Rep. Sci. Res. Challenger 1(6): 1-82.
HAEDRICH, R. L. AND N. R. MERRETT. 1 988. Summary atlas
of deep living demersal fishes in the North Atlantic Basin.
J.Nat. Hist. 22:1325-1362.
HAEDRICH, R. L., G. T. ROWE, AND P. T. POLLONI. 1980. The
megabenthic fauna in the deep sea south of New England,
USA. Mar. Biol. 57:165-179.
HUBBS, C. L., W. I. FOLLETT, AND L. J. DEMPSTER. 1979. List
of the fishes of California. Occ. Pap. Calif. Acad. Sci. (133):
1-51.
JENSEN, A. S. 1902. Ichthyologiske studier. HI. Om nogle ny
arter af slaegten Lycodes. Vidensk. Meddel. Naturh. Foren.
Kbhvn. 1901:205-214.
JORDAN, D. S. 1920. The genera of fishes, part IV. Stanf.
Univ. Publ., Univ. Ser. (43):41 1-576.
LEVTTON, A. E. AND R. H. GIBBS, JR. 1988. Standards in
herpetology and ichthyology. Standard symbolic codes for
institution resource collections in herpetology and ichthy-
ology. Supplement No. 1 : additions and corrections. Copeia
1988(l):280-282.
LEVITON, A. E., R. H. GIBBS, JR., E. HEAL, AND C. E. DAWSON.
1985. Standards in herpetology and ichthyology: part I.
Standard symbolic codes for institutional resource collec-
tions in herpetology and ichthyology. Copeia 1985(3):802-
832.
MARKLE, D. F. 1 980. A new species and a review of the deep-
sea fish genus Asquamiceps (Salmoniformes, Alepocephali-
dae). Bull. Mar. Sci. 30(l):45-53.
MARKLE, D. F. AND G. R. SEDBERRY. 1978. A second spec-
imen of the deep-sea fish, Pachycara obesa, with a discussion
of its classification and a checklist of other Zoarcidae off
Virginia. Copeia 1978(l):22-25.
MCALLISTER, D. E. AND E. I. S. REES. 1964. A revision of
the eelpout genus Melanostigma, with a new genus and with
comments on Maynea. Natl. Mus. Canada, Bull. 199:85-
109.
MERRETT, N. R. AND N. B. MARSHALL. 1981. Observations
on the ecology of deep-sea bottom living fishes collected off
northwest Africa (08°-27°N). Prog. Oceanogr. 9:185-244.
MUSICK, J. A. 1979. Community structure of fishes on the
continental slope and rise off the middle Atlantic coast of
the United States. Va. Inst. Mar. Sci., Spec. Sci. Rept. (96):
1-53.
NIELSEN, J. G. 1973. Parabrotulidae. In Check-list of the
fishes of the north-eastern Atlantic and of the Mediterra-
nean. J.-C. Hureau and T. Monod, eds. Unesco, Paris. 1:
548-549.
. 1986. Parabrotulidae. In Fishes of the north-eastern
Atlantic and the Mediterranean. P. J. P. Whitehead, M.-L.
Bauchot, J.-C. Hureau, J. Nielsen, and E. Tortonese, eds.
Unesco, Paris. 3:1151-1152.
NORMAN, J. R. 1938. Coast fishes, pt. HI. The Antarctic
region. Discovery Repts. 18:1-104.
PAPPENHEIM, P. 1912. Die Fische der Deutschen Siidpolar-
Expedition 1901-1903. 1. Die Fische der Antarktis und Su-
bantarktis. Deutsche Sudpolar Exped. 13:160-182.
PAULL, C. K., B. HECKER, R. COMMEAU, R. P. FREE-
MAN-LYNDE, C. NEUMANN, W. P. CORSO, S. GOLUBIE, J. E.
HOOK, E. SIKES, AND J. CURRAY. 1984. Biological com-
munities at the Florida Escarpment resemble hydrothermal
vent taxa. Science 226(4677):965-967.
PEARCY, W. G., D. L. STEIN, AND R. S. CARNEY. 1982. The
deep-sea benthic fish fauna of the northeastern Pacific Ocean
on Cascadia and Tufts Abyssal Plains and adjoining conti-
nental slopes. Biol. Oceanogr. l(4):375-428.
REGAN, C. T. 1913. The Antarctic fishes of the Scottish Na-
tional Antarctic Expedition. Trans. Royal Soc. Edinburgh
49(2):229-292.
RICHARDSON, J. 1855. Account of the fish. In Last of the
Arctic voyages. E. Belcher, ed. 2:1-30.
ROSENBLATT, R. H. AND D. M. COHEN. 1986. Fishes living
in deepsea thermal vents in the tropical eastern Pacific, with
descriptions of a new genus and two new species of eelpouts
(Zoarcidae). Trans. San Diego Soc. Nat. Hist. 21(4):71-79.
ROULE, L. 1916. Notice preliminaire sur quelques especes
nouvelles ou rares des poissons provenant de croisieres de
S.A.S. le Prince de Monaco. Bull. Inst. Oceanogr. Monaco
(320): 1-32.
. 1919. Poissons provenant des campagnes du yacht
"Princesse Alice" (1891-1913) et du yacht "Hirondelle II"
(1914). Result. Camp. Sci. Prince Albert I 52:1-191.
SPRINGER, V. G. AND W. C. FRIEHOFER. 1976. Study of the
monotypic fish family Pholidichthyidae (Perciformes).
Smithson. Contr. Zool. (2 16): 1^*3.
VAILLANT, L. 1888. Expeditions scientifiques du Travailleur
etdu Talisman pendeant les annees 1880, 1881, 1882, 1883.
Poissons. G. Masson, Paris. 406 pp.
WATTE, E. R. 1916. Fishes. Sci. Rept., ser. C, Australasian
Ant. Exped. 1911-1914. 3(l):l-92.
WHTTLEY, G. P. 1931. New names for Australian fishes. Austr.
Zool. 6(4):3 10-334.
ZUGMAYER, E. 191 la. Diagnoses des poissons nouveaux
provenant des campagnes du yacht "Princesse Alice" ( 1 90 1-
1910). Bol. Inst. Oceanogr., Monaco (193):1-14.
. 1 9 1 1 b. Poissons provenant des campagnes du yacht
"Princesse Alice." Result. Camp. Sci. Prince Albert I 35:1-
174.
CALIFORNIA ACADEMY OF SCIENCES
Golden Gate Park
San Francisco, California 94118
PROCEEDINGS
OF THE
CALIFORNIA ACADEMY OF SCIENCES
Vol. 46, No. 11, pp. 243-277, 13 figs., 3 tables.
December 20, 1989
A PHYLOGENETIC REVISION OF THE PHALLOSTETHID FISHES
(ATHERINOMORPHA, PHALLOSTETHIDAE)
By
Lynne R. Parent!
California Academy of Sciences, Golden Gate Park,
San Francisco, California 94118
ABSTRACT: The Phallostethidae (including Neostethidae) comprises a group of small to minute atherino-
morph fishes distinguished from all other teleosts by the presence in males of a complex, bilaterally asymmetric,
subcephalic copulatory organ, the priapium. I recognize 19 species in four genera: Phallostethus Regan, 1913,
Neostethus Regan, 1916, Gulaphallus Herre, 1925, and Phenacostethus Myers, 1928. This includes one new
species, Neostethus robertsi, from Luzon. Phallostethids are widely distributed throughout coastal brackish
waters from Thailand, Singapore, Malaysia, and the Philippines, as well as inland freshwater streams of
Luzon and Borneo.
The three objectives of this review are to: (1) describe the anatomy and variation of the priapium and
propose synonymy and probable homology of its many and varied osteological and myological modifications;
(2) present taxonomic synonymies and diagnoses of the 19 species in four genera; and (3) hypothesize phy-
logenetic relationships among the species by grouping into higher categories based primarily on the presence
of shared derived characters of the priapium.
Received November 18, 1988. Accepted September 2, 1989.
TABLE OF CONTENTS
Abstract
Introduction
Materials and Methods
Relationships of Phallostethids to other
Atherinomorphs
Synonymy and Homology of Priapial
Structures
Phylogenetic Analysis
Key to Phallostethidae
Classification of Phallostethidae
Systematic Accounts
Family Phallostethidae
Subfamily Phallostethinae
Tribe Phallostethini
Genus Phallostethus Regan, 1913
Phallostethus dunckeri Regan,
1913 267
243 Genus Phenacostethus Myers, 1928 267
244 Phenacostethus smithi
245 Myers, 1928 268
Phenacostethus posthon
247 Roberts, 1971 268
Phenacostethus trewavasae
248 Parenti, 1986 268
255 Tribe Neostethini 269
264 Genus Neostethus Regan, 1916 ... 269
265 Neostethus lankesteri
266 Regan, 1916 269
266 Neostethus bicornis
267 Regan, 1916 269
267 Neostethus amaricola
267 (Villadolid and Manacop, 1935) 270
[243]
244
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1 1
Neostethus borneensis
Herre, 1939 270
Neostethus villadolidi
Herre, 1942 271
Neostethus zamboangae
Herre, 1942 271
Neostethus palawanensis
(Myers, 1935) 271
Neostethus thessa (\urich, 1937) 272
Neostethus ctenophorus
(Aurich, 1937) 272
Neostethus robertsi, new species 272
Subfamily Gulaphallinae 273
Tribe Gulaphallini 273
Genus Gulaphallus Herre, 1925 ... 273
Gulaphallus eximius
Herre, 1925 274
Gulaphallus mirabilis
Herre, 1925 274
Gulaphallus falcifer
Manacop, 1936 274
Gulaphallus bikolanus
(Herre, 1926) 275
Gulaphallus panayensis
(Herre, 1942) 275
Conclusions 275
Acknowledgments 276
Literature Cited 276
INTRODUCTION
Regan (1913) named the first known phallo-
stethid species Phallostethus dunckeri in honor
of Georg Duncker (1904), who first reported this
unusual group of fishes from Johore on the Malay
Peninsula. During the following three decades,
attention focused on description of new species
and genera (e.g., Regan 1916; Herre 1925, 1926;
Myers 1928; Villadolid and Manacop 1935;
Manacop 1936), anatomy of the complex cop-
ulatory organ, the priapium (e.g., Bailey 1936;
Aurich 1937; TeWinkel 1939; Woltereck 1942a,
b), and consideration of the relationships of phal-
lostethids to other fishes (e.g., Smith 1 927; Myers
1935; Hubbs 1944).
Presence of a priapium led to speculation that
phallostethids must be live bearers (e.g., Regan
1913). However, Smith (1927) observed female
phallostethids in Thailand depositing eggs. Vil-
ladolid and Manacop (1935) confirmed that fe-
males of Gulaphallus mirabilis lay fertilized eggs
following copulation. Structure of ovaries and
testes, presence of sperm bundles (spermato-
zeugmata), and mode of internal fertilization are
currently under investigation (Grier and Parenti,
in prep.).
Early descriptions of phallostethids followed
the convention of naming morphologically dis-
tinctive populations not just as new species but
as new genera as well. In the last comprehensive
review of phallostethids, Herre (1942) recog-
nized 18 species classified in 10 genera, one of
which had two subgenera (Table 1).
Interest in phallostethid systematics was re-
newed with description of a new species, Phena-
costethus posthon Roberts, 197 la and a detailed
osteological study of Neostethus bicornis Regan
(formerly in Ceratostethus) (Roberts 1 97 1 b). De-
spite extensive work by Roberts, numerous ques-
tions remained concerning the relationships of
phallostethids to other atherinomorphs, limits
of, and relationships among, genera, and total
number of recognizable species. It was with the
goal of answering these questions that I began
my studies of phallostethids.
A preliminary review resulted in detailed os-
teology of Phallostethus dunckeri and the pro-
posal that phallostethids are the sister group of
the Indo-Pacific atherinoid, Dentatherina mer-
ceri Patten and Ivantsoff, 1983 (Parenti 1984). I
described a new species, Phenacostethus trewa-
vasae Parenti, 1 986a and brought the total num-
ber of then recognized species to 20. Homology
of priapial structures, always a challenge to stu-
dents of phallostethids, was considered by ex-
amining development of the pelvic fins of one
species, Gulaphallus falcifer Manacop (formerly
in Manacopus), in which there are bilaterally
asymmetric adult females as well as males (Pa-
renti 1986b). I (Parenti 1986c) reviewed the as-
sociation between bones and teeth in teleost fish-
es because in some phallostethid species there is
a perichondrally ossified lower jaw bone, the
paradentary, that may possess teeth.
Much of the information compiled on phal-
lostethid fishes has been inaccessible to, or ig-
nored by, systematists because of the array of
peculiar names applied to parts of the complex
priapium and because of the small size of phal-
lostethids. Further, no previous classifications
summarize cladistic relationships adequately;
rather, they emphasize gross phenetic differences
among species.
Phallostethid fishes are distributed throughout
coastal mainland and insular Southeast Asia from
Thailand to the Philippines (Fig. 1), entirely west
PARENTI: PHALLOSTETHID FISHES
245
of Wallace's Line, one of several hypothetical
boundaries separating the Asian from the Aus-
tralian biota (see Whitmore 1987). Cladistic re-
lationships of phallostethids can be used to pre-
sent a working hypothesis of the distributional
history of one segment of the Indo-Australian
biota (see methodology in Humphries and Pa-
renti 1986). Results from this study will be com-
bined with those of other analyses to prepare a
historical biogeographic analysis of the Indo-
Australian atherinomorph fishes (Parenti, in
prep.).
This review has three objectives: (1) to de-
scribe the anatomy and variation of the priapium
and propose synonymy and probable homology
of its numerous osteological and myological
modifications; (2) to present taxonomic synon-
ymies and diagnoses of the herein recognized 1 9
species, with one described as new, classified in
four genera; and (3) to hypothesize relationships
among the species, grouping them into higher
categories, primarily on synapomorphies of the
priapium.
MATERIALS AND METHODS
I observed osteological characters on speci-
mens counterstained with alcian blue and aliza-
rin red S (Dingerkus and Uhler 1977) or stained
solely with alizarin. I examined specimens with
a Zeiss SV8 dissecting microscope and recorded
data using a drawing tube and photomicrography
attachments. Representatives of 1 8 of the 1 9 cur-
rently recognized phallostethid species were ex-
amined as part of my recent studies. No speci-
mens of Neostethus ctenophorus (Aurich, 1937)
were available; characters are from the original
description.
Meristic data (except for scale counts) were
tabulated from cleared and stained specimens.
In recording meristic data (Table 2), the terminal
half-centrum is counted as one vertebra, and all
fin spines and rays are counted, including, for
example, the last two anal rays which articulate
with just one pterygiophore.
Institutional abbreviations follow the Stan-
dard Symbolic Codes for Institutional Research
Collections in Herpetology and Ichthyology
(Leviton et al. 1985). The following abbrevia-
tions are defined: Dist., district; Is., island; Prov.,
province; R., river; and SL, standard length. Cat-
alog numbers for phallostethid material exam-
ined are given in the text.
TABLE 1. Comparison of Herre's (1942) classification of
Phallostethids with that presented herein.
Herre(1942)
Present paper
Phallostethus dunckeri
Phenacostethus smithi
Mirophallus bikolanus
Solenophallus thessa
Gulaphallus eximius
Gulaphallus mirabilis
Ceratostethus bicornis
Manacopus falcifer
Neostethus (Neostethus)
amaricola
Neostethus (Neostethus)
lankesteri
Neostethus (Neostethus)
siamensis
Neostethus (Neostethus)
villadolidi
Neostethus (Sandakanus)
borneensis
Neostethus (Sandakanus)
coronensis
Neostethus (Sandakanus)
panayensis
Neostethus (Sandakanus)
zamboangae
Ctenophallus ctenophorus
Plectrostethus
palawanensis
Phallostethus dunckeri
Phenacostethus smithi
Phenacostethus posthon
Phenacostethus trewavasae
Gulaphallus bikolanus
Neostethus thessa
Gulaphallus eximius
Gulaphallus mirabilis
Neostethus bicornis
Gulaphallus falcifer
Neostethus amaricola
Neostethus lankesteri
synonym of Neostethus
lankesteri
Neostethus villadolidi
Neostethus borneensis
synonym of Neostethus
borneensis
Gulaphallus panayensis
Neostethus zamboangae
Neostethus robertsi
Neostethus ctenophorus
Neostethus palawanensis
The theory of phylogenetic reconstruction fol-
lowed is that commonly referred to as phyloge-
netic systematics or cladistics, as outlined by
Hennig (1966) and modified by numerous au-
thors (see review by Wiley 1981). Genera and
other higher taxa are defined as monophyletic by
the sharing of homologous derived characters,
termed synapomorphies. Species are defined as
the smallest, recognizable, morphologically dis-
tinctive groups of males and females which may
share homologous derived characters, termed
autapomorphies.
For priapial characters, a state that is unique
to a particular phallostethid species, or group of
species, is treated as derived. This is a special
case of the method of outgroup comparison (see
Maddison et al. 1 984) which is used to determine
polarity of non-priapial characters. The follow-
ing atherinomorph specimens were examined for
outgroup comparison: Bedotia sp., aquarium
material, CAS 44367 (1 of 3 counterstained);
Dentatherina merceri, Moluccas, USNM 230374
246
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1 1
INDIAN
OCEAN
10
FIGURE 1. Distribution of phallostethids. Solid circles, type localities of the 19 currently recognized phallostethid species: 1.
Phallostethus dunckeri Regan, Muar R.; 2. Phenacostethus smithi Myers, Bangkok; 3. Phenacostethus posthon Roberts, a. Khlong
Kla Sohm, b. Khlong Langu; 4. Phenacostethus trewavasae Parenti, Baram R.; 5. Neostethus lankesteri Regan, a. Muar R., b.
Singapore, c. Chantabun R.; 6. Neostethus bicornis Regan, Kuala Langat; 7. Neostethus palawanensis (Myers), Ulugan Bay; 8.
Neostethus thessa (Aurich), Lake Mainit; 9. Neostethus borneensis Herre, a. Sandakan Bay, b. Coron; 10. Neostethus villadolidi
Herre, a. Misamis, b. Dumaguete; 1 1 . Neostethus zamboangae Herre, Zamboanga; 1 2. Neostethus amaricola (Villadolid and
Manacop), Pasay; 13. Neostethus robertsi, new species, Calasiao R.; 14. Neostethus ctenophorus (Aurich), Laguna de Bay; 15.
Gulaphallus eximius Herre, Nueva Vizcaya; 1 6. Gulaphallus mirabilis Herre, Ibo R.; 1 7. Gulaphallusfalcifer Manacop, Pampanga
and Nueva Ecija; 18. Gulaphallus bikolanus (Herre), lakes Bato and Lanigay; and 19. Gulaphallus panayensis (Herre), Panay.
Solid triangle, unsubstantiated report of unidentified phallostethid from off northwestern Sumatra by Aurich (1937).
(2 of 42 counterstained); Hypoatherina ovalaua,
Vanuatu, Malekula Is., CAS 62564 (2 of 1 0 coun-
terstained); Melanotaenia affinis, northern New
Guinea, CAS 40573 (4 of 3 1 counterstained), and
Pseudomugil signifer, New South Wales, Austra-
lia, CAS 62565 (6 of 62 counterstained).
The pattern of reduction or loss of a character,
which is common in atherinomorphs, is useful
as a source of information for phylogenetic re-
construction when correlated with other, more
complex characters (Parenti 1 986b). Ontogenetic
data are used to infer homology among character
states (Patterson 1 982) and as an additional source
of information for character polarity (Nelson
1978). Ontogenetic transformation and pattern
of reduction or loss are the most reliable sources
of information on polarity of states of the pria-
pium. Outgroup comparison is limited because
a priapium is not found, even in a simplified
condition, outside the Phallostethidae.
PARENTI: PHALLOSTETHID FISHES
247
TABLE 2. Meristic characters of phallostethids. Dl = first dorsal rays, D2 = second dorsal rays, A = anal rays, PI = pectoral
rays, BR = branchiostegal rays, V = vertebrae, SLS = scales in lateral series.
Species
Dl
D2
A
PI
BR
V
SLS
N. lankesteri
2
6-7
16-17
11-12
5
34-35
31-32
N. bicornis
1
5-6
15-16
10-11
5-6
36-37
31-34
N. palawanensis
1
5-6
15-16
11
5
34-35
30-32
N. thessa
0
8-10
17-22
12-13
5
36-37
32-34
N. borneensis
1-2
5-6
13-16
10-12
5
31-32
25-27
N. robertsi
2
6
17-18
10-11
5
36
30
N. villadolidi
2
5-6
14-16
9
5-6
35-36
30-32
N. zamboangae
2
5-6
15-17
9
5-6
35-36
29-32
N. amaricola
1-2
5-6
16
10-11
5-6
35-36
32
N. ctenophorus*
2
5-7
14-17
10-11
—
—
31-33
G. eximius
1-2
7-8
15-18
11
5
36-37
52-58
G. mirabilis
1-2
7
17-18
10-12
5
36-37
36-38
G.falcifer
1-2
8
17
10
5
35
32
G. bikolanus
0
6
14-17
11-13
5
35-36
32
G. panayensis
1-2
6-7
13-15
11
5
31-33
28-30
P. dunckeri
0
8-10
26-28
9-10
4
40
b
P. trewavasae
1
6
14-15
9-10
5
34
b
P. position
1
5-6
14-15
9-10
4
34-35
b
P. smilhi
1
6-7
14-15
9-10
4-5
33-35
b
"From Aurich( 1937).
b Because Phallostethini scales are small and deciduous, it is difficult to record an accurate count.
RELATIONSHIPS OF PHALLOSTETHIDS TO
OTHER ATHERINOMORPHS
None has questioned monophyly of the phal-
lostethids, i.e., all phallostethid males have a
priapium, and none has argued that the priapium
is not a uniquely evolved structure. I present no
argument here to contradict phallostethid mono-
phyly. However, I acknowledge that information
from some future study could lead us to conclude
that not all priapia are homologous.
Perhaps the only currently controversial aspect
of phallostethid systematics is their relationship
to other acanthopterygian fishes. The following
four issues, with references for arguments pro
and con, summarize the debate:
1. Atherinomorphs are monophyletic, al-
though their relationship to other acanthomorph
teleosts remains obscure (Rosen 1964; Rosen and
Parenti 1981; Collette et al. 1984; Parenti 1981,
1984; Grier and Collette 1987).
Since Rosen (1964) formally recognized the
series Atherinomorpha, which comprises the
atherinoids, cyprinodontiforms, and beloni-
forms, the group has been accepted by ichthy-
ologists, although with some reservation (e.g.,
Roberts 1971b). However, it is generally agreed
that systematic differences between the Ather-
inidae and Cyprinodontidae are not as great as
had been thought by Hubbs (1944). Traditional
groups such as the Percesoces— which contains
the mullets (Mugiloidea), polynemids (Polyne-
moidea), and phallostethids (Phallostethoidea),
according to the classification of Myers (1937)—
are no longer recognized. However, possible ath-
erinomorph relatives may be among constituents
of these taxa, such as the mugilids. Confirmation
of the sister group of the Atherinomorpha, as
well as other acanthomorph taxa, is the subject
of ongoing studies by numerous systematic ich-
thyologists, including myself, and will not be dis-
cussed further here.
2. Phallostethids are atherinomorphs (Rosen
and Parenti 1981; Parenti 1984; White etal. 1984;
Ivantsoff et al. 1987), or may be related to poly-
nemids (Bailey 1936) or to gobioids (Springer
1983).
Determining whether phallostethids are ath-
erinomorphs requires a well-corroborated defi-
nition of the series Atherinomorpha; phallo-
stethids either fit the definition or they do not.
Rosen and Parenti (1981) listed defining char-
acters of the Atherinomorpha, which includes a
derived ethmoid region and hyobranchial ap-
paratus. Recent workers have corroborated the
248
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1 1
proposal that the Atherinomorpha is monophy-
letic (e.g., Grier and Collette 1987), and that
phallostethids are atherinomorphs (Parent! 1 984;
Ivantsoffetal. 1987).
Bailey (1936) hypothesized a close relationship
between phallostethids and polynemids based on
the shared close association of the pelvic and
pectoral girdles. Springer (1983) cited shared re-
ductions between phallostethids and gobioids to
suggest, in a footnote, their possible close rela-
tionship. To endorse the hypothesis that phal-
lostethids are atherinomorphs requires that the
characters shared by phallostethids and poly-
nemids or gobioids be treated as convergent.
However, Springer's (1983) proposal expands the
list of possible atherinomorph close relatives to
polynemids, mullets, and gobioids.
3. Atherinoids, including the silversides (or
hardyheads), rainbowfishes, and phallostethids,
are not monophyletic (Rosen and Parenti 1981;
Parenti 1984; Ivantsoffetal. 1987)oratherinoids
are monophyletic (White et al. 1984).
Rosen and Parenti (1981) and Parenti (1984)
proposed that atherinoids do not constitute a
monophyletic group because they exhibit trans-
formation series for certain characters, such as
transformation of the first dorsal fin from ple-
siomorphic atherinoids (e.g., Bedotia and Me-
lanotaenid) in which it is well developed, to phal-
lostethids in which it is reduced or absent. This
view was supported by Ivantsoffet al. (1987).
White et al. (1984) proposed two developmental
characters to support their view that atherinoids
are monophyletic. This issue will be clarified when
there is a well-corroborated proposal of relation-
ships among the major groups of atherinomorph
fishes, a study under progress.
4. Phallostethids are the sister group of Den-
tatherina Patten and Ivantsoff, 1983 (Parenti
1984), or the close relationship of phallostethids
to Dentatherina is not supported (Ivantsoffet al.
1987).
Parenti (1984) proposed a close relationship
between Dentatherina merceri and phallosteth-
ids based on, among other characters, presence
of a paradentary bone. At issue is whether the
paradentary elements— ossifications of the labial
ligament— are homologous in phallostethids and
Dentatherina. The phallostethid paradentary
bone is composed of acellular bone surrounding
hyaline cartilage (Parenti 1986c; Ivantsoffet al.
1987). The Dentatherina paradentary bone is
composed of hyaline cartilage with a calcified
core (Ivantsoffet al. 1987).
Ivantsoffet al. (1987) cited the histological dif-
ferences between the phallostethid and Dentath-
erina paradentary elements to conclude that the
bones are not homologs and, therefore, cannot
be used to propose a close relationship between
two taxa. They also rejected homology based on
the conclusion that calcification or ossification
in a labial ligament could result as a response to
stress. I concur that if the bones are not homol-
ogous they cannot be used to propose a close
relationship between two taxa. However, I reject
the functional argument against homology. The
labial ligament could become calcified or ossified
in any taxon in which it occurs; within atherino-
morphs, it is reported ossified only in phallos-
tethids and Dentatherina. In my opinion, addi-
tional characters are needed to accept or refute
the proposal of a phallostelhid-Dentatherina sis-
ter group.
Phallostethids need no additional support to
be accepted as a monophyletic group; a seem-
ingly endless number of derived characters could
be listed to define them (see Synonymy and Ho-
mology of Priapial Structures). Further research
should focus on the question: to which group of
Old World atherinoids are phallostethids most
closely related? Hubbs (1944:70) summarized the
importance of this type of question in system-
atics: "Whether or not a suborder or an order be
recognized for the Phallostethidae, on the basis
of the bizarre copulatory organ (the priapium)
and of other structures, is, like most questions
of ranking, of no great inherent significance.
Whether the relationship of the family is with
Cyprinodontidae or with the Atherinidae, or with
neither, is a problem of greater meaning."
SYNONYMY AND HOMOLOGY OF
PRIAPIAL STRUCTURES
Since Regan (1913) first described the complex
anatomy of the phallostethid priapium, students
of this unusual group of fishes have struggled to
attain a stable, appropriate set of terms for the
numerous priapial structures (Regan 1916; Bai-
ley 1936; Aurich 1937). Nearly every paper dis-
cussing priapial anatomy contains errors of iden-
tification or proposed homology. A common
mistake is to refer to the same structure by more
than one name in a single paper.
PARENTI: PHALLOSTETHID FISHES
249
Proliferation of names for priapial structures
was encouraged by Regan (1916), who believed
that the priapium was a new morphological
structure with no known homolog among teleost
fishes. For example, Regan gave the adjectival
name "anterior infrasulcar" to a bone which I
consider the homolog of one of the pelvic bones.
No doubt, a stable nomenclature has not been
attained because homology of the priapial struc-
tures has not been well understood, and because
structure of the priapium may vary greatly from
one phallostethid genus to another. Bailey (1 936)
concluded that priapia of Phenacostethus smithi
and Gulaphallus mirabilis, and hence of all phal-
lostethids, correspond bone-for-bone. Further-
more, he believed that these structures could be
identified in polynemids, close relatives of phal-
lostethids according to Bailey. However, he did
not examine a species of Neostethus, which has
bony elements not found in either Phenacoste-
thus or Gulaphallus. Also, a close relationship
between polynemids and phallostethids has not
been supported (see Relationships of Phallo-
stethids to other Atherinomorphs).
The priapium may be thought of as having
three structural and functional components: sus-
pensory, holding or clasping, and papillary (Bai-
ley 1936). The suspensory component consists
of expanded anterior pleural ribs, including an-
teriorly expanded cleithra in Neostethus and Gu-
laphallus, as well as modified pelvic bones. The
holding or clasping component consists of the
externalized ctenactinia and/or toxactinium, the
movement of which is controlled mainly by a set
of longitudinal muscles. The papillary compo-
nent consists of the seminal and infrasulcar pa-
pillae and papillary bones used in the transfer of
sperm bundles from males to females.
To describe the diverse and detailed priapia
of phallostethids, I have compiled a synonymy
of priapial structures (Table 3 and below). In
addition, I propose, where possible, the homolog
of each element as found in relatively unmodified
atherinoid fishes. Listed is the name of a struc-
ture, first and subsequent citations, including
misidentifications, in major papers on phallo-
stethids, brief description, probable homolog,
known synonyms, and citations.
antepleural cartilage or bone (Figs. 2, 3)— Regan
(1916:7), Roberts (1971b:408), Parenti (1986a:
228). A large block of cartilage, sometimes par-
tially or fully ossified, that lies anterior to the
distal tips of the first pleural ribs in males.
aproctal side or aspect— Regan (1913:552, 1916:
5), Herre (1 942: 1 38), Parenti (1986a:225). That
side of the male's body without the anal open-
ing. Male phallostethids are bilaterally asym-
metric such that the anus and seminal papilla
are offset to opposite sides of the body (see
Parenti 1986a).
axial bone-Regan (1916:7), Bailey (1936:464),
TeWinkel (1939:61), Villadolid and Manacop
(1935:215). An elongate bone with rodlike
spine and broad, dorsal crest. The aproctal and
proctal axial bones are homologous with the
pelvic bones (Regan 1913, 19 16; Bailey 1936).
aproctal axial bone (Figs. 2-5)— Roberts
(1971b:408-409). The axial bone on the
aproctal side of the body. Synonyms: an-
terior infrasulcar bone (Regan 1916:8,
Bailey 1936:466), desmactinium (Aurich
1937:267), ctenactinium (misidentifica-
tion, Herre 1925:508, 1926:539; Myers
1937:142), second ctenactinium (mis-
identification, Bailey 1936:463), short
ctenactinium (misidentification, Villado-
lid and Manacop 1935:195).
proctal axial bone (Figs. 3-5) — Roberts
(1971b:408-409). The axial bone on the
proctal side of the body, with which the
toxactinium and ctenactinia articulate.
Synonym: axial bone (Regan 1916:7;
Myers 1935:5; Bailey 1936:464; Parenti
1986a:228, 1986b:308).
ctenactinium— Regan (19 13: 5 52, 1916:5), Myers
(1928:1, 1935:5), Bailey (1936:465-466),
Manacop (1936:376), Aurich (1937:264),
TeWinkel (1939:62), Herre (1939:140, 1942:
138), Woltereck(1942b:339), Roberts (197 la:
6, 1971b:393), Parenti (1984:6, 1986b:309).
One of two curved, rodlike priapial bones that
articulate with the posterior end of the proctal
axial bone. Homologous with a pelvic fin ray
(Bailey 1936).
first ctenactinium (Figs. 3-5)— The promi-
nent, elongate externalized priapial bone
in Neostethus and Gulaphallus. Usually
called the ctenactinium, referred to herein
as the first ctenactinium to distinguish it
unambiguously from the second ctenac-
tinium. Synonym: toxactinium (misiden-
250
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1 1
TABLE 3. Summary of names of priapial structures and
synonyms. See text for definitions and references.
TABLE 3. Continued.
Priapial structure
Synonyms
Priapial structure
Synonyms
antepleural cartilage or
bone
aproctal side or aspect
axial bone
aproctal axial bone
proctal axial bone
anterior infrasulcar bone, des-
mactinium, ctenactinium,
second ctenactinium, short
ctenactinium
axial bone
pulvinulus
seminal papilla
toxactinium
uncus
shieldlike pad, pulvinular pad
genital papilla, penis
outer pulvinular bone, pul-
vinular bone, toxactinial
bone, inner pulvinular
bone, modified pulvinulus
ctenactinium
first ctenactinium
second ctenactinium
efferent groove
epididymis
fringe
glandular groove
infrasulcar bones
anterior infrasulcar
bone
posterior infrasulcar
bone
infrasulcar prominence
or papilla
keel
longitudinal muscles
ctenactinial muscle
outer muscle of aproctal
side
inner muscle of proctal
side
muscle of proctal side
papillary bones
penial bone
basipenial bone
papillary bone
prepapillary bone
cristate bone
pleuro-priapial muscle
priapial process of cleithra
priapial ribs
aproctal priapial rib
proctal priapial rib
priapium
proctal side or aspect
pulvinular appendage
pulvinular spine
pulvinular bones
inner pulvinular bone
outer pulvinular bone
toxactinium
Priapklaue, "pulvinular
spine," anterior infrasulcar
bone
terminal coil of vas deferens
comblike projections, comb-
like cilia, cilia, comb
aproctal axial bone
embryonic fin-fold, fringe
penial bones
first pleural ribs
first pleural rib on aproctal
side
first pleural rib on proctal
side
toxactinium
tification, Herre 1925:508, 1926:539; Vil-
ladolid and Manacop 1935:200).
second ctenactinium (Figs. 2, 4, 5, 9)— The
shorter of the two externalized priapial
bones that may have serrations, as in
Phallostethus, and N. zamboangae. Syn-
onym: Priapklaue (Aurich 1937:267),
"pulvinular spine" or tip of anterior in-
frasulcar bone (misidentification, Herre
1942:138).
efferent groove— Regan (1916:5). A groove on
the aproctal side of the body, posterior to the
glandular groove and separated from it by the
seminal papilla and infrasulcar prominence or
papilla.
epididymis (Fig. 5) -Regan (1913:555), Roberts
( 1 97 1 a:6), Parenti ( 1 986a:23 1 ). The coiled vas
or ductus deferens enclosed in a membranous
sac at posterior extent of the priapium. Syn-
onym: terminal coil of vas deferens (Regan
1916:4).
fringe (Figs. 3, 13)-Regan (1916:5), Roberts
(1971b:408). Posterior ends of pelvic fin rays
that extend beyond the posteroventral outline
of the epididymis. Synonyms: comblike pro-
jections (Myers 1928:8; Villadolid and Man-
acop 1935:197), comblike cilia (Herre 1939:
140), cilia (Herre 1942:138), comb (Roberts
1971b:408).
glandular groove— Regan (1916:5). A groove on
the aproctal side of the body at the boundary
between the priapium proper and the body
wall.
infrasulcar bones— Regan (1916:8). Two bones,
referred to by Regan ( 1 9 1 6) as the anterior and
posterior infrasulcar, that, in part, support the
infrasulcar and seminal papillae.
PARENTI: PHALLOSTETHID FISHES
251
SCA
RAD
PEN
ANT
PAX
APAX
FIGURE 2. Diagrammatic representation of modified pelvic and pectoral fins in the priapium of a dextral male, Phenacostethus
smithi (CAS-SU 35957 and modified in part from Bailey 1936), slightly exploded view from proctal side. Anterior to left.
Dashed line, outline of pulvinulus; dotted and dashed line, anterior extent of cleithrum. Dense stippling, cartilage; light stippling,
bone. Abbreviations: ant, antepleural bone and cartilage; apax, aproctal axial bone; bp, basipenial bone; cl, cleithrum; cte, second
ctenactinium; fpr, first pleural ribs; pap, papillary bone; pax, proctal axial bone; pen, penial bone; pin, posterior infrasulcar
bone; pp, prepapillary bone; ptt, posttemporal bone; rad, radials; sea, scapulocoracoid; scl, supracleithrum; u, uncus.
anterior infrasulcar bone— Regan (1916:8).
See aproctal axial bone.
posterior infrasulcar bone (Fig. 2)— Regan
(19 16:8), Bailey (1936:467). A rod-shaped
bone at the base of the aproctal axial bone
with a laminar projection that lies ventral
to the aproctal axial bone, reaching the
base of the infrasulcar papilla or promi-
nence.
infrascular prominence or papilla— Regan (1916:
5, 8), Herre (1939:141). A papilliform projec-
tion anterior to the seminal papilla, supported
internally by a posterior projection of the
aproctal axial bone.
keel-Regan (1913:550), Myers (1935:5), Rob-
erts (1971b:410). A median, abdominal,
slightly frayed, fleshy ridge between the uro-
genital opening and anterior rays of the anal
fin. Homologous with embryonic fin-fold
(Roberts 1971b). Synonym: fringe (alternate
term not homologous with fringe as defined
above, Regan 1916:2; Myers 1928:5).
longitudinal muscles (Fig. 6)— Regan (1916:13,
fig. 10; 1916:14). Four fusiform muscles—
ctenactinial muscle, outer muscle of the aproc-
tal side, inner muscle of the aproctal side, and
muscle of the proctal side— that originate on
the axial bones and control copulatory move-
ments of the priapium. Precise homology with
pelvic fin muscles (Winterbottom 1974) is un-
clear.
ctenactinial muscle (Fig. 6)— Regan (1916:
14). Largest of the longitudinal priapial
muscles, with an origin at the anterior end
of the dorsal crest of the aproctal axial
bone and insertion at the proximal end of
the ctenactinium.
outer muscle of the aproctal side (Fig. 6)—
Regan ( 1 9 1 6: 1 4). An elongate muscle that
originates on the concave, aproctal side of
the main crest of the proctal axial bone
252
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1 1
SCL
SCA
RAD
IPUL
FIGURE 3. Diagrammatic representation of modified pelvic and pectoral fins in priapium of a sinistral male, Neostethus
lankesteri (CAS-SU 67162). Anterior to left. Dense stippling, cartilage; light stippling, bone. Open circles, fibrous connective
tissue of pulvinular appendage. Dotted line approximates outline of epididymis. Abbreviations: apax, aproctal axial bone; bp,
basipenial bone; cl, cleithrum; cte, second ctenactinium; ctn, first ctenactinium; fpr, first pleural ribs; ipul, inner pulvinular bone;
opul, outer pulvinular bone; pap, papillary bone; pax, proctal axial bone; pen, penial bone; pfr, pelvic fin rays; ptt, posttemporal
bone; rad, radials; sea, scapulocoracoid; scl, supracleithrum.
and inserts on the base of the papillary
bone.
inner muscle of the proctal side (Fig. 6)—
Regan ( 1 9 1 6: 1 4). An elongate muscle that
originates at the junction of the main crest
of the proctal axial bone, internal to the
outer muscle of the aproctal side, and in-
serts on a process of the proximal end of
the toxactinium or outer pulvinular bone.
muscle of the proctal side (Fig. 6)— Regan
(1916:14). A muscle that originates at the
extreme anterior end of the axial bone,
with a few fibers at the posterior base of
the toxactinium, posteriorly produced into
a long tendon that inserts on the base of
the papillary bone.
papillary bones (Figs. 2-4)— Regan (1916:8),
Parenti (1984:6). Five small posterior priapial
bones. Synonym: penial bones (Bailey 1936:
467).
penial bone (Figs. 2-4)- Bailey (1936:467).
A thin, laminar bone that projects into the
seminal papilla.
basipenial bone (Figs. 2-4)— Bailey (1936:
465, 468). A rodlike bone that articulates
with the base of the penial bone and pos-
terior border of the cristate bone.
papillary bone (Figs. 2-4)— Regan (1916:8),
Bailey (1936:468), Herre (1939:143). A
slender, curved bone that supports the
seminal papilla.
prepapillary bone (Fig. 2)— Bailey (1936:
468). A slender, rodlike bone that artic-
ulates proximally with the basipenial bone
and distally with the papillary bone.
cristate bone -Bailey (1936:469). A flat-
tened, platelike bone that lies between the
papillary bone and the aproctal axial bone
and projects into the base of the seminal
papilla.
pleuro-priapial muscle (Fig. 6) — Regan (1916:12).
A thin muscle that originates on the enlarged
PARENTI: PHALLOSTETHID FISHES
253
SCA
RAD
OPUL
APAX
PEN
FIGURE 4. Diagrammatic representation of modified pelvic and pectoral fins in priapium of a sinistral male, Gulaphallus
mirabili? (CAS-SU 38904 and modified in part from Bailey 1936). Anterior to left. Dense stippling, cartilage; light stippling,
bone. Abbreviations: apax, aproctal axial bone; bp, basipenial bone; cl, cleithrum; ctn, first ctenactinium; fpr, first pleural ribs;
opul, outer pulvinular bone; pap, papillary bone; pax, proctal axial bone; pax-pro, process of the proctal axial bone; pen, penial
bone; ptt, posttemporal bone; rad, radials; sea, scapulocoracoid; scl, supracleithrum.
transverse process of the third or fourth ver-
tebra on the proctal side and inserts on the
proctal axial bone at the point of articulation
of the first ctenactinium.
priapial process of cleithra (Figs. 3, 4)— Regan
(1913:553-554). Anterior prolongation of the
cleithra in Neostethus and Gulaphallus.
priapial ribs (Figs. 2-4)-Bailey (1936:463-464).
The first pleural ribs on the third or fourth
vertebra, modified to support the priapium.
aproctal priapial rib— Bailey (1936:464). The
first pleural rib on the aproctal side of the
body.
proctal priapial rib— Bailey (1936:463). The
first pleural rib on the proctal side of the
body.
priapium (Figs. 2-5, 9)-Regan (1 9 13:551,1916:
4), Myers (1928:1), Bailey (1936:453), Aurich
(1937:263), TeWinkel (1939:59), Woltereck
(1942a:254, 1942b:338).Thesubcephaliccop-
ulatory organ of phallostethid fishes compris-
ing a bony portion used to hold the female
during mating, a fleshy copulatory organ or
seminal papilla through which sperm bundles
APROCTAL AXIAL
BONE
SEMINAL PAPILLA
EPIDIDYMIS
FIGURE 5. Diagrammatic representation of head and an-
terior portion of body, Gulaphallus mirabilis, sinistral male
(CAS 50720). Anterior to left. Dense stippling, first ctenactin-
254
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1 1
ROCTAL TRANSVERSE PROCESS
VENTRAL BODY PROFILE
APROCTAL AXIAL BONE
'\
FIRST
CTENACTINIUM
CTM
FIGURE 6. Schematic diagram of priapial muscles (modified from Regan 1916:fig. 10). Anterior to right. Stippling, bone.
Abbreviations: ppm, pleuro- priapial muscle; ctm, ctenactinial muscle; mp, muscle of proctal side; ima, inner muscle of aproctal
side; oma, outer muscle of aproctal side.
pass, and expanded cleithra and pleural ribs
for suspension.
proctal side or aspect— Regan (1913:551, 1916:
5), Herre (1 942: 1 38), Parenti (1 986a:233). That
side of the body with the anal opening.
pulvinular appendage (Fig. 3)— Regan (1916:5,
8), Herre (1942:137). A subconical, com-
pressed mass of fibrous tissue that lies just pos-
terior to the pulvinulus and is supported by
the inner pulvinular bone in Neostethus.
pulvinular spine (Fig. 3)— Regan (1916:5, 9). A
pointed, curved process of the toxactinium
(=outer pulvinular bone).
pulvinular bones (Figs. 2-4)— Regan (191 6:8-9).
Two bones, termed an inner and an outer pul-
vinular, that articulate with the anterior end
of the proctal axial bone,
inner pulvinular bone (Fig. 3) — Regan (1916:
8-9). A longitudinal, chondral bone that
lies anteroventral to the proctal axial bone,
outer pulvinular bone (Figs. 2-4)— Regan
(1916:9). See toxactinium.
pulvinulus (Figs. 2, 9)— Regan (1916:5, 8), Myers
(1928:6, 1935:5), Villadolid and Manacop
(1935:194), Manacop (1936:376), Aurich
(1937:264), Herre ( 1939:1 40), TeWinkel( 1939:
62), Roberts (197 la: 7). Fibrous tissue sur-
rounding the inner pulvinular bone and/or
toxactinium (=outer pulvinular bone) cover-
ing point of articulation of the toxactinium,
cleithra, and proctal axial bone. Synonyms:
shieldlike pad (Bailey 1936:454); pulvinular
pad (Parenti 1986a:230).
seminal papilla (Fig. 5)— Regan (1916:5), Herre
(1939:143). A papilliform projection lying
posterior to the infrasulcar prominence, sup-
ported internally by the posterior infrasulcar
bone. Synonyms: genital papilla (TeWinkel
1939:62), penis (Roberts 1971a:l; Parenti
1986a:228).
toxactinium (Figs. 2-4)-Regan (1913:551),
Myers (1928:1), Bailey (1936:462), Woltereck
(1942b:339), Herre (1939:140, 1942:138),
Roberts (1971a:6), Parenti (1986a:228). A
rodlike, longitudinal bone, the posterior end
PARENTI: PHALLOSTETHID FISHES
255
of which movably articulates with the sym-
physis of the cleithra and anterior end of the
proctal axial bone. In Phallostethus and Phe-
nacostethus, the anterior, distal end curves be-
low the head towards the aproctal side of the
body. Called the outer pulvinular bone in all
other phallostethids, in which the anterior end
articulates with the anterior end of the urohyal
and the first basibranchial bone. Synonyms:
outer pulvinular bone (Regan 1916:9), pulvin-
ular bone (Bailey 1936:463), toxactinial bone
(Bailey 1 936:47 1), inner pulvinular bone (mis-
identification, Roberts 1971b:393), modified
pulvinulus (Parenti 1984:6).
uncus (Fig. 2)-Bailey (1936:465, 467). A
U-shaped chondral bone that articulates with
the proctal axial bone posteriorly.
PHYLOGENETIC ANALYSIS
External morphology and osteology of the
priapium has been reported in detail for a variety
of phallostethid species (e.g., Regan 1913, 1916;
Bailey 1936; Aurich 1937; TeWinkel 1939; Rob-
erts 197 la, b; and Parenti 1984, 1986a). How-
ever, this represents the first attempt to sum-
marize derived priapial and other characters
within all phallostethid fishes, and to use cor-
roborated hypotheses of synapomorphies to di-
agnose monophyletic genera and groups of gen-
era.
Aurich (1937:285) perhaps came the closest to
the goal of diagnosing groups of genera by tab-
ulating identifying characters of his three
subfamilies (but called "Familien")— Phallo-
stethinae, Neostethinae, and Gulaphallinae—
recognized herein at the tribal level with modi-
fied definition and composition. Aurich did not
treat relationships among his three subfamilies,
nor did he indicate which of the listed characters
are what we would today recognize as derived or
apomorphic.
Herre (1942) provided a synopsis of all genera
and species, except Phallostethus and Phenaco-
stethus, both then monotypic. I find many of
Herre's (1942) characters, especially of scales,
unworkable in a phylogenetic analysis. For ex-
ample, number of scales on the opercle was used
to divide Neostethus species into subgenera
Neostethus and Sandakanus. Herre (1942:153)
described several new species, including N. pa-
nayensis, treated as closely related to N. zam-
boangae Herre, 1 942 and N. coronensis Herre,
1 942, with which it shares two scales on the oper-
cle. However, in priapial morphology, N. panay-
ensis is a member of the monophyletic Gula-
phallus as defined below, and not closely related
to N. zamboangae or TV. coronensis (=borneensis,
herein).
Roberts (1971b:396) published a diagram in
which relationships among the then recognized
nine genera were summarized. It was not Rob-
erts's purpose to provide a phylogenetic revision
of all phallostethids; however, like Herre, Rob-
erts depicted the monotypic Manacopus as closely
related to Neostethus; Manacopus is considered
a synonym of Gulaphallus herein.
Monophyly of Phallostethidae
As stated in the section Relationships of Phal-
lostethids to other Atherinomorphs, phallosteth-
id fishes have been considered monophyletic since
Regan's (1916) review of the then three known
species: by definition, all phallostethid males have
a priapium, a complex structure not found in any
other group of fishes. It is perhaps impossible to
tabulate all the characters in phallostethids that
have undergone transformation from the state in
plesiomorphic atherinoids. Over 30 modifica-
tions of pelvic and pectoral bones, muscles, and
cartilage can be identified (see Synonymy and
Homology of Priapial Structures).
Further, presence of a keel, homologous with
the ventral embryonic fin-fold as reported by
Takita and Nakamura (1986) in larval Hypoath-
erina bleekeri, and other paedomorphic charac-
ters in all phallostethids suggest that evolution
of a priapium is correlated with maturation at a
relatively small size. Congruence (Mickevich
1978) exists between patterns of loss and reduc-
tion associated with small size and development
of some complex priapial characters, especially
within phallostethins.
Relationships of Subfamilies and Tribes
The following classification summarizes phy-
logenetic relationships among the tribes and
subfamilies of phallostethids I recognize:
Family Phallostethidae
Subfamily Phallostethinae
Tribe Phallostethini
Tribe Neostethini
Subfamily Gulaphallinae
Tribe Gulaphallini
256
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1 1
Phallostethini
Neostethini Gulaphallini
Phallostethini Neostethini Gulaphallini
FIGURE 7. Two alternative cladograms of relationships
among the three phallostethid tribes. Characters numbered as
in Phylogenetic Analysis.
I use the following vernacular terms: phallo-
stethins for Phallostethini, neostethins for Neo-
stethini, and gulaphallins for Gulaphallini. To
avoid confusion, no vernacular is used for Phal-
lostethinae. Monophyly of each tribe is consid-
ered below. Two alternative relationships are:
Phallostethini and Neostethini are sister tribes
(Fig. 7a); and Neostethini and Gulaphallini are
sister tribes (Fig. 7b). I have found no support
for the hypothesis that Gulaphallini and Phal-
lostethini are sister tribes; that possible phylog-
eny will not be discussed further.
The first alternative is supported by at least
four derived characters, two of the priapium, one
of the oral jaws, and one of the gill arches. Char-
acters have the same numbers as in the clado-
grams of Figures 7, 8, and 11.
1. Second ctenactinium present (Figs. 2, 3, 9).
The second ctenactinium is a small, curved,
externalized priapial bone that may have bony
serrations (Regan 1916:fig. 3). It is present in
Neostethus (Figs. 3, 9) and Phallostethus, and is
rudimentary in Phenacostethns (Fig. 2). The bone
is absent in Gulaphallus (Fig. 4), contrary to Au-
rich (1937) who misidentified the large, partially
externalized papillary bone as a second ctenac-
tinium.
2. Shieldlike pulvinulus present (Figs. 2, 9)
and outer pulvinular bone robust (Figs. 2, 3).
The oval pulvinulus of Phallostethus, Phena-
costethus (Fig. 2), and Neostethus (Fig. 9) forms
a shieldlike covering of the articulation point be-
tween the outer pulvinular and proctal axial
bones. The outer pulvinular bone, although pres-
ent in gulaphallins (Fig. 4), is larger and more
robust in phallostethins and neostethins. Bailey
(1936) and Aurich (1937) reported a slender, long
pulvinulus paralleling the longitudinal ctenactin-
ial muscles in Gulaphallus. However, the struc-
ture in Gulaphallus appears to be simply a fold
of skin covering the muscles and is viewed here
as doubtfully homologous with the distinctive
pulvinulus of neostethins and phallostethins.
3. Highly protrusible oral jaws.
Premaxillary ascending processes are long, and
a short, rodlike block (submaxillary element of
Parenti 1984) lies between the maxilla and vo-
mer in Neostethini and Phallostethini, both
characters contributing to the highly protrusible
oral jaws (Roberts 1971b; Ivantsoffet al. 1987).
The block may be ossified in Phallostethus and
some species of Neostethus; otherwise it is car-
tilaginous. In Gulaphallini, there is a thin con-
nective tissue meniscus between the maxilla and
vomer, except in G. panayensis in which the me-
niscus is enlarged.
4. Reduction of hyobranchial apparatus.
The entire hyobranchial apparatus is reduced
in phallostethids relative to other atherinoids (see
Rosen and Parenti 1981). Teeth on the fourth
ceratobranchial toothplate are few or absent in
phalostethins and neostethins; they are present
in outgroup taxa such as Dentatherina merceri
and Pseudomugil signifer, as well as in Gula-
phallus.
The alternative that Neostethini and Gula-
phallini are sister tribes is supported by what may
be described as one complex priapial character.
5. Anterior elongation of priapial suspensory
system, including anteriorly expanded cleithra
and transverse processes of fourth vertebrae of
males, with the proctal transverse process larger
than the slightly modified aproctal transverse
process (Figs. 3, 4), and elongate first ctenactin-
ium, the prominent externalized priapial bone
(Figs. 3, 4, 9).
Parsimony may be used to choose the first phy-
logenetic hypothesis (Fig. 7a) over the second
(Fig. 7b), but that criterion alone is not satisfac-
tory. Outgroup comparison is of limited use in
deciding polarity of priapial structures, and like-
wise, complex priapial characters may be treated
PARENTI: PHALLOSTETHID FISHES
257
as one or several characters. I choose the hy-
pothesis that Phallostethini and Neostethini are
sister tribes for reasons of parsimony and also
because derived characters other than those of
the priapium are shared. This hypothesis re-
quires that anterior elongation of the priapium
in neostethins and gulaphallins (character 5,
above) be viewed as plesiomorphic for all phal-
lostethids.
Monophyly of Tribes
Phallostethini.— Four characters distinguish
phallostethins from all other phallostethids.
6. A large, fleshy seminal papilla.
The seminal papilla is enlarged relative to the
papilla in neostethins or gulaphallins (Fig. 5).
The large papilla may be smooth (Roberts 1 97 1 a:
fig. 5) or ruffled (Roberts 1971a:fig. 4). An en-
larged seminal papilla refutes the hypothesis that
the diminutive phallostethins can be described
solely as paedomorphic phallostethids.
7. The curved toxactinium the largest exter-
nalized bony priapial element (Fig. 2).
Treating the distinctive curved toxactinium as
a derived character of Phallostethus and Phena-
costethus is consistent with the interpretation
(above) that the elongate ctenactinium and an-
terior elongation of the priapium in neostethins
and gulaphallins are primitive for phallostethids.
8. Extremely small and delicate, with decid-
uous scales.
The largest phallostethid, Gulaphallus exi-
mius, reaches a recorded 37 mm SL (see char-
acter 47, below). TeWinkel (1939) gives an un-
supported range of 14-40 mm SL. Neostethins
and gulaphallins generally mature at over 20 mm
SL; the exception, N. borneensis, is considered
secondarily reduced. Small adult size and the
deciduous scales of phallostethins are probably
paedomorphic characters correlated with char-
acter 9.
9. A translucent, membranous dome on dor-
sal surface of head.
The membranous dome (Roberts 1971a:figs.
2, 3) on the dorsal surface of the head may best
be described as a larval characteristic of atheri-
noids (as in Hypoatherina bleekeri; see Takita
and Nakamura 1986) retained in adults.
Neostethini. — Two characters distinguish
Neostethus from all other phallostethids.
10. An inner pulvinular bone (Fig. 3).
The inner pulvinular bone is a large, well-de-
veloped chondral bone lying just posterior to,
and articulating with, the outer pulvinular bone.
It has no identified homolog in phallostethins or
gulaphallins.
1 1 . A high number of thin, bony projections
on the papillary bone (Fig. 3).
The papillary bone has a high number (re-
portedly more than 80; Roberts 1971b) of thin,
bony projections in Neostethus (Aurich 1937:fig.
6). These are the projections used by Myers (1935)
to diagnose a new genus, Plectrostethus (=Neo-
stethus, herein). Myers apparently compared his
new species with phallostethins and gulaphallins,
which lack projections. In the neostethins, Neo-
stethus thessa and N. zamboangae, the papillary
bone has only rudimentary projections.
Gulaphallini. — Gulaphallus species are togeth-
er distinguished from all other phallostethids by
three derived characters.
12. A perforated gular flap of skin through
which the anterior end of the first ctenactinium
may pass and remain while at rest (Fig. 5; Aurich
1937:fig. 2).
Neostethins are the other group of phallosteth-
ids with an elongate first ctenactinium; however,
neostethins do not possess a large gular flap
through which the anterior end of the ctenactin-
ium may pass. This character, mentioned and
figured by Aurich (1937:fig. 2), is uniquely de-
rived in Gulaphallus.
13. The aproctal axial bone projects beyond
ventral body profile (Figs. 4, 5).
In no other phallostethids does a priapial bone
project beyond the body profile and away from
the body of the priapium.
1 4. A small, dorsal crest posteriorly on proctal
axial bone (Fig. 4).
In other phallostethids (Figs. 2, 3) the proctal
axial bone has a uniform dorsal outline.
Species Relationships Within Tribes
Phallostethini
Relationships among, and definitions of, the
four phallostethin species Phallostethus dun-
ckeri, and Phenacostethus smithi, P. posthon, and
P. trewavasae were treated in detail by Parenti
(1986a) and need not be repeated here. Defining
characters of each species are listed in the Sys-
tematic Accounts.
258
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1 1
40
139
138
137
>36
33
32
291
141 I
1421
135 1
131 ,
28
1
134 1
130 ,
27
1
1
1
^
v
1
1
i
• rb
• 1
I
| LJ
An
.
r^ i JL
126
125
24
FIGURE 8. Cladogram of relationships among species in the
Neostethini. Solid squares, synapomorphies; open squares,
symplesiomorphies. Characters numbered as in Phylogenetic
Analysis.
Neostethini
Hypothesized relationships among the ten
Neostethus species are depicted in Figure 8. Neo-
stethus lankesteri is considered to be the most
plesiomorphic neostethin; it is excluded from a
monophyletic group comprising Neostethus
amaricola, N. villadolidi, N. robertsi, N. pala-
wanensis, N. bicornis, N. borneensis, N. zam-
boangae, and N. thessa, denned by two shared
derived characters.
15. An enlarged, prominent, second ctenac-
tinium.
Presence of a second ctenactinium is treated
as a derived character of neostethins and phal-
lostethins (see character 1, above). The second
ctenactinium is a small curved bone in phallo-
stethins (Fig. 2) and N. lankesteri (Fig. 3). It is a
much larger, prominent, externalized priapial
bone in all other species of Neostethus, including
N. ctenophorus, as illustrated by Aurich (1937:
fig. 6). Variation in shape of the second ctenac-
tinium and in its association within other pria-
pial structures is the primary source of infor-
mation on morphological species recognition
within neostethins.
16. A short, bladelike aproctal axial bone.
The aproctal axial bone is nearly as long as the
proctal axial bone and reaches the pulvinulus in
phallostethins (Fig. 2), Neostethus lankesteri (Fig.
3), N. ctenophorus (Aurich 1937:fig. 6), and Gu-
laphallus (Fig. 4). In all other species of Neos-
tethus (Roberts 1971b:fig. 12), the aproctal axial
bone is shortened and somewhat bladelike; it
does not reach the pulvinulus.
Neostethus ctenophorus, of which no speci-
mens are available, is treated as forming an un-
resolved trichotomy with N. lankesteri and all
other Neostethus (Fig. 8). I divide these remain-
ing species of Neostethus into two monophyletic
groups, which I call the borneensis and the bi-
cornis groups, diagnosed as follows:
The borneensis group, comprising Neostethus
borneensis, N. thessa, and N. zamboangae, is di-
agnosed by two shared derived characters that
could be treated as one complex priapial char-
acter.
17. Strongly curved, thin, second ctenactin-
ium.
The second ctenactinium is a strongly curved,
thin bone (Aurich 1 937:fig. 4) as opposed to being
a stout fin ray as in Neostethus amaricola and N.
villadolidi (Fig. 9) and N. bicornis (Roberts 1 97 Ib:
fig. 12).
18. Papillary and penial bones curved and
elongate, lying parallel to dorsal surface of second
ctenactinium.
Position of the papillary and penial bones just
dorsal to the second ctenactinium means that the
seminal papilla (supported by the papillary and
penial bones) empties just dorsal to the second
ctenactinium, as opposed to a more posterior
position in those phallostethids with smaller
papillary and penial bones.
The bicornis group, comprising Neostethus bi-
cornis, N. palawanensis, N. robertsi, N. amari-
cola, and N. villadolidi, is diagnosed by one com-
plex, derived character.
19. Both the proctal and aproctal transverse
processes of the fourth vertebra in males are en-
larged and oriented anteriorly.
It was concluded (character 5, above) that the
anterior elongation of the entire priapial appa-
ratus is a plesiomorphic character for phallo-
stethid fishes. This character complex is modi-
PARENTI: PHALLOSTETHID FISHES
259
fied in the bicornis group such that the aproctal
as well as the proctal transverse processes of the
fourth vertebra in males are drawn so far ante-
riorly that the anterior tip of both processes lies
parallel to the first vertebra.
Within the borneensis group, Neostethus bor-
neensis is considered plesiomorphic to N. thessa
and N. zamboangae, diagnosed as sister species
by sharing two derived characters.
20. Anterior aproctal ossification present.
In all species of Neostethus except N. lankes-
teri, the aproctal axial bone is short and does not
reach the pulvinulus (see character 16, above).
In N. thessa and N. zamboangae, there is a thin,
elongate ossification just anterior to the aproctal
axial bone but separate from it. Aurich (1937:
fig. 4) figured the aproctal axial bone and anterior
aproctal ossification as one long bone. A similar
bone in N. bicornis was called the "pulvinular
osselet" by Roberts (1971b:fig. 12). However,
because of differences in size and shape of the
bones, I do not consider that minute ossification
in N. bicornis to be homologous with the ossi-
fication in N. thessa and N. zamboangae.
2 1 . Reduction in number of papillary projec-
tions.
The thin, elongate papillary and penial bones
of the borneensis group (character 18, above) are
even further reduced within N. thessa and N.
zamboangae so that the fine, bony projections
of the papillary bone (Fig. 3) are few or absent.
Within the bicornis group, I recognize a tri-
chotomy among N. bicornis and two sister-group
pairs, N. palawanensis and N. robertsi, and N.
villadolidi and N. amaricola.
Neostethus palawanensis and N. robertsi share
one derived character.
22. Claw-shaped, anteriorly bifurcated, sec-
ond ctenactinium (Fig. 1 3).
The anterior extent of the second ctenactinium
in all other phallostethids that have the bone is
either blunt (Fig. 3) or sharply pointed (Fig. 9).
Hence, the claw-shaped, anteriorly bifurcated
second ctenactinium of N. palawanensis and N.
robertsi (Fig. 1 3), much larger and better devel-
oped in the latter species, is judged to be uniquely
derived within phallostethids.
Neostethus villadolidi and N. amaricola share
one derived character.
2 3 . Sharply pointed second ctenactinium (Fig.
9).
Neostethus villadolidi and N. amaricola are two
small, narrow-bodied phallostethid species dis-
FIGURE 9. Diagrammatic representation of head and an-
terior portion of body, a. Neostethus villadolidi, sinistral male,
CAS 60258, b. Neostethus amaricola, sinistral male, CAS-SU
38898. Anterior to left. Arrow points to second ctenactinium.
tinguished from all other phallostethids by their
sharply pointed, either anteriorly or dorsally di-
rected, second ctenactinia.
Autapomorphies of Each Species of Neostethus
Neostethus lankesteri
24. Males with a small, slightly pointed, tri-
angular, ventral bony projection on the elongate
first ctenactinium (Fig. 3).
The first ctenactinium is an elongate, uniform-
ly rod-shaped bone in all other phallostethids in
which it occurs. Neostethus lankesteri is unique
among phallostethids in having a small, slightly
pointed, triangular, bony projection on the ven-
tral surface of the first ctenactinium (Fig. 3).
25. Distal tips of the first pleural ribs project
beyond ventral outline of priapium.
260
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1 1
FIGURE 10. Diagrammatic representation of head and an-
terior portion of body, adult female, a. Neostethus thessa (CAS-
SU 36546), b. Neostethus villadolidi(CAS 60258), c. Neostethus
lankesteri (CAS-SU 67 1 62), d. Neostethus bicornis (CAS 35783).
Abbreviations: a, anus; u, urogenital opening.
The first pleural ribs in all phallostethid males
are offset to the proctal side of the priapium (Fig.
3). In TV. lankesteri, the distal tips of the ribs
project beyond the ventral outline of the pria-
pium when it is viewed from the proctal side.
26. Females have a large, hooklike, urogenital
papilla (Fig. lOc).
In all female phallostethids, the anus is ante-
rior to the urogenital openings (Fig. 10). Most
species are like N. thessa (Fig. lOa), in which
females have no anal or urogenital papilla or
covering. Adult female N. lankesteri are readily
distinguished from all other phallostethid fe-
males by a large, hooklike papilla that lies pos-
terior to the anal opening and anterior to the
urogenital openings, towards which it curves (Fig.
lOc).
Neostethus bicornis
21. Mature males with two, elongate ctenac-
tinia.
The second ctenactinium of TV. bicornis is the
longest of any phallostethid, at least one-half the
length of the first ctenactinium in adult males
(Roberts 1971b:fig. 12).
28. Immature males with a brown blotch on
the proctal pelvic rays that fades with growth and
maturity.
Immature males can be readily identified as
sinistral or dextral (see Parenti 1986a and char-
acter 44, below) by a small brown blotch present
on the proctal pelvic fin rays. This blotch be-
comes diffuse, hence less recognizable, with
growth.
29. Females with a fleshy, hoodlike flap cov-
ering the urogenital openings (Fig. 1 Od).
Adult female TV. bicornis are readily distin-
guished from all other phallostethids by the pres-
ence of a unique fleshy, hoodlike flap that extends
from the anus to the urogenital openings (Fig.
lOd). Although females of TV. lankesteri and TV.
bicornis have fleshy coverings of the urogenital
openings, I do not recognize this as a shared
derived character because form of the urogenital
covering is different in the two species, and nu-
merous other characters refute a close relation-
ship between the two.
Neostethus amaricola
30. Distal tips of the first pleural ribs of fe-
males project beyond the ventral body profile.
Herre ( 1 942: 1 48) used "Female with two spurs
or papillae on sides of throat" as a key character
of TV. amaricola. The spurs or papillae are the
distal tips of the first pleural ribs that project
beyond the ventral body profile on either side of
the urogenital openings.
31. A sharply pointed, anteriorly directed,
second ctenactinium (Fig. 9b).
Males of TV. amaricola are readily distin-
PARENTI: PHALLOSTETHID FISHES
261
guished from those of its sister species, N. vil-
ladolidi, by a sharply pointed second ctenactin-
ium that is directed anteriorly (Fig. 9).
Neostethus borneensis
Neostethus borneensis is a relatively short-
bodied species that is distinguished from all oth-
er neostethins by two, probably correlated, char-
acters (see G. panayensis, below, for a similar
definition).
32. Low number of vertebrae (31-32) (Table
2) in males and females.
Vertebrae in other species of Neostethus range
from 34 to 37 (Table 2), in Gulaphallus from 35
to 37, except for G. panayensis in which the num-
ber of vertebrae is also reduced, and in phallo-
stethins from 33 to 40 (Table 2). Vertebrae num-
ber 40—43 in Dentatherina merceri (fide Patten
and Ivantsoff 1983:332), and 29-30 in Pseudo-
mugil signifer (CAS 62565), two outgroup species.
The plesiomorphic vertebral number for phal-
lostethids is ambiguous. I treat the vertebral
number in N, borneensis as derived because it is
low compared to all other neostethins and phal-
lostethins.
33. Low number of scales in a lateral series
(25-27) (Table 2) in males and females.
Scales in a lateral series in other species of
Neostethus range from 29 to 34, and in Gula-
phallus from 28 to 58 (Table 2). Scales in a lateral
series number 40-43 in Dentatherina merceri (fide
Patten and Ivantsoff 1983:332), and 29-32 in
Pseudomugil signifer (CAS 62565), two outgroup
species. The plesiomorphic number of scales in
a lateral series in phallostethids is ambiguous. I
hypothesize that the number in TV. borneensis is
derived because it is low compared to outgroup
species and other phallostethids.
Neostethus villadolidi
34. A fleshy urogenital protuberance (Fig. lOb)
in females.
Adult female N. villadolidi have a fleshy pro-
tuberance just posterior to the anus and anterior
to the urogenital openings. It could be described
as an incipient or rudimentary papilla; however,
I do not recognize a transition series from the
state of the urogenital papilla in TV. lankesteri or
the urogenital flap in TV. bicornis, as discussed
above for character 29.
35. A sharply pointed, dorsally directed sec-
ond ctenactinium (Fig. 9a).
Male specimens of TV. villadolidi are readily
distinguished from male specimens of TV. amari-
cola, its proposed sister species, by a sharply
pointed second ctenactinium that is directed dor-
sally (Fig. 9).
Neostethus zamboangae
36. A small, hooked, second ctenactinium with
two to seven bony serrations.
Neostethus zamboangae adult males have two
to seven bony serrations on the ventral surface
of the hooked second ctenactinium. Regan (1913:
figs. 1, 3) illustrated serrations on what I identify
as the second ctenactinium in Phallostethus
dunckeri; however, I do not propose a close re-
lationship between N. zamboangae and P.
dunckeri because other characters lead me to
conclude that they are members of different,
monophyletic tribes.
Neostethus thessa
37. Absence of first dorsal fin rays and most
of the first dorsal pterygiophore in males and
females.
The first dorsal fin is absent in three phallo-
stethid species: N. thessa, Phallostethus dunckeri
(Parenti 1986a), and Gulaphallus bikolanus. I
view this absence as the individual loss of the
first dorsal fin in each species. A first dorsal fin
is plesiomorphic for atherinomorphs, being pres-
ent in outgroups Bedotia, Melanotaenia, Hy-
poatherina, Dentatherina, and Pseudomugil and
all other phallostethids. Each of the three phal-
lostethid species that lacks a first dorsal fin is a
member of its own well-corroborated monophy-
letic group, and a remnant of the first dorsal pte-
rygiophore is present in some specimens.
38. Oral dentition reduced to several small
unicuspid teeth on premaxilla and dentary in
males and females.
Oral teeth, in one or two rows in other Neo-
stethus species, are reduced to a few randomly
distributed unicuspid teeth in TV. thessa.
39. A relatively long anal fin ( 1 7-22 rays) (Ta-
ble 2) in males and females.
Anal fin rays in other Neostethus and Gula-
phallus number from 1 3 to 18 (Table 2), in Phe-
nacostethus from 14 to 15 (Table 2). Phalloste-
thus dunckeri is the one other phallostethid species
having a relatively long anal fin with 26-28 rays
(Table 2). In one outgroup species, Dentatherina
262
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1 1
5O
151 1
152 1
149 1
154
1
146 1
153
i
1
r
147
4 r
346
]45
]44
143
P
c
1 [
1
14
• '3
1112
FIGURE 1 1 . Cladogram of relationships among species in
the Gulaphallini. Solid squares, synapomorphies; open squares,
symplesiomorphies. Characters numbered as in Phylogenetic
Analysis.
merceri, anal fin rays number from 1 5 to 17 (Pat-
ten and Ivantsoff 1983).
40. A relatively long second dorsal fin (8-10
rays) (Table 2) in males and females.
Second dorsal fin rays in other Neostethus range
from 5 to 7, in Gulaphallus from 6 to 8, in Phena-
costethus from 5 to 7, and in Phallostethus, 8-
10 (Table 2). Absence of the first dorsal fin and
increase in the number of second dorsal fin rays
in both N. thessa and P. dunckeri invites spec-
ulation that the two dorsal fins have become con-
fluent in development. However, absence of the
first dorsal fin in Gulaphallus bikolanus with no
increase in number of second dorsal fin rays re-
futes this as a general phenomenon in phallo-
stethids.
Neostethus palawanensis
4 1 . A relatively straight first ctenactinium with
a broad, fleshy ventral ridge.
The first ctenactinium is the prominent exter-
nalized priapial bone in neostethins and gula-
phallins (see character 5, above). It is typically a
long, slender, curved bone (Figs. 3, 4) except in
TV. palawanensis in which it is relatively straight
and has a thick, fleshy ventral ridge (Myers 1 935).
I consider the fleshy ventral ridge uniquely de-
rived in N. palawanensis.
Neostethus robertsi
42. A large, well-developed, claw-shaped
second ctenactinium (Fig. 1 2).
Neostethus robertsi and N. palawanensis are
considered sister species because they share the
uniquely derived character of a claw-shaped, an-
teriorly bifurcated, second ctenactinium (see
character 22, above). The claw-shaped second
ctenactinium is much larger and better devel-
oped in N. robertsi than in N. palawanensis.
Neostethus ctenophorus
There are no identifiable autapomorphies of
TV. ctenophorus (see Systematic Accounts).
Gulaphallini
Relationships among the five Gulaphallus
species are depicted in Figure 1 1 . Four species,
G. eximius, G. mirabilis, G. falcifer, and G. bi-
kolanus, are considered to be most closely related
because they share one derived character.
43. Large, well-developed aproctal axial bone
(Fig. 4).
The aproctal axial bone is relatively large and
has a well-developed dorsal and ventral ridge in
G. eximius, G. mirabilis, G. falcifer, and G. bi-
kolanus (Fig. 4), as opposed to the condition in
G. panayensis in which the bone is thin and elon-
gate.
Gulaphallus bikolanus and G. eximius are con-
sidered sister species because they share one de-
rived character.
44. Adult males exclusively dextral.
All adult male phallostethids are bilaterally
asymmetric: a sinistral male has the seminal pa-
pilla offset to the left side of the body and the
anus offset to the right (Parenti 1986a). The re-
verse is true for dextral males. Sinistral and dex-
tral males occur in about equal numbers in all
phallostethid species except the sister species
Phenacostethus trewavasae, which is sinistral,
Phenacostethus posthon, which is dextral, and the
two species of Gulaphallus. I (Parenti 1986a) re-
viewed the phenomenon of bilateral asymmetry
PARENTI: PHALLOSTETHID FISHES
263
in phallostethids and speculated that the Phen-
acostethus sister species evolved from an ances-
tral species in which there were both sinistral and
dextral males and concluded that such state-
ments of ancestry are untestable. My hypothesis
that G. eximius and G. bikolanus are sister species
because males of each are exclusively dextral does
not refute this argument about speciation. The
proposal requires that the common ancestor of
the two Gulaphallus species was exclusively dex-
tral. If that ancestral species had an exclusively
sinistral sister species, it is unknown to us now.
Gulaphallus mirabilis and G. falcifer are con-
sidered to be sister species because they share
two derived characters.
45. Antepleural element ossified in males and
females.
The antepleural cartilage surrounds the distal
tips of the first pleural ribs. It is ossified in G.
mirabilis and G. falcifer and cartilaginous in oth-
er species of Gulaphallus.
46. A brown blotch (concentration of mela-
nophores) at midventral surface just anterior to
anal fin in males and females.
Parenti ( 1 986b) used this pigmentation pattern
as an identifying character of G. falcifer. How-
ever, it is found also in G. mirabilis and hence
used as a derived character to support a close
relationship between the two species.
Autapomorphies of Each Species of Gulaphallus
Gulaphallus eximius
47. Relatively large body size and scales in a
lateral series more than 50 (range 52-58) (Table
2) in males and females.
Gulaphallus eximius is the largest known phal-
lostethid, reaching a recorded 37 mm SL (Herre
1942). Although large size may be considered
plesiomorphic for phallostethids, G. eximius is
perhaps best described as secondarily large. One
possible sister taxon, Dentatherina merceri, is
relatively small, maturing at under 25 mm SL
(Patten and Ivantsoff 1983). Correspondingly, G.
eximius has a high number of scales in a lateral
series, ranging from 52 to 58, as opposed to a
range of 25-34 for all other neostethins and
gulaphallins except G. mirabilis (see character
50, below). Also, outgroup species, D. merceri
and Pseudomugil signifer have fewer scales than
does G. eximius (see character 33, above).
48. A brown blotch (concentration of mela-
nophores) surrounds the anus and urogenital
openings in immature females.
This pigmentation pattern is not found in oth-
er phallostethids nor in outgroup species and,
therefore, is considered uniquely derived in G.
eximius.
49. Rudimentary pelvic girdle represented by
two small scale-like discs of bone in adult fe-
males.
Just posterior to the urogenital openings in
females is a pair of ossified discs inferred to be
a rudimentary pelvic girdle. Similar pelvic girdle
elements are found in adult female Phenacoste-
thus posthon (see Parenti 1986b:fig. 4).
Gulaphallus mirabilis
50. Scales in lateral series range from 36 to 38
(Table 2).
The high number of scales in a lateral series
in G. mirabilis is judged to be an independent
increase (see character 47, above).
Gulaphallus falcifer
5 1 . Adult females with an elongate pelvic fin
ray projecting beyond body profile on either the
left or right side of the urogenital opening.
Parenti (1986b) described development of the
elongate asymmetric bony element in females of
G. falcifer, concluding that it is homologous with
a pelvic fin ray, not a postcleithrum. Gulaphallus
falcifer is the only phallostethid in which bilateral
asymmetry has been described in females.
Gulaphallus bikolanus
52. First dorsal fin absent in males and fe-
males.
Absence of the first dorsal fin in G. bikolanus
is hypothesized to be independent of its absence
in N. thessa and P. dunckeri (see character 37,
above) for reasons of parsimony.
Gulaphallus panayensis
Gulaphallus panayensis is a relatively small
Gulaphallus species defined by two, probably
correlated, characters (see N. borneensis, above,
for a similar definition).
53. A low number of vertebrae (3 1 -3 3) (Table
2) in males and females.
264
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1 1
The low number of vertebrae in G. panayensis
is considered a reduction from a higher, plesio-
morphic number for phallostethids (see charac-
ter 32, above).
54. A low number of scales in a lateral series
(28-30) (Table 2) in males and females.
The low number of scales in a lateral series is
considered a reduction from a higher, plesio-
morphic number for phallostethids (see charac-
ter 33, above).
KEY TO PHALLOSTETHIDAE
The following key is based primarily on char-
acters discussed in the Phylogenetic Analysis.
However, characters used to identify the species
of Phallostethus and Phenacostethus are from
Parenti(1986a).
1A. Second ctenactinium present (Figs. 2, 3,
9); shieldlike pulvinulus present (Figs. 2,
9); outer pulvinular bone robust (Figs.
2, 3); oral jaws highly protrusible; few or
no teeth on fourth ceratobranchial
toothplate; no perforation in skin of gu-
lar region; aproctal axial bone does not
project beyond ventral body profile; no
dorsal crest on proctal axial bone
Phallostethinae 2A
1 B. No second ctenactinium; no pulvinulus;
outer pulvinular bone small; oral jaws
not highly protrusible; teeth on fourth
ceratobranchial toothplate; perforated
gular flap of skin through which anterior
end of first ctenactinium may pass (Fig.
5); aproctal axial bone projects beyond
ventral body profile (Figs. 4, 5); small,
dorsal crest posteriorly on proctal axial
bone (Fig. 4) Gulaphallini 1 5 A
2A. Large, fleshy, smooth or ruffled, seminal
papilla; large, externalized, curved tox-
actinium; translucent, membranous
dome on dorsal surface of head; no inner
pulvinular bone; no bony projections on
papillary bone Phallostethini 3A
2B. Seminal papilla small; no externalized
toxactinium; no translucent, membra-
nous dome on dorsal surface of head;
inner pulvinular bone present (Fig. 3);
high number of thin, bony projections
on papillary bone (Fig. 3) or just rudi-
mentary projections Neostethini 6A
3A. Anal fin rays 26-28; vertebrae 40; sec-
ond ctenactinium serrated; oral jaws
equal; no first dorsal fin; second dorsal
fin rays 8-10
Phallostethus dunckeri (p. 267)
3B. Anal fin rays 14-15; vertebrae 33-35;
second ctenactinium greatly reduced and
not serrated; lower oral jaw projects be-
yond upper oral jaw; first dorsal fin ray
1 ; second dorsal fin rays 5-7
Phenacostethus 4A
4A. Distal end of seminal papilla ruffled;
slightly curved toxactinium (Fig. 2), pe-
nial bone present
Phenacostethus smithi (p. 268)
4B. Distal end of seminal papilla smooth;
greatly curved toxactinium; penial bone
absent 5A
5 A. First dorsal fin origin posterior to base of
last anal fin ray; males dextral; eye-lens
large — Phenacostethus posthon (p. 268)
5B. First dorsal fin origin opposite midpoint
of anal fin; males sinistral; eye-lens small
to minute
Phenacostethus trewavasae (p. 268)
6A. Second ctenactinium small; small,
slightly pointed, ventral bony projection
on elongate first ctenactinium (Fig. 3);
posterior tips of first pleural ribs project
beyond body of priapium; females with
large, hooklike genital papilla (Fig. 1 Oc)
Neostethus lankesteri (p. 269)
6B. Second ctenactinium large; ventral mar-
gin of first ctenactinium smooth, no
pointed projection; posterior tips of first
pleural ribs do not project beyond body
of priapium; females with urogenital pa-
pilla that may be fleshy but does not form
distinct hook 7A
7A. Long aproctal axial bone, reaches pul-
vinulus . . Neostethus ctenophorus (p. 272)
7B. Short, bladelike aproctal axial bone, does
not reach pulvinulus 8A
8A. Strongly curved, thin, second ctenactin-
ium; papillary and penial bones curved
and elongate, lying parallel to dorsal sur-
face of second ctenactinium; proctal
transverse processes of fourth vertebra
of males enlarged and oriented anterior-
ly borneensis group 9A
8B. Straight, stout, second ctenactinium;
papillary and penial bones short and
straight; proctal and aproctal transverse
processes of fourth vertebra of males en-
PARENTI: PHALLOSTETHID FISHES
265
larged and oriented anteriorly
bicornis group 1 1A
9A. Anterior aproctal ossification present;
few or no bony projections on papillary
bone; vertebrae 35-37; scales in lateral
series 29-34 10A
9B. No anterior aproctal ossification; nu-
merous bony projections on papillary
bone; vertebrae 31-32; scales in lateral
series 25-27
Neostethus borneensis (p. 270)
1 0 A. Small, hooked, second ctenactinium with
2-7 bony serrations; first dorsal fin rays
2; second dorsal fin rays 5-6; complete
row of teeth on premaxilla and on den-
tary; anal fin rays 1 5-1 7
Neostethus zamboangae (p. 271)
I OB. Small, hooked second ctenactinium with
no bony serrations; no first dorsal fin;
second dorsal fin rays 8-10; several small
unicuspid teeth on premaxilla and on
dentary; anal fin rays 1 7-22
Neostethus thessa (p. 272)
II A. Mature males with two elongate cte-
nactinia; immature males with brown
blotch on proctal pelvic rays; females
with fleshy, hoodlike flap covering uro-
genital openings (Fig. lOd)
Neostethus bicornis (p. 269)
11B. Mature males with one elongate cte-
nactinium; immature males with clear
pelvic rays; urogenital openings of fe-
males exposed, not covered by fleshy,
hoodlike flap 12A
1 2 A. Claw-shaped, anteriorly bifurcated, sec-
ond ctenactinium 1 3A
12B. Sharply pointed second ctenactinium
(Fig. 9) 14A
13 A. Straight first ctenactinium with broad,
fleshy, ventral ridge; small, claw-shaped
second ctenactinium
Neostethus palawanensis (p. 271)
13B. Slightly bowed first ctenactinium with
smooth ventral outline, no fleshy ridge;
large, claw-shaped second ctenactinium
(Fig. 13) Neostethus robertsi (p. 272)
1 4 A. Distal tips of first pleural ribs of females
project beyond ventral body profile;
sharply pointed, anteriorly directed sec-
ond ctenactinium (Fig. 9b); no fleshy
protuberance anterior to urogenital
openings of females
Neostethus amaricola (p. 270)
1 4B. Distal tips of first pleural ribs of females
do not project beyond ventral body pro-
file; sharply pointed, dorsally directed
second ctenactinium (Fig. 9a); fleshy
protuberance anterior to urogenital
openings in females (Fig. lOb)
Neostethus villadolidi (p. 271)
1 5A. Thin, elongate aproctal axial bone; ver-
tebrae 31-33; scales in lateral series 28-
30 Gulaphallus panayensis (p. 275)
15B. Large, well-developed aproctal axial
bone (Fig. 4); 35-37 vertebrae; scales in
lateral series 32 or more 16A
1 6A. Males dextral; antepleural element car-
tilaginous in adults; midventral surface
with tiny, scattered melanophores . . 1 7 A
16B. Males sinistral or dextral; antepleural
element ossified in adults; brown blotch
at midventral surface just anterior to anal
fin ISA
17 A. Scales in lateral series 52-58; brown
blotch surrounds anus and urogenital
openings in immature females; rudi-
mentary pelvic girdle represented by two
small scalelike discs of bone in adult fe-
males; first dorsal fin present with rays
1-2 Gulaphallus eximius (p. 274)
1 7B. Scales in lateral series 32; anus and uro-
genital openings in females surrounded
by tiny, scattered melanophores; no ru-
dimentary pelvic girdle in adult females;
no first dorsal fin
Gulaphallus bikolanus (p. 275)
1 8 A. Scales in lateral series 36-38; pelvic rays
do not project beyond ventral body pro-
file Gulaphallus mirabilis (p. 274)
18B. Scales in lateral series 32; an elongate
pelvic fin ray projecting beyond body
profile on either the left or right side of
urogenital opening in adult females . . .
Gulaphallus falcifer (p. 274)
CLASSIFICATION OF PHALLOSTETHIDAE
The following classification of phallostethids
reflects some conclusions of the Phylogenetic
Analysis. All higher taxa are monophyletic.
Species are listed in order of year of description.
Family Phallostethidae Regan, 1913
Subfamily Phallostethinae Regan, 1913
Tribe Phallostethini new status
Genus Phallostethus Regan, 1913
P. dunckeri Regan, 1913
266
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1 1
Genus Phenacostethus Myers, 1 928
P. smithi Myers, 1928
P. posthon Roberts, 1971
P. trewavasae Parent!, 1986
Tribe Neostethini new status
Genus Neostethus Regan, 1916
N. lankesteri Regan, 1916
N. bicornis Regan, 1916
N. palawanensis (Myers, 1935), new
combination
TV. amaricola (Villadolid and Manacop,
1935)
N. thessa (Aurich, 1937), new combi-
nation
N. ctenophorus (Aurich, 1937), new
combination
N. borneensis Herre, 1939
TV. villadolidi Herre, 1942
N. zamboangae Herre, 1 942
N. robertsi new species
Subfamily Gulaphallinae Aurich, 1937
Tribe Gulaphallini new status
Genus Gulaphallus Herre, 1925
G. eximius Herre, 1925
G. mirabilis Herre, 1925
G. bikolanus (Herre, 1926), new com-
bination
G. falcifer Manacop, 1936
G. panayensis (Herre, 1 942), new com-
bination
SYSTEMATIC ACCOUNTS
Family Phallostethidae Regan, 1913
Type genus Phallostethus Regan, 1913
Phallostethinae Regan, 1913 (as subfamily of Cyprinodonti-
dae)
Neostethinae Aurich, 1937 (as subfamily of Phallostethidae,
type genus Neostethus Regan)
Gulaphallinae Aurich, 1937 (misprinted Gulaphalinae, as
subfamily of Phallostethidae, type genus Gulaphallus Herre)
DIAGNOSIS.— Small to minute, laterally com-
pressed, nearly transparent, atherinomorph fish-
es, largest size recorded 37 mm SL. Anus and
urogenital openings anterior, below pectoral fin
base, in both sexes. A median, abdominal, slight-
ly frayed, fleshy ridge or keel from urogenital
openings to anterior rays of anal fin. First dorsal
fin with one or two short spines or thickened
rays, or no rays but pterygiophore may be pres-
ent.
Copulatory organ of males a priapium. The
following characters are of males, unless other-
wise noted. Main bony priapial support the proc-
tal axial bone. Main externalized bone(s) a tox-
actinium, a toxactinium and a ctenactinium, or
one or two ctenactinia. Anterior ramus of clei-
thrum elongate anteriorly or not. Postcleithra ab-
sent or present as small slivers of bone in both
sexes. First pair of pleural ribs expanded antero-
posteriorly and dorsoventrally, ventral tips
meeting and enclosed in block of cartilage and/
or bone, the antepleural element. Large, fibrous
pad, the pulvinulus, if present, covers point of
articulation of toxactinium, proctal axial bone,
and anterior extent of cleithra. Fertilization in-
ternal, sperm bundles passed to females through
seminal papilla, which may be large or small,
smooth or ruffled, and supported internally by
papillary and penial bones. Development ovipa-
rous; females lay fertilized eggs.
Males bilaterally asymmetric: anus and sem-
inal papilla offset to opposite sides of body, proc-
tal and aproctal, respectively. Prominent exter-
nalized priapial bones arise on left or right side
of body. Proctal, or proctal and aproctal, trans-
verse processes of fourth vertebrae expanded and
oriented anteriorly or not. Females bilaterally
symmetric except those of Gulaphallus falcifer,
in which one pelvic ray is enlarged on either left
or right side of body at urogenital openings. Pel-
vies otherwise absent or represented by bundle
of fin rays, called fringe, surrounding epididymis.
Small postanal papillae may contain rudimen-
tary pelvic girdles and rays in adult female Phe-
nacostethus posthon and Gulaphallus eximius.
Hyobranchial apparatus with reductions from
typical atherinoid condition; teeth on fourth cer-
atobranchial toothplate present or absent. Bran-
chiostegal rays 4 to 6.
DESCRIPTION. — Background uniformly pale
yellow or straw-colored in alcohol, with small,
dark melanophores on dorsal surface of head and
body, midlateral intermuscular septum, along
anal fin base, dorsal and ventral midline. Seminal
papilla and externalized bones of priapium may
be covered with scattered melanophores. Scales
cycloid, relatively small to medium-sized, and
deciduous. Dorsal scales with faint dark margins.
Eye-lens large to minute. Dorsal surface of head
with translucent, membranous dome or not.
Mouth slightly to highly protractile with thin or
no meniscus, cartilaginous rod, or ossified sub-
maxillary element between maxilla and vomer.
Unicuspid teeth in single or double row, with
large teeth on lateral ramus of premaxilla, or few,
PARENTI: PHALLOSTETHID FISHES
267
small teeth. Second dorsal fin with 5-10 rays.
Anal fin unmodified, with 13-28 rays. Pectoral
fin with 9-13 rays. Caudal fin emarginate, form-
ing incipient lobes. Caudal skeleton with two
epurals, autogenous parhypural, and dorsal and
ventral hypural plate. Vertebrae 31-40.
REMARKS. -Regan (19 13, 19 16), Myers (1928),
Bailey (1 936), Herre( 1942), and Roberts (1 97 lb)
listed both diagnostic and descriptive characters
of phallostethid fishes. This diagnosis, like the
ones that follow, is phylogenetic, as in Weitzman
and Fink (1985) for genera of xenurobryconin
characid fishes. Only characters or character states
hypothesized to be derived are considered di-
agnostic of taxa. Comparisons among characters
used to diagnose species, tribes, and subfamilies
are given in the Phylogenetic Analysis and dif-
ferential diagnoses.
Subfamily Phallostethinae Regan
DIAGNOSIS.— Priapium with a reduced second
ctenactinium, and a prominent, shieldlike pul-
vinulus covering articulation point of robust tox-
actinium and proctal axial bone. Jaws of males
and females highly protrusible with elongate pre-
maxillary ascending processes and distinct sub-
maxillary cartilages or bones, and a reduced hyo-
branchial apparatus with few or no teeth on fourth
ceratobranchial toothplate.
COMPOSITION.— Two tribes, Phallostethini and
Neostethini, both family-group names used for
the first time at the tribal level.
Tribe Phallostethini new status
DIAGNOSIS.— Males with a large, fleshy sem-
inal papilla that may be smooth or ruffled; tox-
actinium relatively straight to curved. Adults of
both sexes extremely small and delicate, with
deciduous scales; a translucent, membranous
dome on dorsal surface of head.
COMPOSITION AND DISTRIBUTION.— Two gen-
era, Phallostethus Regan from coastal peninsular
Malaysia, and Phenacostethus Myers, from
coastal peninsular Thailand and Malaysia and
Sarawak, as diagnosed below.
Genus Phallostethus Regan, 1913
Phallostethus Regan, 1913:548-549 (type species: Phalloste-
thus dunckeri Regan, 1 9 1 3 by monotypy).
DIAGNOSIS.— As for P. dunckeri Regan, 1913,
below.
COMPOSITION AND DISTRIBUTION.— One species
from brackish water, mouth of Muar R., Johore,
Malaysia (see Roberts 1971a:fig. 1).
Phallostethus dunckeri Regan, 1913
(Fig. i)
Phallostethus dunckeri Regan, 1913:548-555, figs. 1-4 (origi-
nal description, mouth of Muar R., Johore, Malay Penin-
sula); 1916:1, 16-25, figs. 12a, 13-15, pis. la, Ilia, IVa, b
(characters, comparison with Neostethus). Myers 1928:1-6
(characters, comparison with Phenacostethus); 1937:137
(characters). Herre 1939:140 (characters). Roberts 1971a:l-
20 (relationships). Parenti 1984:1-12, figs. 1-4 (osteology,
relationships); 1986a:229-235, tables 1, 2 (relationships,
characters); 1986c:41 (characters).
DIFFERENTIAL DIAGNOSIS. — Phallostethus
dunckeri has an anal fin longer (26-28 rays, as
opposed to 1 3-22) and vertebral number higher
(40, as opposed to 31-37) than any other phal-
lostethid. Males are readily separable from those
of Phenacostethus by a serrated second ctenactin-
ium. Both sexes distinct from Phenacostethus by
lacking first dorsal fin, and a second dorsal fin
with 8-10 rays, rather than 5-7.
MATERIAL.— Lectotype (designated by Ladiges et al. 1958):
ZMH 193, dextral 3, mouth of Muar R., Johore, Malaysia,
1 902, G. Duncker, coll. Paralectotypes: BMNH 1 9 1 3.5.24: 1 8-
22 (5 spec.: 1 sinistral <5, 3 9, 1 sex undet.) and ZMH 194-195
(19 spec.: 2 dextral S, 14 9, 3 juv.), collected with lectotype.
REMARKS.— Ladiges et al. (1958) gave lecto-
type and paratype (=paralectotype) status to ZMH
193 and 194, respectively. Because ZMH 195
and the BMNH specimens are of the original
syntypic series, they are also paralectotypes. Pa-
renti (1 986a) called all these specimens syntypes.
Genus Phenacostethus Myers, 1928
Phenacostethus Myers, 1928:6 (type species: Phenacostethus
smithi Myers, 1928 by original designation).
DIFFERENTIAL DIAGNOSIS.— A lower jaw pro-
truding beyond upper jaw and greatly reduced
second ctenactinium distinguishes Phenacoste-
thus from all other phallostethids. Anal and sec-
ond dorsal fin rays and vertebrae are fewer than
in Phallostethus. First dorsal fin with one ray as
opposed to first dorsal fin absent in Phallostethus.
COMPOSITION AND DISTRIBUTION.— Phenaco-
stethus smithi Myers, 1 928 and P. posthon Rob-
erts, 197 la, in fresh and brackish waters, Thai-
land and peninsular Malaysia; and P. trewavasae
Parenti, 1986a, in freshwater, Sarawak, Malay-
sian Borneo.
268
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1 1
Phenacostethus smithi Myers, 1 928
(Figs. 1, 2)
Neostethus lankesteri (non Regan). Smith 1927:353-355 (re-
port of specimens from Bangkok, Thailand).
Phenacostethus smithi Myers, 1 928:6-8, figs. 1 , 2 (original de-
scription, Bangkok, Thailand). Smith 1929:14 (characters,
distribution). Myers 1937:137-138 (characters). Bailey 1936:
470_474; pis. 3, 4 (anatomy, comparison with Gulaphallus).
TeWinkel 1939:60-68, figs. 1-5 (anatomy, comparison with
Gulaphallus). Herre 1939:140, 142 (characters, listed). Hubbs
1944:73 (characters). Smith 1945:475-476 (ecology). Rob-
erts 1971a:4-18, figs. 1, 2, 4, 7 (mislabelled as fig. 6) (anat-
omy, relationships). Parenti 1986a:229-234, fig. 4a, tables
1, 2 (characters, relationships).
Phenacostethus thai Fowler, 1937:219-220, figs. 189, 190
(original description, Bangkok, Thailand).
DIFFERENTIAL DIAGNOSIS. — Unlike other
phallostethins, males of Phenacostethus smithi
have a ruffled, rather than smooth, distal end of
enlarged seminal papilla.
MATERIAL.— Holotype: AMNH 9247, dextral 6, 13.5 mm
SL, Bangkok, Thailand, Dec. 1926, H. M. Smith, coll. Para-
types: AMNH 9248 (1 2), USNM 92979 (1 1 spec.: 4 sinistral
S, 1 dextral 6, 5 2, 1 sex undet.), USNM 202564 (1 2), USNM
104417 (1 sinistral 3), CAS-SU 26520 (8 spec.: 2 sinistral 3, 3
dextral 6, 2 2, 1 juv.), ANSP 108764 (ex. SU 26520) (2 spec.:
1 dextral 3, 1 2) and USNM 92297 (2 spec.: 1 dextral 3, 1 2),
collected with holotype.
Holotype and paratypes of Phenacostethus thai: ANSP 51352,
ANSP 51353-60 (9 spec.: 2 sinistral 3, 2 dextral 3, 5 2), Bang-
kok, Thailand, 2 Aug. 1923, H. M. Smith, coll.
OTHER MATERIAL. -THAILAND: UMMZ 87345 (40 spec.:
10 sinistral 3, 14 dextral 3, 1 5 2, 1 juv.), MCZ 47055 (13 spec.:
6 sinistral 3, 2 dextral 3, 5 2), USNM 93506 (100+), USNM
93507 (50+), USNM 93508 (36 spec.: 5 sinistral 3, 12 dextral
3, 17 2, 2 juv.), USNM 88668 (100+), USNM 88667 (100+
spec.; 4 juv. cleared and stained), USNM 88659 (3 spec.: 1
sinistral 3, 1 dextral 3, 1 2), BMNH 1927.12.29:1-10(31 spec.:
6 sinistral 3, 4 dextral 3, 102, 11 juv.; 2 cleared and stained),
ZMA 100.652 (35), and CAS-SU 35957 (112 spec.: 5 cleared
and stained), Bangkok, and MCZ 47299 (18 spec.; 1 sinistral
3, 1 dextral 3 and 2 2 cleared and stained), Chantaburi City.
MALAYSIA: USNM 266124 (50+), Trengganu; USNM
266125 (3 spec.: 1 sinistral 3, 2 2), USNM 266126 (15 spec.:
3 sinistral 3, 5 dextral 3, 7 2), USNM 266127 (1 2), Johore.
REMARKS. — In the original description of
Phenacostethus smithi, Myers (1928:8) stated that
"... nineteen paratypes, 1 1 males and 8 females,
are deposited in the American Museum of Nat-
ural History, the United States National Mu-
seum, and my own collection." I assume that
Myers's "own collection" was deposited, in part,
in the Stanford University (SU) collection, now
CAS-SU 26520 and ANSP 108764. However, I
have no reason to exclude any of the above
USNM paratypes of Phenacostethus smithi from
type status.
In his description of Phenacostethus thai, Fow-
lers (1937) did not separate what he called the
holotype from other type specimens and I am
unable to separate that specimen from the rest
of the type material. See also Bohlke (1984: 140).
DISTRIBUTION.— Coastal fresh and brackish
waters, Thailand and peninsular Malaysia (Rob-
erts 1971a:fig. 1).
Phenacostethus posthon Roberts, 1971
(Fig. i)
Phenacostethus posthon Roberts, 197 la: 12-1 7, figs. 3, 5, 6
(mislabelled as fig. 7) (original description, Pungah and Lan-
gu Town, Thailand). Parenti 1986a:225-235, fig. 5, tables
1, 2 (characters, relationships); 1986b:308-309, figs. 4, 5
(anatomy).
DIFFERENTIAL DIAGNOSIS.— Phenacostethus
posthon is unique among phallostethins, having
a first dorsal fin origin posterior to base of last
anal fin ray rather than opposite midpoint of anal
fin so that first and second dorsal fins are rela-
tively close together. Males are dextral, as in Gu-
laphallus eximius and G. bikolanus, as opposed
to sinistral, as in P. trewavasae, or sinistral or
dextral, as in all other phallostethids.
MATERIAL.— Holotype, MCZ 47300, 3, 16.7 mm SL, Khlong
Kla Sohm, about 15 km SW of Pungah Town, Pungah Prov.,
Thailand, 29 June 1970, T. R. Roberts, coll. Paratypes: MCZ
47301 (2 3 cleared and stained), MCZ 47301 A (22 2), MCZ
4730 IB (32 3), same data as holotype, and MCZ 47302 (20),
Khlong Langu at Langu Town, 48 km NW of Satul Town,
Thailand, 26 June 1970, T. R. Roberts, coll.
OTHER MATERIAL. -MALAYSIA: USNM 229302 (182 spec.:
82 3, 77 2, 23 juv.; 6 cleared and stained), Muar R., Johore.
DISTRIBUTION.— Coastal fresh and brackish
waters, Thailand and peninsular Malaysia (Rob-
erts 1971a:fig. 1).
Phenacostethus trewavasae Parenti, 1986
(Fig. 1)
Phenacostethus trewavasae Parenti, 1986a:226-229, figs. 1-3,
4b, tables 1, 2 (original description, Baram R., Sarawak,
Malaysian Borneo).
DIFFERENTIAL DIAGNOSIS.— Phenacostethus
trewavasae is unique among phallostethids, hav-
ing a minute eye-lens, and males sinistral rather
than dextral, as in P. posthon, Gulaphallus exi-
mius and G. bikolanus, or sinistral or dextral, as
in all other phallostethids.
MATERIAL.— Holotype: ROM41826.3, 14.1 mmSL,Sungei
Kejin Tugang, tributary of Sungei Kejin, Baram R., Sarawak,
Malaysia, 3 Aug. 1981, D. Watson, coll. Paratypes: ROM 44289
( 1 9), ROM CSS 1 2 (3 spec. : 2 3, 1 2, cleared and stained), ROM
41827 (1), collected with holotype. ROM 41829 (1), ROM
PARENTI: PHALLOSTETHID FISHES
269
4 1830 (13), CAS 55454 (5 spec.; 1 cleared and stained), BMNH
1 984.7. 1 2: 1-5 (5), AMNH 55570 (5), USNM 267266 (5), Sun-
gei Kejin, station at confluence of Kejin Tugang and Kejin R.,
Baram R., Sarawak, Malaysia, 1 1 Feb. 1980, D. Watson, coll.
OTHER MATERIAL. -MALAYSIA: ROM 41828 (1 1), ROM
44290 (29 spec.: 2 $ and 2 9 cleared and stained) and ROM
44291 (50), Sarawak.
DISTRIBUTION.— Tributaries of Baram R., Sa-
rawak, Malaysian Borneo (Fig. 1).
Tribe Neostethini new status
DIAGNOSIS.— Priapium with an inner pulvin-
ular bone, and thin bony projections numbering
80 or more on papillary bone.
Genus Neostethus Regan, 1916
Neostethus Regan, 1916:1, 2 (type species: Neostethus lankes-
teri Regan, 1916 by original designation, use of "gen. et sp.
n." for one of two included new species).
Ptectrostethns Myers, 1935:5 (type species: Plectrostethus pa-
lawanensis Myers, 1935, by original designation).
Ceratostethus Myers, 1937:141 (type species: Neostethus bi-
cornis Regan, 1 9 1 6, by original designation).
Solenophallus Aurich, 1937:264 (unavailable; no type species
designated of the two treated under genus).
Ctenophallus Herre, 1939:144 (type species: Solenophallus
ctenophorus Aurich, 1937, by original designation).
Sandakanus Herre, 1942:151 (type species: Neostethus bor-
neensis Herre, 1 939, by original designation, as a subgenus
of Neostethus Regan).
Solenophallus Herre (ex Aurich), 1953:242 (type species: So-
lenophallus thessa Aurich, 1 937, by original designation).
DIAGNOSIS.— As for the tribe Neostethini,
above.
COMPOSITION AND DISTRIBUTION. — Ten
species: N. lankesteri Regan, 1916, Thailand,
peninsular Malaysia, Singapore, and Borneo; N.
bicornis Regan, 1916, peninsular Malaysia, Sin-
gapore, Thailand, Borneo, and Palawan, Phil-
ippines; N. palawanensis (Myers, 1935), Pala-
wan, Philippines; N. thessa (Aurich, 1937),
Mindanao, Philippines; N. ctenophorus (Aurich,
1937), Luzon, Philippines; N. borneensis Herre,
1939, Borneo and Coron, Philippines; N. zam-
boangae Herre, 1942, Mindanao and Luzon,
Philippines; N. robertsinew species, Luzon, Phil-
ippines; N. amaricola (Villadolid and Manacop,
1 935), and N. villadolidi Herre, 1 942, throughout
Philippines.
Neostethus lankesteri Regan, 1916
(Figs. 1,3, 10c)
Neostethus lankesteri Regan, 1916:2-14, figs. 1-10, 12B, pis.
IB, 2, 3A, 4C (original description, Muar R., Johore, pen-
insular Malaysia, and Singapore). Weber and DeBeaufort
1922:381-383, fig. 103 (characters). Myers 1928:8 (charac-
ters). Smith 1929:13 (distribution). Myers 1937:139 (listed).
Herre 1939:143 (listed). Roberts 1971a:8 (characters). Ivan-
tsoff et al. 1987:651-652, figs. 2A-C, 3A (anatomy, com-
parisons with atherinoids).
Neostethus siamensis Myers, 1937:139-140 (original descrip-
tion, Chantabun R., Thailand). Herre 1 939: 1 43 (listed). Smith
1945:477 (compiled). Roberts 197 la:2-3 (collection report).
Parenti 1986c:39^»l, fig. 1 (osteology, histology).
Neostethus (Neostethus) lankesteri. Herre 1942: 149 (characters,
relationships, distribution).
Neostethus (Neostethus) siamensis. Herre 1 942: 1 50 (characters,
probable synonym of N. lankesteri Regan).
DIFFERENTIAL DIAGNOSIS.— Neostethus lan-
kesteri males can be recognized by a small, slight-
ly pointed, triangular, ventral bony projection on
elongate first ctenactinium, and posterior tips of
first pleural ribs projecting beyond body of pria-
pium. Females are unique among phallostethids
having a large, hooklike, posteriorly projecting,
urogenital papilla.
MATERIAL.— Lectotype (here designated): BMNH 1937.12.9:
7, dextral 3, 26.0 mm SL, Muar R., peninsular Malaysia, 18
Feb. 1902, G. Duncker, coll. Paralectotypes: BMNH 1937.12.9:
8-10 (3 spec.: 1 dextral 3, 1 sinistral 6, 1 9) Singapore, G.
Duncker, coll. Holotype of Neostethus siamensis: USNM
102140, adult 9, 28.7 mm SL, Chantabun R., Thailand, Apr.
1933, H.M. Smith, coll.
OTHER MATERIAL. -SINGAPORE: BMNH 1970.7.22:71-
77 (7 spec.: 1 sinistral 3, 1 dextral 3, 5 9), R. Tengah; BMNH
1970.7.22:78-79 (2 9), R. Serangoon; RMNH 26425 (69 spec.:
22 sinistral 3, 29 dextral 3, 18 9), R. Berih; CAS-SU 67162
(105 spec.; 5 cleared and stained), R. Selatar. MALAYSIA:
FMNH 51653 (1 dextral 3), East Coast Residency, CAS-SU
61762 (10 spec.: 6 sinistral 3, 2 dextral 3, 2 9), Sarawak. BRU-
NEI: CAS-SU 61763 (1). INDONESIA: AMS 1.19355-041 (26
spec.: 7 sinistral 3, 7 dextral 3, 9 9, 3 juv.), East Kalimantan.
THAILAND: MCZ 76857 (375 spec.; 6 cleared and stained),
near Tha Chalap on road to Chantaburi, Chantaburi Prov.;
MCZ 47200 (3), Satul Prov., CAS 63194 (2), Rayong Prov.;
CAS 63195 (17), Chumporn Prov.
DISTRIBUTION. — Coastal localities of Thai-
land, peninsular Malaysia, Singapore, and Bor-
neo.
Neostethus bicornis Regan, 1916
(Figs. 1, 10d)
Neostethus bicornis Regan, 1916:14-16, fig. 11 (original de-
scription, Kuala Langat, peninsular Malaysia). Myers 1928:
9 (characters). Herre and Myers 1937:70 (ecology, distri-
bution).
Ceratostethus bicornis (Regan). Myers 1937:141-142 (rela-
tionships, referral to new genus). Herre 1939:142 (distri-
bution); 1942:145 (characters, relationships, distribution);
1953:240 (synonymy, distribution). Roberts 1971a:6-10
(characters). Roberts 1971b:397^409, figs. 2-12 (osteology,
functional anatomy). Rosen and Parenti 1981:20, fig. 19A
(dorsal gill arch anatomy).
270
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1 1
DIFFERENTIAL DIAGNOSIS.— Mature males have
two elongate ctenactinia rather than a long first
ctenactinium and short second ctenactinium as
in other Neostethus. Immature males have a
brown blotch on proctal pelvic fin rays that fades
with growth and maturity. Females have a fleshy,
hoodlike flap that covers urogenital openings.
MATERIAL. -Lectotype (here designated): BMNH 1937.12.9:
4, immature sinistral 3, 20.9 mm SL, Kuala Langat, peninsular
Malaysia, G. Duncker, coll. Paralectotypes: BMNH 1 937. 12.9:
5-6 (2 9, 16.7-19.1 mm SL), collected with lectotype.
OTHER MATERIAL. -THAILAND: MCZ 47304 (142 spec.;
2 9 cleared and stained); MCZ 76856 (71 spec.; 1 sinistral $
cleared and stained); MCZ 47305 (93 spec.; 1 dextral 3 cleared
and stained); MCZ 52364 (100+); MCZ 47306 (85 spec.; 1
dextral 3 and 1 9 cleared and stained); MCZ 6008 1 (70), near
Tha Chalap on road to Chantaburi, Chantaburi Prov.; CAS
63 1 96 (2), Chumporn Prov. SINGAPORE: BMNH 1970.7.22:
69-70 (2 spec.: 1 dextral 3, 1 9), R. Poyen; BMNH 1970.7.22:
63-68 (6 spec.: 3 dextral 3, 3 9), R. Tengah; CAS-SU 67161
(10 spec.: 4 sinistral 3, 5 dextral 3, 1 juv.), R. Seletar; CAS-SU
35783 (51 spec.; 10 cleared and stained); MNHN 38-14 (4),
R. Kranji; CAS-SU 31 133 (1 1 spec.: 9 sinistral 3, 1 dextral 3,
1 9); USNM 102142 (8 spec.: 1 sinistral 3, 4 dextral 3, 3 9).
BRUNEI: CAS-SU 61764 (1). MALAYSIA: USNM 189012
(5), mangrove swamps on island opposite Sandakan; FMNH
5 1 726-5 1 729 ( 1 73), East Coast Residency, Kinabatangan Dist;
and private collection of Maurice Kottelat (5), Sungei Seburut
near Matang, Sarawak. PHILIPPINES: USNM 98839 (3) and
USNM 150632 (2), Nakoda Bay; USNM 98840 (1), Malam-
paya R.; USNM 98838 (3), Palawan. No locality: MCZ 60049
(50+).
DISTRIBUTION.— One of the most widespread
phallostethids; coastal localities of Thailand,
peninsular Malaysia, Singapore, Borneo, and Pa-
lawan, Philippines.
Neostethus amaricola
(Villadolid and Manacop, 1935)
(Figs. 1, 9b)
Gulaphallus amaricola Villadolid and Manacop, 1935:194-
196, pi. 1 (original description, Pasay, Rizal Prov., Luzon,
Philippines).
Neostethus amaricola (Villadolid and Manacop). Myers 1937:
139 (characters, referral of G. amaricola to Neostethus). Herre
1939:143 (distribution). Hubbs 1944:73-74 (characters).
Herre 1953:241 (distribution).
Neostethus (Neostethus) amaricola. Herre 1942:149 (charac-
ters, relationships, distribution).
DIFFERENTIAL DIAGNOSIS. — Distal tips of first
pleural ribs of females of Neostethus amaricola
project beyond ventral body profile. Males may
be recognized by a sharply pointed second cte-
nactinium directly anteriorly, rather than dor-
sally as in N. villadolidi.
MATERIAL. -PHILIPPINES: USNM 98835 (6 spec.: 1 si-
nistral 3, 5 9; 13-20 mm SL), Ragay R., Ragay Gulf; FMNH
40771-^40785 (15 spec.: 2 sinistral 3, 5 dextral 3, 6 9, 2 juv.;
13-21 mm SL), Navotas; CAS-SU 32354 (3 spec.: 2 sinistral
3, 1 9; 20-23 mm SL), Nasugbu; CAS-SU 32926 (6 spec.: 1
sinistral 3, 2 dextral 3, 3 9; 23-24 mm SL), Manila; USNM
98834 (2 spec.: 1 sinistral 3, 1 dextral 3), Port San Vicente;
USNM 98833 (3 sinistral 3), Manila Bay; CAS-SU 38898 (200+
spec.; 1 3 cleared and stained), Navotas; CAS 50724 (266), Nipa
creek S. of Masinloc, Luzon. CAS-SU 38899 (190 spec.; 5
cleared and stained), Mindoro.
In the original description, Villadolid and
Manacop (1935:195) stated that there were to be
at least one male and one female, as well as sev-
eral additional cotypes collected in Pasay, Rizal
Prov., Luzon, in the zoological museum of the
College of Agriculture, University of the Phil-
ippines. These specimens are presumed lost.
DISTRIBUTION AND REMARKS.— A Neostethus
species widespread in coastal, brackish-water
habitats throughout the Philippines, including
Luzon and Mindoro. Herre (1942, 1953) re-
ported N. amaricola from Leyte; however, I have
reidentified two USNM lots from that locality as
the closely related N. villadolidi.
Neostethus borneensis Herre, 1939
(Fig. i)
Neostethus borneensis Herre, 1939:143 (original description,
report from around Sandakan Bay, North Borneo); 1 940b:
14-17, pis. 10, 11 (description, Kabili R., Sandakan Bay,
Segaliud R., North Borneo).
Neostethus (Sandakanus) borneensis. Herre 1942:151-152
(characters, relationships, distribution); 1953:242 (listed as
type species of subgenus Sandakanus).
Neostethus (Sandakanus) coronensis Herre, 1942:152-153
(original description, Coron, Busuanga, Philippines); 1953:
242 (distribution).
Bohlke (1953) followed Herre (1942) in treat-
ing Herre's (1940b:14-17) detailed discussion,
which included the heading "Neostethus bor-
neensis, new species," as the original description.
However, Herre's 1940b publication was pre-
dated by a compilation and brief statement of
identifying characters of phallostethid species
(Herre 1 939), in which the name, characters, and
statement of distribution are recorded. This con-
stitutes the original description of TV. borneensis.
DIFFERENTIAL DIAGNOSIS.— Neostethus bor-
neensis is a robust, relatively short-bodied species
with vertebrae (31-32, as opposed to 34-37), and
scales in a lateral series (25-27, as opposed to
29-34), fewer than in congeners.
PARENTI: PHALLOSTETHID FISHES
271
MATERIAL.— Holotype: CAS-SU 33018, sinistral<3, 16.5 mm
SL, Kabili R., Sandakan Dist, North Borneo, 30 Jan. 1937,
A. W. C. T. Herre, coll. Paratypes: CAS-SU 33022 (116 3; 2
cleared and stained), CAS-SU 33019 (45 2), FMNH 51537-
51538 (12 spec.: 6 S and 6 2). BMNH 1938.12.1:155-164 (19
spec.: 8 3, 3 2, 8 juv.), collected with holotype. CAS-SU 69021
(7), CAS-SU 69022 (6 cleared and stained), CAS-SU 33020
(9) Segaliud R., North (Malaysian) Borneo, A. W. C. T. Herre,
coll. Holotype ofNeostethus coronensis: CAS-SU 36542, dex-
tral 6, 22.5 mm SL, Coron, Busuanga, Philippines, 28 June
1 940, A. W. C. T. Herre, coll. Paratypes ofN. coronensis: CAS-
SU 36543 (109 spec.: 23 sinistral 3, 25 dextral 3, 52 2, 9 juv.;
3 cleared and stained) collected with holotype.
OTHER MATERIAL. -MALAYSIA: FMNH 51654 (16), East
Coast Residency; USNM 189015 (23), mangrove swamps on
island opposite Sandakan.
DISTRIBUTION.— Borneo and Coron, Busuan-
ga, Philippines.
Neostethus villadolidi Herre, 1 942
(Figs. 1, 9a, lOb)
Neostethus (Neostethus) villadolidi Herre, 1942:150 (original
description, Misamis Oriental, Mindanao, and Negros Ori-
ental, Philippines).
Neostethus villadolidi. Herre 1953:241-242 (distribution).
DIFFERENTIAL DIAGNOSIS.— Neostethus villa-
dolidi is a slender-bodied species closely allied
with TV. amaricola, but differing from it and all
other phallostethids by having a sharply pointed,
dorsally directed, second ctenactinium, and a
fleshy urogenital protuberance in females.
MATERIAL.— Holotype: CAS-SU 36537 (misprinted 23637
in original description), sinistral S, 22.5 mm SL, mangrove
swamp next to Fishery Experiment Station, Cagayan de Misa-
mis, Misamis Oriental Prov., Mindanao, Philippines, 20 Aug.
1940, A. W. C. T. Herre, coll. Paratypes: CAS-SU 36538 (59
spec.: 14 sinistral 3, 13 dextral 3, 26 2, 6 juv.), collected with
holotype.
OTHER MATERIAL. -PHILIPPINES: CAS-SU 52344 (36
spec.; 4 cleared and stained), CAS-SU 36566 (19 spec.: 2 si-
nistral 3, dextral 3, 1 1 2, 5 juv.), and CAS-SU 32355 (111 spec.:
5 sinistral 3, 3 1 dextral 3, 58 2, 1 7 juv.; 7 cleared and stained),
CAS 60258 (20 spec.: 10 3 and 10 2), MNHN 38-15 (4), ROM
54986 ( 1 58), ROM 54987 ( 1 3), ROM 54988 (80), ROM 50489
(1), ROM 51854 (426), Negros Oriental; AMS 1.21938-016
(39), Mactan Island, Cebu; CAS 54645 (12 spec.: 6 dextral 3,
5 2, 1 juv.), northern end of Cuyo, Palawan Prov.; USNM
98836 (1 2, 21 mm SL), USNM 98837 (2 spec.: 1 2 and 1
dextral 3), Leyte.
Herre (1 942: 1 50) mentioned two lots from Ne-
gros Oriental without catalogue numbers, ques-
tionably to be included in the description of TV.
villadolidi. I list here several lots from Negros
Oriental; however, like Bohlke (1953), I recog-
nize none as type material.
DISTRIBUTION. — Mindanao, Cuyo, Mactan,
Negros Oriental and Leyte, Philippines.
Neostethus zamboangae Herre, 1 942
(Fig. i)
Neostethus (Sandakanus) zamboangae Herre, 1942:153-154
(original description, Zamboanga, Mindanao, Philippines);
1953:242 (distribution).
DIFFERENTIAL DIAGNOSIS.— Males of Neoste-
thus zamboangae are readily distinguished from
congeners by two prominent ctenactinia: an elon-
gate first ctenactinium, and a smaller, hooked
second ctenactinium with from two to seven bony
serrations. Phallostethus dunckeri males also have
a serrated second ctenactinium; however, nu-
merous derived characters that distinguish phal-
lostethins from neostethins support a conclusion
that these similar second ctenactinia are inde-
pendently derived.
MATERiAL.-Holotype: CAS-SU 36544, dextral 3, 22 mm
SL, Zamboanga, Mindanao, Philippines, 6 Sep. 1940, A. W.
C. T. Herre, coll. Paratypes: CAS-SU 36545 (242 spec.: 30
sinistral 3, 45 dextral 3, 1172, 50 juv.; 4 cleared and stained)
collected with holotype.
OTHER MATERIAL. -PHILIPPINES: CAS 50725 (141 spec.;
4 cleared and stained); UMMZ 211666 (37), Sorsogon Bay,
Buhatan R., 5 km E of Sorsogon Town, Luzon.
DISTRIBUTION. — Coastal localities of Zam-
boanga, Mindanao and Sorsogon, Luzon, Phil-
ippines.
Neostethus palawanensis (Myers, 1935),
new combination
(Fig. 1)
P lectrostethus palawanensis Myers, 1935:5-6 (original descrip-
tion, mouth of the Caiholo R., Ulugan Bay, west coast of
Palawan, Philippines). Myers 1937:141 (characters). Herre
1 939: 1 44 (distribution); 1 942: 1 55 (characters, relationships,
distribution). Hubbs 1944:71, 73 (characters). Herre 1953:
242 (distribution).
Plectrostethus palawensis (misspelling of P. palawanensis Myers).
Ivantsoffet al. 1987:653-654, fig. 41 (anatomy, comparisons
with atherinoids).
DIFFERENTIAL DIAGNOSIS.— Neostethus pala-
wanensis is a slender-bodied species readily dis-
tinguished by a relatively straight first ctenactin-
ium with a broad, fleshy ventral ridge.
MATERIAL.— Holotype: USNM 93421, sinistral 3, 22.5 mm
SL, mouth of Caiholo R., Ulugan Bay, west coast of Palawan,
Palawan Prov., Philippines, 29 Dec. 1908, U.S.S. "Albatross."
Paratypes: USNM 93422 (2, 19 mm SL, allotype = paratype);
USNM 93423 (8 spec.: 5 dextral 3, 3 2), collected with holotype.
272
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1 1
OTHER MATERIAL. - PHILIPPINES: USNM 93424 (eggs only)
collected with holotype and paratypes (no type status). CAS
63200 (17 spec.: 5 sinistral 6, 4 dextral 3, 8 9; 4 cleared and
stained), Cuyo, Palawan Prov.
DISTRIBUTION.— Coastal localities of Palawan
and Cuyo, Palawan Prov., Philippines.
Neostethus thessa (Aurich, 1937),
new combination
(Figs. 1, 10a)
Solenophallus thessa Aurich, 1937:264-272, table 1, figs. 1A,
3-5, 71 (original description, Lake Mainit, Mindanao, Phil-
ippines). Herre 1939:142 (distribution); 1953:242 (valid de-
scription of Solenophallus).
DIFFERENTIAL DIAGNOSIS.— Neostethus thessa
is readily distinguished from congeners by: no
dorsal fin rays (remnant of first dorsal pteryg-
iophore present in some specimens); oral den-
tition reduced to several small, unicuspid teeth
on premaxilla and dentary; anal fin with 1 7-22
rays, as opposed to 13-18; and second dorsal fin
with 8-10 rays, as opposed to 5-7.
MATERIAL. -PHILIPPINES: CAS-SU 36546 (46 spec.: 9 si-
nistral 3, 9 dextral 6, 26 9, 2 juv.; 9 cleared and stained), Lake
Mainit, Mindanao (type locality, no type status).
This is the only known material of TV. thessa.
Aurich (1937) indicated that he based his original
description on at least one male (26 mm SL) and
one female (29 mm SL), without giving place of
deposition of any type or other material. No
specimens of Neostethus thessa are present in
collections of the ZMB (H. J. Paepke, pers.
comm.) and ZMH (H. Wilkens, pers. comm.).
However, I do not designate a neotype from the
CAS-SU material, even though it is from the type
locality, because Aurich's material may even-
tually be located upon a thorough search of pos-
sible collections, and, to my knowledge, there has
never been confusion or concern among ichthy-
ologists over identification of the species asso-
ciated with the name Solenophallus thessa Au-
rich.
DISTRIBUTION. — Lake Mainit, Mindanao,
Philippines (Fig. 1).
Neostethus ctenophorus (Aurich, 1937),
new combination
(Fig. 1)
Solenophallus ctenophorus Aurich, 1937:272-275, table l,figs.
6, 7II-V (original description, Laguna de Bay, Luzon, Phil-
ippines).
Ctenophallus ctenophorus (Aurich). Herre 1939:144 (referral
to new genus); 1942:154-155 (characters, distribution, re-
lationships); 1953:240 (synonymy).
REMARKS.— Herre (1942:154) diagnosed Cte-
nophallus as distinct ". . . from other phallos-
tethids by the presence of a dense fringe along
both margins of the open groove on the pria-
pium." This statement was extracted from Au-
rich's (1 937) original description based on at least
one male (24 mm SL) and one female (20 mm
SL). Place of deposition of type or other material
was not given by Aurich. No specimens of TV.
ctenophorus are present in collections of the ZMB
(H. J. Paepke, pers. comm.) or the ZMH (H.
Wilkens, pers. comm.).
The "dense fringe" is formed by the numerous
bony projections of the papillary bone, diagnos-
tic of Neostethus. The same character was also
used by Myers (1 935) to diagnose his new genus,
Plectrostethus, literally a "quill on the chest,"
placed here in synonymy of Neostethus.
I have examined no specimens of TV. cteno-
phorus; however, without this species, the genus
Neostethus is paraphyletic. Therefore, I synon-
ymize Ctenophallus with Neostethus. Meristic
data summarized in Table 2 are from Aurich
(1937).
DISTRIBUTION.— Laguna de Bay, Luzon, Phil-
ippines.
Neostethus robertsi new species
(Figs. 1, 12, 13)
HOLOTYPE.— CAS 50723, dextral 3, 22.3 mm SL, Calasiao
R., about 12 km N of San Carlos City, Pangasinan Prov.,
Luzon, Philippines, 16 Mar. 1976, T. R. Roberts, coll.
PARATYPES. -CAS 64254 (40 spec.: 12 sinistral 3, 9 dextral
3, 10 9, 9 juv. or sex undet., 9.0-23.0 mm SL; 1 sinistral 3, 1
dextral 3 and 1 9 cleared and stained), collected with holotype.
DIFFERENTIAL DIAGNOSIS.— Neostethus rob-
ertsi, like N. palawanensis, has a second cten-
actinium which is bifurcated distally, but distin-
guished from that species by having a larger,
better-developed, claw-shaped second ctenactin-
ium, and a slightly bowed second ctenactinium
that lacks a fleshy, ventral ridge.
DESCRIPTION.— A small, laterally compressed
species, largest size recorded 23.0 mm SL. No
vestigial pelvic fin rays or bones in adult females;
males with pelvic and parts of pectoral fin mod-
ified into priapium that is either dextral (holo-
type and eight <5 paratypes) or sinistral ( 1 3 <5 para-
types). Prominent externalized subcephalic bone
a first ctenactinium, arising on either left or right
PARENTI: PHALLOSTETHID FISHES
273
FIGURE 12. Neostethus robertsi new species. Holotype (CAS 50723), dextral male, 22.3 mm SL.
side of body and articulating with posterior ex-
tent of left or right (proctal) axial bone, curving
gently along the left or right side of the head and
the tip lying just ventral to lower jaw. First cte-
nactinium smooth, without fleshy, ventral ridge.
Large, well-developed, claw-shaped second cte-
nactinium. Fibrous pulvinulus lateral to, and
covering articulation point of, inner pulvinular
and proctal axial bones. Pleural ribs of fourth
vertebra in males expanded anteroposteriorly,
their distal tips meeting on right side of proctal
axial bone in sinistral males and left side in dex-
tral males. First pleural rib on third vertebra in
females.
Ventral dermal keel extending from posterior
extent of priapium in males or urogenital open-
ing in females, to anal fin origin. Scales on body
of moderate size, deciduous. Color in alcohol like
most other phallostethids: ground coloration very
pale yellow or light brownish (straw-colored).
Dark brown melanophores scattered on dorsal
surface of head, along midlateral intermuscular
septum, on operculum and priapium, and along
basal portion of anal fin, dorsal and ventral mid-
line. Dorsal scales with faint dark margins. Eye-
lens large.
Skull and jaws like those illustrated for Neo-
stethus bicornis by Roberts (1971b), with little
modification. Jaw teeth small, conical, in a single
uneven row or grouped together in indistinct rows.
Caudal skeleton with two epurals, autogenous
parhypural, and a dorsal and a ventral hypural
plate. Caudal fin rays broken in nearly all spec-
imens examined. Pectoral fin narrow and elon-
gate, with 10-1 1 rays. Two dorsal fins, the first
with two spines or thickened rays supported by
a single pterygiophore, the second with 6 rays.
Anal fin rays 17-18, vertebrae 36, scales in a
lateral series 30, branchiostegal rays 5 (see Table
2).
DISTRIBUTION. — Pangasinan Prov., Luzon,
Philippines (Fig. 1).
ETYMOLOGY.— The specific name, robertsi, in
honor of Tyson R. Roberts, collector of the type
series and student of phallostethid fishes.
Subfamily Gulaphallinae Aurich
Tribe Gulaphallini new status
DIAGNOSIS.— Adult males with a perforated
gular flap of skin through which anterior end of
first ctenactinium projects and remains while at
rest; and an aproctal axial bone that projects be-
yond ventral body profile with a small postero-
dorsal crest.
Genus Gulaphallus Herre, 1925
Gulaphallus Herre, 1925:508 (type species: Gulaphallus exi-
mius Herre, 1925, by subsequent designation of Myers 1928:
9).
FIGURE 13. Diagrammatic representation of head and an-
terior portion of body, Neostethus robertsi, dextral male, CAS
50723. Anterior to the right. Arrow points to second ctenac-
tinium.
274
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1 1
Mirophallus Herre, 1926:539 (type species: Mirophallus bi-
kolanus Herre, 1926, by monotypy).
Acanthostethus Herre, 1929:141-142 (type species: Gulaphal-
lusfalcifer Manacop, 1936, by original designation; preoc-
cupied by Acanthostethus Smith, 1869 in Hymenoptera).
Manacopus Herre, 1940a:141 (replacement name for Acan-
thostethus Herre, 1939, and therefore taking same type species:
Gulaphallus fakifer Manacop, 1936).
DIAGNOSIS.— As for the tribe Gulaphallini,
above.
COMPOSITION AND DISTRIBUTION. — Five
species, four from freshwater hill-streams, Lu-
zon, G. eximius Herre, 1925, G. mirabilis Herre,
1925, G. fakifer Manacop, 1936, and G. biko-
lanus (Herre, 1 926), and one from coastal brack-
ish to marine localities, Panay, G. panayensis
(Herre, 1942), Philippines.
Gulaphallus eximius Herre, 1925
(Fig. i)
Gulaphallus eximius Herre, 1925:509-510, pis. 1, 2; figs. 1, 2
(original description, Nueva Vizcaya Province, Luzon, Phil-
ippines). Myers 1928:9 (characters). Herre 1939:142 (char-
acters, distribution). Hubbs 1944:73 (characters). Herre 1953:
241 (distribution).
DIFFERENTIAL DIAGNOSIS. — Gulaphallus exi-
mius, the largest phallostethid species, reaches
37 mm SL, and has more than 50 (range 52-58)
scales in lateral series, as opposed to 25-38, in
all other phallostethids. A brown blotch sur-
rounds anus and urogenital openings in imma-
ture females. Rudimentary pelvic girdle repre-
sented by two small scale-shaped discs of bone
in adult females. Males dextral as in G. bikola-
nus, but unlike that species, a first dorsal fin is
present in both sexes.
MATERIAL.— Lecto type (designated by BQhlke 1953): CAS-
SU 24474, 3 33.5 mm SL, creek, Santa Fe, Nueva Vizcaya
Prov., Luzon, Philippines, 1 8 May 1 924, R. C. McGregor, coll.
Paralectotypes: CAS-SU 18149 (3 spec.: 1 3, 2 9), collected
with lectotye.
OTHER MATERIAL. -PHILIPPINES: BMNH 1933.3.11.168
(1 3, 32.2 mm SL), FMNH 40593-40596 (4 spec.: 1 3, 2 2, 1
immature 3), CAS-SU 38900 (72 spec.; 1 1 cleared and stained),
CAS-SU 36832 (4 spec.: 1 3, 2 2, 1 juv.), Luzon.
DISTRIBUTION. — Mountain streams, Luzon,
Philippines.
Gulaphallus mirabilis Herre, 1925
(Figs. 1,4,5)
Gulaphallus mirabilis Herre, 1 925:5 1 1 , pi. 2, figs. 3-5 (original
description, Ibo creek, tributary of Angat R., Bulacan Prov-
ince, Luzon, Philippines). Myers 1928:10 (characters). Vil-
ladolid and Manacop 1 9 3 5 : 1 96-2 1 7 , pis. 1 -4, figs. 1 -3 (anat-
omy, development). Bailey 1936:457-470, pis. 1, 2
(osteology). Aurich 1937:275-279, figs. 8, 9 (anatomy). Myers
1937:142 (characters). TeWinkel 1939:63-67 (anatomy, de-
velopment). Herre 1939: 142 (characters, distribution); 1942:
144-145 (characters, distribution, relationships). Hubbs 1944:
73 (characters). Herre 1953:241 (distribution). Roberts 1971b:
41 1 (characters). Grier et al. 1980:332, table 1 (testis struc-
ture). Rosen and Parenti 1981:20, fig. 19B (dorsal gill arch
anatomy).
DIFFERENTIAL DIAGNOSIS. — Gulaphallus mi-
rabilis is readily distinguished from congeners by
scales in lateral series ranging from 36 to 38 as
opposed to from 28 to 32, or more than 50.
MATERIAL. -PHILIPPINES: Luzon: ANSP 91045 (4 spec.:
dextral 3, 9, 2 juv., 12-25 mm SL), BMNH 1933.3.11.169-
186 (18 spec.: 3 dextral 3, 2 sinistral 3, 1 1 9, 2 immature; 1
dextral $ and 1 9 cleared and stained), FMNH 76769 (6 spec.:
1 dextral S, 5 juv.), CAS 727 (1), CAS 728 (1), CAS 731-735
(5), CAS 740-743 (4), CAS 747-751 (5), CAS-SU 30720 (165),
CAS-SU 38903 (64 spec.; 5 cleared and stained), CAS-SU
35959 (25), CAS 54938 (2 spec.: 1 sinistral 6, 1 9), CAS-SU
35958 (121 spec.: 1 1 sinistral 3, 15 dextral 3, 32 2, 63 juv.),
FMNH 46946 (9 spec.: dextral 3, 4 2, 4 juv.), USNM 94293
(26 spec.: 8 dextral 3, 4 9, 14 juv.), USNM 104412 (2), MCZ
33904 (10), MNHN 27-192, 193 (25, labelled syntypes, but
not from type locality), MNHN 32-204 (4), ZMA 100.651 (4
spec.: 1 dextral 3, 1 9, 2 juv., 17-29 mm SL), Molawin Creek,
CAS-SU 38209 (7), CAS 50721 (279), UMMZ 211664 (54),
Olo Creek, CAS-SU 38901 (7), CAS-SU 38902 (1 cleared and
stained), Sison, and CAS 50719 (230), Barabac Creek, Pan-
gasinan Prov.; CAS-SU 38904 (43 spec.: 4 sinistral 3, 6 dextral
3, 24 9, 9 juv., 13-31 mm SL; 2 dextral 3 and 2 9 cleared and
stained), FMNH 40623^10627 (4 spec.: 2 sinistral 3, 2 9), Santa
Maria R.; MCZ 33904 (10), Los Banos, Laguna Prov.; USNM
197728 (31 juv.), Central Luzon; CAS 50720 (23 spec.: 3 si-
nistral 3, 2 dextral 3, 8 9, 10 juv.), Porac R., Pampanga Prov.
DISTRIBUTION AND REMARKS. —One of the best-
known and frequently collected phallostethid
species because of its abundance in Molawin
Creek, which flowed through the campus of the
College of Agriculture, Luzon, Philippines (Herre
1 942). Known also from other localities through-
out Luzon.
Gulaphallus falcifer Manacop, 1936
(Fig. i)
Gulaphallus falcifer Manacop, 1936:375-379, pis. 1,2 (original
description, Barrio Laput, Mexico, Pampanga Prov., and
Nampicuan, Nueva Ecija, Luzon, Philippines).
Acanthostethus falcifer (Manacop). Herre 1939:141-143 (re-
ferral to new genus).
Manacopus falcifer (Manacop). Herre 1940a:141 (referral to
new genus); 1942:146-147 (characters, distribution, rela-
tionships); 1953:241 (synonymy). Parenti 1986b:306-309,
figs. 1-3 (anatomy, development).
DIFFERENTIAL DIAGNOSIS.— Gulaphallus fal-
cifer is the only phallostethid species in which
PARENTI: PHALLOSTETHID FISHES
275
females as well as males exhibit bilateral asym-
metry: adult females with an elongate pelvic fin
ray, on either left or right pelvic girdle, that pro-
jects beyond body profile to left or right of uro-
genital openings.
MATERIAL. -PHILIPPINES: Luzon: CAS-SU 32356 (42
spec.: 9 sinistral 6, 14 dextral 6, 10 sinistral 2, 3 dextral 2, 6
juv.; 7 cleared and stained), Barrio Laput, Pampanga Prov.
(type locality). No locality: CAS 63 1 98 (2 cleared and stained).
REMARKS.— Manacop (1936) lists the follow-
ing type material: Holotype: Philippine Fish and
Game Administration 31778, adult <3, 27.5 mm
SL, gourami pond at Barrio Laput, Mexico, Pam-
panga Prov., Luzon, Philippines, 27 Aug. 1935,
C. Sandiko, coll. Allotype (=paratype): Philip-
pine Fish and Game Administration 3 1 779 (adult
9 26.5 mm SL). Paratypes: Philippine Fish and
Game Administration 31780 (17 spec.: 8 2 and
9 <5). This material is presumed unavailable for
study. Bohlke (1953) did not designate a neotype
from the CAS-SU material.
DISTRIBUTION. — Barrio Laput, Pampanga
Prov., Luzon, Philippines.
Gulaphallus bikolanus (Herre, 1926),
new combination
(Fig. 1)
Mirophallus bikolanus Herre, 1926:540-541, pi. 3 (original
description, Lake Bato, Camarines Sur Province, and Lake
Lanigay, Albay Province, Luzon, Philippines). Myers 1928:
10-1 1 (characters). Aurich 1937:279-282, fig. 10 (anatomy).
Myers 1937: 142 (characters). Herre 1939: 142 (distribution);
1942:142 (characters, distribution, relationships). Hubbs
1944:73-74 (characters). Herre 1953:241 (distribution). Pa-
renti 1986a:234, table 2 (bilateral asymmetry).
DIFFERENTIAL DIAGNOSIS.— Males of Gula-
phallus bikolanus are dextral as in G. eximius,
but unlike that species, both males and females
of G. bikolanus lack a first dorsal fin and have
32 scales in a lateral series, as opposed to more
than 50.
MATERIAL AND REMARKS. — Lectotype (designated by Bohlke
1953). CAS-SU 24475, immature 3, 19.5 mm SL, Lake Bato,
Camarines Sur Prov., Luzon, Philippines, 30 Jan. 1926, G. A.
Lopez, coll. Paralectotypes: CAS-SU 18148 (2 9, 22-23 mm
SL), MNHN 27-194 (2, 21.1-23.8 mm SL), collected with
lectotype. Bohlke (1953) was apparently unaware of the MNHN
specimens, collected with the syntypes, which also have para-
lectotype status.
OTHER MATERIAL. -PHILIPPINES: AMNH 50592 (24 spec.:
13 dextral 3, 9 9, 2 juv.), CAS 50722 (194 spec.: 131 dextral
3, 55 9, 8 juv.), Guinobatan R.; CAS 53165 (70 spec.: 24 dextral
3, 37 8, 9 juv.), UMMZ 211665 (12 spec.: 5 dextral 3, 7 9),
Cabangan R., Albay Prov., Bikol Region, Luzon.
DISTRIBUTION. — Camarines Sur and Albay
provinces, Luzon, Philippines.
Gulaphallus panayensis (Herre, 1 942),
new combination
(Fig. 1)
Neostethus (Sandakanus) panayensis Herre, 1942: 1 53 (original
description, Capiz and Estancia, Panay, Philippines); 1953:
242 (distribution).
DIFFERENTIAL DIAGNOSIS. — Gulaphallus pa-
nayensis is a small, slender species known only
from limited type material, with vertebrae 31-
33, as opposed to 35-37, and scales in lateral
series 28-30, as opposed to 32-58 in congeners.
The largest male (holotype) does not have a com-
pletely developed priapium; however, I con-
clude, primarily from examination of cleared and
stained specimens, that panayensis is a Gula-
phallus.
MATERIAL.— Holotype: CAS-SU 36539, dextral 3, 19.0 mm
SL, Capiz, Panay, Philippines, 3 Aug. 1940, A. W. C. T. Herre,
coll. Paratypes: CAS-SU 36540 (22 spec.: 1 sinistral 3, 3 dextral
3, 18 2; 2 cleared and stained) collected with holotype, and
CAS-SU 36541 (22 spec.: 5 dextral 3, 159, 2 juv.), Estancia,
Panay, Philippines, 27 July 1940, A. W. C. T. Herre, coll.
DISTRIBUTION.— Capiz and Estancia, Panay,
Philippines (Fig. 1).
CONCLUSIONS
1 . The Phallostethidae comprise a monophy-
letic group of 1 9 species, classified in four genera,
of three monophyletic tribes— Phallostethini,
Neostethini, and Gulaphallini. Phallostethini
(containing Phallostethus and Phenacostethus)
and Neostethini (containing solely Neostethus)
are sister taxa comprising the subfamily Phallo-
stethinae. Subfamily Gulaphallinae is coexten-
sive with tribe Gulaphallini and genus Gula-
phallus.
2. Of the 19 recognized species, one, Neo-
stethus robertsi, from Luzon, Philippines, is de-
scribed as new. Neostethus siamensis Myers, 1 937
is placed in synonymy of Neostethus lankesteri
Regan, 1916; Neostethus coronensis Herre, 1 942
is placed in synonymy of Neostethus borneensis
Herre, 1939. Five new combinations are pro-
posed: Plectrostethus palawanensis Myers, 1935,
Solenophallus thessa Aurich, 1937 and S. cteno-
phorus Aurich, 1937, are placed in Neostethus;
Mirophallus bikolanus Herre, 1 926 and Neoste-
276
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 1 1
thus panayensis Herre, 1 942 are placed in Gu-
laphallus.
3. All known names proposed for parts of the
bilaterally asymmetric priapium are listed and
defined; synonyms and presumed homologs are
identified. Derived characters of the priapium
provide the primary information for species group
recognition. However, bilateral asymmetry in fe-
males of one species (G. falcifer), elaborate uro-
genital coverings in females of three species (N.
lankesteri, N. borneensis, and N. villadolidi), and
ontogenetic changes in pigmentation pattern of
the anus and urogenital openings in females of
one species (Gulaphallus eximius) indicate that
the complex evolution of the male priapium has
been paralleled by numerous modifications in
the reproductive system of females.
ACKNOWLEDGMENTS
This study could not have been completed
without the generous initial support of NSF grant
BSR 83-15258, which is gratefully acknowl-
edged. Additional support from CAS, including
the fine curatorial assistance of D. Catania and
A. Snyder, and NSF, through grant BSR 87-
0035 1 , facilitated completion of the project. Dis-
cussions with T. R. Roberts (CAS), E. Trewavas
(BMNH), and R. Winterbottom (ROM) contrib-
uted to my understanding of phallostethid anat-
omy and distribution. W. Ivantsoff (Macquarie
University) kindly donated atherinoid material
to CAS and discussed atherinomorph system-
atics with me. T. R. Roberts read and com-
mented on an earlier version of the manuscript.
I also benefited from access to the Genera of
Fishes database being compiled by W. N. Esch-
meyer (CAS). The typescript was prepared using
a personal computer purchased in part with a
generous donation from the San Francisco
Aquarium Society.
The following provided loans of specimens,
information, and/or assistance during visits to
their institutions: M. N. Feinberg and G. Nelson
(AMNH), J. Paxton (AMS), B. Chernoff (ANSP),
G. J. Howes and A. Wheeler (BMNH), T. Grande
and D. J. Stewart (FMNH), K. E. Hartel and K.
F. Liem (MCZ), M.-L. Bauchot and M. De-
Soutter (MNHN), M. van Oijen (RMNH), E. J.
Crossman and R. Winterbottom (ROM), R. R.
Miller (UMMZ), S. Jewett and R. P. Vari
(USNM), H. Nijssen (ZMA), H. J. Paepke (ZMB),
H. Wilkens (ZMH), and M. Kottelat (private col-
lection).
Comments of several reviewers greatly im-
proved the presentation of ideas.
LITERATURE CITED
AURICH, H. 1937. Die Phallostethiden (Unterordnung Phal-
lostethoidea Myers). Intl. Rev. Ges. Hydrobiol. Hydrogr.
34:263-286.
BAILEY, R. J. 1936. The osteology and relationships of the
phallostethid fishes. J. Morph. 59:453-483.
BOHLKE, E. B. 1984. Catalog of type specimens in the ich-
thyological collection of the Academy of Natural Sciences
of Philadelphia. Acad. Nat. Sci. Philad., Spec. Publ. 14:1-
246.
BOHLKE, J. 1953. A catalogue of the type specimens of recent
fishes in the Natural History Museum of Stanford Univer-
sity. Stanford Ichthyol. Bull. 5:1-168.
COLLETTE, B. B., G. E. McGowEN, N. V. PARIN, AND S. MITO.
1984. Beloniformes: development and relationships. Pp.
334-354 in Ontogeny and systematics of fishes. H. G. Moser,
W. J. Richards, D. M. Cohen, M. P. Fahay, A. W. Kendall,
Jr., and S. L. Richardson, eds. Am. Soc. Ichthyol. Herpetol.
Spec. Publ. No. 1.
DINGERKUS, G. AND L. D. UHLER. 1977. Enzyme clearing of
alcian blue stained whole small vertebrates for demonstra-
tion of cartilage. Stain Tech. 52:229-232.
DUNCKER, G. 1904. Die Fische der malayischen Halbinsel.
Mitt. Nat. Mus. Hamburg 21:133-207.
FOWLER, H. W. 1937. Zoological results of the third de-
Schauensee Siamese Expedition, Part VIII— fishes obtained
in 1936. Proc. Philad. Acad. Nat. Sci. 89:125-264.
GRIER, H. J. AND B. B. COLLETTE. 1987. Unique spermato-
zeugmata in testes of halfbeaks of the genus Zenarchopterus
(Teleostei: Hemiramphidae). Copeia 1987:300-311.
GRIER, H. J., J. R. LINTON, J. F. LEATHERLAND, AND V. L. DE
VLAMING. 1 980. Structural evidence for two different tes-
ticular types in teleost fishes. Am. J. Anat. 159:331-345.
HENNIG, W. 1966. Phylogenetic systematics. Univ. Illinois
Press, Urbana, Illinois. 263 pp.
HERRE, A. W. C. T. 1925. Two strange new fishes from
Luzon. Philipp. J. Sci. 27:507-513.
. 1 926. Four new Philippine fishes. Philipp. J. Sci. 3 1 :
533-543.
. 1939. The genera of Phallostethidae. Proc. Biol. Soc.
Wash. 52:139-144.
. 1940a. Manacopus, a new name for a genus of Phal-
lostethidae. Copeia 1940:141.
. 1 940b. New species of fishes from the Malay Pen-
insula and Borneo. Bull. Raffles Mus. 16:5-26.
. 1942. New and little known phallostethids, with keys
to the genera and Philippine species. Stanford Ichthyol. Bull.
2:137-156.
. 1953. Check list of Philippine fishes. U.S. Fish Wildl.
Ser. Res. Rep. 20.
HERRE, A. W. C. T. AND G. S. MYERS. 1937. A contribution
to the ichthyology of the Malay Peninsula. Bull. Raffles Mus.
13:5-75.
HUBBS, C. L. 1 944. Fin structure and relationships of the
phallostethid fishes. Copeia 1944:69-79.
HUBBS, C. L. AND L. C. HUBBS. 1945. Bilateral asymmetry
and bilateral variation in fishes. Pap. Mich. Acad. Sci. Arts
Let. 30:229-310.
HUMPHRIES, C. J. AND L. R. PARENTI. 1986. Cladistic bio-
geography. Oxford Monographs on Biogeography No. 2.
Clarendon Press, Oxford, England. 98 pp.
PARENTI: PHALLOSTETHID FISHES
277
I VANTSOFF, W., B. SAID, AND A. WILLIAMS. 1987. Systematic
position of the family Dentatherinidae in relationship to
Phallostethidae and Atherinidae. Copeia 1987:649-658.
LADIGES, W., G. VON WAHLERT, AND E. MOHR. 1958. Die
Typen und Typoide der Fischsammlung des Hamburgischen
Zoologischen Staatsinstituts und Zoologischen Museums.
Mitt. Hamburg Zool. Inst. 56:155-167.
LEVTTON, A. E., R. H. GIBBS, JR., E. HEAL, AND C. E. DAWSON.
1985. Standards in herpetology and ichthyology: part I.
Standard symbolic codes for institutional resource collec-
tions in herpetology and ichthyology. Copeia 1985:802-832.
MADDISON, W. P., M. J. DONOGHUE, AND D. R. MADDISON.
1 984. Outgroup analysis and parsimony. Syst. Zool. 33:83-
103.
MANACOP, P. R. 1 936. A new phallostethid fish with notes
on its early development. Philipp. J. Sci. 59:375-381.
MICKEVICH, M. F. 1978. Taxonomic congruence. Syst. Zool.
27:143-158.
MYERS, G. S. 1928. The systematic position of the phallo-
stethid fishes, with diagnosis of a new genus from Siam. Am.
Mus.Novit. 295:1-12.
. 1935. A new phallostethid fish from Palawan. Proc.
Biol. Soc. Wash. 48:5-6.
. 1937. Notes on phallostethid fishes. Proc. U.S. Natl.
Mus. 84:137-143.
NELSON, G. 1978. Ontogeny, phylogeny, paleontology, and
the biogenetic law. Syst. Zool. 27:324-345.
PARENTI, L. R. 1981. A phylogenetic and biogeographic anal-
ysis of cyprinodontiform fishes (Teleostei, Atherinomor-
pha). Bull. Am. Mus. Nat. Hist. 168:335-557.
. 1 984. On the relationships of phallostethid fishes
(Atherinomorpha), with notes on the anatomy of Phallo-
stethusdunckeriReg»n, 1913. Am. Mus. Novit. 2779:1-12.
-. 1986a. Bilateral asymmetry in phallostethid fishes
(Atherinomorpha), with description of a new species from
Sarawak. Proc. Calif. Acad. Sci. 44:225-236.
. 1986b. Homology of pelvic fin structures in female
phallostethid fishes (Atherinomorpha, Phallostethidae). Co-
peia 1986:305-310.
. 1986c. The phylogenetic significance of bone types
in euteleost fishes. Zool. J. Linn. Soc. 87:37-51.
PATTEN, J. M. AND W. IVANTSOFF. 1983. A new genus and
species of atherinid fish, Dentatherina merceri from the west-
em Pacific. Japan. J. Ichthyol. 29:329-339.
PATTERSON, C. 1982. Morphological characters and homol-
ogy. Pp. 21-74 in Problems of phylogenetic reconstruction.
K. A. Joysey and A. E. Friday, eds. Academic Press, New
York.
REGAN, C. T. 1913. Phallostethus dunckeri, a remarkable new
cyprinodont fish from Johore. Ann. Mag. Nat. Hist. 12:548-
555.
. 1916. The morphology of the cyprinodont fishes of
the subfamily Phallostethinae, with descriptions of a new
genus and two new species. Proc. Lond. Zool. Soc. 1916:1-
26.
ROBERTS, T. R. 197 la. The fishes of the Malaysian family
Phallostethidae (Atheriniformes). Breviora 374:1-27.
. 1971b. Osteology of the Malaysian phallostethid fish
Ceratostethus bicornis, with a discussion of the evolution of
remarkable structural novelties in its jaws and external gen-
italia. Bull. Mus. Comp. Zool. 142:393-418.
ROSEN, D. E. 1 964. The relationships and taxonomic position
of the halfbeaks, killifishes, silversides, and their relatives.
Bull. Am. Must. Nat. Hist. 127:217-268.
ROSEN, D. E. AND L. R. PARENTI. 1981. Relationships of
Oryzias, and the groups of atherinomorph fishes. Am. Mus.
Novit. 2719:1-25.
SMITH, H. M. 1927. The fish Neostethus in Siam. Science 65:
353-355.
. 1929. Notes on some Siamese fishes. J. Siam Soc.
Nat. Hist, Suppl. 8:11-14.
. 1945. The fresh-water fishes of Siam, or Thailand.
Bull. U.S. Natl. Mus. 188:1-622.
SPRINGER, V. G. 1983. Tyson belos, new genus and species
of western Pacific fish (Gobiidae, Xenisthminae), with dis-
cussions of gobioid osteology and classification. Smithson.
Contrib. Zool. 390:1-40.
TAKITA, T. AND K. NAKAMURA. 1986. Embryonic develop-
ment and prelarva of the atherinid fish, Hypoatherina blee-
keri. Japan. J. Ichthyol. 33:57-61.
TEWINKEL, L. E. 1939. The internal anatomy of two phal-
lostethid fishes. Biol. Bull. 76:59-69.
VILLADOLID, D. V. AND P. R. MANACOP. (1934) Issued 1935.
The Philippine Phallostethidae, a description of a new species,
and a report on the biology of Gulaphallus mirabilis Herre.
Philipp. J. Sci. 55:193-220.
WEBER, M. AND L. F. DEBEAUFORT. 1922. Phallostethidae.
Pp. 381-383 in The fishes of the Indo-Australian Archipel-
ago IV. Leiden, The Netherlands.
WEITZMAN, S. H. AND S. V. FINK. 1985. Xenurobryconin
phylogeny and putative pheromone pumps in glandulocau-
dine fishes (Teleostei: Characidae). Smithson. Contrib. Zool.
421:1-121.
WHITE, B. N., R. J. LAVENBERG, AND G. E. McGowEN. 1984.
Atheriniformes: development and relationships. Pp. 355-
362 in Ontogeny and systematics of fishes. H. G. Moser, W.
J. Richards, D. M. Cohen, M. P. Fahay, A. W. Kendall, Jr.,
and S. L. Richardson, eds. Am. Soc. Ichthyol. Herp. Spec.
Publ. No. 1.
WHITMORE, T. C., ED. 1987. Biogeographical evolution of
the Malay Archipelago. Oxford Monographs on Biogeog-
raphy No. 4. Clarendon Press, Oxford, England. 147 pp.
WILEY, E. O. 1981. Phylogenetics; the theory and practice of
phylogenetic systematics. John Wiley and Sons, New York.
439 pp.
WINTERBOTTOM, R. 1 974. A descriptive synonymy of the
striated muscles of the teleostei. Proc. Acad. Nat. Sci. Philad.
125:225-317.
WOLTERECK, R. 1 942a. Stufen der Ontogenese und der Evo-
lution von Kopulationsorganen bei Neostethiden (Perce-
soces, Teleostei). Intl. Rev. Ges. Hydrobiol. Hydrogr. 42:
253-268.
. 1942b. Neue Organe, durch postembryonale
Umkonstruktion aus Fischflossen entstehend. Intl. Rev. Ges.
Hydrobiol. Hydrogr. 42:317-355.
CALIFORNIA ACADEMY OF SCIENCES
Golden Gate Park
San Francisco, California 94 1 1 8
PROCEEDINGS
OF THE
CALIFORNIA ACADEMY OF SCIENCES
Vol. 46, No. 12, pp. 279-287, 3 figs.
September 11, 1990
NEW AND RECONSIDERED MEXICAN ACANTHACEAE. IV.
By
Thomas F. Daniel
Department of Botany, California Academy of Sciences, Golden Gate Park,
San Francisco, CA 94118
ABSTRACT: Dyschoriste mcvaughii, characterized by its diminutive habit, red corollas, and inconspicuously
spurred thecae, is newly described from western Jalisco. The monotypic genera Ixtlania and Tabascina are
considered to be congeneric with Justicia and the new names, /. ixtlania and J. tabascina are provided. A
new combination, Schaueria parviflora, is proposed for the species resulting from the taxonomic merger of
Streblacanthus parviflorus and Schaueria calycobractea.
Received October 10, 1989. Accepted January 10, 1990.
INTRODUCTION
Ongoing studies of the more than 350 species
of Acanthaceae in Mexico continue to necessitate
the description of new and the reevaluation of
previously described taxa. In this report, a re-
markable red-flowered species of Dyschoriste is
described for the first time. Recent collections
and studies of two monotypic genera, Ixtlania
M. E. Jones and Tabascina Baillon, reveal them
to be congeneric with the large and polymorphic
genus Justicia L. Schaueria calycobractea R. Hil-
senbeck & D. Marshall, a species recently de-
scribed from Veracruz, is shown to be correctly
classified generically but conspecific with the pre-
viously described Streblacanthus parviflorus
Leonard from Guatemala and thus in need of
nomenclatural renovation. Detailed descriptions
are provided for each of these taxa.
Dyschoriste mcvaughii T. F. Daniel, sp. nov.
(Figures 1, 2f)
TYPE.— MEXICO. Jalisco: between Ayutla and Mascota near
summit of pass, 7-8 mi NW of Los Volcanes, 30 April 1951,
R. McVaugh 12187 (Holotype: MICH!; isotype: US!).
Herba perennis usque ad 1 dm alta. Folia sessilia vel sub-
sessilia petiolo usque ad 2 mm longo, elliptica vel obovata, 4-
1 8 mm longa, 3-7 mm lata, 1 . 1-4-plo longiora quam latiora.
Dichasia 1-3-flora, in axillis foliorum distalium sessilia vel
subsessilia. Bracteolae et bracteolae secundae lineares vel lin-
eares-ellipticae vel lineares-oblanceolatae, 5.5-1 1 mm longae,
1-3 mm latae. Calyx 9-12 mm longus lobis tubo 1.4-3-plo
longioribus. Corolla rubra, 34-42 mm longa. Stamina 9-14
mm longa, thecae 1 .8-2.4 mm longae, basi inconspicuae cal-
caratae vel muticae. Stylus 27-38 mm longus. Capsula 7 mm
longa, glabra.
Perennial herb from woody rhizome to 1 dm
tall, with numerous woody roots. Stems
subquadrate to quadrate-sulcate, densely pubes-
cent with straight to flexuose eglandular tri-
chomes 0. 1-0.3(-0.5) mm long. Leaves sessile or
subsessile with petioles to 2 mm long, blades
elliptic to obovate, 4-18 mm long, 3-7 mm wide,
1.1-4 times longer than wide, rounded to acute
at apex, rounded to acute to cuneate at base,
surfaces pubescent (especially along veins) like
stems (although the trichomes tending to be more
antrorse), proximal leaves reduced in size, 1.5-
3 mm long. Inflorescence of 1-3 -flowered di-
chasia borne in axils of distal leaves forming a
terminal spicate thyrse, dichasia sessile or sub-
[279]
280
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 12
FIGURE 1. Dyschoriste mcvaughii. a, habit, x 1.2 (McVaugh 12187); b, leaf, x4 (Daniel & Bartholomew 4828); c, flowering
node with leaf removed showing 3-flowered dichasium, x3 (McVaugh 12187); d, calyx, x5 (McVaugh 12187); e, androecium
in opened corolla, x5 (Daniel & Bartholomew 4828); f, distal portion of stamen, x 12 (McVaugh 12187); g, distal portion of
style and stigma, x 15 (Daniel & Bartholomew 4828).
DANIEL: MEXICAN ACANTHACEAE
281
FIGURE 2. Scanning electron micrographs of pollen, a, Ruellia petiolaris (Nees) T. Daniel (Daniel & Breedlove 4930); b,
Justicia tabascina (Cowan 2860), polar view; c, Justicia ixtlania (Daniel 2070), equatorial view; d, /. tabascina (Cowan 2860),
equatorial view showing surface between 2 trema regions; e, J. tabascina (Cowan 2860), equatorial view showing 1 trema region;
f, Dyschoriste mcvaughii (Daniel 4828), equatorial view; g, Schaueria parviflora (Trigos 198), equatorial view; h, S. parviflora
(Herrera 72), subequatorial view; i, S. parviflora (Contreras 11180), equatorial view. Scale in c-i same as in b.
282
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 12
sessile with peduncles to 1.5 mm long. Bractlets
and secondary bractlets linear to linear-elliptic
to linear-oblanceolate, 5.5-11 mm long, 1-3 mm
wide, pubescent like leaves. Calyx 9-12 mm long,
tube 3-5 mm long, hyaline between veins, lobes
subulate, 6-9 mm long, 1 .4-3 times longer than
tube, aristate and becoming stiff apically, pubes-
cent like leaves. Corolla red, 34-42 mm long,
externally pubescent with flexuose eglandular tri-
chomes, tube 24-33 mm long, slightly curved
and gradually ampliate distally but not differ-
entiated into a well-defined throat, upper lip 9-
1 1 mm long with 2 subcirculate lobes 4.5-6 mm
long, 4.5-7 mm wide, lower lip 8.5-12 mm long
with 3 lobes obovate-elliptic to obovate, 8-10
mm long, 5-7 mm wide. Stamens didynamous,
9-14 mm long, the two pairs fused for 1-2 mm
at base, thecae 1.8-2.4 mm long, muticous or
with blunt to pointed appendages 0.05-0.1 mm
long at base, pollen prolate, 3-colporate, inter-
colpal regions multi-striate with 5-14 pseudo-
colpi of irregular lengths, exine minutely verru-
cate. Style 27-38 mm long, pubescent with
eglandular trichomes, stigma filiform, 1.3-1.5
mm long. Capsule ellipsoid, 7 mm long, gla-
brous. Seeds not seen.
DISTRIBUTION AND HABITAT.— Central and
western Jalisco (Fig. 3); in disturbed areas of oak
and oak-pine forest at elevations from 1,600 to
1,900m.
PHENOLOGY.— Flowering: March-June.
Emery Leonard annotated McVaugh's collec-
tion as an undescribed species of Ruellia L. On
the basis of many of its macromorphological at-
tributes, this species might indeed be mistaken
for Ruellia. Dyschoriste Nees is usually easily
distinguished from Ruellia by the presence of
conspicuous pointed appendages at the base of
each theca which are not known among Amer-
ican Ruellia. In D. mcvaughii these appendages
are inconspicuous or absent. Characters of the
pollen and calyx preclude the placement of this
species in Ruellia, however. Pollen of Ruellia
(Fig. 2a) is spherical and three-porate (Raj 1961,
1973). It is best characterized by its homobro-
chate-reticulate exine. Pollen of Dyschoriste (Fig.
2f) is prolate and three-colporate. The exine is
verrucate with minute, rounded projections and
striate with multiple pseudocolpi between the
colpi. In Mexican Dyschoriste, the calyx tube is
often prominently angled and hyaline between
the major veins which extend into the lobes. The
tube commonly ruptures at maturity in these weak
regions. In most species of Dyschoriste, the calyx
lobes are long attenuate to aristate and become
stiff at maturity. Although there is considerable
variation in calyx form among Mexican species
of Ruellia, none have a partially hyaline tube
with stiff, aristate lobes. In features of both pollen
and calyx form, the plants described here con-
form to other species of Dyschoriste.
The reduced thecal appendages of/), mcvaughii
are an unusual feature in the genus. Anthers of
McVaugh 12187 vary from having blunt ap-
pendages to lacking appendages altogether. Dan-
iel and Bartholomew 4828, collected in the same
general region, has anthers with the pointed ap-
pendages typical of other species of Dyschoriste.
Elsewhere in the genus, variation of thecal ap-
pendages is evident among the Chiapas collec-
tions of D. ovata (Cav.) Kuntze cited by Daniel
(1986). In these collections, thecal appendages
vary from one or more hairlike projections to a
single stout mucro.
Kobuski (1928) revised the American species
of Dyschoriste and recognized 2 1 species as oc-
curring in Mexico, none of which are similar to
D. mcvaughii. Among the 40 American species
he treated, only D. pringlei Greenman from Ja-
lisco has corollas similar in size to those of D.
mcvaughii. Recent study of D. pringlei for
McVaugh's Flora Novo-Galiciana (Daniel, un-
published) reveals that this species differs from
D. mcvaughii by having conspicuously flattened
young stems, the dichasia crowded at or near the
stem apex resulting in a headlike floral cluster,
and blue corollas. A closer relative is undoubt-
edly D. rubiginosa Ramamoorthy & Wassh., a
species with large reddish corollas recently trans-
ferred to Dyschoriste from Hygrophila R. Br.
(Ramamoorthy and Wasshausen 1985). These
two species can be distinguished from all other
Mexican Dyschoriste by their reddish corollas
and they can be distinguished from one another
by the following couplet:
Diminutive perennial herb to 1 dm tall; distal
leaves 1 . 1^4 times longer than wide; flowers
1-3 in leaf axils; bractlets straight; corolla
34-42 mm long; thecal appendages absent
or up to 0. 1 mm long D. mcvaughii
Perennial herb or shrub to 1.5 m tall; distal
leaves 5-15 times longer than wide; flowers
more than 3 (up to 16) in leaf axils; bractlets
DANIEL: MEXICAN ACANTHACEAE
283
D DYSCHORISTE MCVAUGHII
A JUSTICIA IXTLANIA
• JUSTICIATABASCINA
• SCHAUERIA PARVIFLORA
FIGURE 3. Map showing distribution of Dyschoriste mcvaughii, Justicia ixtlania, J. tabascina, and Schaueria parvijlora.
usually conspicuously curved; corolla (20-)
25-34 mm long; thecal appendages 0.1-0.2
mm long D. rubiginosa
PARATYPES.— MEXICO. Jalisco: between Ameca and Aten-
guillo, 1 4.6-1 6.2 mi W of Mixtlan, T. Daniel & B. Bartholomew
4828 (CAS); Chiquilistlan, M. Jones 378 (POM, US); La Pal-
ma, M. Jones s.n. (POM).
Justicia ixtlania T. F. Daniel, nom. nov.
Ixtlania acicularis M. E. Jones, Contr. West. Bot. 15:151.1 929.
TYPE.— MEXICO. Nayarit: Ixtlan [del Rio], 19 February
1927, M. E. Jones 23534 (Holotype: POM!, photo and frag-
ments: US!; isotype: MO!). Not /. acicularis Sw. (1788).
Erect to reclined perennial herb to 3 dm tall.
Stems subterete to subquadrate, sulcate, pubes-
cent with flexuose to recurved eglandular tri-
chomes 0.5-1.5 mm long, the trichomes evenly
disposed or usually concentrated in 2 vertical
lines, mature stems often glabrate. Leaves short-
petiolate with petioles 2-1 1 mm long, blades
somewhat coriaceous, lanceolate to ovate to el-
liptic (sometimes narrowly so) to subcirculate to
obovate, 12-85 mm long, 1 1-47 mm wide, 1.1-
5 times longer than wide, rounded to acute at
apex, acute to rounded to truncate to subcordate
at base, surfaces glabrous or pubescent, margin
entire to subcrenate, white-callose. Inflorescence
of axillary and/or terminal, densely bracteate,
often clustered spikes (or spikelike thyrses) to 5
cm long, rachis pubescent with an understory of
straight glandular and eglandular trichomes to
0. 1 mm long and an overstory of longer flexuose
eglandular trichomes, flowers 1 per node. Bracts
sometimes conduplicate, ovate-lanceolate to
lance-subulate, (3-)5-10 mm long, 0.8-1.5 mm
wide, abaxial surface pubescent with an under-
story of straight glandular and eglandular tri-
chomes to 0. 1 mm long and an overstory of flex-
uose eglandular trichomes 0.2-1 mm long, margin
subscarious and ciliate. Bractlets lanceolate to
lance-subulate, (4-)5.5-9 mm long, 0.8-1.5 mm
wide, pubescent like bracts. Calyx 5-lobed, 6.5-
9 mm long, lobes divided nearly to base, lan-
ceolate to lance-subulate, pubescent like bracts,
margins scarious, posterior lobe reduced in size.
Corolla entirely white to pink-purplish with pur-
ple markings on lower lip, 12-16 mm long, ex-
ternally pubescent with straight to flexuose eglan-
dular trichomes to 0.6 mm long, tube 8-1 1 mm
284
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 12
long, slightly ampliate distally, upper lip 3-4 mm
long, bilobed with rounded lobes 0.1-0.5 mm
long at apex, lower lip 3.5-5 mm long, with 3
subcirculate to obovate lobes 3-4 mm long, 2.5-
3.5 mm wide, central lobe larger than lateral lobes.
Stamens 3.5-4 mm long, thecae 0.6-1 mm long,
unequally inserted, the lower (sometimes slightly
larger) with a blunt calcarate basal appendage
0.2-0.3 mm long, pollen prolate, 2-colporate with
each colpus flanked by 2 pseudocolpi, exine re-
ticulate. Style 9-1 1 mm long, glabrous, stigma
unlobed, subcirculate, oblique, 0.2-0.3 mm long.
Capsule short-stipitate, 3-5 mm long, stipe 0.5-
1.3 mm long, head ellipsoid, 2.5-3.7 mm long,
externally pubescent with downward pointing to
appressedeglandulartrichomes 0.2-0.3 mm long
and straight glandular trichomes up to 0. 1 mm
long (mostly near apex). Seeds 4 per capsule,
somewhat flattened, subcordate to subcirculate
to elliptic in outline, 1-1.3 mm long, 0.8-1 . 1 mm
wide, surface papillose (n = 14, Daniel and Bar-
tholomew 4788).
DISTRIBUTION AND HABITAT.— West-central
Mexico (Sinaloa, Nayarit, Jalisco, and Michoa-
can) (Fig. 3); along streambanks in deciduous or
subdeciduous forest at elevations from 150 to
1,250m.
PHENOLOGY.— Flowering and fruiting: March-
April.
Leonard (1945) accepted Ixtlania as a viable
genus, considered a probable relationship with
Beloperone Nees, and noted a superficial resem-
blance to Chaetothylax (erroneously cited as
"Chaetochlamys") umbrosus Nees of South
America. Chaetothylax Nees differs from Ixtla-
nia by having four (vs. five) calyx lobes, dimor-
phic thecae with the lower theca reduced or ap-
pearing abortive and separated from the upper
theca by about 1 mm (vs. homomorphic thecae
that are in contact with one another), and pu-
bescent (vs. glabrous) seeds.
In a recent delimitation and infrageneric clas-
sification of Justicia, Graham (1988) included
both Beloperone and Chaetothylax within Jus-
ticia. She did not examine material or discuss
the disposition of Ixtlania, however. On the basis
of the diagnostic characters of the corolla (i.e.,
rugulate) and androecium (i.e., two stamens with
bithecous anthers, the lower thecae with basal
appendages), it is clear that Ixtlania is likewise
congeneric with the large and polymorphic Jus-
ticia. The chromosome number noted above for
this species is also the most common number in
Justicia (Daniel et al. 1984, in press).
Using Graham's (1988) provisional key to in-
frageneric taxa of Justicia, J. ixtlania appears to
be most closely affiliated with sect. Leucoloma
V. Graham from South America. Justicia ixtla-
nia differs from the two species of this section
by the appendages of the lower thecae which are
more than one-fifth of the length of the thecae.
It resembles Graham's description of this section
in most other aspects, however, including the
pollen form (Fig. 2c). In Justicia two-colporate,
four-pseudocolpate pollen is known only in sect.
leucoloma and two sections restricted to the Old
World. This type of pollen has also been reported
in Mexacanthus T. Daniel (Daniel 1981), a
monotypic genus of Odontoneminae endemic to
Mexico.
Considerable variation in corolla color was
noted within living populations of this species.
The specimen from Michoacan differs from those
from Jalisco, Nayarit, and Sinaloa by its narrow-
er, lanceolate to narrowly elliptic to oblanceolate
leaves. In other respects it is identical to the more
northerly collections.
ADDITIONAL SPECIMENS EXAMINED.— MEXICO. Jalisco: vie.
of Arroyo de Chorillo (tributary of Rio de Quimixto), ca. 14
km SW of Puerto Vallarta, A. Carter & F. Chisaki 1249 (MEXU,
MICH, UC); Mpio. Cuautitlan, 1 .5-2 km W de Tequesquitlan,
R. Cuevas G. 1052 (CAS); along Hwy. 200 S of Puerto Vallarta,
7.6 mi S of El Tuito, T. Daniel 2070 (ASU, CAS, DUKE, K,
MEXU, MO); between Tomatlan and Talpa de Allende, ca.
1 1 mi N of Tomatlan, T. Daniel 2086 (ASU); along Hwy. 15
ca. 1 mi E of Nayarit border, T. Daniel & B. Bartholomew
4788 (CAS); Mpio. La Huerta, Est. Biol. Chamela, Arroyo
Chamela, E. Lott 967 (ASU, CAS, DUKE, MO); Mpio. Te-
quila, Barranca de los Tanques, desv. al camino de San Martin
de las Canas, L. Puga 6114 (IBUG). Michoacan: Los Chorros
del Varal, Los Reyes, B. Guillen Jimenez s.n. (IBUG). Nayarit:
30 km NE of Tepic, 2 km E of Poicho titan, M. Baker 1104
(ASU); La Bahada, SE of San Bias, E. Lehto 24208 (ASU);
San Felipe to Arroyo San Jose del Conde, Y. Mexia 1930 (A,
CAS, F, GH, MICH, UC, US); Potrero la Taberna, km 28 de
Amatlan de Canas, carr. Tepic-Barranca del Oro, B. Ahuet
Juan R. s.n. (IBUG). Sinaloa: along Hwy. 40 between Villa
Union and Cd. Durango, 26 mi SW of Tropic of Cancer, T.
Daniel 4019 (CAS); near Microondas tower near Villa Union,
E. Lehto 24273A (ASU); Mpio. Concordia, along Hwy. 40 ca.
5 mi NE of Concordia, A. Sanders et al. 4996 (CAS); Mpio.
Cosala, arroyo NE de Guadalupe de Los Reyes, P. Tenorio L.
et al. 8345 (CAS).
Justicia tabascina T. F. Daniel, nom. nov.
Tabascina lindenii Baillon, Hist. PL 10:445. 1891. (as T. lin-
deni) TYPE.— MEXICO. Tabasco: forets de Teapa, October,
DANIEL: MEXICAN ACANTHACEAE
285
J. Linden s.n. (Holotype: P!( photographs: CAS, F). Not J.
/iiwfe«/iHoullet(1870).
Shrub to 2.5 m tall. Young stems subquadrate
to quadrate, evenly pubescent with antrorse
eglandular trichomes to 0.6 mm long, internodes
often constricted just above nodes. Leaves pet-
iolate with petioles to 33 mm long, blades ovate-
elliptic to elliptic, 36-165 mm long, 12-83 mm
wide, 2-3 times longer than wide, acuminate at
apex, acute to subattenuate at base, surfaces pu-
bescent with cauline type trichomes mostly or
entirely restricted to major veins. Inflorescence
of 1-4 terminal racemose thyrse(s) to 7.5 cm
long, flowers paired at nodes, sessile to short (to
1 mm) pedicellate in axil of 2 bractlets, the flow-
ers and bractlets pedunculate in axil of a bract
with peduncles to 12 mm long, rachis and pe-
duncles evenly pubescent with flexuose to an-
trorse eglandular trichomes to 0.6 mm long.
Bracts often caducous, linear, 4-7 mm long, 0.6-
1.5 mm wide, pubescent like rachis. Bractlets
linear, 2.5-6 mm long, 0.8-1 mm wide, pubes-
cent like rachis. Calyx 5-lobed, 10-16 mm long,
tube 1-2 mm long, lobes valvate or nearly so,
anterior lobes often connivent along edges,
asymmetrically elliptic to lance-elliptic, 8-1 1 mm
long, 2-3.2 mm wide, posterior lobe ovate-ellip-
tic to elliptic, 10-13 mm long, 4-5 mm wide, all
lobes pubescent with straight to flexuose eglan-
dular trichomes to 0.5 mm long. Corolla yellow,
24 mm long, externally pubescent with glandular
trichomes 0.05-0.2 mm long, tube 13 mm long,
upper lip 1 1 mm long, entire, internally rugulate,
lower lip 10 mm long with 3 rounded lobes to 3
mm long and 2.5 mm wide. Stamens 9 mm long,
filaments glabrous, thecae subsagitate, unequally
inserted, 5.3 mm long, rounded and lacking basal
appendages, pollen prolate, 4-porate (to 4-sub-
colporate) with pores in a trema region contain-
ing 2 longitudinal rows of circular insulae, exine
reticulate. Style 15-18 mm long, glabrous, stigma
more or less capitate, 0.2 mm long. Capsule 19-
22 mm long, glabrous, stipe 8-9 mm long, head
11-13 mm long. Seeds 4, flattened, subcirculate
in outline, 4.2-5 mm long, 3.8-4.8 mm wide, the
surfaces smooth.
DISTRIBUTION AND HABITAT.— South-central
Tabasco (Fig. 3); in lower montane rain forest at
an elevation of about 140 m.
PHENOLOGY. —Flowering: March and October;
fruiting: March.
The recent rediscovery of plants resembling
the type, and only known collection, of Tabas-
cina lindenii allows for a reevaluation of this
monotypic genus. Cowan's collection was made
in the region of the type locality and contains
fruits which were unknown to Baillon (1891).
Lindau (1895) maintained Tabascina as a genus
of Justicieae, differing from Justicia by its broad,
leaflike (vs. narrow) calyx lobes and flowers in
terminal cymes (vs. flowers solitary or in spikes
or contracted panicles). The rugulate corolla, an-
droecium of two bithecous stamens, and pollen
with trema regions studded with circular insulae
(Fig. 2d, e) all suggest placement of this species
within Justicia. The large, foliaceous, and het-
eromorphic calyx lobes that are nearly or com-
pletely valvate constitute an unusual feature of
the species; however, among Mexican Justicia,
J. caudata A. Gray also has valvate, or nearly
valvate, and partially connivent calyx lobes. The
inflorescence of/, tabascina consists of a single
terminal thyrse (i.e., an indeterminate main axis
with determinate lateral axes), or a cluster of
these. The lateral axes consist of dichasia in which
usually only a single peduncled flower develops.
Additional floral buds of the dichasia are evident
on the youngest portions of the thyrse but are
represented only by prominent scars on the older
portions. In view of the diversity of calyx forms
and inflorescence types documented for species
of Justicia, these characters would not appear to
warrant recognition of Tabascina as distinct from
Justicia.
Graham (1988) noted that four-aperturate pol-
len has been reported in a few New World species
of Justicia but she was unable to confirm its oc-
currence. Justicia tabascina, with its four-aper-
turate pollen (Fig. 2b), can not be readily clas-
sified in any of the infrageneric taxa recognized
by Graham (1988). Pollen of this species differs
from Graham's (1988) type 3 pollen, which is
characteristic of section Drejerella (Lindau) Gra-
ham, only by the presence of four rather than
three apertures. Justicia tabascina exhibits sev-
eral macromorphological features (e.g., unap-
pendaged thecae) that exclude it from this sec-
tion, however.
ADDITIONAL SPECIMEN EXAMINED.— MEXICO. Tabasco:
Mpio. Tacotalpa, Cerro de Madrigal, 7 km de las Est. Tacotalpa
hacia Tapijulapa, C. Cowan et al. 2860 (CAS).
286
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 12
Schaueria parviflora (Leonard) T. F. Daniel,
comb. nov.
Streblacanthus parviflorus Leonard, J. Wash. Acad. Sci. 31:
1941. TYPE.— GUATEMALA. Izabal: Escoba, across bay
(west) from Puerto Barrios, 3 May 1939, P. Standley 72949
(Holotype: F!).
Schaueria calycobractea Hilsenbeck & Marshall, Brittonia 35:
362. 1983. TYPE.-MEXICO. Veracruz: Estacion de Biolo-
gia Tropical Los Tuxtlas, 22 July 1978, D. Marshall 1 (Ho-
lotype: US fide Hilsenbeck and Marshall 1983 but not lo-
cated there in 1989; isotypes: MEXU, TEX!, US).
Perennial herb to shrub to 2 m tall. Young
stems subquadrate to quadrate-sulcate (often
somewhat compressed), sparsely pubescent in 2
decussate lines with antrorse to flexuose to re-
trorse eglandular trichomes 0.1-0.4 mm long,
soon glabrate. Leaves petiolate with petioles to
80 mm long, (lance-ovate to) ovate to ovate-
elliptic, 50-180 mm long, 14-90 mm wide, 2-
3.6 times longer than wide, acuminate to subfal-
cate at apex, (subcordate to) rounded to acute
and usually somewhat asymmetric at base, sur-
faces glabrous (or pubescent along midvein of
adaxial surface with cauline type trichomes). In-
florescence of terminal, sessile to pedunculate
spikes to 55 mm long, peduncles to 1 5 mm long,
rachis pubescent with antrorse to flexuose to re-
trorse eglandular trichomes 0.1-0.2 mm long
either in 2 lines or more or less evenly disposed,
flowers opposite in axils of bracts (sometimes in
axils of distalmost pair of leaves as well). Bracts
linear-lanceolate to lance-subulate, 12-22 mm
long, 1.2-2.2 mm wide, abaxial surface glabrous
to very sparsely pubescent with antrorse eglan-
dular trichomes 0. 1 mm long (especially distally)
and rarely with glandular trichomes 0.05-0. 1 mm
long as well, margin sparsely ciliate with similar
trichomes. Bractlets linear-setaceous to lance-
subulate, 7-30 mm long, 0.8-1 mm wide, pu-
bescent like bracts although always with glan-
dular trichomes. Calyx 5-lobed, 10-25 mm long,
lobes linear-setaceous to subulate, 9-24 mm long,
0.5-1.5 mm wide, irregularly unequal (with 2 or
more lobes up to 25% longer and 50% wider than
the others), 9-24 times longer than the tube, pu-
bescent like bractlets. Corolla yellow (or white
in Guatemala), 22-28 mm long, externally pu-
bescent with eglandular trichomes 0.1-0.2 mm
long, tube 14-18.5 mm long, ampliate at base
and apex, upper lip 4.5-10 mm long, emarginate
with 2 rounded lobes 0.3-0.7 mm long, lower lip
5.5-9.5 mm long with 3 lobes, lateral lobes linear
to elliptic to oblanceolate-elliptic, 5-7 mm long,
1.2-2.8 mm wide, central lobe elliptic to oblan-
ceolate, 5.5-7 mm long, 1.5-3 mm wide. Sta-
mens attached 3.5-5 mm proximal to mouth of
corolla, 8-9 mm long, thecae parallel, equally
inserted, subequal, 1.4-2.5 mm long, pointed at
base, pollen subspheric, 3-colporate with each
colpus flanked by 2 pseudocolpi, pseudocolpi
sometimes fused in intercolpal regions forming
a pseudocolpal ring, exine reticulate, reticulum
of intercolpal regions sometimes irregularly di-
vided or fragmented. Style 19-23 mm long, gla-
brous, stigma bilobed with lobes 0.2 mm long.
Capsule clavate, 10.5-16 mm long, glabrous (or
pubescent with sparse glands to 0.05 mm long
in Guatemala), stipe 4.5-8.5 mm long, head 6-
9 mm long, retinacula 2-2.8 mm long. Seeds 4,
flat, subcirculate to subcordate in outline, 3.5-
5.2 mm long, 3-4.9 mm wide, surfaces and mar-
gin papillose.
DISTRIBUTION AND HABITAT.— Southeastern
Mexico (southern Veracruz) and east-central
Guatemala (Izabal) (Fig. 3); in lowland to lower
montane rainforest (primary and secondary
growth) from near sea level to 700 m.
PHENOLOGY. — Flowering: January, March-
April, July, and October; fruiting: March, June-
July, and October.
In the protologue of Streblacanthus parviflo-
rus, Leonard (1941) noted a relationship with S.
cordatus Lindau, a species of southern Central
America. Gibson (1974) noted the occurrence of
four calyx lobes in Streblacanthus Kuntze and
maintained S. parviflorus as the northernmost-
occurring species in the genus. In 1983, Hilsen-
beck and Marshall described Schaueria calyco-
bractea from Veracruz and noted that it appears
similar to Streblacanthus parviflorus from Gua-
temala. They further noted that Streblacanthus
could be distinguished from Schaueria Nees by
its four-lobed (vs. five-lobed) calyx and unequal-
ly (vs. nearly equally) inserted thecae, but that
these generic distinctions had not always been
followed. They concluded that these genera are
very similar and may not deserve separation. I
agree with them that further study of the generic
limits of Schaueria and Streblacanthus is desir-
able. I also concur with placement of the plants
from the vicinity of the Estacion de Biologia
Tropical Los Tuxtlas in Schaueria based on the
distinctions provided above. However, the type
of Streblacanthus parviflorus and other collec-
DANIEL: MEXICAN ACANTHACEAE
287
tions from Guatemala likewise have five-lobed
calyces and subequally inserted thecae. There-
fore, they should also be treated in Schaueria.
Since the description of Schaueria calycobrac-
tea (Hilsenbeck and Marshall 1983), in which
two collections were cited, additional collections
of this species have been made in the vicinity of
San Andres Tuxtla and in the Uxpanapa region
about 1 50 km to the southeast. The three known
collections from Guatemala differ from the Mex-
ican collections by the presence of sparsely dis-
tributed glandular trichomes (vs. glabrous in
Mexico) on the capsules and white (vs. yellowish
in Mexico) corollas. Given the overall similarity
in all other characters, these distinctions seem
relatively minor.
Pollen of Schaueria parviflora from Veracruz
(Fig. 2g, h) is similar to that of plants from Gua-
temala (Fig. 2i) and agrees with a description of
the type, S. calycotricha (Link & Otto) Nees, from
South America (Raj 1961).
ADDITIONAL SPECIMENS EXAMINED.— GUATEMALA. Iza-
bal: El Estor, E. Contreras 11180 (LL); between Escobas and
Montana Escobas, across bay from Puerto Barrios, /. Steyer-
mark 39324 (F). MEXICO. Veracruz: Mpio. San Andres Tux-
tla, cerca de Laguna Escondida, entre Sontecomapan y Mon-
tepio, J. Beaman 5763 (F); Mpio. San Andres Tuxtla, Estacion
de Biologia Tropical Los Tuxtlas, J. Calzada 764 (F, US), A.
Gdmez-Pompa 4484 (F), G. Ibarra M. 540 (MO), G. Ibarra
M. & S. Sanaca C. 1776 (MO), G. Martinez C. 1708 (CAS,
F), R. C. Trigos 198 (F, US), G. Webster 20895 (TEX); Mpio.
Minatitlan, Zona de Uxpanapa, 1-2 km SE of La Chinantla
(17°15'30"N, 94°26'30"W), T. Daniel & T. Wendt 5810 (CAS,
CHAPA, MEXU); Mpio. San Andres Tuxtla, al sur de la Ebi-
trolotu, A. Villegas H. 72 (CAS, F, US); Mpio. Hidalgotitlan,
desde el Poblado 6, al S por la brecha y la vereda al horcajo
oriental del Rio Cuevas, 17°15'N, 94°30'W, T. Wendt et al.
2604 (CAS).
ACKNOWLEDGMENTS
I am grateful for the financial assistance pro-
vided by the National Science Foundation (BSR-
8609852), the American Philosophical Society,
Arizona State University, and the California
Academy of Sciences. Mary Ann Tenorio pro-
vided the line drawing and SEM micrographs.
My field work in Veracruz was made possible by
Tom Wendt. Loans and/or other courtesies were
provided by the herbaria cited in the text.
LITERATURE CITED
BAILLON, H. 1891. Histoire des plantes, Vol. 10. Librairie
Hachette, Paris.
DANIEL, T. F. 1981. Mexacanthus, a new genus of Acantha-
ceae from western Mexico. Syst. Bot. 6:288-293.
. 1986. Acanthaceae. Pp. 27-30 in Listados floristicos
de Mexico. IV. Flora de Chiapas. D. E. Breedlove, ed. In-
stituto de Biologia, UNAM, Mexico.
DANIEL, T. F., B. D. PARFITT, AND M. A. BAKER. 1984. Chro-
mosome numbers and their systematic implications in some
North American Acanthaceae. Syst. Bot. 9:346-355.
DANIEL, T. F., T. I. CHUANG, AND M. A. BAKER. In Press.
Chromosome numbers of American Acanthaceae. Syst. Bot.
GIBSON, D. N. 1974. Acanthaceae. Pp. 328^61 in Flora of
Guatemala (Fieldiana, Bot. 24 [pt. 10]). P. C. Standley et
al., eds. Field Museum of Natural History, Chicago.
GRAHAM, V. A. 1988. Delimitation and infra-generic clas-
sification ofJusticia (Acanthaceae). Kew Bull. 43:551-624.
HILSENBECK, R. A. AND D. L. MARSHALL. 1983. Schaueria
calycobractea (Acanthaceae), a new species from Veracruz,
Mexico. Brittonia 35:362-366.
KOBUSKI, C. E. 1928. A monograph of the American species
of the genus Dyschoriste. Ann. Missouri Bot. Gard. 15:9-
91.
LEONARD, E. C. 1941. New Acanthaceae from Guatemala.
J. Wash. Acad. Sci. 31:96-105.
. 1945. Acanthaceae. P. 1 1 5 in Mexican phanerogams
described by M. E. Jones (Contr. U.S. Natl. Herb. 29:87-
116). C.V.Morton, ed.
LINDAU, G. 1895. Acanthaceae. Pp. 274-354 in Die Natttr-
lichen Pflanzenfamilien, Vol. 4 (3b). H. G. A. Engler and K.
A. E. Prantl, eds. Wilhelm Englemann, Leipzig.
RAJ, B. 1961. Pollen morphological studies in the Acantha-
ceae. Grana Palynol. 3:3-108.
. 1973. Further contribution to the pollen morphology
of the Acanthaceae. J. Palynol. 9:91-141.
RAMAMOORTHY, T. P. AND D. C. WASSHAUSEN. 1985. Anew
name in Dyschoriste (Acanthaceae). Brittonia 37:358-359.
CALIFORNIA ACADEMY OF SCIENCES
Golden Gate Park
San Francisco, California 94 1 1 8
PROCEEDINGS
OF THE
CALIFORNIA ACADEMY OF SCIENCES
Vol. 46, No. 13, pp. 289-297, 5 figs. September 11, 1990
THREE NEW SPECIES OF
JUSTICIA (ACANTHACEAE) FROM PANAMA
By
Thomas F. Daniel
Department of Botany, California Academy of Sciences,
Golden Gate Park, San Francisco, California 94118
and
Dieter C. Wasshausen
Department of Botany, National Museum of Natural History,
Smithsonian Institution, Washington, D.C. 20560
ABSTRACT: Three new species of Justicia, J. fortunensis, J. veraguensis, and J. readii, are described from
Panama. The three species are illustrated and their known distributions are mapped. A key to these and a
related species in Panama, /. refractifolia, is provided. A chromosome number of n = 14 is reported for J.
fortunensis. With 23 species now known from Panama, Justicia is the largest genus of Acanthaceae in that
country.
Received December 11, 1989. Accepted March 14, 1990.
common with J. refractifolia (Kuntze) Leonard
INTRODUCTION . . ' _/ A11\. ...
of Costa Rica and Panama. All four have axillary
The Acanthaceae of Panama comprise 109 and terminal inflorescences of conspicuously and
species in 34 genera (Durkee 1988). Justicia is densely bracteate spikes with the bracts and-co-
the largest and taxonomically most complex ge- rollas less than 20 mm long. These characters
nus of Acanthaceae with between 420 (Mabber- distinguish this group of species from all other
ley 1987) and 600 or so (Graham 1988) species Panamanian Justicia. A key to these four species
worldwide. Durkee (1988) noted the occurrence is presented below,
of 2 1 species of Justicia in Panama. Recent ex-
plorations in the species-rich montane regions 1. Young stems pubescent with whitish to light
near Fortuna Dam and on Cerro Tute in western brown trichomes (and bracts not drying dark
Panama have resulted in the collection of two red); calyx pubescent with eglandular tri-
undescribed species of Justicia. Another species, chomes; filaments pubescent,
previously known only from Cerro Jefe, and 2. Bracts dark pink (often drying brown-
treated under the name of a species from the Old ish); corolla 8-10 mm long, externally
World, is recognized as new and its range within pubescent with glandular and eglandu-
Panama is expanded. lar trichomes; filaments proximally pu-
These three species share several features in bescent with eglandular trichomes, dis-
[289]
290
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 13
tally glabrous; capsule pubescent with
glandular and eglandular trichomes;
seeds smooth J. fortunensis
2. Bracts green (drying dark green or
brown); corolla 10-17 mm long, exter-
nally pubescent with eglandular tri-
chomes; filaments pubescent with glan-
dular trichomes throughout length;
capsule pubescent with eglandular tri-
chomes; seeds roughened with minute
rounded protrusions J. refractifolia
1. Young stems glabrous or pubescent with
golden trichomes (if trichomes light brown
or straw-colored as sometimes in J. readii,
then bracts drying dark red); calyx pubes-
cent with glandular and eglandular tri-
chomes; filaments glabrous.
3. Young stems glabrous; bracts drying pale
to dark greenish; bractlets ovate-elliptic
to elliptic to oblanceolate to obovate;
calyx 5-6 mm long; corolla 10-13 mm
long, externally pubescent with glan-
dular and eglandular trichomes; seeds
smooth /. veraguensis
3. Young stems evenly covered with a
dense layer of (flexuose to) antrorse to
antrorsely appressed golden (to light
brown or straw-colored) eglandular tri-
chomes; bracts drying dark red; bract-
lets linear-lanceolate to linear to linear-
elliptic; calyx 7.5-10 mm long; corolla
15-19 mm long, externally pubescent
with eglandular trichomes; seeds papil-
lose to verrucose J. readii
Although Leonard (1958) recognized more than
70 species of Justicia in Colombia and Durkee
(1986) treated 27 species from Costa Rica, none
of the taxa from these adjacent regions appears
to be conspecific with the three species newly
described below. Justicia is the largest genus of
Acanthaceae in Panama with 23 species now
known from the country.
Justicia fortunensis sp. nov.
(Fig. 1)
TYPE.— PANAMA. Chiriqui: vicinity of Fortuna Dam, along
trail across valley of Rio Homito, 12 March 1988, T. Daniel,
F. Almeda, and G. McPherson 5550 (Holotype: CAS!; isotypes:
DUKE!, F!, GH!, K!, MICH!, MO!, NY!, PMA!, US!).
Suflrutex vel frutex usque ad 1 m altus. Caules juniores
trichomatibus cglandulosis antrorsis (vel retrorsis) 0.2-0. 5 mm
longis in 2 lineis dispositis pubescentes. Folia petiolata, lam-
inae lanci-ellipticae vel ovato-ellipticae vel ellipticae, 45-152
mm longae, 10-56 mm latae, 2.1-5-plo longiores quam lati-
ores. Inflorescentia spicata dense bracteata, rhachis trichoma-
tibus eglandulosis pubescens. Bracteae atroroseae saepe in sic-
co brunneolas late ellipticae vel subcirculares vel obovatae 7-
13(-17) mm longae 6-11.5 mm latae. Bracteolae ellipticae.
Calyx 4.5-6.5 mm longus extus eglandulosus. Corolla 8-10
mm longa extus glandulosa. Stamina 2.8-3 mm longa fila-
mentis pubescentibus. Stylus 4.5-5 mm longus. Capsula 10
mm longa extus glandulosa. Semina 2-3 mm longa 2-2.5 mm
lata laevia.
Perennial herb to shrub to 1 m tall. Young
stems subquadrate to quadrate, pubescent with
antrorse (or retrorse), whitish to light brown (and
with conspicuous darker brown septae), eglan-
dular trichomes 0.2-0.5 mm long, mostly con-
centrated in 2 decussate lines, older stems be-
coming glabrate. Leaves petiolate, petioles to 14
mm long, pubescent like young stems, blades
lance-elliptic to ovate-elliptic to elliptic, 45-152
mm long, 10-56 mm wide, 2.1-5 times longer
than wide, acuminate to long-acuminate at apex,
acute to attenuate at base, surfaces sparsely pu-
bescent with cauline type trichomes along major
veins, otherwise glabrous. Inflorescence of pe-
dunculate, densely bracteate spikes to 60 mm
long, 1 0-20 mm in diameter near midspike, borne
singly (or sometimes 2) in axils of distal leaves,
peduncles to 45 mm long, pubescent like stems,
rachis more or less evenly pubescent with straight
to flexuose to antrorse, eglandular trichomes 0. 1-
0.4 mm long. Bracts dark pink (basal pair rarely
somewhat leaflike and partly green), drying pink-
ish or brownish, broadly elliptic to subcirculate
to obovate, 7-1 3(-l 7) mm long, 6-11.5 mm wide,
rounded or obtuse (and often apiculate) to acute
at apex, either abruptly constricted or tapering
into a short (to 3 mm long) petiole at base, abax-
ial surface pubescent with prominent or incon-
spicuous, straight to flexuose eglandular and stip-
itate glandular trichomes to 0.5 mm long, margin
prominently ciliate with coarse, conspicuously
multi-celled, eglandular trichomes to 1.4 mm
long. Bractlets elliptic, 6-1 2 mm long, 1.8-4 mm
wide, gradually tapering into a petiole at base,
pubescent like bracts. Calyx 4.5-6.5 mm long,
5-lobed, lobes dark pink (at least distally), di-
vided nearly to base of calyx, lanceolate to sub-
ulate, 0.5-0.8 mm wide, pubescent with eglan-
dular trichomes. Corolla white to very pale pink,
8-10 mm long, externally pubescent with stipi-
tate glandular and eglandular trichomes 0.1-0.3
DANIEL AND WASSHAUSEN: NEW PANAMANIAN JUSTICIA
291
2mm
FIGURE 1. a-d, Justicia fortunensis (Daniel et al. 5550). a, Habit; b, Bract; c, Bract, bractlets, calyx, and corolla; d, corolla
with stamens, e-h, Justicia readii (Read et al. 85-6b). e, Habit; f, Bract; g, Bracts, bractlet, calyx, and young corolla; h, corolla
with stamens.
292
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 13
\
20
p m
FIGURE 2. Camera-lucida drawing of meiotic chromo-
somes of/. fortunensis (Daniel etal. 5500, n = 14)atmetaphase
II.
mm long, tube 4-5.5 mm long, internally densely
pubescent, upper lip triangular, internally rugu-
late, 3.5-4 mm long with 2 lobes 0.4-0.7 mm
long, lower lip 4-4.5 mm long with 3 elliptic
lobes 3-3.5 mm long, 2-3.5 mm wide, central
lobe larger than lateral lobes. Stamens attached
near apex of tube, 2.8-3 mm long, exserted up
to 2 mm from mouth of corolla, filaments 1.5-
2 mm long, pubescent with eglandular trichomes
proximally, glabrous distally, thecae superposed,
the upper 1.4-1.5 mm long, the lower 1.5-1.8
mm long (including a blunt basal appendage to
0.2 mm long), pollen diporate (to subdicolporate)
with trema region traversed by 2 rows of 6-7
insulae. Style 4.5-5 mm long, distally glabrous,
proximally pubescent with upward pointing,
eglandular trichomes. Capsule 1 0 mm long, pu-
bescent with glandular and eglandular trichomes
0.1-0.3 mm long, stipe 4 mm long, head 6 mm
long. Seeds 4 per capsule, flattened, subcircular
to subelliptic in outline, 2-3 mm long, 2-2.5 mm
wide, surface smooth, (n = 14, Daniel et al. 5550,
Fig. 2)
DISTRIBUTION AND PHENOLOGY.— Known only
from Pacific slopes of the Cordillera Central in
western Panama (north-central Chiriqui) in the
watershed of the Rio Chiriqui (Fig. 3). Plants
occur in cloud forests at elevations from about
1025 to 1250 m. Flowering: January- April and
August; fruiting: January-April.
Justicia fortunensis is not readily identifiable
with any subgeneric taxa using the keys and de-
scriptions provided by Graham (1988). It has
pollen ("Type 5," Fig. 4) like that found in five
sections recognized by Graham. The chromo-
some number reported here for the type is the
most commonly encountered number in the ge-
nus and occurs in various subgeneric taxa (Daniel
and Chuang 1989; Daniel et al. 1984, 1990).
The Fortuna Dam area, the only region from
which this species is known, has received con-
siderable attention by plant collectors in the last
• FORTUNENSIS
A READII
• VERAGUENSIS
too
200 km
FIGURE 3. Map showing distribution of Justicia fortunensis, J. readii, and J. veraguensis.
DANIEL AND WASSHAUSEN: NEW PANAMANIAN JUSTICIA
293
FIGURE 4. Scanning electron micrographs of pollen ofJusticia. a, /. readii (Read et al. 85-6b), view of "trema region"; b, J.
readii (Read et al. 85-6b), view between 2 "trema regions"; c, /. veraguensis (Croat 34153), view of "trema region"; d, J.
fortunensis (Daniel et al. 5550), view of "trema region."
294
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 13
FIGURE 5. Justicia veraguensis (Croat 34153). a, Habit; b, Bract; c, Bractlets, calyx, and gynoecium; d, Corolla with stamens.
few years and has yielded a wealth of new taxa
(Dwyer 1985).
PARATYPES.— PANAMA. Chiriqui: Edwin Fa-
brega Dam and Reserve in Fortuna along trail
to Rio Hornito above Los Planes, F. Almeda et
al. 6320 (CAS, MO, PMA, US); W side of Rio
Hornito, W. D'Arcy 16003 (MO); La Fortuna
hydroelectric project, along Chiriqui River, B.
Hammel 2292 (MEXU); between Los Planes de
Hornito and Fortuna Lake, 8°41'N, 82°13'W, R.
Hampshire & C. Whitefoord 701 (BM); ca. 5 km
E of Fortuna Dam, along trail crossing Rio Hor-
nito, 8°45'N, 82°15'W, S. Thompson 5012 (CAS).
Justicia veraguensis, sp. nov.
(Fig. 5)
TYPE.— PANAMA. Veraguas: Cerro Tute, above Santa Fe
beyond Escuela Agricola Inter-americana, 1 .8 mi beyond fork
in rd., 5 April 1976, T. Croat 34153 (Holotype: NY!).
Frutex usque ad 2 m altus. Caules juniores glabri. Folia
petiolata, laminae ovato-ellipticae vel ellipticae 65-165 mm
longae 21-63 mm latae 2.6-3.4-plo longiorcs quam latiores.
DANIEL AND WASSHAUSEN: NEW PANAMANIAN JUSTICIA
295
Inflorescentia spicata dense bracteata, rhachis trichomatibus
eglandulosis et plerumque glandulosis pubescens. Bracteae in
sicco subviridulae vel atroviridulae obovatae vel late ellipticae
vel subcirculares vel deltatae 8-1 3 mm longae 6-8.5 mm latae.
Bracteolae ovato-ellipticae vel ellipticae vel oblanceolatae vel
obovatae. Calyx 5-6 mm longus extus glandulosus. Corolla
10-13 mm longa extus glandulosa. Stamina 5.5-6 mm longa
fi la mentis glabris. Stylus 8 mm longus. Capsula 9-1 1 mm longa
extus glandulosa. Semina 3-3.3 mm longa 2.7-3.7 mm lata
laevia.
Shrub to 2 m tall. Young stems quadrate, gla-
brous, older stems glabrous. Leaves petiolate,
petioles to 50 mm long, glabrous, blades ovate-
elliptic to elliptic, 65-165 mm long, 21-63 mm
wide, 2.6-3.4 times longer than wide, acuminate
at apex, acute to subattenuate at base, surfaces
glabrous. Inflorescence of axillary and terminal,
pedunculate, densely bracteate spikes to 65 mm
long, 13-21 mm in diameter near midspike, pe-
duncles to 15 mm long, glabrous or pubescent
in 2 lines with retrorse eglandular trichomes to
0.3 mm long, rachis pubescent with straight to
flexuose eglandular and glandular (sometimes not
evident) trichomes 0.1-0.3 mm long. Bracts
drying pale to dark greenish, obovate to broadly
elliptic to subcirculate to deltate, 8-13 mm long,
6-8.5 mm wide, rounded to acute at apex, ses-
sile or petiolate (petiole to 2 mm long), con-
stricted and/or tapering at base, abaxial surface
pubescent with straight to flexuose eglandular
and glandular trichomes 0. 1-0.3 mm long or be-
coming glabrate, margin ciliate with trichomes
like those of abaxial surface and with an over-
story of flexuose eglandular trichomes to 1 .2 mm
long. Bractlets ovate-elliptic to elliptic to oblan-
ceolate to obovate, 6-10 mm long, 1.7-2.7 mm
wide, tapering and often petiolate at base, pu-
bescent like bracts. Calyx 5-6 mm long, 5-lobed,
lobes divided nearly to base of calyx, lanceolate,
0.7-1.7 mm wide, pubescent like bracts. Corolla
white (to greenish white) with purplish markings,
10-13 mm long, externally pubescent with flex-
uose eglandular and glandular trichomes 0. 1-0.3
mm long, tube 6-6.5 mm long, internally gla-
brous, upper lip triangular, internally rugulate,
4-7 mm long, emarginate with lobes 0.2-0.3 mm
long, lower lip 4-7 mm long with 3 broadly el-
liptic to subcirculate lobes 2.5-4 mm long, 2-4.5
mm wide, central lobe larger than lateral lobes.
Stamens attached near apex of tube, 5.5-6 mm
long, exserted up to 5 mm beyond mouth of
corolla, filaments 4.5-5.2 mm long, glabrous,
thecae superposed, the upper 1.2-1.5 mm long,
the lower 1.7-2 mm long (including a basal spur
0.3-0.4 mm long), pollen diporate (to subdicol-
porate) with trema region traversed by 2 rows of
6-7 insulae. Style 8 mm long, distally glabrous,
proximally pubescent with eglandular trichomes.
Capsule 9-1 1 mm long, pubescent with glandular
and eglandular trichomes 0. 1-0.2 mm long, stipe
3.5-4 mm long, head 5.5-7 mm long. Seeds 4
per capsule, flattened, subcircular in outline, 3-
3.3 mm long, 2.7-3.7 mm wide, surface smooth.
DISTRIBUTION AND PHENOLOGY.— Known only
from the mountains of Veraguas in western Pan-
ama. The type and one paratype were collected
on the Pacific slopes of Cerro Tute in the Cor-
dillera Central of northern Veraguas (Fig. 3). We
have not been able to determine the exact lo-
cation in Veraguas of the other paratype; how-
ever, T. Croat (personal communication) sug-
gests that the Rio Segundo Brazo is the second
river along the road to the lowlands on the Ca-
ribbean slope of Cerro Tute. Plants occur on
forested slopes and on rocks in watercourses at
elevations from 630 to 1200 m. Flowering:
March-April and September; fruiting: April.
Justicia veraguensis exhibits a suite of char-
acters that precludes its placement into any of
the sections recognized by Graham (1988). It has
pollen (Fig. 4) similar to that of J. fortunensis,
which is undoubtedly a close relative.
PARATYPES.— PANAMA. Veraguas: 5 mi W of
Santa Fe on rd. past Escuela Agricola Alto Pie-
dra, R. Liesner 852 (GH, MO); Rio Segundo
Brazo, P. Maas & R. Dressier 1660 (F, U, US).
Justicia readii, sp. nov.
(Fig. 1)
TYPE.— PANAMA. Panama: Cerro Jefe, 5 January 1972,
J. Dwyer 9500 (Holotype: MO!; isotype: NY!).
Frutex usque ad 1.5 m altus. Caules juniores trichomatibus
eglandulosis aureis (vel pallidis brunneolis vel stramineis) (flex-
uosis vel) antrorsis vel appressis 0.4-2 mm longis aequaliter
pubescentes. Folia petiolata, laminae ellipticae vel obovatae
45-185 mm longae 17-78 mm latae 2.3-3.8-plo longiores quam
latiores. Inflorescentia spicata dense bracteata, rhachis tri-
chomatibus glandulosis et eglandulosis pubescens. Bracteae
virides in sicco atrovinosae obovatae vel subcirculares vel el-
lipticae 8-16 mm longae 4.5-11.5 mm latae. Bracteolae line-
ares-lanceolatae vel lineares vel lineares-ellipticae. Calyx 7.5-
10 mm longus extus glandulosus. Corolla 15-19 mm longa
extus eglandulosa. Stamina 5-6.5 mm longa filamentis glabris.
Stylus 11-16 mm longus. Capsula 9-12.5 mm longa extus
glandulosa. Semina 2.3-2.6 mm longa, 2.3-2.5 mm lata papil-
losa vel verrucosa.
Shrub to 1.5 m tall. Young stems subquadrate
to quadrate, densely and evenly pubescent with
296
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 13
(flexuose to) antrorse to antrorsely appressed,
golden (to light brown or straw-colored), eglan-
dular trichomes 0.4-2 mm long, older stems re-
maining pubescent. Leaves petiolate, petioles to
30 mm long, pubescent like stems, blades elliptic
to obovate, 45-185 mm long, 17-78 mm wide,
2.3-3.8 times longer than wide, acute to acu-
minate at apex, acute to attenuate at base, mid-
vein pubescent like stems, surfaces sparsely pu-
bescent with similar trichomes. Inflorescence of
axillary and terminal, pedunculate, densely brac-
teate spikes to 75 mm long, 12-21 mm in di-
ameter near midspike, peduncles to 1 4 mm long,
pubescent like stem, rachis pubescent with
straight to flexuose (to antrorsely appressed, es-
pecially toward base of spike) eglandular and
glandular trichomes to 1 mm long. Bracts green,
drying dark red, obovate to subcirculate to ellip-
tic, often somewhat asymmetrical, 8-1 6 mm long,
4.5-11.5 mm wide, rounded and sometimes
apiculate to acute at apex, sessile, rounded or
constricted and tapering (to subpetiolate) at base,
abaxial surface pubescent with straight to flex-
uose glandular and eglandular trichomes 0.2-0.5
mm long (sometimes not persisting on older
bracts), and usually with coarser, antrorse eglan-
dular trichomes to 0.8 mm long as well, margin
ciliate with flexuose eglandular (usually coarse)
and glandular trichomes up to 2 mm long. Bract-
lets linear-lanceolate to linear to linear-elliptic,
8-13 mm long, 1.5-2.1 mm wide, tapering little,
if at all, and sessile at base, pubescent like bracts.
Calyx 7.5-10 mm long, 5-lobed, lobes divided
nearly to base of calyx, subulate, 0.5-0.8 mm
wide, pubescent with glandular and eglandular
trichomes. Corolla whitish to lavender, 15-19
mm long, externally pubescent with flexuose
eglandular trichomes to 0.6 mm long, tube 9-12
mm long, internally glabrous, upper lip trian-
gular, internally rugulate, 4.5-8 mm long, emar-
ginate with lobes 0.2-0.3 mm long, lower lip 5-
7.5 mm long with 3 subelliptic lobes 2-4.5 mm
long, 1.6-2.8 mm wide, central lobe larger than
lateral lobes. Stamens attached near apex of co-
rolla tube, 5-6.5 mm long, exserted up to 5.5
mm beyond mouth of corolla, filaments 4-5 mm
long, glabrous, thecae superposed, the upper 1 .2-
1.5 mm long, the lower 1 .5-2 mm long (including
a basal spur 0.2-0.3 mm long), pollen triporate
(to subtricolporate) with trema region traversed
by 2 rows of 7 insulae. Style 11-16 mm long,
distally glabrous, proximally pubescent with
eglandular trichomes. Capsule 9-12.5 mm long,
pubescent with glandular and eglandular tri-
chomes 0.1-0.4 mm long, stipe 3.5-6 mm long,
head 5-6.5 mm long. Seeds 4 per capsule, flat-
tened, subcircular in outline, 2.3-2.6 mm long,
2.3-2.5 mm wide, surface papillose (when im-
mature) and papillose to verrucose (when ma-
ture).
DISTRIBUTION AND PHENOLOGY.— Cloud for-
ests of eastern Panama (Panama, San Bias, and
Darien) at elevations from 500 to 1300 m (Fig.
3). Flowering November-January; fruiting: De-
cember-January.
The type and most other collections of this
species have golden, antrorsely-appressed cau-
line trichomes to 0.8 mm long, relatively small
leaves (blades 45 to 90 mm long, 17 to 31 mm
wide with petioles to 1 7 mm long), and relatively
short spikes (to 55 mm long) with small bracts
(8-12 mm long, 4.5-7 mm wide). Four collec-
tions (Dressier 3221, Folsom et al. 6296, Maas
& Dressier 677, and Mori & Kallunki 3424) differ
most conspicuously from those specimens re-
sembling the type by having light brown to straw-
colored, flexuose to antrorse-appressed cauline
trichomes to 2 mm long, larger leaves (55 to 185
mm long, 18-78 mm wide with petioles to 30
mm long), and somewhat longer spikes (to 75
mm long) with larger bracts (11 to 16 mm long,
5-1 1.5 mm wide). Although these two forms are
readily recognizable, the general overlap in the
character states noted above and the occurrence
of both in the same region (i.e., Cerro Jefe) sug-
gests that they be treated as a single taxon at this
time.
Justicia readii also cannot be classified in Gra-
ham's (1988) scheme. It has trilaterally sym-
metric pollen ("Type 2," Fig. 4) which is char-
acteristic of Graham's section Tyloglossa of Africa
and Arabia. Species in this section differ from J.
readii by their inflorescence structure and bract
form. In spite of the obvious difference in pollen,
/. readii is more similar to /. fortunensis and J.
veraguensis than to any other species in Panama.
Durkee (1978) treated this species as /. ad-
hat oda L., a widely cultivated species native from
the Indian subcontinent to southeastern Asia.
Justicia adhatoda, which is not presently known
from Panama, can be distinguished from /. readii
by its longer (23-30 mm) corollas; larger (12-34
mm long and 7-12 mm wide) and eglandular
bracts that do not dry dark red; linear-elliptic,
DANIEL AND WASSHAUSEN: NEW PANAMANIAN JUSTICIA
297
eglandular, and wider (1.5-2.3 mm) calyx lobes;
longer (3.5-4.5 mm) thecae, both of which bear
basal spurs; and bilaterally symmetric pollen.
Justicia readii has been rather well collected
on Cerro Jefe, a mountain northeast of Panama
City that has been much visited since it became
readily accessible to plant collectors in the mid-
1960's (Dwyer 1967, 1985). It is not surprising
that the species was recently collected along the
relatively new El Llano-Carti Road in the Ser-
rania de San Bias. There is a continuous ridge
system linking Cerro Jefe with the Serrania de
San Bias. The southeasternmost station for this
species, in the Serrania de Pirre in Darien, is
rather isolated from the other known occur-
rences.
PARATYPES.— PANAMA. Darien: ridgetoparea
N of Cerro Pirre, between Cerro Pirre top and
Rancho Plastico, /. Folsom et al. 6296 (MO).
Panama: Cerro Jefe, along main rd. before turn-
off to summit, T. Croat 13054 (MO); La Eneida,
region of Cerro Jefe, R. Dressier 3 221 (MO); Cer-
ro Jefe, /. Dwyer & A. Gentry 8489 (MO); near
top of Cerro Jefe to 1 mi. beyond, A. Gentry et
al. 3440 (MO); La Eneida, Cerro Jefe region, P.
Maas & R. Dressier 677 (MO); 5-10 km NE of
Altos de Pacora, "Campamento Quatro," Gor-
gas Memorial Labs yellow fever research camp,
S. Mori & J. Kallunki 3424 (MO, NY); Cerro
Jefe, near jet. of rds. to Cerro Jefe and Altos de
Pacora, S. Mori et al. 3776 (MO); rd. from Cerro
Jefe to Cerro Azul, E. Tyson 5306 (MO); Cerro
Jefe between Cerro Azul and La Eneida, ca. 1 5
mi NE of Panama City, R. Wilbur et al. 15 54 IB
(DUKE, MO). San Bias: Serrania de San Bias,
Nusagandi along El Llano-Carti Rd., R. Read et
al. 85-6B (US).
ACKNOWLEDGMENTS
We are grateful to Frank Almeda for assistance
in the field and for making a special attempt to
recollect plant material, to Gordon McPherson
for assistance in the field and for facilitating the
senior author's stay in Panama, Sue Thompson
for providing a collection of one species, Tsan
and Fei-Mei Chuang for making the chromo-
some count, Alice Tangerini for skillfully pre-
paring the line drawings, and the curators of the
following herbaria for loans or other courtesies:
BM, CAS, DUKE, F, GH, MEXU, MO, NY, U,
US.
RESUMEN
Se describen e ilustran tres especies nuevas de
Justicia de Panama: J.fortunensis, J. veraguensis,
y J. readii. Se presenta un mapa que muestra la
distribucion de cada una de las especies. Las tres
especies nuevas parecen tener afinidades con /.
refractifolia y una clave de estas cuatro especies
es presentada. Se reporta un numero de cro-
mosomas de n = 14 para J.fortunensis. Con 23
especies en el pais, Justicia es el genero mas gran-
de de Acanthaceae en Panama.
LITERATURE CITED
DANIEL, T. F., B. D. PARFITT, AND M. A. BAKER. 1984. Chro-
mosome numbers and their systematic implications in some
North American Acanthaceae. Syst. Bot. 9:346-355.
DANIEL, T. F. AND T. I. CHUANG. 1989. Chromosome num-
bers of some cultivated Acanthaceae. Baileya 23:86-93.
DANIEL, T. F., T. I. CHUANG, AND M. A. BAKER. 1990. Chro-
mosome numbers of American Acanthaceae. Syst. Bot. 15:
13-25.
DWYER, J. D. 1967. A new herbarium in the Canal Zone.
Taxon 16:158-159.
. 1985. The history of plant collecting in Panama. Pp.
179-183 in The Botany and Natural History of Panama: La
Botanica e Historia Natural de Panama. W. D'Arcy and M.
Correa A., eds. Monographs in Systematic Botany from the
Missouri Botanical Garden 10.
DURKEE, L. H. 1978. Acanthaceae. Pp. 155-284, in Flora of
Panama. R. Woodson et al., eds. Ann. Missouri Bot. Gard.
65.
. 1986. Acanthaceae. Pp. 1-92 in Flora Costaricensis.
W. Burger, ed. Fieldiana, Bot. 18.
. 1988. A checklist of Acanthaceae in Costa Rica,
Nicaragua, and Panama. Acanthus 3:3-4.
GRAHAM, V. A. 1988. Delimitation and infra-generic clas-
sification of Justicia (Acanthaceae). Kew Bull. 43:551-624.
LEONARD, E. C. 1958. The Acanthaceae of Colombia, III.
Contr. U.S. Natl. Herb. 31:323-781.
MABBERLEY, D. 1987. The Plant-Book. Cambridge Univ.
Press, Cambridge.
CALIFORNIA ACADEMY OF SCIENCES
Golden Gate Park
San Francisco, California 94 1 1 8
PROCEEDINGS
OF THE
CALIFORNIA ACADEMY OF SCIENCES
Vol. 46, No. 14, pp. 299-326, 12 figs. September 11, 1990
NEW SPECIES AND NEW COMBINATIONS IN BLAKEA AND
TOPOBEA (MELASTOMATACEAE), WITH AN HISTORICAL
PERSPECTIVE ON GENERIC LIMITS IN
THE TRIBE BLAKEEAE
By
Frank Almeda
Department of Botany, California Academy of Sciences, Golden Gate Park,
San Francisco, California 94118
ABSTRACT: Blakea P. Browne and Topobea Aublet, the two currently recognized genera in the Blakeeae have
been variously combined and segregated during their long history. A synoptic taxonomic history of these
genera is presented with special attention given to characters used to combine or separate them. In keeping
with traditional generic concepts, Blakea and Topobea are maintained and a revised generic key is provided
to accommodate anomalous or unusual Mesoamerican species. Descriptions, diagnostic illustrations, discus-
sions, and distribution maps are presented for four new species of Blakea (B, gregii, B. hammelii, and K.
herrerae from Panama; and K. scarlatina from Costa Rica and Nicaragua) and three new species of Topobea
(T. fragrantissima, T. hexandra, and T. suaveolens from Panama). The study of recently collected flowering
material of B. crassifolia and K. parvifolia necessitates their transfer to Topobea and the adoption of a new
name, T. caliginosa, for the narrow Panamanian endemic previously known as B. micrantha.
Received February 2, 1990. Accepted March 14, 1990.
The striking morphological isolation of the
INTRODUCTION „, , . .. r A JT ......
Blakeeae within the family and the homogeneity
The largely pantropical Melastomataceae of its constituent taxa undoubtedly account for
comprise some 200 genera and approximately the fact that its tribal status has not been seriously
4,800 species. This comparatively large natural challenged since it was established by Bentham
family has traditionally been divided into four- and Hooker (1867). However, Blakea P. Browne
teen tribes based on fruit characters, staminal and Topobea Aublet, the two currently recog-
features, and seed morphology. Of the eight tribes nized genera in the tribe, have been variously
restricted to the western hemisphere, only the combined and segregated during their long his-
Blakeeae and Miconieae are berry-fruited. The tory.
Blakeeae are readily recognized by their axillary Recent fieldwork done in connection with the
flowers that are individually subtended by two preparation of a treatment of Melastomataceae
pairs of decussate bracts, ovoid to pyramidal for Flora Mesoamericana has resulted in the dis-
smooth seeds, and wood with multiseriate rays covery of several undescribed species in the
and frequent occurrence of druses (Koek-Noor- Blakeeae. An effort to devise a rational basis for
man et al. 1979; ter Welle and Koek-Noorman assigning these taxa to one genus or the other has
1981). necessitated another look at generic limits in the
[299]
300
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 14
tribe. I recently expressed serious reservations
about the continued recognition of Topobea and
even suggested that it be submerged in the older
and larger genus Blakea (Almeda 1 989). Addi-
tional study of this problem, coupled with the
paucity of information on pollen morphology
(Patel et al. 1 985) and chromosome numbers (Solt
and Wurdack 1980), has led me to refrain from
merging Blakea and Topobea without a compre-
hensive study throughout the extensive neotrop-
ical range of these genera.
The following summary provides a synoptic
taxonomic history of Blakea and Topobea to-
gether with an assessment of the diagnostic fea-
tures used to combine or separate them. This is
followed by descriptions of four new species of
Blakea, three new species of Topobea, and three
generic transfers. Working within the constraints
of a geographically denned study, my decision
to recognize Blakea and Topobea is, of necessity,
tempered by the historical consensus of opinion
among specialists in the family and the need for
more data. Nevertheless, a slight modification of
original generic concepts is necessary to accom-
modate the accretion of new and unusual Central
American species.
SYNOPTIC HISTORY OF
BLAKEA AND TOPOBEA
The genus Blakea was first defined by Patrick
Browne (1756) for a Jamaican species that was
subsequently given the binomial B. trinervia by
Linnaeus (1759). Browne's circumscription of
Blakea included a plate and a brief diagnosis
describing it as a shrub with nitid trinerved el-
liptic leaves, 6-merous axillary flowers, and tri-
angular laterally connate anthers. The plate is
generalized and provides no details of anther
morphology. The terse description is sufficient,
however, to preclude confusion with the allied
genus Topobea, which was described and illus-
trated by Aublet (1775) from material collected
in French Guiana. Although many of Aublet's
published drawings are mixtures and some of his
descriptions are composites of discordant ele-
ments (Howard 1983), both the plate and de-
scription of T. parasitica emphasize the oblong
subulate anthers that are typical of so many
species in this genus. Curiously, Aublet over-
looked or chose not to describe and illustrate the
broad, gaping, dorsally inclined, confluent anther
pores. This is another diagnostic character that
figures prominently in decisions to maintain To-
pobea as a distinct genus. In choosing the epithet
parasitica, Aublet was responsible for perpetrat-
ing a misunderstanding of the epiphytic habit
that is common to so many species of Blakea
and Topobea.
The younger Linnaeus (1781) and de Jussieu
(1789) maintained Blakea and Topobea as de-
fined by Aublet. By retaining Blakea quinque-
nervia Aublet [=Bellucia grossularioides (L.)
Triana] in Blakea, it is clear that their interpre-
tation was a broad one that included at least two
genera now relegated to two different berry-fruit-
ed tribes.
David Don (1823), in an early paper devoted
to the Melastomataceae, was the first to part ranks
with his predecessors by merging Topobea and
Blakea. His decision, however, was based on
characters other than anther morphology. The
weakness of Don's argument is best conveyed by
his own words: "Although the Topobea of Aublet
recedes somewhat from Blakea, in its being par-
asitical; yet, notwithstanding, in the Lambertian
Herbarium are several unpublished species, from
Don Jose Pavon, natives of Peru, and not par-
asitical, which agree with Topobea in every es-
sential point; and these also, accord well with
Blakea, except in having four, instead of six scales,
surrounding the calyx, which, however, is a vari-
able character; and, therefore I think myself jus-
tified in uniting these two genera." De Candolle
(1828) also adopted this interpretation in his
Prodromus. Like Don and his immediate pre-
decessors, de Candolle's broad generic concept
of Blakea included species now recognized in
Belinda.
Naudin (1852) rejected this inclusive dispo-
sition and emphasized diversity in anther mor-
phology and habit as reasons for resurrecting To-
pobea. He provided no involved explanation to
justify his decision, yet he described a new genus,
Pyxidanthus, to accommodate three new species
that he considered closely related to Blakea. The
concept of Blakea and Topobea espoused by
Naudin was faithfully followed by Bentham and
Hooker (1867) without comment.
Triana (1871), as astute specialist in the Me-
lastomataceae, synonymized Pyxidanthus under
Blakea along with Valdesia Ruiz and Pavon,
another minor generic segregate. Pyxidanthus,
Valdesia, and the subsequently described Ama-
raboya Linden are now universally accepted syn-
onyms of Blakea. As recognized by Triana (1 87 1),
ALMEDA: NEW BLAKEA AND TOPOBEA
301
Blakea included those species with short, oval
or elliptic, apically blunt or rounded anthers with
two minute well-separated pores. Topobea in-
cluded those species with linear-oblong or subu-
late anthers with two confluent dorsally inclined
pores. Baillon ( 1879), who was noted for his broad
view of generic limits, was unimpressed by these
differences in anther morphology. He returned
Topobea to the synonymy of Blakea but provid-
ed no convincing reasons aside from a comment
describing the anthers of Topobea as more elon-
gate and narrower than those in Blakea. For a
botanist who excelled in character analysis, Bail-
Ion's brief treatment of these genera reflects a
superficial understanding of the anther differ-
ences between them.
The most comprehensive and perhaps most
influential accounts of generic limits in the Blake-
eae are those of Cogniaux (1888, 1891). Cog-
niaux was undoubtedly influenced by Triana.
Unlike Triana, however, he was the first to pro-
vide the following explicit key clarifying the dis-
tinctions between Blakea and Topobea as he
understood them:
A. Staminum filamenta crassiuscula, anthe-
rae breves, obtusae a latere compressae,
connective crassisimo, postice calcarato ....
Blakea
B. Staminum filamenta filiformia; antherae
lineari vel oblongo-subulatae, rostratae,
connective mediocri, interdum ecalcarato
Topobea
These differences amply characterized the
species then known to Cogniaux, but they cannot
be used for a strict characterization of the two
genera at present. The short, obtuse, laterally
compressed anthers of Blakea (Fig. 2 A, C, E; Fig.
3A, E) versus the linear to oblong-subulate ros-
trate anthers of Topobea (Fig. IE, F; 2B; 3B, C)
are the only modal contrasts that appear to be
valid for distinguishing the majority of species
now assigned to these genera. The differences in
filament thickness and presence or absence of
dorsal appendages on the connective are incon-
sistent criteria for differentiating Blakea and To-
pobea. Like his predecessors, Cogniaux neglected
to emphasize that Blakea has anthers with two
typically minute and well-separated pores and
Topobea has two (rarely one) broad approximate
or confluent dorsally inclined pores. The generic
concepts promoted by Cogniaux have been
adopted by all subsequent specialists (Almeda,
in press; Gleason 1935, 1958; Wurdack 1973,
1980) and other authors of major regional treat-
ments of the Melastomataceae (Standley 1924,
1 938; Standley and Williams 1 963), except Mac-
bride (1941). Invoking a rationale similar to that
of Baillon, Macbride combined Blakea and To-
pobea and commented on the inconsistency of
the filament character used as a diagnostic fea-
ture by Cogniaux.
In a paper describing many new species of
Blakea and Topobea, Gleason (1945) sympa-
thized with, but rejected, Macbride's view. The
close relationship between these two genera is
readily apparent in their many parallel and over-
lapping variations in habit, leaves, indument,
floral bracts, petals, calyx lobes, ovary cell num-
ber, and seeds. Although Baillon (1879) and
Macbride (1941) intimated that anther charac-
ters provide insufficient grounds on which to base
the recognition of genera, Gleason (1945, 1947),
who defended their use, pointed out that speci-
mens can be sorted precisely into one genus or
the other when staminal material is available.
According to Patel (in litt, 18 July 1985), pre-
liminary results of a numerical taxonomic anal-
ysis of pollen (using SEM data) in the Melasto-
mataceae reveal that Blakea and Topobea sort
out into different clusters. Because pollen of Me-
lastomataceae is diverse exomorphically (Patel
et al. 1985), an expanded survey is needed to
extend and refine these observations.
Until more information about micromor-
phological characters such as chromosome num-
bers, pollen morphology, and chemistry is avail-
able, it seems preferrable to adhere to traditional
generic concepts until we can rest decisions on a
broader data base than was available to our pre-
decessors.
This, however, does not ease the current dif-
ficulty of assigning the few unusual species to one
genus or the other. In the Mesoamerican region,
at least four different species groups are problem-
atic in determining where to draw the line be-
tween Blakea and Topobea. I will consider the
relationships among Blakea wilburiana Almeda
and allied species in a future paper. The other
three species groups are mentioned here because
at least one species from each is newly described
below.
The first of these, Blakea gregii, belongs to a
group of five closely related species characterized
by nectar-producing pendant flowers with green
petals. The anthers of B. gregii are elliptic-oblong
302
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 14
FIGURE 1 . Representative stamen morphology in Blakea and Topobea. A, B. wilburiana Almeda; B, B. austin-smithii Standley;
C, B. purpusii Brandegee; D, T. crassifolia (Almeda) Almeda; E, T. pittieri Cogn.; F, T. brenesii Standley. (A from Almeda et
al 6525, CAS; B from Almeda et al. 2145, CAS; C from Breedlove & Almeda 47747, CAS; D from de Nevers 5943, CAS; E
from Almeda et al. 3161, CAS; F from Wilbur et al. 15809, DUKE.)
and laterally compressed with two dorsally in-
clined approximate pores (Fig. 4G). Anther con-
formation is like that of Blakea but the dorsal
inclination of the apical pores is more like that
encountered in Topobea. The pore width, how-
ever, is narrower than that typically found in
many species of the latter genus. In deciding to
place this species in Blakea, I was influenced by
ALMEDA: NEW BLAKEA AND TOPOBEA
303
FIGURE 2. Representative stamen morphology in Blakea and Topobea. A, B. elliptica (Gleason) Almeda; B, T. calycularis
Naudin; C, B. woodsonii Gleason; D, T. storkii Standley; E, B. tuberculata J. D. Smith. (A from Almeda et al. 6261, CAS; B
from Breedlove 50940, CAS; C from Almeda et al. 3046, CAS; D from Almeda et al. 2884, CAS; E from Almeda & Nakai 4529,
CAS.)
the somewhat more typical Blakea-type anthers
of its close relatives, B. austin-smithii Standley
(Fig. IB) and B. purpusii Brandegee (Fig. 1C).
Topobea hexandra has been the most difficult
to place genetically. It is part of an unusual as-
semblage in the Blakeeae distinguished by small
sessile or short-pedicellate flowers with only six
antesepalous stamens. Its close relatives include
304
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 14
C I
FIGURE 3. Representative stamen morphology in Blakea and Topobea. A, B. litoralis L. O. Williams; B, T. parvifolia (Gleason)
Almeda; C, T. albertieae Wurdack; D, B. hirsuta Triana var. rotundata Markgraf; E, B. anomala J. D. Smith. (A from Almeda
& Nakai 4189, CAS; B from Almeda et al. 5898, CAS; C from Almeda et al. 6153, CAS; D from Lugo 4140, CAS; E from
Almeda 2655, CAS.)
T. cordata Gleason, T. caliginosa Almeda, and
T. crassifolia (Almeda) Almeda. Stamens of T.
cordata were unknown to Gleason (1950) when
he proposed it as a new species. His decision to
place it in Topobea was based on its superficial
resemblance to some sessile-leaved species of that
genus. Although each of the species in this com-
plex is known from very few flowering collec-
tions, available anther material exhibits inter-
esting variations on the Topobea theme. With
the exception of T. hexandra, all have anthers
with two broad confluent pores. In T. cordata,
the anthers are oblong-subulate (5-5.5 mm long)
and the confluent pores are somewhat ventrally
inclined. In T. caliginosa and T. crassifolia the
anthers are linear-oblong but shorter (2-2.5 mm)
ALMEDA: NEW BLAKEA AND TOPOBEA
305
with truncate to somewhat dorsally inclined pores
(Fig. 1 D). The anthers of T. hexandra are also
short (2 mm) and oblong, but they are laterally
compressed and the two apical pores, although
broad, are ventrally inclined and separated from
one another (Fig. 1 1 F). In the context of varia-
tion exhibited by its close relatives, the sum-
mation of anther characters in this species argues
for placement in Topobea. Without these in-
sights, one might be inclined to assign it to Blak-
ea.
The last of the problematic species to be con-
sidered here is T. suaveolens Almeda. This species
is unusual in having fragrant pendant flowers
with brightly colored petals and connivent an-
thers that form a ring around the exserted style
(Fig. 1 2C). The essentially elliptic shape and con-
spicuous lateral compression of the anthers (Fig.
12E) are reminiscent of Blakea-lype anthers but
the broad solitary dorsally inclined apical pore
is a feature known only in Topobea. Again, my
generic placement of this species is influenced by
its close relationship to T. fragrantissima Al-
meda (Fig. 10), a species with androecial char-
acters that can only be attributed to Topobea.
Interestingly, the anomalous androecial mor-
phology in the cases discussed above all appear
to represent derived syndromes. Each is associ-
ated with a unique pollen presentation mecha-
nism (T. suaveolens), a specialized pollinator
spectrum (see discussion under B. gregii), or great
reduction in flower size (T. hexandra). Blakea,
with about 1 00 described species, and Topobea,
with over 60 species, are large and diverse, closely
related genera. If more unusual anther forms come
to light, it may become increasingly difficult to
separate them without establishing arbitrary lim-
its or fueling counterarguments to submerge To-
pobea. For floristic purposes, I find it expedient
to recognize both genera. This is done for con-
venience and to preclude extensive nomencla-
tural changes pending a comprehensive study of
the tribe. It is not done out of a firm conviction
that generic status is necessarily the best way to
reflect relationships based on currently perceived
differences. Accordingly, the following revised
key to these genera is offered in an attempt to
incorporate the most obvious Mesoamerican
anomalies discovered to date:
Stamens 1 2 in number; anthers 2-pored, oval,
oblong, or elliptic, compressed laterally,
bluntly obtuse or broadly rounded at the
summit with two typically well-separated
(often minute) apical pores Blakea
Stamens 6, 8, or 12 in number; anthers 1 -pored
or 2-pored, linear-oblong to oblong-subu-
late or rostrate, usually not compressed lat-
erally (if conspicuously compressed then
1 -pored or dorso-basally appendiculate) with
approximate or confluent dorsally inclined
pores (if pores are prominently inclined ven-
trally then flowers are hexandrous or pedi-
cels are beset with spreading brown hairs 1-
2.5 mm long) Topobea
NEW TAXA AND NEW COMBINATIONS
Blakea gregii Almeda, sp. nov.
(Fig. 4)
TYPE: — PANAMA. Chiriqui: Cerro Pate Macho, windswept
ridge, 8°49'N, 82°24'W, elev. 2,100 m, 17 Jan. 1986, deNevers
& McPherson 6840 (holotype: CAS! isotypes: MEXU!, MO!,
PMA!, US!).
Frutex ca. 2—4 m altus. Ramuli primum quadrangulati de-
mum teretes sicut petioli laminarum subtus pedicelli bracte-
aeque pilis 1-2.5 mm longis induti. Petioli 1.8-3.5 cm longi;
lamina 9.3-14.5 x 5-9 cm elliptica vel elliptico-ovata apice
acuminata basi obtusa vel rotundata vel late acuta, 5-nervata
vel 5-plinervata, coriacea et integra, nervis secundariis 3-6 mm
inter se distantibus. Flores 6-meri penduli in quoque nodo
superiore plerumque 1-4, pedicellis 1.2-2 cm longis; bracteae
exteriores 7-8 x 5-8 mm ellipticae vel ovato-ellipticae ad
basim paulo (2 mm) coalitae; bracteae interiores omnino libe-
rae 5-6 x 4-7 mm ovatae vel suborbiculares apice truncato-
rotundato. Hypanthium (ad torum) 6 mm longum extus gla-
brum vel sparse subamorpho-furfuraceum; calycis tubus 2-3
mm longus, lobis 2-4 mm longis extus sparsiuscule strigosis.
Petala 13-14 x 9-11 mm obovata apice rotundato sparse ca-
duceque glanduloso-ciliolata. Filamenta 6-7 mm longa; an-
therae 4-4.5 x 2 cm subulatae lateraliter non cohaerentes,
dorsaliter biporosae; connectivum dorsaliter ca. 0.5 mm supra
thecarum basim inconspicue tuberculatum. Stylus 1.2-1.3 cm;
ovarium 6-loculare et Vi inferum apice glabro (cono et collo
non evoluto).
Terrestrial shrubs 2-4 m tall. Young inter-
nodes quadrate, becoming rounded with age.
Distal branchlets, vegetative buds, petioles, and
pedicels densely covered with a brown indument
of appressed, ± curved subulate hairs mostly 1-
2.5 mm long. Mature leaves of a pair essentially
equal in size; petioles 1.8-3.5 cm long; blades
coriaceous, 9.3-14.5 cm long and 5-9 cm wide,
elliptic to elliptic-ovate, apex acuminate, base
obtuse to rounded but sometimes varying to
broadly acute, margin entire and somewhat revo-
lute when dry, 5 -nerved or 5-plinerved, the in-
nermost pair of primaries diverging from the me-
dian nerve 4-6 mm above the blade base, the
306
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 14
2.6cm
B
FIGURE 4. Blakea gregii Almeda. A, habit; B, representative leaf with enlargement (abaxial surface); C, representative flower
(natural posture) with petals and stamens removed; D, outer floral bract (abaxial surface); E, inner floral bract (abaxial surface);
F; petals (adaxial surface); G, stamens, lateral view (left), ventral view (middle), dorsal view (right). (A from Almeda et al. 6137,
CAS; B-G from the holotype.)
ALMEDA: NEW BLAKEA AND TOPOBEA
307
elevated network of transverse secondaries spaced
3-6 mm apart at the widest portion of the blade,
the young unexpanded leaves beset with a de-
ciduous indument of white wooly hairs but typ-
ically becoming glabrous adaxially at maturity
and moderately to copiously covered abaxially
with smooth appressed to ± curved brown hairs
1.5-2.5 mm long. Flowers pendant, solitary,
paired, or in fascicles of 4 in the leaf axils of
distal branches; pedicels 1.2-2 cm long. Floral
bracts thick and semisucculent, sessile and en-
tire; outer bracts connate basally for 2 mm, 7-8
x 5-8 mm, elliptic to elliptic-ovate, apex acute,
densely covered abaxially with appressed to in-
curved smooth brown hairs; inner bracts free to
the base, 5-6 x 4-7 mm, ovate to subrotund,
apex rounded to truncate varying to retuse with
a bluntly acute median lobe, sparingly to mod-
erately beset with appressed brown hairs that are
typically concentrated around or restricted to the
median apex. Hypanthia (at anthesis) campan-
ulate to suburceolate, 6 mm long to the torus and
7-8 mm in diameter, glabrous or sparingly brown
furfuraceous toward the base. Calyx tube 2-3
mm long; calyx lobes erect, 2-4 mm long and 4-
5 mm wide basally, ovate to deltoid-ovate with
a blunt callose-thickened tooth at the abaxial apex
of each lobe, roughened along interlobe sinuses,
adaxially glabrous, abaxially covered with a sparse
indument of tardily deciduous short incurved
brown hairs that are restricted to or largely con-
centrated near the apex. Petals 6, glabrous, con-
nivent to somewhat imbricate when fully ex-
panded, 13-14 x 9-11 mm, green, obovate,
apically rounded, entire but irregularly glandu-
lar-ciliate. Stamens 1 2, isomorphic, free and en-
circling the exserted style; filaments complanate
and glabrous, 6-7 mm long; anthers 4-4.5 mm
long, 2 mm wide, elliptic-oblong, laterally com-
pressed, each anther tipped with two dorsally
inclined approximate pores; connective dilated
dorso-basally ca. 0.5 mm above the filament in-
sertion into a blunt callose spur. Ovary Vi infe-
rior, 6-celled, glabrous and lobulate at the apex
but not expanded into a cone or collar. Style
straight, glabrous, 1.2-1.3 cm long, conspicu-
ously exserted beyond the petals and calyx lobes;
stigma truncate. Berry globose, about 6 mm long
and 8 mm in diameter. Seeds clavate to narrowly
pyriform, ca. 1 mm long, beige with a smooth
glossy testa and a conspicuous lateral raphe.
PHENOLOGY.— Flowering specimens have been
collected in January; the only fruiting specimen
was collected in March.
DISTRIBUTION.— A little-collected cloud forest
or elfin forest species known only from Chiriqui
Province, Panama in the region extending from
Cerro Horqueta to Cerro Pate Macho at 1,800-
2,200 m (Fig. 5).
ADDITIONAL SPECIMENS EXAMINED.— PANAMA. Chiriqui:
S slopes of Cerro Horqueta, 21 Jan. 1971, Wilbur et al. 13486
(CAS, DUKE, MO); Cerro Pate Macho, 8°49'N, 82°24'W, 13
Mar. 1988, Almeda et al. 6137 (CAS, CR, MO, NY, PMA,
TEX).
Blakea gregii is closely related to four other
species characterized by cryptic, pendant flowers
with green, connivent petals and nocturnal or
crepuscular nectar production (Almeda 1980,
198 la). The green-flowered species are typically
cloud forest hemiepiphytes that germinate ter-
restrially, ascend nearby trees by adventitious
roots, and ultimately become epiphytic by losing
root contact with the soil. Both B. chlorantha
and B. gregii, however, are often conspicuous
terrestrial shrubs on windswept ridges of the con-
tinental divide where the vegetation is common-
ly reduced to elfin woodland.
Intensive field studies by Lumer (1981) and
Lumer and Schoer (1986) have shown that B.
austin-smithii, B. chlorantha, and B. penduliflora
are visited and pollinated by six species of ro-
dents belonging to four genera: Oryzomys, Pero-
myscus, Reithrodontomys, and Scotinomys. Lu-
mer and Schoer (1986) suggest that rodents may
indeed play a significant role as pollinators of
flowering plants in the harsh windy cloud forest
environment. Nothing is known about the pol-
lination of B. gregii, but similarities in floral mor-
phology and habitat suggest that small rodents
may also play an important role in the repro-
ductive biology of this species.
To facilitate comparison, diagnostic characters
of the five species of Blakea with pendant green
flowers are summarized in the following key:
1 . Leaf blades modified adaxially at the pet-
iole-laminar junction into flap-like saccate
pouches, the basal and lateral margins of
which are free from but conspicuously de-
current on the petiole.
2. Pubescence of uppermost internodes
consisting of hairs 1-2 mm long; leaf
blades bluntly denticulate and revolute
on drying; outer floral bracts linear-Ian-
308
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 14
A B. gregii
• B. scarlatina
• T. fragrantissima
-ioc
100 200 km
FIGURE 5. Distributions of Blakea gregii, B. scarlatina, and Topobea fragrantissima.
ceolate, 1 5-2 1 mm long, equaling or ex-
ceeding the calyx lobes at anthesis; Cos-
ta Rica (Cerro Chompipe, Volcan Barva,
and Volcan Irazu)
_ B. austin-smithii Standley
2. Pubescence of uppermost internodes
consisting of hairs mostly less than 0.5
mm long; leaf blades entire and revolute
on drying; outer floral bracts ovate to
elliptic-ovate, 5.5-9 mm long, conspic-
uously shorter than the calyx lobes at
anthesis; Costa Rica (Cordillera de Ti-
laran) B. chlorantha Almeda
Leaf blades not modified adaxially at the
petiole-laminar junction into flap-like sac-
cate pouches.
3. Older petioles, pedicels, and floral bracts
essentially glabrous; outer floral bracts
13-17 mm wide on fruiting hypanthia;
Costa Rica (Cordillera de Talamanca
and Volcan Barva)
B. penduliflora Almeda
3. Older petioles, pedicels, and floral bracts
moderately to copiously covered with
appressed or spreading brown hairs;
outer floral bracts 5-9 mm wide on
fruiting hypanthia.
4. Uppermost internodes and elevated
primary nerves on lower leaf sur-
faces moderately to densely covered
with spreading brown hairs mostly
less than 0.5 mm long; pedicels (20-)
28-38 mm long; outer floral bracts
13-20 mm long; Mexico (Chiapas)
and western Guatemala (Quezalten-
ango and San Marcos) _
B. purpusii Brandegee
4. Uppermost internodes and elevated
primary nerves on lower leaf sur-
faces densely covered with ap-
pressed brown hairs 1-2.5 mm long;
pedicels 1 2-20 mm long; outer floral
bracts 7-8 mm long; western Pana-
ma (Chiriqui) B. gregii Almeda
Among the green-flowered species, B. gregii is
probably most closely related to B. chlorantha.
They are separable even in sterile condition be-
cause B. gregii lacks the flap-like saccate pouches
at the petiole-laminar junction that are typical
of B. chlorantha. The latter also differs consis-
tently in a number of other characters. Pubes-
cence of the upper cauline internodes consists of
± flattened subulate hairs mostly less than 0.5
mm long and the floral bracts are beset with a
mixture of spreading shaggy hairs (these mostly
ALMEDA: NEW BLAKEA AND TOPOBEA
309
less than 0.5 mm long) and appressed stellate
hairs. In addition, the pedicels of B. chlorantha
are shorter (0.4-1 cm), the calyx lobes are tri-
angular instead of bluntly rounded, and the pet-
als are about half the size (6-8.5 x 5-6 mm) of
those in B. gregii.
This species is named for Greg de Nevers
(195 5-), a student of Arecaceae, who collected
the type and many other important additions to
the melastome flora of Mesoamerica in the course
of his extended field work in Panama from 1984
to 1986.
Blakea hammelii Almeda, sp. nov.
(Fig. 6)
TYPE.— PANAMA. Chiriqui: 3.5 mi. NE of Boquete, end of
rd. on slope S of Rio Palo Alto, 17 Nov. 1978, Hammel 5688
(holotype: CAS!; isotype: MO).
Frutex epiphyticus vel terrestris. Petioli 0.7-1.7 cm longi;
lamina 3.5-7.5 x 1.6-4.5 cm elliptica vel elliptico-ovata apice
caudata vel cuspidata basi acuta ad maturitatem subcoriacea
et glabra, 5-nervata, nervi in axillis acarodomatiis instructi,
nervis secundariis 0.25 mm inter se distantibus. Flores 6-meri
in quoque nodo superiore singuli vel bini, pedicellis 2.6-3.5
cm longis, bracteae omnino liberae; bracteae exteriores 3.5-5
x 2-3 mm ovatae vel ovato-ellipticae extus sparse subamor-
pho-furfuraceae demum glabratae; bracteae interiores 3-4 x
3-3.5 mm oblongo-ovatae extus sparse inconspicueque fim-
briolato-ciliolatae. Hypanthium (ad torum) 3-4 mm longum
extus glabrum; calycis tubus 1-1.5 mm longus, lobis 1-1.5 x
2.5-3. 5 mm. Petala glabra 1.4-1.5 x 0.8-0. 9 cm obovata apice
rotundato. Filamenta 4.5-5.5 mm longa; antherae 3.5-4 x 1-
1.5 mm oblongae lateraliter non cohaerentes apicaliter bipo-
rosae; connectivum dorsaliter ca. 1-1.5 mm supra thecarum
basim tuberculatum. Stylus 10-1 1 mm; ovarium 5-6-loculare
et % inferum, cono glabro (collo non evoluto).
Epiphytic or terrestrial shrubs or small trees
2.5-6 m tall. Young vegetative buds deciduously
furfuraceous, otherwise glabrous, the distal
branchlets glabrous and bluntly quadrate with
thickened interpetiolar lines or ridges. Mature
leaves of a pair equal to unequal in size, glabrous
throughout; petioles 0.7-1 .7 cm long; blades sub-
coriaceous, 3.5-7.5 cm long and 1.6-4.5 cm wide,
elliptic to elliptic-ovate, apex caudate to cuspi-
date, base acute, margin entire, 5 -nerved abax-
ially with well-developed, irregularly ruptured
domatia formed in the angles between the me-
dian vein and each of the two proximal lateral
veins, the striolate transverse secondary veins
spaced mostly less than 0.25 mm apart at the
widest portion of the blade. Flowers erect, soli-
tary or paired in leaf axils of distal branches;
pedicels 2.6-3.5 cm long, glabrous. Floral bracts
sessile, entire, and free to the base; outer bracts
3.5-5 x 2-3 mm, ovate to ovate-lanceolate or
elliptic-ovate, apex bluntly acute to rounded,
sparingly and deciduously brown furfuraceous-
puberulent; inner bracts 3-4 x 3-3.5 mm, ob-
long-ovate, apex rounded, essentially glabrous
but deciduously fimbriolate-puberulent on the
margins. Hypanthia (at anthesis) campanulate,
3-4 mm long to the torus and 6-7 mm in di-
ameter, glabrous. Calyx tube 1-1.5 mm long; ca-
lyx lobes 1-1.5 mm long and 2.5-3.5 mm wide
basally, broadly ovate with a blunt callose-thick-
ened tooth at the abaxial apex of each lobe, mar-
gin entire, glabrous on both surfaces. Petals 6,
glabrous, 1.4-1.5 x 0.8-0.9 cm, white flushed
with pink unilaterally on the abaxial surface, ob-
ovate, apically rounded, entire. Stamens 1 2, iso-
morphic, free and declined to one side of the
flower opposing the style; filaments complanate
and glabrous, 4.5-5.5 mm long; anthers 3.5—4
mm long, 1-1.5 mm wide, yellow, linear-oblong
in dorsal and ventral view, narrowly ovoid in
profile view, truncate at the apex with two some-
what dorsally inclined pores; connective slightly
thickened dorsally and dilated into a blunt spur
ca. 1-1.5 mm above the filament insertion. Ovary
% inferior, 5-6-celled, glabrous at the summit
which is distended into a prominent cone that
becomes rounded with age. Style declinate and
somewhat sigmoid, glabrous, 10-11 mm long;
stigma truncate. Berry globose, 5-6.5 mm long
and 4-7 mm in diameter. Seeds irregularly pryi-
form to pyramidate, 0.5-1 mm long, beige with
a smooth testa and conspicuous lateral raphe.
PHENOLOGY.— Flowering specimens have been
collected from January through April and in No-
vember; the only known fruiting specimens were
collected in January and April.
DISTRIBUTION. — Local and uncommon in cloud
forests on the slopes of Cerro Pate Macho and
along the Rio Palo Alto in Chiriqui Province,
Panama at 1,600-1,900 m (Fig. 8).
ADDITIONAL SPECIMENS EXAMINED. — PANAMA. Chiriqui:
S slopes of Cerro Pate Macho, along trail E of Jaramillo Arriba,
8°49'N, 82-23' W, 12 Apr. 1984, Churchill & Kuijt 5103 (CAS);
trail to Cerro Pate Macho, 8°49'N, 82°24'W, 17 Jan. 1986, de
Nevers & McPherson 6830 (CAS); back side of mountain, Bo-
quete, 20 Mar. 1977, Folsom 2185 (CAS); along trail between
N fork of Rio Palo Alto and Cerro Pate Macho, ca. 6 km NE
of Boquete, 8°48'N, 82°23.5'W, 6 Feb. 1986, Grayum et al.
6371 (CAS); road along Rio Palo Alto ca. 3 km NE of Boquete
to the end, 15 Apr. 1982, Huft 1855 (CAS); Rio Palo Alto, 28
Apr. 1983, Schmalzel 1478 (CAS).
310
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 14
FIGURE 6. Blakea hammelii Almeda. A, habit; B, representative leaf with enlargement showing domatium (abaxial surface);
C, representative flower (natural posture); D, petal (adaxial surface); E, stamens, lateral view (left), dorsal view (middle), ventral
view (right); F, berry with persistent decussate bracts; G, seeds. (A, D, E from Churchill & Kuijt 5103; B from Huft 1855; C
from de Nevers & McPherson 6830; F, G from Schmalzel 1478.)
ALMEDA: NEW BLAKEA AND TOPOBEA
311
Because of its elliptic to elliptic-ovate leaves
with finely striolate secondary veins, erect long-
pedicellate flowers, and white petals that are
flushed with pink abaxially, B. hammelii is readi-
ly mistaken for B. pauciflora Gleason, to which
it seems most closely related. The two differ pri-
marily in androecial details and the nature of the
calyx. In B. pauciflora the anthers, which are
fused laterally for most of their length, are conic
to rounded apically with two minute subterminal
pores positioned on the ventral side of the anther
sacs. Many specimens of B. pauciflora and B.
hammelii are collected only in fruiting condition,
but the flaring, essentially truncate, calyx of the
former provides a consistent character by which
the two species can be distinguished.
Although the domatia of B. hammelii and B.
pauciflora are superficially similar to the pocket
domatia ofClidemia hammelii Almeda (Almeda
1 989), they appear to be modified versions of the
pit or marsupiform (pocket-shaped) domatia de-
scribed and illustrated by Jacobs (1 966) and Stace
(1965), respectively. The term domatium was
coined by Lundstrom (1887) when he hypothe-
sized a mutualistic relationship between mites
and plants bearing these structures. Because of
insufficient field study and the lack of experi-
mental data, the functional significance of foliar
domatia has remained conjectural for over a cen-
tury. Based in part on Lundstrom's seminal pa-
per, Pemberton and Turner (1989) recently pre-
sented new data to "support the hypothesis of a
widespread facultative mutualism in which leaf
domatia serve as shelters and nurseries for ben-
eficial mites, which in turn reduce the number
of phytophagous arthropods and pathogens using
the plants." Interestingly, Pemberton and Turner
found three genera of predaceous mites in 32%
of the domatia of an unidentified Panamanian
species ofBlakea examined specifically for their
study.
This species is named for Barry E. Hammel
(1946-), avid student of the Mesoamerican flora,
who collected the type of this and several other
recently described species of Melastomataceae.
Blakea herrerae Almeda, sp. nov.
(Fig. 7)
TYPE. — PANAMA. Comarca de San Bias: El Llano-Carti
road at about km 1 9. Ina Igar trail in the vicinity of Nusagandi,
elev. 350 m, 1 Feb. 1989, Almeda et al. 6507 (holotype: CAS!;
isotypes: AAU!, BM!, BR!, CR!, DUKE!, F!, G!, MA!, MEXU!,
MICH!, MO!, NY!, P!, PMA!, TEX!, US!, WIS!).
Frutex epiphyticus. Ramuli subquadrangulares demum te-
retes primum sicut folia novella obscure furfuracei mox glabrati .
Petioli 1-2 cm longi; lamina 6.5-1 1 x 3.2-5.8 cm clliptica vel
elliptico-obovata apice caudata vel cuspidata basi acuta,
3-nervata vel 3-plinervata, subcoriacea et integra, nervis se-
cundariis 0.25 mm inter se distantibus. Flores 6-meri in quo-
que nodo superiore l-2(-3), pedicellisO.5-1.2 cm longis, brac-
teae omnino liberae; bracteae exteriores 2.5-5 x 1.5-2.5 mm
oblongo-lanceolatae acutae primum extus sparse vel modice
subamorpho-furfuraceae demum glabratae; bracteae interiores
2-4 x 2-3 mm ellipticae vel ovato-ellipticae apice late acuto
extus glabrae. Hypanthium (ad torum) 3-4 mm longum extus
sparsiuscule caduceque subamorpho-furfuraceum; calycis tu-
bus 1.5-2.5 mm longus, lobis 1-2 mm longis. Petala 6-7 x 3-
5 mm obovato-oblonga apice rotundato. Filamenta 3-4 mm
longa; antherae 2.5-3 x 0.75-1 mm oblongae inter se late-
raliter cohaerentes subapicaliter minute biporosae; connecti-
vum nee prolongatum nee appendiculatum. Stylus 6-7 mm;
ovarium 6-loculare et omnino inferum apice glabro (cono et
collo non evoluto).
Epiphytic shrubs 2-4 m tall. Distal branchlets
subquadrate, becoming rounded with age. Veg-
etative buds and very young leaves deciduously
scurfy-puberulent. Mature leaves of a pair equal
to somewhat unequal in size, glabrous through-
out; petioles 1-2 cm long; blades subcoriaceous,
6.5-1 1 cm long and 3.2-5.8 cm wide, elliptic to
elliptic-obovate, apex caudate to cuspidate or
rarely varying to acuminate, base acute, margin
entire, 3-nerved or 3-plinerved with an addi-
tional submarginal pair of inconspicuous veins,
the striolate transverse secondary veins spaced
mostly 0.25 mm apart at the widest portion of
the blade. Flowers erect, solitary or paired, rarely
in fascicles of three in each leaf axil of the distal
branches; pedicels 0.5-1.2 cm long, sparingly
scurfy-puberulent to glabrous. Floral bracts ses-
sile, entire, and free to the base; outer bracts 2.5-
5 x 1.5-2.5 mm, oblong-lanceolate, apex acute,
sparingly to moderately scurfy-puberulent to gla-
brate; inner bracts 2—4 x 2-3 mm, elliptic to
elliptic-ovate, apex acute, essentially glabrous but
fimbriolate-puberulent on the margins. Hypan-
thia (at anthesis) campanulate, 3—4 mm long to
the torus and 3-4 mm in diameter, sparingly and
deciduously scurfy-puberulent. Calyx tube 1.5-
2.5 mm long; calyx lobes 1-2 mm long and 2.5-
3 mm wide basally, triangular to triangular-ovate,
margin entire or deciduously ciliate, glabrous on
both surfaces. Petals 6, glabrous, 6-7 x 3-5 mm,
pale greenish-white, oblong-obovate, apically
rounded, entire. Stamens 12, isomorphic; fila-
ments 3-4 mm long, complanate, glabrous and
declined to one side of the flower opposing the
style; anthers 2.5-3 mm long, 0.75-1 mm wide,
yellow, laterally connate for their entire length,
312
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 14
B
FIGURE 7. Blakea herrerae Almeda. A, habit; B, C, representative leaves (abaxial surface) showing variation in shape and
enlargement (Code population) showing pubescence details; D, representative flower and bud (natural posture) with enlargement
showing hypanthial pubescence; E, petal (adaxial surface); F, berry with persistent decussate bracts; G, seeds. (A-E from the
holotype; F, G from Sytsma et al. 2603.)
ALMEDA: NEW BLAKEA AND TOPOBEA
313
A B. hammelii
• B. herrerae
• T. hexandra
T. suaveolens
0 100 200 km
-10°
FIGURE 8. Distributions of Blakea hammelii, B. herrerae, Topobea hexandra, and T. suaveolens.
linear-oblong, with two minute pores positioned
ventrally just below the truncate apex; connec-
tive slightly thickened dorsally but not prolonged
or appendaged at the filament insertion. Ovary
completely inferior, 6-celled, glabrous at the
summit surrounding the stylar scar but not dis-
tended into a prominent cone or collar. Style
somewhat declinate and slightly incurved dis-
tally, glabrous, 6-7 mm long; stigma truncate.
Berry globose and pink at maturity, 9-10 mm
long and 10-12 mm in diameter. Seeds clavate
to narrowly pyriform or pyramidate, 1-1.5 mm
long, beige with a smooth testa and conspicuous
lateral raphe.
PHENOLOGY.— Flowering and fruiting occur
sporadically throughout the year.
DISTRIBUTION.— Low rainforests from the Ca-
ribbean slope of central Panama (Code) east to
the Llano-Carti road in the Nusagandi region
(Comarca de San Bias) at 100-400 m (Fig. 8).
ADDITIONAL SPECIMENS EXAMINED. — PANAMA. Code: be-
tween La Junta and Limon, 5 hr. walk N of Alto Calvario, 1 1
Oct. 1977, Folsom 5878 (CAS); trail between Rio Blanco and
Continental Divide N of El Cope and El Petroso sawmill,
8°38'N, 80°36'W, 14 Dec. 1980, Sytsma et al. 2603 (CAS).
Comarca de San Bias: Llano-Carti road, km 1 6, trail to creek
on Atlantic drainage, 2 Feb. 1989, Almeda et al. 6523 (CAS,
PMA); Nusagandi, Llano-Carti road, 28 Jul. 1984, de Nevers
& Todzia 3540 (CAS); along Llano-Carti road, W of road 1 3.8
km to 15.8 km from Interamerican Hwy., 9°19'N, 78°55'W,
24 Aug. 1984, de Nevers et al. 3746 (CAS); Llano-Carti road,
13.8 km to 19 km from Interamerican Hwy., 9°19'N, 78°55'W,
3 Sep. 1 984, de Nevers & Porras 3822 (CAS); Llano-Carti road,
km 19.1, 9°19'N, 78°56'W, 9 Nov. 1984, de Nevers & Herrera
4242 (CAS); Llano-Carti road, km 19.1, 9°19'N, 78°55'W, 1 1
Mar. 1985, de Nevers & Herrera 5100 (CAS). Comarca de San
Bias/Panama Border: on Kuna divide trail N of Llano-Carti
road, 9°20'N, 79°00'W, 16 Dec. 1987 ', McPherson 11 879 (MO).
Panama: Llano-Carti road, 9°16'N, 78°58'W, 12 Sep. 1980,
Sytsma 1077 (CAS); headwaters of Rio Chagres, Rio Esperan-
za, and Rio Piedras, 9°20'N, 79°20' W, 1 7 Oct. 1 984, de Nevers
et al. 4088 (CAS).
First collected in 1977, Blakea herrerae, is
another species that has been confused with, and
is probably derived from, B. paudflora. Both
species have finely striolate secondary foliar ve-
nation and laterally fused anthers that are de-
clined to one side of the flower to form a semi-
circle. Blakea paudflora differs from B. herrerae
in a number of floral characters. The pedicels of
B. paudflora are longer (2.5-5.3 cm), the calyx
is essentially truncate and flaring, the style is
longer (9-12 mm) and flexuous, and the larger
petals (13- 19 x 8- 10 mm) are white flushed with
pink instead of pale greenish- white. Blakea her-
rerae is also distinctive in that some mature leaves
on all collections examined have cryptic tufts of
short stiff hairs on the lower surface in the angles
between the median vein and each of the two
314
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 14
proximal lateral veins. These may well function
as acarodomatia, like the concentrated foliar hair
tufts now known to harbor beneficial mites in
many plant groups (Pemberton and Turner 1 989).
There is some variation in the size and shape
(Fig. 7B, C) of mature leaves, but none of this
can be correlated with geography, elevation, or
other diagnostic characters. Despite the limited
range of this species, there is some variation that
appears to be geographically correlated. The two
cited collections from Code Province west of the
Canal Area (Fig. 8) differ from populations east
of the Canal in having elevated primary leaf veins
below that are sparsely beset with a mixture of
scurfy hairs and spreading glandular hairs. The
lower leaf surfaces of these two collections are
also unusual in having a liberal scattering of mi-
nute red glands on and between the striolate sec-
ondary veins. More material from Code and oth-
er intervening areas is needed to better assess the
significance of these differences in pubescence.
A population from Cerro Jefe, Panama (rep-
resented by D'Arcy 12181, McPherson 7435, and
Sytsma 1414, all CAS), is similar to B. herrerae
in foliar venation, details of the floral bracts, and
anther morphology. These collections differ from
other cited material in having thicker leaves,
larger petals (12-15 x 8-9 mm) that are pink,
and longer styles (1 1-12 mm long). Because plants
of this population differ from typical specimens
of Blakea herrerae in characters that are com-
monly diagnostic, they are excluded from my
circumscription of this species pending addi-
tional study.
Field study of this species in the Nusagandi
region revealed an unusual method of vegetative
reproduction that to my knowledge has never
been reported for an epiphytic member of the
Melastomataceae. Blakea herrerae is commonly
a massive rainforest epiphyte with lax spreading
branches that skirt host trees. In two individuals
examined, the outermost branches were attached
to adjacent trees by clinging adventitious roots.
One individual had colonized four neighboring
host trees in this fashion.
Blakea herrerae also appears to be unique
among the erect-flowered species of the genus in
having greenish-white flowers that produce a
pleasant spicy fragrance but lack nectar. The fra-
grance is reminiscent of that produced by neo-
tropical orchids that are pollinated by male eu-
glossine bees. The few species of Blakea and
Topobea with showy white and/or pink flowers
that have been studied in the field are buzz-pol-
linated by many species of bees (including two
euglossine genera) seeking pollen as a floral re-
ward (Lumer 1981; Renner 1989). Several eu-
glossine genera are among the common buzz-
pollinating bees in the neotropics (Buchmann
1983; Roubik 1989). However, only female eu-
glossines are known to collect pollen from neo-
tropical Melastomataceae (Renner 1 989). In the
male euglossine pollination syndrome of the Or-
chidaceae, the flowers produce no nectar and the
pollen is not available as a food source. It will
be interesting to learn if euglossine bees buzz-
pollinate the poricidal anthers of B. herrerae and
if its floral fragrance serves only as a secondary
attractant, as suggested by Renner (1989) for
melastomes in general. In addition to the Or-
chidaceae, floral fragrance collection by male
euglossines has been found in some species of
Anthurium and Spathiphyllum (both in the Ara-
ceae), one species of Gloxinia (Gesneriaceae),
Cyphomandra (Solanaceae), Dalechampia (Eu-
phorbiaceae), and a scattering of species in other
families of flowering plants (Ackerman 1986;
Williams 1983). Because pollen presentation in
Cyphomandra provides a parallel to the porici-
dally dehiscent anthers of Blakea herrerae, ad-
ditional fieldwork is needed to document details
of its pollination biology.
This species is named for Heraclio Herrera
(195 9-), a student of the Panamanian flora who
kindly directed me to a flowering population of
this species at Nusagandi, Panama.
Blakea scarlatina Almeda, sp. nov.
(Fig. 9)
TYPE.— COSTA RICA. Alajuela: 20-30 km SE of Cataratas
de San Ramon, elev. 2,150-2,500 ft. (655-762 m), 20 Mar.
1978, Almeda et al. 4308 (holotype: CAS!).
Frutex epiphyticus ca. 4 m altus. Ramuli primum subquad-
rangulati demum teretes sicut folia novella pedicelli hypanthia-
que sparse vel densi furfuracei demum glabrati. Petioli 1.7-
4.5(-6.3) cm longi; lamina 8.5-19.7 x 5.4-10 cm elliptica vel
elliptico-ovata apice acuminata basi acuta vel obtusa vel ro-
tundata, 3-nervata, coriacea et integra, nervis secundariis 1-2
mm inter se d istanti bus. Flores 6-meri in quoque nodo superio-
re singuli, pedicellis 1-1.8 cm longis; bracteae exteriores 2-2.7
x 2.2-3.6 cm late ovatae vel suborbiculares ca. 1-1.5 cm coa-
litae apice truncato-rotundato plerumque mucronato; bracteae
interiores 1-2 cm omnino coalitae. Hypanthium (ad torum)
1-1.3 cm longum; calycis tubus 5-6 mm longus, lobis 5-10
mm longis. Petala 3-4.5 x 2.5-4.3 cm obovata vel cuneata
apice truncate vel rotundato. Filamenta 9-13 mm longa; an-
therae 6-8 x 4-6 mm inter se non cohaerentes, poris duobos
minutis terminalibus; connectivum ad basim dorsaliter dente
ALMEDA: NEW BLAKEA AND TOPOBEA
315
FIGURE 9. Blakea scarlatina Almeda. A, habit; B, representative leaf (abaxial surface) with enlargement showing pubescence
detail; C, outer floral bracts; D, inner floral bracts enveloping hypanthium; E, hypanthium (top view) showing ovary summit,
torus and calyx lobes; F, petal (adaxial surface); G, stamens, lateral view (left), ventral view (right). (A, B from the holotype;
C-G from Hammel 9297, CAS.)
3-5 mm armatum. Stylus 2-2.1 cm; ovarium 6-iocuiare et buds, juvenile leaves, pedicels, and young hy-
omnino inferum cono 5 mm longo (coiio inciuso) giabro. panthia sparsely to densely furfuraceous-puber-
Massive epiphytic shrubs to 4 m tall with coarse ulent but becoming glabrate with age. Mature
spreading branches 0.5-8 m long. Vegetative leaves of a pair essentially equal or only slightly
316
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 14
unequal in size; petioles 1.7-4.5(-6.3) cm long;
blades thick and coriaceous, 8.5-19.7 cm long
and 5.4-10 cm wide, elliptic to elliptic-ovate,
apex acuminate, base acute to obtuse or rounded,
margin entire, 3 -nerved with an additional pair
of inconspicuous intramarginal veins, the trans-
verse secondary veins mostly 1-2 mm apart at
the widest portion of the blade, essentially gla-
brous but whitish puncticulate above, sparsely
furfuraceous-puberulent on the elevated primary
veins below or almost glabrous. Rowers erect,
solitary in the axils of uppermost leaves; pedicels
somewhat compressed, 1-1.8 cm long. Floral
bracts thick and coriaceous, sessile and entire;
outer bracts connate basally for about half their
length, 2-2.7 x 2.2-3.6 cm, broadly ovate to
suborbicular, apex rounded to subtruncate, typ-
ically mucronate, deciduously furfuraceous-pu-
berulent abaxially; inner bracts connate for their
entire length to form a shallowly lobed, glabrous,
bowl-like collar 1-2 cm high that envelops the
hypanthium but is usually concealed by the outer
bracts. Hypanthia (at an thesis) campanulate, 1-
1.3 cm long to the torus and 1.3-1.6 cm in di-
ameter. Calyx tube 5-6 mm long; calyx lobes
erect or slightly spreading, 5-10 mm long and
9-14 mm wide basally, triangular-ovate and
apiculate, margin entire, moderately to sparingly
furfuraceous-puberulent abaxially. Petals 6, fleshy
and glabrous when fresh, 3-4.5 x 2.5-4.3 cm,
brilliant red but white for the basal '/4 of their
length, obovate to cuneate, apically truncate to
broadly rounded, entire. Stamens 12, isomor-
phic, free but connivent and forming a somewhat
declinate nearly semicircular ring around the style;
filaments complanate, white, fleshy and gla-
brous, 9-13 mm long; anthers 6-8 mm long and
4-6 mm wide, yellow, oblong-ovoid, laterally
compressed and tipped with two minute pores;
connective thickened and prolonged dorsally at
the base into an acute spur 3-5 mm long that
assumes a pseudoterminal position because of
the pronounced adaxial tilt of each anther apex.
Ovary completely inferior, 6-celled, glabrous at
the gently fluted summit and dilated apically into
a cone and stylar collar 5 mm long. Style decli-
nate and incurved distally, glabrous, 2-2.1 cm
long and 0.5-1 mm wide, stigma truncate to cap-
itellate. Mature berry and seeds not seen.
PHENOLOGY.— Flowering specimens have been
collected in every month except December and
January; of the nearly 40 collections available
for this study none had mature fruits.
DISTRIBUTION. —Widespread but local and un-
common in wet evergreen forests from southern
Nicaragua (Zelaya) south to Costa Rica from Par-
que Nacional Rincon de la Vieja (Guanacaste)
and the Cordillera de Tilaran (Alajuela) south-
east to the vicinity of Turrialba (Cartago) from
sea level to 1 ,450 m (Fig. 5).
ADDITIONAL SPECIMENS EXAMINED.— COSTA RICA. Ala-
juela: 3 km N of Rio Cataratas on road to Bajo Rodriquez, 23
Feb. 1 978, Almeda & Nakai 3884 (CAS, CR); 1 8-35 km NNW
of San Ramon on road to Cataratas, 21 Mar. 1986, Almeda el
al. 5652 (CAS); La Palma de San Ramon, 6 Nov. 1923, Brenes
3920 (F); La Palma de San Ramon, 29 Sep. 1925, Brenes 4466
(F); La Palma de San Ramon, 24 Oct. 1926, Brenes 5031 (F);
La Palma de San Ramon, 24 Nov. 1926, Brenes 5115 (F); La
Palma de San Ramon, 24 Nov. 1926, Brenes 5138 (F); La
Palma de San Ramon, 22 Aug. 1927, Brenes 5671 (F); La
Palma de San Ramon, 5-8 Aug. 1935, Brenes 20627 (F); Re-
serva Forestal de San Ramon above and in valley of Rio San
Lorencito, 10°13'N, 84°37'W, 12-14 Mar. 1987, Burger et al.
12105 (CAS, F); Monteverde Reserve, Atlantic slope, Rio Pe-
nas Blancas valley, 19 Mar. 1986, Haber 1519 (CAS); San
Carlos, margin of Rio Penas Blancas at elevation of San Pedro,
29 Jun. 1985, Haber & Bella 1744 (CAS); San Carlos, Penas
Blancas, 9 Jul. 1985, Haber & Bella 1873 (CAS); Rio Penas
Blancas, San Carlos, 1 1 Aug. 1 985, Haber & Bella 2306 (CAS);
finca Don Bolivar Ruiz on road between the Reserva de San
Ramon station and the road to Palmarena, Feb. 1987, Herrera
508 (CAS); 6 km S of Ciudad Quesada, near Rio La Vieja, 10
Sep. 1967, Lent 1265 (F); 8 km NE of Villa Quesada, near
Artezalea, 17 Feb. 1966, Molina et al. 17258 (F); near San
Isidro de Penas Blancas, 22 Oct. 1965, Schnell 231 (F); La
Marina, near Ciudad Quesada, 23 Oct. 1965, Schnell 243 (F);
Villa Quesada, 21 Feb. 1939, Smith HI 603 (F); 5.8 km N of
Cataratas de San Ramon, 1 1 Apr. 1976, J. & K. Utley 4606
(CAS, DUKE). Cartago: Instituto Interamericano de Ciencias
Agricolas, Turrialba, 4 Oct. 1950, Ledn 2790 (MO); 2.5 km E
of Tuis along banks of Rio Tuis, 9 Mar. 1978, Utley 6006
(CAS, DUKE). Cartago/San Jose Border: Estacion Carrillo,
Parque Nacional Braulio Carrillo, 28 Jul. 1984, Zamora &
Elizondo 672 (CAS). Guanacaste: Parque Nacional Rincon de
la Vieja, Hacienda Santa Maria, SE of Mirador on road to
Volcan Santa Maria, 10°48'N, 85°19'W, Herrera & Robles 756
(CAS, MO). Heredia: Finca La Selva on Rio Puerto Viejo, just
E of its junction with the Rio Sarapiqui, along Far Loop Trail,
1,000 m S, 21 Jul. 1980, Hammel 9297 (CAS, DUKE, F);
Finca La Selva on Rio Puerto Viejo just E of its junction with
Rio Sarapiqui, ridge above Q. Esquina, 30 Aug. 1981, Smith
135 (CAS, DUKE). Limon: hills 2 airline km SSE of Islas
Buena Vista in Rio Colorado, 10°40'N, 83°40'W, 13-14 Sep.
1986, Davidse & Herrera 31127 (CAS); Hacienda Tapezco-
Hda. La Suerte, 29 airline km W of Tortuguero, 10°30'N,
83°47'W, 28 Aug. 1979, Davidson & Donahue 8913 (CAS);
path beyond Rio Sucio, Braulio Carrillo, May 1984, Gomez
et al. 22760 (CAS); ridge between Rio Chirripo and Q. El
Molinete, 10°12'N, 83°54'W, 19 Jul. 1984, Grayumetal. 3553
(CAS); Cerro Coronel, E of Laguna Danto, 10°41'N, 83°38'W,
1 5-20 Sep. 1986, Stevens & Montiel 24614 (CAS). Puntarenas:
Monteverde community, windbreak next to road, 10°20'N,
84°50'W, 22 Jul. 1986, Haber 5729 (CAS). NICARAGUA.
Zelaya: 4 km N of Nueva Guinea, between Rio Plata and San
Antonio, 1 1°44'N, 84°26'W, 7 Sep. 1983, Nee 27883 (CAS).
ALMEDA: NEW BLAKEA AND TOPOBEA
317
Blakea scarlatina is unique among its Central
American congeners in having large flowers (8-
9 cm across at anthesis) with brilliant red petals
that are white for the basal '/4 of their length. For
an obligate epiphyte this species has a compar-
atively broad geographic and elevational distri-
bution; however, all available collections exhibit
remarkable uniformity in floral and vegetative
features.
Blakea scarlatina was first collected by A. M.
Brenes in 1 923. Many additional specimens have
accumulated in herbaria since that time, but most
have repeatedly been misidentified as B. gran-
diflora Hemsley. Blakea grandiflora, which is
known only from higher elevations ( 1 ,400-2,350
m) in Costa Rica, differs in having 5-plinerved
leaves with an additional intramarginal pair of
veins, longer pedicels (3-5 cm), flowers borne in
clusters of 2-4 in the upper leaf axils, petals that
are white flushed with pink distally, and anther
connectives that are modified dorso-basally into
short deflexed spurs. Vegetatively, B. scarlatina
has an aspect most reminiscent of B. cuneata
Standley, a little-collected species of Belize, Gua-
temala, and Honduras. Striking similarities in
foliar morphology and pubescence details ini-
tially led me to assign many collections of B.
scarlatina to B. cuneata. Study of the type of B.
cuneata and more recent collections, however,
leaves no doubt that these two allopatric species
are readily separated by a number of diagnostic
characters. In B. cuneata there is a consistent
tendency for the elliptic leaves to be narrower
(3.5-6 cm), and the free portions of the outer
floral bracts are elliptic-lanceolate and bluntly
acute at the apex. Blakea cuneata also differs in
having smaller petals (1.8-2.1 x 1.5-2 cm) that
are pink, shorter filaments (5-6 mm), and linear-
oblong anthers that are short (5-5.5 mm), erect,
and unappendaged dorso-basally.
In anther morphology B. scarlatina most
closely approaches B. tuberculata]. D. Smith and
B. cuatrecasii Gleason. Blakea tuberculata also
has anthers that are connivent and free, but they
form a complete circle around the straight style.
Blakea cuatrecasii has stamens that form a nearly
circular ring opposing the declinate style, but the
anthers are laterally fused for nearly half of their
length. Because both of these species differ from
B. scarlatina in so many other characters, I see
no compelling reason to suggest that they form
a particularly close alliance based solely on an-
droecial morphology.
The flowers of B. scarlatina are some of the
most spectacular among neotropical epiphytes.
They are also some of the most enigmatic when
attempting to make inferences about likely pol-
linators. The casual observer might be inclined
to suggest bird pollination because of the brilliant
red coloration of the petals. This seems highly
unlikely. I have found no detectable nectar in
any flowers examined in the field during morning
and late afternoon hours. These observations are
supported by Stein and Tobe (1989), who found
no anatomical basis for nectar production in this
species (referred to as Blakea sp. nov.).
Although pollinators of B. scarlatina remain
unknown, a color slide photograph in the frag-
ment packet of Burger et al. 12105 shows the
brown anther discoloration (bruising) typically
found on yellow-anthered melastome species that
have been visited by buzz-pollinating bees. Field
studies will be needed to determine the pollinator
spectrum of this species and to evaluate conflict-
ing reports on the nature of its floral fragrance.
The flowers of B. scarlatina have been described
as having a musky odor (Haber 1519), a distinct
delicate fragrance (Almeda et al. 5652), and a
strange, unpleasant odor (Nee 27883).
The name for this species is derived from scar-
latinus, Latin for scarlet, in reference to the vivid
red petal color.
Topobea caliginosa Almeda, nom. nov.
Blakea micrantha Almeda, Rhodora 82:614. 1980.
TYPE. — PANAMA. Veraguas: Cerro Tute ca 10 km NW of
Santa Fe on ridgetop in cloud forest above 1 ,000 m, 1 9 Jun.
1975, Mori 6765 (holotype: CAS!; isotype: MO!).
A study of anther morphology in recently col-
lected flowering material of this species has led
to the conclusion that it is closely related to T.
cordata, T. crassifolia, and T. hexandra. This has
necessitated a reassessment of the proper generic
placement for this specialized group of species.
For reasons discussed in the synoptic history at
the beginning of this paper, it seems appropriate
to transfer Blakea micrantha to Topobea. Adop-
tion of a nomen novum is needed because the
epithet micrantha is pre-empted in Topobea (Al-
meda 1981b:307). For additional comments and
an enumeration of its diagnostic features, see the
discussion under T. hexandra. The new name
for this species is derived from caligo, Latin for
fog, mist, or darkness, in reference to its fog-
shrouded habitat at the summit of Cerro Tute.
318
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 14
Topobea crassifolia (Almeda) Almeda, comb,
nov.
Blakea crassifolia Almeda, Rhodora 82:612. 1980.
TYPE:— PANAMA. Code: La Mesa above El Valle in forest
on both sides of junction with road to Cerro Pilon, ca 800 m,
21 Jul. 1974, Croat 25430 (holotype: CAS!; isotypes: MO!,
US!).
The rationale for transferring B. micrantha to
Topobea also applies to B. crassifolia. Among
the four hexandrous species, T. crassifolia is most
like T. cordata in having ovate to elliptic-ovate
leaves and a 4-celled ovary. The leaves of T.
crassifolia are petiolate whereas those of T. cor-
data are invariably sessile. Anther differences be-
tween these two species are discussed above un-
der the synoptic history. With a distribution that
extends to Costa Rica, T. crassifolia is the only
hexandrous species known to occur outside of
Panama.
Topobea fragrantissima Almeda, sp. nov.
(Fig. 10)
TYPE:— PANAMA. Chiriqui: vicinity of Fortuna Dam, along
trail across valley of Rio Homito, elev. 1,100-1,250 m, 12
Mar. 1988, Almeda et al. 6086 (holotype: CAS!; isotypes: CR!,
F!, MO!, PMA!, TEX!, US!).
Frutex epiphyticus vel terrestris. Petioli 0.5-3 cm longi; lam-
ina 1.5-5.5 x 1.6-3 cm elliptica vel elliptico-obovata apice
acuminata basi acuta ad maturitatem glabra, 3-nervata vel
3-plinervata, subcoriacea et integra nervis secundariis 0.25 mm
inter se distantibus. Flores 6-meri in quoque nodo superiore
singuli vel bini, pedicellis 2-3 cm longis, bracteae omnino
liberae ellipticae vel obovatae; bracteae exteriores 5-11 x 3-
5 mm apice rotundato; bracteae interiores 4-6 x 4-5 mm apice
rotundato. Hypanthium (ad torum) 3-4 mm longum extus
glabrum; calycis tubus 1 mm longus, lobis 1 mm longis. Petala
1.2-1.4 x 1-1.2 cm obovato-oblonga apice rotundato. Fila-
menta 5-6 mm longa; antherae 3.5-5 x 1 mm oblongae la-
teraliter non cohaerentes, poro unico dorsaliter inclinato; con-
nectivum ad basim dorsaliter dente 0.25 mm longo descendenti
armatum. Stylus 1-1.1 cm; ovarium 4-loculare et lh inferum
apice glabro (cono et collo non evoluto).
Epiphytic or terrestrial shrubs or small trees
1.5-4 m tall. Distal branchlets subquadrate and
glabrous with interpetiolar ridges or lines. Veg-
etative buds and young leaves sparingly and de-
ciduously lepidote-furfuraceous. Mature leaves
of a pair equal to somewhat unequal in size, gla-
brous throughout; petioles 0.5-3 cm long; blades
subcoriaceous, 1.5-5.5 cm long and 1.6-3 cm
wide, elliptic to elliptic-obovate, apex acumi-
nate, base acute, margin entire, 3-nerved or
3-plinerved abaxially with an additional sub-
marginal pair of inconspicuous veins, the striol-
ate transverse secondary veins spaced mostly 0.25
mm apart at the widest portion of the blade.
Flowers erect, solitary or paired in leaf axils of
distal branches; pedicels 2-3 cm long, glabrous.
Floral bracts sessile, entire and free to the base;
outer bracts 5-11 x 3-5 mm, elliptic or rarely
varying to obovate, glabrous, apex rounded; in-
ner bracts 4-6 x 4-5 mm, obovate, glabrous,
apex broadly rounded. Hypanthia (at anthesis)
campanulate, 3-4 mm long to the torus and 4-
5 mm in diameter, glabrous. Calyx tube 1 mm
long; calyx lobes 1 mm long and 1-1.5 mm wide
basally, ovate to deltoid-ovate with a blunt cal-
lose-thickened tooth on the abaxial apex of each
lobe, margin entire but sometimes roughened
along interlobe sinuses, glabrous on both sur-
faces. Petals 6, glabrous, 1.2-1.4 x 1-1.2 cm,
white flushed with pink unilaterally, obovate,
apically rounded, entire. Stamens 12, isomor-
phic, free and strongly declined to one side of
the flower opposing the style; filaments com-
planate and glabrous, 5-6 mm long; anthers 3.5-
5 mm long, 1 mm wide, yellow, linear-oblong
and tipped with a solitary dorsally inclined pore;
connective dilated dorso-basally into a deflexed
spur 0.25 mm long. Ovary Va inferior, 4-celled,
glabrous at the summit but not distended into a
cone or collar. Style declinate and somewhat sig-
moid, glabrous, 10-11 mm long; stigma punc-
tiform. Berry globose, 7-10 mm long and 10 mm
in diameter. Seeds clavate to narrowly pyriform
or pyramidate, 1 mm long, beige with a smooth
testa and conspicuous lateral raphe.
PHENOLOGY.— Flowering and fruiting speci-
mens have been collected in January, March,
April, and July.
DISTRIBUTION.— Local and uncommon in cloud
forests from the Boquete region of western Pan-
ama (Chiriqui) to the slopes bordering the Rio
Hornito above Los Planes east to Cerro Colorado
at 1,000- 1,300m (Fig. 5).
ADDITIONAL SPECIMENS EXAMINED.— PANAMA. Chiriqui:
Edwin Fabrega Dam and Reserve along trail to Rio Hornito
above Los Planes, 8°45'N, 82°15'W, 18 Jan. 1989, Almeda et
al. 6309 (CAS, MO, PMA); Monte Rey, above Boquete, 21
Jul. 1971, Croat & Porter 15692 (CAS, MO); trail to Zarzo,
between Los Planes de Hornito and Fortuna Lake, 8°41'N,
82°13'W, Hampshire & W hitefoord 689 (BM, CAS); ca. 5 km
E of Fortuna Dam along trail crossing Rio Hornito, 8°45'N,
82°15'W, 26 Apr. 1988, Thompson 5021 (CAS, CM). Chiriqui/
Bocas del Tore Border: windswept cloud forest off the road to
Cerro Colorado, 26 Jan. 1989, Almeda et al. 6418 (CAS, CR,
DUKE, MO, PMA, US).
Topobea fragrantissima is distinguished by its
glabrous leaves with finely striolate secondary
ALMEDA: NEW BLAKEA AND TOPOBEA
319
FIGURE 10. Topobea fragrantissima Almeda. A, habit; B, representative leaf (abaxial surface); C, D, enlargement of foliar
venation (abaxial surface); E, petal (adaxial surface); F, stamens, ventral view (left), lateral view (middle), dorsal view (right);
G, berry with persistent decussate bracts; H, outer floral bract (adaxial surface); I, inner floral bract (adaxial surface); J, seeds.
(A-I from the holotype; J from Croat & Porter 15692, CAS.)
venation, long-pedicellate, erect flowers, linear-
oblong anthers with solitary dorsally inclined
pores, and a 4-celled ovary. Although it has been
described as an epiphytic shrub or small tree, my
field observations and the label information on
most collections suggest that this species typi-
cally becomes a small, free-living tree. The style
in this species is decimate like many other species
320
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 14
of Topobea, but the posture and orientation of
its androecium is unlike that of other congeners
with free anthers. In the erect-flowered species
of Topobea with free anthers, the filaments are
declined to one side of the flower opposing the
style and the anthers are typically erect or some-
what incurved apically. In T.fragrantissima this
orientation is taken to such an extreme that both
filaments and anthers lie flat on the surface of
the petals to form a semicircular configuration
(Fig. 10A). In this unusual androecial orientation
the ventral surface of the anther faces upward
and the dorsal surface (including the dorsally in-
clined pore) face downward toward the petal sur-
face.
Field observations of pollination in this species
should be of special interest because this pollen
presentation mechanism is coupled with the pro-
duction of a perfume-like fragrance. The unusual
anther posture and orientation of the pore may
require some peculiar manipulative behavior on
the part of bees if this species is buzz-pollinated
like the majority of species in the family with
poricidally dehiscent anthers.
Topobea fragrantissima is similar to T. suave-
olens (also described below) in a number of char-
acters, and it seems likely that these two species
are closely related. The shape and venation of
mature leaf blades are nearly identical. In the few
known collections of T. suaveolens, however, fo-
liar pocket domatia (Fig. 1 2B) are formed abax-
ially in the angle between the median vein and
the innermost pair of lateral veins; domatia are
lacking in T. fragrantissima. The differences be-
tween these species are most readily apparent
when fresh flowers are examined. In T. suaveo-
lens the connivent petals give the pendant flowers
a bell-like conformation, the free stamens form
a ring around the straight style, and the ovary is
6-celled. The most distinctive feature shared by
the two species is the solitary anther pore. Within
the Blakeeae, this derived character is known in
other species. Topobea acuminata Wurdack, T.
caudata Wurdack, T. dodsonorum Wurdack, and
T. pittieri Cogn. are a few notable examples. The
fact that solitary anther pores occur in more than
one species complex within Topobea suggests that
this character has arisen more than once in the
course of its evolutionary history.
The name for this species is derived fromfra-
gro, Latin for sweet smelling, in reference to the
perfume-like fragrance of fresh flowers.
Topobea hexandra Almeda, sp. nov.
(Fig. 11)
TYPE.— PANAMA. Panama: Cerro Jefe, along summit road
and along trail into the Chagres Valley, elev. ca. 900 m, 19
Feb. 1988, Almeda et al. 5837 (holotype: CAS!; isotypes: CR!,
DUKE!, F!, MO!, NY!, PMA!, TEX!, US!).
Frutex hemiepiphyticus. Ramuli primum sulcato-quadran-
gulati demum teretes glabri (in nodis caduce puberuli pilis
castaneis ca. 0.5-1 mm longis). Petioli 5-14 mm longi; lamina
2-3.9 x 1.1-3 cm subrotundata vel elliptico-ovata apice ro-
tundato vel obtuso basi late acuta vel obtusa vel rotundata,
3-nervata coriacea et integra, nervis secundariis nervulisque
invisis. Flores 6-meri sessiles vel subsessiles in quoque nodo
superiore singuli vel bini; bracteae omnino liberae; bracteae
exteriores 5-6.5 x 3-5 mm ovatae vel ovato-ellipticae apice
obtuso vel mucronato; bracteae interiores 4-5 x 4-6 mm late
ovatae vel suborbiculares apice rotundato. Hypanthium (ad
torum) 3 mm longum extus sparsiuscule caduceque stellulato-
furfuraceum; calycis tubus 1 mm longus, lobis 2 mm longis.
Petala 6.5-7 x 4 mm obovato-elliptica apice obtuso. Fila-
menta 3 mm longa; antherae 6, ca. 2 x 1 mm oblongae inter
se non cohaerentes, ventraliter biporosae; connectivum ad ba-
sim dorsaliter dente 0.25 mm descendenti armatum. Stylus
5.5 mm; ovarium 2-loculare et omnino inferum apice glabro
(cono et collo non evoluto).
Hemiepiphytic shrubs to 1 m tall adhering to
the bark of host trees by nodal and internodal
adventitious roots. Distal branchlets quadrate to
quadrisulcate, glabrous or sparsely beset with
spreading, deciduous, glandular hairs 1-2 mm
long like the young petioles and upper and lower
surfaces of the juvenile leaves; older branches
rounded with leaf scars that are typically swollen
and nodular in appearance. Uppermost nodes
copiously beset with brown spreading hairs. Veg-
etative buds copiously covered with a deciduous
brown stellate-lepidote indument. Leaves of a
pair equal in size, glabrous throughout; petioles
5-14 mm long; mature blades coriaceous, 2-3.9
cm long and 1.1-3 cm wide, suborbicular to el-
liptic-ovate, apex rounded or varying to obtuse,
base obtuse to rounded, rarely varying to acute,
margin entire, 3 -nerved, often with an additional
intramarginal pair of depressed veins. Flowers
erect, solitary or paired in the leaf axils of distal
branches, sessile or subsessile with short (to 1
mm) ill-defined pedicels formed by the com-
pressed bases of the outer floral bracts. Floral
bracts thick and semisucculent, sessile, entire and
free to the base, sparingly stellulate-furfuraceous
abaxially; outer bracts 5-6.5 x 3-5 mm, con-
cave, ovate to elliptic-ovate, apex obtuse to
bluntly mucronate; inner bracts 4-5 x 4-6 mm,
broadly ovate to suborbicular, apex rounded.
Hypanthia (at anthesis) narrowly campanulate,
ALMEDA: NEW BLAKEA AND TOPOBEA
321
VJ
D
FIGURE 11. Topobea hexandra Almeda. A, habit; B, representative leaf (abaxial surface), C, enlargement of distal node
showing spreading hairs; D, representative flower (natural posture); E, petal (abaxial surface); F, stamens, ventral view (left),
lateral view (right); G, berry with persistent floral bracts removed; H, abaxial surface of inner floral bract (left), abaxial surface
of outer floral bract (right); I, seeds. (A-H from the holotype; I from Witherspoon 8552.)
322
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 14
3 mm long to the torus and 4-5 mm in diameter,
sparingly stellate-lepidote or stellulate-furfura-
ceous. Calyx tube 1 mm long; calyx lobes erect,
2 mm long and 2-2.5 mm wide basally, ovate to
deltoid-ovate, entire but irregularly roughened
along interlobe sinuses, sparingly stellate-lepi-
dote. Petals 6, liberally covered with disc-shaped
hyaline processes, 6.5-7 x 4 mm, pink, elliptic-
obovate, apically obtuse, entire. Stamens 6, iso-
morphic and free; filaments complanate and gla-
brous, 3 mm long, declinate but incurved distally
bringing the anthers to an incurved position op-
posing the style; anthers 2 mm long, 1 mm wide,
pale yellow, turning brownish orange with age,
oblong with two ventrally inclined pores at the
broadly rounded apex; connective slightly thick-
ened and dilated dorsally at the filament inser-
tion into a short blunt spur up to 0.25 mm long.
Ovary completely inferior, 2-celled, sparingly
beset with stellate-lepidote hairs, glabrous at the
summit surrounding the stylar scar but not dis-
tended into a prominent cone or collar. Style
declinate, incurved distally, glabrous, 5.5 mm
long; stigma punctiform. Berry globose, 5-6 mm
long and 4-7.5 mm in diameter. Seeds bluntly
deltoid, 1-1.5 mm long, beige with a smooth
glossy testa and conspicuous lateral raphe.
PHENOLOGY. — The only known flowering
specimens were collected in February; fruiting
collections have been made in February, Septem-
ber, October, and December.
DISTRIBUTION. —Endemic to the low cloud for-
ests on Cerro Jefe in central Panama at 900-
1,000 m (Fig. 8).
ADDITIONAL SPECIMENS EXAMINED.— PANAMA. Panama:
summit and S facing slopes of Cerro Jefe, 9 Feb. 1978, Almeda
& Nakai 3459 (CAS); Cerro Jefe, along trail on ridge running
NE from summit, 18 Dec. 1974, Mori & Kallunki 3755 (MO);
Cerro Jefe, 29 Oct. 1980, Sytsma 2007 (MO); Cerro Jefe, road
leading N from summit, 26 Sep. 1975, /. T. & F. WUherspoon
8552 (MO).
Collections of T. hexandra are few despite the
fact that it is one of the most common shrubby
epiphytes in the forest at the summit of Cerro
Jefe. In the field, fertile material is evidently
overlooked without the aid of binoculars, be-
cause the small flowers are produced on upper-
most branchlets positioned high on host trees
where access to sunlight is optimal.
Topobea hexandra, together with T. cordata,
T. caliginosa and T. crassifolia, form a closely
related species group best interpreted as a spe-
cialized evolutionary line within the genus. They
are distinguished from other congeners by a com-
bination of specialized features associated with
great reduction in flower size. These are: (1) flow-
ers sessile or short-pedicellate (2-3 mm) at an-
thesis; (2) diminutive petals (4. 5-9 x 1.5-6 mm);
(3) each flower has only six stamens, each of
which is attached to the torus opposite a calyx
lobe; and (4) the ovary is 2-celled or 4-celled.
The stamen number in T. crassifolia and T. cali-
ginosa was unknown to me when I described
these taxa, because all available study material
had flowers with detached stamens (Almeda
1980). Recent collections of these two species
and T. cordata clearly show that they are all hex-
androus. The reduction in stamen number ex-
hibited by this group of species is otherwise un-
known in the tribe.
Among its close allies, T. hexandra is most
similar to T. caliginosa. Both species have
3-nerved leaves, translucent petals that have a
liberal scattering of hyaline disc-shaped gland-
like processes, and 2-celled ovaries. Topobea ca-
liginosa differs in having glabrous uppermost
nodes, oblanceolate to spatulate, basally atten-
uate leaf blades, confluent anther pores, and un-
appendaged anther connectives.
Within its restricted range, T. hexandra shows
inconstancy in characters of the indument. Distal
branchlets can either be glabrous or sparsely be-
set with spreading glandular hairs. When present
these hairs are found on distal internodes, peti-
oles of juvenile foliage, and on upper and lower
surfaces of some young leaves, but most of the
hairs appear to fall away with age. All specimens
of the type collection are devoid of glandular
hairs. Mori & Kallunki 3755 and WUherspoon
8552 are beset with hairs as described above, but
Almeda & Nakai 3459, which consists of two
branches, has hairs on one branch but not on the
other. Unfortunately, T. hexandra is known from
too few collections to determine which, if any,
of these forms constitutes the prevalent condi-
tion.
The epithet for this species is derived from the
Greek words hex, six, and andros, male, in ref-
erence to the 6-stamened flowers.
Topobea parvifolia (Gleason) Almeda, comb. nov.
Blakea parvifolia Gleason, Phytologia 3:357. 1950.
TYPE.— PANAMA. Code: crest of Cerro Pajita, El Valle de
Anton, 1,100 m, Allen 3761 (holotype: NY; isotype: MO!).
ALMEDA: NEW BLAKEA AND TOPOBEA
323
In the protologue, Gleason described the an-
thers ofBlakea parvifolia as acute and emargin-
ate at the tip with each anther sac opening by a
separate dorso-terminal pore (Fig. 3B). Because
anthers of this kind can only be interpreted as
those of Topobea, Gleason appears to have erred
in placing this species in Blakea and subsequent-
ly maintaining it in that genus for his treatment
of the family in the Flora of Panama (Gleason
1958). Gleason's brief discussion in the proto-
logue also stated that "the one flower remaining
had lost part of its petals and stamens but there
was no evidence that there had been more than
four of the one and eight of the other." My study
of this species in the field and herbarium shows
that it has stamens that are coherent in a ring,
as noted by Gleason, but the flowers are consis-
tently 6-merous with 1 2 stamens.
Despite the heightened collecting activity in
Panama during the past decade, T. parvifolia is
still known only from the windswept slopes and
ridges in the vicinity of El Valle de Anton.
Topobea suaveolens Almeda, sp. nov.
(Fig. 12)
TYPE: PANAMA. Veraguas: along trail to summit of Cerro Tute
about '/2 mile above the Escuela de Agriculture Alto Piedra
near Santa Fe, elev. 900-1,100 m, 29 Jan. 1989, Almeda et al.
6484 (holotype: CAS!; isotypes: AAU!, BM!, BR!, CR!, DUKE!,
F!, MEXU!, MICH!, MO!, NY!, PMA!, TEX!, US!).
Arbor epiphytica 4 m. Ramuli primum sulcato-quadran-
gulati demum teretes glabri; linea interpetiolaris paulo elevata
evoluta. Petioli 1.1-1.8 cm longi; lamina 3.6-5.3 x 1.6-2.5
cm ell i plica apice acuminata vel caudato-acuminata basi acuta,
3-plinervata, nervi in axillis acarodomatiis instructi, subcori-
acea et integra, nervis secundariis 0.25 mm inter se distantibus.
Flores 6-meri penduli in quoque nodo superiore singuli, pedi-
cellis, 2.3-3 cm longis; bracteae omnino liberae; bracteae ex-
teriores 4.5-7.5 x 3-4 mm ellipticae vel ovato-ellipticae apice
acuto vel rotundato; bracteae interiores 4-4.5 x 4 mm ovatae
vel suborbiculares apice rotundato. Hypanthium (ad torum) 4
mm longum extus glabrum; calycis tubus 1 .5 mm longus, lobis
1 mm longis. Petala 1.2-1.5 x 0.9-1.1 cm obovata apice ro-
tundato. Filamenta 2.5-3.5 mm longa; antherae 2.5 x 1 mm
oblongae inter se lateraliter non cohaerentes, poro unico dor-
saliter inclinato; connectivum nee prolongatum nee appendi-
culatum. Stylus 8.5-9 mm; ovarium 6-loculare et '/2 inferum
apice glabro (cono et collo non evoluto).
Epiphytic trees to 4 m tall, often obscuring and
overtaking the crowns of host trees. Distal
branchlets quadrate to quadrisulcate and gla-
brous with well-defined interpetiolar ridges or
lines. Vegetative buds deciduously lepidote-fur-
furaceous. Mature leaves of a pair equal or slight-
ly unequal in size, glabrous throughout; petioles
1.1-1.8 cm long; blades subcoriaceous, 3.6-5.3
cm long and 1.6-2.5 cm wide, elliptic, apex acu-
minate to caudate-acuminate, base acute, margin
entire, 3-plinerved with an additional incon-
spicuous pair of submarginal veins and numer-
ous striolate transverse secondary veins spaced
mostly 0.25 mm apart at the widest portion of
the blade, pocket domatia typically formed abax-
ially in the angle between the median vein and
the two proximal lateral veins. Flowers pendant
and solitary in the leaf axils of uppermost
branches; pedicels 2.3-3 cm long, glabrous. Flo-
ral bracts sessile, glabrous, entire, and free to the
base; outer bracts 4.5-7.5 x 3-4 mm, elliptic to
elliptic-ovate, apex bluntly acute to rounded; in-
ner bracts 4-4.5 x 4 mm, ovate to suborbicular,
apex broadly rounded. Hypanthia (at anthesis)
campanulate, 4 mm long to the torus and 5.5-6
mm in diameter, glabrous. Calyx tube 1.5 mm
long; calyx lobes 1 mm long and 4 mm wide
basally, broadly ovate to deltoid-ovate with a
blunt callose-thickened tooth on the abaxial apex
of each lobe, margin entire, glabrous on both
surfaces. Petals 6, glabrous, connivent to some-
what imbricate and bell-like when fully expand-
ed, 1.2-1.5 x 0.9-1.1 cm, white flushed with
dark pink along a broad continuous margin, ob-
ovate, apically rounded, entire. Stamens 12, iso-
morphic, free and encircling the exserted style;
filaments complanate and glabrous, 2.5-3.5 mm
long; anthers 2.5 mm long, 1 mm wide, yellow,
laterally compressed, oblong in ventral view and
narrowly ovoid in profile view with a shallow
dorso-basal depression at the filament insertion,
tipped with a solitary, dorsally inclined pore 0.75
mm in diameter; connective simple. Ovary Vz
inferior, 6-celled, dilated at the glabrous summit
into a smooth gentle dome 1-1.5 mm high. Style
straight, glabrous, 8.5-9 mm long; stigma trun-
cate. Berry globose, 5-6 mm long and 6-7 mm
in diameter. Seeds narrowly and irregularly pyr-
iform, 0.75-1 mm long, pale brown with a smooth
testa and prominent lateral raphe.
PHENOLOGY.— The three known collections of
this species which are in flower and fruit were
collected in January, February, and March.
DISTRIBUTION.— Known only from the cloud
forests of Cerro Tute in west-central Panama at
850-1, 100m (Fig. 8).
ADDITIONAL SPECIMENS EXAMINED.— PANAMA. Veraguas:
vicinity of Cerro Tute, along trail to summit, 08°30'N, 8 1°07'W,
19 Mar. 1987, McPherson 10654 (CAS); near Cerro Tute-
Arizona, above Santa Fe and Alto de Piedra, 8°30'N, 8 TIO'W,
5 Feb. 1988, McPherson 12043 (CAS).
324
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 14
FIGURE 12. Topobea suaveolens Almeda. A, habit; B, representative leaf (abaxial surface) with enlargement showing do-
matium; C, flower (natural posture); D, petal (adaxial surface); E, stamens, lateral view (left), dorsal view (middle), ventral view
(right); F, berry with persistent decussate bracts; G, seeds. (A-F from the holotype; G from McPherson 10654.)
ALMEDA: NEW BLAKEA AND TOPOBEA
325
Topobea suaveolens is unusual in having flow-
ers that are bell-like and pendant with free con-
nivent anthers that form a ring around the exsert-
ed straight style. This floral syndrome is not found
in the other species of Topobea with brightly
colored petals. Topobea suaveolens is clearly a
sister species of T.fragrantissima (also described
above). In addition to differences noted in the
discussion of the latter, T. suaveolens is distinc-
tive in at least two other floral features. Its fila-
ments are conspicuously broadened at the base
(Fig. 1 2E), and its shorter, laterally compressed
anthers are narrowly ovate in profile view and
have a conspicuous dorso-basal depression.
The strong vegetative similarities between T.
suaveolens and T.fragrantissima can make iden-
tification of fruiting specimens difficult. Al-
though T. suaveolens is known from few collec-
tions, even sterile material can be distinguished
by the peculiar foliar pocket domatia that are
formed abaxially in the angle between the mid-
vein and the two proximal lateral veins (Fig. 1 2B).
Perhaps the most remarkable feature of this
species is its agreeable floral fragrance. I de-
scribed it as fruity (Almeda 6484), but a more
telling comparison likens it to grape-flavored
chewing gum (McPherson 12043). The name for
this species is derived from suavis, Latin for sweet,
in reference to the sweet-smelling flowers.
ACKNOWLEDGMENTS
This study was supported, in large part, by
U.S. National Science Foundation Grant BSR
8614880 (Rora Mesoamericana) and the G.
Lindsay Field Research Fund of the California
Academy of Sciences. For field assistance and/
or special technical support in the herbarium I
thank Bruce Bartholomew, Thomas F. Daniel,
Greg de Nevers, Gordon McPherson, Kei Nakai,
Orbelia R. Robinson, and Colleen Sudekum. For
logistical support in the field, special thanks go
to the Museo Nacional de Costa Rica, the Mis-
souri Botanical Garden, the Organization for
Tropical Studies, and the Smithsonian Tropical
Research Institute. I am also grateful to John J.
Wurdack for a review of the manuscript; Varsha
C. Patel for information on pollen morphology;
Ellen del Valle for the line drawings; Gina Umana
Dodero for reviewing the Spanish summary; and
the curators and staffs of the following herbaria
(acronyms fide Holmgren et al. 1981) for loans,
gifts, or special assistance during study visits:
BM, BR, CM, CR, DUKE, F, K, MEXU, MO,
NY, P, PMA, TEX, US, WIS.
RESUMEN
Blakea y Topobea, los unices generos en la
tribu Blakeeae, estan intimamente relacionados,
pero las anteras de Topobea tienen tecas lineal-
oblongas, con uno o dos poros dorso-apicales o
dos poros ventro-apicales. Topobea es un genero
con mas de 60 especies descritas, incluido en
Blakea solamente por Don, Baillon y Macbride,
pero es muy util reconocerlo hasta que dispon-
gamos de una monografia. Se presenta una his-
toria taxonomica de los dos generos con un co-
mentario sobre las caracteristicas de las anteras
empleado originalmente por Browne, Aublet,
Cogniaux, y otros botanicos. Las diferencias ge-
nericas a menudo se han alterado un poco para
reflejar mejor las caracteristicas inusuales de al-
gunas especies que se encuentran en las monta-
nas de Costa Rica y Panama. Se provee una clave
neuva para los dos generos. Se describen cuatro
especies nuevas de Blakea (B. gregii, B. ham-
melii, y B. herrerae de Panama; y B. scarlatina
de Costa Rica y Nicaragua), y tres especies nue-
vas de Topobea (T.fragrantissima, T. hexandra,
y T. suaveolens de Panama) y se transfieren tres
especies de Blakea a Topobea. Estos cambios se
basan en caracteristicas de las anteras. Se pro-
veen descripciones, ilustraciones y discusiones
sobre las afinidades entre las especies nuevas.
Ademas se presentan notas sobre polinizacion y
mapas de distribucidn.
LITERATURE CITED
ACKERMAN, J. D. 1986. Coping with the epiphytic existence:
pollination strategies. Selbyana 9:52-60.
ALMEDA, F. 1 980. Central American novelties in the genus
Blakea (Melastomataceae). Rhodora 82:609-615.
. 1 98 1 a. Blakea penduliflora (Melastomataceae): a new
green-flowered species from Costa Rica. Brittonia 32(4):508-
511.
. 1981b. New and reconsidered species of Miconia
(Melastomataceae) from Costa Rica and Panama. Proc. Cal-
if. Acad. Sci. 42(10):303-314.
. 1989. Five new berry-fruited species of tropical
American Melastomataceae. Proc. Calif. Acad. Sci. 46(5):
137-150.
. In press. Melastomataceae. In Flora de Nicaragua.
W. D. Stevens, ed. Missouri Botanical Garden, St. Louis.
AUBLET, J. B. C. F. 1775. Histoire des plantes de la Guiane
francoise. London and Paris. 4 vo\s. (Topobea, 1:476, t. 189.)
BAILLON, H. E. 1879. Melastomacees. Hist. PI. 7:1-65. (En-
glish transl. 7:1-65. 1881.)
BENTHAM, G. AND J. D. HOOKER. 1867. Melastomaceae.
Genera plantarum l(2):725-773.
326
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 14
BROWNE, P. 1756. The civil and natural history of Jamaica
in three parts. London. (Blakea, p. 323, t. 34.)
BUCHMANN, S. L. 1 983. Buzz pollination in angiosperms. Pp.
73-113 in Handbook of experimental pollination biology.
C. E. Jones and R. J. Little, eds. Van Nostrand Reinhold
Co., Inc., New York, New York.
CANDOLLE, A. P. DE 1828. Melastomaceae. In Prodromus
systematis naturalis regni vegetabilis 3:99-202.
COGNIAUX, A. 1888. Melastomaceae (Blakeae). In C. F. P.
von Martius, Flora Brasiliensis 14(4):558-564.
. 1891. Melastomacees. In A. and C. de Candolle, eds.
Monographiae phanerogamarum 7: 1-1 256. G. Masson, Paris.
DON, D. 1823. An illustration of the natural family called
Melastomataceae. Mem. Wern. Nat. Hist. Soc. 4:276-329.
GLEASON, H. A. 1935. Melastomaceae. In A. Pulle, ed. Flora
of Surinam 3:178-281. J. H. de Bussy Ltd., Amsterdam.
. 1945. On Blakea and Topobea. Bull. Torrey Bot.
Club 72(4):385-393.
. 1 947. Additional notes on Blakea and Topobea. Phy-
tologia 2:279-281.
. 1950. Observations on tropical American mela-
stomes. Phytologia 3:345-360.
. 1958. Melastomataceae. In R. E. Woodson, Jr. and
R. W. Schery, eds. Flora of Panama. Ann. Missouri Bot.
Gard. 45:203-304.
HOLMGREN, P. K., W. KEUKEN, AND E. K. SCHOFIELD. 1981.
Index Herbariorum. Part 1, 7th ed. The herbaria of the world.
Regnum Veg. 106:1-152.
HOWARD, R. 1983. The plates of Aublet's Histoire des plantes
de la Guiane francoise. J. Arnold Arbor. 64:225-292.
JACOBS, M. 1966. On domatia— the viewpoints and some
facts. I, II, III. Proc. Koninkl. Nederl. Akad. Wetensch. C69:
275-316.
JUSSIEU, A. L. DE. 1789. Genera plantarum. Paris.
KOEK-NOORMAN, J., P. HOGEWEG, W. H. M. VAN MAANEN,
AND B. J. H. TER WELLE. 1979. Wood anatomy of the
Blakeeae (Melastomataceae). Acta Bot. Neerl. 28:21^*3.
LINNAEUS, C. 1759. Systema naturae. Vol. 2, 10th ed. Stock-
holm. (Facsimile ed. by J. Cramer, Weinheim, 1964.)
LINNAEUS, C., VON FILIUS 1781. Supplementarum plantarum.
Braunschweig.
LUMER, C. 1981. Rodent pollination of Blakea (Melasto-
mataceae) in a Costa Rican cloud forest. Brittonia 32(4):
512-517.
LUMER, C. AND R. D. SCHOER. 1986. Pollination of Blakea
austin-smithii and B. penduliflora (Melastomataceae) by small
rodents in Costa Rica. Biotropica 18(4):363-364.
LUNDSTROM, A. N. 1 887. Pflanzenbiologische studien II. Die
anpassungen der pflanzen an thiere, I. Von domatia. Nova
Acta Regiae Soc. Sci. Upsal. (Ser. 3) 13(10):l-88.
MACBRIDE, J. F. 1 94 1 . Melastomataceae. In Flora of Peru.
Field Mus. Nat. Hist., Bot Ser. 13:249-521.
NAUDIN, C. 1852. Melastomacearum monographicae de-
scriptionis. Ann. Sci. Nat. Ill, 18:85-154.
PATEL, V. C, J. J. SKVARLA, AND P. H. RAVEN 1985. Pollen
characters in relation to the delimitation of Myrtales. Ann.
Missouri Bot. Gard. 71:858-969.
PEMBERTON, R. W. AND C. E. TURNER 1989. Occurrence of
predatory and fungivorous mites in leaf domatia. Amer. J.
Bot. 76(1): 105- 11 2.
RENNER, S. S. 1 989. A survey of reproductive biology in
neotropical Melastomataceae and Memecylaceae. Ann. Mis-
souri Bot. Gard. 76:496-518.
ROUBIK, D. W. 1989. Ecology and natural history of tropical
bees. Cambridge University Press, Cambridge, England.
SOLT, M. L. ANDj.J. WURDACK 1980. Chromosome numbers
in the Melastomataceae. Phytologia 47:199-220.
STAGE, C. A. 1965. Cuticular studies as an aid to plant tax-
onomy. Bull. Brit. Mus. (Nat. Hist.), Bot. 4(1): 1-78.
STANDLEY, P. C. 1 924. Melastomaceae. In Trees and shrubs
of Mexico. Contr. U.S. Natl. Herb. 23(4): 1046-1074.
. 1938. Melastomaceae. In Flora of Costa Rica. Field
Mus. Nat. Hist., Bot Ser. 18(3):783-845.
STANDLEY, P. C. AND L. O. WILLIAMS. 1963. Melastomaceae.
In Flora of Guatemala. Fieldiana, Bot. 24:407-525.
STEIN, B. A. AND H. TOBE. 1989. Floral nectaries in Melas-
tomataceae and their systematic and evolutionary implica-
tions. Ann. Missouri Bot. Gard. 76:519-531.
TRIANA, J. 1871. Les Melastomacees. Trans. Linn. Soc. Lon-
don 28:1-188.
WELLE, B. J. H. TER AND J. KOEK-NOORMAN. 1981. Wood
anatomy of the neotropical Melastomataceae. Blumea 27:
335-394.
WILLIAMS, N. H. 1983. Floral fragrances as cues in animal
behavior. Pp. 50-72 in Handbook of experimental polli-
nation biology. C. E. Jones and R. J. Little, eds. Van Nos-
trand Reinhold Co., Inc., New York, New York.
WURDACK, J. J. 1973. Melastomataceae. In T. Lasser, ed.
Flora de Venezuela 8:1-819.
. 1980. Melastomataceae. In G. Harling and B. Sparre,
eds. Flora of Ecuador 13: 1^06.
CALIFORNIA ACADEMY OF SCIENCES
Golden Gate Park
San Francisco, California 941 18
PROCEEDINGS
OF THE
CALIFORNIA ACADEMY OF SCIENCES
Vol. 46, No. 15, pp. 327-335, 4 figs.
September 11, 1990
THREE NEW SPECIES OF CONOSTEGIA
(MELASTOMATACEAE: MICONIEAE) FROM
SOUTHERN CENTRAL AMERICA
By
Frank Almeda
Department of Botany, California Academy of Sciences, Golden Gate Park,
San Francisco, California 94118
ABSTRACT: Three new species of Conostegia, a well-defined berry-fruited genus of Melastomataceae with a
center of diversity in Costa Rica and Panama, have been collected in little-explored regions of southern Costa
Rica and western Panama during the past decade. Descriptions, discussions, diagnostic illustrations, pheno-
logical notes, and a distribution map are provided for Conostegia fragrantissima, C. muriculata, and C.
orbeliana. Attention is drawn to the need for recording field data on taxonomically useful floral and fruit
characters.
Received February 22, 1990. Accepted April 10, 1990.
INTRODUCTION
Conostegia, a neotropical genus of about 50
species is one of approximately 30 berry-fruited
genera assigned to the New World tribe Mico-
nieae. Unlike some genera in this large tribe of
1,800-1,900 species, Conostegia is a natural as-
semblage readily denned by its terminal inflo-
rescence, calyptriform calyx that is circumscissle
at or near the torus, and isomorphic stamens
with lorate anthers and unappendaged connec-
tives. Conostegia is also noteworthy among neo-
tropical Melastomataceae because most (but not
all) of its species have flowers that are anisomer-
ous and typically pleiostemonous. Among other
genera in the Miconieae only a few species of
Clidemia and Miconia have flowers with unequal
numbers or proportions of serially homologous
structures.
Variation in several species of Conostegia re-
mains imperfectly understood because of inad-
equate geographic sampling and fragmentary
preservation of floral characters on herbarium
specimens. Petal and stamen numbers, the pos-
ture and orientation of the style, stigma mor-
phology, and ovary cell number provide useful
information for the delimitation of species in
Conostegia. These characters are often difficult
to evaluate after specimens have been pressed
and dried. Ideally, notes on these features should
be recorded in the field when the preservation of
flowers and fruits in a liquid medium is not fea-
sible.
The three species described below all come
from rich, but little-known, areas of Costa Rica
and Panama. Over 33 species of Conostegia oc-
cur in this small isthmian region, making it the
center of diversity for the genus and a promising
area for continued botanical exploration in the
Mesoamerican region.
Conostegia fragrantissima Almeda, sp. nov.
Fig. 1
TYPE.— PANAMA. Bocas del Toro. Fortuna Dam area, along
continental divide trail bordering Chiriqui Province at
8°45'04"N, 82°15'04"W, 1,200-1,300 m, 10 Mar. 1988, Al-
[327]
328
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 15
FIGURE 1. Conostegia fragrantissima Almeda. A, habit, x ca. l/y, B, representative leaf (abaxial surface), x 1; C, flower bud
when fresh, x ca. 4; D, young berry with detached calyptriform calyx, x 5; E, flower bud when dry, x ca. 4; F, representative
flower showing reflexed petals, declinate style and stigma enlargement (right), x ca. 3; G, petal (adaxial surface), x ca. 5; H,
stamens, ventral view (left) and lateral view (right), x ca. 8; I, seeds, x ca. 30. (A-I from the holotype.)
meda et al. 6064 (holotype: CAS!; isotypes: AAU!, BM!, BR!, 1-2.6 cm longi; lamina 4.5-9.5 x 1.5^4.2 cm elliptica vel
CR!, DUKE!, F!, MEXU!, MO!, NY!, PMA!, TEX!, US!). elliptico-lanceolata apice acuminata vel caudato-acuminata basi
acuta, 3(-5)-plinervata, membranacea et Integra supra glabra,
Arbor 4-11 m. Ramuli primum quadrangulati demum tere- subtus sparsiuscule caduceque lepidota. Inflorescentia 4-7.5
tes glabri; linea interpetiolaris paulo elevata evoluta. Petioli cm longa multiflora; pedicellis (ad anthesim) 2.5-7 mm longi;
ALMEDA: NEW SPECIES OF CONOSTEGIA
329
C. fragrantissima
C. muriculata
C. orbeliana
too 200 km
FIGURE 2. Distributions of Conostegia fragrantissima, C. muriculata, and C. orbeliana.
bracteolis 0.25-1.5 mm longis mox caducis. Alabastra matura
acuminata vel apiculata 5-7 x 2-3 mm; calyx calyptriformis
3-4.5 mm longus. Petala 6-7, glabra 5-7 x 3.5-4.5 mm, obo-
vata apice irregulariter rotundato. Stamina 15-19, isomor-
phica glabra; filamenta 2-2.5 mm longa; antherarum thecae
1.5-2 x 0.5 mm oblongae, poro ventraliter inclinato; connec-
tivum nee prolongatum nee appendiculatum. Stylus 5-5.5 mm
glaber; stigma capitatum 1.5 mm diam.; ovarium 6-loculare
et omnino inferum apice glabro (cono et collo non evoluto).
Trees 4-1 1 m tall. The distal branchlets gla-
brous and bluntly quadrate with darkened inter-
petiolar lines or ridges. Leaves of a pair equal to
somewhat unequal in size; petioles 1-2.6 cm long;
blades membranaceous, 4.5-9.5 cm long and 1.5-
4.2 cm wide, elliptic to elliptic-lanceolate, apex
acuminate to caudate-acuminate, base acute,
margin entire, 3(-5)-plinerved, the innermost pair
of elevated primaries diverging from the median
nerve in opposite fashion (l-)3-5 mm above the
blade base, glabrous above, sparsely and decid-
uously furfuraceous-lepidote below. Inflores-
cence a terminal paniculiform dichasium 4-7.5
cm long typically branching above the node ini-
tiating the inflorescence, the rachis inconspic-
uously and deciduously furfuraceous-lepidote;
bracteoles sessile to subsessile, early deciduous,
narrowly triangular to subulate or linear-oblong,
0.25-1.5 mm long, 0.25 mm wide, glabrous,
margin entire. Pedicels 2.5-7 mm long, decidu-
ously furfuraceous-lepidote. Flower buds ob-
long-ellipsoid, becoming horizontally constrict-
ed at the torus and truncate at the base when
dry, 5-7 mm long, 2-3 mm wide, smooth and
glabrous or sparsely furfuraceous-lepidote, the
calyptriform calyx 3-4.5 mm long, acuminate to
apiculate; torus glabrous adaxially. Petals 6-7,
glabrous, white with a horizontal red band near
the base, obovate, reflexed, irregularly lobed api-
cally, 5-7 mm long and 3.5-4.5 mm wide dis-
tally. Stamens 15-19, gently declined to one side
of the flower opposing the style; filaments gla-
brous, complanate, 2-2.5 mm long; anthers 1.5-
2 mm long, 0.5 mm wide, orange, linear-oblong,
truncate to broadly rounded at the apex with a
somewhat ventrally inclined terminal pore 0.25
mm in diam.; connective thickened dorsally but
not prolonged or appendaged at the filament in-
sertion. Ovary (at anthesis) completely inferior,
6-celled, ovoid, the apex glabrous, smooth and
lacking an elevated cone or collar. Style decli-
nate, glabrous, 5-5.5 mm long; stigma capitate,
1.5 mm in diam. with 6-7 papillose lobes. Berry
depressed-globose, 3 mm long and 4 mm in diam.
Seeds oblong, oblong-ovoid, or narrowly pyram-
idate, 0.5-0.75 mm long, beige, smooth and pol-
ished on the convex face.
PHENOLOGY.— Flowering specimens have been
330
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 15
collected in the months of February through May;
fruiting specimens have been collected in Feb-
ruary and April.
DISTRIBUTION.— Local in wet evergreen and
elfin forests of western Panama from Cerro Pate
Macho (Chiriqui) to the Cerro Colorado region
(Bocas del Toro) at 1,200-2,000 m (Fig. 2).
ADDITIONAL SPECIMENS EXAMINED. — PANAMA. Bocas del
Toro: Fortuna Dam area, along continental divide trail,
8°45'04"N, 82°15'04"W, 10 Mar. 1988, Almeda et al. 6068
(CAS); Fortuna Dam region near trail along continental divide,
8°45'N, 82°15'W, 11 Feb. 1986, McPherson 8406 (CAS); vi-
cinity of Cerro Colorado on trail along creek 8.6 miles from
Camp Chami, 8°35'N, 81°45'W, 14 Apr. 1986, McPherson
8900 (CAS); vicinity of Cerro Colorado mine above San Felix,
8°35'N, 8°50'W, 26 Jan. 1988, McPherson 12024 (CAS). Chi-
riqui: SE slopes and summit of Cerro Pate Macho, 4 km NE
of Boquete, 26 May 1981, Sytsma et al. 4884 (CAS).
This species is distinguished by a combination
of indument and floral characters. The lower leaf
surfaces, branchlets of the inflorescence, pedi-
cels, and distal cauline internodes are sparsely
beset with deciduous furfuraceous-lepidote hairs.
Although many specimens superficially appear
glabrous to the unaided eye, the inconspicuous
indument is readily detected on one or all of the
above-mentioned structures when examined with
a dissecting microscope. Aside from the decli-
nate style and lobulate depressed-capitate stig-
ma, most of the distinctive floral features are not
readily gleaned from herbarium specimens with-
out good field notes. The flowers are fragrant, the
anthers are bright orange in color, and each con-
spicuously reflexed white petal has a red hori-
zontal band near the base. This distinctive petal
coloration is, to my knowledge, otherwise un-
known in Conostegia.
The closest relative of C. fragrantissima is
probably C. pittieri Cogn., a species known only
from Costa Rica and Nicaragua. They both have
elliptic to elliptic-lanceolate, caudate-acuminate,
plinerved leaves and flowers with similar petal
and stamen numbers. Conostegia pittieri is gla-
brous throughout, and its leaves are prevailingly
undulate-denticulate. It also differs from C. fra-
grantissima in having longer floral buds (9-13
mm), larger petals (9-13 x 5-8 mm), more ovary
cells (8-10), and a crateriform capitate stigma.
In Gleason's treatment of Conostegia for the
Flora of Panama (Gleason 1958), C. fragrantis-
sima keys closest to C. montana (Sw.) D. Don
ex DC., a widespread species ranging from south-
ern Mexico (Chiapas) and the West Indies (How-
ard 1989) south through Central America to Co-
lombia, Venezuela, and Ecuador. Conostegia
fragrantissima shares many features with C.
montana including foliar shape, petal and sta-
men numbers, and the number of ovary cells,
but the latter has stellate hairs on the inflores-
cence rachis, pedicels, and elevated primary veins
on lower leaf surfaces. In addition, the flowers
of C. montana produce no detectable odor, the
pedicels are shorter (1 mm), the petals are com-
pletely white, the anthers are pale yellow, and
the style (when dry) is abruptly expanded just
below the subtruncate unlobed stigma. The style
and stigma enlargement of C. fragrantissima (Fig.
IF) was drawn from liquid-preserved flowers.
Upon drying, the style retains its linear outline,
but the capitate stigma becomes gently fluted and
bluntly obconic in profile view.
The fragrant flowers of C. fragrantissima are
unusual but not unique among its congeners. The
flowers of C. pittieri are also fragrant (fide Lumer
1148, CAS), but information of this kind is rarely
reported on herbarium labels. In choosing the
epithet for this species, which is derived from
fragro, Latin for sweet smelling, I draw attention
to the very fruity fragrance of the fresh flowers.
Conostegia muriculata Almeda, sp. nov.
(Fig. 3)
TYPE. — PANAMA. Bocas del Toro: Above Chiriqui Grande,
1 0 road-miles from the continental divide and 2 miles along
road to the east, 8°55'N, 82°10'W, 300 m, 6 Aug. 1988,
McPherson 12836 (holotype: CAS!; isotypes: CR!, DUKE!,
MO!, PMA!, US!).
Frutex vel arbor parva 2.5-3.5 m. Ramuli primum acute
quadrangulati demum teretes glabri; linea interpetiolaris ob-
scure evoluta. Petioli 1.5-7 cm longi; lamina 8-25.5 x 4.5-
14 cm elliptica vel obovato-elliptica apice acuminata basi acu-
ta, 5-nervata vel 5-7-plinervata, membranacea et integra, su-
pra glabra, subtus sparsiuscule caduceque lepidota. Inflores-
centia 12-30 cm longa (pedunculo 4-10 cm longo incluso) laxa
multiflora; pedicellis (ad anthesim) 0.5-3 mm longis; bracteolis
0.5-1 mm longis mox caducis. Alabastra matura acuminata
6-10 x 3.5-4.5 mm; calyx calyptriformis 3-5 mm longus.
Petala 5, glabra, 5-6 x 4_5(_6.5) mm, obovata apice rotun-
dato. Stamina (8-)9-10, isomorphica glabra; filamenta 2.5-3
mm longa; antherarum thecae 2-3 x 0.75 mm oblongae, poro
ventraliter inclinato; connectivum nee prolongatum nee ap-
pendiculatum. Stylus 3-4 mm glaber; stigma capitatum 0.75
mm diam. : ovarium 6-loculare et omnino inferum apice glabro
(cono et collo non evoluto).
Shrubs or small trees 2.5-3.5 m tall. The distal
branchlets glabrous with interpetiolar lines,
quadrate and carinate on the angles but becom-
ALMEDA: NEW SPECIES OF CONOSTEGIA
331
B
FIGURE 3. Conostegia muriculata Almeda. A, habit, x ca. Vi; B, representative leaf (abaxial surface), x '/2; C, flower bud
(left) and bud with detached calyptriform calyx and floral parts removed (right), x 4; D, representative flower showing spreading
petals, arching style, and stigma enlargement (right), x 5; E, petal (adaxial surface), x 6; F, stamens, ventral view (left) and
lateral view (right), x 7; G, seeds, x 26. (A-F from the holotype; G from Almeda et al. 6328.)
332
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 15
ing rounded-quadrate with age. Leaves of a pair
equal or unequal in size (if unequal then the
smaller one 2/3 to 3/5 the size of the larger one);
petioles 1.5-7 cm long, sparsely to moderately
lepidote when young; blades membranaceous, 8-
25.4 cm long and 4.5-14 cm wide, elliptic to
elliptic-obovate, apex acuminate, base acute,
margin entire, 5 -nerved or inconspicuously 5-7-
plinerved with a prominulous network of sec-
ondary and higher order veins, glabrous above,
sparsely and deciduously lepidote below. Inflo-
rescence a terminal paniculiform dichasium 1 2-
30 cm long on a well-defined arching or deflexed
peduncle mostly 3.5-10 cm long, the rachis gla-
brous; bracteoles sessile, early deciduous, nar-
rowly triangular to subulate, 0.5-1 mm long,
0.25-0.5 mm wide, sparingly lepidote to glabrate
abaxially, margin entire. Pedicels 0.5-3 mm long,
sparsely to moderately lepidote. Flower buds ob-
long-ellipsoid, 6-10 mm long 3.5-4.5 mm wide,
sparsely to moderately lepidote basally, the ca-
lyptriform calyx 3-5 mm long, acuminate; torus
glabrous adaxially and callose thickened along
the adjacent circumscissle line. Petals 5, gla-
brous, obovate, reflexed to widely spreading,
gently lobed to somewhat oblique apically, 5-6
mm long, 4-5(-6.5) mm wide distally. Stamens
(8-)9-10; filaments glabrous, complanate, 2.5-3
mm long; anthers 2-3 mm long, 0.75 mm wide,
pale yellow, linear-oblong with a ventrally in-
clined terminal pore 0.25 mm in diam.; connec-
tive thickened dorsally but not prolonged or ap-
pendaged at the filament insertion. Ovary (at
anthesis) completely inferior, 6-celled, ovoid, the
apex glabrous, smooth, and lacking an elevated
cone or collar. Style conspicuously curved dis-
tally, glabrous, 3-4 mm long; stigma capitate,
0.75 mm in diameter, copiously papillose but
lacking well-defined lobes. Berry purple at ma-
turity, depressed-ovoid, 5-6 mm long and 5-7
mm in diameter. Seeds narrowly pyriform, 0.75
mm long, yellowish-white, irregularly muriculate
on the convex face.
PHENOLOGY.— Flowering specimens have been
collected in April, July, August, and September;
specimens in young fruit have been collected in
September, but mature fruiting collections have
been made only in January.
DISTRIBUTION. — A local and uncommon
understory tree of wet evergreen forests from
southeastern Costa Rica (Limon) southeast to the
Caribbean slopes of Panama, from Bocas del Toro
province to Cerro Tute (Veraguas) and the region
north of El Cope (Code) from 90-1,200 m (Fig.
2).
ADDITIONAL SPECIMENS EXAMINED.— COSTA RICA. Limon:
Hitoy Cerere Biological Reserve, SW of Valle La Estrella along
Rio Cerere to ca. 1 km upstream from Quebrada Barrera,
9°40'30"N, 83°02'W, 31 Jul. 1985, Grayum & Hammel 5773
(CAS). PANAMA. Bocas del Toro: above Chiriqui Grande on
a side road ca. 10 miles below the continental divide ca. 2'/2
miles E on that road, 8°55'N, 82°10'W, 19 Jan. 1989, Almeda
et al 6328 (CAS, CR, MO, NY, PMA). Code: New Works, 7
km N of El Cope, 18 Aug. 1977, Folsom 4938 (CAS, PMA);
forest on continental divide above El Cope, 8°38'N, 80°38'W,
27-29 Apr. 1985, Hammel 13663 (CAS). Veraguas: Cerro Tute,
trail past agricultural school near Santa Fe, 17 Sep. 1979, An-
tonio 1840 (CAS).
Specimens of C. muriculata were first collected
in 1977 but the distinctiveness of this species
became evident only after study of the flowering
collection designated here as the type. The dis-
tinguishing features of C. muriculata are its
quadrate carinate distal branchlets; sparingly
lepidote indument on lower leaf surfaces, pedi-
cels, and flower buds; distinctly pedunculate pan-
iculiform dichasia; and muriculate seed coat.
Collectors have described the petals of this species
as deep blue (Folsom 4938), pale blue- violet
(Grayum & Hammel 5773), lavender (Mc-
Pherson 12836), and purple (Antonio 1840,
Hammel 13663). All of these collectors, except
Folsom and McPherson, have also noted that the
inflorescence is pendant or deflexed. Because these
characters are unusual in Conostegia and poten-
tially diagnostic, additional field study is needed
to determine whether they are consistent.
The close relatives of C. muriculata among
described congeners are not readily apparent. It
superficially resembles C. tenuifolia J. D. Smith
and C. rhodopetala J. D. Smith. The former, de-
scribed from Costa Rica (Smith 1899) but now
also known from Nicaragua and Panama, differs
in having narrower leaves (3.6-7.4 cm) with
abruptly caudate-acuminate apices, ovoid buds
that are conspicuously subtruncate at the base
(when dry), 16-22 stamens per flower, an ovary
with 10-12 locules, and seeds with a smooth
testa. In having oblong-ellipsoid acuminate flow-
er buds and a 6-celled ovary, C. muriculata is
also similar to the Costa Rican endemic C. rho-
dopetala. In the field, C. rhodopetala is readily
distinguished from other species of Conostegia
by the bright pink coloration of inflorescence
branches, flower buds, and petals. The most no-
table characters separating it from C. muriculata
ALMEDA: NEW SPECIES OF CONOSTEGIA
333
include its sparse and deciduously stellulate-fur-
furaceous indument on the inflorescence rachis,
longer petals (8-10 mm), larger number of sta-
mens per flower (12-17), and smooth seeds.
The name for this species is the diminutive of
muricatus, Latin for short and tubercular ex-
crescences. It emphasizes the finely sculptured
seeds. The seeds of C. polyandra Benth. are de-
picted as muriculate in the plate included with
the protologue (Bentham 1 844). However, an ex-
amination of several recent collections of C. po-
lyandra from throughout its range reveals that
its seeds are completely smooth.
Conostegia orbeliana Almeda, sp. nov.
(Fig. 4)
TYPE.— PANAMA. Bocas del Toro: vicinity of Cerro Col-
orado mine above San Felix, along trails N of road along
continental divide, 8°35'N, 81°50'W, 1,500 m 26 Jan. 1988,
McPherson 12014 (holotype: CAS!; isotypes: CR!, DUKE!,
MEXU!, MO!, PMA!, TEX!, US!).
Arbor 6-7 m. Ramuli quadrangulati demum teretes sicut
foliorum subtus venae primarieae inflorescentia hypanthiaque
dense pilis stellatis et pilis stipitato-stellatis induti. Petioli 1.2-
3 cm longi; lamina 5-1 1.2 x 3-5.5 cm elliptica vel elliptico-
ovata, apice acuminata vel cuspidata, basi acuta, 3-5-pliner-
vata, membranacea et obscure serrulata vel subintegra. Inflo-
rescentia 4-7(-10) cm longa laxa pauciflora; pedicellis (ad an-
thesim) 8-20 mm longis; bracteolis 1-2 mm longis mox caducis.
Alabastra matura acuta vel acuminata 12-15 x 9-11 mm;
calyx calyptriformis 6-8 mm longus. Petala 8, glabra, 11-15
x 9-13 mm, obovata apice rotundato vel rotundato-subtrun-
cato. Stamina 19-22, isomorphica glabra; filamenta 5-5.5 mm
longa; antherarum thecae 3-3.5 x 0.75 mm oblongae, poro
dorsaliter inclinato; connectivum nee prolongatum nee appen-
diculatum. Stylus 7 mm glaber, stigmate subpeltato centre
concavo costato; ovarium 12-13-loculare et omnino inferum,
collo 2 mm alto glabro paulo costulato.
Trees 6-7 m tall. Uppermost internodes quad-
rate becoming rounded with age, moderately to
copiously covered with a mixture of sessile-stel-
late and short stipitate-stellate hairs. Leaves of
a pair equal to somewhat unequal in size; petioles
1.2-3 cm long; blades membranaceous, 5-11.2
cm long and 3-5.5 cm wide, elliptic to elliptic-
ovate, apex acuminate to cuspidate, base acute,
margin inconspicuously serrulate to subentire, 3-
5-plinerved, the innermost pair of elevated pri-
maries diverging from the median nerve in op-
posite or subalternate fashion 5-9 mm above the
blade base, glabrous above at maturity, moder-
ately covered with a mixture of sessile-stellate
and short stalked-stellate hairs on the elevated
primary and secondary veins below. Inflores-
cence a terminal paniculiform dichasium 4-7
(-10) cm long, the rachis copiously beset with
sessile stellate and short stalked stellate hairs;
bracteoles sessile, early deciduous, subulate to
linear-oblong, 1-2 mm long, 0.25-0.5 mm wide
with pubescence like that of the rachis on the
abaxial surface, margin entire. Pedicels 8-20 mm
long, copiously beset with sessile-stellate and
short, stalked, stellate hairs. Rower buds ellip-
soid, 12-15 mm long, 9-11 mm wide, coarsely
verrucose for the basal half of their length and
moderately covered with tardily deciduous stel-
late hairs throughout, the calyptriform calyx 6-
8 mm long, acute to acuminate; torus glabrous
adaxially. Petals 8, glabrous, connivent, white,
obovate, widely spreading, apically rounded to
subtruncate with irregularly lobed undulate mar-
gins, 11-15 mm long, 9-13 mm wide distally.
Stamens 19-22, typically positioned in a ring
around the style; filaments glabrous, complanate,
5-5.5 mm long; anthers 3-3.5 mm long, 0.75
mm wide, yellow, linear-oblong, truncate to
broadly rounded at the apex with a somewhat
dorsally inclined terminal pore 0.25 mm in di-
ameter; anther connective thickened dorsally but
not prolonged or appendaged at the filament in-
sertion. Ovary (at anthesis) completely inferior,
12- 13 -celled, depressed-globose, the apex gla-
brous, gently fluted or costulate with an elevated
stylar collar 2 mm long. Style straight, glabrous
7 mm long; stigma flared, subpeltate and crater-
iform with 12-13 longitudinal riblike lobes. Ma-
ture berry and seeds not seen.
PHENOLOGY. —The two known collections, both
of which are in flower, were made in January.
DISTRIBUTION.— A local and uncommon ele-
ment of cloud forest vegetation on slopes and
valleys in the vicinity of Cerro Colorado mine
above San Felix in western Panama at 1,450-
1,500m (Fig. 2).
ADDITIONAL SPECIMENS EXAMINED.— PANAMA. Bocas del
Toro/Chiriqui Border: slopes and valleys of Cerro Colorado
region, 27 Jan. 1989, Almeda et al. 6445 (CAS, MO, NY,
PMA).
Judging from the few specimens collected, this
species, which was first gathered in 1988, is ap-
parently rare and localized. Conostegia orbeliana
has a combination of features that readily set it
apart from its congeners. The uppermost cauline
internodes, pedicels, and inflorescence rachis are
mostly densely beset with a mixture of sessile
334
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46, No. 15
B
FIGURE 4. Conostegia orbeliana Almeda. A, habit, x ca. 'A, B, representative leaf (abaxial surface) with enlargement (right),
x ca. %; C, flower bud (left) and bud with detached calyptriform calyx and floral parts removed (right), x ca. 1 lh; D, representative
flower showing spreading petals, straight style, and stigma enlargement (right), x ca. 4/5; E, petals (adaxial surface), x 3; F,
stamens, ventral view (left) and lateral view (right), x 6. (A-C from the holotype; D-F from Almeda et al. 6445.)
stellate and short-stalked stellate hairs. Its in- Other notable features of C. orbeliana are the
conspicuously serrulate to subentire leaves are coarsely verrucose hypanthial cups (Fig. 4C) and
unusually small for a species with comparatively the dorsally inclined anther pore (Fig. 4F).
large flowers that measure ca. 3 cm in diameter. Overlapping similarities in foliar pubescence,
ALMEDA: NEW SPECIES OF CONOSTEGIA
335
petal and stamen numbers, style posture, stigma
morphology, and the number of ovary cells sug-
gest that C. orbeliana is closely related to C. vol-
canalis Standley & Steyerm. of southern Mexico
(Chiapas, Guerrero, Jalisco), Guatemala, Hon-
duras, and Nicaragua. I define C. volcanalis in a
broad sense, perhaps too broadly, to accom-
modate what I interpret as regional variation in
leaf and petal size, bud shape, and foliar pubes-
cence density. Some populations, for example,
have leaves that are nearly glabrous below. Oth-
ers have lower leaf surfaces beset with a mixture
of sessile stellate and long-stalked stellate hairs,
whereas others have only long-stalked stellate
hairs mostly confined to the elevated primary
and secondary veins. Even in its broadest cir-
cumscription, C. volcanalis is readily separated
from C. orbeliana by its larger (12-27 x 6-15
cm) mostly undulate-dentate leaves, longer pet-
ioles (3.5-6.7 cm), smaller floral buds (8-10 mm),
and completely smooth pedicels and hypanthial
cups.
This species is named for Orbelia R. Robin-
son, friend and fellow student of the Melasto-
mataceae, in grateful recognition of all the vol-
unteer technical and research assistance she has
rendered in the herbarium of the California
Academy of Sciences during the past decade.
ACKNOWLEDGMENTS
This study was supported, in part, by U.S. Na-
tional Science Foundation Grant BSR 8614880
(Flora Mesoamericana) and the Nathan Jay and
Virginia Friedman Fund of the California Acad-
emy of Sciences. For special technical and/or field
assistance during the course of this study I thank
Thomas F. Daniel, Gerrit Davidse, Greg de Nev-
ers, Gordon McPherson, and Orbelia R. Rob-
inson. My special thanks go to the Missouri Bo-
tanical Garden and the Smithsonian Tropical
Research Institute for logistical support in the
field; Ellen del Valle for the diagnostic illustra-
tions; Gina Umana Dodero for reviewing the
Spanish summary; and the curators and staffs of
the following herbaria (acronyms fide Holmgren
et al. 1 98 1 ) for loans, gifts, and/or assistance dur-
ing study visits: CR, DUKE, F, MO, PMA, US.
RESUMEN
Conostegia es un genero natural de amplia dis-
tribution pero con gran concentration en Costa
Rica y Panama, 33 de sus 50 especies occurren
alii. La relativa inaccesibilidad a ciertas areas
tales como las selvas de la parte sur de Costa
Rica (Limon) y de las provincias de Bocas del
Toro y Chiriqui en Panama, se reflejan en el
numero de especies recolectados hasta el presen-
te. El numero de los petalos y estambres, la po-
sition del estilo, la forma y tamano del estigma,
y el numero de loculos en el ovario son a menudo
caracteres taxonomicos muy utiles en Conoste-
gia. Se proveen descripciones, illustraciones y
discusiones referentes a las afinidades de tres es-
pecies nuevas: Conostegia fragrantissima y C.
orbeliana de Panama y C. muriculata de Costa
Rica y Panama. Tambien se presenta un mapa
de distribution para todas ellas.
LITERATURE CITED
BENTHAM, G. 1844. The botany of the voyage of H.M.S.
Sulphur, under the command of captain Sir Edward Belcher
. . . during the years 1836-1842. London.
GLEASON, H. A. 1958. Melastomataceae. In Flora of Panama.
R. E. Woodson, Jr. and R. W. Schery, eds. Ann. Missouri
Bot. Gard. 45:203-304.
HOLMGREN, P. K., W. KEUKEN, AND E. K. SCHOFIELD. 1981.
Index Herbariorum. Part 1, 7th ed. The herbaria of the world.
Regnum Veg. 106:1^52.
HOWARD, R. A. 1 989. Melastomataceae. In Flora of the Less-
er Antilles. 5(2):532-579. Arnold Arboretum, Harvard Uni-
versity, Jamaica Plain, Massachusetts.
SMITH, J. D. 1 899. Undescribed plants from Guatemala and
other Central American republics XXI. Bot. Gaz. 27:331-
339.
CALIFORNIA ACADEMY OF SCIENCES
Golden Gate Park
San Francisco, California 94118
INDEX TO VOLUME 46
(Compiled by Tomio Iwamoto)
New names
Blakea fuchsioides
Blakea gregii
Blakea hammelii
Blakea herrerae
Blakea scarlatina
Calicina arida
Calicina basalta
Calicina conifera
Calicina diminua
Calicina dimorphica
Calicina galena
Calicina mesaensis
Chiloglanis harbinger
Chiloglanis niger
Chiloglanis polypogon
Chiloglanis reticulatus
Chiloglanis sanagaensis
Clidemia hammelii
Conostegia fragrantissima
Conostegia muriculata
Conostegia orbeliana
Drapetis arnaudi
Drapetis cerina
Drapetis destitute
Drapetis Solaris
Drapetis torulosa
Dyschoriste mcvaughii
Hologr aphis caput-medusae
Holographis leticiana
Holographis tolantongensis
Justicia fortunensis
Justicia ixtlania
Justicia readii
Justicia tabascina
Justicia veraguensis
Miconia calocoma
Miconia dissitiflora
Miconia grayumii
Miconia ibarrae
Miconia iteophylla
Miconia ligulata
Miconia peltata
Neostethus robertsi
Pachycara crossacanthum
Pachycara gymninium
Pachycara lepinium
Pachycara mesoporum
Pachycara pammelas
Pachycara rimae
Pachycara shcherbachevi
Pachycara sulaki
Tococa croatii
Topobea caliginosa
Topobea fragrantissima
Topobea hexandra
Topobea suaveolens
New names in boldface type
Abies concolor 10
Acacia greggi 1 2
Acanthaceae 73, 79, 81, 279, 289-290, 297
Acanthodactylus sp. 197
Acanthostethus 274
falcifer 274
Adenostoma
fasciculatum 12
sparifolium 12
Aegotheles bennettii 196
Aegothelidae 196
Agama agama 194
Agamidae 195, 197
Agave
aurea 12
shawii 12
sp. 27, 29
spp. 12-13
Amaraboya 300
Amblyarrhena 146, 148
Aneides 47
lugubris 48, 58
Anguidae 3 1
Anniella
geronimensis 31-32, 45-46, 48, 59
pulchra 32, 44, 48
Anniellidae 31
Anoplopoma fimbria 222
Anthurium 314
Antimora 93
microlepis 93
rostrata 93
Aphelandra scabra 79
Aplonis sp. 196
Araceae 150, 314
Archoleptoneta 131
[337]
338
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46
Arctostaphylos
glauca 12
patula 12
pringlei 12
pungens 12
Arizona
elegans 45, 49
elegans eburnata 33
elegans occidentalis 33
elegans pacata 33
Artemisia californica 1 2
Assam iidae 123
Atherinidae 247-248
Atherinomorpha 243, 247-248, 254
Atopochilus 152, 163-164, 176
savor gnani 153, 164
sp. undet. a 153, 163
sp. undet. b 153, 163
Austrolycichthys 221, 223
brachycephalus 240
Sagarins
bagarius 163
yarrelli 163
Banksula96, 107, 122, 131
Batrachoseps 16, 60-61, 95, 129-130
attenuatus 129
pacificus 16, 59
pacificus major 15-16, 44, 48
pacificus spp. 15-16, 44, 48
Bedotia24&, 261
sp. 245
Bellucia 300
grossularioides 300
Beloperone 284
Betula glandulosa 188
Biantidae 123
fl/pes 60
biporus 48-49, 52,61
Blabomma 131
137, 140, 299-311, 323, 325
anomala 304
austin-smithii 302-303, 307-308
chlorantha 307-309
crassifolia299, 318
cuatrecasii 3 1 7
c lineal a 317
elliptica 303
fuchsioides 137-140
grandiflora 3 1 7
gregii 299, 301, 305-309, 325
hammelii 299, 309-31 1, 313, 325
herrerae299, 311-314, 325
hirsuta var. rotundata 304
litoralis 304
micrantha299, 317
parvifolia 299, 322-323
pauciflora 311, 313
penduliflora 307-308
purpusii 302, 308
quinquenervia 300
scarlatina 299, 308, 314-315, 317, 325
trinervia 300
tuberculata 303, 317
wilburiana 301-302
woodsonii 303
Blakeeae 299, 301, 303
Boidae 32
193-194, 198-203
angulata 194, 198, 201
blandingi 194, 197-201, 203
ceylonensis 194, 198, 200-201
195, 200-201, 203
n 195, 198-201, 203
dendrophila 193, 195, 198-200, 202-203
dightoni 195, 201
drapiezii 195, 201, 203
w 196, 200-201, 203
196, 201
hexagonatus 197
irregularis 193-194, 196, 198-203
jaspidea 197, 201, 203
kraepelini 197-198, 201, 203
multifasciata 197
multomaculata 197, 200-201
nigriceps 197, 201, 203
nuchalis 195
ocellata 197, 201, 203
ochracea 197-198, 201
pulverulenta 197-198, 201, 203
schultzei 197-198, 203
trigonata 197-199, 201, 203
Brotulinae 226
Bufo 19
alvarius 63
boreas 19, 38, 44, 47^18, 58
boreas halophilus 1 7
cognatus 63
microscaphus 19, 44, 48
microscaphus calif ornicus 18
punctatus 20, 45, 48
woodhousei 63
Bufonidae 17
Buparellus 122
Bupares 122
jamaicensis 27
Caddidae 134
Caddo pepper ella 134
Calicina 95-102, 112, 117, 121-123, 126, 129-135
arida95, 100, 111-112, 124-126, 130-133
antfa subgroup 99-100, 111-112, 122, 124, 127,
129, 132
basalta95, 116, 119, 126, 128-129, 132-133
INDEX
339
breva95, 116, 119, 126, 128-129, 132-133
cloughensis 95, 111-112, 124-126, 130-134
conifera95, 105, 126, 132-133
digita 95, 106, 122, 125-126, 130-134
digita group 99, 106, 113, 122-125, 129, 132, 134
digita subgroup 99, 106-107, 122-123, 127, 129,
132
diminua95, 99, 120-121, 125-126, 130-133
dimorphica 95, 116-117, 119, 126, 128, 132-133
ensata95, 101, 114, 117, 125-126, 128-129, 132-
133
galena 95, 109, 126, 130-134
kaweahensis 95, 108, 124-127, 130, 132-135
kaweahensis group 1 24
kaweahensis subgroup 99, 107-108, 122, 132
keenea95, 110, 126, 130, 132-133
macula 95, 114, 117, 119, 126, 128, 132-133
mariposa 95, 99-102, 105-106, 108-111, 122-
126, 132-134
mariposa group 99- 100, 105, 113, 122-123, 125,
129-132
mariposa subgroup 99, 105, 124, 127, 132
mesaensis95, 101, 116, 119, 126, 128, 132-133
minor95, 101, 113-114, 116, 125-127, 130, 132-
133
minor subgroup 105, 113, 122, 124, 132
morroensis95, 110, 126, 129-130, 132-133, 135
palapraeputia 95, 99, 102, 112-113, 122, 124-
127, 130, 132-135
palapraeputia group 100, 112, 123, 132
palapraeputia subgroup 100, 112, 124, 132
piedra95, 114, 116, 125-126, 128-129, 132-133,
135
polina95, 116-117, 125-126, 128-129, 132-133,
135
sequoia 95, 120-121, 125-126, 130-133, 135
sequoia subgroup 105, 120-122, 124-125, 127,
132
serpentinea95, 102, 114, 116, 122, 126, 128-134
serpentinea group 99, 100, 113, 116-117, 122-
123, 125, 129-130, 132, 134
serpentinea subgroup 105, 114, 119, 122, 124-
125, 127-130, 132
sierra95, 106-107, 126, 130-134
topanga95, 102, 109-110, 126-127, 129-134
topanga subgroup 100, 109-110, 124, 132
yosemitensis 95, 105, 126, 132-133
Calicina species group 99
Callisaurus 52
draconoides 23, 45^6, 48^9, 62-63
Calloselasma rhodostoma 195
Calotes 199
cristatelus 195
sp. 194-195, 197
versicolor 195, 197
Calymmaria 131
Carlowrightia arizonica 19
Cassia foment osa 192
Caudata 15
Ceanothus
cordulatus 12
greggi 12
leucodermis 12
Ceratostethus 244, 269
bicornis 245, 269
Cericidium microphyllum 12
Chaetochlamys 284
Chaetothylax 284
umbrosus 284
Charasia dorsal is 195
Chilmeniscus complex 62
Chiloglanis 151-166, 168-169, 175-176
batesii 151-163, 165, 168-169, 171-172
benuensis 151, 153, 157, 164, 167-168
brevibarbis 152, 155, 163-164
cameronensis 151, 153, 157, 162-163, 165, 168,
176
congicus 152, 156-157, 163
deckenii 152, 163-164
disneyi 151-159, 161-163, 165, 168, 172-173
harbinger 151-153, 155, 157, 161-163, 165, 175
lamottei 157
macropterus 162
marlieri 157
micropogon 151-152, 155, 162, 169, 172
neumanni 152
niger 151-163, 165, 173-174
niloticus 153, 162
normani 162
occidentalis 153, 160, 162-163, 172
polyodon 153, 157, 163
polypogon 151-164, 167-168
reticulatus 151-160, 162-166, 168
sanagaensis 151-163, 165, 169-170
sardinhai 163
savorgnani 163
voltae 151, 153-155, 161, 163-166, 168
Chilomeniscus 52
Chilomeniscus
cinctus 33-34, 45-46, 48-49, 63
punctissimus 48
savagei 48
stramineus 48
Chilomeniscus complex 48
Chilopsis linearis 12,14
Chionactis
occipitalis 44, 48, 63
occipitalis annulata 34
Chiropodomys gliroides 197
Chrysemys scripta 48-49, 6 1
Chrysopelea ornata 195
Clemmys marmorata 22, 44, 47-48, 58
Clidemia 137, 327
costaricensis 142
hammelii 137, 140-143,311
reitziana 142
340
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46
Clusiaceae 143
Cnemidophorus 52
ceralbensis 48, 60
deppei complex 48
hyperythrus 30-31, 45-46, 48, 60, 62
hyperythrus beldingi 30
hyperythrus hyperythrus 30
hyperythrus schmidti 30
labialis 30, 45^6, 48, 59, 62
multiscutatus 3 1
tigris 45
tigris multiscutatus 30
tigris tigris 30
Cobitis taenia 158
Coleonyx 52
switaki 23, 48, 62
variegatus 45
variegatus abbotti 23
variegatus peninsularis 23
variegatus variegatus 23
Coluber
"lateralis complex" 58
aurigulus 48-49
barbouri 48-49
flagellum 34, 45, 49
flagellum fuliginosus 34
flagellum piceus 34
lateralis 44, 48^49
later alis complex 48
lateralis lateralis 34
Colubridae 33, 193, 195, 198
Conostegia 327, 330, 332, 335
cinnamomea 148
fragrantissima 327-330, 335
montana 330
muriculata 327, 329-332, 335
orbeliana 327, 329, 333-335
pittieri 330
polyandra 333
rhodopetala 332
tenuifolia 332
volcanalis 335
Coryphaenoides armatus 93
Coturnix chinensis 196
Crocidura sp. 1 96
Crocodylus 6 1
Crossopalpus 179-180, 188
Crossostomus chilensis 238
Crotalus 52
insularis 46
mitchellii 45-46, 48-49, 62-63
mitchellii mitchellii 39
mitchellii pyrrhus 39
ndjCT- 45-^6, 48, 62
afrax afro* 38
flfrax complex 48
catalinensis 48, 52
cerastes 44, 48, 63
cerastes laterorepens 38
39, 45-46, 48, 62
39
enyofurvus 39
tortugensis 48
vinrfts 44, 48-49
v/rafts Ae//eri 39-40
Crotaphopeltis 200
Crotaphytus 52
insularis 45, 48, 62
insularis vestigium 24
Ctenophallus 269, 272
ctenophorus 245, 272
Ctenosaura hemilopha 48^9, 52, 60-61
Cyphomandra 314
Cyprinodontidae 247-248, 266
Dalechampia 3 1 4
Dentatherina 248, 261
merceri 244-245, 256, 261-263
Diadophis
punctatus 44, 48
punctatus similis 34
Dicliptera resupinata 79
Dipsadoboa 200
Dipsosaurus 52
rforsafo 45, 49, 62-63
dorsalis dorsalis 24
Diptera 179
Draco sp. 1 94
Drapetis 179, 188
(Crossopalpus) discalis \ 84
(Eudrapetis) discalis 1 84
arcuata 179
arnaudi 179, 181-183
assimilis 191
assimilis group 179-181
cerina 179, 181-184
destituta 179, 181-182, 184-185
</woi/w 179-182, 184-186, 188
divergens 180
rfwVfaa 180
exi'/w 180
infumata 179, 181-182, 185-186, 188, 190
ingrica 179
latipennis 179-182, 188-190
Ha/oz 188
na/ca group 1 80-1 8 1
simulans 179
Solaris 179, 181-182, 190-191
torulosa 179-182, 191-192
Dudleyaspp. 12-13
Dyschoriste 279, 282
mcvaughii 279-283
INDEX
341
pringlei 282
rubiginosa 282-283
Echinocactus sp. 29
Elaphe rosaliae 34-35, 48, 62
Elaphropeza 179-180
Elapidae 198
ElgariaSS, 61-62
cedrosensis 48-49
kingi 48, 62
multicarinata 44, 48-49
multicarinata webbi 3 1
panamintina 48, 62
paucicarinata 48-49, 62
Elgaria complex 48
Elytraria imbricata 79
Emoia cyanura 196
Empididae 179
Emydidae 22
Ensatina eschscholtzii 47-48, 58
Eonycteris spelea 195, 199
Ephedra californica 1 2
Eptesicus tenuipennis 1 94
Eridiphas 60
sleveni 52
s/ev/m'48^9, 61
Eriogonum fasciculatum 12
Euchilichthys 152, 163-164, 176
dybowskii 153, 163-164
royauxi 153, 163
Eudrapetis 179-180, 188
discalis 184
Eumeces 58, 61
#/&?/*/ 44, 48
gilberti rubricaudatus 29
lagunensis 48-49
skiltonianus 29-30, 44, 48-49
skiltonianus complex 48
skiltonianus skiltonianus 29
Euphorbiaceae 314
Fa/co sparverius 30
mutilata 196-197
sp. 196
acanthodes 12
sp. 22
Fourquiera splendens 1 2
chenopodiafolia 12
dumosa 12
Fuchsia 139
Gallusgallus 195
Gamasida 143
Gambelia
wislizenii 45
wizlizenii 49
wizlizenii copei 24
wizlizenii wizlizenii 24
195
vittatus 196
Gekkonidae 23, 195
Geochelone 6 1
Gesneriaceae 314
Gloxinia 314
Gochnatia hypoleuca 75
modestus 196
sp. 195
Gopherus agassizii 63
Graffenrieda gracilis 216
Gulaphallinae 244, 254, 266, 273
Gulaphallini 244, 254, 256-257, 264, 273-275
Gulaphallus 243-245, 249, 253-254, 256-258, 261-
263, 266, 268, 273, 275-276
amaricola 270
bikolanus 244-247, 261-263, 265-266, 268, 274-
275
ex/mms 244-247, 257, 262-263, 265-266, 268,
273-276
/afa/er 244-247, 262-263, 265-266, 274, 276
mirabilis 244-247, 249, 253, 262-263, 265-266,
274
panayensis 244-247, 256, 261-266, 274-275
Gymnelus 93
Henrya insularis 79
Heteromeles arbutifolia 12
Holographs 73, 76-77, 81
anisophylla 73, 75
argyrea 73, 79
caput-medusae 73-74, 76-79
ehrenbergiana 74
hintonii 74, 79
ilidfolia 74
Ieticiana73, 75, 77, 80-81
/?a///Vfo 74-75
par ay ana 73-75, 81
peloria 73-74
pueblensis 73-74
tamaulipica 74
tolantongensis 73-77
virgata 74
websteri 73-74
Hoplocephalus 202
Hygrophila 282
58,61
arenicola 49
cadaverina 20, 22, 38, 44, 48^9
exima 49
microscaphus 49
regilla22, 44, 48-^9, 61
342
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46
regilla deserticola 22
regilla hypochondriaca 21-22
Hylidae 20
Hypoatherina 26 1
bleekeri 254, 257
ovolaua 246
Hypsiglena 52
torquata 35, 45, 48
torquata deserticola 35
torquata klauberi 35
Iguanidae 23
Imantodes 202
cenchoa 202
Ischyropsalidae 134
Ixtlania 279, 284
acicularis 283
Japalura variegata 197
Juniperus 75
calif ornica 12
Jwtf/da 279, 284-285, 289-297
acicularis 283
adhatoda 296
caudata 285
fortunensis 289-293, 295-297
ixtlania 279, 281, 283-284
lindenii 285
readii 289-293, 295-297
refractifolia 289-290, 297
tabascina 279, 281, 283-285
veraguensis 289-290, 292-297
sect. Drejerella 285
sect. Leucoloma 284
sect. Tyloglossa 296
Laceraria 146, 148
Lacertidae 197
Laelapidae 143
Lamprolepis smaragdinum 195-196
Lampropeltis 52
catalinensis 48
g#M/MS 35, 45, 48
getulus californiae 35
getulus complex 48
getulus yumensis 35
w'nVfa 48
pyromelana 49
zomtfa 35, 44, 48-49, 59
zonata agalma 35
Laniatores 95
Larrea tridentata 12
Leandra 216
consimilis209, 216
Leiocassis 162-163
Leptotyphlopidae 32
Leptotyphlops 52
, 45, 48, 52
humilis cahuilae 32
humilis humilis 32
Libocedrus decurrens 1 2
Lichanura 52
roseofusca 32
trivirgata 32, 45^6, 48-49, 62-63
trivirgata gracia 32-33
trivirgata roseofusca 32-33
trivirgata trivirgata 32-33
Lycenchelys 86
"D" 91, 235
"E" 88, 234
crassiceps 222, 228
monstrosa 238
Lycodapus mandibularis 93
Lj*ofes 82, 221,228
atlanticus 228
brachycephalus 82, 222-223, 240
crassiceps 228
diapterus 93
macrops 228
mucosus 228
sp. 88, 91, 228, 234-235
Lycodryas 202
Manacopus 244, 254, 274
./&/«/«• 245, 274
MayAira bulbiceps 82, 84, 223, 227
Melanostigma 226
Melanotaenia 248, 261
qffinis 246
Melastomataceae 137, 209, 219, 299-301, 311, 314,
327
Mexacanthus 284
A//a»iifl 137, 150, 209, 214, 216, 327
arbor icola 139
Irenes/7 2 11-2 12
calocoma 137, 144-146, 148
centrodesma 148
consimilis 2 1 6
dissitiflora 137, 146-148
fulvostellata 214
grayumii 209-2 12
ibarrae209, 212-214
iteophylla 209, 214-217, 219
Hgulata209, 212, 216-217
peltata209, 216-219
rupticalyx 146
wagneri 146
Miconieae 150, 216, 327
Micrurus fulvius 198
Mirophallus 274
bikolanus 245, 274-275
Mochokidae 151, 164
Molossidae 194
Monochaetum 148
Motacillasp. 195, 197
Motacillidae 195, 197
INDEX
343
Mugiloidea 247
Muridae 195, 197
Mus musculus 196
Muscicapidae 197
Mystus 163
Nectarina famosa 194
Nemastomatidae 134
Neostethidae 243
Neostethinae 254, 266
Neostethini 243, 254, 256-257, 264, 266, 269, 275
Neostethus 243, 249, 253-254, 256-257, 259, 261-
262, 266-267, 269-270, 272, 275
(Neostethus) 254
(Neostethus) amaricola 245, 270
(Neostethus) lankesteri 245, 269
(Neostethus) siamensis 245, 269
(Neostethus) villadolidi 245, 271
(Sandakanus) 254, 270
(Sandakanus) borneensis 245, 270
(Sandakanus) coronensis 245, 270
(Sandakanus) panayensis 245, 275
(Sandakanus) zamboangae 245, 271
amaricola 243, 245-247, 258-26 1 , 265-266, 269-
271
bicornis 243-247, 258-261, 265-266, 269, 273
borneensis 244-245, 247, 254, 257-259, 26 1 , 263-
266, 269-270, 275-276
coronensis 254, 271, 275
ctenophorus 244-247, 258, 262, 264, 266, 269,
272
lankesteri 243, 245-247, 252, 258-261, 264-266,
268-269, 275-276
palawanensis 244-247, 258-259, 262, 265-266,
269, 271-272
panayensis 254, 275-276
robertsi 243-247, 258-259, 262, 265-266, 269,
272-273, 275
siamensis 269, 275
thessa 244-247, 257-266, 269, 272
villadolidi 244-247, 258-261, 265-266, 269-271,
276
zamboangae 244-247, 250, 254, 257-259, 261,
265-266,269,271
Neozoarcinae 226
Nerodia 60
valida 48^9, 52
\allida 61
Odontonema auriculatum 79
Odontoneminae 284
Olneyatesota 12, 28
Ololaelaps 137
sp. 143
Onagraceae 139
Opiliones95, 134
Opuntia
littoralis 12
prolifera 12
sp. 27
spp. 12
Orchestina 131
Orchidaceae 314
Oryzomys 307
Pachycara 82, 84-89, 91, 221, 223, 226-227, 231, 233,
235, 238-240
brachycephalum 222, 227, 240-241
bulbiceps 84-87, 92-93, 221-222, 224-225, 227-
228, 241
crassiceps 222-223, 227-230
crossacanthum 221, 224, 226-227, 230
gymninium 82, 84, 86-89, 92-93, 222, 225, 227,
234-235
lepinium 82, 84, 86-87, 90-91, 93, 222, 227, 235
mesoporum 221, 224-226, 235, 237, 241
obesa 84, 222-223, 227-228
obesum 82, 84, 86
pammelas 221, 227, 235, 239
rimae 221, 223, 225-226, 233, 235-236, 241
shcherbachevi 221-222, 226, 233-234, 241
sp. 85
sulaki 221, 225, 227, 231, 238, 241
suspectum 84, 86, 88-89, 222, 226, 235
suspectus 87
Pachycarichthys 223
Pachycereus pringlei 12, 22
Pachychara 223
obesa 84
Pachycormus discolor 1 2
Pangasius 163
Parabeloniscus 122
Parabrotulidae 226
Parabuteo unicinctus 21
Passer sp. 195
Passeridae 195
Pelobatidae 16
Pentanychidae 134
Peramelidae 196
Percesoces 247
Peromyscus 307
sp. 34
Petrosaurus 24, 52
mearnsi 24-25, 45^6, 48
repens 48
thalassinus 48
Petrosaurus complex 48, 62
Phalangodidae 95-96, 98
Phalangodinae 98, 123
Phallostethidae 243, 246-248, 254, 264-266, 275
Phallostethinae 243, 254, 256, 264-267
Phallostethini 243, 247, 254, 256-257, 264-265, 267,
275
Phallostethus 243-244, 250, 254, 256-257, 262, 264-
267, 275
dunckeri 243, 245-247, 257, 262, 265, 267, 271
344
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46
Phasianidae 195-196
Phenacostethini 264
Phenacostethus243, 249, 254, 256-257, 262-264, 267,
275
position 243-247, 257, 262-264, 267-268
smithi 243, 245-247, 249, 251, 257, 262, 264,
267-268
thai 268
trewavasae 243-247, 257, 262, 264, 267-268
Philautus sp. 195
Phrynosoma 52
cerroense 48
coronatum 25, 46, 48, 60, 62
coronatum complex 48
coronatum schmidti 25
coronatus 45
mcallii 25, 48^9
platyrhinos 25, 44, 48
platyrhinos calidiarum 25
Phucocoetes suspectus 82, 88, 235
Phyllodactylus 52, 60
angelensis 48
apricus 48
bugastrolepis 48
homolepidurus nolascoensis 48
nocticolus 23, 45-46, 48
nocticolus nocticolus 23
nocticolus sloani 23
partidus 48
santacruzensis 48
tt' 48
Phyllodactylus complex 62
Phyllorhynchus 52
decurtatus 45, 48^19, 62-63
decurtatus decurtatus 35
decurtatus perkinsis 35
Physocarpus 188
Pilocosta 2 19-220
campanensis 2 1 9-220
2 19-220
oerstedii 219
oerstedii campanensis 209, 219
Pwzws 131
P//2WS
Jeffrey i 12
lambertiana 12
monophylla 12
murrayana 10
pinceana 75
quadrifolia 12
Pituophis
melanoleucus 45, 49
melanoleucus affinis 36
melanoleucus annectens 36
melanoleucus bimaris 36
Platanus 190
racemosa 12
Plectrostethus 257, 269, 272
palawanensis 245, 269, 271, 275
palawensis 27 1
Plesiositagra cucullatus 194
Plethodontidae 15
Polynemoidea 247
Popw/us
balsamifera 188
fremontia 12
sp. 35
tremuloides 10
Praomys sp. 197
Prosopis juliflora 12, 14
Psammodynastes pulverulentus 195
Pseudemys script a 52
Pseudomugil 26 1
signifer246,256,26\, 263
Pseustes poecilonotus 202
Pteropodidae 195
Ptychosoma 122
Pyxidanthus 300
dumosa 12
opaca 75
spp. 12
turbinella 12
22
awrora 22, 44, 47, 58
aurora draytoni 22
boylii 22, 44, 48
limnocharis 194-195
sp. 196
Ranidae 195-196
exulans 196
rattus 196
sp. 195
Reithrodontomys 307
Rhacophoridae 195
Rhamnus californica 1 2
Rhinocheilus
lecontei 45
lecontei lecontei 36
leconti 49
.R/zws ovata 1 2
7?osa californica 1 2
RuelliaZl, 282
albiflora 79
petiolaris 28 1
horsfieldi 195
Salientia 16
Sa&cspp. 12, 14
INDEX
345
Salvadora
hexalepis 36, 45, 49
hexalepis hexalepis 36-37
hexalepis klauberi 36-37
hexalepis virgultea 37
Salvia
apiana 12
pachyphylla 12
Sandakanus 269
angustus 48, 52
grandaevus 48
spp. 49
Sa/or complex 48, 61
Sauna 23
Sauromalus 52
ater 48
australis 25, 48
hispidus 48
klauberi 48
o&esttf 25-26, 45^6, 48, 63
obesus obesus 25-26
slevini 48
var/ws 48
Sauromalus complex 48, 62
Saxicoloides fulica 197
Scaphiella 131
Scaphiopus 47
couchii 16, 48
hammondii 16, 44, 48
Sceloporus 52
graciosus 26-27, 44, 48
graciosus vandenburghianus 26
hunsakeri 48
/«#/ 48
magister 26-27, 45
magister complex 48, 57, 62
magister lineatulus 48
magister monserratensis 48
magister rufidorsum 26-27, 48
magister uniformis 26-27, 49
magister zosteromus 48
occidentalis 26-27, 44, 48
occidentalis biseriatus 27
orcMtt/ 27, 45, 48
orcutti complex 48, 62
Schaphiopus couchii 45
Schaueria 286-287
calycobractea 279, 286-287
calycotricha 287
cordatus 286
parviflora 279, 281, 283, 286-287
Scincidae 29
Scotinomys 307
Scotolemon 122
balearicus 122
espanoli 122
Serpentes 32
Stem* 106
Siphonoglossa 8 1
Sitalces californica 95-96
Sitalcina 95-98, 106, 122, 131, 134-135
borregoensis 97
californica 97-98
cloughensis 1 1 2
cockerelli 121, 131, 135
digitus 106
ensata 1 1 7
kaweahensis 108
keenea 1 10
macula 1 1 7
madera 97
mariposa 98, 105
minor 1 1 3
morroensis 1 10
palapraeputia 112
p/«/ra 1 1 6
polina 1 1 6
scapula 97
sequoia 121
serpentinea 1 1 6
106
6reva 1 1 9
swra 97
topanga 109
yosemitensis 105
Solanaceae 314
Solenophallus 269, 272
ctenophorus 269, 272, 275
fAassa 245, 269, 272, 275
Sonora 52
bancroftae 37
mosaueri 37
semiannulata 37
semiannulata bancroftae 45-46, 48
semiannulata mosaueri 48
semiannulata semiannulata 45, 49
Spathiphyllum 314
Stenandrium 73
Streblacanthus 286
parviflorus 279, 286
Sturnidae 196
Synodontis 155, 164
Tabascina 279, 285
lindenii 284
Tachydromiinae 179
Tachydromus sexlineatus 195
Tadarida sp. 194
Tantilla 52
planiceps 37, 45, 48
Teiidae 30
Teleostei 82, 221
Telescopus 200-202
rf/zara 200
semiannulatus 200
346
PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 46
Testudinata 22
Tetramerium
nervosum 79
oaxacanum 8 1
tenuissimum 79
Texella 122
Thamnophis
digueti 48-49
elegans 37-38, 44, 48, 198
elegans hueyi 37
hammondii2\, 38, 44, 48-49
hammondii complex 48
marcianus 63
sirtalis 198
spp. 37
Thermarces 221, 226, 240
cerberus 226
Titiotus 131
137, 150
croatii 137, 148-150
symphyandra 150
ra 140, 299-305, 314, 317-318, 320, 323, 325
acuminata 320
albertieae 304
brenesii 302
caliginosa 299, 304, 317, 322
calycularis 303
caudata 320
co«tofa304, 317-318, 322
crassifolia 302, 304, 317-318, 322
dodsonorum 320
fragrantissima 299, 305, 308, 318-320, 325
hexandra 299, 303-305, 313, 317, 320-322, 325
parasitica 300
parvifolia 304, 322-323
p/tt/m 302, 320
storkii 303
suaveolens 299, 305, 313, 320, 323-325
Tragulidae 195
Tragulus javanicus 195
Triaenonychidae 134
Trimorphodon 52, 202
biscutatus 45, 49, 202
biscutatus lyrophanes 38
biscutatus vandenburghi 38
Trionyx spiniferus 63
62
wotata 28, 48^49
notata notata 28
Urosaurus 52
graciosus 44, 48
graciosus graciosus 28
fcAte/ai 28, 48
microscutatus 28, 45, 48
nigricaudus 48
ornatus 63
Urosaurus complex 48
t/so/z/a 131
[7/a52
antigua 48
nolascensis 48
palmeri 48
squamata 48
stansburiana 45, 48
stansburiana elegans 28
stellata 48
[/to complex 48
Valdesia 300
Varanus salvator 195
Vespertilionidae 194
Viperidae 195
Vipiridae 38
Washingtonia filifera 12
Xantusia 60
gigilis 48
henshawi 45-46, 48, 59, 62
henshawi bolsonae 49
henshawi henshawi 28
spp. 47
v/g/fo 44, 49
v/£//w arizonae 49
v/g/7/5 v/gz7/5 29, 49
wgiV/s wigginsi 29
Xantusiidae 28
schidigera 12, 27
spp. 12
12-13
Zoarcidae 82, 221, 235
Zwma 131