Full text of "Contributions in science"

See other formats

tl. Number 450

-&K 28 March 1995

(Wf

Contributions
in Science

V&tl LI JHHMHHBMHWHIMEHHBBjnHinnaBBnHWni

Tertiary Sawflies of the Tribe Xyelini
(Insecta: Vespida = Hymenoptera: Xyelidae)
and Their Relationship to the
Mesozoic and Modern Faunas

Alexander P. Rasnitsyn

Natural History Museum of Los Angeles County

Serial

Publications

of THE
Natural History
Museum of
Los Angeles
County

Scientific

Publications

Committee

James L. Powell, Museum President
Daniel M. Cohen, Committee Chairman
Brian V. Brown
Kenneth E. Campbell
Kirk Fitzhugh
Edward C. Wilson
Robin A. Simpson, Managing Editor

The scientific publications of the Natural History Mu-
seum of Los Angeles County have been issued at irregular
intervals in three major series; the issues in each series are
numbered individually, and numbers run consecutively,
regardless of the subject matter.

# Contributions in Science, a miscellaneous series of tech-
nical papers describing original research in the life and
earth sciences.

# Science Bulletin, a miscellaneous series of monographs
describing original research in the life and earth sci-
ences. This series was discontinued in 1978 with the
issue of Numbers 29 and 30; monographs are now
published by the Museum in Contributions in Science.

# Science Series, long articles and collections of papers
on natural history topics.

Copies of the publications in these series are sold through
the Museum Book Shop. A catalog is available on request.

The Museum also publishes Technical Reports, a mis-
cellaneous series containing information relative to schol-
arly inquiry and collections but not reporting the results
of original research. Issue is authorized by the Museum’s
Scientific Publications Committee; however, manuscripts
do not receive anonymous peer review. Individual Tech-
nical Reports may be obtained from the relevant Section
of the Museum.

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

Printed at Allen Press, Inc., Lawrence, Kansas
ISSN 0459-8113

Tertiary Sawflies of the Tribe Xyelini

(Insecta: Vespida = Hymenoptera: Xyelidae)
and Their Relationship to the
Mesozoic and Modern Faunas

Alexander P. Rasnitsyn1

ABSTRACT. Ten Tertiary species of Xyelini are currently known. One of them, Enneoxyela ? cenozoica
(Zhang, 1989) comb, nov., from the Middle Miocene of China, is a Late Tertiary relic of a mid-Mesozoic
group, being a member or at least a close relative of the otherwise Late Jurassic genus. All other Tertiary
Xyelini are of Oligocene age and belong to the genus Xyela Dalman, 1819. Subgenus Pinicolites Meunier,
1920, stat. nov. is represented by only one species, X. (P.) graciosa (Meunier, 1920). Subgenus X. {Xyela)
is represented by eight fossil species, including X. (X.) magna Statz, 1936, which belongs to the X. minor
group and is most closely related to living North American species. The others are all attributed to the
X. julii group. Four are extinct — X. (X.) latipennis Statz, 1936, X. (X.) angustipennis Statz, 1936, X. (X.)
florissantensis sp. nov., and X. (X.) micrura sp. nov. — whereas X. (X.) cf. menelaus Benson, 1961, and
X. (X.) cf. julii Brebisson possibly represent the two living European species. All known Tertiary Xyela
come from the highest Upper Oligocene Rott Formation of Germany, except X. (X.) florissantensis , from
the Lower Oligocene Florissant Formation of Colorado, USA. All known Tertiary Xyelini were buried
in lacustrine, tuffaceous deposits.

INTRODUCTION

The sawfly family Xyelidae has special significance
for the history of hymenopterous insects (order
Vespida; see Rasnitsyn, 1988, and references there-
in for name justification). It appeared in the fossil
record as early as the Middle or early Late Triassic,
whereas other hymenopteran fossils are unknown
before the earliest Jurassic (Rasnitsyn, 1988). Xyelid
morphology suggests an ancestral position for the
family with respect to other Flymenoptera, al-
though this claim is not entirely accepted (compare
Konigsmann, 1976, and Rasnitsyn, 1980).

The three subfamilies of Xyelidae are compa-
rable in size and diversity. The ancestral Archex-
yelinae Rasnitsyn, 1964, is known only from the
Triassic, while the Macroxyelinae Ashmead, 1898,
and Xyelinae Newman, 1834, range from the Early
Jurassic to the present. During Mesozoic time, the
Xyelidae was a prominent hymenopteran family,
especially in areas with relatively cooler climates,
where it often was dominant.

In the Tertiary (or perhaps the Late Cretaceous;
there are insufficient data for that period), the pat-
tern changed radically. Xyelidae became a rare group
with a poor fossil record, confined mostly to the
temperate regions of the Northern Hemisphere.
Only 17 Tertiary specimens are known worldwide.

1. Paleontological Institute, Russian Academy of Sci-
ences, 117647 Moscow, Russia.

Contributions in Science, Number 450, pp. 1-14
Natural History Museum of Los Angeles County, 1995

Thirteen belong to Xyela Dalman, 1819, s.l; of
these, 10 were described by Statz (1936) (paralec-
totype 4019 of Xyela latipennis is not a hymenop-
teran but a possible trichopteran) and 3 were de-
scribed by Meunier (1920) and Zhang (1989) and
in the present paper. Three are members of Me-
gaxyela Ashmead, 1898: one was described by Brues
(1908; redescribed by Zhelochovtzev and Rasnit-
syn, 1972), the second by Zhang (1989), and the
third is an undescribed species of Megaxyela from
the Lower Miocene of Sikhote-Alin, Maritime
Province, kept in the Paleontological Institute, Rus-
sian Academy of Sciences, Moscow. One specimen
belongs to Xyelecia Ross, 1932 (Zhang, 1989). The
Lower Cretaceous deposits of Siberia and Mongolia
have yielded more than 80 xyelid specimens, now
housed at the Paleontological Institute in Moscow.

Unlike the Jurassic and Cretaceous xyelid faunas,
which were composed of comparable diversities of
both Xyelinae and Macroxyelinae, the Tertiary fau-
na was strongly dominated by Xyelinae, particularly
Xyela Dalman. There are only three known species
of Megaxyela Ashmead and one of Xyelecia Ross
belonging to Tertiary Macroxyelinae. The domi-
nance of Xyela among the Tertiary Xyelidae is sim-
ilar to the contemporary fauna which comprises a
total of 48 xyelid species, 31 of which belong to
Xyela (Smith, 1978).

The distribution of the fossils is enigmatic. The
larval stage of Xyela develops within staminate cones
of pines and feeds on the unripe pollen. The adults
readily visit other, flowering plants for pollen, but

ISSN 0459-8113

nevertheless they spend much of their time on pine
trees. These habits have persisted since at least the
Early Cretaceous (Krassilov and Rasnitsyn, 1982).
Xyela are intimately associated with the genus Pi-
nus Linne, 1753, both genera being widespread over
the Northern Hemisphere. Xyela is particularly
abundant and diverse in southwestern North Amer-
ica (Burdick, 1961; Rasnitsyn, 1971), where pine
trees are also common and represented by a number
of species. Still, the Tertiary fossil record is poor
for Xyela in the above region. A single Xyela spec-
imen has been found among the tens of thousands
of fossils collected at Florissant, Creed, and Green
River. Even more unexpected, the only known Ter-
tiary fauna rich in Xyela comes from the Rott For-
mation (latest Oligocene diatomite deposits accu-
mulated in a mountain lake at Rott near Bonn in
Germany), which is poor in Pinus fossils (Weyland,
1937, 1948). Nevertheless, the Rott hymenopteran
assemblage is even richer in xyelids than most Cre-
taceous ones, being represented by 11 specimens
among 14 sawflies (79%) and 76 solitary hymenop-
terans (14.5%; figures from Statz, 1936). It is un-
likely that the rare local pine trees could have housed
all of them, and the diversity of Xyela suggests a
considerable diversity of host pine tree species. In-
deed, the extant closely related species of Xyela
rarely exploit the same host plant species (Burdick,
1961; Rasnitsyn, 1965, 1971). The insects may have
been transported to the Rott lake by winds from
nearby mountains as has been described for extant
Xyela (Fridolin, 1936).

The above considerations do not fully explain
the composition of the Rott assemblage, however,
because there are many other mountain lake de-
posits with rich insect assemblages but few, if any,
Xyela. Yet this is not the most perplexing aspect
of the fossil record of the group. Even more difficult
to explain is the absence of Xyela among hundreds
of thousands of insect inclusions in Baltic amber,
a fossil pine resin. Their absence from Baltic amber
could not be because these insects are able to escape
entrapment. Indeed, I have identified two specimens
of X. ussuriensis Rasnitsyn in spruce resin collected
at the Sikhote-Alin Mountains (cf. Zherikhin and
Sukacheva, 1989).

Also enigmatic is the appearance of a represen-
tative of the otherwise Late Jurassic genus En-
neoxyela Rasnitsyn, 1966, or a closely related ge-
nus, in the Tertiary (Middle Miocene) of China (see
below).

The above review of the Xyela fossil record shows
that it deserves exploration. This became possible
after a visit to the Natural History Museum of Los
Angeles County (Los Angeles, California) and the
National Museum of Natural History (Washing-
ton, D.C.) in 1989-1990. The first institution keeps
the Georg Statz collection of the Rott insect fossils
(Sphon, 1973); the latter has the only specimen of
Xyela from the Lower Oligocene of Florissant, Col-
orado.

2 ■ Contributions in Science, Number 450

SPECIMENS EXAMINED

Material utilized in the study is from the following col-
lections: Invertebrate Paleontology Section, Natural His-
tory Museum of Los Angeles County, Los Angeles, Cal-
ifornia (LACMIP); Arthropoda Laboratory, Paleontological
Institute, Russian Academy of Sciences, Moscow, Russia
(PIN); Department of Entomology, U.S. National Mu-
seum of Natural History, Washington, D.C. (NMNH);
and Linqu Paleontological Museum, Linqu, Shandong
Province, P.R. China (LPM) (only a photograph of the
Chinese specimen was examined).

TAXONOMY

I am following a moderately splitting approach to Xyela
taxonomy developed in my earlier publications (Rasnit-
syn, 1965, 1971). I consider the widespread synonymi-
zation of allopatric species by Benson (1961, 1962) as
possibly correct but premature given the present state of
our knowledge. In my opinion, more data on Xyela mor-
phology and distribution should be accumulated before
these geographically disjunct populations with slight mor-
phological differences are lumped together. Otherwise we
risk mixing and losing important information.

Family Xyelidae Newman, 1834
Subfamily Xyelinae Newman, 1834
Tribe Xyelini Newman, 1834
Genus Xyela Dalman, 1819
Subgenus Pinicolites
Meunier, 1920, stat. nov.

Pinicolites: Meunier, 1920: 896; Burdick, 1959: 121;
Rasnitsyn, 1971: 192. Type species: Pinicolites
graciosus Meunier, 1920: 896; monobasic.
Pleroneura (partim): Statz, 1936: 262; Sphon, 1973:
60.

DIAGNOSIS. Pinicolites (Figs. 1-4) similar to X.
( Mesoxyela Rasnitsyn, 1965) (Fig. 4), Xyela (Xyela)
(Figs. 5-20), and Pleroneura Konow, 1897. Differs
from the largely Mesozoic genera Eoxyela Rasnit-
syn, 1965, Enneoxyela, and Spathoxyela Rasnitsyn,
1969 in having R sinuate resulting in costal space
widened at midlength, and probably also in having
mesonotum impunctate (in that respect similar ad-
ditionally to Spathoxyela). Similar to Xyela (Xyela)
and X. (Mesoxyela) and differing from Pleroneura
in having thin and flat (saw-like instead of needle-
like) ovipositor and short or lacking 1 r-m in fore-
wing (RS and M scarcely or not at all separated).
Similar to X. (Mesoxyela) and Pleroneura in having
wide pterostigma. Similar to Xyelisca Rasnitsyn,
1969, Pleroneura, and X. (Mesoxyela) and differing
from Eoxyela, Enneoxyela, Spathoxyela, and X.
(Xyela) in having antennal funicle shorter than seg-
ment 3. Similar to all Xyelini except Xyela s.str. in
having free SC stalk in forewing and, except Xyela
s.str. and Pleroneura, in having fore SC branch long,
reaching level of RS base. Similar to Pleroneura and
unlike all other Xyelini in having 2r cell short in
the forewing. Similar to many X. (Xyela) and dif-
fering from Pleroneura in color pattern of meso-

Rasnitsyn: Tertiary Sawflies of the Tribe Xyelini

Figure 1. Line drawing of Xyela ( Pinicolites ) graciosa (Meunier, 1920) traced after photograph of holotype, with
forewings displaced to show venation of the hind pair. Scale bar = 1 mm.

notum, which, instead of being almost uniformly
dark, is light with dark spots marking areas of mus-
cle attachment. Similar to all Xyelini except Pie -
roneura in having hindwing lacking free apex of Ax
(unknown for Xyelisca).

SPECIES INCLUDED. Type species only.

SYSTEMATIC AND PHYLOGENETIC PO-
SITION. The characters used to identify the sys-
tematic and phylogenetic position of Pinicolites
within the tribe Xyelini are presented in the fol-
lowing list; their distribution is shown in Table 1:
1. Antennal funicle: 0-short (shorter than 3rd seg-

Contributions in Science, Number 450

Rasnitsyn: Tertiary Sawflies of the Tribe Xyelini fl 3

Figure 3. LACMIP 4010: Wing venation combined from
left and right wings. Scale bar = 1 mm.

ment unless the latter is short itself), 1-longer than
3rd segment. A short antennal funicle is considered
plesiomorphic because it is found in less advanced
Xyelini [Xyelisca, Xyela ( Mesoxyela )], as well as in
Triassic Archexyelinae as exemplified by Dinoxyela
armata Rasnitsyn (Rasnitsyn, 1969, Fig. 24).

2. Mesonotum: 0- -punctate, 1 — impunctate. A
punctate mesonotum is probably plesiomorphic be-
cause Triassic and Jurassic Xyelidae all retained this
character state.

3. Mesonotum: 0— uniformly dark, 1 — light with
dark spots marking muscle attachment sites.

4. Forewing R: 0— sinuate before RS base, 1 —
straight or gently curved, at most slightly bent at

Figure 4. LACMIP 4010: Photograph of impression. Scale bar = 1 mm.

4 ■ Contributions in Science, Number 450

Rasnitsyn: Tertiary Sawflies of the Tribe Xyelini

Table 1. Data matrix for groundplan characters of taxa
of Xyelini as discussed in the text, with Liadoxyelini
taken as an outgroup.

1.11111
123456789 012345

. Ancestor

000000000

000000

Liadoxyelini

100010000

01????

Eoxyela

000100000

00010?

Enneoxyela

000100000

000010

Spathoxyela

110100000

000010

Xyelisca

O000000?0

00????

Mesoxyela

010000000

00010?

Xyela

111001110

000000

Pinicolites

011000011

000020

Pleroneura

010000101

101121

RS base. A sinuate R is considered piesiomorphic
because it is characteristic for the oldest xyeiid
group, the Triassic Archexyelinae. It is also found
in most of the predominantly Jurassic Liadoxyelini
Rasnitsyn, 1966 (Xyclinae), and the majority of Ma
croxyelinae (Rasnitsyn, 1969).

5. Forewing SC: 0- -free, 1— appressed to R except
apically. A free SC is considered piesiomorphic be-
cause it is a groundplan character state for Ptery-
gota. It is present also in the majority of Xyelidae,
including most Archexyelinae, Macroxyelinae, and
Liadoxyelini.

6. Forewing SC: 0 — reaching, 1 - far from reaching
level of RS base. The first character state is probably
piesiomorphic because it is present in most Paleo-
zoic insects and most Xyelidae, including most Ar-
chexyelinae, Macroxyelinae, and Liadoxyelini.

7. First abscissa of forewing RS: 0 — much longer
than, 1 — subequal to 1st abscissa of M. A long first
abscissa of RS is considered piesiomorphic because
archetypically holometabolous insects have RS and
M connected by an r-m crossvein positioned distad
of RS and M bases. Only secondarily have these
veins become fused, and when becoming longer,
the fusion often results in shortening of the basal
abscissa of RS.

8. Pterostigma: 0- -narrow, 1— wide. The pteros-
tigma is narrow in all Triassic and Jurassic Xyelidae
except the aberrant Jurassic Lydoxyela Rasnitsyn,
1966 (Liadoxyelini). Thus, it is considered here as
the piesiomorphic character state.

9. Forewing cell 2r: 0— -long, 1— -short. A long 2r
is found in all Xyelidae except Pinicolites and Ple-
roneura and, thus, is considered piesiomorphic.

10. Forewing crossvein Ir-m: 0 — short or absent,
1— -long (cells Ir and Im-cu distant). A long Ir-m
widely separating RS and M is undoubtedly the
groundplan character state in holometabolous in-
sects. However, the same is not necessarily true in
the Hymenoptera, in which Pleroneura is practi-
cally unique in having 1 r- m long. Another example

Contributions in Science, Number 450

Figure 5. Cladogram calculated using Hennig86 from
the matrix displayed in Table 1. When present below a
node, the numbers indicate syn- and autapomorphies list-
ed in the text.

is the highly modified wing venation of the horntail
genus Sirex Linne. A reversion may have taken place
here, with the result that the short or absent 1 r-m
is piesiomorphic, and Pleroneura and Sirex are con-
sidered homoplastically apomorphic instead of
sympl esiomorpbic.

11. Forewing crossvein Im-cu: 0 — long, 1— -short.
A long Im-cu correlates with the less angulated Cu,
which is evidently piesiomorphic for the winged
insects.

12. Hindwing A.: 0 — lacking free apex, 1- — with
free apex. This case is similar to that of the crossvein
Ir-m (No. 10, above). A free As is undoubtedly a
groundplan character state for the winged insects
in general and for Xyelidae in particular (found in
the Triassic Archexyelinae; Rasnitsyn, 1969, figs.
36, 39, 41). However, it is not found in higher
Xyelidae (Xyelinae + Macroxyelinae) except in Ple-
roneura, which is otherwise a relatively specialized
form, and it is unlikely that this character state was
inherited directly from the Triassic ancestor. In-
stead, I hypothesize that the free A, apex has been
re-acquired here.

Figure 6. Phylogenetic tree of the subfamily Xyelinae.
Black boxes show the fossils recorded for the respective
time intervals, double lines designate the hypothesized
(not confirmed by fossils) existence of taxa, and thin lines
indicate ancestry. Geochronological units are abbreviated
as follows: J, = Early Jurassic, J2 = Middle Jurassic, J3 —
Late Jurassic, K, = Early Cretaceous, K2 = Late Creta-
ceous, P, = Paleocene, P2 = Eocene, P3 = Oligocene, Nj
— Miocene, N, = Piiocenc, R = the present time.

Rasnitsyn: Tertiary Sawflies of the Tribe Xyelini ■ 5

Figure 7. Xyela ( Mesoxyela ) mesozoica Rasnitsyn, 1965, PIN 3064/1924; Eastern Siberia, upper Vitim River in 45
air km upstream from Romanovka Village, Baissa Locality, Bed 31; Neocomian, Lower Cretaceous, Zaza Formation.

Figures 8-15. Line drawing of ovipositor of lectotype
of Xyela (Xyela) latipennis Statz, 1936 (8), paralectotypes
of X. (X.) angustipennis Statz, 1936 (LACMIP 4012 and
4014, respectively) (9, 10), X. (X.) sp. cf. menelaus Benson,
1960 (11 = LACMIP 4017, 12 = LACMIP 4016), X. (X.)
cf. julii (Brebisson, 1818) (LACMIP 4018) (13), holotype
of X. (X.) micrura sp. nov. (14), and holotype of X. (X.)
magna Statz, 1936 (15).

13. Ovipositor: 0 — saw-like, 1— needle-like. A saw-
like ovipositor probably is plesiomorphic because
it is the only ovipositor type found in the Mesozoic
Xyelidae. The mechanically more efficient needle-
like ovipositor is typical for hymenopterans boring
shoots and wood. It is uncommon among those
developing in the pine tree staminate cones, such
as Xyela ( Mesoxyela ) and X. (X.) alpigena , X. (X.)
concava, X. (X.) linsleyi, and X. (X.) longula groups.

14. Ovipositor: 0 — downcurved, 1 — straight, 2 —
upcurved. An upcurved ovipositor is considered the
most apomorphic in the transformation series be-
cause it is not found in the Mesozoic Xyelidae and
is known only for a few Cenozoic forms [Plero-
neura, Pinicolites , and X. (X.) concava group]. For
the two other character states, a straight ovipositor
is present in Triassic Archexyelinae, while the oth-
erwise less advanced Xyelini, including X. (Mesox-
yela), have it downcurved. Eoxyela is among the
genera with a straight ovipositor.

15. Larva: 0 — feeding on pollen in staminate cones
of pine trees, 1 — boring fir shoots. The former char-
acter state is considered plesiomorphic for the rea-
sons I have discussed previously (Rasnitsyn, 1980,
1988). For fossils, the forms with a saw-like ovi-
positor are hypothesized to feed in the staminate

6 ■ Contributions in Science, Number 450

Rasnitsyn: Tertiary Sawflies of the Tribe Xyelini

Figures 16 19. 16. Xyela ( Xyela ) latipennis Statz, 1936, photograph of lectotype. 17-19. Xyela ( Xyela ) angustipennis

Statz, 1936, photographs of LACMIP lectotype 401 1 (17), LACMIP paralectotype 4012 (18), and LACMIP paralectotype

Contributions in Science, Number 450

Rasnitsyn: Tertiary Sawflies of the Tribe Xyelini ■ 7

cones, whereas the shoot-borers normally have a
needle-like ovipositor.

Based on the data from Table 1, a cladogram has
been derived using the “ie” option of Hennig86
(Farris, 1988). When no a priori weighting is ap-
plied, the result is an overflow of trees (tree length
25, consistency index 0.64, retention index 0.50)
and a completely unresolved consensus tree, show-
ing no subclades except the terminal groups. In
contrast, when a weight of 2 is applied a priori to
character 4, and the only multistate character (No.
14) is coded as nonadditive, the result is a single
cladogram (tree length 24, consistency index 0,70,
retention index 0.61). This cladogram essentially
agrees with my intuitive assessment of the relation-
ships within the group (Fig. 5). The phylogenetic
tree (Fig. 6) is modified from the cladogram to re-
flect the geological succession of the taxa involved,
as well as the fact that some of them seemingly
lack autapomorphies. Until autapomorphies are
found, these taxa are considered to be paraphyletic.

I consider paraphyletic taxa to be legitimate (Ras-
nitsyn, 1987, 1988, and references therein) and feel
no need to discard Xyelisca, Enneoxyela, Mesox-
yela, and Xyela s.l. or to reduce any of them to a
parataxon plesion (Patterson and Rosen, 1977). I
feel also that Pinicolites is roughly equidistant phe-
netically from Xyela s.str. and Mesoxyela and more
distant from Pleroneura. That is why I prefer to
reduce Pinicolites to subgeneric rank rather than
to make Mesoxyela a full genus or to lump Ple-
roneura under Xyela s.l. Either of the latter two
decisions would obscure the close similarity among
the three subgenera of Xyela and the phenetically
distant position of Pleroneura.

Pinicolites graciosus Meunier, 1920
Figures 1-4

Pinicolites graciosus Meunier, 1920: 896, figs. 4, 5;

Burdick, 1959: 121; Rasnitsyn, 1969: 38, fig. 60.
Pleroneura graciosa: Statz, 1936: 262, Abb. 1.

MATERIAL EXAMINED. Published photo-
graph of the holotype (Meunier, 1920, fig. 5), and
LACMIP 4010 described and figured by Statz (1936);
Rott Formation, Rott near Bonn, Germany, Latest
Oligocene.

DESCRIPTION. Structure as figured (Figs. 1-3).
Wing : ovipositor ratio 0.8:1, sheath : basal plate ra-
tio 3.4:1. In LACMIP 4010, length of body without
head as preserved (inflated because of postmortem
decomposition) 7.5 mm, forewing about 4 mm,
ovipositor 3.1 mm, sheath 2.4 mm. According to
Meunier (1920), length of holotype body 5.5 mm,
calculated forewing length about 3.5 mm.

The two known specimens differ slightly in their
forewing length; more significantly, there are dif-
ferences in the position of the hindwing crossvein
lr with respect to the RS base and possibly also in
the form of the forewing pterostigma and 2r cell.

8 ■ Contributions in Science, Number 450

These do not seem sufficient, however, to rule out
their conspecificity.

Subgenus Xyela Dalman, 1819

Rasnitsyn (1965) subdivided this subgenus into spe-
cies groups and later (Rasnitsyn, 1971) transformed
them into sections. One of these sections, which
was based on the relatively short wings and long
body, comprised all fossil species known up to that
time. This distinction seems to be partially real and
partially the result of different postmortem changes
in both living and fossil specimens. Pinned museum
specimens have the abdomen shortened because of
desiccation, while fossils usually have it inflated
because of decomposition. As a result, this char-
acter has been proven to be misleading. Other di-
agnostic characters of the section are not known,
so the latter has to be discarded. The position of
the included species is discussed below.

According to the suggestion by the editorial ad-
viser, I abandon here the concept of the section
which is not accepted by the ICZN and return to
the species group.

cf. Xyela julii Group
Xyela (Xyela) latipennis Statz, 1936
Figures 8, 16

Xyela latipennis Statz, 1936: 263, Abb. 2; Burdick,

1959: 121; Sphon, 1973: 60.

Xyela ( Xyela e.g. magna) latipennis: Rasnitsyn,

1965: 491, fig. 6.

MATERIAL EXAMINED. LACMIP lectotype
4015 [specimen described and figured by Statz
(1936); designated (as holotype) by Sphon (1973)];
Rott Formation, Rott near Bonn, Germany, Latest
Oligocene. Note: Two of the paralectotypes (“para-
types” by Sphon, 1973) proved to belong in part
to other Xyela species (see below), and the third
probably is a caddisfly (see Introduction).

DESCRIPTION. Structure similar to that figured
by Statz (1936), though wing venation as well as
some other details is difficult to confirm because
the specimen has since faded. Antennal segment 3
light, head and mesonotum with characteristic col-
or pattern (Fig. 2). Ovipositor flat (saw-like), wide,
gently downcurved, sheath with sides almost
straight, weakly converging caudally, roundly nar-
rowed subapically toward subacute, symmetrical
apex. Wing : ovipositor ratio 1.6-1. 9:1. Length of
body with ovipositor, as preserved, 4.8 mm, fore-
wing 3.3 mm, ovipositor 1. 8-2.1 mm, sheath 1.3
mm. Length of ovipositor basal plate, as well as
sheath : basal plate ratio cannot be determined with
certainty.

TAXONOMIC POSITION. The flat, gently
downcurved ovipositor is typical for the subgenus
Xyela. Assignment of the species should be con-
sidered only tentative, because the characters most

Rasnitsyn: Tertiary Sawflies of the Tribe Xyelini

reliable for group discrimination, viz. those dealing
with the fine structure of the ovipositor stylets, are
not preserved in the fossil. This is true for other
fossil Xyelini as well. Within the species group,
there are no other species described with a sym-
metrical ovipositor apex. Moreover, species with a
sheath as wide (X. bakeri Konow, X. menelaus
Benson) have an ovipositor that is not downcurved.

Xyela ( Xyela ) angustipennis
Statz, 1936
Figures 9, 10, 17=19

Xyela angustipennis Statz, 1936: 264, Abb. 3; Bur-
dick, 1959: 121 (as possibly conspecific with Xye-
la latipennis ); Sphon, 1973: 60.

Xyela (Xyela e.g. magna ) latipennis : Rasnitsyn,
1965: 491.

MATERIAL EXAMINED. LACMIP lectotype
4011 [specimen described and figured by Statz
(1936); designated (as holotype) by Sphon (1973)]
and LACMIP paraiectotypes 4012, 4014 [designat-
ed as paratypes by Sphon (1973)]; Rott Formation,
Rott near Bonn, Germany, Latest Oligocene. Note:
One further paralectotype (“paratype” of Sphon,
1973) proved to belong to a different species of
Xyela (see below).

DESCRIPTION. This description is based most-
ly on paraiectotypes because the lectotype is faded
and has been partly destroyed and its ovipositor is
seen only from above. As a result, the lectotype
shows no important characters except the color
pattern and sheath length which are similar to those
of paraiectotypes.

Structure generally similar to that figured by Statz
(1936). Color dark, including mesonotum, with an-
tennal segment 3 and probably legs light-colored.
Antennal funicle not preserved. Ovipositor flat (saw-
like), wide, gently downcurved, sheath not down-
curved, parallel-sided basally, almost rectilinear ta-
pering from somewhat between midlength and last
third toward subacute, asymmetrically placed apex,
with lower margin practically straight. Small, widely
spaced denticles seen on the lower sheath margin
of one paralectotype probably belong to lower ovi-
positor stylet. Basal sheath truncation subvertical.
Wing : ovipositor ratio in lectotype (with wing length
calculated from Statz’s drawing) 1.6:1, sheath : basal
plate ratio 1.9 (lectotype)-2.1:l. Length of body
with ovipositor, as preserved, 4.3-4.9 mm (lecto-
type 4.8 mm), forewing length 2.8 mm (lectotype),
ovipositor 1.6-1. 7 mm (lectotype), sheath 1. 0-1.1
(lectotype) mm.

TAXONOMIC POSITION. Assignment of this
species to the X. (X.) julii group is based on similar
grounds as the assignment of the previous species.
Within the group there are no other species de-
scribed with a similar sheath contour. In addition,
the wing size is among the smallest in the genus.

Contributions in Science, Number 450

cf. Xyela (Xyela) menelaus
Benson, 1960
Figures 11, 12, 20, 21

Xyela latipennis Statz, 1936: 263 (p.p.); Sphon, 1973:

60 (p.p.).

MATERIAL EXAMINED. LACMIP 4017-4016
[paraiectotypes of Xyela latipennis Statz, designat-
ed as paratypes by Sphon (1973)]; Rott Formation,
Rott near Bonn, Germany, Latest Oligocene.

DESCRIPTION. LACMIP 4017. Color pattern
generally typical for the genus (Fig. 2), though rather
dark. Third antennal segment dark dorsally. Ovi-
positor short, slightly downcurved, with sheath
straight, weakly tapering toward subacute, slightly
beak-like downward directed apex, obliquely trun-
cated basally. Wing : ovipositor ratio about 2.8:1,
sheath : basal plate ratio 1.1:1. Length of body with
ovipositor, as preserved, 4.5 mm, wing about 2.8
mm, ovipositor 1.6 mm, sheath 0.9 mm.

LACMIP 4016. Color comparatively light (pos-
sibly faded), possibly with typical pattern (Fig. 2),
although less developed. Ovipositor as above, ex-
cept sheath narrow, possibly because of some de-
formation. Length of body with ovipositor, as pre-
served, 4.9 mm, ovipositor about 1.7 mm, sheath
0.9 mm.

TAXONOMIC POSITION. LACMIP 4017. The
ovipositor form is typical for the X. julii group and
similar to X. (X.) menelaus Benson. The precise
relationship to X. (X.) menelaus cannot be deter-
mined, however, because its description mentions
length of neither wing nor ovipositor (Benson, 1960).
LACMIP 4016 and 4017 probably are conspecific,
unless the difference in ground color and sheath
width is real and not due to postmortem changes.

cf. Xyela (Xyela) julii
(Brebisson, 1818)

Figures 13, 22

Xyela latipennis Statz, 1936: 263 (p.p.); Sphon, 1973:

60 (p.p.).

MATERIAL EXAMINED. LACMIP 4018,
paralectotype of Xyela latipennis Statz [designated
as paratype by Sphon (1973)]; Rott Formation, Rott
near Bonn, Germany, Latest Oligocene.

DESCRIPTION. Color pattern generally typical
for the genus (Fig. 2), head possibly light (what
appears to be a dark orbit is probably the internal
eye apodeme). Ovipositor moderately long, gently
downcurved, with sheath straight, weakly tapering
toward subacute, slightly beak-like downward di-
rected apex, obliquely truncated basally. Wing : ovi-
positor ratio about 1.8:1, sheath: basal plate ratio
1.9:1. Length of body with ovipositor, as preserved,

5.5 mm, wing 3.8 mm, ovipositor 2.1 mm, sheath

1.5 mm.

TAXONOMIC POSITION. The shape of the

ovipositor is typical for the subgenus Xyela and

Rasnitsyn: Tertiary Sawflies of the Tribe Xyelini ■ 9

Figures 20-22. 20, 21. Xyela ( Xyela ) cf. menelaus Benson, 1960, photographs of LACMIP 4017 (20) and LACMIP

4016 (21). 22. Xyela {Xyela) cf. julii (Brebisson, 1818), photograph of LACMIP 4018.

10 ■ Contributions in Science, Number 450

Rasnitsyn: Tertiary Sawflies of the Tribe Xyelini

similar to that of X. (X.) julii (Brebisson, 1818),
differing only in that the sheath is narrower subapi-
cally. In addition, head is possibly lighter in color.
Both of these differences are rather subtle and might
be preservational. Assignment of the fossil to X.
(X.) julii cannot be rejected at present, although
additional material is necessary for certain identi-
fication.

Xyela ( Xyela ) florissantensis
sp. nov.

Figures 23, 24

MATERIAL EXAMINED. NMNH holotype
127677; Early Oligocene, Florissant, Colorado.

DESCRIPTION. Color pattern not preserved.
Head with maxillary palp large, leg-like, though less
developed than in ordinary X. (Xyela) species. Fore
femur short and thick, especially apically (due to
discoloration of femoral apex). Forewing venation
unusual in having short first abscissa of RS com-
bining with short RS + M, otherwise ordinary. Ovi-
positor long, flat, gently downcurved, sheath ta-
pering from beyond middle toward subacute apex
situated at dorsal sheath margin. Wing: ovipositor
ratio 1.4:1, sheath: basal plate ratio 2.2:1. Length
of body with ovipositor, as preserved, 6.7 mm,
wing 3.8 mm, ovipositor 2.7 mm, sheath 1.9 mm.

DIAGNOSIS. The new species differs from all
other Xyela in having a forewing with the first ab-
scissa of RS short. It possibly also differs by having
a short and thick fore femur, unless this is due to
diagenetic discoloration as a result of the seeming
loss of the femoral apex. It is similar to, and can
be tentatively assigned to, the X. (X.) julii group
because of its flat, gently curved ovipositor. Within
the group it differs from all other species by the
dorsal position of the ovipositor apex. In its sheath :
basal plate ratio, the species is similar to X. (X.)
julii but differs in the wing : ovipositor ratio.

ETYMOLOGY. The species is named after the
type locality.

Xyela (Xyela) micrura sp. nov.

Figures 14, 25

Xyela angustipennis Statz, 1936: 264 (p.p.); Sphon,
1973: 60.

MATERIAL EXAMINED. LACMIP holotype
4013 [paralectotype of Xyela angustipennis Statz,
1936, designated as paratype by Sphon (1973)]; Rott
Formation, Rott near Bonn, Germany, Latest Oli-
gocene.

DESCRIPTION. Color pattern of metanotum
typical of genus (Fig. 2). Ovipositor flat (saw-like),
wide, short, more or less straight, sheath probably
parallel-sided basally, almost rectilinear tapering to-
ward dorsally situated apex, with lower margin con-
vex and dorsal margin apparently straight, basal
sheath truncation possibly oblique. Wing : ovipos-
itor ratio about 2.6:1, sheath : basal plate ratio about
1.5:1. Length of body with ovipositor, as preserved,

Contributions in Science, Number 450

Figure 23. Xyela (Xyela) florissantensis Rasnitsyn sp. nov.,
line drawing of holotype.

4.2 mm, forewing length about 3.5 mm, ovipositor
about 1.3 mm, sheath about 0.7 mm.

DIAGNOSIS. The new species is similar to, and
can be tentatively assigned to, the X. (X.) julii group
because it has a flat, straight ovipositor. Within the
group it is similar to X. (X.) bakeri and X. (x.)
menelaus in having a short ovipositor but differs in
the dorsal position of apex. Additionally, the new
species differs from X. (X.) bakeri in having the
sheath relatively short and from X. (X.) menelaus
in having the sheath longer.

ETYMOLOGY. The species epithet is Greek for
“short tail” and refers to the short ovipositor.

cf. Xyela (Xyela) minor Group
Xyela (Xyela) magna Statz, 1936
Figures 15, 26

Xyela magna Statz, 1936: 264, Abb. 4; Burdick,
1959: 121; Sphon, 1973: 61.

Xyela ( Xyela e.g. magna) magna : Rasnitsyn, 1965:
498, fig. 5.

Xyela (Xyela) magna: Rasnitsyn, 1969: 38, fig. 63.
Xyela (Xyela: Sect. Magnixyela) magna: Rasnitsyn,
1971: 193.

MATERIAL EXAMINED. LACMIP holotype
4020; Rott Formation, Rott near Bonn, Germany,
Latest Oligocene.

DESCRIPTION. Visible structure similar to that
figured by Statz, but many details are indiscernible
because of fading. Antenna light, head and meso-
notum with characteristic color pattern (Fig. 2).
Forewing venation poorly preserved, differing from
Statz’s (1936) drawing in lacking proximal (super-

Rasnitsye: Tertiary Sawflies of the Tribe Xyelini ■ 11

Figures 24-26. 24. Xyela ( Xyela ) florissantensis

sp. nov., photograph of holotype. 25. Xyela ( Xye-
la) micrura sp. nov., photograph of holotype. 26.
Xyela {Xyela) magna Statz, 1936, photograph of
holotype.

12 ■ Contributions in Science, Number 450

Rasnitsyn: Tertiary Sawflies of the Tribe Xyelini

numerary) vein between RS and M + Cu and in free
SC branch situated near RS base (more distally than
figured). Ovipositor long, fiat (saw-like), down-
curved, with sheath: basal plate ratio 2:1, wing:
ovipositor ratio 1.5:1. Length of body with ovi-
positor, as preserved, 7 mm, forewing length 4.5
mm, ovipositor 3.1 mm, sheath 2.1 mm.

TAXONOMIC POSITION. In Rasnitsyn’s (1965)
key, this fossil keys to the X. minor group and
particularly to X. minor Norton, 1868, and X. pini
Rohwer, 1913. It differs from both in having a wide-
ly rounded sheath apex and an intermediate sheath :
basal plate ratio. Additionally, it differs from the
latter species in having a shorter ovipositor and
sheath. Xyela magna is most similar to North
American species.

Genus cf. Enneoxyela
Rasnitsyn, 1966

^Enneoxyela cenozoica
(Zhang, 1989) comb. nov.

Figure 27

Xyela cenozoica Zhang, 1989: 211, figs. 205, 206,
pi. 57, fig. 1.

MATERIAL EXAMINED. Photograph of LPM
holotype 820138; Shanwang Formation, 22 km east
of Linqu, Shandong Province, China; Middle Mio-
cene.

DESCRIPTION. Color dark. Head narrow. An-
tenna with 3rd segment and funicle both short,
subequal in length, funicle figured originally as about
8-segmented. Forewing with SC free, meeting R
well before RS base, meeting C clearly beyond RS
base. R almost straight, gently curved at (not before)
RS base. First abscissa of RS much longer than that
of M, RS + M shorter than the latter. Pterostigma
narrow, with 2r-rs at its midlength. Ovipositor long,
upcurved, tapering caudally, sheath probably half
as long as forewing or a little longer. Length of
body with ovipositor, as preserved, 9.3 mm, fore-
wing length 5.3 mm, sheath more than 2.3 mm.

TAXONOMIC POSITION. Characters seen on
the photograph of the holotype basically are di-
agnostic of Enneoxyela, which is known from three
species from the Late Jurassic of the Karatau Range
in Southern Kazakhstan (northern central Asia)
(Rasnitsyn, 1966, 1969). There are differences, how-
ever, particularly in the relatively short SC and in
the ovipositor which is upcurved and tapering rath-
er than being straight and parallel-sided. These dif-
ferences warrant separation of the fossil in a new,
closely related genus. I prefer to postpone this until
study of the actual holotype is possible.

ACKNOWLEDGMENTS

My work in the USA in 1989-1990 was supported by
grants from the Smithsonian Institution, Washington, D.C.,
the Museum of Comparative Zoology, Harvard Univer-
sity, Cambridge, Massachusetts, and the California Acad-
emy of Sciences, San Francisco, California. They were

Figure 27. ? Enneoxyela cenozoica (Zhang, 1989), line
drawing traced after the photograph of holotype.

initiated by Dr. Karl V. Krombein, Prof. James M. Car-
penter, and Dr. Wojciech J. Pulawski, respectively. Fa-
cilities for studying the fossils were provided by Dr. David
R. Smith at the Smithsonian Institution and by Dr. Edward
C. Wilson at the Natural History Museum of Los Angeles
County, in addition, Dr. Wilson and Dr. Smith have sup-
plied me with photographs of the Xyela fossils kept at
their institutions. Dr. Zhang Junfeng of the Shandong
Museum, Jinan, P.R. China, supplied me with the pho-
tograph of the holotype of Xyela cenozoica Zhang. Dr.
Vladimir E. Gokhman of the Moscow State University
helped me to calculate the cladogram of Xyelinae using
Hennig86.

I am deeply thankful to Dr. Edward C. Wilson, Dr.
Brian V. Brown, and Roy Snelling at the Natural History
Museum of Los Angeles County and to two anonymous
reviewers who read early versions of the manuscript and
improved it considerably.

LITERATURE CITED

Ashmead, W.H. 1898. Classification of the homtails and
sawflies, of the suborder Phytophaga. Canadian En-
tomologist 30:205-213.

Benson, R.B. 1960. Two new European species of Xyela
Dalman (Hymenoptera: Xyelidae). Proceedings of
the Royal Entomological Society of London B 29(7-
8):110-112.

. 1961. The sawflies (Hymenoptera Symphyta) of

Contributions in Science, Number 450

Rasnitsyn: Tertiary Sawflies of the Tribe XyeliniB 13

the Swiss National Park and surrounding area. Er-
gebnisse der wissenschaftlichen Untersuchungen des
schweizerischen Nationalparks 7{N.F.)(44): 163-1 91.

. 1962. Holarctic sawflies (Hymenoptera: Sym-

phyta). Bulletin of the British Museum (Natural His-
tory) Entomology 12(8):381-409.

Brebisson, L.A., de. 1818. Sur un nouveau genre d’in-
secte de l’ordre des Hymenopteres (Pinicola). Nou-
veau Bulletin des Sciences par la Societe Philoma-
tique de Paris, 116-117.

Braes, C.T. 1 908. New phytophagous Hymenoptera from
the Tertiary of Florissant, Colorado. Bulletin of the
Museum of Comparative Zoology 51:257-276.

Burdick, D.J. 1959. Notes on the taxonomy of the fossil
Xyelidae (Hymenoptera: Symphyta). Journal of the
Kansas Entomological Society 32(3): 120- 122.

— . 1961. A taxonomic and biological study of the

genus Xyela Dalman in North America. University
of California Publications in Entomology 17(3):285-
356.

Dalman, J.W. 1819. Nagra nya Genera och Species af
Insekter, beskrifna. Kunglige Svenska Vetenskap-
sakademiens Hadlingar 40:117-127.

Farris, J.S. 1988. Hennig86. Computer program and ref-
erence manual. 18 pp.

Fridolin, W.Yu. 1936. The animal-plant community of
the mountain region Khibiny. The biocenotic re-
search during 1930-1935. Transactions of the Kola
Station, Academy of Sciences of the U.S.S.R. 3:3-
295 (in Russian).

Konigsmann, E. 1976. Das phylogenetische System der
Hymenoptera. I. Deutsche Entomologische Zeit -
schrift, N.F. 23(4/5):253-279.

Konow, F.W. 1897. Uber die Xyelini. Entomologische
Nachrichten 23:55-58.

Krassilov, V.A., and A.P. Rasnitsyn. 1982. Unique find-
ing: Pollens in gut of the Early Cretaceous sawflies.
Paleontologicheskiy Zhurnal 4:83-96 (in Russian).

Linne, C., von. 1753. Sy sterna Naturae sy sterna Regna
tria Naturae, in Classes et Ordines, Genera et Spe-
cies redacta tabulisque aeneis illustrata. Editio 8.
Holmiae, 136 pp.

Meunier, F. 1920. Quelque insectes de 1’Aquitainen de
Rott, Sept.-Monts (Prusse rhenane). Koninklijke
Adademie van Wettenschappen te Amsterdam, Pro-
ceedings, Section of Science 22(2):891-898.

Newman, E. 1834. Attempted division of British insects
into natural orders. Entomological Magazine 2:379-
431.

Norton, E. 1868. Catalogue of the described Tenthre-
dinidae and Uroceridae of North America. Trans-
actions of the American Entomological Society 2:
321-368.

Patterson, C., and D.E. Rosen. 1977. Review of ichtyo-
dectiform and other Mesozoic teleost fishes and the
theory and practice of classifying fossils. Bulletin of
the American Museum of Natural History 158:81-
172.

Rasnitsyn, A.P. 1964. New Triassic Hymenoptera from
Middle Asia. Paleontologicheskiy Zhurnal 1:88-96
(in Russian).

. 1965. Notes on the biology, systematics and

phylogeny of Xyelinae (Hymenoptera, Xyelidae).
Polskie Pismo Entomologiczne 35(12):483-519 (in
Russian, with English summary).

. 1966. New Xyelidae (Hymenoptera) from the

Mesozoic of Asia. Paleontologicheskiy Zhurnal 4:
69-85 (in Russian; translated into English in Inter-
national Geological Review, 1967, 9:723-737).

— . 1969. Origin and evolution of Lower Hyme-
noptera. T ransactions of the Paleontological Insti-
tute, Academy of Sciences of the U.S.S.R. 123:1-
196 (in Russian; translated into English in 1979 by
Amerind Co., New Delhi).

— -. 1971. Evolution of Xyelidae (Hymenoptera). In

Current problems in paleontology. Transactions of
the Paleontological Institute, Academy of Sciences
of the U.S.S.R. 130:187-19 6 (in Russian).

— . 1980. Origin and evolution of Hymenoptera.

T ransactions of the Paleontological Institute, Acad-
emy of Sciences of the U.S.S.R. 174:1-192 (in Rus-
sian).

— . 1987. The importance of [not] being a cladist.

Sphecos 14:23-25.

. 1988. Outline of evolution of the hymenop-

terous insects (order Vespida). Oriental Insects 22:
115-145.

Rohwer, S.A. 1913. A synopsis, and description of the
Nearctic species of sawflies of the genus Xyela, with
descriptions of other new species of sawflies. Pro-
ceedings of the U.S. National Museum 45:265-281.

Ross, H.H. 1932. The hymenopterous family Xyelidae
in North America. Annals of the Entomological So-
ciety of America 25:153-169.

Smith, D.R. 1978. Suborder Symphyta. In Hymenop-
terorum Catalogus, Pars 14, ed. J. Van der Vecht
and R.D. Shenefeldt, Hague: Dr. W. Junk, 193 pp.

Sphon, G.G. 1973. Additional type specimens of fossil
invertebrata in the collections of the Natural History
Museum of Los Angeles County. Contributions in
Science, Natural History Museum of Los Angeles
County 250:1-75.

Statz, G. 1936. Uber alte und neue fossile Hymenop-
terenfunde aus den tertiaren Ablagerangen von Rott
am Siebengebirge. Decheniana 93:256-312.

Weyland, H. 1937. Beitrage zur Kenntnis der rheinische
Tertiarflora. II. Palaeontographica Abt. B 83:67-
122.

— . 1948. Beitrage zur Kenntnis der rheinische Ter-
tiarflora. VII. Palaeontographica Abt. B 88:113-188.

Zhang, Jun feng. 1989. Fossil insects from Shanwang,
Shandong, China. Jinan: Shandong Science and
Technology Publishing House, 459 pp. (in Chinese,
with English summary).

Zhelochovtzev, A.N., and A.P. Rasnitsyn. 1972. On some
Tertiary sawflies (Hymenoptera: Symphyta) from
Colorado. Psyche 79(4):3 15-327.

Zherikhin, V.V., and I.D. Sukacheva. 1989. Patterns of
insect burial in resins. In Sedimentary cover of the
earth in time and space. Stratigraphy and paleon-
tology, ed. B.S. Sokolov, 84-92. Moscow: Nauka
Press (in Russian, with English summary).

Accepted 30 August 1994.

14 ■ Contributions in Science, Number 450

Rasnitsyn: Tertiary Sawflies of the Tribe Xyelini

Number 451
28 March 1995

Contributions
in Science

Systematics and Biology of the
Bee Genus Xeralictus
(Hymenoptera: Halictidae, Rophitinae)

Roy R. Snelling and Gerald I. Stage

A Revision of the Nearctic Melittidae:
The Subfamily Melittinae
(Hymenoptera: Apoidea)

Roy R. Snelling and Gerald I. Stage

^*5S Natural History Museum of Los Angeles County

Serial
Publications
of THE
Natural History
Museum of
Los Angeles
County

Scientific

Publications

Committee

James L. Powell, Museum President
Daniel M. Cohen, Committee Chairman
Brian V. Brown
Kenneth E. Campbell
Kirk Fitzhugh
Robin A. Simpson, Managing Editor

• Contributions in Science, a miscellaneous series of tech-
nical papers describing original research in the life and
earth sciences.

• Science Bulletin, a miscellaneous series of monographs
describing original research in the life and earth sci-
ences. This series was discontinued in 1978 with the
issue of Numbers 29 and 30; monographs are now
published by the Museum in Contributions in Science.

• Science Series, long articles and collections of papers
on natural history topics.

Copies of the publications in these series are sold through
the Museum Book Shop. A catalog is available on request.

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

Printed at Allen Press, Inc., Lawrence, Kansas
ISSN 0459-8113

Systematics and Biology of the
Bee Genus Xeralictus
(Hymenoptera: Halictidae, Rophitinae)

Roy R. Snelling1 and Gerald I. Stage2

ABSTRACT. Formal justification for the removal of Xeralictus from subfamily Halictinae to subfamily
Rophitinae is presented. The genus Xeralictus, as both adults and larvae, is described and a key separates
the two known species, X. timberlakei Cockerell and X. bicuspidariae new species. Important taxonomic
features of the two Xeralictus species are illustrated.

The genus is limited primarily to desert regions of southern California and adjacent Nevada, Arizona,
and Baja California. Its two species are oligolectic on the loasaceous plant genera Mentzelia and, to a
lesser extent, Eucnide. Xeralictus species are part of a pollinator guild that includes Megandrena mentzeliae
Zavortink (Andrenidae) and two species of Hesperapis (Melittidae); other bees associated with these
flowers, especially those of the andrenid genus Perdita, appear to be scavengers that probably are not
effective pollinators. The relationships of these bees and flowers are briefly discussed.

INTRODUCTION

Xeralictus Cockerell, 1927, is a genus of halictid
bees known primarily from rocky canyons in the
deserts of the southwestern United States and ad-
jacent Mexico. The two known species, one pre-
viously undescribed, are part of a pollinator guild
centered around the loasaceous genus Mentzelia.
Other bees in this guild include Hesperapis , sub-
genus Xeralictoides (Melittidae) and Megandrena ,
and subgenus Eyrthrandrena (Andrenidae), as well
as several species of smaller bees that appear to be
principally scavengers.

Although Xeralictus had previously been placed
in the subfamily Halictinae (Cockerell, 1927; Mich-
ener, 1944), it was removed to the Dufoureinae
(now Rophitinae) by Eickwort (1969), on the
strength of our statements to him at that time. The
present paper provides the formal justification for
that decision.

SPECIMENS EXAMINED

Material utilized in this study is from the following
institutional and private collections: American Mu-
seum of Natural History (AMNH); Bee Biology and
Systematics Laboratory, USDA, Logan, Utah (BBSL);

1. Research Associate, Natural History Museum of Los
Angeles County, 900 Exposition Boulevard, Los Angeles,
California 90007.

2. RFD #1, Bowles Road, Stafford Springs, Connect-
icut 06076.

Contributions in Science, Number 451, pp. 1-17
Natural History Museum of Los Angeles County, 1995

California Academy of Sciences (CAS); Museum of
Comparative Zoology (MCZ); Central Texas En-
tomological Institute (CTMI); Natural History Mu-
seum of Los Angeles County (LACM); National
Museum of Natural History (USNM); Gerald I.
Stage, personal collection (GISC); University of Cal-
ifornia at Berkeley (UCB), Davis (UCD), and Riv-
erside (UCR); University of Kansas (UKAN); and
Thomas J. Zavortink, personal collection (TjZC).

TERMINOLOGY

In general, the morphological terminology em-
ployed here follows that established by Michener
(1944) and most subsequent authors. Puncture sizes
and the distances between punctures are as defined
by Snelling (1985). Distinction is here made be-
tween true abdominal segments (numbered in ro-
man numerals in the larval description) and meta-
somal segments (numbered in arabic numerals in
the adult descriptions).

Anterior ocellus diameter (OD) is the transverse
diameter of the anterior (median) ocellus.

Facial length (FL) is measured with the head in
frontal view, along the midline from the apical clyp-
eal margin to the anterior (median) ocellus.

Head length (HL) is measured with the head in
frontal view, from the apical (lower) margin of the
clypeus to the vertexal (upper) margin of the head.

Head width (HW) is the greatest width of the
entire head, including the eyes, in frontal view.

Interocellar distance (IOD) is the minimum dis-
tance between the posterior (lateral) ocelli.

Lower interocular distance (LID) is the distance

between the inner eye margins at the level of the
lateral angle of the clypeus.

Ocellocular distance (OOD) is measured as the
least distance between a lateral ocellus and the ad-
jacent compound eye.

Ocellovertexal distance (OVD) is measured with
the apical clypeal margin and dorsal vertexal margin
on the same plane; OVD is the shortest distance
between the posterior (upper) margin of the pos-
terior ocelli and the vertexal margin.

Upper interocular distance (UID) is the minimum
distance between the inner eye margins (at about
level of ocelli).

SPECIMEN DATA

Because published data on the distribution, sea-
sonality, and floral preferences of Xeralictus are
virtually nonexistent, we have cited full label data
for all specimens we have seen.

SYSTEMATICS

When Cockerell (1927) described and named Xera-
lictus, he allied it with genera now included within
the subfamily Halictinae. Michener (1944) also
placed Xeralictus in that subfamily but noted some
anomalies in that assignment. Xeralictus remained
within the Halictinae until Eickwort (1969) re-
moved it to the Dufoureinae (now Rophitinae) on
the strength of data that we had provided to him.

Michener (1944) correctly observed that in Xe-
ralictus females the prepygidial fimbria is not di-
vided (divided in Halictinae), the labrum does not
bear an apical process (present in Halictinae), and
the scopa is confined to the metatibia and meta-
basitarsus (scopa includes metafemur in Halictinae).
These three features, anomalous within the Ha-
lictinae, are common within the Rophitinae. More
recently, Michener et al. (1994) stated that place-
ment of Xeralictus in the Rophitinae is tentative
but presented no reasons why inclusion therein
should be so considered. In our view, the assign-
ment of Xeralictus to the Rophitinae is fully jus-
tified for reasons presented below.

Xeralictus shows obvious affinities with the Ro-
phitinae and shares those features of the Rophitinae
that separate them from the Halictinae. Pronotal
humeri are normally present in Halictinae, but in
Rophitinae occur only in Conanthalictus, Sphe-
codosoma, and a few species of Dufourea. In Ha-
lictinae, a fine carina extends diagonally from the
humerus across the side of the pronotum (Fig. 10);
such a carina is absent in Xeralictus (Fig. 9) and all
other rophitines except Conanthalictus (Fig. 11),
in which it is incomplete.

Halictine females possess a fan-like brush of api-
cally curved hairs on the posteroapical corner of
the metabasitarsus (Fig. 8). All Rophitinae lack such
a brush. Females of Xeralictus and some Dufourea
have a long, spatuliform posteroapical process (Fig.

6), and other species of Dufourea (Fig. 5) and all
Micralictoides (Fig. 7) have the posteroapical cor-
ner produced.

The prepygidial fimbria is divided in Halictinae
but not in Rophitinae (partially divided in some
Dufourea and in Xeralictus ). The pygidial plate in
halictine females is broad, flat, and with the apical
margin convex. In contrast, the rophitine pygidial
plate is narrow, there is a sharply elevated secondary
plate, and the apex is acute.

Metasomal structures of male rophitines are
markedly different from those of halictine males.
Within the Halictinae, the distal margins of the
exposed sterna are generally simple, either trans-
verse or weakly concave. Projections from either
the discs or distal margins are common among male
Rophitinae but rare in Halictinae. Sternum 7 of
rophitine males has one or (rarely) two pairs of
distal lobes (apparently lacking in the Chilean genus
Penapis ), and sternum 8 possesses a single, elongate
apical process. In the Halictinae, sterna 7 and 8 are
short and transverse, without processes.

The genital capsule of Halictinae is characterized
by short, broad gonocoxites, gonostyli, penis valves,
and volsellae; one or more of these may be con-
torted into bizarre shapes. In the Rophitinae, these
structures are usually elongate and slender (some-
what broadened in Protodufourea and Sphecodo-
soma). The genital foramen of Halictinae is very
large, is often broader than long, and occupies most
of the ventral surface of the gonobase. The genital
foramen is longer than broad in Rophitinae, except
in Conanthalictus, in which it is about as broad as
long.

The Xeralictus larva in most respects is a typical
rophitine and will key to that subfamily in the key
by McGinley (1981); the mature larva is described
in detail below.

Because Xeralictus shares such a wide array of
characteristics with the Rophitinae, and so few with
the Halictinae, it seems inarguable that it should
be included within that subfamily rather than the
Halictinae. It should be noted, for the sake of com-
pleteness, that the characteristics by which the Ro-
phitinae are separable from the Halictinae will
mostly serve to separate them also from the other
halictid subfamily, the Nomiinae. The only major
feature shared between the Rophitinae and the
Nomiinae is the lack of a completely divided pre-
pygidial fimbria in the female.

The Halictinae and Nomiinae form a homoge-
neous assemblage with a great many shared fea-
tures, including a general tendency in the females
to be polylectic. This agrees well with the opinion
expressed by Torchio et al. (1967) based on com-
parative biological data. By contrast, the Rophitinae
differ strikingly in many morphological characters,
as both adults and larvae, and biologically, includ-
ing a tendency toward oligolecty, from these two
subfamilies. We believe that the inclusion of the
Rophitinae within the family Halictidae should be
reevaluated, but that problem is beyond the scope
of the present study.

2 ■ Contributions in Science, Number 451

Snelling and Stage: The bee genus Xeralictus

Genus Xeralictus Cockerell

Xeralictus Cockerell, 1927:41. Type species: Xera-
lictus timberlakei Cockerell, 1927; monobasic and
original designation.

DIAGNOSIS

Moderate-sized to large Rophitinae with blackish
integument (metasoma may be dull reddish), sparse
pilosity, three submarginal cells in forewing, long
basal face of propodeum, and mesosoma subpol-
ished to polished between distinct, well-separated
punctures; male inner eye margins moderately di-
vergent below, clypeus about 2.3-3. 1 times as broad
as long, and pygidial plate absent.

DESCRIPTION

Moderate-sized to large, total length more than 7.5
mm; integument non-metallic, blackish, except me-
tasoma may be dull reddish in female, generally
smooth and shiny between distinct punctures; pos-
terior margins of metasomal terga broadly de-
pressed, depressed bands with sparse, fine piligerous
punctures; marginal cell of forewing long, distance
from apex of pterostigma to apex of marginal cell
distinctly longer than distance from marginal cell
to wing tip; three submarginal cells present, middle
cell much shorter than first or third and receiving
first recurrent vein near its apex; first transverse
cubital vein interstitial with basal vein.

Pilosity generally sparse on head and body except
hairs long, moderately dense and concealing integ-
ument on gena, side of propodeum, and outer side
of metatibia and tarsi, and except for apical fimbria
of metasomal tergum 6 and all of tergum 7 of male;
female with long, curled hairs along lower margin
of mandible, hairs of lower genal area long and
curled, those of side of propodeum short and not
concealing surface, pro- and mesotibiae and tarsi
densely pubescent, scopal hairs of metatibia and
metabasitarsus long, dense, and reclinate, metaso-
mal terga 3 and 4 with short basal bands of short,
white hairs. Tergum 5 covered by compact mass of
brownish hairs that is divided in middle, partially
exposing pygidial plate on following segment.

MALE: Head (Fig. 1) distinctly broader than long;
antennal sockets slightly below midlength of head;
inner eye margins moderately divergent below, LID
about 1.3 times UID. Vertexal margin nearly straight
in frontal view and strongly elevated above top of
eye; anterior ocellus below line drawn between tops
of eyes. Clypeus at least twice as broad as long,
longer than labrum; separated from antennal socket
by about a socket diameter; apical margin broadly
convex between distinct sublateral angles that are
nearer lateral angles than each other. Labrum more
than twice as broad as long; median tubercle nar-
row and high at base, lower and broader distally,
its margins sharply carinate, especially basad; latero-
basal portion of labrum concave and polished.

Frontal suture weak or absent. In profile, greatest
width of gena distinctly greater than width of eye;
thick, slightly curved process present near base of

mandible (Fig. 2). Hypostomal carina sharply ele-
vated at base and abruptly reduced at about one-
third distance from base, angle mesad of mandible
base prominent and acute; hypostoma about twice
as broad as base of mandible.

Mandible (Fig. 1) about 2.5 times as long as basal
width, apical tooth stout; preapical tooth at right
angle to long axis of mandible, dorsal margin be-
tween it and subbasal angle long, with concave
shiny facets basad and distad of subbasal angle,
distal facet long and tapering toward preapical tooth;
ventral margin with prominent, convex flange be-
ginning at about midlength and tapering toward
apex.

Ocellar triangle broad; OOD greater than IOD
in frontal view, OVD equals or exceeds OD.

Antennal scape more than three times as long as
broad, about as long as following three segments
combined, extending to level of anterior ocellus;
first flagellar segment longer than broad and longer
than nearly quadrate second segment; flagellar seg-
ments 3-10 longer than broad, somewhat flattened
beneath, without hair tufts or obvious sensory
structures.

Pronotal collar narrow and high, deeply de-
pressed in middle; pronotal side without humeral
angles, ridges, or carinae. Mesoscutum about as
long as wide. Dorsal face of propodeum about twice
as long as metanotum and about one-half as long
as posterior face when viewed in profile.

Metasomal T7 densely pilose, without pygidial
plate. Sterna 2 and 3 translucent, shiny, posterior
margin slightly projecting across middle one-third;

54 (Figs. 22, 23) with large, flat, shiny apical process;

55 (Figs. 24, 27) with high, curved, sublateral la-
mella-like ridges and large, apically broadened me-
dian process; S6 (Figs. 25, 28) short, with two slen-
der, curved, submedian apical processes; S7 (Figs.
26, 29) bifurcate at apex; S8 (Figs. 30, 32) with basal
apodeme broad, apical process long, narrow, and
pilose. Genitalia (Figs. 34-37): genital foramen lon-
ger than broad; capsule elongate; volsella promi-
nent.

FEMALE: Similar to male except usual sexual
differences and the following. Head (Fig. 3) slightly
broader than long or about as long as broad. Clyp-
eus about twice as broad as long; sublateral angles
more prominent than in male and margin between
them more strongly convex. Labral tubercle a sim-
ple median convexity, neither prominent in profile
nor with sharp margins. Preapical tooth of man-
dible broadly rounded; subbasal angle absent; ven-
tral convexity absent. Greatest width of gena only
slightly greater than that of eye; no ventral process
near base of mandible. Hypostomal carina nearly
uniform in height throughout its length; hypostoma
only slightly wider than base of mandible.

Base of stipes with dense brush of long, simple
hairs; galea of maxilla short and broad, apex nar-
rowly rounded and extending only slightly beyond
third segment of maxillary palpus; maxillary palpus
six-segmented, segments elongate and cylindrical,
last segment reaching tip of extended glossa; labial

Contributions in Science, Number 451

Snelling and Stage: The bee genus Xeralictus ■ 3

palpus four-segmented, segments elongate and cy-
lindrical, basal segment distinctly curved at base,
last segment extending slightly beyond apex of seg-
ment 4 of maxillary palpus.

Ocellar triangle broad and ocelli below level of
tops of eyes in frontal view; OOD slightly greater
than IOD; OVD about twice as great as OD.

Antennal scape about as long as following four
segments combined; first flagellar segment distinctly
longer than broad and longer than second; seg-
ments 2-10 broader than long.

Mesotibial spur with seven to nine large, coarse
teeth; outer metatibial spur with three to six low,
oblique, coarse, well-separated teeth; basitibial plate
large, marginate, flat, densely pubescent; metaba-
sitarsus with posteroapical, translucent spatuliform
process (Fig. 6). Scopa (Fig. 19) copious, with many
long, distally curled hairs.

Prepygidial fimbria of metasomal tergum 5 par-
tially divided posteriorly; pygidial plate largely con-
cealed by pilosity of tergum 5; metasomal sterna
unmodified.

Features of pilosity, limited to female, that are
presumably associated with pollen gathering and
transport, in addition to the metatibial scopa, in-
clude the brush of long, curled hairs on the lower
gena; the long, simple, apically curled hairs of the
probasitarsus; the marginal fringes of long hairs on
metasomal sterna 2-4.

LARVA: The following description is compar-
ative to that of the mature larva of Sphecodosoma
dicksoni (Timberlake) in Rozen (in prep.). See that
paper for references to descriptions of larvae of
taxa named below.

Diagnosis: Mature larvae of the Rophitinae can
be recognized by the following combination of
characters: dorsolateral body tubercles conical (i.e.,
not transverse), present on most body segments;
those of pronotum noticeably smaller than those
of following segments; labiomaxillary region more
or less produced; salivary lips projecting, transverse;
abdominal segment IX strongly produced medially.
Because those features by which known mature
larvae of Rophitinae may be distinguished from one
another are given in Rozen (in prep.), they are not
repeated here.

Head (Figs. 12, 15): Integument of capsule with
scattered sensilla that are small and not obviously
setiform; integument somewhat pigmented; inter-
nal ridges and mandibular apices more darkly pig-
mented.

Head (Fig. 13) small compared to remainder of
body; head capsule distinctly wider than length
measured from top of vertex to lower clypeal mar-
gin in frontal view. Tentorium well developed, pos-
sessing well-developed dorsal arms; anterior ten-
torial pits normal in position, not immediately
adjacent to anterior mandibular articulations; pos-
terior tentorial pits in normal position at junction
of posterior margin of head and hypostomal ridges;
posterior thickening of head capsule moderately
developed, not curving forward medially as seen in

dorsal view; posterior margin of head in normal
position; median longitudinal thickening of capsule
absent except at summit; hypostomal ridge well
developed, arching upward in middle, without ra-
mus, of moderate length, forming approximately
90° angle with posterior margin as seen in lateral
view (Fig. 15); pleurostomal ridge well developed;
epistomal ridge moderately developed, extending
part way between anterior tentorial pits; epistomal
depression not pronounced (see Remarks). Parietal
bands evident. Antennal prominence moderately
weak (Fig. 15) (see Remarks); antennal disc mod-
erately small (Fig. 12); antennal papilla small (Fig.

12) , bearing three sensilla. Vertex evenly rounded
in side view (Fig. 15), without unusual projections;
clypeus moderately wide, of normal length (i.e., not
short so that clypeolabral suture almost in line with
anterior mandibular articulations as seen in frontal
view, as in Sphecodosoma dicksoni, Dufourea mul-
leri (Cockerell), and D. novaeangliae (Robertson));
frontoclypeal area in lateral view (Fig. 15) projecting
somewhat beyond la brum (see Remarks). Labrum
in profile not projecting beyond clypeus; labral
sclerite not evident; labral tubercles virtually absent
(Figs. 12, 15); epipharyngeal surface spiculate.

Mandible (Figs. 16-18) robust at base, tapering
to simple apex; dorsal surface with a few small,
sharp-pointed spicules; outer surface with large,
conspicuous tubercle; dorsal adoral surface with
numerous large, sharp-pointed teeth near cusp; dor-
sal apical edge dentate; ventral apical edge with
small teeth; apical concavity weakly developed, ap-
parently represented by shallow groove extending
from apex along ventral surface, this area nonspic-
ulate. Labiomaxillary region (Fig. 15) only some-
what produced, not as much as in Sphecodosoma
and Dufourea (see Remarks). Maxillary apex not
produced mesally; sclerotized cardo and stipes pig-
mented, clearly visible; articulating arm of stipital
sclerite quite evident because of pigmentation; ga-
lea not evident; maxillary palpus moderately elon-
gate, longer than basal diameter. Labium divided
into prementum and postmentum but not as strong-
ly so as in Sphecodosoma and Dufourea ; premental
sclerite unpigmented; labial palpus slightly smaller
than maxillary palpus. Salivary lips developed, a
projecting, narrow, transverse slit. Hypopharynx a
bulging, non-bilobed, spicule-bearing surface; hy-
popharyngeal groove extending between apices of
articulating arms of stipital sclerites, separating hy-
popharynx from dorsal labial surface.

Body: Integument pigmented, rigid at least on
postdefecating larva, without setae or setiform sen-
silla; sides of pronotum, dorsal and ventral areas
of most body segments spiculate; dorsolateral tu-
bercles and apex of abdominal segment X micro-
scopically irregularly roughened compared to rath-
er smooth integument elsewhere. Body form (Fig.

13) moderately robust, not greatly elongate; inter-
segmental lines well incised; intrasegmental lines
apparently evident on some abdominal segments as
lines extending down from front of dorsolateral

4 ■ Contributions in Science, Number 451

Snelling and Stage: The bee genus Xeralictus

tubercles; paired dorsolateral body tubercles con-
spicuous on meso- and metathorax, and abdominal
segments I- VIII, all but absent on IX, compara-
tively reduced in size on prothorax, and absent on
abdominal segment X; dorsolateral body tubercles
conical rather than transverse, many appearing trun-
cated, possibly as result of small brood cell (see
Remarks); other tubercles absent; venter of abdom-
inal segment IX strongly produced medially; seg-
ment X attached dorsally to IX, its apex without
ridges or other modifications, but flattened (see Re-
marks); anus presumably apical on X as seen in
lateral view (Fig. 13). Spiracles (Figs. 13, 14) mod-
erately small, not on tubercules, subequal in size,
without sclerites; peritreme present but narrow;
atrium projecting above body wall, with rim, glo-
bose; atrial wall smooth; primary tracheal opening
with collar; subatrium normally short, with about
nine chambers. Male sex characters unknown; fe-
male with two darkly pigmented cuticular scars on
venter of each of abdominal segments VII-IX.

Material Studied: 1 postdefacating larva, Big
Morongo Canyon, Riverside Co., California, April
1967 (G.I. Stage; AMNH).

Remarks: This larva was examined by Rozen,
who has been studying the rophitine larvae, and
most of the preceding description is drawn directly
from his remarks. Additionally, he noted that the
specimen revealed certain features that appear to
be the direct result of its having been enclosed in
a confining brood cell during development, as men-
tioned in the following Biology section. The ap-
parent resultant modifications include the follow-
ing: front of head and perhaps antennal prominences
somewhat flattened; labrum somewhat recessed un-
der flattened clypeus; many dorsolateral body tu-
bercles apically truncate (Fig. 13); abdominal seg-
ment X short, apically flattened. While some of
these features may actually be innate to this species
and not the result of confined quarters during
growth, there is little doubt that the truncate dor-
solateral tubercles, short abdominal segment X, and
flattened face would not occur in a larva recovered
from a larger cell. We do not know if all cells of
this species are so confining.

We cannot state with certainty whether or not
Xeralictus spins a pupal cocoon, even though the
larva described above is a postdefecation specimen;
in other rophitines, feces and cocoon fabric are
closely connected. If this bee does not spin a co-
coon, it is the first known rophitine to exhibit this
trait. Such a possibility is supported by some ana-
tomical features of the larva: the more rigid, pig-
mented body integument, the less produced la-
biomaxillary region, and the less distinct division
of the labium into prementum and postmentum
when compared to other known rophitines. On the
other hand, the strongly projecting (but narrow)
salivary lips suggest cocoon spinning. If this species
produces no cocoon, this habit must be recently
evolved. Larvae of lineages that abandon cocoon
spinning quickly lose projecting salivary lips, as ev-

Contributions in Science, Number 451

idenced in the anthophorid genus Exomalopsis,
where some species spin cocoons and have well-
developed salivary lips and projecting labiomaxil-
lary regions, whereas other species have lost co-
coon-spinning features (McGinley, 1987).

BIOLOGY

Little is known of the nesting biology of Xeralictus ;
only a single nest of X. timberlakei has been found.
That nest was located in Big Morongo Canyon in
the Little San Bernardino Mountains, Riverside
County, California, and was excavated by GIS. The
site was on a talus-covered slope of about 45° over-
lying a layer of decomposed granite and fine clay
soil. The nest entrance was in an abandoned ver-
tebrate burrow. The female Xeralictus was seen to
fly directly into the burrow, flying slowly until out
of sight.

When excavated, the tunnel was irregular and
meandered along cracks and around stones in the
soil, vertical at some points, horizontal at others.
The tunnel was circular in cross-section, not filled
with soil and lined with fine soil, with no indication
of any lining secretion or wax. A total of four cells
was found, two with mature larvae, one with frag-
ments of an adult female, and one, the terminal
cell, empty. No measurements are available for these
cells, which were stout and abruptly truncate-ovoid
in shape; the cell closures were flat. Although no
waxy lining was present (surface dull), the fine soil
lining of the interior of the cells was apparently
treated in some manner since the cells were re-
moved intact. The larval meconium was evenly de-
posited along the bottom of the cell. Mature larvae
fit snugly in the cells, with only the tubercles in
contact with the cell walls.

The preceding information was reconstructed
from fragmentary notes by GIS. To this may be
added that females are often seen flying about in
“searching” patterns on talus slopes. They explore
shadowed areas, cracks, and holes.

Females of both species of Xeralictus are oli-
goleges on flowers of some Mentzelia species in the
Bicuspidaria Section (Darlington, 1934; Thompson
and Roberts, 1974) and are part of a complex of
bees associated with these Mentzelia and the mor-
phologically convergent Eucnide mens, the sole
member of the section Mentzeliopsis (Thompson
and Ernst, 1967) of Eucnide; both are genera within
the Loasaceae. Mohavea confertiflora (Scrophular-
iaceae) probably should be included here also, but
it has not been studied as intensively as the Loa-
saceae species. Four closely related allopatric spe-
cies of Mentzelia are involved: M. hirsutissima, M.
involucrata, M. tricuspis, and M. tridentata. These
species have exserted stigmas that tend to preclude
pollination except by bees of moderate size that
must orient in a specific manner when entering the
blossom; the blossom of Eucnide mens is mor-
phologically parallel and must be entered in the

Snelling and Stage: The bee genus Xeralictus ■ 5

same fashion. Another species in the Bicuspidaria
Section is M. reflexa ; it (and some populations of
each of the other three species) has a short stigma
that terminates on the same level as the anthers and
may be pollinated by nearly any visiting insects and
is also capable of self-pollination.

In addition to the two Xeralictus species treated
here, the pollinator guild includes two Hesperapis
species (Melittidae) belonging to the subgenus Xe-
ralictoides, H. laticeps Crawford, 1917, and an un-
described species (Stage and Snelling, in prep.), and
one species of Megandrena (Andrenidae), M. merit-
zeliae Zavortink, 1972. The floral visiting behavior
of M. mentzeliae has been presented in some detail
by Zavortink (1972), who observed this species at
its type locality in Clark County, Nevada. Other
bees, belonging to the genera Perdita (Andrenidae)
and Lasioglossum ( Dialictus ) (Halictidae), are com-
monly encountered on these plants but appear to
be scavengers, even though such species as P. koe-
belei Timberlake appear to be oligoleges on these
plants.

At his Clark County site, Zavortink (1972) found
three members of this pollinator guild visiting
Mentzelia tricuspis: Megandrena mentzeliae, Xe-
ralictus bicuspidariae, and Hesperapis new species.
Zavortink noted that the Megandrena is more ac-
tive in the forenoon than the other two species and
tended to visit freshly opened flowers, and that the
other two species visited older blossoms in which
the mass of stamens had already been loosened due
to aging. He observed,

When the stamens are tightly appressed to the style, as
they are in the fresh flowers visited by Megandrena ment-
zeliae, the Xeralictus and Hesperapis are not able to force
their way into the pollen chamber. They are, in fact, only
rarely seen in such flowers. If, as is apparently the case,
the stigma of Mentzelia tricuspis is receptive when the
flower first opens, then pollination has occurred before
the flower is visited by Xeralictus and Hesperapis, and
the latter do little more than collect residual pollen.

While this may be true of the populations at that
locality, it certainly is not true in other areas where
both Xeralictus and Hesperapis utilize M. tricuspis
and other species of Mentzelia in the absence of
Megandrena.

Xeralictus females enter Mentzelia blossoms in
a characteristic fashion. The females plunge directly
into a blossom and, even when the stamens are still
tightly compressed together in a newly opened blos-
som, immediately force entry. If the bee is a female
collecting pollen, the tip of the metasoma is hooked
over the style and the bee usually rotates around
the interior of the blossom, pivoting on the style.
While the pollen is being collected, a clearly audible
scratching sound may be heard. Presumably this is
the result of the bee raking pollen from the pollen
chamber or from the anthers.

The female periodically backs out of the mass
of stamens and falls sideways within the blossom;
in this position she then removes pollen from the
pollen brush on the underside of the head and from

the forelegs and transfers it to the metatibial scopa.
She may then burrow back into the blossom to
gather more pollen or, less often, fly to another.

One puzzling aspect of the biology of these bees
is whether or not floral nectar is used in provisioning
the nest cells. The blossoms of the Mentzelia ap-
parently do not produce nectar, although the au-
thors are in disagreement on this matter: Stage
maintains that they do produce nectar, while Snell-
ing has the opposite view. The latter view is also
that of H.J. Thompson and T.J. Zavortink (pers.
comm.), both of whom have studied these flowers
in the field.

In areas where Megandrena mentzeliae is absent,
Hesperapis and Xeralictus enter freshly opened
blossoms, although Hesperapis, the smallest bees
in this pollinator guild, do so with some difficulty.
Once the stamens are loosened, the large amount
of pollen produced by the anthers in then available
for exploitation by true scavenger bees. These are
mostly various species of Perdita, especially P. koe-
belei Timberlake (1964), but including also Ancy-
landrena timberlakei Zavortink (1974), Agaposte-
mon spp., and Lasioglossum ( Dialictus ) spp.
Individuals of Perdita species are often present in
great numbers and may form veritable “clouds”
around the Mentzelia plants. Females of P. koebelei
have been observed attempting to “steal” pollen
from the scopae of Xeralictus females.

Mating occurs on the flowers. Males apparently
are territorial, at least to some degree, but further
study of this must be conducted. They usually perch
within blossoms and await the arrival of foraging
females. Within the blossoms males assume a dis-
tinctive “guarding” posture (as in Fig. 38, on a blos-
som of Mohavea confertiflora ), possibly to intim-
idate other males. Any female that enters the
blossom is immediately grappled and a mating at-
tempt ensues. If the mating attempt is successful,
the bees remain coupled for up to about 30 seconds,
commonly lying on their sides within the blossom.
After separation the female either resumes her for-
aging activity on that flower or moves to another.
The male, after mating, flies off to make a “round”
of his territory. Such males may pounce on females
within blossoms or attack males in other blossoms.
Such attacks result in furious buzzing and energetic
combat. Often the combatants fall from the blos-
som; the combat terminates as soon as the bees
reach the ground and both return to their routines.

The distribution of the pollinator guild associ-
ated with Section Bicuspidaria is complex but, in
general, at any given site there are only two of these
bees present. In general, also, of the two species
that may be present, there is one in which the fe-
male has a red metasoma and one in which the
female has a black metasoma. Additionally, two
species of the same genus are seldom present at the
same site.

A few localities do not fit this pattern. At Phoenix
(Arizona) and at Isla Angel de la Guarda (Baja Cal-
ifornia), only a single species has been found: a dark

6 ■ Contributions in Science, Number 451

Snelling and Stage: The bee genus Xeralictus

Figures 1-4. Xeralictus timberlakei, frontal and lateral views of male (1, 2) and female (3, 4) head. Figures by Ruth
Ann DeNicola.

phase of X. bicuspidariae. Zavortink’s Clark Coun-
ty (Nevada) site has three species, one of each genus;
this site is further exceptional in that females of all
three species have red metasomata.

Other localities with atypical populations include
northern Mohave County (Arizona): Willow Beach;
9.6 and 13.7 mi SE Hoover Dam; Kingman Wash;
Lone Mountain Road. At each site, X. bicuspi-
dariae and Hesperapis new species are sympatric.
Although most populations of the Hesperapis have
females with red metasomata, many at these sites
have very dark reddish, almost brown, metasomata
and others are normal. At these sites, the Hesper-

Contributions in Science, Number 451

apis species forage almost exclusively at Eucnide
urens, and X. bicuspidariae is found only on M.
tricuspis. At several dozen other sites, we have found
this species of Hesperapis to be an oligolege of
Mentzelia, and it is only at these four sites that
females with dark metasomata are found.

While there are no known sites where the two
species of Hesperapis occur together, two Califor-
nia localities are known where the two Xeralictus
species are sympatric: El Paso Mountains (Kern
County) and the Chuckwalla Mountains (Riverside
County). At both locations the two species forage
on M. involucrata in about equal numbers.

Snelling and Stage: The bee genus Xeralictus ■ 7

Figures 5-11. 5-8. Apex of metatibia of female of Dufourea virgata (5), Xeralictus timberlakei (6), Micralictoides

ruficaudus (7), and Lasioglossum fuscipenne (8). 9-11. Lateral view of pronotum of X. timberlakei (9), Halictus ligatus
(10), and Conanthalictus bakeri (11). Figures by Ruth Ann DeNicola.

Of the two species of Xeralictus, X. timberlakei
has the more restricted distribution, and its range
is primarily confined to the western Colorado and
Mojave Deserts of southern California. Within this
area it appears to be an oligolege of M. involucrata,
the most widely distributed member of the Bicus-
pidaria Section of Mentzelia. Why X. timberlakei
should have a distribution that is largely limited to
the western portion of the range of its host plant
is puzzling, but presumably some factor of com-
petition is involved.

More widely distributed is X. bicuspidariae, the
range of which encompasses much of that of M.
involucrata but includes also M. tricuspis, M. tri-
dentata, and M. hirsutissima. Although X. bicus-
pidariae is most often associated with M. involu-
crata, each of the other species is also utilized as
a pollen source.

The distribution of various members of the
Mentzelia Bicuspidaria Section exceeds that pres-
ently known for Xeralictus. Whether or not these
bees are truly absent from these areas remains to

8 ■ Contributions in Science, Number 451

Snelling and Stage: The bee genus Xeralictus

Figures 12-14. Xeralictus timberlakei, mature larva. 12. Frontal view of head. 13. Lateral view of head and body.
14. Spiracle. Figures by Ruth Ann DeNicola.

be determined. Further collecting is obviously nec-
essary, especially in Baja California.

KEY TO SPECIES OF XERALICTUS

1 Male, antenna 1 3- segmented and rnetaso-
mal sterna 4 and 5 with prominent median
process on apical margins ............ 2

Female, antenna 12-segmented and rnetaso-

mal sterna without median processes .... 3

2(1) Median process of S4 distinctly angulate on
each side (Fig. 22); process of S5 with sharp
apicolateral comer (Fig. 24); large species,
head width at least 2.75 mm, usually more
than 3.00 mm .... timberlakei Cockerell

— Mediae process of S4 not angulate on each
side (Fig. 23); process of S5 with rounded
apicolateral comers (Fig. 27); smaller spe-
cies, head width less than 2.70 mm .....

............. bicuspidariae new species

3(1) Metasoma dark brown, tergal margins often
broadly dusky ferruginous; large species,

head width at least 2.50 mm and usually
over 2.70 mm; head a little broader than

long timberlakei Cockerell

— Metasoma ferruginous, tergum 2 usually
with dark brown spot on each side (rarely
entire metasoma dusky ferruginous); smaller
species, head width no more than 2.60 mm
and usually less than 2.50 mm; head slightly

longer than broad

............. bicuspidariae new species

Xeralictus timberlakei Cockerell

Figures 1-4, 6, 9, 19, 20, 22,

24-26, 30, 31, 35, 37, 38

Xeralictus timberlakei Cockerell, 1927:42; 6.

DIAGNOSIS

Male. Median process of metasomal tergum 4
angulate on each side and process of sternum 5
with sharp apicolateral comer; head width at least
2.75 mm and usually over 3.00 mm. Female. Meta-
somal terga dark brown across discs; head width

Contributions in Science, Number 451

Snelling and Stage: The bee genus Xeralictus M 9

Figures 15-18. Xeralictus timberlakei, mature larva. 15. Lateral view of head. 16. Right mandible, ventral view. 17.
Same, dorsal view. 18. Same, mesial view. Figures by Ruth Ann DeNicola.

at least 2.50 mm, usually over 2.70 mm, and a little
greater than head length.

DESCRIPTION

Male, measurements (mm): HW 2.81-3.60; FL
1. 9-2.2; wing length (WL) 7.3-8. 7; total length (TL)
12.0-13.6.

Head about 1.2 times broader than long. Inner
eye margins moderately divergent below, LID about
1.2 times UID. IOD about 2.5 times OD; OOD
about 2.7 times OD. Clypeus moderately shiny be-
tween scattered fine to moderate punctures; middle
of supraclypeal area moderately shiny and impunc-
tate, but laterally with subcontiguous punctures that
become sparser laterad; lower paraocular area shiny
between scattered fine punctures that become mod-
erate to coarse adjacent to antennal sockets; punc-
tures of vertex moderate, variably spaced from sub-
contiguous behind ocelli to close or sparse laterad
and near vertexal margin. Gena shiny between mod-
erate subcontiguous to dense punctures.

Mesoscutum shiny, punctures fine to moderate,
subcontiguous at side, becoming sparse in center;
scutellum shiny, very weakly tessellate, with very
widely scattered fine punctures over most of disc,
some subcontiguous coarse punctures at extreme
side; metanotum slightly duller, with sparse mod-
erate punctures in middle and subcontiguous coarse
punctures laterad. Mesepisternum slightly shiny and
roughened between coarse subcontiguous or dense
punctures; metepisternum dull, finely, contiguously
rugosopunctate. Basal area of propodeum slightly
shiny and distinctly roughened, most of basal area
with fine, longitudinal striae; side slightly shiny and
distinctly roughened, most of basal area with fine,
longitudinal striae; side slightly shiny and distinctly
roughened between minute subcontiguous to dense
punctures. Outer metatibial spur with four to five
coarse, suberect teeth.

Metasomal terga moderately shiny and finely
roughened between fine, dense to subcontiguous
punctures; apical margins of segments 1-5 broadly
depressed with finer, more obscure punctures than

10 ■ Contributions in Science, Number 451

Snelling and Stage: The bee genus Xeralictus

Figures 19-23. Xeralictus timberlakei. 19. Scopa of female. 20, 21. Ventral view of male metasoma of X. timberlakei
(20) and X. bicuspidariae (21). 22, 23. Male S4 of X. timberlakei (22) and X. bicuspidariae (23). Figures 19-21 by Ruth
Ann DeNicola.

on discs. Sterna 2 and 3 shinier, sparsely and finely
punctate, but with broad apical margins nearly
transparent, polished and shiny. Apical process of
S4 (Fig. 22) broad and with distinct lateral angle;
in ventral view, apicolateral angle of median process
of S5 (Fig. 24) narrowly rounded. Process of S8
(Fig. 30) long, evenly narrowed distad to convex
apical margin.

Female, measurements (mm): HW 2.5-3.0; FL
1. 9-2.2; WL 6. 5-7.6; TL 10.9-12.4.

Fiead about 1.2 times as long as wide. Inner eye
margins weakly divergent below, LID about 1.1
times UID. IOD about 2.06 times OD; OOD about

Contributions in Science, Number 451

2.4 times OD; OVD about 1.5 times OD. Clypeus
smooth and shiny, punctation as described for male.
Remainder of cephalic punctation as in male.

Mesosoma as in male, but mesepisternal punc-
tures less dense and more obscured by roughening
of interspaces. Outer metatibial spur with five to
six coarse suberect teeth.

Metasoma similar to that of male other than
usual sexual differences (six segments, sterna simple,
etc.); T 6 pygidial plate usually hidden under pre-
pygidial fimbria of T5; discs of S2-S6 transversely
roughened, moderately shiny; distal one-half or
more of S2 with sparse, coarse, piligerous punc-

Snelling and Stage: The bee genus Xeralictus ■ 11

Figures 24-29. Xeralictus spp., male metasomal sterna 5, 6, and 7, respectively, of X. timberlakei (24-26) and X.
bicuspidariae (27-29).

tures; S3-S5 each with preapical bands of coarse,
piligerous punctures, the discs without definite
punctures. Terga dark reddish brown with yellow-
ish hyaline margins.

TYPE MATERIAL

The type male is from Salt Creek [San Bernardino
Co.], California, 20 Mar. 1927 (P.H. Timberlake),
on flowers of [Nuttallia] = Mentzelia involucrata,
and is deposited in the California Academy of Sci-
ences. We have examined the type and it agrees
with the current concept of this species.

SPECIMENS EXAMINED

UNITED STATES, California, Kern Co.: 2 29, Iron Cyn.,
El Paso Mts., 17 Apr. 1966 (G.I. Stage; GISC), on Ment-
zelia involucrata ; 2 22, 6 66, same locality, 17 Apr. 1962
(C.A. Toschi; GISC), on M. involucrata ; 1 6, E branch

Last Chance Cyn., El Paso Mts., 10 Apr. 1960 (C.A. Tos-
chi; GISC); 2 66, Last Chance Cyn., El Paso Mts., 12 Apr.
1964 (R.R. Snelling; LACM), on M. involucrata ; 4 22, 6
66, same except 15 Apr. 1964; 2 22, 13 66, same locality,
6 Apr. 1966 (G.I. Stage; GISC), on M. involucrata; 6 66,
same locality, 6 Apr. 1968 (T.J. Zavortink; TJZC), on M.
involucrata (5 66) and Malacothrix (1 6); 4 66, Red Rock
Cyn., El Paso Mts., 12-14 Apr. 1966 (G.S. Daniels; LACM),
on M. involucrata. Riverside Co.: 1 6, Beal’s Well, 13
Apr. 1949 (P.H. Timberlake; UCR), on Aster abatus; 1 6,
Berdoo Cyn. Rd., 3.9 mi E Dillon Rd., 3 Apr. 1985 (T.J.
Zavortink, S.S. Shanks; TJZC), on M. involucrata; 3 22,
4 66, Blythe, Apr. 1941 (G.E. Bohart; BBSL), on Echi-
nocactus; 1 6, 11 mi S Hwy. 60, Blythe-Niland Rd., 13
Apr. 1949 (R.C. Dickson; UCR), on M. involucrata; 4 22,
4 66, Box Cyn., E of Mecca, 4 Apr. 1937 (P.H. Timberlake;
UCR), on M. involucrata (4 22, 1 6) and Mohavea con-
fertiflora (1 6); 2 22, same locality, 24 Mar. 1953 (P.H.
Timberlake; UCR), on M. involucrata; 1 <5, 1 mi W Corn
Springs Recreation Site, 20 Apr. 1973 (E.M. Fisher; LACM);
76 22, 38 66, Corn Springs Wash, Chuckwalla Mts., 6-8
mi SSE Desert Center, 22 & 25 Mar. 1970, 11-12 Apr.

12 ■ Contributions in Science, Number 451

Snelling and Stage: The bee genus Xeralictus

Figures 30-37. Xeralictus males. Metasoma sternum 8, ventral and lateral views, respectively, of X. timberiakei (30,
31) and X. bicuspidariae (32, 33). 34, 35. Lateral view of genital capsule of X. bicuspidariae (34) and X. timberiakei
(35). 36, 37. Genital capsule, ventral (left half) and dorsal (right half) views of X. bicuspidariae (36) and X. timberiakei
(37). Figures 30-33 by Ruth Ann DeNicola.

1970 (T.J. Zavortink; TJZC), on M. involucrata; 9 22, 6
58, same locality, 26 Apr. 1973 (T.J. Zavortink; TJZC),
on M. involucrata; 2 $8, 3.2 mi W Corn Springs, 1900 ft,
7 Apr. 1994 (R.R. Snelling; LACM), on M. involucrata ;
1 2, 1 3, same except 11 Apr. 1994; 1 2, Cottonwood
Mts., 21 mi E Indio, 18 Mar. 1966 (G.S. Daniels; GISC),
on M. involucrata ; 1 <3, Cottonwood Springs, 26 Apr.
1949 (J.E. Gillaspy; UCB); 4 22, 99 88, 6 mi S Cottonwood
Springs, 23 Mar. 1966 (J.W. MacSwain, G. Salt, P.D. Hurd;
UCB), on M. involucrata; 2 88, 21 mi SW Cottonwood
Springs, 9 Apr. 1952 (R.F. Smith; UCB), on Mentzelia; 1
3, 2 mi W Desert Center, 14 Mar. 1960 (P.H. Raven;
UCB), on M. involucrata; 1 8, Colorado River Aqueduct,
3.6 mi NW Desert Hot Springs, 11 Apr. 1971 (T.J. Za-
vortink; TJZC), on M. involucrata; 20 2$, 8 88, same
locality, 2 Apr. 1985 (T.J. Zavortink, S.S. Shanks; TJZC),
on M. involucrata; 9 $9, 8 88, Midway Cyn., 4.5 mi NW
Desert Hot Springs, 22 Mar. 1967 (G.L Stage; USNM),
on M. involucrata; 10 2$, 12 88, same locality, 29 Mar.

Contributions in Science, Number 451

1967 (R.R. Snelling, G.L Stage; LACM, USNM), on M.
involucrata (8 $2, 12 88) and Mohavea confertiflora (2
22); 5 2$, Whitehouse Cyn., 4.5 mi NW Desert Hot Springs,
20 Mar. 1967 (G.L Stage; USNM); 1 2, 7 88, canyon
between Midway and Whitehouse Cyns., 4.5 mi NW Des-
ert Hot Springs, 13 Mar. 1968 (G.S. Daniels; LACM), on
M. involucrata; 5 22, 4 88, same locality, 6 Apr. 1967 (G.L
Stage; USNM), on M. involucrata (5 22, 2 88), Encelia sp.
(1 8), and Malacothrix sp. (1 8); 102 22, 19 88, same
locality, 18 Apr. 1967 (G.S. Daniels, G.I. Stage; LACM,
USNM), on M. involucrata; 1 <3, same locality, 27 Apr.

1967 (G.L Stage; USNM), on M. involucrata; 16 22, 2 88,
same locality, 1 May 1967 (G.S. Daniels; LACM), on M.
involucrata; 5 22, 14 88, Dry Morongo Wash, 10 Mar.

1968 (1 8), 23 Mar. 1968 (6 88), 24 Mar. 1968 (5 22), 3
May 1968 (7 88) (all G.I. Stage; USNM), on M. involu-
crata; 2 88, Hidden Spring, 2 Mar. 1927 (T. Craig; CAS);
2 33, Indio, 1 Mar. 1958 (G.H. Nelson; UCD); 1 2, 1 3,
15 mi E Indio, 18 Mar. 1958 (E.G. Linsley; UCB), on

Snelling and Stage: The bee genus Xeralictus M 13

Figures 38. Xeralictus timberlakei. Male in Mohavea confertiflora blossom in typical “guarding” posture. Photo by
G.I. Stage.

Mentzelia; 2 66, same locality, 13 Apr. 1949 (P.H. Tim-
berlake; UCR), on M. involucrata; 4 66, 20 mi E Indio,
26 Apr. 1963 (E.I. Schlinger; UCR), on M. involucrata ;
60 $9, 63 66, Little San Bernardino Mts., NW of Desert
Hot Springs, 1 mi E Hwy. 62, 3 8c 10 May 1969 (T.J.
Zavortink; TJZC), on M. involucrata ; 5 99, same except
0.75 mi E Hwy. 62, 3 May 1969, on M. involucrata ; 1 <3,
Mecca, 9 Apr. 1952 (W.H. Lange; UCD); 4 99, same lo-
cality, 18 Feb. 1964 (W.H. Ewart; UCR), on M. involu-
crata; 1 6, Morongo Wash, 3 mi S Morongo Valley, 24
Mar. 1972 (T.J. Zavortink; TJZC), on Datura meteloides;

1 6, Shaver’s Well, 16 Feb. 1964 (R.R. Snelling; LACM),
on M. involucrata ; 11 99, 5 66, same except 21 Mar. 1988,

2 99, 4 66, same except 26 Mar. 1966; 1 9, 2 66, same
except 16 Apr. 1966 (G.I. Stage; GISC); 6 66, 2 mi N
Shaver’s Well, 9 Apr. 1952 (J.W. MacSwain; UCB), on
Mentzelia; 1 9, 2 66, 3 mi W Shaver’s Well, 1 Mar. 1964
(R.R. Snelling; LACM), on M. involucrata; 1 9, 1 6, same
except 28 Mar. 1966; 1 9, 5 66, 4 mi E Shaver’s Well, 9
Apr. 1952 (E.G. Linsley; UCB), on M. involucrata; 1 9,
“Edom” (=Thousand Palms), 14 Mar. 1947 (E.G. Linsley;
UCB), on Geraea; 3 99, 2 66, Thousand Palms Cyn., 15
Mar. 1988 (R.R. Snelling; LACM), on M. involucrata; 1
9, Whitewater, 9 Mar. 1940 (R.M. Bohart; UCD), on
Geraea canescens; 3 99, same locality, 25 Mar. 1934 (C.M.
Dammers; UCR), on Cactaceae; 2 99, same locality, 19
Apr. 1934 (P.H. Timberlake; UCR), on M. involucrata;
1 6, 10 mi E Whitewater, 18 Mar. 1960 (R.M. Bohart;
UCD). San Bernardino Co.: 1 9, 1 6, Baker, 15 Mar. 1935
(AMNH); 8 99, 3 66, Morongo Pass, 22 Apr. 1937 (P.H.
Timberlake; UCR), on M. involucrata; 5 99, 15 66, Mo-

rongo Valley, 17-21 Apr. 1957 (R.R. Snelling and M.D.
Stage; LACM), on M. involucrata; 22 99, 4 66, 14 mi S
Twentynine Palms, 14 Apr. 1935 (P.H. Timberlake; UCR),
on M. involucrata.

DISCUSSION

Apparently little, if any, of the range of X. timber-
lakei lies east of the Colorado River, and the species
appears to be uncommon east of the Chuckwalla
Mountains in Riverside County. The principal part
of the distribution includes those desert mountain
ranges that mark the western edge of the Colorado
Desert. This distribution extends north into the
Mojave Desert, at least as far as El Paso Mountains
in Kern County and the Salt Creek area of San
Bernardino County, approximately 20 mi north of
Baker; there are no records for the Panamint Range
where the host plant is common.

Xeralictus bicuspidariae
new species

Figures 21, 23, 27-29, 32-34, 36

DIAGNOSIS

Male. Median process of metasomal sternum 4
simple, without lateral angles; apicolateral angles

14 ■ Contributions in Science, Number 451

Snelling and Stage: The bee genus Xeralictus

of sternum 5 broadly rounded; head width less than
2.70 mm. Female. Head width not exceeding 2.60
mm and usually less than 2.50 mm and head less
than 1.10 times as long as broad; metasoma fer-
ruginous in populations sympatric with X. timber-
lakei.

DESCRIPTION

Male, measurements (mm): HW 2.10-3.01; FL
1.5-1. 9; WL 5.4-6. 8; TL 8.0-10.3.

Shape and sculpture of head about as described
for X. timberlakei, but labral tubercle less elevated;
genal process near base of mandible lower and dis-
tinctly transverse; hypostomal carina high and la-
in ell if orm but not strongly reflexed laterad.

Mesosoma as described for X. timberlakei but
outer metatibial spur with two,, rarely three, fine,
strongly reclinate teeth.

Metasoma about as in X. timberlakei except:
median process of S4 (Fig. 23) without lateral an-
gles; apicolateral angles of process of S5 (Fig. 27)
broadly rounded in ventral view; process of S8 acute
(Fig. 32).

Female, measurements (mm): HW 2.00- 2.51; FL
1.45-1.94; WL 5. 1-6. 3; TL 7.8-10.9.

Head 1.00-1.05 times as long as broad. Inner eye
margins weakly divergent below, LID about 1.05
times UID. IOD about 2.0 times OD; OOD about
2.1 times OD; OVD about 1.4 times OD. Head
otherwise about as in female X. timberlakei.

Mesosoma as in X. timberlakei. Outer metatibial
spur with two, or rarely three, coarse suberect teeth
near midlength.

Metasoma as in X. timberlakei but terga and
sterna light reddish, tergum 2 usually with distinct
lateral brown spots (see later Discussion).

TYPE MATERIAL

Holotype male: Last Chance Canyon, El Paso Mts.,
Kern Co., California, 15 Apr. 1964 (R.R. Snelling),
on Mentzelia involucrata; in LACM. Paratypes (all
from El Paso Mts.): 8 $2, 45 66, same data as ho-
lotype; 2 22, Iron Canyon, 15 Apr. 1964 (R.R. Snell-
ing), on M. involucrata; 3 22, Red Rock Canyon,
12 Apr. 1966 (G.S. Daniels), on M. involucrata; 5
$6, Mesquite Canyon, 4 May 1969 (T.J. Zavortink),
on M. involucrata (2 66) and Encelia virginiensis
(3 66); 17 22, 7 66, Last Chance Canyon, 6 Apr. 1968
(T.J. Zavortink), on M. involucrata. Paratypes in
AMNH, BBSL, CAS, LACM, USNM, TjZC, CISC,
and UCR.

ADDITIONAL SPECIMENS (not paratypes)

MEXICO, Baja California: 4 2$, 2 66, Isla Angel de la
Guarda, 3, 7, & 25 Mar. 1966 (G.S. Daniels; LACM), on
Mentzelia hirsutissima; 1 6, same locality, 2 Apr. 1973
(H.j. Thompson; TJZC), on M. hirsutissima; 3 22, 5 66,
Isla Mejia, 2 Apr. 1973 (H.J. Thompson; TJZC), on M.
hirsutissima.

Contributions in Science, Number 451

UNITED STATES, Arizona, La Paz Co.3: 4 22, 1 6,

9.1 mi S Quartzsite, 21 Apr. 1966 (P.D. Hurd; UCB), on
M. involucrata. Maricopa Co.: 1 2, 14 mi N Ajo, 20 Mar.
1968 (W.J. Hanson; BBSL), on Mentzelia sp.; 1 2, 1 6, 12
mi SW Gillespie Dam, Gila Bend Mts., 29 Mar. 1969 (T.J.
Zavortink; TJZC), on M. involucrata; 1 2, Phoenix, 21
Apr. 1935 (R.H. Crandall; LACM). Mohave Co.: 9 22, 4
66, Hwy. 93, 5.8 mi S Hoover Dam, 25-26 Mar. 1960
(H.J. Thompson; LACM), on M. tricuspis ; 2 22, 1 6, 9.6
mi SE Hoover Dam, 19 Apr. 1967 (G.I. Stage; USNM),
on M. tricuspis ; 1 2, 1 6, 13.7 mi SE Hoover Dam, 20
Apr. 1967 (G.I. Stage; USNM), on M. tricuspis (2) and
Eucnide urens {6); 2 22, 1 6, Willow Beach, 19 Apr. 1967
(G.I. Stage; USNM), on M. tricuspis; 8 22, 3 66, 2.8 mi E
Willow Beach, 9 Apr. 1967 (G.I. Stage, G.S. Daniels;
USNM), on M. tricuspis; 2 $2, 1 6, same except 20 Apr.
1967 (G.I. Stage; USNM); 2 22, 5 66, same except 2 May
1967 (G.S. Daniels; LACM); 16, same except 9 May 1967
(G.S. Daniels; LACM). Yavapai Co.: 1 2, Hwy. 93, 3.5
mi N Santa Maria River, 26 Mar. 1960 (H.J. Thompson;
LACM), on M. involucrata. Yuma Co.: 2 $2, 3 66, Palm
Canyon, Kofa Mts., 31 Mar. 1968 (R.M. Bohart; BBSL,
UCD); 1 6, 29 mi S Quartzsite, 23 Mar. 1970 (T.J. Za-
vortink; TJZC), on M. involucrata; 2 22, 31 mi S Quartz-
site, 23 Mar. 1970 (T.J. Zavortink; TJZC), on M. invo-
lucrata; 1 6, 30 mi S Quartzsite, 29 Mar. 1969 (T.J.
Zavortink; TJZC), on M. involucrata; 4 22, 2 66, same
except 12 Apr. 1969, on M. involucrata (3 22, 2 66) and
Opuntia basilaris (1 2); 6 22, 6 66, 32 mi S Quartzsite, 23
Mar. 1968 (R.W. Rust, D.R. Miller, R.L. Brumley; BBSL);
23 22, 29 66, 34 mi S Quartzsite, 21 Mar. 1966 (P.D. Hurd,
J.W. MacSwain, W.J. Turner; UCB), on M. involucrata;
1 5, W'elkon, 6 Apr. 1935 (A.L. Meiander; MCZ).

California, Imperial Co.: 1 5, Chocolate Mts., 14 mi
NE Glamis, 18 Mar. 1966 (G.S. Daniels; LACM), on M.
involucrata; 1 2, 2 55, Fossil Cyn., Coyote Mts., 3.5 mi
NNW Ocotillo, 26 Apr. 1970 (T.J. Zavortink; TJZC), on
AT. involucrata; 1 2, 3 55, same except 16 Mar. 1973; 1
5, Glamis, 8 Apr. 1964 (R.M. Bohart; UCD), on Mentzelia
sp.; 1 2, 1 5, 10 mi N Glamis, 30 Mar. 1973 (R.M. Bohart,
C. Goodpasture; UCD); 2 22, 1 5, 2 mi N Midway Well
turnoff, Hwy. 78, 18 Mar. 1966 (G.S. Daniels; LACM),
on M. involucrata; 1 2, 3 66, 2 mi SE Mountain Spring,
26 Apr. 1970 (T.J. Zavortink; TJZC), on M. hirsutissima;
12 22, 5 55, same except 25 Apr. 1973; 2 22, 15 mi N
Ogilby, 10 Mar. 1968 (G.S. Daniels; LACM), on M. in-
volucrata; 1 2, 1 5, Painted Gorge, Coyote Mts., 5 mi N
Ocotillo, 26 Apr. 1970 (T.J. Zavortink; TJZC), on AT.
involucrata ; 2 55, same locality, 17 Mar. 1966 (G.S. Dan-
iels; LACM), on AT. involucrata; 1 6, Picacho Rd., 0.7 mi
N All-American Canal, 10 Mar. 1968 (G.S. Daniels;
LACM), on M. involucrata. Inyo Co.: 1 5, Surprise Cyn.,
Panamint Range, 6 Apr. 1961 (R.P. Allen; LACM), on
Eucnide urens; 2 66, same locality, 9 May 1958 (A. Menke;
UCD). Riverside Co.: 1 5, Chuck walla Mts., 13 mi S Hwy.
60, 13 Apr. 1949 (P.H. Timberlake; UCR), on AT. invo-
lucrata; 2 66, Corn Springs Wash, Chuckwalla Mts., 6-8
mi SSE Desert Center, 22 Mar. 1970 (T.J. Zavortink; TJZC),
on AT. involucrata; 2 22, same except 26 Apr. 1973; 1 2,

3.2 mi W Com Springs, 1900 ft, 7 Apr. 1994 (R.R. Snelling;
LACM), on AT. involucrata; 1 6, Palm Springs, 27 Mar.
1964 (D.F. Veirs; UCD). San Bernardino Co.: 2 66, Calico

3. In 1986, Yuma County, Arizona, was divided into
two counties, the northern one newly created as La Paz
County. The following specimens are, therefore, labeled
“Yuma Co.” but are from localities now in La Paz County.

Snelling and Stage: The bee genus Xemlictus ■ 15

Mts., 7.0 mi NE Barstow, 27 Apr. 1973 (T.J. Zavortink;
TJZC), on M. tridentata; 1 <3 , same except 14.0 mi ENE
Barstow; 2 9$, Daggett, 13 Mar. 1968 (G.S. Daniels, LACM),
on M. tricuspis var. brevicornuta; 1 <3, Needles, 3 Apr.
1951 (J.W. MacSwain; UCB); 2 99, 14 mi S Twentynine
Palms, 14 Apr. 1935 (H.L. McKenzie; BBSL), on M. in-
volucrata. San Diego Co.: 1 9, 1 <3, Borrego Springs, 30
Mar. 1976 (J.L. Neff; CTMI), on M. “ tricuspis ”4; 1 9, 2
<3<3, 3 mi SE Borrego Springs, 15-18 Apr. 1976 (P. Lincoln;
CTMI), on M. “ tricuspis .”4

Nevada, Clark Co.: 1 8, Boulder Dam, 8 Apr. 1973
(F.D. Parker; BBSL); 1 9, Lake Mead, 8 Apr. 1959 (G.E.
Bohart; BBSL), on Encelia farinosa; 1 9, 1 8, same locality,
18 Apr. 1949 (BBSL), on Platyopuntia; 5 99, 1 <3, Lake
Mead Blvd., 8.9 mi E Hwys. 91 6 c 93, 14 May 1969 (T.J.
Zavortink; TJZC), on M. tricuspis; 14 99, 1 8, Springs
Mts., 5 mi N Las Vegas, 3300 ft, 25-26 May 1969 (R.R.
Snelling; LACM), on M. tricuspis ; 446 99, 95 88, Spring
Mts., 13 mi NW Las Vegas, 3000-3400 ft, various data
between 22 Apr. and 25 May, various years (T.J. Zavor-
tink; TJZC), on M. tricuspis.

ETYMOLOGY

The name of this species reflects the association of
X. bicuspidariae with the Bicuspidaria Section of
the genus Mentzelia.

DISCUSSION

This species, like X. timberlakei, is very consistent
in its morphological features and is superficially
very similar to that species. Males of the two species
are especially similar, except in size and in the form
of the metasomal sterna. The features of the male
sterna are so distinctive that an examination of the
metasomal venter is sufficient to distinguish be-
tween the two species (Figs. 20, 21). Females also
differ in size and usually in the color of the meta-
soma.

The metasomal segments of X. timberlakei fe-
males are dark brown with translucent yellowish
margins. Occasional specimens may have the meta-
soma light reddish brown. Typically, females of X.
bicuspidariae have a distinctly red metasoma, usu-
ally with a distinct brown spot on each side of T2.
This characteristic is especially obvious in those
areas where the two species occur together, as well
as in areas where the ranges are adjacent. We know
of only two California sites where the two species
of Xeralictus coexist (El Paso Mountains, Kern
County; Chuckwalla Mountains, Riverside Coun-
ty), but this may result from inadequate collecting.
However, a survey made by one of us (GIS) in 1967
to locate such sites was unfruitful.

The range of X. bicuspidariae is more extensive
than that of X. timberlakei, ranging from above
36°N south to 29°N in Baja California, Mexico.
Eastward, X. bicuspidariae extends to Clark Coun-

4. Because Mentzelia tricuspis is not known to occur
in San Diego County (Prigge, 1993), we presume that these
records are based on misidentification of some other spe-
cies of Mentzelia.

ty, Nevada, and Phoenix, Maricopa County, Ari-
zona. Over most of this range the females are char-
acterized by the distinctly red metasoma. Specimens
from the area of the Colorado River may also have
red legs, at least in part. The one female from Phoe-
nix, Arizona, and those from Isla Angel de la Guar-
da are atypical in that the metasoma is dark reddish
brown, thus similar to X. timberlakei. Some females
from Imperial County, California, and Yuma Coun-
ty, Arizona, have the metasoma dark reddish, but
not as dark as those from Phoenix and Isla Angel
de la Guarda. The two females from San Diego
County, California, are also characterized by darker
red metasoma, but not as dark as those from Im-
perial County. Unfortunately, no specimens from
the area between the California-Mexico border and
Isla Angel de la Guarda are available. Presumably
such specimens would continue the trend toward
increased metasomal darkening.

ACKNOWLEDGMENTS

For the loan of material utilized during the course of this
study we are indebted to the following: P.H. Arnaud, Jr.
(CAS); G.E. Bohart and T.L. Griswold (BBSL); H.E. Evans
(MCZ); the late P.D. Hurd, Jr. (UCB); C.D. Michener
(UKAN); J.L. Neff (CTMI); J.G. Rozen, Jr. (AMNH); the
late R.O. Schuster (UCD); the late P.H. Timberlake and
S.I. Frommer (UCR); and T.J. Zavortink (TJZC). We are
especially grateful to Gil Daniels, Henry Thompson, and
Tom Zavortink for the diligence with which they col-
lected so much of the material on which this study is
based.

An earlier draft of this study was reviewed and mate-
rially improved by constructive comments by J.G. Rozen,
Jr., T.J. Zavortink, and the late C.L. Hogue; T.L. Griswold
and T.J. Zavortink reviewed the final draft and their crit-
icisms are deeply appreciated; Rozen’s assistance with the
larval characterization was particularly helpful.

A special “thank you” is extended to Ruth Ann
DeNicola for her fine illustrations.

LITERATURE CITED

Cockerell, T.D.A. 1927. Two new types of desert bees.

Pan-Pacific Entomologist 4:41-44.

Crawford, J.C. 1917. New Hymenoptera. Proceedings
of the Entomological Society of Washington 19:165-
172.

Darlington, J. 1934. A monograph of the genus Ment-
zelia. Annals of the Missouri Botanical Gardens 21:
106-226.

Eickwort,G.C. 1969. Tribal positions of Western Hemi-
sphere green sweat bees, with comments on their
nest architecture (Hymenoptera: Halictidae). Annals
of the Entomological Society of America 62:652-
660.

McGinley, R.J. 1981. Systematics of the Colletidae, based
on mature larvae with phenetic analysis of apoid
larvae (Hymenoptera: Apoidea). Publications in En-
tomology of the University of California 91:1-307.
— — — . 1987. In Immature insects, ed. F.W. Stehr, 689-
704. Dubuque, Iowa: Kendall/Hunt Publishing
Company.

Michener, C.D. 1944. Comparative external morphol-
ogy, phylogeny, and a classification of the bees (Hy-

16 ■ Contributions in Science, Number 451

Snelling and Stage: The bee genus Xeralictus

menoptera). Bulletin of the American Museum of
Natural History 82:151-326.

Michener, C.D., R.J. McGinley, and B.N. Danforth. 1994.
The bee genera of North and Central American (Hy-
menoptera: Apoidea). Washington, D.C.: Smithson-
ian Institution Press, 209 pp.

Prigge, B. 1993. Loasaceae. In The Jepson manual —
Higher plants of California, ed. J.C. Hickman, 740-
745. Berkeley: University of California Press.

Rozen, J.G., Jr. In prep. Nesting biology and immature
stages of the rophitine bee Sphecodosoma dicksoni
with biological notes on Rophites trispinosus (Hy-
menoptera: Apoidea: Halictidae).

Snelling, R.R. 1 985. The systematics of the hylaeine bees
(Hymenoptera: Colletidae) of the Ethiopian zoo-
geographical region: The genera and subgenera with
revisions of the smaller groups. Contributions in
Science 361:1-33.

Stage, G.I., and R.R. Snelling. In prep. A revision of the
Nearctic Melittidae: The subfamily Dasypodinae
(Hymenoptera: Apoidea).

Thompson, H.J., and W. E. Ernst. 1967. Floral biology
and systematics of Eucnide (Loasaceae). Journal of
the Arnold Arboretum 48:56-88.

Thompson, H.J., and J. Roberts. 1974. In A flora of
southern California, ed. P.A. Munz, 550-554.
Berkeley: University of California Press.

Timberlake, P.H. 1964. A revisional study of the bees
of the genus Perdita F. Smith, with special reference
to the fauna of the Pacific Coast (Hymenoptera,
Apoidea). Part VI. University of California Publi-
cations in Entomology 28:125-388.

Torchio, P.F., J.G. Rozen, Jr., G.E. Bohart, and M. Fav-
reau. 1967. Biology of Dufourea and of its clep-
toparasite, Neopasites (Hymenoptera: Apoidea).
Journal of the New York Entomological Society 75:
132-148.

Zavortink, T.J. 1972. A new subgenus and species of
Megandrena from Nevada, with notes on its for-
aging and mating behavior (Hymenoptera: Andren-
idae). Proceedings of the Entomological Society of
Washington 74:61-75.

. 1974. A revision of the genus Ancylandrena

(Hymenoptera: Andrenidae). Occasional Papers of
the California Academy of Sciences 109:1-36.

Submitted 24 May 1994; accepted 28 October 1994.

Contributions in Science, Number 451

Snelling and Stage: The bee genus Xeralictus ■ 17

A Revision of the Nearctic Melittidae:
The Subfamily Melittinae
(Hymenoptera: Apoidea)

Roy R. Snelling1 and Gerald I. Stage2

ABSTRACT. Of the two melittine genera present in North America, Melitta occurs also in the Palearctic
and Subsaharan regions; Macropis is Holarctic and especially speciose in the Palearctic Region. We treat
the monobasic Nearctic Dolichochile as a subgenus of Melitta ; all four Nearctic species of Melitta are
separated by a key. One new species, M. eickivorti is described from the eastern United States; the type
locality is in New York State.

Macropis is likewise represented by four Nearctic species. These are separated in a key; brief notes
on their biologies, especially host plant data, and their distributions are presented. A neotype for M.
longilingua Provancher is designated and the name placed in synonymy with M. ciliata Patton; M. clypeata
Swenk is a synonym of M. nuda (Provancher).

Taxonomically significant morphological features of the species in both genera are illustrated.

INTRODUCTION

This is the first of two papers that will revise the
species-level taxonomy of the melittid bees of the
Nearctic Region, the only portion of the Western
Hemisphere where these bees are known to be pres-
ent. Michener (1981) revised the worldwide higher
classification of the Melittidae. Three subfamilies
were recognized: Meganomiinae (four genera, re-
stricted to eastern and southern Africa), Melittinae
(five genera, Holarctic and African), and Dasypodi-
nae (eight genera in three tribes, Holarctic and Af-
rican). The genera Ctenoplectra and Ctenoplectri-
na, formerly associated with the Melittidae, were
removed to the new family Ctenoplectridae by
Michener and Greenberg (1980).

SPECIMENS EXAMINED

Specimens examined for this study are from the
following institutional and private collections:
American Museum of Natural History (AMNH);
Bee Biology and Systematics Laboratory (USDA) at
Utah State University (BBSL); California Academy
of Sciences (CAS); Canadian National Collection
(CNC); Cornell University (CORN); Division of Plant
Industry, Florida Department of Agriculture (DPIF);
Michigan State University (MSU); Museum of Com-
parative Zoology (MCZ); Natural History Museum
of Los Angeles County (LACM); Snow Entomo-
logical Museum, University of Kansas (SEMC); G.I.

1. Entomology Section, Natural History Museum of
Los Angeles County, 900 Exposition Boulevard, Los An-
geles, California 90007.

2. R.F.D. #1, Bowles Road, Stafford Springs, Con-
necticut 06076.

Stage, personal collection (GIS); United States Na-
tional Museum of Natural History (USNM); Uni-
versity of California, Berkeley (UCB) and Riverside
(UCR); and University of Nebraska (UNEB).

SYSTEMATICS

There are two subfamilies of Melittidae represent-
ed within the Nearctic Region. They have been
separated by Michener (1981) as follows:

Melittinae: “Paraglossa densely hairy; forewing
with two or three submarginal cells, second (if only
two cells) or second plus third as long as or longer
than first, first transverse cubital (— base of second
submarginal cell) slanting, usually well separated
from first recurrent vein. Larvae spin cocoons.”
Dasypodinae: “Paraglossa largely bare, usually
markedly shorter than suspensorium, hairs largely
limited to apex, or paraglossa absent; forewing with
two submarginal cells, second usually shorter than
first, first transverse cubital (=base of second sub-
marginal cell) usually close to first recurrent vein.
Known larvae do not spin cocoons.”

Of the five genera of Melittinae recognized by
Michener, three occur in North America (although
we here treat one of these as a subgenus) and may
be separated by the following key, modified from
that of Michener.

KEY TO

NORTH AMERICAN GENERA
OF MELITTINAE

a. Forewing with three submarginal cells; male with
neither pygidial plate nor yellow face marks;
male gonostylus broadly fused with gonocoxite
(Figs. 13-19); male sternum 8 ending in beveled
area simulating a pygidial plate Melitta

Contributions in Science, Number 451, pp. 19-31
Natural History Museum of Los Angeles County, 1995

Figures 1-4. Male metasomal sterna 6 and 7, respectively, of Melitta americana 1, 2) and M. eickworti (3, 4). Figures
1 and 2 by Ruth Ann DeNicola.

b. Forewing with two submarginal cells; male with
pygidial plate and yellow face marks; male gon-
ostylus long, slender at base, and articulated with
gonocoxite (Figs. 20-27); male sternum 8 with-
out modified beveled area Macropis

Genus Melitta Kirby

Melitta Kirby, 1802:117. Type-species: Melitta tri-
cincta Kirby, 1802; designated by Richards, 1935.
Cilissa Leach, 1815:155. Type-species: Andrena

20 ■ Contributions in Science, Number 451

haemorrhoidalis Fabricius, 1775; designated by
Westwood, 1840.

Kirbya Lepeletier, 1841:145. Type-species: Melitta
tricincta Kirby, 1802; designated by Sandhouse,
1943. Preoccupied.

Pseudocilissa Radoszkowski, 1891:241. Type-spe-
cies: ( Cilissa robusta Radoszkowski, 1876) =
Melitta dimidiata Morawitz, 1876. Monobasic.

Melitta subg. Brachycephalapis Viereck, 1909:47.
Type-species: Melitta ( Brachycephalapis ) cali-
fornica Viereck, 1909. Monobasic and original
designation.

Snelling and Stage: North American Melittinae

Figures 5-8. Male metasomal sterna 6 and 7, respectively, of Melitta calif ornica (5, 6) and M. melittoides (7, 8).
Figures by Ruth Ann DeNicola.

Dolichochile Viereck, 1909:49. Type-species: Dol-
ichochile melittoides Viereck, 1909. Monobasic
and original designation.

The following are diagnostic characters for Melitta:
Melittine bees with three submarginal cells; mouth-
parts ordinary for the group, maxillary palpus two-
to six-segmented; scopa on female metatibia and
metabasitarsus simple, these segments slender; pro-
podeal triangle large, dull; seventh metasomal ster-

Contributions in Science, Number 451

num of male with large disc and insignificant apical
lobes.

This primarily Holarctic genus is most diverse in
the Palearctic Region. A few species occur in south-
ern Africa. Four species are known in North Amer-
ica. Little is known of the biology of any of these
Nearctic species. The three species found in the
eastern United States are apparently oligoleges on
various Ericaceae, while the single western species

Snelling and Stage: North American Melittinae ■ 21

Figures 9-15. Male metasomal sternum 8 and genitalia (lateral view) of Melitta americana (9, 13), M. eickworti (10),
M. californica (11, 15), and M. melittoides (12, 14). Figures 8 and 11-15 by Ruth Ann DeNicola.

appears to be oligolectic on Sphaeralcea (Malva-
ceae).

Although Michener (1981) treated Dolichochile
as a genus apart from Melitta , we prefer to regard
it as a subgenus within Melitta . In our view, the
very numerous character states shared between the

22 ■ Contributions in Science, Number 451

two (all discussed by Michener) outweigh the few
apomorphies that characterize Dolichochile, all of
which are modifications of the female mouthparts
that are presumably related to foraging behavior.
We believe that the structural distinctiveness of
Dolichochile is insufficient to warrant separate ge-

Snelling and Stage: North American Melittinae

Figures 16-19. Male genitalia, ventral view of Melitta melittoides (16), M. californica (17), M. eickworti (18), and
M. americana (19). Figures 16, 17, and 19 by Ruth Ann DeNicola.

neric status. Michener admitted that recognition of
Dolichochile as a genus rendered Melitta paraphy-
letic. Like Michener, we are not bothered by para-
phyletic genera when there is a practical reason for
their recognition. But Dolichochile, with but a sin-
gle species, does not, in our view, represent a jus-
tifiable case for such recognition. We are in agree-
ment with Michener et al. (1994) in treating Doli-
chochile as a subgenus of Melitta .

Contributions in Science, Number 451

KEY TO

SPECIES OF MELITTA

1 Maxillary palpus six-segmented; female man-
dible shorter than eye length and with sub-
apical tooth; labrum with wedge-shaped,
slightly depressed, median impunctate area
partially dividing smooth area (subg. Melitta)
2

Snelling and Stage: North American Melittinae ■ 23

- Maxillary palpus five-segmented; female man-
dible slightly longer than eye, distal half a
long, flattened, pointed blade, with two small
teeth on inner margin; labrum uniformly con-
vex and impunctate except near apical margin
(subg. Dolichochile ) . . . melittoides (Viereck)

2(1) Male, antenna 13-segmented and sternum 8
with disk-shaped pygidium-like process visible
at apex of metasoma 3

- Female, antenna 12-segmented and sternum 8

not visible, but tergum 6 with triangular py-
gidial plate 5

3(2) Distal margin of metasomal sternum 6 broadly
concave (Figs. 1-3); pygidium-like apex of ster-
num 8 smooth and shiny; with head in full
frontal view, distance from lateral ocellus to
upper head margin a little greater than diam-
eter of anterior ocellus 4

- Distal margin of metasomal sternum 6 straight
or convex (Fig. 5); pygidium-like apex of ster-
num 8 dull and conspicuously roughened; with
head in full frontal view, distance from lateral
ocellus to upper head margin less than diam-
eter of anterior ocellus . californica Viereck

4(3) Apical disc of sternum 8 with longitudinal me-
dian impression, disc convex on either side;
disc of metasomal terga 2-4 subcontiguously
to densely punctate .... americana F. Smith

- Apical disc of sternum 8 flat to concave, with-

out median impressed line; discs of metasomal
terga 2-4 sparsely punctate

eickworti, new species

5(2) Punctures on upper two-thirds of clypeus shal-
low and indistinct, interspaces dull and tes-
sellate, some exceeding one puncture diame-
ter; inner eye margins and vertex without
blackish pilosity; pale hairs of head and me-
sosoma more or less fulvous 6

- Punctures on upper two-thirds of clypeus deep

and distinct, mostly subcontiguous, interspac-
es smooth; inner eye margins and vertex with
conspicuous blackish pilosity; hairs of head
and mesosoma definitely whitish

californica Viereck

6(5) Punctation of discs of metasomal terga 2-4
subcontiguous to dense americana F. Smith

- Punctation of discs of metasomal terga 2-4
sparse to scattered . . eickworti, new species

Melitta ( Melitta ) americana

(F. Smith)

Figures 1, 2, 9, 13, 19

Cilissa americana F. Smith, 1853:123; 9.

Melitta americana: Cockerell, 1906:5-6; 9. Mitch-
ell, 1960:522-524; 9 6.

Melitta americaniformis Viereck, 1909:50; 9.

This species of the eastern United States ranges
from Massachusetts to Florida and Mississippi and
flies from April to July. Mitchell (1960) recorded
M. americana as a visitor to flowers of Polycodium

24 ■ Contributions in Science, Number 451

and Rubus, and Michener (1947) found it on Gay-
lussacia dumosa in southern Mississippi.

In addition to the types of C. americana (BMNH)
and M. americaniformis (USNM), we have exam-
ined the following specimens of M. americana :

FLORIDA, Alachua Co.: 1 5, Austin Carey Memorial
Forest, 31 Apr. 1975 (G.B. Fairchild; LACM); 1 9, same
locality, 1-2 May 1975 (G.B. Fairchild; LACM). Franklin
Co.: 1 6, McIntyre, 16 Apr. 1982 (L.L. Pechuman; CORN).
Jackson Co.: 1 9, Florida Caverns State Park, 22 Apr. 1972
(H. Greenbaum; SEMC), Malaise trap. Lake Co.: 1 9,
Alexander Spring Camp, Ocala National Forest, 31 Mar.
1974 (G.C. Eickwort; CORN). Suwanee Co.: 4 99, 3 66,
Suwanee River State Park, 12-25 Apr. 1977 (J.R. Wiley;
LACM, UTSU). Wakulla Co.: 2 66, Sopchoppy, 2-3 Apr.
1981 (L.L. Pechuman; CORN). GEORGIA, Richmond
Co.: 1 9, Fort Gordon, 13 Apr. 1958 (R.R. Snelling; LACM),
on Vaccinium; 2 66, same locality and collector, 8 May
1958, on Vaccinium; 1 9, same locality and collector, 30
May 1958. MASSACHUSETTS, Barnstable Co.: 11 99,
Eastham, 27 June 1908 (C.W. Johnson; USNM). MISSIS-
SIPPI, Forrest Co.: 1 9, Hattiesburg, 16 Apr. 1944 (C.D.
Michener; SEMC). NEW JERSEY, Burlington Co.: 1 9,
1 6, Browns Mills, 10 June 1921 (AMNH); 1 9, Browns
Mills, 29 June 1921 (AMNH). Ocean Co.: 1 9, Lakehurst,
19 May (AMNH). NORTH CAROLINA, Harnett Co.:
1 9, (no further locality), 10 May 1933 (T.B. Mitchell;
BMNH). Onslow Co.: 1 9, Holly Shelter, 18 May 1950
(T.B. Mitchell; BBSL). Sampson Co.: 1 6, Ivanhoe, 12
Apr. 1945 (T.B. Mitchell; BBSL).

Melitta (Melitta)
californica Viereck
Figures 5, 6, 11, 15, 17

Melitta ( Brachycephalapis ) californica Viereck,
1909:47; 6.

Melitta wilmattae Cockerell, 1937:3; 9.

Melitta maritima Cockerell, 1941:344; <5.

Melitta californica: Michener, 1981:120.

Michener (1981), after examining relevant type
specimens, established the preceding synonymy. He
also cited data for the few known specimens of
this species, presently known only from desert
regions of southwestern Arizona, southeastern Cal-
ifornia, and Lower California, Mexico. The several
specimens collected by G.E. Bohart at Constitu-
cion, Baja California Sur, possess more numerous
blackish hairs, especially on the mesoscutum and
apical metasomal terga, than do specimens from
more northern localities. They are otherwise much
like the specimens collected near San Felipe and in
southern California. In addition to the records cited
by Michener, we can add the following:

New records: MEXICO, BAJA CALIFORNIA, 3 66,
3 mi S San Quintin, 8-12 Mar. 1960 (D.P. Gregory; UCB),
on Lycium parishii ; 1 6, same, except on Encelia cali-
fornica; 21 99, 26 66, San Felipe, 24-28 Mar. 1963 (G.I.
and K.N. Stage; GIS, LACM), on Sphaeralcea orcuttii (20
99, 26 66) and Dalea megacarp a (1 6); 1 6, 3 mi N San
Felipe, 25 Mar. 1964 (J.C. Hall; UCR); 1 6, 22 km N
Punta Prieta, 26 Mar. 1979 (E.M. Fisher; LACM), on
Viscainoa geniculata. BAJA CALIFORNIA SUR, 1 9, 3
66, Constitucion, 22 Feb. 1974 (G.E. Bohart; BBSL).

Snelling and Stage: North American Melittinae

UNITED STATES , ARIZONA, Pinal Co.: 3 <$<$, 10
km W Maricopa, 13 Mar. 1989 (R.L. Minckley and W.T.
Wcislo; SEMC), on Sphaeralcea; 5 88, same, except 20
Mar. 1989; 1 <5, same, except 21 Mar. 1989. CALIFOR-
NIA, Imperial Co.: 1 8, Westmoreland, 6 Apr. 1949 (P.D.
Hurd; UCB); 1 2, 2 88, 3 mi NW Glamis, 4 Mar. 1972
(A.R. Hardy; BBS, LACM), on Sphaeralcea. Riverside
Co.: 1 <5, 18 mi W Blythe, 22 Mar. 1974 (F. Parker and
R. Bitner; BBSL). San Diego Co.: 1 8, Coronado, 15 May
1890 (F.E. Blaisdell; CAS); 2 88, Torrey Pines State Park,
no date (A.R. Moldenke), on Coreopsis maritima.

Melitta ( Melitta ) eickworti,
new species
Figures 3, 4, 10, 18

Melitta americana: Cane et al., 1985:135-142.
DIAGNOSIS

This species most closely resembles M. americana;
both sexes are separable by the much sparser punc-
tation of the mesoscutum, scutellum, and basal
metasomal terga, as noted in the key. From the
other three North American species, M. eickworti
is separable by the features noted in the key.

DESCRIPTION

FEMALE. Measurements (mm). Head width 2.9-
3.3, head length 2.6-2.8, wing length 7.7-8.5, total
length 10.5-12.1.

Structure and Punctation. Head. 1.12-1.19 times
as broad as long; inner eye margins slightly diver-
gent below, upper interocular distance 0.86-0.93
times lower interocular distance; in frontal view,
vertex margin strongly arched above ocelli, distance
from ocelli to margin distinctly greater than di-
ameter of anterior ocellus. Interocellar distance
slightly greater than diameter of anterior ocellus;
ocellocular distance about 2 times diameter of an-
terior ocellus. First flagellar segment about 2 times
diameter of anterior ocellus. First flagellar segment
distinctly longer than broad and about one-third
longer than second segment, second segment dis-
tinctly broader than long.

Basal two-thirds (approximately) of clypeus
densely tessellate and slightly shiny between mod-
erate punctures (about 0.05 mm diameter) that are
mostly separated by about one puncture diameter
or less, but with narrow, median, nearly impunctate
line; apical one-third with broader impunctate area,
median portion lightly tessellate and somewhat
shiny, grading to smooth and shiny toward sides.
Paraocular area shiny between fine (0.03 mm di-
ameter) to moderate punctures that are mostly sub-
contiguous, but with irregular interspaces exceed-
ing a puncture diameter. Frons densely tessellate
and dull, punctures fine and obscure, mostly sep-
arated by one puncture diameter or less; side of
face shinier and less sharply tessellate between fine
close to sparse punctures; vertex with shiny, nearly
impunctate area between ocelli and eyes, otherwise

Contributions in Science, Number 451

tessellate and slightly shiny between fine, mostly
subcontiguous punctures; gena slightly shiny, densely
tessellate between sparse, obscure, fine punctures.

Mesosoma. Middle one-half or more of meso-
scutum shiny and polished between moderate
punctures separated by 1.0-2. 5 puncture diameters,
punctures becoming subcontiguous only anteriorly
and laterally, where interspaces become more or
less distinctly tessellate and dull. Scutellum shiny,
with scattered fine punctures on anterior two-thirds;
posterior one-third, and along midline nearly to
base, subcontiguously to contiguously punctate with
moderate punctures. Mesepisternum dull and
sharply tessellate between subcontiguous to con-
tiguous moderate punctures. Metepistemum mod-
erately shiny and less sharply tessellate, virtually
impunctate. Side of propodeum anteriorly similar
to metepistemum, becoming dull and subcontig-
uously punctate distad; basal area with irregular,
widely spaced rugulae anteriorly, interspaces mod-
erately shiny, remaining area distinctly tessellate and
less shiny; disc slightly shiny, reticulate-punctate.

Wings transparent, slightly brownish, darker be-
yond cells; stigma and veins light brown.

Metasoma. Tergum 1, anterior to marginal im-
punctate band with a narrow zone of sparse, mod-
erate, piligerous punctures, basad of which disc is
smooth and shiny between widely scattered fine
punctures; disc of tergum 2 smooth and shiny be-
tween sparse to scattered minute to fine punctures;
tergum 3 similar to second, but some punctures
moderate in size; tergum 4 smooth and shiny be-
tween scattered, moderate punctures; tergum 5
slightly shiny and distinctly tessellate between sub-
contiguous to close, moderate punctures.

Pilosity. Mostly whitish, somewhat yellowish on
sides and apical margin of clypeus and on lower
gena; yellowish red on mandible; hairs of mesoscu-
tum mostly pale but with sparse blackish hairs; those
in center entirely blackish, but no dark hairs be-
tween parapsidal line and margin; scutellum largely
pale pubescent, but a few blackish hairs in center.
Metasomal tergum 1 without definite preapical band
of pale hairs; terga 2-4 with narrow apical bands
of appressed, white, plumose hairs that are weak
or interrupted at middle; discal hairs short, simple,
and pale on terga 2-4, long, plumose, and blackish
on terga 5 and 6.

Color. Head and body blackish, legs dark brown-
ish; flagellum dark reddish brown beneath; tegula
clear yellowish.

MALE. Measurements (mm). Head width 2.5-
2.9, head length 2.4-2.7, wing length 7.0-8. 1.

Structure and Punctation. Head. 1.03-1.11 times
as long as broad; inner eye margins slightly con-
vergent below, upper interocular distance 1.03-1.06
times lower interocular distance; in frontal view,
vertex margin strongly elevated behind ocelli, dis-
tance from ocelli to margin distinctly greater than
diameter of anterior ocellus. Interocellar and ocel-
locular distances about 2 times diameter of anterior
ocellus. First flagellar segment about 0.75 times

Snelling and Stage: North American Melittinae ■ 25

length of second, second about 0.80 times length
of third.

Clypeus finely rugosopunctate but with shiny,
narrow, impunctate median line and transverse im-
punctate preapical band. Supraclypeal area subcon-
tiguously punctate, grading to slightly larger, dense
punctures on frons; punctures of paraocular area
mostly subcontiguous, interspaces shiny; preoccip-
ital area similar but interspaces lightly tessellate;
vertex, adjacent to lateral ocelli, with large, smooth,
shiny, impunctate area.

Mesosoma. Similar to that of female, but me-
soscutum almost entirely smooth and shiny, with-
out anterior and lateral zones of subcontiguous
punctures.

Wings as described for female.

Metasoma. Similar to that of female. Hidden
sterna and genitalia similar to those of M. ameri-
cana but differing as illustrated (Figs. 3, 4, 10, 18),
but apical disc of sternum 8 concave and with dis-
tinct longitudinal median impression.

Pilosity. Similar to that of female, but clypeus
hidden beneath dense prostrate plumose hairs and
face generally with hairs denser and longer; vertex
and gena with some long, fuscous hairs among the
pale hairs; center of mesoscutum with few or no
fuscous hairs; hair bands of metasomal terga 2-4
very weak and interrupted in middle; terga 5-7
largely dark pubescent, without pale hair bands,
but with some pale hairs at sides.

Color. Similar to that of female.

TYPE MATERIAL

Holotype 9 from South Hill Preserve, vicinity of
Ithaca, Tompkins County, New York, 10 June 1981
(G.C. Eickwort), on Vaccinium stramineum. Para-
types: 8 99, 4 65, Ithaca vicinity (South Hill Preserve,
South Hill Swamp, Dawes Hill), 9-19 June (G.C.
Eickwort), on V. stramineum ; 10 99, 5 66, Dawes
Hill, 2 mi SSW West Danby, Tompkins Co., New
York, 11-15 June (G.C. Eickwort), on V. strami-
neum; 3 99, 6 66, Hector Land Use Area, near Reyn-
oldsville, Schuyler Co., New York, 12-30 June (G.C.
Eickwort), on V. stramineum. Holotype and most
paratypes in CORN; 2 99, 2 66 paratypes in LACM.

ADDITIONAL MATERIAL

In addition to the paratypes listed above, we have
seen the following nonparatypic material of M.
eickworti.

GEORGIA, County unknown: 1 5, Indian Grave Gap,
21 May 1952 (P.W. Fattig; USNM). MARYLAND, Mont-
gomery Co.: 13 92, 10 66, Plummers Island, 26 May 1972
(P.D. Hurd, Jr.; USNM), on V. stramineum. NEW JER-
SEY, Morris Co.: 1 2, Newfoundland, 26 May 1910
(AMNH). NORTH CAROLINA, Buncombe Co.: 1 6,
Black Mountains, 9 May 1927 (J.C. Crawford; USNM);
1 2, same locality and collector, 17 May 1927, on V.
stramineum; 1 6, Black Mountain, 30 May 1911 (AMNH),
on Polycodium; 1 6, same, except 4 June 1928; 3 66, same

26 ■ Contributions in Science, Number 451

locality, “1911 Expedition” (AMNH). County unknown:
1 6, Mt. Greybeard, 23 May (N. Banks; USNM). TEN-
NESSEE, Morgan Co.: 1 2, Burrville, 29 May 1959 (B.
Benesh; CORN).

ETYMOLOGY

This new species is named for, and dedicated to,
the late George C. Eickwort, so tragically killed in
an automobile accident in Jamaica on 11 July 1994;
George had originally recognized the possible nov-
elty of this species and called it to our attention.

DISCUSSION

The available specimens show little variation be-
yond that noted in the preceding description. Both
sexes of the series collected by Hurd at Plummers
Island have the mesosomal pubescence yellower
than do the types, and the females have more black-
ish hairs on the disc of the mesoscutum.

There is no doubt that this species has been con-
fused with M. americana in the past. In general,
M. americana is a more southern species, uncom-
monly encountered north of North Carolina. In
the northern United States, M. americana is largely
replaced by M. eickworti , which ranges south to
Mississippi through the southern mountains.

Cane et al. (1985), in their description of the
pollination ecology of Vaccinium stramineum, cit-
ed this species as M. americana.

Melitta ( Dolichochile )
melittoides Viereck
Figures 7, 8, 12, 14, 16

Dolichochile melittoides Viereck, 1909:49; 9. Cock-
erell, 1911:672; 9. Michener, 1981:41-42; 9 6.
Melitta ( Dolichochile ) melittoides: Michener, 1951:
1134. Hurd, 1979:1979.

Melitta melittoides: Mitchell, 1960:524-525; 9 6.

This is another species of the eastern United States,
ranging from New Hampshire south to Tennessee
and Georgia. Specimens have been recorded
(Mitchell, 1960) from flowers of Polycodium, Xol-
isma, and Zenobia, all Ericaceae. Although most
records are from May and June, M. melittoides has
been taken as late as September in Virginia accord-
ing to Mitchell (1960).

Genus Macropis Klug

Megilla Fabricius, 1804:328. Type-species: Megilla
labiata Fabricius, 1805; designated by West-
wood, 1840. Suppressed by International Com-
mission on Zoological Nomenclature, Opinion
1383, 1986. Also: Apis acervorum Linne, 1758,
designated by Richards, 1935. Macropis Klug,
1809:107, no. 16. Type-species: Megilla labiata
Fabricius, 1805. Monobasic.

The International Commission on Zoological No-
menclature, Opinion 1383 (1986), designated Apis
pilipes Fabricius, 1775, an anthophorine bee, as the

Snelling and Stage: North American Melittinae

type species of Megilla Fabricius, 1805, thus effec-
tively rendering Megilla a junior synonym of An-
thophora Latreille, 1803. Macropis, in the generally
accepted sense, was validated and placed on the
Official List of Generic Names in Zoology and Me-
gilla labiata on the Official List of Specific Names
in Zoology.

The characters cited in the key to genera of Me-
littinae will easily separate Macropis from others
in this subfamily. This is a Holarctic genus but
much more diverse in the Paiearctic Region, where
there are three subgenera recognized (Michener,
1981). The few North American representatives all
belong to the nominate subgenus.

The 11 known species of Macropis are obli-
goleges on the genus Lysimacbia (Primulaceae).
The female bees gather pollen and the floral oils
of Lysimacbia when provisioning their nest cells.
The floral oil of Lysimacbia is secreted by tri-
chomous elaiophores located basad on the floral
petals and stamens. The oil is used not only in
nest provisioning but also in lining the cells (Cane
et ah, 1983).

KEY TO

NEARCTIC SPECIES
OF MACROPIS

1 Male, antenna 13-segmented and face yellow-
maculate 2

- Female, antenna 12-segmented and face whol-
ly dark 6

2(1) Metasomal dorsum dull, punctures coarse and
distinct, separated by less than a puncture di-
ameter 3

- Metasomal dorsum polished, with scattered

minute punctures 4

3(2) Outer surface of metabasitarsus polished be-
tween sparse piliferous punctures

steironematis Robertson

- Outer surface of metabasitarsus dull and mi-
nutely roughened between punctures

s. opaca Michener

4(2) Face mostly yellow below level of antennal
sockets (i.e. conspicuous supraclypeal and lat-
eral face marks are present); metatibia with
one or two distal tooth-like processes on inner
surface at base of apical spurs and one or both
apical spurs reduced (Figs. 28, 30) ....... 5

- Supraclypeal and lateral face marks reduced

or absent; metatibia without distal tooth-like
processes on inner surface and both apical spurs
normally developed (Fig. 29)

nuda (Provancher)

5(4) Metatibia with two tooth-like distal processes
and both spurs reduced (Fig. 28); labrum dark
ciliata Patton

- Metatibia with one tooth-like distal process

and only outer spur reduced (Fig, 30); labrum
pale patellata Patton

6(1) Metasomal terga polished between scattered,
minute punctures 7

Contributions in Science, Number 451

- Metasomal terga dull, punctures deep and dis-

tinct, separated by less than a puncture di-
ameter steironematis Robertson

7(6) Disc of clypeus shiny between distinctly sep-
arated punctures, some interspaces as large as,
or larger than, one puncture diameter; pos-
terior face of propodeum mostly smooth and
shiny, with sparse, obscure, fine punctures;
process of labrum high, conspicuous, convex,
and strongly carinate 8

- Disc of clypeus uniformly subcontiguously

punctate and only moderately shiny; posterior
face of propodeum, especially dorsocephalad,
with punctures deep and distinct, mostly sep-
arated by one puncture diameter or less; pro-
cess of labrum low, inconspicuous, and weak-
ly carinate patellata Patton

8(7) Hairs on outer face of meso- and metabasi-
tarsus dark brownish; anterior rim of propo-
deal triangle roughened and with fine, oblique
rugules nuda (Provancher)

- Hairs on outer face of meso- and metabasitarsi

(except apical brush) whitish; anterior rim of
propodeal triangle smooth and shiny, at least
across middle one-third ciliata Patton

Macropis ciliata Patton

Figures 20, 24, 28

Macropis ciliata Patton, 1880:31; 9. Michener, 1938:
135; 6 9. Mitchell, 1960:526; 9 6.

Macropis longilingua Provancher, 1888:424; 9.
NEW SYNONYMY.

Provancher (1888) described M. longilingua from
a female specimen. This specimen is apparently lost;
it has never been identified among the specimens
in Provancher’s collection. The description is in-
adequate, and there is no certainty that it is based
on a Macropis. If it is, in fact, a Macropis, then it
could be either M. ciliata or M. patellata. Because
both species are present in eastern Canada, there is
no way to determine which of the two species Pro-
vancher may have had before him or if, in fact, his
specimen was correctly assigned to Macropis.

To settle the status of Provancher’s name, we
have chosen a specimen of M. ciliata and desig-
nated it to be the neotype of M. longilingua, thus
rendering Provancher’s name a synonym of the old-
er Patton name. The neotype female and two neo-
paratype females of M. longilingua are from Flat-
bush, New York, collected 20 June 1896 by J.L.
Zabriskie; the neotype and one neoparatype are
deposited in the collections of the AMNH, and
one neoparatype is deposited in LACM.

Males of this species are easily recognized by the
short metatibial spurs (Fig. 28) and the wholly ase-
tose genitalic structures (Figs. 20, 24). Females are
most similar to those of M. nuda because in both
species the posterior face of the propodeum is mostly
smooth and shiny and both have similar labral tu-
bercles. However, females of M. ciliata have the
hairs on the outer face of the metabasitarsus (except

Snelling and Stage: North American Melittinae ■ 27

Figures 20-30. Macropis males. Right half of genitalia, dorsal view, and profile of gonostylus of M. ciliata (20, 24),
M. nuda (21, 25), M. patellata (22, 26), and M. steironematis (23, 27). Metatibial spurs of M. ciliata (28), M. wwdtf
(29), and M. patellata (30).

the distal brush) whitish rather than dark brown to
blackish.

The recorded distribution of this species encom-
passes an area from Wisconsin to Quebec and
Maine, south to Georgia. According to Mitchell

28 ■ Contributions in Science, Number 451

(1960), it has been taken on flowers of Apocynum,
Houstonia, and Hydrangea. In Maryland we found
M. ciliata at flowers of Rhus typhina and Lysi-
machia quadrifolia ; females were observed col-
lecting pollen only from the latter plant species. We

Snelling and Stage: North American Melittinae

have also seen a female from Black Mountain, North
Carolina, collected on Ceanothus americanus.

Macropis nuda (Provancher)

Figures 21, 25, 29

Eucera nuda Provancher, 1882:174; 9.

Macropis ciliata: Provancher, 1888:320; 9. MIS-
IDENTIFICATION.

Macropis {Macropis) morsei Robertson, 1897:338;
6 9.

Macropis clypeata Swenk, 1907:293; <3. NEW
SYNONYMY.

Macropis morcei (sic!): Michener, 1938:135; <3 9.
Macropis nuda: Mitchell, 1960:527; 9 <3. Rozen and
Jacobsen, 1980:1-11. Cane et al., 1983:257-264.

The type of M. clypeata has been examined and
found to be inseparable from males of M. nuda as
we understand the species.

Females of M. nuda are easily recognized by the
dark hairs on the meso- and metabasitarsi and the
smooth, sparsely punctate posterior face of the pro-
podeum. Dark-haired basitarsi are characteristic of
M. steironematis also, but in that species the pro-
podeal disc is closely and sharply punctate and the
metasomal terga are densely and sharply punctate.
Males of M. nuda have yellow face marks largely
confined to the clypeus. The labrum is dark, and
the mandible usually is without a basal yellow spot;
a lateral face mark is sometimes present adjacent
to the clypeus, but it never fills the area between
the eye and the clypeus. An irregular, small supra-
clypeal spot may be present but is usually lacking.
Males of the other species have the face largely
yellow below the level of the antennal sockets, with
large supraclypeal marks and the side of the face
yellow between the clypeus and the inner eye mar-
gin. The male gonostylus is distinctive in the pres-
ence of numerous short setae on the outer face
(Figs. 21, 25).

The nesting biology of M. nuda was described
by Rozen and Jacobson (1980), based on obser-
vations made in the Edmund Niles Huyck Reserve,
Albany County, New York. They observed several
nest concentrations along a roadway embankment.
“Nest entrances may be hidden by objects on the
ground, and the surface of the nesting site tends to
be partly covered with moss or other low vegeta-
tion. . . . Nests are shallow with main tunnels trav-
eling considerably laterally.” Cells were usually
placed in linear series of two, but series of three or
four cells were sometimes found.

Females were taking pollen from a strand of Ly-
simachia ciliata , located about 25 m from the nest-
ing area. Rozen and Jacobsen’s observations were
mostly made between 13 and 29 July; at the latter
time, no nesting activity was seen, although a few
M. nuda were still visiting flowers. Larvae and co-
coons were described and figured.

Cane et al. (1983) reported their observations on
male behavior, foraging, grooming, and nest struc-
ture and provisioning by females, at two sites in

Contributions in Science, Number 451

New York State; one of these sites is that reported
previously by Rozen and Jacobsen (1980). The sum-
mary of the study by Cane et al. (1983) is taken
from their abstract: the first demonstrated use of
floral oils in nest linings and nest provisions is re-
ported for M. nuda. “The floral oil chemistry is
closely analogous to the cell-lining lipids secreted
from the Dufour’s glands of many other bee genera
[but not by Macropis, in which the glands are poor-
ly developed]. Details of oil and pollen collection
and transfer behaviors at Lysimachia ciliata are
compared with those of other bee taxa. The si-
multaneous pollen and oil collection behaviors are
reminiscent of Melitta ‘buzz pollination’ {sensu
Buchmann, 1978).”

The range of M. nuda extends from southern
Canada, where it is transcontinental, through the
northeastern United States, at least as far south as
New Jersey, and west to Montana, Colorado, and
Idaho (Moser Camp Ground, Cub River Canyon,
Franklin Co., 30 June to 25 July, on Lysimachia
thyrsiflora; Mapleton, Franklin Co., 15 July 1978;
all BBSL). Mitchell (1960) recorded floral visits to
species of Apocynum, Geranium, Rubus, and Vac-
cineum. As already noted, Rozen and Jacobsen
found Lysimachia ciliata to be the pollen source
at their site. To these floral records we can add
Aralia hispida, Lactuca pulchella, and Ceanothus
americana.

Macropis patellata Patton

Figures 22, 26, 30

Macropis patellata Patton, 1880:33; 6. Michener,

1938:135; <5 9. Mitchell, 1960:528-529; 9 3.

Males of M. patellata are easily separated from
those of other American species because only one
metatibial spur is greatly reduced and partially hid-
den by a broadly tooth-like process of the tibial
apex (Fig. 30). The male genitalia (Figs. 22, 26) are
also distinctive. Females are less easily character-
ized, though the subcontiguously punctate clypeal
disc will distinguish this species from M. nuda and
M. ciliata. The clypeal punctation is similar in M.
steironematis, but in that species the metasomal
terga are sharply and densely punctate, rather than
smooth and shiny.

Macropis patellata ranges from Vermont to
North Carolina, west to Nebraska and Iowa. The
only floral records we have seen are from Lysi-
machia (= Steironema on label) ciliata in Mary-
land.

Macropis steironematis
Robertson
Figures 23, 27

Macropis steironematis Robertson, 1891:63; 9 <3.

Michener, 1938:135; 6 9. Mitchell, 1960:529-

530; 9 a.

Snelling and Stage: North American Melittinae ■ 29

This is one of the less frequently collected species,
but certainly the most easily recognized. Both sexes
may be immediately recognized by the sharply and
subcontiguously punctate metasoma; in all other
North American species, the metasoma is polished,
with only very scattered, obscure, minute punc-
tures. Males are further distinguished by the char-
acteristic profile of the gonostylus and the presence
of a number of flattened, blade-like setae on the
outer face of the penis valve (Figs. 23, 27).

Mitchell (1960) gave the range as “Iowa and Mis-
souri, east to Virginia, North Carolina and Geor-
gia.” We have seen specimens from southern Illi-
nois, Nebraska, and Kansas. Recorded floral visits
include Apocynum cannabinum, Ceanothus amer-
icanus, Lysimachia (= Steironema) sp., Melilotus
alba, and Seriocarpus lineifolius (Mitchell, 1960).

Macropis steironematis opaca
Michener

Macropis steironematis subsp. opaca Michener,
1938:134; <5.

This form was described from a single male col-
lected at “Morgan’s Ferry, Yakima River, Wash-
ington, July 1, 1882.” The type is in the Museum
of Comparative Zoology and is the only known
specimen. The status of this form cannot be de-
termined in the absence of additional material, and
we have elected to accord it the status originally
proposed by Michener; the differences between this
form and the nominate form are as those cited by
Michener in the original description and noted ear-
lier in the key.

ACKNOWLEDGMENTS

We are indebted to the following individuals for the loan
of material utilized in this study: P.H. Amaud, Jr. (CAS);
G.C. Eickwort (CORN); H.E. Evans (MCZ); S.I. Frommer
and the late P.H. Timberlake (UCR); the late P.D. Hurd,
Jr. (UCB); L. Masner (CNC); C.D. Michener and R.W.
Brooks (SEMC); T.L. Griswold and F.D. Parker (BBSL);
B.C. Ratcliffe (UNEB); J.G. Rozen, Jr., and M. Favreau
(AMNH); V. Scott and F.W. Stehr (MSU); and C. Vardy
(BMNH).

We thank Brian V. Brown for reviewing and com-
menting on an early draft of the manuscript. Review of
the final version was ably accomplished by Rob Brooks
and Jack Neff.

The illustrations for Melitta (except M. eickworti) are
the work of Ruth A. DeNicola, to whom we express our
appreciation and thanks.

LITERATURE CITED

Buchmann, S.L. 1978. Vibratile (“buzz”) pollination in
angiosperms with poricidally dehiscent anthers. Ph.
D. thesis, University of California, Davis. 228 pp.
Cane, J.H., G.C. Eickwort, F.R. Wesley, and J. Spielholz.
1983. Foraging, grooming and mate-seeking behav-
iors of Macropis nuda (Hymenoptera, Melittidae)
and use of Lysimachia ciliata (Primulaceae) oils in
larval provisions and cell linings. American Midland
Naturalist 110:257-264.

30 ■ Contributions in Science, Number 451

. 1985. Pollination ecology of V accinium sta-

mineum (Ericaceae: Vaccinioidea). American Jour-
nal of Botany 72:135-142.

Cockerell, T.D.A. 1906. North American bees of the
genera Andrena and Melitta in the British Museum,
part 1. Psyche 13:5-10.

— . 1911. Descriptions and records of bees. 39.

Annals and Magazine of Natural History (ser. 8)
8:660-673.

. 1937. Bees collected in Arizona and California

in the spring of 1937. American Museum Novitates
948:1-15.

. 1941. Observations on plants and insects in Baja

California, Mexico, with descriptions of new bees.
Transactions of the San Diego Society of Natural
History 9:337-352.

Fabricius, J.C. 1775. Sy sterna entomologie. Flensburg
and Leipzig, xxx + 832 pp.

. 1804. Sy sterna Piezatorum. Braunsweig, 439 +

30 pp.

Hurd, P.D., Jr. 1979. Family Melittidae. In Catalog of
Hymenoptera in America north of Mexico, K.V.
Krombein et al. Smithsonian Institution, Washing-
ton, D.C., vol. 2, 1978-1981.

International Commission on Zoological Nomenclature.
1986. Opinion 1383. Apis pilipes Fabricius, 1775
(Insecta, Hymenoptera): Designated as type species
of Megilla Fabricius, 1805. Bulletin of Zoological
Nomenclature 43:121-122.

Kirby, W. 1802. Monographia apum angliae, vol. 1.
Ipswich, xxii + 258 pp.

Klug, J.C.F. 1809. In Faunae Insectorum Germaniae
Initia, vol. IX, G.W.F. Panzer. Niirnberg.

Leach, W.E. 1815. Entomology, In Brewster’s Edin-
burgh Encyclopedia, vol. 9, 57-172. Edinburgh.

Lepeletier de Saint-Fargeau, A. 1841. Histoire naturelle
des insectes. Hymenopteres, vol. 2. Paris, 680 pp.

Linne, K. von. 1758. Systema naturae. Editio decima,
vol. 1. Stockholm, 824 + iii pp.

Michener, C.D. 1938. A review of the American bees
of the genus Macropis (Hymen., Apoidea). Psyche
45:133-135.

. 1947. Bees of a limited area in southern Mis-
sissippi. American Midland Naturalist 38:443-455.

. 1951. Family Melittidae, In Hymenoptera of

America north of Mexico — Synoptic catalog, C.F.W.
Muesebeck et al. United States Department of Ag-
riculture Monograph, no. 2, 1134-1136.

. 1981. Classification of the bee family Melitti-
dae, with a review of species of Meganomiinae. Con-
tributions of the American Entomological Institute
13(3):1-135.

Michener, C.D., and L. Greenberg. 1980. Ctenoplectri-
dae and the origin of long-tongued bees. Zoological
Journal of the Linnean Society 69:183-203.

Michener, C.D., R.J. McGinley, and B.N. Danforth. 1994.
The bee genera of North and Central America (Hy-
menoptera: Apoidea ). Washington, D.C.: Smithson-
ian Institution Press, 209 pp.

Mitchell, T.B. 1960. Bees of the eastern United States.
North Carolina Agricultural Experiment Station
Technical Bulletin 141, vol. 1, 538 pp.

Patton, W.H. 1880. Description of the species of Ma-
cropis. Entomologists Monthly Magazine 17:31-35.

Provancher, L. 1882. Faune Canadienne. Naturaliste
Canadienne 13:161-175.

. 1888. Additions et corrections. Faune ento-

mologique du Canada, hymenopteres, vol. 2. Que-
bec, 475 pp.

Snelling and Stage: North American Melittinae

Radoszkowski, O. 1891. Revision des armures copu-
latrices des males des genres Cilissa et Pseudocilissa.
Horae Societe Entomologique Rossicae 25:236-243.

Richards, O.W. 1935. Notes on the nomenclature of
the aculeate Hymenoptera, with special reference to
the British genera and species. T ransactions of the
Royal Entomological Society of London 83:143-
176.

Robertson, C. 1891. Descriptions of new species of North
bees. T ransactions of the American Entomological
Society 18:49-65.

. 1897. North American bees — Descriptions and

synonyms. T ransactions of the St. Louis Academy
of Sciences 7:315-3 56.

Rozen, J.G., and N.R. Jacobson. 1980. Biology and
immature stages of Macropis nuda, including com-
parisons to related bees (Apoidea, Melittidae). Amer-
ican Museum Novitates 2702:1-11.
mature stages of Macropis nuda, including com-

parisons to related bees (Apoidea, Melittidae). Amer-
ican Museum Novitates 2702:1-11.

Sandhouse, G.A. 1943. The type species of the genera
and subgenera of bees. Proceedings of the United
States National Museum 92:519-619.

Smith, F. 1853. Catalogue of hymenopterous insects in
the collection of the British Museum, part 1: An-
drenidae and Apidae. London: The British Muse-
um, 197 pp.

Swenk, M.H. 1907. The bees of Nebraska, part 3. En-
tomological News 18:293-300.

Viereck, H.L. 1909. Descriptions of new Hymenoptera.
Proceedings of the Entomological Society of Wash-
ington 11:42-51.

Westwood, J.O. 1840. An introduction to the modern
classification of insects, vol. 2: Synopsis of the gen-
era of British Insects. London, 587 pp.

Received 24 May 1994; accepted 12 December 1994.

Contributions in Science, Number 451

Snelling and Stage: North American Melittinae ■ 31

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

rax

Number 452
9 June 1995

Contributions
in Science

A Review of the Cranes (Aves: Gruidae) of
Rancho La Brea, With the
Description of a New Species

Kenneth E. Campbell, Jr.

UN 3 0 m>

^^L^aries

Natural History Museum of Los Angeles County

Serial
Publications
of THE
Natural History
Museum of
Los Angeles
County

Scientific

Publications

Committee

James L. Powell, Museum President
Daniel M. Cohen,
Committee Chairman
Brian V. Brown
Kenneth E. Campbell
Kirk Fitzhugh
Robin A. Simpson, Managing Editor

• Contributions in Science, a miscellaneous series of tech-
nical papers describing original research in the life and
earth sciences.

• Science Series, long articles and collections of papers
on natural history topics.

Copies of the publications in these series are sold through
the Museum Book Shop. A catalog is available on request.

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

Printed at Allen Press, Inc., Lawrence, Kansas
ISSN 0459-8113

A Review of the Cranes (Aves: Gruidae) of
Rancho La Brea, with the
Description of a New Species

Kenneth E. Campbell, Jr.1

ABSTRACT. A review of all known specimens of cranes (Aves: Gruidae) from the asphalt deposits at
Rancho La Brea, California, revealed that three species of Grus were present in the Los Angeles Basin in
the late Pleistocene. These included the living species G. americana and G. canadensis and a new, smaller
species described herein. The new species is distinguished most readily by cranial proportions, which
indicate a longer, more slender skull with a relatively enlarged basicranial region. Numerous postcranial
elements are also referable to this species. The 502 crane specimens from Rancho La Brea represent a
minimum of 77 individuals.

INTRODUCTION

The fossil bird collection from the upper Pleisto-
cene asphalt deposits at Rancho La Brea, California,
housed in the George C. Page Museum of La Brea
Discoveries, is the largest collection of fossil birds
in existence. The estimated size of the collection
is approximately 350,000-400,000 specimens, and
it includes both cranial and postcranial material.
To date, 139 species of birds, representing 32 fam-
ilies of 14 orders, have been identified from the
Rancho La Brea (RLB) collections. Of these, 24
species are extinct. The most thorough study to
date of the Rancho La Brea paleoavifauna was that
of Howard (1962a), wherein she analyzed the avian
assemblages from individual pits and presented the
most complete taxonomic list available for the col-
lection.

The entire Rancho La Brea vertebrate collection
has been reorganized since the opening of the George
C. Page Museum in 1977, resulting in large numbers
of fossil birds coming to light that had not previ-
ously been integrated into the avian collection. In
addition, major new collections of RLB fossil birds
from the ongoing excavation at Pit 91 and various
salvage operations at adjacent construction sites have
added to the original avian collection studied by
Dr. Howard. Several new species have now been
recognized in these collections, including the new
crane described herein, a giant eagle, a lapwing, and
a giant passerine of as yet unknown affinities. The
last three species are being described elsewhere.
Numerous additional new species are expected to
be identified as work on the entire collection pro-
ceeds.

1. Vertebrate Paleontology /Rancho La Brea Section,
Natural History Museum of Los Angeles County, 900
Exposition Boulevard, Los Angeles, California 90007.

Contributions in Science, Number 452, pp. 1-13
Natural History Museum of Los Angeles County, 1995

The cranes of Rancho La Brea comprise a small,
but nonetheless significant, portion of the overall
paleoavifauna. The first announcement of fossil
cranes from the site was the report of three spec-
imens by L. Miller (1910). Two of these specimens
were referred to Grus canadensis, but a third spec-
imen was described as a new species, G. minor L.
Miller 1910. The species name minor was to in-
dicate the supposed much smaller size of the new
paleospecies compared to that of G. canadensis.
In a later paper, however, L. Miller (1925) an-
nounced that G. minor was an invalid species. He
explained that his earlier description of the species
was based on a comparison with a single modern
specimen that later proved to be misidentified to
species. In addition, rather than being smaller than
G. canadensis, G. minor proved to be larger than
specimens of the former species available to L. Mil-
ler (1925:77) for comparison.

Subsequently, there have been no additional pa-
pers in which the cranes of RLB played a significant
role. As with so many avian species from RLB with
modern representatives, the occurrence of cranes
was noted only in general summaries of the com-
plete avifauna, and then only briefly (e.g., Howard,
1930, 1962a, 1962b). The most notable informa-
tion recorded over the years was that the least num-
ber of cranes from RLB increased from 29 individ-
uals of Grus canadensis and 1 of G. americana
cited in Howard’s (1930) first census to 41 individ-
uals of G. canadensis and 2 of G. americana in her
last (Howard, 1962a). However, part of this in-
crease was undoubtedly attributable to a new meth-
od of determining least numbers of individuals
(Howard, 1962a:7), wherein the least number of
individuals was determined for each pit and the
results totalled. This was in contrast to the original
method that treated all of the pits as one collection.
The current study raises the least number of indi-
viduals reported from RLB to 52 for G. canadensis.

8 for G. americana, and 17 for the new species
described herein. The new total of 77 individuals
for all species of Grus is 80 percent greater than
that last reported by Howard (1962a).

A number of the crane specimens in the RLB
collection were noted by Howard (notes with spec-
imens) to be particularly small cranes; presumably
these were some of those referred to as representing
the Little Brown race of Grus canadensis (Howard,
1962b). When I first began identifying the speci-
mens of cranes newly added to the collections, I
noted that there were subtle, but constant, com-
binations of osteological characters that distin-
guished the smallest of the crane specimens from
counterpart elements of G. canadensis, including
the smallest of modern comparative specimens. The
discovery of a crane cranium with proportions no-
tably different from those of G. canadensis con-
firmed that a third species of crane was present at
Rancho La Brea.

MATERIALS AND METHODS

Prior to this study, there were 440 catalogued specimens
assigned to two species of Grus, or just to the genus, from
Rancho La Brea. Of these, all but seven were found in
the collection. The missing specimens include one prox-
imal and one distal ulna, one carpal phalanx, one distal
tibiotarsus, one distal tarsometatarsus, and two pedal pha-
langes. In addition, 15 of the specimens referred to Grus
were found to be incorrectly identified to genus, leaving
a total of 418 specimens previously identified as crane
present and accounted for in the collection. Each collec-
tion of unidentified bird fossils from Rancho La Brea was
searched for specimens of cranes, including those made
both early in this century and in the past few years. This
resulted in the identification of 86 new specimens of Grus
spp., or an increase of 21 percent over those listed in the
pre-existing catalogue.

Fourteen complete and two partial skeletons of Recent
Grus canadensis were used for comparative purposes. An
effort was made to use small individuals of G. canadensis
for comparison because the new species overlapped the
smaller size range of the former species. Measurements
of these and the fossil material were made using dial
calipers accurate to 0.1 mm. Anatomical terminology is
from Howard (1980) and Baumel (1993). Least numbers
of individuals represent the sum of the least numbers for
each pit (Howard, 1962a:6).

SYSTEMATICS

Order Ralliformes
(Reichenbach 1852)

Family Gruidae Vigors 1825
Genus Grus Pallas 1766
Grus americana Linnaeus 1758

MATERIAL.

Clavicule: Right partial, K7349.

Coracoids: Left complete, FI 121; left fragment, B9517,

K3379; right complete, F327.

Scapulae: Left partial, H5651; right partial, H5652.

Carpometacarpi: Left complete, F350; right complete,
F810.

Femur: Left distal, B8597.

Tibiotarsi: Left proximal, B5750; left shaft and distal,
PMS 518; right proximal, R50795; right shaft and dis-
tal, PMS 519; right distal, F569.

Fibula: Left complete, PMS 522.

Synsacrum: Partial, B6073.

Tarsometatarsi: Left complete, PMS 523; left proximal,
F652; left distal, F654; right complete, PMS 524; right
proximal, G4893; right distal, F651.

Metatarsal I: Right fragment, PMS 527.

Pedal digit 2 phalanx 1: Left, J1959; right, J1960, PMS
531.

Pedal digit 2 phalanx 2: Left, PMS 533; right, J5305.
Pedal digit 2 phalanx 3: Left, PMS 538.

Pedal digit 3 phalanx 1: Left, J1 932, PMS 530; right, F260,
K2642.

Pedal digit 3 phalanx 2: Left, J5415, PMS 534.

Pedal digit 3 phalanx 3: Left, J5315, PMS 526.

Pedal digit 3 phalanx 4: Left, PMS 537.

Pedal digit 4 phalanx 1: Left, J3589, PMS 532; right,
K7001.

Pedal digit 4 phalanx 2: Left, PMS 535.

Ungual phalanx: Left, PMS 537.

A total of 45 specimens from seven pits represents
a minimum of eight individuals. However, 16 of
the specimens were from the articulated distal hind-
limbs of a single individual recoverd from the Page
Museum Salvage site.

Grus canadensis Linnaeus 1758

MATERIAL.

Crania: Partial, F387, F433, F471, F477, R38926.
Mandibles: Symphysis with partial dentary, F106, F920,
F2950; dentary fragment, F102, F922; articular, K2359.
Quadrate: Right, F3377.

Claviculae: Left partial, J6532, K7188.

Sterna: Partial, F463, F475, F526, F536, F567.

Synsacra and pelves: Partial, C7263, E698, F697.
Scapulae: Left, F481, H5655, H6517, H6530, K5182,
K5183; right, D5338, E212, F354, F784, F2051, H5654,
K7351.

Coracoids: Left complete, F329, F345, F445, F528, F570,
F607, F663, H3691; left proximal, B8891, H3693,
K2733, K5187, K5190; left distal, B8996, F1067; left
shaft, F145; right complete, B9291; D6510, F434, F483,
F512, F573, F580, F1064, H3690, H3694, H4409,
K5186, R40363; right proximal, F793, K5189.
Humeri: Left proximal, G2144, G2145, G2146, J9185,
K5151, K5152; left distal, C4430, K5153; right proxi-
mal, C9991, R50827; right distal, C2502, D4593, D5166,
D5471.

Ulnae: Left proximal, C2479, C2907, C7967, K5180; left
distal, C4826, D5707, G6813, G6814, K5178; right
complete, C967, C2986, F628; right proximal, C1903;
right distal, C4970, D4845, F506, K2795, K2797.
Radii: Left proximal, K5185; left distal, H8338; right dis-
tal, H8360, H8384.

Carpometacarpi: Left complete, F182, F187, F455, F566,
F977, F1065, FI 103, H1675, H1678, K2809; right
complete, E9559, F204, F205, F355, F458, F979, F1069,
FI 129; right proximal, K5176; right distal, C5557.
Carpal digit 2 phalanx 1: Left, C6009, D1415, D2711,
D5563, F2024, J783, J786, K2828, K7337; right, C8053,
C9341, D2709, J883.

2 ■ Contributions in Science, Number 452

Campbell: Cranes of Rancho La Brea

Carpal digit 2 phalanx 2: Left, J 1415; right, D2730.
Femora: Left complete, B9442, F348, F538, K5156; left
proximal, E7759, K5160; left shaft, K5157; left distal,
F4629, K5161; right complete, E6332, E6946, F549,
F738, K5158; right distal, F4598, K5159.

Tibiotarsi: Left complete, K7330; left proximal, D5131,
F67, F685, F961, F984, K5170; left distal, F149, F263,
F442, F457, F460, F487, F507, F521, F534, F544, F568,
F584, F598, F740, F787, F967, F980, F1009, F1070,
FI 134, F6735, F6736, F6738, F6765, F6772, K5173,
K7334, R12073, R13199, R35165, R40182; left shaft,
F597; right complete, K7329; right proximal, D5010,
FI 144, K7335; right distal, E8236, F32, F148, F332,
F476, F480, F510, F517, F556, F559, F857, F919, FI 148,
F6737, F6762, F6763, F6764, F6766, K950, K1642,
K2568, K2960, K5171, K5172, K7333, K7347, K7348,
R42148.

Fibulae: Left proximal, H8521, K7331; right proximal,
H8520, H8522.

Tarsometatarsi: Left complete, F577, F780, G4871, G4878,
K3137, K7307; left proximal, E5980, E8444, F330, F421,
F459, F462, F513, F551, F939, F963, F1061, F1127,
G4889, G6224, K5162, K5164, K5166, K7346; left dis-
tal, E7922, E8282, F30, F461, F493, F509, F515, F537,
F545, F564, F710, F834, F865, F933, FI 131, G4880,
G4895, G5970, G6116, G6171, K5163; right complete,
G4875, G4881, K3138, K3139, K3140, K7306, R29018;
right proximal, F514, F519, F525, F587, F684, F781,
F791, F903, G5952, K746, K2013, K2014, K2524,
K5168; right distal, E5574, E6791, E7212, E7676,
E8269, E8488, E8574, F431, F522, F547, F571, F969,
F1910, G4872, G4879, G4892, G4894, G6025, G6181,
K2000, K5167, K5169, R10312, R15026, R18315; shaft
fragments, E9962, F438, F516, F543, F1030, FI 143.
Pedal digit 1 phalanx 1: Left, K1558.

Pedal digit 2 phalanx 1: Left, B8836, D916, D7785, D8204,
D8272, F927, J1952, J1953, J1954, J1955, J1956,
R39012; right, B5787, D2805, D5273, E7407, F162,
F163, F924, J 1 957, J 1 958, K495, K1449, K1461, K7343,
R52247.

Pedal digit 2 phalanx 2: Left, D6363, F996, J5263, J5736;

right, B8843, J5264, J5265, J5437, K1778.

Pedal digit 3 phalanx 1: Left, B5988, E8034, F505, J3278,
J371 3, J3714, J3715, J3727, J3728, J3732, J3734, K2644,
K7338, K7339; right, B9840, D4369, D7527, F360, F447,
F911, FI 105, J1933, J3711, J3712, J3729, J3730, K2643,
R12194.

Pedal digit 3 phalanx 2: Left, C8235, J5222, J5227, J5382,
K1462; right, C5620, J5225, J5385, J5386, J9957, K1269,
K1463.

Pedal digit 3 phalanx 3: Left, D9062, J5490, K1288; right,
K2137.

Pedal digit 4 phalanx 1 : Left, D3213, D3305, J3588, J4560,
K1488, K1489; right, J4417,J4559, J4561, J4562, K7340.

A total of 417 specimens from 19 pits represents a
minimum of 52 individuals.

Grus pagei, new species

Figure 1

HOLOTYPE. Partial cranium, F735.
DIAGNOSIS. The holotype cranium represents
a species of Grus that differs from that of G. amer-
icana by being of much smaller size and from that
of G. canadensis by having (1) width across ossa
frontales at orbits narrower, whereas widths across
the ala parasphenoidalis (= ala tympanica) and the

lamina parasphenoidalis (= basitemporal plate) are
greater; (2) distance from prominentia cerebellaris
to end of ossa frontales at nasofrontal hinge greater,
with ossa frontales sloping more gently toward hinge
area; and (3) condylus occipitalis larger, although
overall cranium smaller.

TYPE LOCALITY. Pit A, Rancho La Brea, Han-
cock Park, Los Angeles, California, USA.

TYPE HORIZON AND AGE. Asphalt-intruded
upper Pleistocene (Rancholabrean Land Mammal
Age) alluvial deposits of the Los Angeles Basin,
previously designated Submember C of Member C
of the Palos Verdes Sand (Woodard and Marcus,
1973).

MEASUREMENTS. See Tables 1 and 2.

REFERRED MATERIAL.

Scapulae: Right partial, E4928, H5653, K5184.
Coracoids: Left partial, F966, K5188; right partial, F486.
Humeri: Left proximal, G2141; left distal, K5155; right
distal, D7861.

Ulna: Right distal, K5179.

Carpometacarpi: Left complete, FI 83, HI 679; right com-
plete, F76, HI 680; right distal, K7344.

Carpal digit 2 phalanx 1: Right, D7171.

Femur: Right distal, F4636.

Tibiotarsi: Left distal, F563, K5174; right complete, K3133;
right proximal, F1017; right distal, F494, F557, K5177,
K7336, K7345.

Tarsometatarsi: Left proximal, E8298, E8352, G4877,
K4100, K5165; left distal, E8073; right complete, F452,
G4874; right distal, E8486, G4873, K4097.

Pedal digit 2 phalanx 1: Right, K7342.

Pedal digit 3 phalanx 1: Left, J3703; right, E8101.

Pedal digit 4 phalanx 1: Left, K7341.

A total of 42 specimens, including the holotype,
from 11 pits represents a minimum of 17 individ-
uals.

ETYMOLOGY. Named for George C. Page, in
recognition of his gift of the George C. Page Mu-
seum of La Brea Discoveries to the people of Los
Angeles and the world. His generosity brought the
study of the complete avifauna of Rancho La Brea
into the realm of possibilities.

DESCRIPTION. All elements of Grus pagei, new
species, average smaller in size than the respective
elements of G. canadensis, whereas the only other
North American species, G. americana, and all oth-
er Recent species of Grus average much larger than
G. canadensis (Johnsgard, 1983). Therefore, G. pa-
gei will be compared only to G. canadensis. For
measurements of all elements, see Table 1.

Skull. The outstanding feature of the cranium of
Grus pagei is how it differs in its proportions from
those of G. canadensis. This can be seen visually
(Figs. 1-3) and through ratios (Table 2). Although
the maximum width of the cranium of G. pagei
overlaps the lower size range of G. canadensis, the
ossa frontales are much narrower, whereas the width
across the ala parasphenoidalis and lamina para-
sphenoidalis is greater. The cumulative effect of
these differences is a proportionately larger basi-
cranial region and a more slender, elongated frontal

Contributions in Science, Number 452

Campbell: Cranes of Rancho La Brea ■ 3

Table 1. Measurements of bones of species of Grus from Rancho La Brea and Recent specimens. OR = observed
range; M - mean; N = number of specimens.

Element

Grus americana

Grus canadensis

Grus pagei

Rancho
La Brea

Recent

Rancho
La Brea

Recent

Rancho
La Brea

Cranium

Width across ossa

17.3-21.4

14.9-18.3

frontales at orbits

18.6

17.3

14.5

Distance from

60.5-72.0

prominentia cerebellaris

72.9

66.3

71.8

to nasofrontal hinge

Width across alae

30.7-33.8

26.8-31.7

parasphenoidalis

31.8

28.9

31.9

Width across lamina

13.7-15.6

12.8-14.6

parasphenoidalis

14.3

13.5

15.5

Maximum width of cranium

38.1-42.3

35.3-41.0

40.4

38.2

37.9

Width across ossa

11.7-14.5

9.5-12.0

frontales between ossa

13.3

11.3

13.1

lacrimales

Distance from center of

37.2-44.5

39.5-47.9

nasofrontal hinge to

42.0

43.2

48.7

posterior edge of orbit

Coracoid

Head to angulus

58.2-74.8

51.8-69.0

54.4-58.4

medialis

74.1

72.9

67.0

60.9

56.4

Head to external end of

79.2-86.9

64.1-86.6

facies articularis stemalis

91.5

94.9

82.7

74.8

68.7

Least width of shaft

15.3-15.6

12.5-15.7

10.7-13.7

11.9-12.8

15.4

15.3

14.1

11.7

12.4

Head through facies

27.4-38.2

25.4-34.4

30.1-30.7

articularis scapularis

37.9

39.5

32.7

29.7

30.4

Scapula

Length of facies

16.4-1 6.6

12.0-16.2

11.0-15.6

10.1-11.1

articularis humeralis

16.5

13.7

13.0

10.6

Acromion length

17.6-18.2

11.7-15.6

10.9-14.7

17.9

13.8

13.2

12.6

Maximum depth through

25.7-26.9

20.0-24.8

18.1-24.0

acromion and facies

26.3

22.9

21.0

19.5

articularis humeralis

Humerus

Total length

177.9-225.2

200.2

4 ■ Contributions in Science, Number 452

Campbell: Cranes of Rancho La Brea

Table 1. Continued.

Element

Grus americana

Grus canadensis

Grus pagei

Rancho
La Brea

Recent

Rancho
La Brea

Recent

Rancho
La Brea

Proximal width

42.0-45.7

33.1-43.1

44.0

38.0

40.6

Depth of head

13.5-14.5

10.6-14.0

13.8

12.2

13.0

Distal width

33.1-34.7

26.8-32.8

34.0

29.7

27.7

Condylus dorsalis depth

18.0-19.4

14.5-18.9

15.3-15.6

18.5

16.5

15.5

Ulna

Total length

241.5-243.1

200.8-247.4

242.3

225.4

Proximal width

22.7-24.9

19.5-22.6

23.5

21.1

Proximal depth

16.5-18.6

14.4-17.9

17.5

16.0

Distal depth

14.5-16.3

12.9-16.5

15.6

14.7

14.2

Distal width

15.6-18.8

14.5-17.6

17.2

16.1

15.4

Carpometacarpus

Total length

137.0-137.2

104.6-125.2

89.9-111.1

94.8-97.1

137.1

117.9

115.4

99.8

96.4

Proximal width

26.8-27.6

20.2-25.3

18.8-21.9

17.6-19.9

27.2

23.7

23.2

20.6

18.8

Os metacarpalis alulare

16.2-17.1

12.0-15.9

11.7-14.4

11.9-12.9

length

16.7

15.0

14.6

13.2

12.4

Distal depth

17.3-18.9

13.9-17.6

13.1-15.0

11.2-13.9

18.1

15.5

16.2

14.3

12.9

Os metacarpale majus

9.0-11.5

8.3-9.9

7.9-8.4

width

11.5

9.8

10.2

9.2

8.2

Carpal digit 2 phalanx 1

Total length

42.7-49.0

39.7-45.1

46.8

42.3

35.0

Maximum width

12.1-14.3

10.8-12.7

13.0

11.8

11.5

Contributions in Science, Number 452

Campbell: Cranes of Rancho La Brea ■ 5

Table 1. Continued.

Element

Grus americana

Grus canadensis

Grus pagei

Rancho
La Brea

Recent

Rancho
La Brea

Recent

Rancho
La Brea

Femur

Total length

100.5-119.2

100.8-129.0

111.7

114.4

Proximal width

22.1-26.1

21.1-25.9

24.7

23.5

Proximal depth

19.5-21.9

17.1-22.0

20.7

19.4

Midshaft width

9.8-12.2

8.9-11.1

13.5

11.1

10.4

8.8

Midshaft depth

9.5-11.9

8.8-11.8

13.7

10.7

9.9

8.3

Distal width

22.4-27.7

21.4-26.3

31.6

25.7

23.5

20.0

Distal depth

21.5-25.4

20.1-25.9

30.0

23.4

22.8

18.6

Tibiotarsus

Total length

251.5-253.1

199.3-293.0

252.3

247.1

226.5

Proximal width

18.6-22.7

13.7-20.9

16.3-16.5

24.1

22.3

20.5

17.9

16.4

Proximal depth

24.0-30.9

21.2-28.1

21.2-22.5

33.5

28.5

27.1

24.7

21.9

Distal width

26.4-27.2

17.9-23.9

16.7-20.6

15.1-17.9

26.9

22.7

20.7

18.8

16.8

Condylus lateralis depth

23.9-24.7

15.7-22.4

15.3-19.5

13.7-15.8

24.4

20.4

19.0

17.2

15.2

Condylus medialis depth

25.0-25.8

16.0-22.9

15.3-19.9

14.5-16.5

25.5

20.6

19.5

18.1

15.9

Tarsometatarsus

Total length

217.2-255.0

175.1-254.1

188.3-194.8

235.7

282.0

229.7

219.6

191.6

Proximal width

28.4-30.1

20.9-25.9

19.0-23.5

18.5-20.2

29.1

25.6

23.3

21.3

19.3

Hypotarsus length

23.1-29.5

14.9-22.1

14.5-19.7

13.8-17.0

26.3

18.6

17.2

15.6

6 ■ Contributions in Science, Number 452

Campbell: Cranes of Rancho La Brea

Table 1. Continued.

Element

Grus americana

Grus canadensis

Grus pagei

Rancho
La Brea

Recent

Rancho
La Brea

Recent

Rancho
La Brea

Distal width

26.0-26.5

18.9-25.1

17.6-21.8

17.6-19.3

26.3

24.9

22.1

20.0

18.5

Trochlea metatarsi tertii

13.4-13.7

9.5-13.0

8.6-13.8

8.5-9. 7

depth

13.6

12.8

11.0

10.5

9.2

Trochlea metatarsi tertii

11.0-11.5

7.7-10.1

6. 9-8. 8

6.9-8. 1

width

11.3

9.9

8.9

7.9

7.5

Pedal digit 1 phalanx 1

Length

10.3-11.2

19.4

10.8

Proximal width

3.3-3. 8

5.0

3.6

Distal width

2.7-3.0

3.9

2.9

Pedal digit 2 phalanx 1

Total length

40.4-43.1

25.8-32.0

21.8-30.5

41.3

28.9

27.3

21.8

Proximal width

9.3-10.6

7.4-9.4

6.2-8.5

10.0

8.4

7.6

6.2

Distal width

6.9-7.6

5. 1-6.7

4.8-6. 6

7.2

6.0

5.5

4.4

Pedal digit 2 phalanx 2

Total length

36.1-37.7

25.4-30.2

19.9-22.8

36.9

27.4

21.4

Proximal width

7 .6-7.7

6. 1-7.5

4.8-5.4

7.7

6.5

5.1

Distal width

6.5-6. 8

5. 1-6.3

4.1-4.6

6.7

5.5

4.4

Pedal digit 3 phalanx 1

Total length

44.3-48.9

31.2-36.9

26.7-35.8

25.4-27.5

46.6

34.0

32.2

26.5

Proximal width

10.5-12.5

8.9-11.8

7.8-9. 7

7.2-7.8

11.5

9.9

8.8

7.5

Distal width

7.8-9. 6

6.3-8. 1

5.3-6. 6

5.2-5. 9

8.5

7.1

6.2

5.6

Contributions in Science, Number 452

Campbell: Cranes of Rancho La Brea ■ 7

Table 1. Continued.

Grus americana Grus canadensis Grus pagei

Element

Rancho
La Brea

Recent

Rancho
La Brea

Recent

Rancho
La Brea

Pedal digit 4 phalanx 1

Total length

31.1-36.0

24.2-27.2

19.1-27.9

34.0

25.8

24.4

20.0

Proximal width

10.0-11.7

8.2-9. 9

7.2-92

11.0

8.9

8.3

7.4

Distal width

6.5-7.8

5.4-6.4

4.6-5. 7

7.2

5.8

5.3

5.1

region. No maxillaries or premaxillaries of G. pagei
have been found; therefore, it is not possible to
determine whether or not the elongation of the bill
continued to its tip.

Coracoid. The coracoid of Grus pagei differs from
that of G. canadensis by having (1) head broader,
relatively more massive, and less pointed craniad
adjacent to tip of impressio lig. acrocoracoideum,
as seen in dorsal and internal view; (2) cotyla scapu-
laris fairly well-defined pit (large, shallow to deep
depression, but not a pit, in G. canadensis ); (3)
facies articularis scapularis prominently developed
posterodorsal to cotyla scapularis, extending only

short distance externad, but extending internad to
processus procoracoideus (not as prominent in G.
canadensis ); (4) processus procoracoideus shorter,
curving ventrad less rapidly, providing for more
open canalis triossealus; (5) distal end of facies ar-
ticularis clavicularis a more prominent projection
that impinges more upon canalis triossealus; (6) fa-
cies articularis humeralis proportionately wider; and
(7) facies articularis sternalis narrower, especially
near angulus medialis.

Scapula. The scapula of Grus pagei differs from
that of G. canadensis by having (1) facies articularis
clavicularis only slightly elevated above margo dor-

Table 2. Ratios of cranial measurements of species of Grus. Top row of each pair consists of actual measurements
(mm), whereas the bottom row presents the ratios of five cranial measurements against the width across the ossa
frontales at the orbits.

Width

across

ossa

frontales
at orbits

Distance from
prominentia
cerebellaris
to naso-
frontal
hinge

Width
across
alae para-
sphenoidalis

Width
across lam-
ina para-
sphenoidalis

Width across
ossa frontales
between ossa
lacrimales

Distance from
center of
nasofrontal
hinge to
posterior
edge of orbit

Grus pagei, new species

Holotype

14.5

71.8

31.9

15.5

13.1

48.7

4.95

2.20

1.07

0.90

3.36

Grus canadensis: Rancho La Brea

Mean of all specimens

18.6

72.9

31.8

14.3

13.3

42.0

3.92

1.71

0.77

0.72

2.26

Most complete specimen

18.2

72.9

33.8

15.6

14.5

44.5

Grus canadensis: Recent specimens

4.01

1.86

0.86

0.80

2.45

Mean of all specimens

17.3

66.3

28.9

13.5

11.3

43.2

3.83

1.67

0.78

0.65

2.50

Smallest skull

14.9

60.5

26.8

13.8

10.6

39.7

4.06

1.80

0.93

0.71

2.66

Largest skull

18.8

68.1

28.8

13.9

9.5

43.1

3.62

1.53

0.73

0.50

2.29

8 ■ Contributions in Science, Number 452

Campbell: Cranes of Rancho La Brea

salis, with latter sloping gradually to former (con-
siderably elevated above margo dorsalis, which turns
abruptly dorsad to meet it, in G. canadensis)', (2)
facies articularis humeralis with anteroventral pro-
jection small, sharply pointed, with facies articularis
coracoideus on facies costalis short and v-shaped,
almost notched (anteroventral projection large,
rounded, with facies articularis coracoideus on fa-
cies costalis long, broadly convex, and wide in G.

canadensis ); and (3) tubercle on facies costalis dor-
sal to pneumatic foramina small, but prominent,
neither completely rounded nor pointed (promi-
nent and fairly sharply pointed in G. canadensis).

Humerus. The humerus of Grus pagei differs
from that of G. canadensis by having (1) sulcus lig.
transversus noticeably more constrained by proxi-
mal extension of the ventral portion of the intu-
mescentia humeri (= bicipital surface); (2) attach-

Figure 1. Holotype cranium of Grus pagei, new species,
George C. Page Museum, Hancock Collection F735, in
dorsal, lateral, and ventral views. Note the narrowness of
the ossa frontales, the greater width across the alae para-
sphenoidalis, and the greater distance from the promi-
nentia cerebellaris to the nasofrontal hinge, in comparison
to the corresponding dimensions of G. canadensis (Figs.
2, 3). xl.

Figure 2. The most complete cranium of Grus cana-
densis from Rancho La Brea, George C. Page Museum,
Hancock Collection F477, in dorsal, lateral, and ventral
views. This specimen has the same proportions as the
crania of modern G. canadensis, xl.

Contributions in Science, Number 452

Campbell: Cranes of Rancho La Brea ■ 9

Figure 3. This small skull of Grus canadensis is approximately of the same maximum cranial width as the holotype
of G. pagei, new species; shown in dorsal, lateral, and ventral views. Note the smaller distance between the alae
parasphenoidalis in this specimen compared to that in the holotype of G. pagei (Fig. 1).

ment of M. proscapulohumeralis brevis a more
elevated bump; (3) tuberculum ventrale (= internal
tuberosity) relatively more massive; (4) condylus
ventralis less bulbous in anterior view, without as
marked a central extension, being more elongated
dorsoventrally; (5) epicondylus ventralis with a more
pronounced posteroventral corner, but overall it
does not extend as far distad and is more com-
pressed dorsoventrally than that of G. canadensis,
especially with the ventral limit of epicondylus ven-
tralis in posterior view closer to the condylus ven-
tralis; and (6) attachments of M. extensor metacarpi
radialis, pars anconalis and palmaris, positioned
slightly more distad and forming shorter, more pro-
nounced shelf.

Ulna. The ulna of Grus pagei differs from that

of G. canadensis by having (1) sulcus tendineus
larger, with proximal end positioned closer to tu-
berculum carpale; (2) tuberculum carpale more
pointed and more slender, but with larger ridge of
bone leading to it from the shaft; and (3) condylus
ventralis ulnae more compressed cranially.

Carpometacarpus. The carpometacarpus of Grus
pagei differs from that of G. canadensis by having

(1) facies articularis ulnocarpalis narrower distally;

(2) external rim of facies articularis ulnaris more
rounded in dorsal (= external) view, without a
“peak” just proximal to the attachment of M. flexor
carpi ulnaris brevis; (3) os metacarpale alulare small-
er, with processus extensorius less projecting, less
twisted ventrad (= internad), and lacking promi-
nent, undercut ridge along dorsal edge distally, i.e.,

10 ■ Contributions in Science, Number 452

Campbell: Cranes of Rancho La Brea

with slight, centrally located depression only; and
(4) facies articularis digiti minoris smaller, not pro-
jecting distad beyond facies articularis digiti majoris
(does project distad beyond facies articularis digiti
majoris in G. canadensis ).

Femur. The femur of Grus pagei differs from
that of G. canadensis by having (1) condylus la-
teralis shorter anteroposteriorly and more undercut
at its proximal posterior extension, more distinct
from posterior surface of shaft, and oriented at
slightly greater angle to shaft; and (2) shaft more
flattened, with less of a ridge leading to posterior
end of condylus lateralis, and broadening more
gradually to meet condylus lateralis. Only distal end
available.

Tibiotarsus. The tibiotarsus of Grus pagei differs
from that of G. canadensis by having (1) crista
cnemialis lateralis closer to exterior articular sur-
face, i.e., flaring less anteriad; (2) facies articularis
medialis smaller; (3) shaft distal to facies articularis
medialis flattened, not rounded; (4) condylus me-
dialis narrower, not as rounded, with distal flange
less flaring mediad; (5) both condylae proportion-
ately shorter anteroposteriorly; and (6) trochlea car-
tilaginis tibialis sloping internad from condylus la-
teralis at a greater angle in distal view, being deeper
adjacent to condylus medialis.

Tarsometatarsus. The tarsometatarsus of Grus
pagei differs from that of G. canadensis by having
(1) eminentia intercondylaris more rounded in an-
terior view and less bulbous in lateral view; (2)
cotyla medialis deeper and slightly narrower, giving
a more constricted form; (3) area between the coty-
lae and hypotarsus a deeper, more enclosed basin,
with posterior edge of cotylae more elevated (opens
internally in G. canadensis ); (4) hypotarsus more
compressed anteroposteriorly, with internal ridge
over enclosed sulcus hypotarsi more elevated and
more separate, curving externad distally; and (5)
trochlea metatarsi tertii slightly more compressed
anteroposteriorly, with distal end rotated slightly
anteriad.

DISCUSSION

The cranes of Rancho La Brea were not specifically
discussed by Howard (1962a) in her analysis of pit
assemblages, but she did list Grus canadensis as a
typical member of the Rancho La Brea Pleistocene
avifauna. Grus canadensis and G. pagei occur to-
gether in 11 of the 19 pits in which the former
species occurs (Table 3), including both older (e.g.,
Pit 4) and younger (e.g., Pit 10) pits. Clearly, these
two species were sympatric at Rancho La Brea in
the late Pleistocene. Grus americana was more lim-
ited in its occurrence, appearing in only 6 of the
19 pits in which G. canadensis occurred and in one
pit where the latter has not yet been found.

As already noted, Grus minor was named by L.
Miller (1910) based on a comparison with a larger
crane misidentified as G. canadensis (L. Miller,
1925). In his first paper, Miller referred only the

holotype to G. minor, and at that time he referred
two other specimens from Rancho La Brea to G.
canadensis. However, in his second paper he re-
ferred all cranes from RLB (four specimens) to the
former species. The holotype of G. minor, UCMP
12533, could not be examined for this study be-
cause all collections of the Museum of Paleontol-
ogy, University of California, Berkeley, were being
transferred to new facilities. However, from the
illustration of the holotype tibiotarsus (L. Miller,
1910, fig. 8), it can be noted that it differs from that
of the more gracile G. pagei by having larger con-
dyles and a more massive shaft. L. Miller (1925:77)
also stated that “osteological distinctions, noted in
the original description [of G. minor], fall to the
ground on comparing the fossil bird with true G.
canadensis. . . .”

The holotype distal tibiotarsus of G. minor mea-
sured 18.8 mm in distal width and 17.9 mm in depth
through the condylus medialis, measurements equal
to and slightly less, respectively, than the mean of
those for Recent G. canadensis and smaller than
the mean of those of contemporary G. canadensis
from Rancho La Brea (Table 1), but well within
the range of the latter. These measurements of G.
minor are outside the range of those of specimens
assigned to G. pagei in this study (Table 1). Thus,
I have no hesitation in agreeing with L. Miller (1925:
77) that the holotype of G. minor is in fact a spec-
imen of G. canadensis and that the former species
is invalid.

The average Grus pagei was about the same size
as the smallest individuals of Recent G. canadensis
(Table 1). Grus pagei averaged much smaller than
contemporary G. canadensis from RLB, the largest
of the former just barely achieving the size of the
smallest of the latter. The proportions of the cra-
nium of G. pagei suggest that its head was larger
and more slender than that of G. canadensis (Table
2), even though the former was of approximately
the same width as smaller examples of the latter.
This may have provided a means of partitioning the
habitat and reducing competition between these
two cranes. Both of these species, of course, are
much smaller than G. americana, which probably
limited direct competition between the larger and
the two smaller species. Both G. americana and G.
canadensis are migratory species, although some
subspecies of G. canadensis, e.g., G. canadensis
pratensis, appear to be sedentary (Johnsgard, 1983).

Cranes take a variety of foods, both plant and
animal, and occur in a variety of habitats. Although
generally associated with marsh lands, they com-
monly feed in upland terrestrial areas; in both types
of habitat, however, they prefer more open, grassy
environs (Johnsgard, 1983). It can only be assumed
that Grus pagei shared the typical habitat prefer-
ences of species of Grus. Thus, by themselves, cranes
do not provide much detailed information about
what the habitat might have been like at Rancho
La Brea in the late Pleistocene.

Intraspecific size variation in species of Grus is

Contributions in Science, Number 452

Campbell: Cranes of Rancho La Brea ■ 11

Table 3. Distribution of species of Grus among the various pits at Rancho La Brea.

Pit number

Grus americana

Grus canadensis

Grus pagei,

new species

Number of
specimens

Least

number of
individuals

Number of
specimens

Least

number of
individuals

Number of
specimens

Least

number of
individuals

Pit 2

Pit 3

Pit 4

Pit 6

Pit 10

Pit 13

Pit 16

122

Pit 36

Pit 37

Pit 60

Pit 61-67

Pit 77

Pit 81

Pit 91

Pond dump1

Bliss 292

Pit A

Academy3

PMS4

16s

No data

Total

417

1 Mixed talings from Pits 3, 4, and 61-67.

2 Bliss 29 includes four separate pits: Pits A-D. For many specimens, the pit of origin was not noted, and these are
lumped together under the name “Bliss 29.” Specimens are separated to pit when this is known.

3 Excavation by Southern California Academy of Science prior to 1913; materials now part of LACM collections. Pit
reopened by LACM in 1913 and designated Pit 17.

4 PMS (Page Museum Salvage) refers to specimens collected during salvage operations during the construction of the
George C. Page Museum from a site under the new museum building.

5 All specimens were from the articulated distal hindlimbs of a single individual.

considerable (Table 1; Johnsgard, 1983), as is in-
trasubspecific size variation. When this variation is
combined with the seasonal mixing of migratory
races that may occur, it is easy to see that the po-
tential for size variation within a species may be
considerable. Indeed, were only a single postcranial
specimen of G. pagei available it would be very
difficult to convincingly demonstrate that it repre-
sented a species different from G. canadensis.
However, the combination of a fairly large number
of specimens, including a cranium, from Rancho
La Brea does make description of the species pos-
sible.

Both Grus americana and G. canadensis are well
represented from numerous Pleistocene and Ho-
locene sites in North America, but G. pagei is the
only known Pleistocene paleospecies of the genus
(Brodkorb, 1967; Cracraft, 1973; Olson, 1985).
There are two described species of Grus from the
Pliocene of North America, G. conferta A.H. Mil-
ler and Sibley 1942 and G. nannodes Wetmore and
Martin 1930. Olson (1985) stated that the generic

placement of G. conferta was suspect, and the shape
of the facies articularis of trochlea metatarsi II of
that species differs quite noticeably from that of G.
pagei. Grus nannodes is based on a fairly undi-
agnostic distal end and shaft of a carpometacarpus
that is smaller than the smallest carpometacarpus
referred to G. pagei, new species. In addition, it
differs by having a fairly pronounced curvature of
the os metacarpale majus opposite the point of its
fusion with the os metacarpale minor, a curvature
that is lacking in G. pagei.

SUMMARY

A review of all collections of fossil birds from Ran-
cho La Brea housed at the George C. Page Museum
has shown that 502 specimens representing at least
77 individuals are properly assignable to three spe-
cies of Grus: G. americana, G. canadensis, and G.
pagei, new species. This represents an increase of
21 percent in the number of specimens and 80
percent in the number of individuals over what had

12 ■ Contributions in Science, Number 452

Campbell: Cranes of Rancho La Brea

been reported previously. The holotype of the pa-
leospecies G. minor from Rancho La Brea is rec-
ognized as being a specimen of G. canadensis, as
previously determined by L. Miller (1925); thus, the
former species is invalid. Compared to G. cana-
densis, G. pagei is characterized by a longer, more
slender cranium, which nonetheless has a propor-
tionately larger basicranium. The differences in pro-
portions between the skulls of G. canadensis and
G. pagei are suggestive of habitat partitioning
through the use of different food resources. Grus
pagei is widespread both temporally and spatially
among the pits at Rancho La Brea, but it was not
found in collections from other late Pleistocene
asphalt deposits of California. Grus pagei is the only
paleospecies of Grus reported from the Pleistocene
of North America, and it is the first paleospecies
to be described from Rancho La Brea since 1948.

ACKNOWLEDGMENTS

I thank F. Hertel (UCLA) and R.B. Payne and R.W. Storer
(University of Michigan Museum of Zoology) for making
available modem comparative specimens. F. Hertel and
L. Martin kindly provided critical comments on drafts of
this paper. I am grateful to the staff and volunteers of the
George C. Page Museum, who have been invaluable in
gathering together and preparing the multitude of avian
specimens at Rancho La Brea. A special effort on the
photographs by R. Meier, LACM photographer, is much
appreciated.

LITERATURE CITED

Baumel, J.J., ed. 1993. Handbook of avian anatomy:
Nomina anatomica avium, 2nd ed. Publications of
the Nuttall Ornithological Club, no. 23, 779 pp.
Brodkorb, P.B. 1967. Catalogue of fossil birds, part 3:

Ralliformes, Ichthyornithiformes, Charadriiformes.
Bulletin of the Florida State Museum, Biological
Sciences 11:99-220.

Cracraft,]. 1973. Systematics and evolution of the Grui-
formes (class Aves), part 3: Phylogeny of the sub-
order Grues. Bulletin of the American Museum of
Natural History 151:1-127.

Howard, H. 1930. A census of the Pleistocene birds of
Rancho La Brea from the collections of the Los
Angeles Museum. Condor 32:81-88.

. 1962a. A comparison of avian assemblages from

individual pits at Rancho La Brea, California. Con-
tributions in Science 58:1-24.

. 1962b. Fossil birds. With especial reference to

the birds of Rancho La Brea. Los Angeles County
Museum, Science Series 17, Paleontology 10:1-44.

. 1980. Illustrations of avian osteology taken from

“The Avifauna of Emeryville Shellmound.” Contri-
butions in Science 330:xxvii-xxxvii.

Johnsgard, P.A. 1983. Cranes of the world. Blooming-
ton: Indiana University Press, 257 pp.

Miller, A.H., and C.G. Sibley. 1942. A new species of
crane from the Pliocene of California. Condor 44:
126-127.

Miller, L.H. 1910. Wading birds from the Quaternary
asphalt beds of Rancho La Brea. University of Cal-
ifornia Publications in Geology 5(21):439-448.

. 1925. The birds of Rancho La Brea. Carnegie

Institution of Washington 349:63-106.

Olson, S. 1985. The fossil record of birds. In Avian
biology, vol. 8, ed. D.S. Farner, J.R. King, and K.C.
Parkes, 79-238. New York: Academic Press.

Wetmore, A., and H.T. Martin. 1930. A fossil crane
from the Pliocene of Kansas. Condor 32:62-63.

Woodard, G.D., and L.F. Marcus. 1973. Rancho La
Brea fossil deposits: A re-evaluation from strati-
graphic and geological evidence. Journal of Pale-
ontology 47(l):54-69.

Received 24 June 1994; accepted 14 December 1994.

Contributions in Science, Number 452

Campbell: Cranes of Rancho La Brea ■ 13

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

59 A

Number 453
9 June 1995

Contributions
in Science

Review of Western Atlantic Species of

COCCULINID AND PSEUDOCOCCULINID LlMPETS,

with Descriptions of New Species
(Gastropoda: Cocculiniformia)

James H. McLean and M. G. Harasewych

Natural History Museum of Los Angeles County

Serial

Publications

of THE
Natural History
Museum of
Los Angeles
County

Scientific

Publications

Committee

James L. Powell, Museum President
Daniel M. Cohen, Committee
Chairman
Brian V. Brown
Kenneth E. Campbell
Kirk Fitzhugh
Robin A. Simpson, Managing Editor

• Contributions in Science, a miscellaneous series of tech-
nical papers describing original research in the life and
earth sciences.

• Science Series, long articles and collections of papers
on natural history topics.

Copies of the publications in these series are sold through
the Museum Book Shop. A catalog is available on request.

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

Printed at Allen Press, Inc., Lawrence, Kansas
ISSN 0459-8113

Review of Western Atlantic Species of

COCCULINID AND PSEUDOCOCCULINID LlMPETS,

with Descriptions of New Species
(Gastropoda: Cocculiniformia)

James H. McLean1 and M. G. Harasewych2

CONTENTS

ABSTRACT 3

INTRODUCTION 3

MATERIALS AND METHODS 3

KEY TO WESTERN ATLANTIC COCCULINIDAE AND PSEUDOCOCCULINIDAE 4

SYSTEMATICS 6

Family Cocculinidae 6

Genus Cocculina Dali, 1882 8

Cocculina rathbuni Dali, 1882 8

Cocculina messingi, new species 11

Cocculina emsoni, new species 13

Genus Coccopigya Marshall, 1986 13

Coccopigya spinigera (Jeffreys, 1883) 13

Coccopigya mikkelsenae, new species 14

Genus Coccocrater Haszprunar, 1987 16

Coccocrater pocillum (Dali, 1890), new combination 16

Coccocrater portoricensis (Dali & Simpson, 1901), new combination 17

Genus Fedikovella Moskalev, 1976 17

Fedikovella caymanensis Moskalev, 1976 17

Fedikovella beanii (Dali, 1882) 19

Family Pseudococculinidae 22

Genus Notocrater Finlay, 1926 24

Notocrater houbricki, new species 24

Notocrater youngi , new species 26

Genus Tentaoculus Moskalev, 1976 27

Tentaoculus eritmeta (Verrill, 1884), new combination 27

Tentaoculus georgiana (Dali, 1927), new combination 28

Genus Caymanabyssia Moskalev, 1986 28

Caymanabyssia spina Moskalev, 1976 28

Genus Kaiparapelta Marshall, 1986 28

Kaiparapelta askewi, new species 28

Genus Copulabyssia Haszprunar, 1988 29

Copulabyssia leptalea (Verrill, 1884) 29

Genus Amphiplica Haszprunar, 1988 29

Amphiplica venezuelensis McLean, 1988 29

REALLOCATED TAX A 30

“Cocculina” conica Verrill, 1884 30

“ Cocculina ” dalli Verrill, 1884 30

“Cocculina” reticulata Verrill, 1885 31

1. Natural History Museum of Los Angeles County, 900 Exposition Boulevard, Los Angeles, California 90007.

2. Department of Invertebrate Zoology, National Museum of Natural History, Smithsonian Institution, Washingt
D.C. 20560.

Contributions in Science, Number 453, pp. 1-33
Natural History Museum of Los Angeles County, 1995

“ Cocculina ” lissocona Dali, 1927 31

“ Cocculina ” rotunda Dali, 1927 32

“ Cocculina ” superba Clarke, 1960 32

ACKNOWLEDGMENTS 32

LITERATURE CITED 32

2 ■ Contributions in Science, Number 453

McLean and Harasewych: Cocculiniform Limpets

ABSTRACT. Western Atlantic members of the limpet families Cocculinidae and Pseudococculinidae are
reviewed. Previously described species are allocated to current genera, and new species are described. A
key to species is provided.

Family Cocculinidae: The genus Coccocrater Haszpmnar, 1987, is recognized in the western Atlantic.
New species are Cocculina messingi, Cocculina emsoni, and Coccopigya mikkelsenae. New combinations
are Coccocrater pocillum (Dali, 1890) and Coccocrater portoricensis (Dali & Simpson, 1901). Moskalev’s
(1976) assignment of Fedikovella beanii (Dali, 1882) is confirmed, and an English translation of the original
description of Fedikovella caymanensis Moskalev, 1976, is given. Lectotypes are designated for Cocculina
rathbuni Dali, 1882, Coccocrater pocillum (Dali, 1890), and Fedikovella beanii (Dali, 1882).

Family Pseudococculinidae: The genera Notocrater Finlay, 1926, Tentaoculus Moskalev, 1976, and
Kaiparapelta Marshall, 1986, are recognized in the western Atlantic. New species are Notocrater houb-
ricki, N. youngi , and Kaiparapelta askewi. New combinations are Tentaoculus eritmeta (Verrill, 1884)
and T. georgiana (Dali, 1927). Waren’s (1991) assignment of Copulabyssia leptalea is confirmed, and an
English translation of Caymanabyssia spina Moskalev, 1976, is given. A lectotype is designated for
Tentaoculus georgiana (Dali, 1927).

Six taxa originally proposed in Cocculina Dali, 1882, are removed from Cocculinidae or Pseudococ-
culinidae and assigned to other families, although not confirmed as valid species.

INTRODUCTION

Limpets of the families Cocculinidae and Pseudo-
cocculinidae occur on biogenic substrates at bathy-
al and abyssal depths. The recent collection of a
number of new species as well as fresh material of
some previously described species has enabled us
to reevaluate all the previously described taxa from
the tropical western Atlantic and the northwestern
Atlantic.

Until recently these families were poorly known.
A worldwide renaissance of interest in the group
started with Moskalev (1976), followed by Hick-
man (1983), Marshall (1986), Haszprunar (1987,
1988a, 1988b), McLean (1987, 1988, 1991, 1992),
and Dantart and Luque (1994). The number of gen-
era has increased: currently there are 6 genera rec-
ognized in the family Cocculinidae and 12 in the
family Pseudococculinidae.

During the 19th century there were nine taxa
proposed in the genus Cocculina from the western
Atlantic. Original descriptions and illustrations of
the species described by Verrill and Dali were cop-
ied by Pilsbry (1890) and Thiele (1909). C.W. John-
son (1934) and Abbott (1974) listed 10 and 11 spe-
cies, respectively, but no new information was add-
ed because few species were reported subsequent
to their initial descriptions.

Although many of the previously described spe-
cies have not been collected again, there is sufficient
information in their original descriptions to allow
most of them to be more precisely allocated to the
genera that are now available. Of the 13 species
originally described in Cocculina, only 5 now re-
main in the family Cocculinidae. Two are now
placed in the Pseudococculinidae. Six other taxa
proposed in Cocculina are removed to genera in
other families but are not allocated at the specific
level.

The four species of Cocculinidae and five species
of Pseudococculinidae described from the western
Atlantic over the last two decades bring the total
to nine species of Cocculinidae and eight species
of Pseudococculinidae, a total that greatly exceeds
the number of species known from the eastern Pa-

Contributions in Science, Number 453

cific. This number can be expected to increase as
studies using research submersibles continue. All of
the recently added new species and records have
come to light from work with submersibles in both
the western Atlantic and eastern Pacific, from either
sampling of biogenic substrates or experimental
work with larval settling.

MATERIALS AND METHODS

Species for which we have freshly collected material are
treated here in detail; those known to us only from the
type material are briefly treated. Revised English trans-
lations of the original Russian descriptions of two species
described by Moskalev (1976) are also included.

Two species described in 1882 by Dali were based on
material from widely separated localities, without desig-
nation of type localities. We have located only some of
the original syntypes in the USNM collection, but we take
this opportunity to designate lectotypes and correspond-
ing type localities for Cocculina rathbuni Dali, 1882, and
Fedikovella beanii (Dali, 1882), in order to facilitate com-
parison with the freshly collected material reported here-
in.

New material, collected using the research submersi-
bles Johnson-Sea-Link I and II and Clelia, has come from
several sources: three species were collected in the course
of experimental work on larval settling conducted off
New Providence Island, Bahamas, by Craig Young, Ro-
land Emson, and Paul A. Tyler; two species were sampled
during studies of stalked crinoids off Grand Bahama Island
by Charles Messing; two species from off St. Vincent,
Lesser Antilles, were found in the collections of the Har-
bor Branch Oceanographic Museum, and one species was
collected off South Carolina by Harasewych.

Radulae were examined with scanning electron mi-
croscopy (SEM) after dissolution of tissues in NaOH, air-
drying, and coating with carbon and gold-palladium. For
specimens in which the preservation was adequate, bodies
were critical-point dried and coated, and the external
anatomy examined with SEM.

Parameters for shell measurements are shown in figure
1. The abbreviations for the measurements are used in
the key to species that follows.

Abbreviations of museum repositories are LACM, Los
Angeles County Museum of Natural History; HBOM,
Harbor Branch Oceanographic Museum, Fort Pierce,
Florida; and USNM, National Museum of Natural His-
tory, Washington, D.C.

McLean and Harasewych: Cocculiniform Limpets ■ 3

Figure 1. Shell measurements. P = apex position, measured from anterior margin of the shell to anterior margin of
the protoconch; SH = shell height, maximum dorsoventral dimension measured perpendicular to the plane of the
aperture; SL = shell length, measured from the anterior margin to the posterior margin of the aperture; SW — shell
width, maximum lateral dimension of aperture measured perpendicular to the SL axis.

KEY TO

WESTERN ATLANTIC COCCULINIDAE
AND PSEUDOCOCCULINIDAE

1) Apical fold of protoconch short and broad;

protoconch sculpture of reticulate or concen-
tric net; radula slightly asymmetrical, 3 inner
lateral teeth (Cocculinidae) 2

- Apical fold of protoconch long and narrow;

protoconch sculpture of pustulose crystals or
anastomosing threads; radula strongly asym-
metrical, 4 inner lateral teeth (Pseudococculin-
idae) . 10

2) Protoconch with concentric sculpture ( Fedi -

kovella ) 3

- Protoconch with reticulate (honeycomb)

sculpture 4

3) Apex overhanging posterior margin of shell . . .

Fedikovella caymanenis Moskalev, 1976

- Apex not overhanging posterior margin of shell
Fedikovella beanii (Dali, 1882)

4) Shell with pits in radial rows; periostracum with

raised hairs or spines ( Coccopygia ) 5

- Shell without pits; periostracum smooth . . 6

5) Shell high (SH/SL > 0.35); apex near posterior

margin of shell (P/SL > 0.8)

. Coccopigya spinigera (Jeffreys, 1883)

4 ■ Contributions in Science, Number 453

- Shell low (SH/SL > 0.30); apex nearly central

(P/SL < 0.6)

Coccopigya mikkelsenae, new species

6) Copulatory organ at tip of right oral lappet

(Cocculina) 7

- Copulatory organ branched from right cephalic

tentacle ( Coccocrater ) 9

7) Shell high (SH/SL > 0.40), with 1 anterior and
2 posterior keels; apex posterior of center (P/
SL > 0.6) . . . Cocculina emsoni, new species

- Shell low (SH/SL < 0.38), lacking keels; apex

near center (P/SL < 0.6) 8

8) Shell with radial sculpture more prominent than
concentric sculpture; pigmented eyes absent . .
Cocculina rathbuni Dali, 1882

- Shell with concentric sculpture more promi-

nent than axial sculpture; pigmented eyes pres-
ent Cocculina messingi, new species

9) Shell high (SH/SL > 0.45); apex near posterior

margin (P/SL > 0.7)

Coccocrater pocillum (Dali, 1890)

- Shell low (SH/SL) < 0.45); apex anterior of

center (P/SL < 0.5)

Coccocrater portoricensis

(Dali &: Simpson, 1901)

10) Shell with interior septum; protoconch finely

McLean and Harasewych: Cocculiniform Limpets

Figures 2-5. Cocculina rathbuni Dali, 1882. 2. Dorsal, lateral, and ventral views of shell (USNM 860363). Scale bar
= 1.0 mm. 3, 4. Dorsal and lateral views of protoconch of specimen in figure 1. Scale bars = 50 jam. 5. Detail of
protoconch sculpture in figure 3. Scale bar = 5 jam.

pitted ( Tentaoculus ) 11

- Shell lacking interior septum; protoconch with-

out pits 12

11) Shell low (SH/SL < 0.40); apex near center

(P/SL < 0.6)

Tentaoculus eritmeta (Verrill, 1884)

- Shell high (SH/SL > 0.40); apex near posterior

margin (P/SL > 0.7)

Tentaoculus georgiana (Dali, 1927)

12) Teleoconch sculpture granulate ........ 13

- Teleoconch sculpture of raised concentric

ridges 16

13) Protoconch with anastomosing threads; teleo-
conch sculpture granulate ( Notocrater ) . . 14

- Protoconch with prismatic crystals; teleoconch
sculpture of anastomosing network ..... 15

14) Protoconch finely pustulose; teleoconch

Contributions in Science, Number 453

coarsely pustulose; eyes present; foot and man-
tle edge with fine brown spots

Notocrater houbricki, new species

- Protoconch coarsely pustulose; teleoconch

finely pustulose; eyes absent; foot and mantle
uniformly beige in color

Notocrater youngi, new species

15) Shell without posterior ridges or indentations
in margin; teleoconch with sharp conical gran-
ules forming diagonally reticulate pattern . . .
..... Caymanabyssia spina Moskalev, 1976

- Shell with 2 posterior ridges producing inden-

tations in margin; teleoconch lacking diago-
nally oriented conical granules .

.......... Kaiparapelta askewi, new species

16) Shell small, length < 4 mm

....... Copulabyssia leptalea (Verrill, 1884)

McLean and Harasewych: Cocculiniform Limpets ■ 5

Figures 6-11. Cocculina rathbuni Dali, 1882. 6. Ventral view of critical-point-dried animal (shell in figures 2-5). Scale
bar = 200 yum. 7. Right lateral view of head. Scale bar = 200 yum. 8. Right epipodial tentacle. Scale bar = 100 yum. 9.
Dorsal view of radular ribbon. Scale bar = 20 yum. 10. Lateral view of longitudinally cut radula, revealing relative heights
of tooth fields. Scale bar = 20 yum. 11. Detail of pluricuspid and marginal teeth. Scale bar = 10 yum.

co = copulatory organ; ct = cephalic tentacle; et = epipodial tentacle; m = marginal teeth; pc = pluricuspid tooth;
pg = pseudoplicatid gill; r = rachidian tooth; 1 = first lateral tooth; 2 = second lateral tooth; 3 = third lateral tooth.

- Shell large, length > 7 mm

. . . . Amphiplica venezuelensis McLean, 1988

SYSTEMATICS

Family Cocculinidae Dali, 1882

DIAGNOSIS. Apical fold of protoconch short
and broad; protoconch sculpture of reticulate net.
Radula slightly asymmetrical, inner lateral teeth 3.
REMARKS. Marshall (1986) defined a number

6 ■ Contributions in Science, Number 4S3

of cocculinid genera on shell and radular charac-
ters; Haszprunar (1987, 1988b) treated anatomy of
cocculinid genera and compared the group to other
cocculiniform families. These works should be
consulted for further details.

Six genera of Cocculinidae are currently recog-
nized: Cocculina Dali, 1882, Coccopigya Marshall,
1986, Coccocrater Haszprunar, 1987, Paracoccu-
lina Haszprunar, 1987, Fedikovella Moskalev, 1976,
and T euthirostria Moskalev, 1976. No species of

McLean and Harasewych: Cocculiniform Limpets

Figures 12-14. Cocculina messingi, new species. 12. Dorsal, lateral, and ventral views of holotype. Scale bar — 1.0
mm. 13, 14. Dorsal and lateral views of protoconch of holotype. Scale bars = 50 fim.

P aracocculina and Teuthirostria are known from
the western Atlantic or eastern Pacific.

Four of the genera (excepting Fedikovella and
Teuthirostria ) have protoconch sculpture of a raised
honeycomb network aligned in longitudinal rows.
The most significant character for separation of
genera is the position of the copulatory organ in
these simultaneous hermaphrodites: modified tip of
right oral lappet in Cocculina, branched from the
base of right cephalic tentacle in Coccopigya and
Coccocrater, from right side of foot in Paracoc
culina, but yet unknown in Fedikovella and Teu-
thirostria. The genus Coccopigya can be recog-
nized by its hirsute periostracum, but no clear sep-
aration of Cocculina, Coccocrater, and Paracoc-
culina can be defined on shell or radular characters.

Contributions in Science, Number 453

Species in most genera have a pair of posterior
epipodial tentacles (figures 5, 8) and a pseudopli-
catid gill (figures 15, 19), as defined by Haszprunar
(1987), on the right side.

Marshall considered the ridges on the edge of
the rachidian tooth to represent vestiges of the first
laterals, numbering the further laterals as 2-4.
However, to avoid ambiguity, we number the slen-
der lateral teeth that have cusps as 1-3 (figure 10,
1-3). The large multicuspid tooth that separates the
laterals from the marginals is here called the plur-
icuspid tooth (figures 10, 11, pc) rather than lateral
tooth 4, as it has no features in common with the
other lateral teeth. The pluricuspid teeth are the
largest teeth and are clearly the most effective teeth
in the row.

McLean and Harasewych: Coeculiniform Limpets ■ 7

Figures 15-19, Cocculina messingi, new species. 15. Ventral view of critical-point-dried animal of holotype. Scale
bar = 200 /im. 16. Anterior margin of oral lappet. Scale bar = 100 /mi. 17. Copulatory organ. Scale bar = 100 n m. 18.
Right epipodial tentacle. Scale bar = 100 /im. 19. Detail of pseudoplicatid gill. Scale bar = 20 Atm.
ct = cephalic tentacle; et = epipodial tentacle; m = mouth; ol = oral lappet; pg = pseudoplicatid gill.

Genus Cocculina Dali, 1882

Type species (subsequent designation, Dali, 1908:

340): Cocculina rathbuni Dali, 1882.

DIAGNOSIS. Protoconch with honeycomb
sculpture, periostracum smooth; teleoconch sculp-
ture of fine radial ribs and growth lines; copulatory
organ at tip of right oral lappet.

REMARKS. Cocculina is represented in the
western Atlantic by the type species and two new
species here described: C. messingi and C. emsoni.
Other western Atlantic taxa described in this genus
are herein reallocated to other genera or enumer-
ated under a heading of species for which the ge-
neric affinity or synonymy has not been established.

8 ■ Contributions in Science, Number 453

In the eastern Pacific, the genus is represented by
C. baxteri McLean, 1987, C. cowani McLean, 1987,
and C. craigsmithi McLean, 1992, the last unique
in the family for its occurrence on whale bone.

Cocculina rathbuni Dali, 1882

Figures 2-11

Cocculina rathbuni Dali, 1882:402; Dali, 1889:347,
pi. 15, figs. 5, 7; Pilsbry, 1890:132, pi. 25, figs. 5,
6 [copy of Dali]; Dali, 1908:340 [type designa-
tion]; Thiele, 1909:6, pi. 2, figs. 1, 2; C.W. John-
son, 1934:66 [checklist only]; Abbott, 1974:34,
fig. 192; McLean, 1987:325, figs. 1-4.

DESCRIPTION. Shell (figure 2) medium to large-

McLean and Harasewych: Cocculiniform Limpets

Figures 20-24. Cocculina messingi, new species. 20. Dorsal view of radular ribbon. Scale bar = 20 /im. 21. Lateral
view of longitudinally cut radula revealing relative heights of tooth fields. Scale bar - 20 ^m. 22. Detail of rachidian,
lateral, and pluricuspid teeth. Scale bar = 10 jim. 23. Detail of marginal teeth. Scale bar = 6 /xm. 24. Detail of distal
ends of outer marginal teeth. Scale bar = 2 ixm.

sized for family (maximum length 11 mm), thin,
not eroded, white, periostracum thin. Shell height
moderate, that of illustrated specimen 0.35 times
length. Anterior and posterior slopes nearly straight,
lateral slopes slightly convex. Outline in dorsal view
elongate-oval, anterior end slightly narrower than
posterior end; aperture not planar, ends raised rel-
ative to sides of shell. Apex slightly posterior to
center, protoconch slightly below highest point of
shell, extending posteriorly. Protoconch length 240
/urn, protoconch sculpture of honeycomb net pat-

Contributions in Science, Number 453

tern, aligned longitudinally in rows (figures 3, 4).
Surface within netted area of protoconch finely pit-
ted (figure 5). Tip of protoconch immersed in pos-
terior slope of shell. Teleoconch sculpture of raised
concentric growth lines and fine radial striae; con-
centric sculpture more prominent than radial sculp-
ture, not raised at intersections with radial striae.
Shell edge thin and sharp.

Dimensions. Length 11, width 6.5, height 2.75
mm (original description); length 5.7, width 3.9,
height 2.0 mm (figure 2).

McLean and Harasewych: Cocculiniform Limpets ■ 9

Figures 25-28. Cocculina emsoni, new species. 25. Dorsal, lateral, and ventral views of holotype. Scale bar =1.0
mm. 26, 27. Dorsal and lateral views of protoconch of holotype. Scale bars = 50 /um. 28. Detail of protoconch sculpture
in figure 26. Scale bar = 5 n m.

External Anatomy (figures 6-8). Eyes lacking,
copulatory organ at tip of right oral lappet, basal
portion enlarged, producing bilobed effect; pseu-
doplicatid gill on right side dorsal to oral lappet;
pair of posterior epipodial tentacles (figures 6, 8),
with tufts of cilia under high magnification (figure
8); area at side of head with tufts of cilia like those
of epipodial tentacles (figure 7).

Radula (figures 9-11). Rachidian broad, outer
edges weakly defined, tip with single small over-
hanging cusp; first lateral with four cusps on outer
edge, second with three, third singly cusped; plur-
icuspid long and broad, with inner and outer cusps.
Marginals similar in size.

NEW RECORD. Off Southwest Reef, New
Providence Island, Bahamas (24°54'04"N,
77°33'14"W), 518 m, Johnson-Sea-Link II, dive
2317, 9 May 1992. Disposition of specimens: USNM

10 ■ Contributions in Science, Number 453

860363, LACM 151187, HBOM 065:03884. On
palmetto ( Sabal palmetto) fronds deployed earlier
for sampling of invertebrate settling. According to
R. Emson (pers. comm.), this species along with C.
emsoni is frequently recruited on palmetto sub-
strates at this locality. Further details will be pro-
vided in a forthcoming paper by C.M. Young, P.A.
Tyler, and R.H. Emson.

REMARKS. Dali mentioned material of Coc-
culina rathbuni from three stations [Massachusetts,
Barbados, and Martinique] in the original descrip-
tion, without designation of a holotype or type
locality. Although the 11 mm specimen (from Mas-
sachusetts?) has not been located, the other two
syntypes are present in the USNM. The Barbados
specimen (Blake sta. 288, USNM 333750) is a par-
tially dissected, dried animal with no shell. The
Martinique specimen (Blake sta. 195, USNM

McLean and Harasewych: Cocculiniform Limpets

Figures 29-35. Cocculina emsoni, new species. 29. Ventral view of critical-point-dried animal of holotype. Scale bar
= 200 jum. 30. Right lateral view of head. Scale bar = 50 /urn. 31. Right epipodial tentacle. Scale bar = 50 /um. 32.
Anterior margin of oral lappet. Scale bar — 100 /um. 33. Dorsal view of radular ribbon. Scale bar = 20 /urn. 34. Detail
of rachidian and lateral teeth. Scale bar = 10 /am. 35. Detail of pluricuspid and marginal teeth. Scale bar = 10 /im.
co = copulatory organ; ct = cephalic tentacle; et = epipodial tentacle.

126807) is correctly labelled C. rathbuni ; it mea-
sures 6.7 mm in length and agrees with our figured
specimen and that of McLean (1987). We here des-
ignate it the lectotype and Martinique the type lo-
cality.

Cocculina messingi, new species

Figures 12-24

DESCRIPTION. Shell (figure 12) medium-sized
for family (maximum length 5.5 mm), thin, not

Contributions in Science, Number 453

eroded, white, periostracum thin. Shell height mod-
erate, that of holotype 0.38 times length. Anterior
and posterior slopes faintly convex, lateral slopes
more markedly convex. Outline in dorsal view elon-
gate-oval, anterior end slightly narrower than pos-
terior end; aperture not planar, ends raised relative
to sides of shell. Apex slightly posterior to center,
protoconch at highest point of shell, extending pos-
teriorly. Protoconch length 200 /tm, protoconch
sculpture of honeycomb pattern, aligned longitu-
dinally in rows. Tip of protoconch immersed in

McLean and Harasewych: Cocculiniform Limpets! II

Figures 36-39. Coccopigya spinigera (Jeffreys, 1883). 36. Dorsal, lateral, and ventral views of syntype, USNM 177890.
Scale bar =1.0 mm. 37, 38. Dorsal and lateral views of protoconch of syntype in figure 36. Scale bars = 50 |iim. 39.
Detail of protoconch sculpture in figure 38. Scale bar = 5 /urn.

posterior slope of shell. Sculpture of raised con-
centric growth lines and fine radial striae; concentric
sculpture more prominent than radial sculpture, not
raised at intersections with radial striae. Shell edge
thin and sharp. Muscle scar and anterior pallial
attachment scar well marked, inner edge of muscle
scar irregular.

Dimensions. Length 5.5, width 3.6, height 2.1
mm (holotype); length 5.3, width 3.8, height 2.1
mm (paratype 1); length 5.4, width 4.0, height 2.4
(paratype 2).

External Anatomy (figures 15-19). Animal trans-
lucent white, showing red buccal musculature, with
large black eyes. Ventral surface of oral lappet with
broad ciliated band (figure 16). Penis bilobed, at tip
of right oral lappet, with open seminal groove (fig-

ure 17). Pseudoplicate gill showing long bands of
cilia (figure 19). Pair of posterior epipodial tentacles
(figure 18).

Radula (figures 20-24). Rachidian broad, tip small,
with single cusp; first lateral with broad base and
three cusps, second lateral narrow, with three cusps,
third with single cusp. Pluricuspid long and broad
with large central denticle, one smaller denticle on
outer and two on inner side. Marginals similar in
size.

TYPE LOCALITY. South of Settlement Point,
Grand Bahama Island, Bahama Islands (26°37'31"N,
78°58'56"W), 412 m, on pencil-sized piece of wood
along with Notocrater houbricki, new species.

TYPE MATERIAL. Three specimens from type
locality, collected by Dr. Charles Messing, using

12 ■ Contributions in Science, Number 453 McLean and Harasewych: Cocculiniform Limpets

deep-submersible Johnson-Sea-Link II, dive 2335,
18 May 1992. Holotype USNM 860353, paratype
1 USNM 860354, paratype 2 LACM 2735.

REMARKS. This species differs from C. rath-
buni as treated above in its coarser concentric sculp-
ture, less prominent radial sculpture, and the pres-
ence of prominent black eyes. Eyes have not pre-
viously been reported in any species of Cocculina.

ETYMOLOGY. The name honors the collector,
Charles Messing of Nova University, Dania, Flor-
ida.

Cocculina emsoni, new species

Figures 25-35

DESCRIPTION. Shell (figure 25) small for family
(maximum length 3.3 mm), thin, not eroded, white,
periostracum thin. Shell height moderately high,
that of holotype 0.48 times length. Anterior slope
convex, posterior slope nearly straight, lateral slopes
slightly convex. Outline in dorsal view elongate-
oval, anterior end with keeled projection producing
concave area close to tip on both sides, posterior
with two projecting keels, forming single concave
embayment in outline. Anterior end slightly nar-
rower than posterior end; aperture not planar, ends
raised relative to sides of shell except that the an-
terior and posterior keeled projections extend
downward. Apex posterior to center, situated at 2A
shell length from anterior end. Protoconch below
highest point of shell, extending posteriorly. Pro-
toconch length 205 jum, protoconch sculpture of
honeycomb pattern, with raised ridges forming ir-
regular hexagons (figures 26, 27). Surface within the
netted area of protoconch finely pitted (figure 28).
Tip of protoconch immersed in posterior slope of
shell. Sculpture of irregular concentric growth lines
and raised radial ribs; secondary radial ribs arising
after shell length of 1.5 mm attained. Midline of
anterior slope with sharply raised anterior ridge on
which there are secondary ribs; posterior slope with
two raised keels, between which there are second-
ary ribs. Shell edge thin and sharp; interior with
grooved areas corresponding to strong anterior ridge
and two posterior ridges.

Dimensions. Length 3.3, width 1.8, height 1.6
mm (holotype).

External Anatomy (figures 29-32). Penis simple,
derived at base of right oral lappet (figure 30); epi-
podial tentacles and pseudoplicated gill present.

Radula (figures 33-35). Rachidian broad, basal
outline hidden, tip with main cusp and two small
lateral cusps; first and second laterals with three
cusps on outer edge, third lateral with single cusp;
pluricuspid broad, with long main cusp and inner
and outer lateral cusps. Marginals similar in size.

TYPE LOCALITY. Off Southwest Reef, New
Providence Island, Bahamas (24°54'04"N,
77°33'14"W), 518 m on palmetto fronds. Further
details will be provided in a forthcoming paper by
R. Emson, C.M. Young, and P.A. Tyler.

Contributions in Science, Number 453

TYPE MATERIAL. Three specimens retrieved
from palmetto fronds placed on bottom for larval
settlement experiments by R. Emson, C.M. Young,
and P.A. Tyler, Johnson-Sea-Link II, dive 2317, 9
May 1992. Holotype USNM 860355, 1 paratype
USNM 860356, 1 paratype LACM 2736.

REMARKS. Cocculina emsoni is remarkable for
its strong anterior ridge and two posterior ridges.
Cocculina angulata Watson, 1886, from the Phil-
ippines (Watson, 1886:30, pi. 4, fig. 2a-c) has an
anterior shell ridge but lacks the two posterior ridg-
es. Direct comparisons of specimens should be made
before commenting further about a possible affinity
between the two species. For C. emsoni it is a
reasonable supposition that the two posterior ridges
serve to shield the two posterior epipodial tentacles,
but the function of the anterior ridge is unknown.

ETYMOLOGY. Named after Roland Emson of
King’s College London, whose experimental work
on larval settling brought this species to light.

Genus Coccopigya Marshall, 1986

Replacement name for Coccopygia Dali, 1889, not
Reichenbach, 1882. Type species by monotypy:
Cocculina spinigera Jeffreys, 1883.

DIAGNOSIS. Protoconch with reticulate sculp-
ture (as in Cocculina ); periostracum thick, hirsute;
teleoconch sculpture of radial ribs and pit rows;
copulatory organ branched from right cephalic ten-
tacle.

REMARKS. The type species is represented in
the northeastern and northwestern Atlantic; a sec-
ond western Atlantic species is described herein.
There are no species known in the eastern Pacific.
Marshall (1986) treated five living and three fossil
species from New Zealand.

Coccopigya spinigera (Jeffreys, 1883)

Figures 36-39

Cocculina spinigera Jeffreys, 1883:393, pi. 44, figs.
1-lc; Verrill, 1884:203; Pilsbry, 1890:125, pi. 25,
figs. 9, 10 [copy Jeffreys]; Abbott, 1974:34 [not
fig. 198].

Cocculina (section Coccopigya ) spinigera; Dali,
1889:348, pi. 31, figs. 7-9.

Cocculina ( Coccopigya ) spinigera; Thiele, 1909:

15, pi. 3, figs. 9, 10.

Coccopigya spinigera ; Marshall, 1986:512, figs. 2B,
3D, E, 12C; Waren, 1991:80, fig. 19A, B, D, F,
H; Dantart and Luque, 1994:278, figs. 1-6, 15,

16, 18.

REMARKS. We illustrate the shell and proto-
conch of a syntype (USNM 177890) from the Outer
Hebrides, Scotland, Triton sta. 10 (59°40'N,
7°21'W), 943 m. The fine pits that occur within the
net pattern on the protoconch are illustrated here
for the first time.

Coccopigya spinigera occurs in the northeastern
and northwestern Atlantic south to North Carolina.

McLean and Harasewych: Cocculiniform Limpets 113

Figures 40-43. Coccopigya mikkelsenae, new species. 40. Dorsal, lateral, and ventral views of holotype, periostracum
removed. Scale bar =1.0 mm. 41. Detail of periostracum at shell edge of paratype 1. Scale bar = 100 n m. 42, 43.
Dorsal and lateral views of protoconch of holotype. Scale bars = 50 jum.

DalPs (1889:pl. 31, fig. 8) original illustrations were
based on specimens received from Jeffreys and in-
cluded a drawing of the copulatory organ. Marshall
(1986) included SEM views of the shell and radula
and a drawing of the copulatory organ, also based
on a syntype specimen. Dali (1889:fig. 9) showed a
broad rachidian with three similar-sized cusps, but
in Marshall’s preparation the rachidian does not
show, suggesting that it folded under in drying.
Waren (1991) gave SEM views of shells from Ice-
land. Dantart and Luque (1994) illustrated material
from Spain and added a species described by Dautz-
enberg and Fischer to the synonymy.

Coccopigya mikkelsenae , new species

Figures 40-47

14 ■ Contributions in Science, Number 453

DESCRIPTION. Shell (figure 40) medium-sized
for family (maximum length 6.2 mm), thin, not
eroded, white under thick periostracum bearing long
hairs (figure 41). Shell height low, that of holotype
0.27 times length. All slopes straight to slightly con-
vex. Outline in dorsal view elongate-oval, sides
nearly parallel; aperture not planar, sides raised
slightly relative to ends. Apex slightly posterior to
center, slightly below level of highest point of shell,
protoconch extending posteriorly. Protoconch
length 210 /Ltm, protoconch sculpture of honey-
comb, nearly rectangular net pattern, aligned lon-
gitudinally in rows (figures 42, 43). Tip of proto-
conch immersed in posterior slope of shell. Sculp-
ture of irregular concentric growth lines and faint
radial striae, and scattered, radially aligned pits. Ra-
dial and concentric sculpture of equal prominence.

McLean and Harasewych: Cocculiniform Limpets

Figures 44-47. Coccopigya mikkelsenae, new species. 44. Oblique view of rachidian, lateral, and pluricuspid teeth.
Scale bar = 20 /urn. 45. Dorsal view of rachidian, lateral, and pluricuspid teeth. Scale bar = 20 /urn. 46. Dorsal view of
radular ribbon. Scale bar = 25 /um. 47. Detail of marginal teeth. Scale bar = 5 /urn.

Shell edge thin and sharp. Muscle scar and anterior
pallial attachment scar well marked.

Dimensions. Length 6.2, width 4.7, height 1.7
mm (holotype, posterior end broken); length 6.9,
width 4.8, height 2.0 mm (paratype).

External Anatomy. Animal lacking pigmented
eyes, penis branching off base of right tentacle, tip
of penis with single, tapering lobe. No gill evident,
but specimens poorly preserved (body not exam-
ined with SEM).

Radula (figures 44-47). Rachidian broad, upper
edge broad, small pointed cusp emerging from up-
per edge; first lateral with secondary cusp on outer
edge, second lateral with four cusps, third lateral
with single cusp and long shaft. Pluricuspid broad,
with large main cusp and inner and outer secondary
cusps. Marginals similar in size.

TYPE LOCALITY. Off Chateau Belair Bay, St.
Vincent, Lesser Antilles (13°10.5'N, 61°15.5'W), 421
m, on wood, with Fedikovella beanii.

Contributions in Science, Number 453

TYPE MATERIAL. Two specimens from type
locality collected by John E. Miller on deep-sub-
mersible Johnson-Sea-Link II, dive 1742, 23 April
1989. Holotype, USNM 860357, 1 paratype HBOM
065:03786. Most of the periostracum on the ho-
lotype was removed for SEM preparation. The un-
figured paratype was also coated for SEM, and the
periostracum is in the process of flaking off.

REMARKS. Coccopigya mikkelsenae meets the
criteria of the genus in having a coarse periostracum
with long hairs, corresponding pits along the radial
ribs, and the penis branched at the base of right
tentacle. The rachidian tooth is broader than that
of other members of the family treated here. The
single small cusp of the rachidian is also unlike that
of other species treated by Marshall (1986).

On shell characters, Coccopigya mikkelsenae dif-
fers from C. spinigera in its lower profile and more
central apex.

ETYMOLOGY. We are pleased to name this

McLean and Harasewych: Cocculiniform Limpets ■ 15

Figures 48, 49. Coccocrater pocillum (Dali, 1890). 48. Dorsal, lateral, and ventral views of lectotype (USNM 87586).
Scale bar =1.0 mm. 49. Detail of periostracum at shell edge of lectotype. Scale bar = 100 jum.

species after Paula Mikkelsen of the Harbor Branch
Oceanographic Museum, Fort Pierce, Florida, who
brought the material to our attention.

Genus Coccocrater
Haszprunar, 1987

Type species by original designation: Cocculina ra-
diata Thiele, 1903.

DIAGNOSIS. Protoconch with reticulate sculp-
ture, periostracum smooth; teleoconch sculpture of
fine radial ribs and growth lines; copulatory organ
branched from right cephalic tentacle.

REMARKS. This genus differs from Coccopigya,
which also has the copulatory organ as a branch
from the base of the right cephalic tentacle, in lack-
ing the hirsute periostracum.

Haszprunar’s (1987:321) diagnosis states that the
copulatory organ is “associated with the right ce-
phalic tentacle.” McLean (1987:330) incorrectly
stated “the enlarged right cephalic tentacle serving
as the penis.”

16 ■ Contributions in Science, Number 453

The genus is represented in the eastern Pacific by
Coccocrater agassizii (Dali, 1908), treated by Ha-
szprunar (1987) and McLean (1987).

Coccocrater pocillum
(Dali, 1890),
new combination

Figures 48, 49

Cocculina ( Coccopigya ) pocillum Dali, 1890:340;

Thiele, 1909:16.

REMARKS. This previously unfigured species was
described from 1600 m off Tobago. The periostra-
cum lacks hairs (figure 49) and each specimen had
a “well marked verge extending from the right ten-
tacle,” a character combination that agrees with
Coccocrater. Dali also reported that the species lacks
epipodial filaments. A lectotype shell (USNM 87586)
is designated and illustrated here (length 4.65, width
3.45, height 2.5 mm). The remaining paralectotype
specimen is recataloged as USNM 860386 (length
5.5, width 3.8, height 3.0 mm).

McLean and Harasewych: Cocculiniform Limpets

Figure 50. Coccocrater portoricensis (Dali &c Simpson, 1901). Dorsal, lateral, and ventral views of holotype (USNM
160496). Scale bar =1.0 mm.

Coccocrater portoricensis
(Dali & Simpson, 1901),
new combination
Figure 50

Cocculina portoricensis Dali and Simpson, 1901:
440, pi. 53, figs. 18, 19; Abbott, 1974:35, fig. 202.
Cocculina ( Coccopigya ) portoricensis ; Thiele, 1909:
16.

REMARKS. This species of 12 mm length from
566 m off San Juan Harbor, Puerto Rico, does not
have a hirsute periostracum and was said to have
a “large verge projecting from the right tentacle”
and the “ctenidium carried over so that it appears
to spring from the right side of the animal.” Again,
this is a character combination in agreement with
Coccocrater. The species remains known only from
the holotype (USNM 160496), reillustrated here
(figure 50).

Genus Fedikovella Moskalev, 1976

Type species by original designation: Fedikovella
caymanensis Moskalev, 1976.

DIAGNOSIS. Protoconch with concentric sculp-
ture; periostracum smooth; teleoconch sculpture
clathrate, apex overhanging concave posterior slope;
cephalic tentacles equal in size.

REMARKS. Moskalev included Cocculina be-
anii Dali, 1882, in Fedikovella because of the small
rachidian tooth figured by Dali, and the Indo-Pacific
C. capulus Thiele, 1925, again citing the small ra-
chidian tooth figured by Thiele. We hesitate to ac-
cept the inclusion of the latter species. Moskalev’s
figure of the radula of the type species was based
on phase contrast optical microscopy; tooth bases

are not shown and the resolution is sufficient only
to place the species in Cocculinidae rather than
Pseudococculinidae.

This genus can be accepted on the distinction
provided by the peculiar undulating, concentric
sculpture of the protoconch shown here (figures
52, 53) for Fedikovella beanii (Dali, 1882). How-
ever, this sculpture is not radically different from
the honeycomb pattern of most cocculinid genera,
from which the difference may be only that the
longitudinal connections of the net are not formed.
Moskalev’s (1976:pl. 2, fig. 2) figure of the type
species Fedikovella caymanensis, which purported
to show the protoconch, was instead an enlarged
view of the clathrate sculpture of the earliest te-
leoconch, as also noted by Marshall (1986:508).

There is still no published information about the
internal anatomy or whether there is a copulatory
appendage of any kind. Dali (1889:347) believed
that all four specimens of “ Cocculina ” beanii that
he examined were females, as it was then not un-
derstood that all cocculinids are simultaneous her-
maphrodites. However, B. Marshall (pers. comm.)
has examined paratype material of F. beanii and
noted a copulatory organ on the right behind the
base of the right cephalic tentacle.

There are no eastern Pacific species allocated to
this genus.

Fedikovella caymanensis Moskalev, 1976

Fedikovella caymanensis Moskalev, 1976:62, fig.

1, pl- 2, figs. 1, 2.

Moskalev’s description was translated by G.V.
Shkurkin in a privately circulated “reprint” dated
December 1978. That translation is repeated here,

Contributions in Science, Number 453

McLean and Harasewych: Cocculiniform Limpets! 17

Figures 51-53. Fedikovella beanii (Dali, 1882). 51. Dorsal, lateral, and ventral views of shell (USNM 860358). Scale
bar =1.0 mm. 52, 53. Dorsal and lateral views of protoconch of specimen in figure 51. Scale bars = 50 yum.

altered to place it in telegraphic style and with ter-
minology slightly changed in places to conform to
that used elsewhere in this paper:

Shell small, high, thin, apex projecting to pos-
terior margin or beyond. Protoconch with concen-
tric sculpture, protoconch pressed against posterior
slope of shell. Anterior slope convex, posterior slope
short, straight. Aperture elliptical, margin whole.
Color opaque cream; periostracum well developed.
Sculpture of intersecting radial and concentric rib-
bing, forming nearly equilateral quadrangles; radial
ribs wider than concentric ribs; ribbing worn in
places. Shell interior yellowish white, exterior
sculpture showing through. Rachidian tooth den-
ticulate, similar to first to third lateral teeth; first
and second lateral teeth denticulate, pluricuspid

18 ■ Contributions in Science, Number 453

tooth largest, having three denticles. Cephalic ten-
tacles equal in size; two posterior epipodial tenta-
cles present. Shell length 1.43-4.10 mm.

TYPE LOCALITY. Western end of Cayman
Trough (19°00'6"N, 80°29'5"W), 6800 m. 33 spec-
imens on wood, collected with research vessel Aka-
demic Kurchatov , cruise 14, sta. 1242 A, 2.5 m Sigs-
bee dredge, 20 March 1967. Additional locality:
east end Cayman Trough (19°38'5"N, 76°37'8"W),
sta. 1267, 6740-6780 m, 5 specimens on various
vegetative remains.

REMARKS. Moskalev’s illustrations do not in-
clude a full view of the shell nor is there any in-
dication of actual shell height. Most Cocculinidae
occur at continental shelf and slope depths, and
this species is unusual in its abyssal occurrence.

McLean and Harasewych: Cocculiniform Limpets

Figures 54-57. Fedikovella beanii (Dali, 1882). 54. Dorsal view of radular ribbon. Scale bar — 25 /*m. 55. Dorsal
view of radular ribbon with marginal teeth folded back. Scale bar = 25 /urn. 56. Detail of rachidian and lateral teeth.
Scale bar = 12.5 /mi. 57. Detail of distal ends or outer marginal teeth. Scale bar = 5 jum.

Fedikovella beanii (Dali, 1882)

Figures 51-57

Cocculina beanii Dali, 1882:403; Dali, 1889:347,
pi. 25, figs. 2, 4, 8; Pilsbry, 1890:132, pi. 25, figs.
23, 24 [copy Dali]; Thiele, 1909:6, pi. 2, figs. 3,
4; C.W. Johnson, 1934:66 [checklist only]; Ab-
bott, 1974:34, fig. 194 [copy Dali].

Fedikovella beanii ; Moskalev, 1976:64 [as beani].

DESCRIPTION. Shell (figure 51) medium-sized
for family (maximum length 8 mm, original de-
scription), thin, not eroded, white, periostracum
thin. Shell moderately high, that of illustrated spec-
imen 0.47 times length. Anterior slope convex, pos-
terior slope concave, lateral slopes nearly straight.
Outline in dorsal view oval, anterior end slightly
narrower than posterior end; aperture planar, ends
not raised relative to sides of shell. Apex posterior
to center, to left of midline; situated at 2A shell
length from anterior end. Protoconch below high-
est point of shell, extending posteriorly. Proto-

conch length 240 jum, protoconch sculpture (at least
near tip where unworn) of parallel, concave ridges
aligned to extend across but not longitudinally. Sur-
face near ridges with fine pits (figures 52, 53). Tip
of protoconch immersed in posterior slope of shell.
Sculpture of raised concentric ridges and radial ribs
of lesser strength, producing beaded effect partic-
ularly on posterior slope and at growth stages great-
er than 3 mm in shell length. Shell edge thin and
sharp. Interior with well marked muscle and pallial
attachment scars.

Dimensions. Length 8, width 5, height 4 mm
(original description); length 5.1, width 3.7, height
2.4 mm (figure 51).

External Anatomy. Dali (1882) reported equal
cephalic tentacles and the gill longer and larger than
that of Cocculina rathbuni. Preservation was poor
in the present material, and it was not used for
critical-point drying and SEM examination. No gill
or penis was apparent; pigmented eyes were lacking;
two posterior epipodial tentacles were present.

Contributions in Science, Number 453

McLean and Harasewych: Cocculiniform Limpets ■ 19

Figures 58-60. Notocrater houbricki, new species. 58. Dorsal, lateral, and ventral views of holotype. Scale bar =1.0
mm. 59, 60. Dorsal and lateral views of protoconch of holotype. Scale bars = 50 jum.

Radula (figures 54-57). Rachidian tooth with
narrow, elevated shaft, overhanging cusp with cen-
tral and two lateral denticles of similar size; base
of shaft bifurcated, superimposed on broader basal
membrane; first and second lateral teeth narrow,
elbowed, with main cusp and one or two lateral
cusps on outer edge, third lateral tooth with single
cusp. Pluricuspid broad, with tapered main cusp
and inner and outer lateral cusps. Marginals of sim-
ilar size.

NEW RECORD. Off Chateau Belair Bay, St. Vin-
cent, Lesser Antilles (13°10.5'N, 61°15.5'W, 421 m,
on wood, with Coccopigya mikkelsenae, new spe-
cies. Eight specimens collected by deep-submersi-
ble Johnson-Sea-Link II, dive 1742, 23 April 1989.
Distribution: 3 specimens USNM 860358, 3 spec-
imens HBOM 065:03787, 2 specimens LACM
151188.

20 ■ Contributions in Science, Number 453

REMARKS. Dali’s original description gave sev-
eral localities and station numbers, including south
of Martha’s Vineyard Island, Massachusetts, and
Martinique, but did not cite a catalog number or
designate a type locality. Here we designate a lec-
totype, USNM 333751 from USFC sta. 997, 335
fms off Martha’s Vineyard Island. The specimen is
6.46 mm in length, chipped at the posterior margin.
A slip marked “Type Fig’d.” accompanies the spec-
imen. This seems to be the specimen illustrated by
Dali (1889). The largest specimen of Fedikovella
beanii in the USNM collection is 7.2 mm in length,
from Blake sta. 195, 502.5 fms, off Martinique.

Dali’s original description emphasized that the
anterior slope was longer than in Cocculina rath-
buni and the sculpture stronger and more cancel-
lated, “even slightly spinous at intersections.” The
rachidian (figured later by Dali, 1889) was said to

McLean and Harasewych: Cocculiniform Limpets

Figures 61-65. Notocrater houbricki, new species. 61. Ventral view of critical-point-dried animal of paratype 1. Scale
bar = 100 /xm. 62. Dorsal view of radular ribbon. Scale bar = 10 Atm. 63. Anterior view of rachidian, lateral, and
pluricuspid teeth. Scale bar =10 jum. 64. Detail of rachidian and lateral teeth. Scale bar = 5 n m. 65. Detail of pluricuspid
and marginal teeth. Scale bar = 5 ^m.

et = epipodial tentacle; pc = pluricuspid tooth; r = rachidian tooth; to = tentacular opening; 1 = first lateral tooth;
2 = second lateral tooth; 3 = third lateral tooth; 4 = fourth lateral tooth.

have a tridentate cusp and bifurcate base. This is
consistent with figure 56 here, in which a more
extended base of the rachidian is revealed, but one
that would have been obscured in the optical mi-
croscopic preparation available to Dali. Dali re-
ported seven to eight cusps on the pluricuspid, com-
pared to a main and two lateral cusps indicated in
figure 54. However, Dali’s preparation may have
been worn and the actual cusp count unclear, as in
figure 55. This possible discrepancy and the fact
that our material showed no gill (due perhaps to
poor preservation) casts some doubt on our con-

Contributions in Science, Number 453

elusion that the present material represents Dali’s
species. However, the profile view of our illustrated
specimen (figure 51) is a good match for the first
5-mm shell length in profile view of the supposed
8-mm specimen figured by Dali.

Fedikovella beanii differs from F. caymanensis
in not having the apex overhang the posterior mar-
gin of the shell, as specified, but not illustrated by
Moskalev. The depth range of F. beanii is consis-
tent with the shelf and slope depths of other coc-
culinids, unlike the abyssal depth reported for the
type species.

McLean and Harasewych: Cocculiniform Limpets ■ 21

Figures 66-69. Notocrater youngi, new species. 66. Dorsal, lateral, and ventral views of holotype. Scale bar =1.0
mm. 67, 68. Dorsal and lateral views of protoconch of holotype. Scale bars = 50 iim. 69. Detail of protoconch sculpture
in figure 68. Scale bar = 5 nm.

Family Pseudococculinidae
Hickman, 1983

DIAGNOSIS. Apical fold of protoconch long
and narrow; protoconch sculpture of pustulose
crystals or anastomosing threads. Radula strongly
asymmetrical, inner lateral teeth 4.

REMARKS. Hickman (1983) was the first to ar-
gue that the radula of Pseudococculina Schepman,
1908, was sufficiently different from that of Coc-
culina to place it in a separate family. Marshall
(1986) further defined the family on shell and rad-
ular characters and Haszprunar (1988a, 1988b)
treated genera and relationships based on study of
anatomy. Genera in Pseudococculinidae have a~
greater range of expression of protoconch sculp-

ture, teleoconch sculpture, and radular and gill con-
ditions than Cocculinidae. The right cephalic ten-
tacle is modified and usually enlarged to function
as the copulatory organ. Gill structures are second-
ary and, if present, are positioned in the pallial
groove on either side of the foot.

The pseudococculinid radula differs from that of
Cocculinidae in having the first lateral tooth large
and triangular, projecting laterally, followed by three
laterals. All four teeth are here numbered 1-4 (fig-
ure 64, 1-4), following Marshall (1986). The large
multicusped tooth is again called the pluricuspid
(figure 64, pc). There is a lateromarginal plate that
is seldom revealed in SEM views (Marshall, 1986).
Because it is not depicted in our illustrations, it is
not mentioned further in our treatment of the pseu-
dococculinid genera.

22 ■ Contributions in Science, Number 453

McLean and Harasewych: Cocculiniform Limpets

Figures 70-75. Notocrater youngi, new species. 70. Ventral view of critical-point-dried animal of holotype. Scale bar
= 250 Atm- 71. Detail of mouth, with cuticular lining. Scale bar = 50 ixm. 72. Right cephalic tentacle. Scale bar = 50
Atm. 73. Dorsal view of radular ribbon. Scale bar — 20 Aim. 74. Detail of pluricuspid and marginal teeth. Scale bar = 5
Atm. 75. Lateral view of rachidian, lateral, and pluricuspid teeth. Scale bar = 5 Atm.

With the exception of the genus Notocrater and
Kaiparapelta and most species of Tentaoculus, the
genera of Pseudococculinidae tend to occur in
deeper water than the Cocculinidae. Many of the
genera are known only from abyssal or hadal depths.

Thirteen genera are now recognized: Pseudococ-
culina Schepman, 1908, Notocrater Finlay, 1926,
Kaiparapelta Marshall, 1986, Kurilabyssia Mos-
kalev, 1976, Caymanabyssia Moskalev, 1976 (and
subgenus Dictyabyssia McLean, 1991), Bandabys-
sia Moskalev, 1976, Tentaoculus Moskalev, 1976,
Mesopelex Marshall, 1986, Colotrachelus Mar-
shall, 1986, Yaquinabyssia Haszprunar, 1988, Co-

Contributions in Science, Number 453

pulabyssia Haszprunar, 1988, Amphiplica Ha-
szprunar, 1988 (with subgenus Gordabyssia Mc-
Lean, 1991), and Punctabyssia McLean, 1991.

The monotypic genus Punctabyssia McLean,
1991, is represented by P. tibbettsi McLean, 1991,
from the eastern Pacific, and the monotypic genus
Yaquinabyssia Haszprunar, 1988, is represented in
the eastern Pacific by Y. careyi McLean, 1988.

The following genera are unknown in either the
western Atlantic or eastern Pacific: Bandabyssia
Moskalev, 1976, Colotrachelus Marshall, 1986,
Kurilabyssia Moskalev, 1976, Mesopelex Mar-
shall, 1986, and Pseudococculina Schepman, 1908.

McLean and Harasewych: Cocculiniform Limpets ■ 23

Figures 76-78. Tentaoculus eritmeta (Verrill, 1884). 76. Dorsal, lateral, and ventral views of holotype. Scale bar =
1.0 mm. 77. Detail of surface sculpture. Scale bar = 100 nm. 78. Detail of interior septum. Scale bar = 100 /xm.

Genus Notocrater Finlay, 1926

Type species by original designation Cocculina cra-
ticulata Suter, 1908 (New Zealand). Synonym:
Punctolepeta Habe, 1958.

DIAGNOSIS. Protoconch sculpture of fine anas-
tomosing threads; teleoconch sculpture of concen-
tric ribs and strong pustules in curving rows. Eyes
present, right cephalic tentacle serving as copula-
tory organ. Inner marginal teeth enlarged, second
the largest in each row.

REMARKS. Notocrater is well represented in
Australasia (Marshall, 1986) and Japan. It is re-
ported living in the New World for the first time
with the description of the following two new spe-
cies. Marshall (1986:526) noted that Cocculina pus -
tulata Woodring, 1928, from the Jamaican Mio-
cene is a Notocrater. Woodring’s illustration
(Woodring, 1928:449, pi. 38, figs. 22, 23) does not

24 ■ Contributions in Science, Number 453

show the pustules, which were described as “min-
ute, closely spaced beads or pustules arranged along
intersecting curved radial lines.”

This genus is an exception among pseudococ-
culinids in living at continental slope depths, a hab-
itat that correlates with the presence of eyes.

Notocrater houbricki, new species

Figures 58-65

DESCRIPTION. Shell (figure 58) small for family
(maximum length 2.6 mm), thin, not eroded, white,
periostracum thin. Shell height moderate, that of
holotype 0.31 times length. Anterior slope convex,
posterior slope concave, lateral slopes slightly con-
vex to straight. Outline in dorsal view elongate-
oval, anterior end slightly narrower than posterior
end; aperture planar, ends not raised relative to
sides of shell. Apex posterior to center, at about 3A
shell length from anterior margin, protoconch be-

McLean and Harasewych: Cocculiniform Limpets

Figures 79-82. Tentaoculus georgiana (Dali, 1927). 79. Dorsal, lateral, and ventral views of lectotype. Scale bar =
1.0 mm. 80, 81. Dorsal and lateral views of protoconch of lectotype. Scale bars = 50 iim. 82. Detail of interior septum.
Scale bar = 100 /xm.

low highest point of shell, extending posteriorly.
Protoconch length 170 jum, protoconch sculpture
of low, densely scattered crystals (figures 59, 60).
Tip of protoconch immersed in posterior slope of
shell. Early sculpture of raised concentric ribs and
fine radial striae. Mature sculpture of elongate pus-
tules on evenly spaced concentric ribs, arranged in
curving rows. Shell edge thin and sharp. Muscle
scar not well marked.

Dimensions. Length 2.6, width 1.5, height 0.8
mm (holotype); the paratype shell is broken.

External Anatomy (figure 61). Eyes large, black;
right cephalic tentacle (copulatory organ) larger than
left, with small lobe (figure 61, to) marking tentac-
ular opening; two posterior epipodial tentacles, both
clearly bifid (figure 61, et). Foot and mantle edge
with minute dark brown spots.

Radula (figures 62-65). Rachidian uncusped,

Contributions in Science, Number 453

tooth rows asymmetrical, right skewed; first lateral
uncusped, shaft large, with laterally projecting el-
bow, second, third, and fourth lateral with bowed
shafts and single cusps. Pluricuspid with broad
overhang and three large, blunt cusps; first three
marginals larger than remaining marginals, second
the largest (figure 65).

TYPE LOCALITY. South of Settlement Point,
Grand Bahama Island, Bahama Islands (26°37'3T'N,
78°58'56"W), 412 m, on pencil-sized piece of wood
along with Cocculina messingi, new species.

TYPE MATERIAL. Two specimens from type
locality, collected by Charles Messing using deep-
submersible Johnson Sea Link II, sta. 2335, 18 May
1992. Holotype USNM 860359, 1 paratype LACM
2737.

REMARKS. Comparisons to N. youngi, new
species, are given under that species.

McLean and Harasewych: Cocculiniform Limpets ■ 25

Figures 83-86. Kaiparapelta askewi, new species. 83. Dorsal, lateral, and ventral views of holotype. Scale bar = 1.0
mm. 84, 85. Dorsal and lateral views of protoconch of holotype. Scale bars = 50 jiim. 86. Detail of protoconch sculpture.
Scale bar = 5 /xm.

ETYMOLOGY. We take pleasure in naming this
striking species after our late friend and colleague
Richard S. (Joe) Houbrick of the Division of Mol-
lusks, National Museum of Natural History.

Notocrater youngi, new species

Figures 66-75

DESCRIPTION. Shell (figure 66) small for family
(maximum length 3.1 mm), thin, not eroded, white,
periostracum thin. Shell height moderate, that of
holotype 0.32 times length. Anterior slope convex,
posterior slope straight, lateral slopes slightly con-
vex to straight. Outline in dorsal view elongate-
oval, anterior end slightly narrower than posterior
end; aperture not planar, sides raised relative to
ends of shell. Apex posterior to center, at about 2A
shell length from anterior margin, protoconch be-
low highest point of shell, extending posteriorly.
Protoconch length 190 jum, protoconch sculpture
of clumped crystals, some forming anastomosing

26 ■ Contributions in Science, Number 453

threads (figures 67-69). Tip of protoconch im-
mersed in posterior slope of shell. Early sculpture
of raised concentric ribs and fine radial striae. Ma-
ture sculpture of elongate pustules on evenly spaced
concentric ribs arranged in curving rows. Shell edge
thin and sharp. Muscle scar not well marked.

Dimensions. Length 3.1, width 2.2, height 1.0
mm (holotype).

External Anatomy (figures 70-72). Cephalic lap-
pets broad, epipodial tentacles two posterior pairs.
Cephalic tentacles with band of cilia and scattered
sensory papillae (figure 72). Mouth with cuticular
lining (figure 71, cl).

Radula (figures 73-75). Rachidian broad, un-
cusped; first lateral triangular, with projecting el-
bow; second, third, and fourth lateral with elbow
and single large overhanging cusp and up to four
serrations on medial side; pluricuspid with four,
similar-sized cusps below the overhang; inner mar-
ginals enlarged compared to outer marginals, sec-
ond marginal the largest.

McLean and Harasewych: Cocculiniform Limpets

Figure 87. Copulabyssia leptalea (Verrill, 1884). Dorsal, lateral, and ventral views of shell (USNM 757345). NE of
Norfolk, Virginia, in 3080-3090 m, R/V Gillis Cruise 75-08, sta. 36. Scale bar =1.0 mm.

TYPE LOCALITY. Off Southwest Reef, New
Providence Island, Bahamas (24°54'04"N,
77°33'14"W), 518 m on palmetto fronds.

TYPE MATERIAL. 17 specimens retrieved from
palm fronds and grass mat collectors placed on
bottom for larval settlement experiments by R. Em-
son, C.M. Young, and P.A. Tyler. Holotype, USNM
860360, Johnson-Sea-Link II, dive 2317, 9 May
1992. 16 paratypes, from same locality and depth
retrieved by Mary Rice, Johnson-Sea-Link I, dive
3463, 10 May 1993; 8 paratypes USNM 860387; 5
paratypes LACM 2738; 3 paratypes HBOM 065:
03885.

REMARKS. On shell characters this species dif-
fers from Notocrater boubricki in its larger size
(length 3.1 compared to 2.6 mm), more anteriorly
located apex, larger and more widely spaced pus-
tules on the protoconch, and less projecting nodes
on the teleoconch.

ETYMOLOGY. This species is named after Craig
M. Young, of Harbor Branch Oceanographic In-
stitution, Fort Pierce, Florida, whose research on
larval recruitment brought this species to light.

Genus Tentaoculus Moskalev, 1976

Type species by monotypy: Tentaoculus perlucida

Moskalev, 1976; New Guinea.

DIAGNOSIS. Shell with or without small septum
near apex on inner surface. Protoconch sculpture
of fine pits and anastomosing ridges; teleoconch
sculpture of fine concentric growth lines and radial
striae.

REMARKS. Marshall (1986:67) used the genus
Tentaoculus for three species from New Zealand,
noting that the genus is unique in having an interior

shell septum. This character provides a clue to the
allocation of an enigmatic northwestern Atlantic
species described originally in the family Fissurel-
lidae by Verrill, 1884, as well as that of a previously
unfigured species described by Dali in Cocculina.

Tentaoculus eritmeta
(Verrill, 1884),
new combination

Figures 76-78

Puncturella ( Fissurisepta ) eritmeta Verrill, 1884:204,
pi. 32, fig. 19; Clarke, 1962:8 [checklist only];
R.I. Johnson, 1989:37 [type specimen].
Puncturella eritmeta ; Pilsbry, 1890:238, pi. 27, figs.
60, 61 [copy of Verrill]; Thiele, 1919:156, pi. 17,
figs. 14, 15 [copy of Verrill].

REMARKS. The original illustration, which has
been copied by Pilsbry (1890) and Thiele (1919),
includes a posterior view with two circular scars
separated by what was intended to represent a sep-
tum. In our interpretation, the upper scar is that
left by loss of the protoconch aperture and the
lower shows the trace of the protoconch tip where
it was engulfed by the posterior slope of the shell.
Verrill (1884) described but did not illustrate an
interior septum: “in the apex there is a minute trans-
verse lamina, forming a small flattened tube.” Shell
proportions and sculpture are consistent with those
of the Tentaoculus species illustrated by Marshall
(1986). The flattened tube described by Verrill is
shown in an enlarged view with SEM, tilted to show
its length (figure 78). This species, which has been
ignored by all recent compilers of the Fissurellidae,
is now added to the list of western Atlantic pseu-
dococculinids.

Contributions in Science, Number 453

McLean and Harasewych: Cocculiniform Limpets ■ 27

Tentaoculus georgiana
(Dali, 1927),
new combination

Figures 79-82

Cocculina georgiana Dali, 1927:121; C.W. John-
son, 1934:66 [listed only]; Abbott, 1976:35 [listed
only].

A lectotype (USNM 108281) of this previously un-
illustrated species is designated and illustrated here
(length 2.5, width 1 .7, height 1 .2 mm). Twenty para-
lectotypes have been recataloged as USNM 860384;
2 paralectotypes LACM 2739. The original type
locality is “off Georgia,” but the printed label reads
“U. S. Fish Com. sta 2415, 440 fm. [805 m], off
Georgia.” Although not mentioned originally by
Dali, it proves to have an interior septum (figure
82) comparable to that of Tentaoculus eritmeta.
The protoconch sculpture is too worn to show the
pits near the terminus, but the anastomosing sculp-
ture illustrated by Marshall (1986:534, fig. 6M) for
T. haptricola Marshall, 1986, from New Zealand
is present. Like the previous species, this species
may safely be allocated to Tentaoculus in the ab-
sence of soft parts. It differs from T. eritmeta in
having a much higher profile and the apex at 3A the
shell length from the anterior, rather than nearly
central.

Genus Caymanabyssia Moskalev, 1986

Type species by original designation: Caymana-
byssia spina Moskalev, 1976.

DIAGNOSIS. Protoconch with columnar prisms;
teleoconch sculpture of sharp pustules on anasto-
mosing network of surface sculpture; central and
lateral teeth degenerate, lacking cusps; right ce-
phalic tentacle enlarged, open seminal groove; gill
leaflets on both sides.

REMARKS. Although the type species was poor-
ly figured, the teleoconch sculpture of strong pus-
tules superimposed on an anastomosing network is
unmistakable. This genus has been used for the
New Zealand species Caymanabyssia rhina Mar-
shall, 1986, and for the eastern Pacific species Cay-
manabyssia vandoverae McLean, 1991. Haszpru-
nar’s (1988:174) anatomical definition of Caman-
abyssia is based on the species C. sinespina Mar-
shall, 1986, which was designated the type species
of the subgenus Dictyabyssia McLean, 1991.

Caymanabyssia spina Moskalev, 1976

Caymanabyssia spina Moskalev, 1976:65, fig. 4,
pi. 2, figs. 7, 8; Marshall, 1986:538.

Moskalev’s original description was translated from
Russian by G.V. Shkurkin in a privately circulated
“reprint” dated December 1978. That translation
is repeated here, altered to place it in telegraphic
style and with terminology slightly revised to con-
form to that used throughout this paper:

Shell small, low, thin, apex at 2A shell length from
anterior margin, protoconch lacking sculpture. An-

28 ■ Contributions in Science, Number 453

terior slope slightly convex, posterior slope straight;
aperture elliptical, margin made deticulate from
projecting surface sculpture. Surface yellowish-
white, semi-transparent. Sculpture of numerous,
similar conelike thorns in regular checker-board
pattern, spaced not less than the diameter of the
base of each thorn. Shell interior white, surface
sculpture visible from inside.

Shell lengths 0.85-2.95 mm, holotype the largest.

Epipodial tentacles clearly visible, a differentia-
tion of the right cephalic tentacle noticeable on 34
specimens and eggs present in 31 specimens.

Rachidian rounded, thickened in the middle by
a horizontal ridge. Subcentral teeth diverging wing-
like from the rachidian, irregularly triangular, with
folds on outer edge. First, second, and third lateral
teeth almost identical, cusps lacking, boomerang
shaped. Fourth lateral tooth [pluricuspid] slightly
smaller than subcentral one, of complex shape with
folds. Fifth lateral tooth [lateromarginal plate]
smallest, its middle part situated at level of lower
margin of fourth lateral tooth.

TYPE LOCALITY. Eastern part of Cayman
Trough (19°38'5"N, 76°38'8"W), 6740-6800 m. A
total of 204 specimens on wood, 7 on other sub-
strates of vegetal origin, collected with research
vessel Akademic Kurchatov, cruise 14, sta. 1267,
2.5 m Sigsby dredge, 24-25 March 1967. Further
locality: western part of Cayman Trough (19°00'6"N,
80°29'5"W), 6800 m, sta. 1242A, 20 March 1967,
2 specimens on wood. In the type series, shells are
broken and bodies deformed in 29 specimens; 11
were used for radular mounts.

REMARKS. Moskalev did not mention the anas-
tomosing background sculpture that occurs be-
tween the sharp pustules, although it is recognizable
in his illustration.

Genus Kaiparapelta Marshall, 1986

Type species by original designation: Kaiparapelta

singularis Marshall, 1986.

DIAGNOSIS. Protoconch sculpture of anasto-
mosing threads, teleoconch sculpture granulate,
profile low.

REMARKS. The new species described here is
the second known species of the genus; the type
species of Kaiparapelta occurs in the New Zealand
Miocene. The genus is still based on shell charac-
ters. For the generic allocation, we are indebted to
A. Waren, who has recognized living material of
this genus from the eastern Atlantic and will report
on the radula and anatomy in a forthcoming paper.

On shell characters the genus differs from No-
tocrater in its lower profile. It also resembles Cay-
manabyssia ( Dictyabyssia ), which has a more oval
and more regular outline. The latter genus has been
reported only from abyssal depths.

Kaiparapelta askewi, new species

Figures 83-86

DESCRIPTION. Shell (figure 83) small for family
McLean and Harasewych: Cocculiniform Limpets

(maximum length 2.7 mm), thin, not eroded, white,
periostracum thin. Shell height moderate, that of
holotype 0.38 times length. Anterior, posterior, and
lateral slopes nearly straight. Outline in dorsal view
broadly oval, margin irregularly undulating; ante-
rior end slightly narrower than posterior end; ap-
erture not planar, ends raised relative to sides of
shell. Apex slightly posterior to center, protoconch
nearly at highest point of shell, extending posteri-
orly. Protoconch length 185 /im, protoconch sculp-
ture of clumped raised threads (figure 86). Tip of
protoconch immersed in posterior slope of shell.
Sculpture of weak concentric growth lines and ir-
regular anastomosing threads or low pustules, pro-
ducing a shagreen surface. Posterior slope with two
faintly indicated, raised ridges terminating at slight
indentations at margin. Anastomosing sculpture
more prominent than radial sculpture, not raised at
intersections with radial striae. Shell edge thin and
sharp. Muscle scar and pallial attachment scar not
well marked.

Dimensions. Length 2.65, width 2.55, height 1.0
mm (holotype); length 3.2, width 2.45, height 1.0
mm (paratype).

Radula and External Anatomy. Unknown.

TYPE LOCALITY. 165 km E of Charleston,
South Carolina (32°43.68'N, 78°05.72'W), 194 m.
This is the locality known as the “Charleston
Lumps,” which is also the type locality of two re-
cently described pleurotomariid gastropods, Pero-
troches charlestonensis Askew, 1988, and P. maur-
eri Harasewych & Askew, 1993.

TYPE MATERIAL. Two dead specimens in sed-
iment sample collected at the type locality by T.M.
Askew and M.G. Harasewych with the research
submersible Clelia, sta. 78, 6 July 1993. Holotype
USNM 860362, paratype LACM 2740.

REMARKS. Shell profile, protoconch sculpture
and teleoconch sculpture agree with that of the type
species. The two posterior ridges and the corre-
sponding indentations at the margin are unique to
this species. The indentations are probably indic-
ative of the position of the posterior epipodial ten-
tacles.

ETYMOLOGY. We are pleased to name this
species after Timothy M. Askew, Director of Ma-
rine Operations, Harbor Branch Oceanographic In-
stitution, Fort Pierce, Florida.

Genus Copulabyssia
Haszprunar, 1988

Type species by original designation: Cocculina

corrugata Jeffreys, 1883.

DIAGNOSIS. Protoconch sculpture of prismatic
crystals, teleoconch sculpture of raised concentric
rings and fine radial striae. Apex below highest point
of shell. Right cephalic tentacle exceptionally large.

REMARKS. Haszprunar (1988a) detailed the
anatomy of the type species and provided SEM
illustrations of the shell, protoconch, and radula of
a Mediterranean specimen identified as the type

species. Dantart and Luque (1994:290) also de-
scribed and illustrated the type species. Waren (1991:
80) noted that the northwestern Atlantic Cocculina
leptalea Verrill, 1884, is also referable to Copula-
byssia on the basis of shell characters, but he did
not illustrate specimens.

Copulabyssia leptalea
(Verrill, 1884)

Figures 87-91

Cocculina leptalea Verrill, 1884:202, pi. 32, figs.
20, 20a, 20b; Pilsbry, 1890:133, pi. 25, figs. 7, 8;
Thiele, 1909:7, pi. 2, fig. 5; C.W. Johnson, 1934:
66 [checklist only]; Abbott, 1974:34 [checklist
only]; R.I. Johnson, 1989:46 [citation of type ma-
terial].

Copulabyssia leptalea; Waren, 1991:80.

REMARKS. As noted by Waren (1991:80), the
holotype (USNM 38079, from USFC sta. 2038, 3700
m off Delaware) is now fragmented. We illustrate
(figure 87) a specimen (USNM 7577345) dredged
off Norfolk, Virginia, 3080-3090 m (length 2.45,
width 1.85, height 0.9 mm). As in the holotype, the
early sculpture is eroded and the protoconch is
replaced by an infilled plug that cannot be com-
pared to that illustrated for the type species by
Haszprunar (1988a). Its radula (figures 88-91) differs
from that of the type species as illustrated by Ha-
szprunar (1988a:fig. ID, E) in having the rachidian
with a single, weakly projecting cusp, rather than
lacking any cusp. The pluricuspid teeth of Copu-
labyssia leptalea have four nearly equal cusps,
whereas those of C. corrugata have two cusps, as
illustrated by Dantart and Luque (1994:fig. 60).

Copulabyssia leptalea differs from C. corrugata
in radular characters as already noted and in having
a more anterior apex.

Genus Amphiplica Haszprunar, 1988

Type species by original designation: Amphiplica
venezuelensis McLean, 1988. Subgenus (or syn-
onym) Gordabyssia McLean, 1991; type species
by original designation Amphiplica ( Gordabys-
sia) gordensis McLean, 1991.

DIAGNOSIS. Shell large, profile low; proto-
conch sculpture of dense net pattern arranged in
longitudinal rows.

REMARKS. Protoconch sculpture was un-
known prior to discovery of the eastern Pacific spe-
cies Amphiplica ( Gordabyssia ) gordensis McLean,
1991. There is a single species in the western At-
lantic.

Amphiplica venezuelensis
McLean, 1988

Amphiplica venezuelensis McLean, 1988:155, figs.
1-7.

REMARKS. This species, from 5057 m in the
Venezuela Basin, attains a maximum length of 14.8

Contributions in Science, Number 453

McLean and Harasewych: Cocculiniform Limpets ■ 29

Figures 88-91. Copulabyssia leptalea (Verrill, 1884). Radula from specimen in figure 87. 88. Dorsal view of radular
ribbon. Scale bar = 25 ^m. 89. Detail of rachidian and lateral teeth. Scale bar = 100 n m. 90. Lateral view of longitudinally
cut radula revealing relative heights of tooth fields. Scale bar = 100 jum. 91. Detail of distal ends of inner and outer
marginal teeth. Scale bar = 5 /urn.

mm and is the largest known pseudococculinid. Its
anatomy was treated by Haszprunar (1988a).

REALLOCATED TAX A

The following taxa were initially described in the
genus Cocculina but are now considered to be
members of families other than Cocculinidae or
Pseudococculinidae. Although most can be allo-
cated to family and genus, their status as species or
synonyms remains to be resolved, pending revision
of the genera. These taxa are arranged in their order
of description.

“ Cocculina ” conica Verrill, 1884

Cocculina conica Verrill, 1884:204; Pilsbry, 1890:
134 [copy Verrill description]; Thiele, 1909:7
[German translation]; C.W. Johnson, 1934:66
[listed only]; Abbott, 1974:34 [listed only]; R.I.
Johnson, 1989:30.

Pilus conica ; Waren, 1993:80, fig. 20A-E.
Although the holotype (USNM 38441) of this orig-

30 ■ Contributions in Science, Number 453

inally unfigured species from 499 fathoms off Nova
Scotia was reported lost by Johnson (1989), Waren
(1993) recognized the species from the detailed
original description and illustrated shell specimens
from deep water off southwestern Iceland. The shell
is less than 1 mm in length and has a posteriorly
overhung apex. This species is the type species for
Waren’s genus Pilus. The radula is unknown, and
the family allocation is therefore uncertain within
the Cocculiniformia.

“ Cocculina ” dalli Verrill, 1884

Cocculina dalli Verrill, 1884:203; C.W. Johnson,
1934:66 [listed only]; Abbott, 1974:34 [listed
only]; R.I. Johnson, 1989:32, pi. 10, fig. 10 [ho-
lotype].

The single known specimen (holotype USNM
38081), from 580 m (39°53'N, 69°47'W) was first
illustrated by Johnson (1989). It is here considered
to be close to the North Atlantic lothia rugosa
(Jeffreys, 1883), family Lepetidae.

McLean and Harasewych: Cocculiniform Limpets

Figures 92-95. Propilidium lissocona (Dali, 1927). 92. Dorsal, lateral, and ventral views of lectotype. Scale bar =1.0
mm. 93, 94. Dorsal and lateral views of the protoconch of the lectotype. Scale bars = 50 /urn. 95. Detail of protoconch
sculpture. Scale bar = 5 n m.

“ Cocculina ” reticulata Verrill, 1885

Cocculina reticulata Verrill, 1885:426; Verrill in
Bush, 1893:240, pi. 2, fig. 6; C.W. Johnson, 1934:
66 [listed only]; Abbott, 1974:34 [listed only];
R. I. Johnson, 1989:62 [citation of holotype only].

The holotype (USNM 44832) from 128 m off Che-
sepeake Bay is here referred to the genus Propili-
dium, family Lepetidae.

“ Cocculina ” lissocona Dali, 1927
Figures 92-95

Cocculina lissocona Dali, 1927:110; C.W. Johnson,
1934:66 [listed only]; Abbott, 1974:35 [listed
only].

We designate and illustrate (figures 92-95) a lec-
totype (USNM 333472, USFC 2668, 538 m off Fer-

nandina, Florida). Fourteen remaining paralecto-
types have been recataloged USNM 860385; two
paralectotypes LACM 2741. This is also referred
to Propilidium, family Lepetidae. It is characterized
by its high profile, radial and concentric sculpture
producing beads at intersections, smooth proto-
conch, and weak interior septum. The septum (fig-
ure 92) is characteristic of Propilidium, being short-
er than that of Tentaoculus.

Marshall (1985:541) illustrated a Tasmanian spe-
cies of Propilidium and gave further notes on the
genus. Propilidium exiguum (Thompson, 1844), the
type species of Propilidium, was treated by Dantart
and Luque (1994:303). Propilidium lissocona and
P. reticulata are indistinguishable from the type
species on shell characters. We suspect that the type
species occurs broadly in the North Atlantic and
that the two western Atlantic taxa should be added
to the synonymy of the type species.

Contributions in Science, Number 453

McLean and Harasewych: Cocculiniform Limpets ■ 31

“ Cocculina ” rotunda Dali, 1927

Cocculina ? rotunda Dali, 1927:115, 121; C.W.
Johnson, 1934:66 [listed only]; Abbott, 1974:35
[listed only].

Type material (holotype USNM 108156), from off
Fernandina, Florida, has a circular outline and a
high, centrally positioned apex. It is here tentatively
referred to the genus Bathysciadium Dautzenberg
& Fischer, 1901 (family Bathysciadiidae).

“ Cocculina ” superba Clarke, 1960

Cocculina superba Clarke, 1960:1, fig. 1.

Clarke illustrated the animal of this abyssal species
from Argentina. Although a radular preparation was
not made, characters of the shell and external anat-
omy are suggestive of the family Lepetidae.

ACKNOWLEDGMENTS

We are grateful to our colleagues who have provided the
specimens on which this study is based. Drs. Craig Young
(Harbor Branch Oceanographic Foundation, Fort Pierce,
Florida), Roland Emson (Kings, College, London), Paul
A. Tyler (University of Southampton, United Kingdom),
and Mary Rice (Smithsonian Marine Station at Link Port,
Port Pierce, Florida) supplied samples of limpets collected
from experimental substrates deployed off New Provi-
dence Island, Bahamas. Their work was supported by NSF
OCE-91 16560 to C.M. Young and NATO Collaborative
Research Grant CRG-900628 to C.M. Young, P.A. Tyler,
and R.H. Emson. Dr. Charles Messing (Nova University,
Dania, Florida) provided samples from his study site at
Grand Bahama Island, Bahamas (NSF EAR-9004232). Paula
Mikkelsen brought to our attention specimens in the col-
lections of the Harbor Branch Oceanographic Museum.
We thank Ms. Susann Braden of the National Museum
of Natural History for her assistance with the scanning
electron microscopy. Alan Kabat of the same institution
helped to verify the status of specimens described by W.
H. Dali. We much appreciate the reviews and helpful
suggestions provided by Bruce Marshall, Winston Ponder,
and Anders Waren.

This is Smithsonian Marine Station at Link Port Con-
tribution No. 352.

LITERATURE CITED

Abbott, R.T. 1974. American seashells, 2nd ed. New
York: Van Nostrand Reinhold, 663 pp.

Askew, T.M. 1988. A new species of pluerotomariid
gastropod from the western Atlantic. The Nautilus
102:89-91.

Clarke, A.H., Jr. 1960. A giant ultra-abyssal Cocculina
(Mollusca, Gastropoda), from the Argentine Basin.
National Museum of Canada, Natural History Pa-
pers 7:1-4.

. 1962. Annotated list and bibliography of the

abyssal marine molluscs of the world. National Mu-
seum of Canada, Bulletin 181:1-114.

Dali, W.H. 1882. On certain limpets and chitons from
the deep waters off the eastern coast of the United
States. Proceedings of the United States National
Museum 4:400-414.

. 1889. Reports on the results of dredging, under

the supervision of Alexander Agassiz, in the Gulf of
Mexico (1877-78) and the Caribbean Sea (1879-80),

32 ■ Contributions in Science, Number 453

by the U. S. Coast Survey Steamer “Blake,”. . .29:
Report on the Mollusca, part 2: Gastropoda and
Scaphopoda. Bulletin of the Museum of Compar-
ative Zoology, Harvard University 18:1-492.

. 1890. Scientific results of explorations by the

U.S. Fish Commission steamer Alabatross, 7: Prelim-
inary report of the collection of Mollusca and Brach-
iopoda obtained in 1887-88. Proceedings of the
United States National Museum 12:219-362.

. 1908. [Reports from ‘Albatross’ dredging op-
erations]. The Mollusca and the Brachiopoda. Bul-
letin of the Museum of Comparative Zoology at
Harvard College 43:205-487.

. 1927. Small shells from dredgings off the south-
east coast of the United States by the United States
Fisheries steamer “Albatross” in 1885 and 1886.
Proceedings of the United States National Museum
70:1-134.

Dali, W.H., and T. Simpson. 1901. The Mollusca of
Porto Rico. United States Fish Commission Bulletin
for 1900, vol. 1, 351-524.

Dantart, L., and A. Luque. 1994. Cocculiniformia and
Lepetidae (Gastropoda: Archaeogastropoda) from
Iberian waters. Journal of Molluscan Studies 60:
277-313.

Harasewych, M.G., and M. Askew. 1993. Perotrochus
maureri, a new species of pleurotomariid from the
western Atlantic (Gastropoda: Pleurotomariidae). The
Nautilus 106:130-136.

Haszprunar, G. 1987. Anatomy and affinities of coc-
culinid limpets (Mollusca, Archaeogastropoda).
Zoologica Scripta 16:305-324.

. 1988a. Anatomy and affinities of pseudococ-

culinid limpets (Mollusca, Archaeogastropoda).
Zoologica Scripta 17:161-179.

. 1988b. Comparative anatomy of cocculiniform

gastropods and its bearing on archaeogastropod sys-
tematics. Malacological Review, supplement 4, 64-
84.

Hickman, C.S. 1983. Radular patterns, systematics, di-
versity, and ecology of deep-sea limpets. The Veliger
26:73-92.

Jeffreys, J.G. 1883. On the Mollusca procured during
the cruise of H.M.S. Triton, between the Hebrides
and Faroes in 1882. Proceedings of the Zoological
Society of London 1883:389-399.

Johnson, C.W. 1934. List of marine Mollusca of the
Atlantic coast from Labrador to Texas. Proceedings
of the Boston Society of Natural History 40:1-204.

Johnson, R.I. 1989. Molluscan taxa of Addison Emery
Verrill and Katharine Jeannette Bush, including those
introduced by Sanderson Smith and Alpheus Hyatt
Verrill. Occasional Papers on Mollusks, Museum
of Comparative Zoology, Harvard University 5:1-
143.

Marshall, B.A. 1986. Recent and Tertiary Cocculinidae
and Pseudococculinidae (Mollusca: Gastropoda)
from New Zealand and New South Wales. New
Zealand Journal of Zoology 12:505-546.

McLean, J.H. 1987. Taxonomic descriptions of coc-
culinid limpets (Mollusca, Archaeogastropoda): Two
new species and three rediscovered species. Zoolo-
gica Scripta 16:325-333.

— -. 1988. Three new limpets of the family Pseu-

dococculinidae from abyssal depths (Mollusca, Ar-
chaeogastropoda). Zoologica Scripta 17:155-160.

-. 1991. Four new pseudococculinid limpets col-
lected by the deep-submersible Alvin in the eastern
Pacific. The Veliger 34:38-47.

McLean and Harasewych: Cocculiniform Limpets

. 1992. Cocculiniform limpets (Cocculinidae and

Pyropeltidae) living on whale bone in the deep sea
off California .Journal of Molluscan Studies 58:401-
414.

Moskalev, L.I. 1976. On the generic classification in
Cocculinidae (Gastropoda, Prosobranchia). Works
of the P. P. Shirshow Institute of Oceanology, The
Academy of Sciences of the USSR 99:59-70. [In
Russian.] Translated for private circulation by G.V.
Shkurkin, 1978.

Pilsbry, H.A. 1890. Manual of conchology, vol. 12.
Academy of Natural Sciences, Philadelphia, 323 pp.

Thiele, J. 1909. Cocculinoidea und die Gattungen Phen-
acolepas und Titiscania. In Systematisches Con-
chylien-Cabinet von Martini & Chemnitz, ed. H.C.
Kuster and W. Kobelt, series 2, vol. 11a, no. 539,
pp. 1-48.

— . 1919. Familia Fissurellidae. In Systematisches

Conchy lien-Cabinet von Martini & Chemnitz, ed.
H.C. Kuster, W. Kobelt, and F. Haas, series 2, vol.
4a, no. 580, pp. 145-68.

Verrill, A.E. 1884. Second catalog of Mollusca, recently
added to the fauna of the New England coast and
the adjacent parts of the Atlantic, consisting mostly

of deep-sea species, with notes on others previously
recorded. T ransactions of the Connecticut Acade-
my 6:139-294.

. 1885. Third catalog of Mollusca, recently add-
ed to the fauna of the New England Coast and the
adjacent parts of the Atlantic, consisting mostly of
deep-sea species, with notes on others previously
recorded. Transactions of the Connecticut Acade-
my 6:395-452.

Waren, A. 1991. New and little known Mollusca from
Iceland and Scandinavia. Sarsia 76:53-124.

. 1993. New and little-known Mollusca from

Iceland and Scandinavia, part 2. Sarsia 78:159-201.

Watson, R.B. 1886. Report on the Scaphopoda and
Gasteropoda collected by H.M.S. Challenger during
the years 1873-1876. Report on the Scientific Re-
sults of the Voyage of the H.M.S. Challenger, 1873-
1876. Zoology 15(part 42):l-680, pis. 1-50.

Woodring, W.P. 1928. Miocene mollusks from Bowden,
Jamaica. Part II. Gastropods and discussion of re-
sults. Carnegie Institution of Washington, Publica-
tion no. 385, vii + 564 pp.

Received 29 March 1994; accepted 10 November 1994.

Contributions in Science, Number 453

McLean and Harasewych: Cocculiniform Limpets ■ 33

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

Number 454
27 July 1995

Contributions
in Science

Systematics of Nearctic Ants of the
Genus Dorymyrmex
(Hymenoptera: Formicidae)

Roy R. Snelling

Natural History Museum of Los Angeles County

Serial
Publications
of THE
Natural History
Museum of
Los Angeles
County

• Contributions in Science, a miscellaneous series of tech-
nical papers describing original research in the life and
earth sciences.

• Science Series, long articles and collections of papers
on natural history topics.

Copies of the publications in these series are sold through
the Museum Book Shop. A catalog is available on request.

Scientific

Publications

Committee

James L. Powell, Museum President
Daniel M. Cohen,
Committee Chairman
Brian V. Brown
Kenneth E. Campbell
Kirk Fitzhugh
Robin A. Simpson, Managing Editor

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

Printed at Allen Press, Inc., Lawrence, Kansas
ISSN 0459-8113

Systematics of Nearctic Ants of the
Genus Dorymyrmex
(Hymenoptera: Formicidae)

Roy R. Snelling1

ABSTRACT. The “neotype” for “ Conomyrma insana (Buckley)” (= Formica insana Buckley, 1866) and
the “lectotype” for “ Conomyrma flava McCook” (= Dorymyrmex insanus var. flavus McCook, 1879),
designated by Johnson (1989), are both shown to be invalid. Appropriate neotypes are designated for
both species; each is redescribed and illustrated.

Dorymyrmex pyr amicus var. smithi Cole, 1936, is removed from the synonymy of D. insanus and
recognized as a distinct species, of which Dorymyrmex medeis (Trager, 1988) is a newly recognized
synonym. Dorymyrmex bureni (Trager, 1988), treated by Johnson (1989) as a synonym of D. flavus , is
reinstated to species rank. Dorymyrmex reginicula (Trager, 1988), treated by Johnson (1989) as a synonym
of D. insanus, is reinstated to species rank. This species is a temporary social parasite in nests of other
Dorymyrmex species. Dorymyrmex wheeleri (Kusnezov, 1952), purported to be a synonym of D. insanus
by Snelling (1973), is reinstated to species rank but remains known only from the type specimens. Two
new species are described and illustrated: D. lipan (Texas) and D. paiute (Utah).

A key is given for the separation of the worker caste of the 12 described species known to occur in
the United States.

INTRODUCTION

No North American ant has been cursed with such
a singularly unfortunate taxonomic history as that
described by Buckley (1866) as Formica insana.
Buckley’s choice for a specific name has proved to
be both remarkably prophetic and apropos. The
original description was inadequate, even by the
standards of the time; no type specimens were des-
ignated, and none of Buckley’s original material is
known to have survived to the present. The identity
of Buckley’s species must, of necessity, rely on spec-
ulation. Much of the same is true of the ant de-
scribed a few years later by McCook (1879) as Do-
rymyrmex insanus var. flavus.

Recently, in an effort to clarify the status of these
names for two of the most common ants across
the southern United States, Johnson (1989) desig-
nated a neotype for F. insana and a lectotype for
D. insanus var. flavus. Regrettably, although the
intent is commendable, neither designation is valid,
as will be demonstrated below. It is my intent to
critique Johnson’s work and to correct his errors,
both of substance and of judgment, by designating
appropriate type specimens for both species.

1. Research Associate, Natural History Museum of
Los Angeles County, 900 Exposition Boulevard, Los An-
geles, California 90007.

Contributions in Science, Number 454, pp. 1-14
Natural History Museum of Los Angeles County, 1995

HISTORICAL RESUME

The original description of Formica insana by S.B.
Buckley (1866) is regrettably brief:

Worker. Length 0.14 inch — Color black or brown-
ish-black, smooth and shining throughout; head sub-
quadrate, the lateral margins slightly curved inwards;
below oval, and rounded above; eyes large, sub-
elliptical, and placed on the anterior portion of the
head near its lateral margins; antennae long, filiform;
mandibles large, curved inwards, the truncated api-
cal ends sharply toothed; head sub-channelled be-
neath, with the anterior and posterior parts deeply
depressed; prothorax about half the width of the
head, rounded above; mesothorax somewhat de-
pressed; metathorax has a rudimentary spine or sharp
protuberance on its upper posterior surface, pedicle
short, inserted near the base of the anterior part of
the abdomen; scale small, wedge-shaped, and in-
clined forwards; abdomen broad-ovate, subacute;
legs small, slender, and rather short.

Female. Length 0.20 inch — Head small, narrower
than the thorax; abdomen broad, oblong and ovate;
color black-bronze, with the margins of the seg-
ments of the abdomen hyaline; thorax raised above
the head; wings not seen; the remainder like the
worker.

Buckley stated that this species is common in cen-
tral Texas.

Edward Norton (1875) appears to have been the
first to recognize that Formica insana belonged to
the then-recently described genus Dorymyrmex
Mayr (1866); H.C. McCook (1879) included it in

ISSN 0459-8113

Dorymyrmex. McCook added little to the under-
standing of D. insanus , but his concept of that
name was based on two of Buckley’s specimens
that he had examined and were then in the collec-
tions of the Academy of Natural Sciences in Phil-
adelphia. Those specimens apparently disappeared
long ago, but McCook did provide figures of D.
insanus , although their accuracy in nearly all re-
spects must be considered suspect.

McCook’s brief discussion of D. insanus was
followed by a description of a new variety, D. in-
sanus var. flavus: “This variety is identical with
insanus , except in the color, which is a uniform
honey yellow, and the contour of the thorax. The
apex of the abdomen and the flagellum of the an-
tennae are tipped with a blackish hue. The variety
appears to be quite permanent, the distinction hold-
ing in a number (25 or 30) of specimens examined.
The cone is evidently higher than the thorax. There
is no tuft under the face.”

McCook stated that he had collected D. insanus
in Texas but made no mention of having found it
elsewhere. He further noted that Norton had ear-
lier suggested that D. insanus might be a synonym
of D. pyramicus, a Brazilian species described by
J. Roger (1863), originally in the genus Prenolepis.
Norton’s suggestion may have been prompted by
Mayr (1870), who recorded specimens from Mex-
ico as D. pyramicus. Of the distribution of D. in-
sanus var. flavus, McCook had little to say other
than that it occurs in the “Southern States” of the
United States.

McCook was apparently the last person to have
examined any of Buckley’s specimens. Formica in-
sana ceased to be a recognized entity when Mayr
(1886) placed it in synonymy with D. pyramicus,
where it was to remain for almost 90 years. This
has had some unfortunate consequences. Why Mayr
chose to put forward this synonymy is uncertain,
since it is clear that he had not seen type material
of either name, nor has any myrmecologist since
then. What is clear is that since Mayr’s time the
identity of D. pyramicus has been based not on
Roger’s type but on the characteristics of the North
American ant that has been treated as its synonym
and perhaps of other blackish or brownish species
as well. Thus, when Forel (1913) named “ Dory-
myrmex pyramicus ” as the type for his new genus
Conomyrma, the specimens that he had before him
were not Roger’s Brazilian ant but specimens from
North America that were assumed to be conspecific
with it. A strong case could be made for the view
that the true type species of Conomyrma should
be Formica insana and not Prenolepis pyramica.
That problem, however, is beyond the purview of
the present study.

Creighton (1950) believed that there was a single
North American species, Dorymyrmex pyramicus,
within which he recognized three infraspecific forms:
D. pyramicus pyramicus, D. p. bicolor W.M.
Wheeler (1906), and D. p. flavopectus M. Smith
(1944). The various species assigned to Dorymyr-

mex were next studied by Kusnezov (1952), who
divided Dorymyrmex into two genera: Dorymyr-
mex and Conomyrma. The North American forms
were all placed in the latter genus. In 1959 he again
dealt with these ants, and the genus Conomyrma
was split into two genera, with C. pyramica and
C. flavopecta remaining in Conomyrma; C. bicolor
and C. wheeleri Kusnezov (1952) were assigned to
a new genus, Biconomyrma.

The status of our forms was reviewed once more
by Snelling (1973). The primary purpose of that
study was to affirm that Conomyrma is a genus
apart from Dorymyrmex and that the North Amer-
ican forms all belong to Conomyrma. At the same
time, Biconomyrma was synonymized with Con-
omyrma.

The most important aspect of the paper, how-
ever, was that Buckley’s Formica insana was res-
urrected from the synonymy of C. pyramica to
replace the latter name in the North American fau-
na; C. pyramica was restricted to South America
and considered to be essentially unidentifiable. Three
North American species of Conomyrma were rec-
ognized: C. bicolor, C. flavopecta, and C. insana.
To the latter were assigned, as synonyms, flavus
McCook, nigra Pergande (1895), antiliana Forel
(1911), smithi Cole (1936), and wheeleri Kusnezov,
regrettably without having examined type material
of nearly all these names. However, had I done so,
there is no reason to believe that my conclusions
would have been different, for my views on these
ants have changed drastically since that time.

When contemplating the possibility of resusci-
tating Buckley’s name, Creighton (1950) had com-
mented that “this looks suspiciously like stepping
out of the frying pan into the fire.” These were
remarkably prophetic words. It is now clear, for
example, that nearly every name that I listed as a
synonym of C. insana almost certainly represents
a valid, separable species. This, however, is not the
most significant problem afflicting our understand-
ing of North American Dorymyrmex. At the core
of all past confusion is the problem of the true
identity of D. insanus. Despite all the changes out-
lined above, since W.M. Wheeler (1902), there has
been a clear consensus on one point: D. insanus is
a dark Texas species with a sharply angulate meso-
nota! profile. I (Snelling, 1973) echoed this in un-
equivocal terms when I asserted that this ant “has
a mesonotum which is sharply declivitous behind.”
As will be shown below, this interpretation is only
partly correct.

In the years following 1973 I became increasingly
uncomfortable with the notion that there was only
a single dark species of Dorymyrmex in North
America. However, since the late W.F. Buren had
expressed an interest in the genus, I did not pursue
the matter. Following Buren’s death, J.C. Trager
continued Buren’s study of Dorymyrmex and in
1988 published the results of his work on the spe-
cies of Florida (placed in the then-recognized genus
Conomyrma). The number of United States species

2 ■ Contributions in Science, Number 454

Snelling: Nearctic Dorymyrmex

was increased from three to nine. In addition to
the three already recognized by Snelling (1973), five
new species were described; one additional species,

C. grandula, originally described by Forel (1922)
as a species of Prenolepis, was recognized as a valid
form of Conomyrma in the southeastern United
States.

Finally, Johnson (1989), also using the generic
name Conomyrma, designated a neotype for D.
insanus and a lectotype for D. flavus in a misguided
effort to establish the identity of D. insanus. At the
same time, he proposed that D. medeis Trager and

D. reginicula Trager be treated as synonyms of D.
insanus, and that D. bureni Trager be similarly
regarded as a synonym of D. flavus. I believe that
Johnson was wrong on nearly every point.

It is my hope that important areas of confusion
that have existed in the past may be corrected and
settled by my actions here. In particular, I wish to
firmly and unequivocably establish the identities of
both D. insanus and D. flavus. The problem of the
identity of D. pyramicus, the type species of the
genus, cannot now be resolved; it is possible that
the types of this name no longer exist. If this proves
to be the case, then a specimen matching the rather
meagre original description will have to be desig-
nated as a neotype; such action is beyond the goals
of the present paper and will have to be deferred
until more material from Brazil is available. How-
ever, there is no evidence to support the view that
a single species ranges from the southern United
States to Argentina. That such confusion still exists
is evidenced by the recent statement by Fowler
(1993) that D. pyramicus “has a distribution which
ranges from the southern United States of America
to Buenos Aires, Argentina .... In particular, its
wide distribution through the Caribbean . . . sug-
gests that this species may have spread through
human commerce, and that it should be adapted
to the harsh conditions of littoral systems.” Fowler
cited papers by Kusnezov (1952, 1959) in support
of this statement, apparently unaware of the more
recent works that show that “D. pyramicus ” is a
composite species (Snelling, 1973; Trager, 1988;
Johnson, 1989). Since the identity of D. pyramicus
remains unsettled, such statements must be regard-
ed with considerable caution.

Most recently, the genus Conomyrma has been
treated as a synonym of Dorymyrmex by Shattuck
(1992). I believe that Shattuck’s conclusions are
sound and that Conomyrma, as a genus apart from
Dorymyrmex, is indefensible.

MATERIALS AND METHODS

The bulk of the specimens used in this study are from
the collections of the Natural History Museum of Los
Angeles County (LACM). Additional material, especially
important type material, was made available from the
collections of the Museum of Comparative Zoology
(MCZ) and the National Museum of Natural History
(USNM). Important specimens were also examined from
the personal collection of J.C. Trager.

The morphological terminology in the descriptions and
comments below are all those conventional to ant sys-
tematics. I do not accept the popular view that the ant
head should be considered prognathous. The head of ants,
as well as of all other Hymenoptera, is essentially hypog-
nathous. Therefore, the so-called “posterior border” of
the head is at the summit of the head and is called the
vertex, a long-established term recognized in nearly all
insect groups. It follows that that portion of the head
bearing the clypeus and the antennal sockets is the front
of the head. Similarly, I abjure the terms “alitrunk,” “trun-
cus,” “trunk,” and “thorax,” preferring instead mesoso-
ma, a term recognized although rarely used by ant sys-
tematists: “thorax” is, of course, morphologically incor-
rect, and the other terms are simply pointless jargon for
which there is no logical justification. The following ac-
ronyms are used to conserve space:

Cephalic index (Cl) — The ratio of head length to head
width as expressed in the formula (HW/HL) (100).

Eye length (EL) — The maximum diameter of the eye
as measured in lateral view.

Head length (HL) — The maximum length of the head,
from the lower (apical) clypeal margin to the summit of
the vertex, not including the mandibles.

Head width (HW) — The maximum width of the head,
in frontal view, exclusive of the compound eyes.

Interocular distance (IOD) — With the head in frontal
view, the minimum distance between the compound eyes.

Interocular ratio (IOR) — The ratio of interocular dis-
tance to eye length as expressed by the formula (IOD/
EL) (100).

Ocular index (OI) — The ratio of the eye length to head
length as expressed by the formula (EL/HL) (100).

Oculomalar ratio (OMR) — The ratio of eye length to
malar area length as expressed by the formula (OMD/
EL) (100).

Oculomandibular distance (OMD) — The length of the
malar area as viewed in profile, from lower eye margin to
uppermost mandibular condyle.

Scape index (SI) — The ratio of scape length to head
length as expressed by the formula (SL/HL) (100).

Scape index2 (SI2) — The ratio of scape length to head
width as expressed by the formula (SL/HW) (100).

Scape length (SL) — The maximum length of the anten-
nal scape, exclusive of the basal condyle, measured from
the basal flange to the apex.

Total length (TL) — The sum of the head length +
length of mesosoma + length of petiole and gaster.

SYSTEMATICS

Dorymyrmex insanus (Buckley)

Figures 1-3

Formica insana Buckley, 1866:165; 92.
Dorymyrmex insanus: Norton, 1875:734. Mc-
Cook, 1879:185-186.

Dorymyrmex pyramicus: Mayr, 1886:433; W.M.
Wheeler, 1902:6-7; 1906:342. Creighton, 1950:
346-349.

Dorymyrmex ( Conomyrma ) pyramicus: Gregg,

1963:432-434 (in part).

Conomyrma insana: Snelling, 1973 (in part).

Since McCook failed to elaborate on the statement
that his new variety flavus differed from D. insanus
in the contour of the mesosoma, it is difficult to
determine by what means Johnson was led to as-

Contributions in Science, Number 454

Snelling: Nearctic Dorymyrmex ■ 3

sume that this meant that D. insanus had a sharply
angulate mesonotal profile while in D. flavus the
profile was non-angulate. Presumably he was fol-
lowing the assumptions of every myrmecologist
since W.M. Wheeler (1902), who fostered this in-
terpretation. In his original description of D. in-
sanus, Buckley stated “mesothorax somewhat de-
pressed.” I do not believe that Buckley would have
used “somewhat” as a qualifier had the depression
been abruptly angulate, and no justification exists
for the assumption that the mesonotum is sharply
angulate. This belief is bolstered by the figure in
McCook’s treatment; in that figure the mesonotum
is rather evenly rounded behind. Although the fig-
ure is clearly a crude one, it still seems unlikely that
an abruptly angulate profile would be rendered as
one that is evenly curved. The neotype specimen
designated by Johnson is a specimen with an abrupt-
ly angulate mesonotal profile. I do not, however,
wish to suggest that the mesonotal profile is evenly
curved as shown in the McCook figure. Instead,
the profile probably should have been rendered as
being broadly angulate or obtuse.

Johnson’s proposed neotype is also at odds with
the original description regarding head shape. Buck-
ley stated that the head of the worker is “rounded
above.” Johnson, on the other hand, described the
“occipital margin” (i.e., the vertex as seen in frontal
view) as clearly to slightly concave. The obvious
discrepancy was passed over without comment.

Of the queen, Buckley stated that the entire in-
sect is “bronze-black” and the margins of the gas-
tral terga “hyaline.” The queen from the same nest
as the neotype designated by Johnson is distinctly
bicolored (reddish, with black gaster), and no evi-
dence was provided to demonstrate that D. insanus
females may be either unicolored or bicolored.
There is once more a clear departure from Buck-
ley’s description.

Part of Johnson’s confusion is clear from his
discussion of the distribution of D. insanus.
Throughout that discussion is the implicit assump-
tion that D. insanus is the only blackish Dory-
myrmex across the southeastern states. This is ev-
idenced by the uncritical acceptance of all prior
literature records of dark Dorymyrmex in these
states as being based on D. insanus ; at least some
of these almost certainly refer to D. grandulus.
Many others, however, are based on the form de-
scribed by Cole (1936) as the var. smithi, here re-
garded as a distinct species with an abruptly de-
clivitous mesonotal profile.

The same uncritical acceptance is reflected in
Johnson’s statement that D. insanus “is fortunately
the only black-bodied Conomyrma described or
reported in or near Texas, its type locality.” At least
two dark-bodied Dorymyrmex species are present
in Texas: D. insanus and D. smithi (= C. insanus,
sensu Johnson). There may well be others. Since
Johnson failed to examine type material of D. smithi,
there is no way that he could have known that this
is a distinct species.

According to the International Code of Zoolog-

ical Nomenclature (1985), Article 75 (d) (4), a neo-
type is validly designated only if there is “evidence
that the neotype is consistent with what is known
of the former name-bearing type from the original
description and from other sources.” Johnson’s
neotype material is clearly discordant with what is
known of the former name-bearing type material,
as described, and from the only other source
(McCook) known to have examined original ma-
terial; the data provided by these two sources are
the only ones that cannot be ignored. Interpreta-
tions subsequent to that of McCook appear to have
been based on other species or a mix of species and
therefore have no bearing on resolving the identity
of this species.

Finally, as a rather minor point, it should be
noted that Johnson did not follow original or-
thography. He designated a type for “ Conomyrma
insana (Buckley)” rather than Formica insana
Buckley. The original author’s orthography should
be used when labelling type material in order to
avoid possible future error.

In order to stabilize the identity of Formica in-
sana Buckley, 1866, 1 designate as neotype a worker
specimen with the following data: Interstate 20, 12
mi E Big Spring, Howard Co., TEXAS, 16 April
1981, collected by W.F. Buren; neoparatype series
consists of 49 workers and one female with the
same data. Neotype deposited in USNM; neopar-
atypes in BMNH, LACM, MCZ, and USNM.

DESCRIPTION

Worker neotype, measurements (mm): HL 0.90;
HW 0.79; SL 0.87; EL 0.26; TL 3.1. Indices and
ratios: Cl 89; SI 97; SI2 110; OI 29; OMR 80; IOR
150.

Head (Fig. 1) broadest slightly above upper eye
margin; in frontal view sides weakly convex and
broadly, evenly rounded onto vertex; vertex margin
mostly weakly convex from side to side, but with
weak median emargination. Eye large, OMD about
one-half as long as distance from upper eye margin
to level of vertex margin.

Promesonotal profile (Fig. 2) weakly convex,
mesonotum broadly subangulate behind, with dis-
tinct sloping posterior face. In profile, basal face of
propodeum distinctly sinuate; dorsal tubercle about
as high as thick at base; declivitous face nearly
straight.

Mandible shiny, costate for nearly entire length.
Head and body moderately shiny, distinctly tessel-
late, but frontal area, clypeus, and gena shinier and
tessellation weak to absent.

Entire head (except clypeus, frontal area, and
gena) densely pubescent, hairs fine and fully ap-
pressed, mostly separated from adjacent hairs by
no more than two times their widths; mesosoma
similarly pubescent, pubescence weaker on sides;
gastral terga similar but hairs slightly denser, coars-
er, and longer. Mandible with short, suberect setae;
clypeus with usual distal row of long curled setae
and several long suberect to erect setae on disc;

4 ■ Contributions in Science, Number 454

Snelling: Nearctic Dorymyrmex

frontal lobes each with one pair of long erect setae.
Pronotal seta pair present, setae short and incon-
spicuous. Gaster with scattered erect setae, es-
pecially on last three segments.

Color medium brown, appendages paler; head
and gaster slightly darker than mesosoma; margins
of gastral terga slightly paler.

Worker neoparatypes, measurements (mm): HL
0.81-0.97; HW 0.68-0.87; SL 0.81-0.94; EL 0.23-
0.27; TL 2. 7-3.4. Indices and ratios: Cl 84-93; SI
96-102; SI2 108-110; OI 27-29; OMR 83-100; IOR
128-162.

Neoparatypes agree generally with above de-
scription; a few lack the pronotal setae.

Female, measurements (mm): HL 1 .03; HW 1 .01;
SL 0.86; EL 0.36; TL 6.0. Indices and ratios: Cl
99; SI 84; SI2 85; OI 35; OMR 61; IOR 171.

Head (Fig. 3) broadest behind eyes, margins gent-
ly convex in frontal view and broadly rounded onto
vertex; vertex margin weakly convex. In fron-
tal view, outer margins of eyes coincident with head
margins; distance from level of eye summit to level
of vertex summit about two times OMD. Meso-
soma, in dorsal view, about as wide as head.

Margins of head (in frontal view) with some short,
fine, suberect to erect hairs; eyes with scattered,
short erect hairs. Pilosity otherwise about as de-
scribed for worker, but frontal lobes each with two
erect setae and an erect seta behind each posterior
ocellus.

Color medium brown, gaster slightly darker and
with apical margins of gastral terga distinctly pallid.

DISCUSSION

Johnson treated the following names as synonyms
of his concept of D. insanus: D. pyramicus var.
smithi (Cole), Conomyrma wheeleri Kusnezov, C.
medeis Trager, and C. reginicula Trager. None of
these names is a synonym of D. insanus as I have
defined that species, and all represent discrete spe-
cies, here reinstated as such.

This species ranges from central Texas north to
Kansas and westward to southern California. The
distribution includes adjacent states across north-
ern Mexico, but the southward extent is unclear,
in part due to the presence of similar-appearing
Mexican species such as D. nigra Pergande and in
part due to inadequate collecting.

Dorymyrmex flavus McCook

Figures 7-9

Dorymyrmex insanus var. flavus McCook, 1879:
186; 9.

Dorymyrmex pyramicus: Creighton, 1950:346-349
(in part).

Dorymyrmex ( Conomyrma ) pyramicus: Gregg,
1963:432-434 (in part).

Conomyrma insana: Snelling, 1973 (in part).

The situation with respect to D. flavus, as treated
by Johnson, is a more difficult and disturbing mat-
ter. The original description was uninformative.

Other than color and contour of the mesosoma,
D. flavus was said to be identical to Buckley’s spe-
cies. The range was merely given as “Southern
States.” While McCook may have meant that the
shape of the mesosoma, including the mesoscutum,
was different between D. flavus and D. insanus in
that one was angulate and one was not, this is
probably not true. Note, for example, McCook’s
enigmatic statement that “the cone is evidently
higher than the thorax.” Possibly he meant that the
propodeal tubercle, the “cone,” is relatively high
when compared to that of D. insanus, and it is
quite possible that this is the true meaning of the
differing “contour of the thorax.” Bearing in mind
that D. flavus was described as a mere variety, he
may equally well have meant simply that the meso-
notum was less conspicuously angled than in the
Buckley specimens of D. insanus. Clearly, however,
there is no justification for the assumption that the
mesonotum is straight in profile. Similarly, the state-
ment by Johnson that McCook “considered his
flavus and Buckley’s insana as non-angular and
angular in mesonotal profile respectively” cannot
be supported by any of McCook’s statements. In
fact, the authentic McCook-determined specimens
of D. flavus from Larissa, Texas, do have a broadly
angulate mesonotal profile, despite Johnson’s claim
to the contrary.

In selecting a lectotype specimen, Johnson had
available to him a few specimens from Larissa,
Cherokee County, Texas, a community that no lon-
ger exists. These specimens were labelled as “Do-
rymyrmex insanus var. flavus McC.” in McCook’s
handwriting. Johnson assumed, perhaps correctly,
that these were some of the original material ex-
amined by McCook. However, this is not abso-
lutely certain, and it is entirely possible that these
were specimens collected and identified at a later
date. The note that the ants were “destroying cot-
ton worms” would seem to suggest that they may,
indeed, be original material. While McCook cited
the distribution to be “Southern States,” he did not
specifically mention any one state. Nor is there any
reason to assume that McCook ever understood
that there are other yellow species of Dorymyrmex
in the southern states. Rather, it seems likely that
with material from several states (implied by his
statement on the distribution of D. flavus ), Mc-
Cook almost certainly had a mixed series.

As in the case of the neotype designation for D.
insanus, Johnson did not use original orthography
when designating his “lectotype”: the specimen is
labelled “ Conomyrma flava (McCook),” rather than
Dorymyrmex insanus var. flavus. I have examined
with great care the few Larissa specimens that com-
prise the putative type series. There are a total of
six specimens mounted on a single card on which
were cut eight points; two specimens obviously now
are lost, as indicated by the glue remnants on the
two empty points.

The condition of these specimens is discouraging.
When originally mounted, all were crushed against
their respective points; gasters were uniformly flat-

Contributions in Science, Number 454

Snelling: Nearctic Dorymyrmex ■ 5

tened. Legs were “scrunched” up against and over
the body, and the specimens are more or less cov-
ered by the adhesive material. In most specimens
the heads are crushed and distorted. Finally, the
specimens are dirty, with particles of sand and other
attached debris; the mouthparts cannot be studied
because they are occluded by sand particles and, in
most specimens, glue as well. It is possible, how-
ever, to determine that the integument of the head
and mesosoma is distinctly tessellate and moder-
ately shiny.

These difficulties are true of all the Larissa spec-
imens except the putative lectotype. In that speci-
men the head and body are not crushed; the gastral
distortion is minor and limited to that resulting
from normal drying of internal tissues. The speci-
men is not crushed flat against the point. The legs
and antennae are fully visible, and no significant
portions of the body are obscured by adhesive, dirt,
or other debris; there is not a sand particle to be
seen anywhere on this specimen, including the
mouthparts.

The “lectotype” differs radically from the other
specimens on the card in one crucial character: in
all the other individuals the mesonotal profile (Fig.
8) is obtusely angulate. Originally this feature was
concealed, but I moved or removed legs so that the
mesonotum was visible in all specimens. In the “lec-
totype” the mesonotum is nearly straight, exactly
as in D. bureni (Fig. 11).

Johnson’s “lectotype” differs from the remaining
specimens on the card in other features as well. The
integument of the head and mesosoma is less sharp-
ly tessellate, hence shinier. Although difficult to
determine in the other five specimens, the head of
the “lectotype” is apparently both longer and rel-
atively narrower; the antennal scape is proportion-
ately longer. The psammophore is slightly nearer
the oral cavity than the occipital foramen in the
“lectotype” but slightly nearer the foramen in the
others.

In short, in terms of preparation technique, phys-
ical condition, and morphological characteristics,
the putative “lectotype” differs sharply from the
other specimens with which it is now associated.
In my opinion, it is conspecific with D. bureni and
not with the other specimens on the card, all of
which can be safely referred to D. flavus as that
name has usually been interpreted. I can only con-
clude that Johnson’s “lectotype” is, in fact, a spec-
imen of D. bureni added to the card at a date later
than the other specimens. As such, it is invalid and
must be set aside.

At the same time, I do not believe it advisable
to select a true lectotype from among the five au-
thentic Larissa specimens, none of which is in suf-
ficiently good condition that the important mor-
phological characteristics are readily visible. Fur-
thermore, while it is possible, even likely, that
McCook had these specimens available when he
described this ant, there is no way to prove that
this is so.

The only obvious solution is to set aside all the

Larissa specimens and to assume that no provable
type material exists. It is then possible to designate
a neotype that is consistent with McCook’s original
description and conforms to the visible morpho-
logical characteristics of the Larissa specimens. I
have, therefore, selected as neotype a worker spec-
imen collected 21 mi W Monahans, Ward Co.,
TEXAS, 6 June 1979, by O.F. Francke, J.V. Moody,
and F.W. Merickel; neotype is deposited in USNM;
neoparatypes (8 females, 91 workers, 3 males) in
BMNH, LACM, MCZ, and USNM.

DESCRIPTION

Worker neotype, measurements (mm): HL 0.87;
HW 0.74; SL 0.85; EL 0.27; TL 2.94. Indices and
ratios: Cl 85; SI 97; SI2 114; OI 31; OMR 140; IOR
130.

Head (Fig. 7) broadest at level of eyes; in frontal
view, sides gently and evenly convex, above broadly
rounded onto vertex; vertex slightly convex in fron-
tal view. Eye large, OMD about one-half as long
as distance from eye summit to level of vertex mar-
gin.

Promesonotal profile (Fig. 8) gently convex,
mesonotum obtusely angulate behind. In profile,
basal face of propodeum weakly sinuate; posterior
tubercle erect and sharp, about as high as long;
posterior declivity nearly straight.

Mandible shiny, costate for nearly entire length.
Clypeus and lower face shiny, head distinctly sha-
greened and less shiny dorsally; mesosoma and gas-
ter similar to upper one-half of head.

Front of head, above level of lower eye margins,
with dense appressed pubescence; mesosoma
(sparser on sides) and gaster similar. Erect setae as
described for D. insanus, except pronotum without
discal seta pair.

Color yellowish, head slightly reddish, and last
three gastral segments medium brown; appendages
yellowish, last flagellar segment brownish.

Neoparatype workers, measurements (mm): HL
0.83-0.91; HW 0.69-0.79; SL 0.83-0.90; EL 0.24-
0.27; TL 2.83-3.08. Indices and ratios : Cl 83-87;
SI 98-100; SI2 112-114; OI 29-30; OMR 79-81;
IOR 130-138.

Neoparatypes agree generally with above de-
scription but largest workers with vertex margin
nearly flat in frontal view; of the 91 paralectotype
workers, 27 have at least one (most often two) erect
submedian setae on the disc of the pronotum.

Female, measurements (mm): HL 1.05-1.09; HW
1.05-1.08; SL 0.95-0.97; EL 0.37-0.40; TL 5.74-
6.13. Indices and ratios: Cl 98-101; SI 87-92; SI2
89-91; OI 35-36; OMR 171-200; IOR 161-172.

Head (Fig. 9) broadest above eyes, margins weak-
ly convex in frontal view, dorsally broadly rounded
onto vertex; vertex margin nearly flat to weakly
convex. Eye margins, in frontal view, coincident
with or slightly exceeding head margin; distance
from eye summit to level of vertex about two times
OMD.

Mesosoma, in dorsal view, about as wide as head.

6 ■ Contributions in Science, Number 454

Snelling: Nearctic Dorymyrmex

Pilosity about as described for D. insanus but
head margins in frontal view with few suberect or
erect hairs.

Color light yellowish brown, with variable slight-
ly duskier blotches on head and mesosoma; gaster
more brownish, terga more yellowish basad and
with broad pallid apical margins.

DISCUSSION

The several males in this series are not described
because of their poor condition and because spe-
cific character states in males are poorly defined.

Johnson included within his concept of D. flavus,
as a synonym, the Floridian species D. bureni Tra-
ger. Since Johnson’s lectotype designation for D.
flavus is invalid and D. flavus as established here
is a different species, it follows that D. bureni must
be reinstated as a valid taxon.

This species has been recently collected in a park
near Blythe, Riverside County, California. Since D.
flavus appears to be absent over most of New Mex-
ico and all of Arizona, it seems likely that this
California record represents an accidental intro-
duction. The county park near Blythe is on the
banks of the Colorado River and is heavily used by
travellers from other states. The ant is now abun-
dant in that park but has not been collected else-
where in California.

Dorymyrmex bureni (Trager)
REVISED STATUS
Figures 10-12

Conomyrma edeni Tryon, 1986:340 ( nomen nu-
dum); Barton, 1986:496 ( nomen nudum).
Conomyrma bureni Trager, 1988:19-21; 92.
Conomyrma flava: Johnson, 1989:187-191; 92 (in
part, misidentification, including “lectotype” of
“ Conomyrma flava .”)

As shown above, Johnson’s “lectotype” of D. fla-
vus is not part of the original series and not con-
specific with authentic specimens identified as D.
flavus by McCook but is conspecific with Trager’s
recently described D. bureni; Johnson’s contention
that D. bureni is a synonym of D. flavus is here
rejected.

Trager (1988) gives the range of D. bureni as
extending from Maryland and Virginia south to
Florida and west to Mississippi. I have seen samples
from several localities in Louisiana and workers
collected at Columbus, Colorado County, Texas.

Dorymyrmex reginicula (Trager)
REVISED STATUS
Figure 15

Conomyrma reginicula Trager, 1988:27-28; 92.
Conomyrma insana: Johnson, 1989:185-187; 92
(in part).

Although Johnson considered D. reginicula to be
a synonym of D. smitbi (misidentified as C. insana ),
I cannot agree and here reinstate D. reginicula. In

my opinion this social parasite differs from D. smi-
tbi in the features originally cited by Trager, and I
see no evidence of hybridization between the two.

In both workers and females of D. reginicula,
the front of the head is shinier because of the much
finer and shorter appressed pubescence. The female
of D. reginicula is distinctly smaller than that of
D. smitbi (HW 0.98-1.03 versus 1.23-1.32 mm,
respectively). In females of D. reginicula, the scapal
pubescence is closely appressed to the shaft, but in
D. smitbi the hairs are distinctly decumbent to sub-
decumbent. Although the scapal pubescence is sim-
ilar in workers of the two species, the scapes of D.
smitbi workers usually have at least some subde-
cumbent to decumbent pubescence along the shaft.
Trager also noted the proportionately longer scape
in workers of D. reginicula (SI over 101 in 80% of
individuals) when compared to those of D. smitbi
(SI less than 101 in 80% of individuals).

Trager established that D. reginicula is a social
parasite in nests of D. bureni and possibly D. bos-
suta and is apparently not known to occur outside
of Florida. The absence of reginicula-\ike speci-
mens from other areas further suggests the specific
distinctness of the two forms.

Dorymyrmex smitbi Cole
REVISED STATUS
Figures 4-6

Dorymyrmex pyr amicus var. smitbi Cole, 1936:
120; 9.

Dorymyrmex pyramicus: G.C. Wheeler and Whee-
ler, 1963:155-159 (misidentification).
Dorymyrmex ( Conomyrma ) pyramicus: Gregg,
1963:432-434 (in part, misidentification).
Conomyrma insana: Snelling, 1973 (in part, misi-
dentification). Johnson, 1989:185-187 (in part,
misidentification).

Conomyrma medeis Trager, 1988:25-26; 92. NEW
SYNONYMY.

I have now no doubt that my decision to declare
Cole’s D. pyramicus var. smitbi a synonym of D.
insanus was in error. Both workers and females of
D. smitbi differ from those of D. insanus in the
distinctly broader head and the concave, rather than
flat to weakly convex, vertex margin.

I have compared paratype workers of D. smitbi
with those of C. medeis and conclude that they are
conspecific. In both series the worker head (Fig. 4)
is broad, and the vertex margin is distinctly to weak-
ly concave in frontal view, rarely flat. In both series
the mesonotal profile (Fig. 5), anterior to its pos-
terior declivity, is flat or nearly so. The propodeal
tubercle is both higher and sharper in workers of
D. smitbi than in those of C. medeis, but these
differences are well within the range of variation
noted in workers of other species.

Although females of D. smitbi were not available
to Cole at the time of the description, those that I
have seen from Nebraska and adjacent states are
essentially inseparable from female paratypes of C.
medeis.

Contributions in Science, Number 454

Snelling: Nearctic Dorymyrmex ■ 7

In both workers and females from Nebraska, the
front of the head is a little duller than in samples
from Florida. Johnson’s “neotype” of “ Conomyr -
ma insana ” and the associated alate female (also
labelled “neotype”!) are like the Florida population
described by Trager as C. medeis. Other workers
and females from more northern and western lo-
calities in Texas resemble more closely the speci-
mens from Nebraska.

The range of D. smithi extends from North Da-
kota to eastern Colorado and New Mexico, east
through Texas to North Carolina and Florida. The
westward extent of the range is uncertain.

Dorymyrmex wheeleri (Kusnezov)
REVISED STATUS
Figures 13, 18

Conomyrma (? Biconomyrma) wheeleri Kusnezov,
1952:438-439; 9.

Conomyrma insana : Snelling, 1973 (in part).

This species was described from two workers col-
lected by P. Klingenberg at “College P. A.,” Tuc-
son, Arizona, on 22 March 1933; a specific holotype
was not designated. A female specimen in the MCZ
collection, perhaps conspecific with the two work-
ers, has a similar label, with identical data, and a
label written by W.M. Wheeler; “ Dorymyrmex
pyramicus n. subsp.”

The locality data as cited by Kusnezov are in-
correct: the first line of the label reads “College P.
K.” on both samples. According to F.G. Werner
(personal communication), P. Klingenberg collect-
ed specimens in Arizona and sent them to W.M.
Wheeler. Apparently, when the specimens were
mounted and the labels written, the notation “coll.
P. K.” (= collector P. Klingenberg) became “Col-
lege P. K.,” misquoted by Kusnezov as “College P.
A.” A further distortion is that of Shattuck (1994),
who translated the citation to read “College Park,
Tucson.”

Of the two syntypes, the larger is here designated
lectotype; the second specimen is the paralectotype.
Measurements (mm) lectotype (paralectotype): HL
0.73 (0.62); HW 0.60 (0.51); SL 0.71 (0.59); EL 0.19
(0.17); TL 2.5 (2.2). Indices and ratios : Cl 82 (83);
SI 97 (96); SI2 117 (115); OI 26 (27); OMR 93 (85);
IOR 153 (146).

When I placed this species in synonymy with D.
insanus in 1973, 1 did so on the basis of Kusnezov’s
original description; nothing in that description sug-
gested that D. wheeleri could be separated from D.
insanus . Kusnezov failed to note a characteristic
that will separate D. wheeleri from all previously
described North American Dorymyrmex: the
sparsely pubescent frons, the hairs of which are
separated from adjacent hairs by several times their
own widths (in the paralectotype many of the hairs
are separated by their own lengths or more). In the
paralectotype the entire frons is smooth and shiny,
but in the lectotype most of the frons is distinctly
shagreened.

The head shape (Fig. 13) is about as in D. insanus:

lateral margins weakly convex, vertex margin so
weakly convex that it is nearly flat. The profile of
the mesonotum (Fig. 18) is flat and with a distinctly
sloping posterior face. The mostly convex basal
face of the propodeum is abruptly depressed be-
hind, to form a distinct transverse crease in front
of the short, acute tubercle.

These two workers appear to be nanitic, and I
have seen no specimens like them in any of the
numerous collections of Dorymyrmex available
from the Tucson area. The female specimen cited
above is probably not conspecific with the worker
types of D. wheeleri. It is similar to a number of
females sent to me by D.E. Wheeler; several of
these are foundresses with nanitic workers that do
not resemble the D. wheeleri types. They may rep-
resent an undescribed species.

Dorymyrmex lipan, new species
Figures 16, 17

DIAGNOSIS

Worker head nearly devoid of appressed pubes-
cence, entirely smooth and shiny.

DESCRIPTION

Worker holotype, measurements (mm): HL 0.81;
HW 0.71; SL 0.79; EL 0.22; TL 2.7. Indices and
ratios: Cl 88; SI 98; SI2 113; OI 27; OMR 94; IOR
159.

Head (Fig. 16) broadest slightly above upper eye
margin; in frontal view, sides nearly straight below
and weakly curved above eyes, broadly and evenly
rounded onto vertex, vertex nearly flat. Eye large,
OMD about 0.60 times distance from upper eye
margin to level of vertex margin.

Promesonotal profile nearly flat, obtusely angu-
late behind (Fig. 17). In profile, basal face of pro-
podeum weakly sinuate, posterior cone erect and
sharp; posterior declivity nearly straight.

Mandible shiny, costate for nearly entire length.
Head and body shiny, integument weakly tessellate
to smooth.

Entire head and body sparsely pubescent, ap-
pressed hairs nowhere obscuring surface and not
imparting sheen across surface. Erect hairs as de-
scribed for D. insanus.

Color medium reddish brown, gaster darkest; an-
tennae and legs yellowish.

Paratype, measurements (mm): HL 0.79-0.86;
HW 0.71-0.77; SL 0.78-0.83; EL 0.19-0.24; TL
2.7-1.9. Indices and ratios: Cl 86-92; SI 92-100;
SI2 107-114; OI 24-28; OMR 93-94; IOR 143-
157.

TYPE MATERIAL

Holotype and 14 paratype workers: TEXAS, Brew-
ster Co.: Basin Road, 4800 ft. elev., Chisos Mts.,
Big Bend National Park, 6 Sept. 1969 (G.C. &C J.
Wheeler, no. TEX-258). Holotype in LACM; para-
types in BMNH, LACM, MCZ, USNM.

8 ■ Contributions in Science, Number 454

Snelling: Nearctic Dorymyrmex

ETYMOLOGY

Named for the Lipan Apache of western Texas; the
name is a noun in apposition.

Dorymyrmex paiute, new species
Figures 14, 23

DIAGNOSIS

Unicolorous brownish species with greatest EL dis-
tinctly shorter than OMD.

DESCRIPTION

Worker (holotype), measurements (mm): HL 0.90;
HW 0.81; SL 0.84; EL 0.21; TL 3.0. Indices and
ratios: Cl 89; SI 97; SI2 110; OI 23; OMR 114; IOR
200.

Head (Fig. 14) broadest at eye level; in frontal
view, margins weakly convex and broadly rounded
onto vertex, vertex weakly concave. Eye small, EL
0.90 times OMD, OMD about 0.64 times distance
from upper eye margin to level of vertex margin.

Mesonotal profile (Fig. 23) flat or slightly con-
cave; base of propodeum long, distinctly sinuate,
posterior tubercle broad at base, sharp.

Sculpture and pilosity as described for D. insanus
but pronotum without erect hairs (several speci-
mens have one pair of setae that are distinctly short-
er than minimum scape thickness).

Color light reddish brown, head more reddish
and gaster more brownish; antennae and legs paler.

Paratypes, measurements (mm): HL 0.81-0.92;
HW 0.71-0.84; SL 0.77-0.87; EL 0.19-0.23; TL
2.9-3.4. Indices and ratios: Cl 88-96; SI 93-100;
SI2 100-114; OI 22-26; OMR 108-129; IOR 171-
200.

TYPE MATERIAL

Holotype and 35 paratype workers: UTAH, Wash-
ington Co.: Zion National Park, 22 July 1932 (W.S.
Creighton). Holotype in LACM; paratypes in
BMNH, LACM, MCZ, USNM.

ETYMOLOGY

Named for the Paiute people who formerly oc-
cupied the Great Basin region; the name is a noun
in apposition.

DISCUSSION

The small eyes and lack of pronotal hairs serve to
separate this species from all others previously de-
scribed from North America. Three specimens do
possess pronotal hairs, but these hairs are distinctly
shorter than the minimum thickness of the antennal
scape. In all other species with pronotal hairs, the
hairs are as long as, or longer than, the maximum
width of the antennal scape.

This species is presently known only from the
type series.

KEY TO WORKERS OF
UNITED STATES SPECIES
OF DORYMYRMEX

1 Mesonotal profile, in all or nearly all in-
dividuals from same nest, evenly convex
(Figs. 11, 20) or flat to weakly concave
(Fig. 22) • 2

- Mesonotal profile, in all or nearly all indi-
viduals from same nest, with distinct dorsal
and declivitous faces that meet in more or
less well-defined angle (Figs. 18, 19) ... 4

2(1) Promesonotal profile convex, base of pro-
podeum angled 145° or less to plane of pos-
terior portion of mesonotum (Figs. 11, 20);
scape surpassing vertex by no more than
0.33 times its length (Fig. 10); color various
3

- Promesonotal profile flat, base of propo-

deum angled 165° or more to plane of meso-
notum (Fig. 22); scape surpassing vertex by
nearly 0.5 times its length; color uniformly
clear yellow (gastral apex sometimes dark-
ened) (Florida) elegans (Trager)

3(2) Head (Fig. 10) broad, Cl over 87 (less in
some minor workers); scape relatively short
(SI less than 112); propodeal tubercle blunt,
posterior face straight or slightly convex (Fig.
11); color largely yellow or reddish yellow,
but if head and gaster infuscated, then me-
sosoma also infuscated (Maryland to Flor-
ida, west to eastern Texas)

bureni (Trager)

- Head narrower, Cl less than 87; scape rel-

atively longer, (SI at least 112 in 95% of
individuals); propodeal tubercle sharper,
posterior face slightly concave (Fig. 20); head
and gaster brown, mesosoma clear yellow-
ish (Florida) flavopectus (M. Smith)

4(1) Front of head and discs of gastral terga 1-
3 shiny, with only obscure sparse to scat-
tered appressed pubescence 5

- Front of head and discs of gastral terga 1-
3 only moderately shiny, with abundant ap-
pressed pubescence that partially obscures
surface and imparts silky sheen to surface

in oblique view 7

5(4) Posterior face of mesonotum distinctly
sloping in profile (Figs. 17, 18) 6

- Posterior face of mesonotum nearly vertical
in profile (Fig. 19) (Florida and Georgia) . .

bossutus (Trager)

6(5) Propodeal tubercle short and preceded by
transverse crease (Fig. 18); disc of gastral
tergum 1 slightly shiny, distinctly roughened
(Arizona) wheeleri (Kusnezov)

- Propodeal tubercle relatively higher and not

preceded by transverse crease (Fig. 17); disc
of gastral tergum 1 smooth and shiny (Tex-
as) lipan, new species

7(4) Head relatively narrow, Cl usually less than
88, rarely as much as 90; vertex margin

Contributions in Science, Number 454

Snelling: Nearctic Dorymyrmex ■ 9

Figures 1-6. Worker head (frontal view), worker head and mesosoma (profile), and female head (frontal view),
respectively, of Dorymyrmex insanus (1-3) and D. smithi (4-6). Figures 1, 2, and 4-6 by Tina Ross.

10 ■ Contributions in Science, Number 454

Snelling: Nearctic Dorymyrmex

Figures 7-12. Worker head (frontal view), worker head and mesosoma (profile), and female head (frontal view),
respectively, of Dorymyrmex flavus (7-9) and D. bureni (10-12). Figures 7, 8, 10, and 11 by Tina Ross.

Contributions in Science, Number 454

Snelling: Nearctic Dorymyrmex Mil

Figures 13-23. 13-16. Head (frontal view) of Dorymyrmex wheeleri (worker) (13), D. paiute (worker) (14), D.

reginicula (female) (15), and D. lip an (worker) (16). 17. Head and mesosoma (profile) of worker D. lipan. 18-23. Worker
mesosoma (profile) of D. wheeleri (18), D. bossutus (19), D. flavopectus (20), D. grandulus (21), D. elegans (22), and
D. paiute (23).

12 ■ Contributions in Science, Number 454

Snelling: Nearctic Dorymyrmex

straight or slightly convex (Figs. 4, 7); eye
relatively large, IOD usually less than 1.50
(never over 1.70) times EL and EL equal to
or exceeding OMD; female unicolorous yel-
lowish or brownish 8

- Head relatively broad, Cl over 90; eye rel-

atively small, IOD at least 1.75 times EL,
OMD at least equal to, and usually greater
than, EL; vertex margin usually distinctly
concave in frontal view, rarely straight; fe-
male head and mesosoma red, gaster black-
ish 10

8(7) Propodeal tubercle relatively prominent;
pronotum usually with discal seta pair, or,
if absent, color distinctly yellowish (west of
Mississippi River) 9

- Propodeal tubercle short (Fig. 21); pronotal

disc without erect setae (Illinois?, Michi-
gan?, New Jersey, south to Florida and Al-
abama) grandulus (Forel)

9(8) Color light to dark brownish, head and gas-
ter commonly darker than mesosoma; pro-
notal seta pair usually present (Kansas to
central Texas, west to southern California)
insanus (Buckley)

- Color clear yellowish to reddish yellow, of-

ten with vertex, mesosomal dorsum, and
apex of gaster infuscated; pronotal seta pair
commonly absent (Kansas and eastern Col-
orado to southern Texas, eastern New
Mexico, and western Louisiana)

flavus (McCook)

10(7) Head and mesosoma black or dark brown
11

- Head and mesosoma red (western Texas to

southern Nevada and California)

bicolor Wheeler

11(10) Vertex, in frontal view, distinctly concave
(Fig. 4) (North Dakota south to eastern New
Mexico and western Texas, east across
southern states to North Carolina, Georgia,
and Florida) smithi Cole

- Vertex, in frontal view, straight or weakly

concave (Fig. 14) (Utah)

paiute , new species

ACKNOWLEDGMENTS

For making important types and other material available,
I thank S. Cover, D.R. Smith, J.C. Trager, and D.E. Whee-
ler. I especially acknowledge my indebtedness to F.G.
Wemer for unravelling “College P. A.” I also wish to
thank Tina Ross for her assistance in the preparation of
some of the figures in this paper. For important review
comments I am indebted to S. Cover and J. Trager.

LITERATURE CITED

Barton, A.M. 1986. Spatial variation in the effect of ants
on an extrafloral nectary plant. Ecology 67:495-504.
Buckley, S.B. 1866. Descriptions of new species of North

American Formicidae. Part 1. Proceedings of the
Entomological Society of Philadelphia 6:152-171.

Cole, A.C., Jr. 1936. Descriptions of seven new western
ants. (Hymenoptera: Formicidae). Entomological
News 47:118-121.

Creighton, W.S. 1950. The ants of North America. Bul-
letin of the Museum of Comparative Zoology at
Harvard College 104:1-585.

Forel, A. 1911. Ameisen des Herrn. Prof. v. Ihering aus
Brasilien (Sao Paulo usw.) nebst einigen anderen aus
Siidamerika und Afrika. Deutsche Entomologische
Zeitschrift, pp. 285-312.

. 1913. Formicides du Congo beige recoltes par

M. M. Bequaert, Luja, etc. Revue Zoologique Af-
ricaine 2:306-351.

. 1922. Glanures myrmecologiques en 1922. Re-
vue Suisse de Zoologie 30:87-102.

Fowler, H.G. 1993. Spatial patterning of Dorymyrmex
pyramica (Hymenoptera, Formicidae) nests in a lit-
toral dune of the State of Sao Paulo, Brazil. Revista
Brasileira Entomologia 37:341-344.

Gregg, R.E. 1963. The ants of Colorado. Boulder: Uni-
versity of Colorado Press, 792 pp.

International Commission on Zoological Nomenclature.
1985. International Code of Zoological Nomen-
clature, 3rd ed., edited by W.D.L. Ride, C.W. Sa-
brosky, G. Bernardi, and R.V. Melville. Hudders-
field, U.K.: H. Charlesworth.

Johnson, C. 1989. Taxonomy and diagnosis of Cono-
myrma insana (Buckley) and C. flava (McCook)
(Formicidae). Insecta Mundi 3:179-194.

Kusnezov, N. 1952. El estado real del grupo Dorymyr-
mex Mayr. Acta Zoologica Lilloana 10:427-448.

. 1959. Die Dolichoderinen-Gattungen von Siid-

Amerika. Zoologische Anzeiger 162:38-51.

Mayr, G. 1866. Myrmecologische Beitrage. Sitzungber-
ichte der Akademie der Wissenschaften in Wien 53:
484-517.

. 1870. Formicidae novogranadenses. Sitzung-

berichte der Akademie der Wissenschaften in Wien
61:370-417.

. 1886. Die Formiciden der Vereinigten Staaten

von Nordamerika. Verhandlunger der K.K. Zoolo-
gisch-botanischen Gesellschaft in Wien 36:419-464.

McCook, H.C. 1879. Formicariae. In Report on cotton
insects, J.H. Comstock. Entomology Report of the
United States Department of Agriculture, 182-189.

Norton, E. 1875. Report upon the collection of For-
micidae, pp. 729-36 of chap. 7: Report upon the
collection of Hymenoptera made in portions of Ne-
vada, Utah, Colorado, New Mexico, and Arizona
during the years 1872, 1873, and 1874. In vol. 5 of
Report upon the geographical and geological explo-
rations and surveys west of the one hundredth me-
ridian, in charge of Lieut. Geo. M. Wheeler. . . .
Washington, D.C.: Engineer Department of the
United States Army.

Pergande, T. 1895. Mexican Formicidae. Proceedings
of the California Academy of Sciences 5:858-896.

Roger, J. 1863. Die neu aufgefiihrten Gattungen und
arten meines Formiciden Verzeichnisses. Berliner
Entomologische Zeitschrift 7:131-214.

Shattuck, S.O. 1992. Generic revison of the ant subfam-
ily Dolichoderinae (Hymenoptera: Formicidae). So-
ciobiology 21:1-181.

. 1 994. Taxonomic catalog of the ant subfamilies

Aneuretinae and Dolichoderinae (Hymenoptera:
Formicidae). University of California Publications
in Entomology 112:1-241.

Contributions in Science, Number 454

Snelling: Nearctic Dorymyrmex 2 13

Smith, M.R. 1944. Additional ants recorded from Flor-
ida, with descriptions of two new subspecies. Florida
Entomologist 27:14-17.

Snelling, R.R. 1973. The ant genus Conomyrma in the
United States (Flymenoptera: Formicidae). Contri-
butions in Science 238:1-6.

Trager, J.C. 1988. A revision of Conomyrma (Hyme-
noptera: Formicidae) from the southeastern United
States, especially Florida, with keys to the species.
Florida Entomologist 71:11-29.

Tryon, E.FT, Jr. 1986. The striped earwig, and ant pred-
ators of sugarcane rootstock borer, in Florida citrus.
Florida Entomologist 69:336-343.

Wheeler, G.C., and J. Wheeler. 1963. The ants of North
Dakota. Grand Forks: University of North Dakota
Press, 326 pp.

Wheeler, W.M. 1902. A consideration of S. B. Buckley’s
“North American Formicidae.” Transactions of the
Texas Academy of Science 4:1-15.

. 1906. The ants of the Grand Canon. Bulletin

of the American Museum of Natural History 22:
329-345.

Received 25 May 1994; accepted 10 January 1995.

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

L'S./JX Number 455
\h!H 27 July 1995

Contributions
in Science

New Miocene Horses from the
Caliente Formation,

Cuyama Valley Badlands, California

Thomas S. Kelly

TP 0 1 1991

1C.O

Natural History Museum of Los Angeles County

Serial

Publications

of THE
Natural History
Museum of
Los Angeles
County

• Contributions in Science, a miscellaneous series of tech-
nical papers describing original research in the life and
earth sciences.

• Science Series, long articles and collections of papers
on natural history topics.

Copies of the publications in these series are sold through
the Museum Book Shop. A catalog is available on request.

Scientific

Publications

Committee

James L. Powell, Museum President
Daniel M. Cohen,
Committee Chairman
Brian V. Brown
Kenneth E. Campbell
Kirk Fitzhugh
Robin A. Simpson, Managing Editor

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

Printed at Allen Press, Inc., Lawrence, Kansas
ISSN 0459-8113

New Miocene Horses from the
Caliente Formation,

Cuyama Valley Badlands, California

Thomas S. Kelly1

ABSTRACT. Three new equid taxa are recognized from the Caliente Formation: the late Hemingfordian
Parapliohippus n. gen., the late Clarendonian Heteropliohippus hulberti n. gen. and sp., and the late
Hemingfordian to late Barstovian Acritohippus quinni n. gen. and sp. Parapliohippus is known only from
the type species, P. carrizoensis (Dougherty), which was previously referred to Merychippus. Paraplio-
hippus is assigned to the tribe Equini and is regarded as the sister taxon of the higher equines ( Pliohippus
s.s., Heteropliohippus, Astrohippus, Ortohippidium, Hippidion, Dinohippus, and Equus). Heteroplio-
hippus is assigned to the tribe Equini and is most closely related to Pliohippus s.s. and Astrohippus. The
acritohippines ( Acritohippus tertius, A. isonesus, and A. quinni ) represent a monophyletic clade whose
relationships to the Equini and the Hipparionini are unresolved. The protohippines (“ Merychippus ”
intermontanus, Calippus, and Protohippus) represent a monophyletic clade whose relationships to the
Hipparionini and the Equini are unclear but are provisionally regarded as the sister group of the Hip-
parionini.

INTRODUCTION

The Cuyama Badlands, which occur along the east-
ern side of Cuyama Valley, Ventura County, Cali-
fornia (Figure 1), consist of nonmarine strata rang-
ing in age from the Oligocene to the Pleistocene,
including the Miocene Caliente Formation. Out-
crops of the Caliente Formation also occur north
of Cuyama Valley along the northeastern flanks of
the Caliente Range and southeast in Lockwood
Valley. The Caliente Formation of the Cuyama Val-
ley Badlands has yielded eight mammalian local
faunas that range in age from the early Miocene,
Hemingfordian North American Land Mammal Age
(NALMA), to the late Miocene, Hemphillian NAL-
MA (Gazin, 1930; Wood, 1937; Stock, 1947; James,
1963; J.P. Quinn, 1984; Madden, 1987; Tedford et
ai, 1987; Kelly and Lander, 1988a, 1988b, 1992;
Kelly, 1992). James (1963) described the small
mammals that occur in these faunas. However, most
of the larger mammals, many representing new taxa,
remain undescribed.

Equid fossils from the Cuyama Badlands are rel-
atively abundant. Kelly and Lander (1992) recog-
nized a total of at least 14 equid taxa from the
Cuyama Badlands with a maximum diversity of 4
taxa within any local fauna (Table 1). Most of the

1. Museum Associate, Vertebrate Paleontology/Ran-
cho la Brea Section, Natural History Museum of Los
Angeles County, 900 Exposition Boulevard, Los Angeles,
California 90007, and 558 Green Acre Drive, Gardnerville,
Nevada 89410.

Contributions in Science, Number 455, pp. 1-33
Natural History Museum of Los Angeles County, 1995

horse taxa from the Cuyama Badlands are known
only from fragmentary specimens consisting of teeth
or isolated appendicular elements. However, three
of these taxa are now represented by relatively com-
plete cranial material from the Caliente Formation
and other Miocene formations of southern Cali-
fornia. Reevaluation of the cranial and dental mor-
phology of these three horses indicates that they
represent new taxa.

The purpose of this report is to: 1) document the
three new Miocene horse taxa from the Caliente
Formation and 2) review the phylogenetic relations
of these new taxa with those of other late Neogene
horse taxa.

METHODS

Measurements of specimens were taken with a vernier
caliper to the nearest 0.1 mm. Upper teeth are indicated
by uppercase letters and lower teeth by lowercase letters.
All teeth were measured along their greatest dimensions.
The term “Neogene hypsodont horses” as referred to
herein not only includes typical hypsodont horses, such
as Equus and Hipparion, but also includes horses tradi-
tionally regarded as mesodont, such as “ Merychippus ”
primus (Osborn, 1918), and differentiates them from other
brachyodont Neogene horses, such as those of the An-
chitheriinae. In the discussions of the new genera named
herein and in the cladistic analyses, I sometimes compare
equid taxa of different hierarchical rank as have other
recent investigators of equid phylogeny (Hulbert, 1988a,
1988b, 1989, 1993; Hulbert and MacFadden, 1991). This
is often necessary because many of the taxa discussed
represent plesions (plesiomorphs) of generic rank (Wiley,
1981); that is, they cannot be assigned to any recognized
equid genus without resulting in paraphyly, and the only

Figure 1. Map showing geographic location of Cuyama Valley Badlands, Ventura County, California. Base maps: U.S.
Geological Survey, Cuyama and Taft 1:250,000, 30 x 60 minute quadrangles.

way to refer to them, other than naming a new genus for
each one, is to include their specific names (e.g. “Mery-
chippus" stylodontus [Merriam, 1915], “Mery chippies”
sp. near “M.” sejunctus [Hulbert and MacFadden, 1991],
“ Merychippus ” goorisi [MacFadden and Skinner, 1981],
“ Dinohippus ” interpolatus [Cope, 1893]). Metric abbre-
viations, dental terminology, and dental formulae follow
standard usage.

All cladistic analyses were performed using version 1.5
of the Hennig86 program (Farris, 1988) and run on a 486
personal computer. Cladograms were generated by the
IE-BB command. The characters were either equally
weighted, the default setting for the Hennig86 program,
or successively weighted. Successive character weighting
was accomplished by using the XSTEPS W command,
which calculates the best fits of each character based on
the product of the character consistency and character
retention indices. The cladograms produced using suc-
cessively weighted characters have more steps, but are
based on more reliable characters, than those produced
using equally weighted characters. The characters and
character states used in the cladistic analyses are presented

in Appendix A, and the character state matrices for the
taxa analyzed are presented in Appendices B and C.

Abbreviations are as follows: APL, greatest antero-
posterior length; CV, coefficient of variation; DPOF, dor-
sal preorbital fossa; L, left; Ma, million years before pres-
ent; MML, metaconid-metastylid length; N, number of
specimens; NALMA, North American Land Mammal Age;
OR, observed range; PBL, preorbital bar length; PRL,
protocone length; PRW, protocone width; R, right; ROC,
radius of curvature; SD, standard deviation; s.s., sensu
stricto; TR, greatest transverse dimension; UDL, I3-P2
diastema length; UTRL, P2-M3 length.

Institutional acronyms are as follows: AMNH, Amer-
ican Museum of Natural History; F:AM, Frick Collection,
American Museum of Natural History; LACM, Natural
History Museum of Los Angeles County; LACM(CIT),
California Institute of Technology specimen and locality
number, specimens now housed at the LACM; UCMP,
University of California, Berkeley, Museum of Paleon-
tology; UCR, University of California, Riverside; V-, UCMP
vertebrate fossil locality; USNM, United States National
Museum.

2 ■ Contributions in Science, Number 455

Kelly: Miocene Horses

SYSTEMATIC PALEONTOLOGY
Class Mammalia Linnaeus, 1758
Order Perissodactyla Owen, 1848
Family Equidae Gray, 1821
Subfamily Equinae Gray, 1821
Tribe Equini Gray, 1821
Parapliohippus new genus

Merychippus, in part, Dougherty, 1940 (pp. ISO-

US).

Merychippus, in part, Buwalda and Lewis, 1955 (pp.

148-150).

TYPE SPECIES. P. carrizoensis (Dougherty,
1940) (previously referred to Merychippus and in-
cludes M. carrizoensis Dougherty and its junior
synonym M. tehachapieneis Buwalda and Lewis).

REFERRED SPECIES. Known only from the type
species, P. carrizoensis.

REFERRED SPECIMENS. Skull, AMNH 17061;
skull, F:AM 110146; skull, F:AM 110129; partial
maxilla with LP2-S, LACM(CIT) 255S; partial
maxilla with RP2-4, UCMP 121890; partial maxilla
with LP2-M2, UCMP 45114; partial maxilla with
RP4-M1 and associated RP2, LACM(CIT) 4919;
partial maxilla with LP4-M1, LACM(CIT) 2559;
partial maxilla with RP4-MS, LACM(CIT) 4760;
partial maxilla with LM1-S, LACM(CIT) 2552; par-
tial maxilla with RM1-S, UCMP 121891; partial
maxilla with RPS-M1, LACM 138110; partial max-
illa with RdPS, LACM 138100; LP2, LACM 138098;
LP2, LACM 138104; RP3, LACM(CIT) 2558; RP3,
LACM 138099; RP3, LACM 101152; LP3, UCMP
21972; LP3, UCR 20859; RP4, LACM(CIT) 2560;
LP4, LACM 30076; RP4, UCMP 22898; LM1-2,
LACM(CIT) 4921; LM1-2, LACM(CIT) 4962;
RM1-2, UCMP 82498; LM1-3, LACM(CIT) 4920;
RM1, LACM(CIT) 2556; LM1, LACM(CIT) 2564;
LM2, LACM(CIT) 4965; RM2, LACM 1350; RMS,
LACM 138097; LM3, UCMP 21762; partial den-
tary with dLp2-4 and Lml-2, UCMP 82486; partial
dentary with Rp2-3, UCR 20856; partial dentary
with Rp2-3, UCMP 21688; partial dentary with
Lp2-m2, UCMP 21692; partial dentary with Lp2-
4, USNM 252734; partial dentary with Rp3-4,
USNM 252773; associated dentaries with Lp3-m3
and Rp2-m2, UCMP 11817; partial dentary with
Lp3-ml, LACM 138084; partial dentary with Lp3,
partial p4, and ml, LACM 138087; partial dentary
with Lml-2, LACM 30114; partial Rp2, LACM
55266; Rp2, LACM(CIT) 2581; Rp3, LACM(CIT)
2573; Rp3, LACM 138091; Rp4, LACM(CIT) 2575;
Rml, LACM(CIT) 2578; Rm2, LACM(CIT) 2576;
Rm2, LACM(CIT) 2579; Lm3, LACM(CIT) 2570;
partial dentary with Lm3, LACM 138080; Rm3,
LACM(CIT) 2574; partial dentary with Rm3, LACM
138085; Lm3, UCMP 21685; Lm3, UCR 14271.

DISTRIBUTION AND AGE. California: Unit 2,
Caliente Formation, Caliente Range, late Fleming-

Table 1. Equidae of local faunas from the Caliente For-
mation, Cuyama Badlands, Ventura County, California.
NALMA included for each local fauna. Taxonomic as-
signment follows Kelly and Lander (1992) and this report.

Hidden Treasure Spring Local Fauna (late Heming-
fordian)

Parahippus sp. indet.

Parapliohippus carrizoensis (Dougherty, 1940)
Acritohippus sp. cf. A. tertius (Osborn, 1918)

West Dry Canyon Local Fauna (latest Hemingfordian)
Parapliohippus carrizoensis (Dougherty, 1940)
Acritohippus quinni n. sp.

Lower Dome Spring Local Fauna (early Barstovian)
Acritohippus quinni n. sp.

Upper Dome Spring Local Fauna (early late Barstovian)
Archeohippus mourning i (Merriam, 1913a)
Acritohippus quinni n. sp.

“ Merychippus ” brevidontus (Bode, 1934)

Doe Spring Local Fauna (late Barstovian)

Acritohippus quinni n. sp.

Mathews Ranch Local Fauna (early Clarendonian)
“Pliohippus” tehonensis (Merriam, 1915)
Heteropliohippus hulberti n. sp.

Hipparion tehonense (Merriam, 1916)

Megahippus sp. indet.

Nettle Spring Local Fauna (late Clarendonian)
Pliohippus or Protohippus sp. indet.

Heteropliohippus hulberti n. sp.

Cormohipparion occidentale (Leidy, 1856)
Megahippus sp. cf. M. matthewi (Barbour, 1914)

Sequence Canyon Local Fauna (Hemphillian)

“ Dinohippus ” sp. cf. “D.” interpolatus (Cope, 1893)

fordian; Flidden Treasure Spring and West Dry
Canyon Local Faunas, Caliente Formation, Cuyama
Valley Badlands, late Hemingfordian; Branch Can-
yon Formation, Santa Barbara Canyon, late Hem-
ingfordian; Phillips Ranch Local Fauna, Bopesta
Formation, southern Sierra Nevada, late Heming-
fordian; Red Division Local Fauna, Barstow For-
mation, Mud Hills, late Hemingfordian; Sunrise
Canyon Local Fauna, Barstow Formation, Calico
Mountains, late Hemingfordian; Yermo Hills Local
Fauna, Barstow Formation, Toomey (Yermo) Hills,
? latest Hemingfordian/earliest Barstovian; Bar-
stow Formation, Alvord Mountain, late Heming-
fordian; Upper Cady Mountains Local Fauna, Hec-
tor Formation, northern Cady Mountains, late
Hemingfordian; Daggett Ridge Local Fauna, Bar-
stow Formation, Daggett Ridge, late Hemingfor-
dian; Units 2 and 3, Punchbowl Formation, Cajon
Valley, late Hemingfordian; Fernwood Member,
Topanga Canyon Formation, Santa Monica Moun-
tains, late Hemingfordian.

Contributions in Science, Number 455

Kelly: Miocene Horses ■ 3

DIAGNOSIS. P arapliohippus is monotypic; di-
agnosis for genus is same as for type species. P ar-
apliohippus is distinguished from all other genera
of the Equinae by having the following suite of
characters: 1) frontal bones flat; 2) DPOF with elon-
gated oval shape, deep depth (> 15 mm), anterior
margin confluent with face, posterior margin with
distinct rim, and posterior pocket present; 3) an-
terior portion of lacrimal bone reduced and effec-
tively removed from DPOF, except at orbital rim,
by extensive posterior development of DPOF; 4)
malar fossa anterodorsally directed, relatively deep,
slightly pocketed, and well separated posteriorly
from DPOF by distinct ridge; 5) relative PEL very
narrow (ratio of PBL to UTRL about 0.05); 6) rel-
ative muzzle length elongated (ratio of UDL to
UTRL about 0.55); 7) cement layer on deciduous
premolars very thin and moderately thick on per-
manent cheek teeth; 8) P3-M2 protocones oval-
shaped (ratio of PRL to PRW = 1. 2-2.0) and con-
nect with protolophs shortly after onset of wear
(about 10% wear); 9) P2-M3 protolophs and me-
talophs remain separate until greater than 50% worn;
10) upper cheek teeth metastyles common but not
well developed; 1 1) Ml -2 protocones connect with
hypocones only in late wear; 12) P2 anterostyle
large and expanded; 13) P2-M3 with plis caballin
absent or rare, external fossette plications rare, if
present single and nonpersistent, and internal fos-
sette plications very simple; 14) P2-M2 hypoconal
grooves close in moderate wear; 15) P3-4 hypo-
conal lakes form with closure of hypoconal grooves;
16) P2-M3 strongly curved (ROC < 40 mm); 17)
dpi very reduced, variably present; 18) dp3-4 and
p3-m3 protostylids absent or may be present only
near base of crowns as anterior cingulids; 19) p2-
m3 metaconids and metastylids well separated only
in early wear and metaconid-metastylid complexes
expanded but not elongated (MML = 45-50% of
APL); 20) p2-m3 plis entoflexid commonly present
in early wear; 21) p2 ectoflexid moderately deep,
partially penetrating isthmus between metaconid
and metastylid, and p3-4 ectoflexids deep, com-
pletely penetrating isthmuses between metaconids
and metastylids; 22) ml-3 metastylids notably
smaller and more labially positioned than meta-
conids; 23) size small (UTRL = 90-100 mm); 24)
cheek teeth mesodont (Ml unworn crown height
about 25 mm); and 25) feet tridactyl.

ETYMOLOGY. From the Greek para : near, be-
side; in reference to morphological similarities with
Pliohippus.

DISCUSSION. Dougherty (1940) described Mer-
ychippus carrizoensis based on the holotype, a par-
tial maxilla with LM1-3 (LACM[CIT] 2552), and
a small sample of additional cheek teeth from the
Caliente Formation. Buwalda and Lewis (1955) de-
scribed Merychippus tehachapiensis based on the
holotype, a partial maxilla with RP4-M1 and as-
sociated RP2 (LACM[CIT] 4919), from the Bopesta
Formation. Miller (1978) and Munthe (1979) re-
evaluated the taxonomic relations of these taxa and

determined that M. tehachapiensis is a junior syn-
onym of M. carrizoensis. J.P. Quinn (1984) pro-
vided a detailed description of the facial and cheek
teeth morphology of this species and determined
that it is more closely related to the pliohippines
than to Merychippus s.s. Other investigators have
also recognized that this species is not referable to
Merychippus s.s. (Woodburne and Tedford, 1982;
Woodburne et al., 1982, 1990; Kelly and Lander,
1988b, 1992; Hulbert, 1989, 1993; Hulbert and
MacFadden, 1991; Skinner vide Macdonald et al.,
1992). The cladistic analyses presented below and
those of Hulbert (1989) and Hulbert and Mac-
Fadden (1991) consistently support recognizing this
taxon as a generically distinct clade of the tribe
Equini. Furthermore, referral of this species to Mer-
ychippus s.s. or any other recognized equid genus
would result in paraphyly, and its continued as-
signment to the waste basket, horizontal taxon
“ Merychippus is unwarranted. Therefore, it is as-
signed to P arapliohippus n. gen.

P arapliohippus exhibits similarities in certain fa-
cial and dental morphologies with Pliohippus s.s.
and Astrohippus but can easily be distinguished
from them. P arapliohippus differs from Pliohippus
s.s. by having the DPOF extensively developed, re-
sulting in the effective removal of the lacrimal bone
from the fossa. In Pliohippus s.s. the lacrimal bone
is not reduced anteriorly and extends well into the
DPOF. Additional characters exhibited by Para-
pliohippus that distinguish it from Pliohippus s.s.
are as follows: 1) the relative PBL is narrower; 2)
the infraorbital foramen is positioned more ante-
riorly on the face; 3) the cheek teeth have much
thinner cement and are much less hypsodont; 4)
the protolophs and metalophs remain separated
longer, uniting when the teeth are more than 50%
worn; 5) the p3-4 ectoflexids are deeper, complete-
ly penetrating the isthmuses between the metacon-
ids and the metastylids; 6) the lateral digits are not
reduced; and 7) the size is much smaller.

The facial morphology of P arapliohippus differs
from that of Astrohippus by having a DPOF and
malar fossa that are pocketed posteriorly, deeper,
and separated from each other by a distinct ridge
of bone. Astrohippus also possesses a faint dor-
soventral ridge that divides the DPOF and malar
fossae into anterior and posterior portions, which
is lacking in P arapliohippus. Additional characters
exhibited by P arapliohippus that distinguish it from
Astrohippus are as follows: 1) the relative muzzle
length is more elongated; 2) the upper cheek teeth
have more curvature and thinner cement and are
much less hypsodont; 3) the protocones are less
elongated, more oval in shape; 4) the hypoconal
grooves close at a later wear stage and lakes are
formed on the P3-4 with closure of the grooves;

5) the p2-4 ectoflexids are deeper, penetrating the
isthmuses between the metaconids and metastylids;

6) the ml-3 metastylids are smaller and more la-
bially positioned than the metaconids; and 7) the
size is much smaller.

4 ■ Contributions in Science, Number 455

Kelly: Miocene Horses

Heteropliohippus new genus

TYPE SPECIES. H. hulberti n. sp.

DISTRIBUTION AND AGE. Nettle Spring and
Mathews Ranch Local Faunas, Caliente Formation,
Cuyama Valley Badlands, California, Clarendonian.

REFERRED SPECIES. Only known from type
species.

DIAGNOSIS. Heteropliohippus is distinguished
from all other late Neogene hypsodont horses by
having the following suite of characters: 1) frontal
bones flat; 2) DPOF shape elongated oval, anterior
margin with distinct rim, posterior margin with pro-
nounced rim, and lacking a posterior pocket; 3)
malar fossa small, shallow in depth, anteroventrally
oriented, and well separated from DPOF; 4) relative
PBL narrow (ratio of PBL to UTRL about 0.08);
5) infraorbital foramen positioned posteriorly, about
over Ml; 6) cement layer moderately thick on de-
ciduous premolars and permanent cheek teeth; 7)
P3-M2 protocone occlusal outlines round (ratio of
PRL to PRW < 1.2) and protocones connect with
protolophs in very early wear; 8) upper cheek teeth
metastyles common but not well developed; 9) P2-
M3 external fossette plications rarely present, if
present nonpersistent, and internal fossette plica-
tions very simple; 10) P2-M2 hypoconal grooves
close in early wear; 11) P3-M2 hypoconal lakes do
not form with closure of hypoconal grooves; 12)
P2-M3 moderately curved (ROC about 50 mm);
13) dp3-4 and p3-m3 protostylids moderately well
developed; 14) p2-m3 metaconids and metastylids
well separated only in early wear and p3-m3 meta-
conid-metastylid complexes expanded but not
elongated (MML = 45-50% of APL); 15) P2-m3
ectoflexids moderately deep, only partially pene-
trating isthmuses between metaconids and metas-
tylids; 16) p3-4 metastylids and metaconids equal
or subequal in size and ml-3 metastylids and me-
taconids equal or subequal in size and position of
their lingual borders; 17) size moderately large
(UTRL = 147 mm); and 18) metacarpal V articulates
primarily with unciform carpal.

ETYMOLOGY. Heteros, Greek for other or dif-
ferent; in reference to proposed relations with
pliohippine horses.

DISCUSSION. The recent discovery of two skulls
from the middle beds of the Caliente Formation
exposed in the Nettle Spring Canyon area, one from
an immature individual and one from an adult with
associated dentaries and partial foreleg, allows re-
evaluation of the taxonomic assignment of these
specimens. Kelly and Lander (1992) tentatively as-
signed the two skulls to '‘‘‘Dinohippus” n. sp. be-
cause they exhibit upper cheek teeth with simple
occlusal patterns and a small shallow malar fossa
that is well separated from the DPOF, somewhat
similar to those of “ Dinohippus ” interpolatus.
However, further study of these specimens and the
cladistic analyses presented below indicate that they
represent a distinct clade more closely related to
the pliohippines ( Pliohippus and Astrohippus).

Heteropliohippus is derived relative to Pliohip-
pus s.s. by having the following character states: 1)
the DPOF is unpocketed and bounded anteriorly
by a relatively distinct rim and the malar fossa is
small, shallow, and unpocketed; 2) the relative PBL
is narrow (ratio of PBL to UTRL about 0.08); 3)
the infraorbital foramen is positioned posteriorly,
about below Ml; 4) the upper cheek teeth are mod-
erately curved; 5) the protostylids are moderately
well developed; and 6) the metaconids and metas-
tylids are about equal in size and position. Addi-
tional characters exhibited by Heteropliohippus that
distinguish it from Pliohippus s.s. are as follows:
1) the cement on the cheek teeth is thinner; 2) the
P3-M2 protocones connect with the protolophs at
a later wear stage; and 3) P3-M2 hypoconal lakes
do not form with closure of the hypoconal grooves.
The above distribution of character states and the
cladistic analysis presented below indicate that Het-
eropliohippus is not referable to Pliohippus s.s.

Heteropliohippus differs from Astrohippus by
having the following characters: 1) the posterior
margin of DPOF with a pronounced rim; 2) the
malar fossa is small, shallow, and well separated
from the DPOF; 3) the cement layer on the per-
manent cheek teeth is thinner; 4) the protocones
are oval-shaped; 5) the P3-M2 protocones connect
with the protolophs in very early wear; 6) the Ml-
3 plis caballin are common, but small and nonper-
sistent; 7) the p3-m3 protostylids are well devel-
oped; and 8) the p2-4 ectoflexids are moderately
deep and partially penetrate the isthmuses between
the metaconids and metastylids. Evander (1993) re-
cently hypothesized that Astrohippus may possess
the autapomorphic character state of abbreviated
metapodials and slender, elongated phalanges.
Evander’s hypothesis was based on his analysis of
faunas containing Astrohippus and not supported
by definitive evidence. However, if proven true with
further study, then this autapomorphy would fur-
ther differentiate Heteropliohippus from Astrohip-
pus.

Heteropliohippus differs from Dinohippus by
having the following characters: 1) the DPOF is
relatively deep and has a distinct rim at the anterior
margin; 2) the relative PBL is narrow (ratio of PBL
to UTRL = 0.08); 3) the cement layer on deciduous
premolars and the permanent cheek teeth is mod-
erately thick; 4) the P3-M2 protocones connect
with the protocones at an earlier wear stage; 5)
upper cheek teeth internal fossette plications are
very simple; 6) the P2-M2 hypoconal grooves close
in early wear; 7) the dp3-4 and p3-m3 protostylids
are moderately well developed; 8) the p2-4 ecto-
flexids are moderately deep and partially penetrate
the isthmuses between the metaconids and metas-
tylids; 9) the ml-3 metastylids and metaconids are
equal or subequal in size and position of their lin-
gual borders; and 10) the size is moderate (UTRL
= 147 mm).

It is well recognized that many equid dental char-
acters exhibit a moderate degree of intraspecific and

Contributions in Science, Number 455

Kelly: Miocene Horses ■ 5

ontogenetic variation (MacFadden, 1984a). At
higher taxonomic levels, such as generic or tribal,
equid dental characters are also prone to homo-
plasy and reversal (Hulbert, 1989; Hulbert and
MacFadden, 1991). However, in studies where large
equid samples were available, certain facial and
dental morphologies were found to be conserva-
tive; that is, they exhibited low degrees of intra-
specific or intrageneric variation. For example,
MacFadden (1984a) performed a comprehensive
statistical analysis on a large quarry sample of Hip-
parion tehonense (Merriam, 1916) to determine the
amount of individual, sexual, and ontogenetic vari-
ation within the sample. He also studied pooled
samples of other equid genera and species.
MacFadden’s (1984a) study resulted in the follow-
ing conclusions: 1) facial fossae morphology is not
significantly influenced by individual variation, sex-
ual dimorphism, or ontogeny, and facial fossae are
taxonomically valid character complexes for ge-
neric determination; 2) qualitative characters of the
upper cheek teeth are not affected significantly by
sexual dimorphism, but many are significantly af-
fected by ontogeny; and 3) most measured char-
acters of the upper cheek teeth are taxonomically
valid. The taxonomic significance of facial fossae
morphology has also been well documented by oth-
er investigators (e.g. Woodburne, 1982, 1989; J.P.
Quinn, 1984; Hulbert, 1988a, 1988b, 1989; Kelly
and Lander, 1988b; Hulbert and MacFadden, 1991).
Although MacFadden (1984a) demonstrated that
many qualitative characters of the upper cheek teeth
are prone to ontogenetic variation, Hulbert (1988a,
1988b, 1989) and Hulbert and MacFadden (1991)
have shown that, after ontogenetic variation has
been accounted for, equid genera can be distin-
guished by a suite of qualitative dental characters.
Because the sample size of Heteropliohippus is small,
the amount of ontogenetic variation for many of
its cheek teeth characters remain undetermined.
However, certain qualitative cheek teeth character
states of Heteropliohippus can be confidently com-
pared with those of other equid genera. For ex-
ample, in Pliohippus s.s., Dinohippus, and Para-
pliohippus, after initial wear, the ml -3 metastylids
are consistently smaller and more labially posi-
tioned than the metaconids, whereas in the sample
of Heteropliohippus, which includes moderately
worn lower molars, the metastylids are about equal
in size and position. In Astrohippus, the p2-4 ec-
toflexids are shallow and do not penetrate the isth-
muses between the metaconids and metastylids re-
gardless of the amount of wear, whereas in the
sample of Heteropliohippus, which includes mod-
erately worn lower premolars, the ectoflexids par-
tially penetrate the isthmuses. Certain other quali-
tative cheek teeth characters are either absent or
present in a genus and, thus, are not affected by
ontogeny. For example, in Pliohippus s.s., Astro-
hippus, Dinohippus, and Parapliohippus, protos-
tylids are absent or may be very weakly developed
as small anterior cingulids near the base of the

crowns, whereas in Heteropliohippus, moderately
well-developed protostylids are present that extend
well up from the base of the crowns. In Pliohippus
s.s., Dinohippus, and Astrohippus, the cement on
the upper cheek teeth is significantly thicker than
that of Heteropliohippus. Even if some of the qual-
itative cheek teeth characters listed in the diagnosis
of Heteropliohippus have to be modified when a
larger sample is available, its distinctive facial mor-
phology, a character complex that has been shown
to be taxonomically significant and not strongly
influenced by ontogeny, sexual dimorphism, or in-
dividual variation in other equid genera, and the
qualitative dental characters noted above still sup-
port recognizing Heteropliohippus as generically
distinct from all other Neogene hypsodont horses.
This conclusion is further supported by the cladistic
analyses presented below, which indicate that Het-
eropliohippus represents a generically distinct clade
that together with Pliohippus s.s. and Astrohippus
form a monophyletic lineage, wherein Pliohippus
s.s. is the closest sister taxon to Heteropliohippus,
and Heteropliohippus is the closest sister taxon to
Astrohippus.

Heteropliohippus hulberti
new species
Figure 2, Tables 2-3

“ Dinohippus ” n. sp. Kelly and Lander, 1992 (p. 4,

appendix 1).

HOLOTYPE. Associated partial skull with RP2-
M3 and partial LdPl, P2-4, M2-3, partial dentaries
with Rp2-m3 and Lp2-m3, partial distal radius,
carpals, partial proximal metapodials, first medial
phalanx, and second medial phalanx, LACM
133452.

TYPE LOCALITY. LACM 6106.

DIAGNOSIS. Same as for genus.

ETYMOLOGY. Named in honor of Richard C.
Hulbert, Jr., of the Georgia Southern University in
recognition of his many contributions to our un-
derstanding of the phylogeny of Neogene horses.

REFERRED SPECIMENS. Partial immature skull
with right and left dPl-4, LACM 134494; partial
upper left cheek tooth, LACM 136055; partial up-
per left cheek tooth, LACM 136056; associated
LdP3-4 and Rdp2, LACM 136054.

DISTRIBUTION AND AGE. Same as for genus.

DESCRIPTION. In the holotype partial skull of
Heteropliohippus hulberti, the left facial region is
preserved from the preorbital bar anteriorly to just
above the P2 and the right facial region from the
posterior aspect of the DPOF anteriorly to just above
the dPl (Figure 2A). Although the posterior portion
of the referred immature skull is badly fractured,
the right and left facial regions are well preserved.
MacFadden (1984a) demonstrated that facial fossa
morphology is not significantly affected by ontog-
eny; thus, the immature skull can be confidently
assigned to H. hulberti because its facial fossae

6 ■ Contributions in Science, Number 455

Kelly: Miocene Horses

Figure 2. Heteropliohippus hulberti n. gen. and sp. A-D, Holotype, LACM 133452: A, partial skull, right lateral view;
B, partial right dentary, lateral view; C, RP2-M3 and partial LdPl, P2-4, M2-3, occlusal view; D, Rp2-m3 and Lp2-
m3, occlusal view. E, Immature skull, LACM 134494, right lateral view. F, RdPl-4, LACM 134494, occlusal view.
Scale = 10 mm.

morphology closely matches that of the holotype.
The facial morphology is characterized by a mod-
erately deep DPOF that is well separated from a
small distinct malar fossa. The preorbital bar is
partially damaged in the holotype but appears to
have been narrow in width, as is the condition in
the referred skull (Figure 2E).

The DPOF (Figures 2A, 2E) is characterized by
having the following: 1) the dorsal and posterior
margins are formed by a continuous, distinct,
rounded rim; 2) the anterior margin is formed by
a low distinct, rounded rim; 3) the ventral margin
is confluent anteriorly with the face, whereas pos-
teriorly it is separated from the malar fossa by a
prominent ridge; 4) the depth is moderate (12-14
mm); 5) a pocket is lacking in the posterior aspect;
and 6) the shape is an elongated oval.

The malar fossa (Figures 2A, 2E) is characterized
by having the following: 1) the dorsal margin is
formed by a distinct ridge separating it from the
DPOF; 2) the posterior margin is bounded by a
low, rounded ridge; 3) the anterior margin is an-
teroventrally oriented and confluent with the facial
crest; 4) the depth is shallow (5-6 mm); 5) a pocket
is lacking in the posterior aspect; and 6) the shape
is oval (APL = 19 mm, TR = 17 mm).

The deciduous PI (Figure 2F) is relatively large
in the immature skull and likewise in the adult as
indicated by the broken crown in the holotype. The
deciduous P2-4 (Figure 2F) are characterized by
the following: 1) the cement layer is moderately
thick; 2) the mesostyles are distinct, but not prom-
inent; 3) the external fossette margins are very sim-
ple with only a slight indication of plis protoloph;

Contributions in Science, Number 455

Kelly: Miocene Horses ■ 7

Table 2. Measurements (in mm) of upper dentition of
holotype of Heteropliohippus hulberti n. sp. from the
Caliente Formation, a = approximate.

Dimension

Right

Left

APL

29.7

29.3

—

21.9

APL

26.7a

26.8

—

27.0

APL

—

26.9

26.0a

26.1

APL

26.5a

—

23.5a

—

APL

23.8

23.7

24.0a

23.9

APL

24.3

24.6

23.6

23.7

P2-4

APL

78.3

78.2a

Ml-3

APL

—

69.3

P1-M3

APL

—

153.5a

UTRL

1 46.3a

146.8

4) the internal fossette margins are very simple with
single plis postfossette that have shallow rounded
outlines; 5) the occlusal outlines of the dP2-3 pro-
tocones are round, whereas the occlusal outline of
the dP4 protocone is an elongated oval (probably
wear-related); 6) the protocones connect with the
protolophs in early wear; 7) the plis caballin are
very small indentations; and 8) the hypoconal
grooves are shallow and nonpersistent, being com-
pletely lost in early wear.

The permanent upper cheek teeth (Figure 2C) are
characterized by having the following: 1) the ce-
ment layer is moderately thick; 2) the mesostyles
are distinct, but not prominent; 3) the internal fos-

Table 3. Measurements (in mm) of lower dentition of
holotype of Heteropliohippus hulberti n. sp. from the
Caliente Formation, a = approximate.

Dimension

Right

Left

APL

24.0a

23.9

—

15.8

APL

24.5

24.7

16.2

APL

25.0

24.8

16.1

APL

22.2

22.6a

15.3

15.5

APL

25.0a

25.4

13.9

APL

28.0a

28.1

12.5

12.6

p2-4

APL

72.4

74.0a

ml-3

APL

73.7a

72.5

p2-m3

APL

147.4

147.0a

sette margins are very simple with the plis postfos-
sette single, weakly developed, and nonpersistent;
4) the external fossette margins are very simple with
the plis protoloph absent or weakly expressed as
small nonpersistent indentations; 5) the hypoconal
grooves are weakly developed and close in early
wear; 6) the plis caballin are small and nonpersis-
tent; 7) the protocones connect with the protolophs
in early wear; and 8) the occlusal outlines of the
protocones are round in the premolars and pro-
gressively become slightly elongated from the Ml
to the M3 (probably wear-related).

Although the dentary anterior to the Rp2 in the
holotype is slightly damaged, there is no alveolus
or root to indicate that a dpi was present. The
lower premolars and molars (Figure 2D) are char-
acterized by the following: 1) the cement layer is
moderately thick; 2) the metaconids and metastyl-
ids are not well separated and the metastylids are
equal in size or slightly smaller than the metaconids;
3) the p3-m2 protostylids are moderately well de-
veloped; and 4) the ectoflexids are moderately deep
and partially penetrate the isthmuses between the
metaconids and metastylids.

The partial lower foreleg of the holotype of Het-
eropliohippus hulberti has the carpals and anterior
portion of the metacarpals preserved. A distinct
articulation facet is present on the unciform carpal
for metacarpal V. The partial metacarpals II and
IV are well developed, which may indicate that the
manus was tridactyl. The associated central first and
second phalanges are of normal equid proportions.

The measurements of the teeth of Heteroplioh-
ippus hulberti are presented in Tables 2 and 3.

DISCUSSION. Merriam (1915) described Pro-
tohippus tehonensis based on a single, well-worn
LM1 (UCMP 21779) from the Santa Margarita For-
mation, Tejon Hills, California. Merriam (1916) also
tentatively referred a lower premolar to this species.
Stock (1935) provisionally assigned a partial right
dentary with a broken dp2, dp3-4, and ml
(LACM[CIT] 1825) from a well core in the Santa
Margarita Formation to this species. Drescher (1941)
referred two additional specimens from the Tejon
Hills to this species, a Lp2-4 (LACM[CIT] 2617)
and a Lml-3 (LACM[CIT] 2618), but regarded this
taxon as belonging to the genus Pliohippus. Savage
(1955) also assigned this species to Pliohippus. Ad-
ditional material of Pliohippus tehonensis from the
early Clarendonian Mathews Ranch Local Fauna
has been identified by James (1963) and Kelly and
Lander (1992). Hulbert (1987a) regarded this taxon
as a separate clade from Pliohippus s.s. and noted
that it exhibits the proper mixture of plesiomorphic
and apomorphic character states that could allow
it to be the sister taxon of some segment of the
Astrohippus-Equus-Dinohippus clade. Recently,
Hulbert (1993, fig. 1) referred this species to “DL
nohippus ” tehonensis. However, the generic status
of this species cannot be confidently determined
because the facial morphology is unknown (Hul-
bert, 1987a). For this reason, it is herein referred

8 ■ Contributions in Science, Number 455

Kelly: Miocene Horses

to “ Pliohippus ” tehonensis. The cheek teeth of
Heteropliohippus hulberti are morphologically
similar to those of “ Pliohippus ” tehonensis, but
they differ from those of “ Pliohippus ” tehonensis
by having the following characters: 1) the upper
cheek teeth have less distinct hypocones, less per-
sistent hypoconal grooves, and less anteroposterior
elongation of the protocones; 2) the lower cheek
teeth are relatively wider transversely and have bet-
ter developed protostylids; and 3) the lower molars
have shallower ectolophids and the metastylids and
metaconids are equal or subequal in size and po-
sition of their lingual borders. The above differ-
ences clearly indicate that Heteropliohippus hul-
berti and “ Pliohippus ” tehonensis represent differ-
ent species, but whether they belong to the same
genus cannot be determined because the facial mor-
phology of “ Pliohippus ” tehonensis is unknown.
However, if the facial morphology of “ Pliohippus ”
tehonensis is determined with future discoveries to
be similar to that of Heteropliohippus hulberti, then
Heteropliohippus hulberti and “ Pliohippus ” teho-
nensis could conceivably be derived from a com-
mon ancestor and tehonensis would

be referable to Heteropliohippus.

Drescher (1941) described Pliohippus leardi based
on the holotype, an isolated LM1 (LACM[CIT]
2645), and a small topotypic sample of upper and
lower cheek teeth from the Chanac Formation, Te-
jon Hills, California. Savage and Russell (1983) and
Hulbert (1993) refer to this species as “ Dinohip -
pus ” leardi. Hulbert (1993, fig. 1) regarded “P/zo-
hippus ” tehonensis as the closest sister taxon to
“ Dinohippus ” leardi. Hulbert (1993) also indicated
that “ Dinohippus ” leardi gave rise to Astrohippus
by cladogenetic speciation and was also the inferred
ancestor that gave rise by cladogenetic speciation
to the Hippidion-Onohippidium clade and to the
“ Dinohippus ” interpolates- ‘Dinohippus ” mexi-
canus-Equus simplicidens clade. Like “ Pliohip-
pus” tehonensis, the generic status of “ Dinohip-
pus” leardi cannot be determined confidently be-
cause the facial morphology is unknown. Hetero-
pliohippus hulberti differs from “ Dinohippus ” leardi
by having the following characters: 1) the cheek
teeth are smaller; 2) the protocones are less an-
teroposteriorly elongated; 3) the hypoconal grooves
are less developed and disappear in an earlier wear
stage; and 4) the lower molars have moderately
well-developed protostylids and shallower ectolo-
phids, and the metastylids and metaconids are equal
or subequal in size and position of their lingual
borders.

Tribe Undetermined

Acritohippus new genus

Hippotherium, in part, Cope, 1889 (pp. 451-454).
Merychippus, in part, Osborn, 1918 (pp. 101-102,
105).

Stylonus, in part, Kelly and Lander, 1988b (p. 4),
1992 (p. 3, appendix 1).

TYPE SPECIES. A. isonesus (Cope, 1889) (pre-
viously referred to Hippotherium, Merychippus,
and Stylonus).

DISTRIBUTION AND AGE. California: Cal-
iente Formation, latest Hemingfordian to late Bar-
stovian; Bopesta Formation, late Barstovian; Bar-
stow Formation, latest Hemingfordian. Florida:
Torreya Formation, early Barstovian. Oregon: Mas-
call Formation, early Barstovian; Sucker Creek For-
mation, late Barstovian. Nebraska: Box Butte For-
mation, late Hemingfordian; Sheep Creek Forma-
tion, late Hemingfordian. Nevada: Virgin Valley
Formation, early Barstovian; Highrock Canyon Se-
quence, early Barstovian. Montana: Six Mile Creek
Formation, early Barstovian.

REFERRED SPECIES. A. tertius (Osborn, 1918)
(previously referred to Merychippus ); A. quinni n.
sp.

DIAGNOSIS. Acritohippus is distinguished from
all other Neogene hypsodont horses by having the
following suite of characters: 1) frontal bones flat;
2) facial crest dorsoventrally compressed; 3) DPOF
shape oval, depth shallow to deep (5 to > 15 mm),
anterior margin confluent with face, ventral margin
lacking pronounced rim, posterior margin with dis-
tinct rim, and posterior pocket shallow or absent;

4) malar fossa shallow in depth (< 10 mm) and
confluent with DPOF; that is, malar fossa and DPOF
only separated posteriorly by low, indistinct ridge;

5) relative PBL very narrow (ratio of PBL to UTRL
about 0.05); 6) muzzle width relative to UTRL broad
(> 36%); 7) relative muzzle length short to mod-
erate (UDL < 55% of UTRL); 8) cement layer thin
on deciduous premolars and thick on permanent
cheek teeth; 9) P3-M2 protocone occlusal outlines
oval (ratio of PRL to PRW = 1. 2-2.0); 10) P2-4
protocones connect with protolophs in early mod-
erate wear and Ml-2 protocones connect with pro-
tolophs in early to early moderate wear; 11) upper
cheek teeth metastyles common but not well de-
veloped; 12) Ml-2 protocones connect with hy-
pocones in late wear; 13) P2 anterostyle large and
unexpanded; 14) P2-M3 plis caballin well devel-
oped, single, and relatively persistent; 15) P2-M3
external and internal fossette plications simple and
relatively nonpersistent; 16) P2-M3 hypoconal
grooves close in moderate to late wear; 17) P3-M2
hypoconal lakes do not form with closure of hy-
poconal grooves; 18) dpi very rarely present, ves-
tigial if present; 19) dp3-4 and p3-m3 protostylids
absent or may be present only near base of crowns
as anterior cingulids; 20) p3-m3 metaconids and
metastylids well separated only in very early to early
wear; 21) p2 ectoflexid moderately deep, partially
penetrates isthmus between metaconid and meta-
stylid; 22) p3-4 ectoflexids deep, completely pen-
etrate isthmuses between metaconids and metas-
tylids; 23) p3-4 metastylids and metaconids equal
or subequal in size and ml -3 metastylids and me-
taconids are equal or subequal in size and position
of their lingual borders; and 24) feet tridactyl.

ETYMOLOGY. Acritos, Greek for mixed or

Contributions in Science, Number 455

Kelly: Miocene Horses ■ 9

confused, in reference to morphological similarities
to both equines and hipparionines; Hippos, Greek
for horse.

DISCUSSION. Hulbert (1988b, 1989) and Kelly
and Lander (1988b) recognized that “ Merychippus ”
isonesus and “ Merychippus ” tertius represent a dis-
tinct clade of Neogene hypsodont horses. Downs
(1956) regarded “M.” isonesus as a junior synonym
of “ Merychippus ” seversus (= Stylonus seversus
Cope, 1879). Accepting Downs’s (1956) synonymy
of “M.” isonesus with “M.” seversus and recog-
nizing that these taxa represent a distinct clade not
referable to Merychippus s.s ., Kelly and Lander
(1988b) assigned them to Stylonus. However, Hul-
bert and MacFadden (1991) noted that S. seversus
represents a different species from “M.” isonesus
and rejected their synonymy. I now agree with Hul-
bert and MacFadden (1991) that these two horses
are not synonymous and, furthermore, regard the
holotype of S. seversus, an isolated upper molar
(AMNH 8180), as specifically indeterminate.
Therefore, S. seversus is a nomen dubium and,
because of the invalidity of the species, Stylonus is
also a nomen dubium.

Although Hulbert (1988b, 1989) and Kelly and
Lander (1988b) recognized “ Merychippus ” isone-
sus and “ Merychippus ” tertius as a distinct clade,
they also included “ Merychippus ” sejunctus (Cope,
1874) in this clade. Hulbert and MacFadden (1991)
demonstrated that “M.” sejunctus and its anage-
netic ancestor “ Merychippus ” sp. near “M.” se-
junctus do not form a monophyletic clade with
“M.” isonesus and “M. tertius but, instead, rep-
resent a separate clade of probable generic rank.
Based on their cladistic analysis, Hulbert and
MacFadden (1991) provisionally included the “M.”
isonesus-tertius clade in the Hipparionini but not-
ed other slightly less parsimonious phylogenetic ar-
rangements, wherein the “M.” isonesus-tertius clade
was the sister group of the Equini plus Hipparionini,
the sister group of the protohippines plus the Hip-
parionini, or the sister group of the pliohippines.
The cladistic analyses presented below also indicate
that the “M.” isonesus-tertius clade is monophy-
letic and represents a generically distinct group of
horses. Based on the cladistic analysis presented
below and those of Hulbert (1989) and Hulbert
and MacFadden (1991), “M.” isonesus , “M.” ter-
tius, and the new species described below are herein
assigned to Acritohippus n. gen.

Acritohippus is derived relative to Pliohippus s.s.
by having the following character states: 1) the ma-
lar fossa is confluent with the DPOF; 2) the relative
PBL is much narrower; 3) the P2-M3 internal fos-
sette plications are slightly more complex and re-
sistant to wear; 4) the plis caballin are better de-
veloped; and 5) the connection of the protocones
to the protolophs occurs in later wear. Additional
characters exhibited by Acritohippus that distin-
guish it from Pliohippus s.s. are as follows: 1) the
cement on the deciduous premolars is thinner; 2)
the dPl is less reduced; 3) the upper cheek teeth
are less curved and less hypsodont; 4) the hypoconal

grooves close at a later wear stage and hypoconal
lakes do not form with closure of the grooves; 5)
the p2-4 ectoflexids are deeper, penetrating the
isthmuses between the metaconids and metastylids;
and 6) the ml -3 metastylids are about equal in size
and position. The above distribution of character
states clearly indicates that Acritohippus represents
a separate genus from Pliohippus s.s.

Although Acritohippus exhibits some morpho-
logical similarities to other pliohippines and the
“ Merychippus ” sp. near “M.” sejunctus clade, it
can be easily distinguished from them. Acritohip-
pus differs from Parapliohippus by having the fol-
lowing characters: 1) a DPOF pocket is commonly
lacking and if present it is relatively shallow in depth;
2) the malar fossa is shallow in depth, crescent-
shaped, and confluent with the DPOF; that is, the
DPOF and the malar fossa are only separated pos-
teriorly by a low, indistinct ridge; 3) the relative
muzzle length is short to moderate (UDL < 55%
of UTRL); 4) the cement layer on the deciduous
cheek teeth is moderately thick; 5) the P3-M2 pro-
tocones connect with the protolophs in early mod-
erate wear; 6) the P2-M3 plis caballin are well
developed and relatively persistent; 7) the P2-M3
internal fossettes have simple plications but are
slightly more complex and persistent relative to
those of Parapliohippus ; 8) P3-4 hypoconal lakes
do not form with closure of the hypoconal grooves;
9) the ml-3 metastylids and metaconids are equal
or subequal in size and position of their lingual
borders; and 10) the UTRL is moderate (105-140
mm). Acritohippus differs from Heteropliohippus
by having the following characters: 1) the malar
fossa is confluent with the DPOF; 2) the P3-M3
protocones connect with the protolophs in later
wear; 3) the P2-M3 plis caballin are much better
developed and relatively persistent; 4) the P3-M3
internal fossette plications have simple plications
but are more complex and persistent relative to
those of Heteropliohippus ; 5) the P2-M2 hypo-
conal grooves close in later wear; 6) the p3-m3
protostylids are absent or weakly developed; and
7) the p3-4 ectoflexids are deeper and completely
penetrate the isthmuses between the metaconids
and metastylids. Acritohippus differs from “Mcr-
ychippus ” sp. near “M.” sejunctus, a separate clade
of probable generic rank, by having the following
characters: 1) the relative muzzle length is shorter
(UDL < 55% of UTRL); 2) the connection of the
P3-4 protocones to the protolophs occurs in later
wear; 3) the P3-M3 internal fossette plications are
simpler and less persistent; 4) the p3-m3 protos-
tylids are absent, or when present are much less
developed; and 5) the p3-m3 metaconids and me-
tastylids are less well separated.

Acritohippus quinni new species

Figure 3, Tables 4-6

Merychippus sumani Merriam, in part, Gazin, 1930
(pp. 50, 62, 69-72, figs. 2-4).

10 ■ Contributions in Science, Number 455

Kelly: Miocene Horses

Figure 3. Acritohippus quinni n. gen. and sp. A-D, Holotype, UCMP 65338: A-B, partial skull, right and left lateral
views; C, RdPl, P2-M3, occlusal view; D, broken LdPl, P2-M3, occlusal view. E, LP2-M3, UCMP 50750, occlusal
view. F, Lp2-m3, LACM 134493, occlusal view. Scale for A-B = 10 mm, C-F = 10 mm.

Merychippus sumani Merriam, in part, James, 1963
(pp. 12, 19, 26-27, tab. 2).

“ Merychippus ” cf. “M.” stylodontus (Merriam)
Woodburne vide Bernor et ai, 1980.

“ Pliohippus ” sp. J.P. Quinn, 1984 (pp. 199-209,
figs. 48-50, tab. 6).

“ Merychippus ” n. sp. J.P. Quinn, 1987 (pp. 23, 27,
tab. 1).

Stylonus n. sp. Kelly and Lander, 1988b (p. 4), 1992
(p. 3, appendix 1).

HOLOTYPE. Partial skull with RdPl, RP2, par-
tial RP3, RP4-M3 and partial LP2, LP3, partial
LM1, LM2-3, UCMP 65338.

TYPE LOCALITY. UCMP V-5823.

DIAGNOSIS. Acritohippus quinni differs from
A. isonesus and A. tertius by having the following
characters: 1) larger size (mean UTRL = 127.3 mm);
2) deeper DPOF (> 15 mm); 3) shorter relative
muzzle length (UDL = 32% of UTRL); 4) more
hypsodont cheek teeth (about 35 mm); and 5) less
curvature of upper cheek teeth (ROC = 45-50 mm).
Further differs from A. isonesus by having the fol-
lowing characters: 1) DPOF with deeper posterior
pocket (> 5 mm); 2) P2-M2 hypoconal grooves
close in earlier wear stage; and 3) P2-M3 fossette
plications slightly less developed. Further differs
from A. tertius by having DPOF posteriorly pock-

Contributions in Science, Number 455

Kelly: Miocene Horses ■ 11

Table 4. Measurements (in mm) of upper dentition of
holotype of Acritohippus quinni n. sp. from the Caliente
Formation, a = approximate.

Dimension

Right

Left

A-P

25.6

25.7

21.2

21.4

A-P

22.2a

—

24.8

—

A-P

22.4

22.6

24.6

24.5

A-P

—

19.5

—

22.7

A-P

21.1

21.3

21.1a

21.7

A-P

20.9

20.3a

19.6

20.1

P2-4

APL

69.3

69.5

Ml-3

APL

61.8

61.5

P1-M3

APL

135.5

UTRL

127.2

127.0

eted and Ml -2 protocones connecting with pro-
tolophs in later wear stage.

ETYMOLOGY. Named in honor of James Pat-
rick Quinn, a research associate of the Natural His-
tory Museum of Los Angeles County, who first
recognized the distinctive characters of this species.

REFERRED SPECIMENS. Partial skull with
RP2-M3 and LP2-M3, UCMP 52525; associated
partial skull with RP2-M3 and LP2-M3, right and
left dentaries with il-m3, and appendicular ele-
ments, UCMP 51000; partial left maxilla with P2-
M3, UCMP 50667; partial left maxilla with P2-
M3, UCMP 50750; partial left maxilla with P2-
M3 and associated right dentary with p2-m3, UCMP
50950; partial left maxilla with P4-M3, UCMP
51180; partial right maxilla with dPl and P2-M3,

UCMP 51300; partial right dentary with p3-m3,
UCMP 50680; associated partial dentaries with Ri2-
m3 and Lil-m3, UCMP 51230; associated partial
dentaries with Rc-il and Lil~m3, UCMP 51260;
partial right dentary with p2-m3, UCMP 52525;
partial skull with RdPl, P2-M3, and LP2-M3,
LACM 15625; skull with RdP4, Ml -2, and partial
LdP3-4 and Ml-2, LACM 134495; associated par-
tial dentaries with Rp3-m3 and Lp2-m3, LACM
134493; associated partial dentaries with Rp3-m3
and Lp2-m3, LACM 138112; partial left dentary
with p2--m3, LACM 138075.

DISTRIBUTION AND AGE. California: West
Dry Canyon, Lower Dome Spring, Upper Dome
Spring, and Doe Spring Canyon Local Faunas, Cal-
iente Formation, Cuyama Valley Badlands, latest
Hemingfordian to late Barstovian; “ Merychippus ”
cf. “M.” intermontanus Range Zone, Bopesta For-
mation, southern Sierra Nevada, late Barstovian.

DESCRIPTION. The facial morphology is well
preserved in several skulls and is characterized by
a moderately deep, posteriorly pocketed DPOF that
is confluent with the malar fossa; that is, the DPOF
and the malar fossa are only separated posteriorly
by a low, rounded, indistinct ridge (Figures 3A-3B).
The preorbital bar between the DPOF and the orbit
is very narrow (6 mm). The buccinator fossa is a
moderately developed depression on the anterior
maxilla and is separated posterodorsally from the
DPOF by a very low, indistinct ridge. The facial
crest is dorsoventrally compressed. The rostrum is
relatively short with the C-P2 diastema about 23-
32 mm in length.

The DPOF is characterized by having the follow-
ing: 1) the dorsal margin is formed by a distinct,
rounded rim; 2) the posterior margin is formed by
a distinct, sharply edged rim; 3) the anteroventral
margin is confluent with the face and somewhat
constricted by a lateral expansion of the face above
the P4-M1; 4) the posteroventral margin is formed

Table 5. Summary of measurements (in mm) of upper dentition of Acritohippus quinni n. gen. and sp. from the
Caliente Formation.

Dimension

Mean

APL

23.6-26.7

25.1

1.4

5.4

19.2-22.3

20.5

1.2

5.8

APL

21.6-23.2

22.3

0.7

3.2

20.6-24.8

22.7

1.5

6.7

APL

20.4-22.5

21.9

0.7

3.4

22.2-24.6

23.2

0.9

3.6

APL

19.4-20.7

20.2

0.6

3.2

21.9-24.7

22.9

0.9

4.1

APL

20.0-21.2

20.6

0.4

2.0

20.6-23.6

22.0

1.0

4.4

APL

20.5-22.3

21.0

0.7

2.9

19.1-21.7

20.0

0.8

4.0

P2-4

APL

61.3-74.1

68.9

4.7

6.8

Ml-3

APL

59.5-64.4

61.4

2.0

3.3

P1-M3

APL

135.5

UTRL

125.2-130.4

127.3

—

12 ■ Contributions in Science, Number 455

Kelly: Miocene Horses

Table 6. Summary of measurements (in mm) of lower dentition of Acritohippus quinni n. gen. and sp. from the
Caliente Formation.

Dimension

Mean

APL

20.5-22.4

22.1

1.2

5.6

10.9-12.4

11.7

0.6

5.1

APL

19.4-22.5

21.6

1.0

4.7

12.0-13.9

13.1

0.8

5.7

APL

20.2-22.7

21.8

1.4

6.5

11.9-13.4

12.8

0.6

4.4

APL

19.4-22.6

20.8

1.3

6.1

10.6-12.6

11.7

0.7

5.7

APL

19.1-23.0

21.5

1.4

6.5

10.0-11.2

10.6

0.6

5.5

APL

23.2-25.4

24.5

0.9

3.5

8.5-10.6

9.4

0.8

8.0

p2-4

APL

61.5-66.7

65.0

2.1

3.2

ml-3

APL

61.0-66.8

64.5

2.3

3.6

p2-m3

APL

120.3-133.2

128.6

4.9

3.8

by a low, indistinct, rounded ridge; 5) the shape is
elongate oval; 6) the depth is moderately deep (>
15 mm); and 7) the posterior aspect is pocketed (6-
8 mm deep).

The malar fossa is characterized by the following:
1) the ventral and posterior margins are bounded
by distinct ridges; 2) the dorsal margin is formed
by the low, indistinct ridge separating it from the
DPOF; 3) the anterior margin is confluent with the
face; 4) the depth is shallow (about 5 mm deep);
and 5) it is not pocketed posteriorly.

The deciduous PI is always present and moderate
in size. The deciduous P2-4 are similar to the adult
premolars except that the morphology of occlusal
outlines of the fossette plications is less complex,
the plis caballin less persistent, and the protocones
connect with the protolophs at an earlier wear stage.

The permanent upper premolars and molars are
characterized by the following (Figures 3C-3E): 1)
the cement layer is moderately thick; 2) the cur-
vature is moderate; 3) the mesostyles are distinct
but not prominent; 4) the internal fossette plica-
tions are simple with the pli protoconule and pli
prefossette single and relatively nonpersistent; 5)
the external fossette plications are very simple with
very poorly developed single plis protoloph that are
nonpersistent and lacking plis hypostyle; 6) the plis
caballin are single, moderately well developed, and
relatively persistent; 7) the protocones have oval
occlusal outlines with small anterolabial spurs in
very early wear and connect with the protolophs
in early moderate wear; that is, the Ml protocone
is connected when the M3 is in early wear; 8) the
hypocones are relatively distinct forming round to
slightly elongated occlusal outlines in early wear;
and 9) the hypoconal grooves are relatively persis-
tent and close without forming hypoconal lakes in
moderate wear.

The dpi is usually absent but, when present, is
vestigial. The lower premolars and molars are char-

acterized by the following (Figure 3F): 1) the cement
layer is moderately thick; 2) the metaconids and
metastylids are only well separated in early wear
and the metastylids and metaconids are equal or
subequal in size and position of their lingual bor-
ders; 3) the preflexids and postflexids have simple
margins and, with wear, become widely separated
with the labial depth of the preflexid very shallow;
4) the ml -3 plis caballinid are commonly present
in early to moderate wear as small indentations;
and 5) the p2 ectoflexid is moderately deep and
partially penetrates the isthmus between the meta-
conid and metastylid, whereas the p3-m3 ectoflex-
ids are deep, completely penetrating the isthmuses
between the metaconids and metastylids.

The measurements of the teeth of Acritohippus
quinni are presented in Tables 4-6.

DISCUSSION. Hulbert and MacFadden (1991)
suggested that the late Hemingfordian Acritohip-
pus tertius and the early Barstovian Acritohippus
isonesus may represent endpoints in a morphocline,
wherein A. tertius gave rise to A. isonesus by an-
agenetic speciation. Furthermore, they noted that
these taxa could eventually be synonymized if this
relationship is confirmed by further study. Even if
A. tertius is eventually proven to be a junior syn-
onym of A. isonesus, Acritohippus quinni cannot
be regarded as synonymous with A. isonesus be-
cause of the following facts: 1) A. quinni exhibits
at least seven apomorphic character states relative
to A. isonesus (see Diagnosis above); 2) both A.
isonesus and A. quinni existed during the same
chronologic interval, the early to late Barstovian;
and 3) each species appears to have been geograph-
ically isolated, with A. quinni known only from
southern California and A. isonesus known pri-
marily from the Northwest and Midwest. A more
probable scenario is that the Hemingfordian A. ter-
tius or a similar plesiomorphic ancestor gave rise
to a southern West Coast population that evolved

Contributions in Science, Number 455

Kelly: Miocene Horses ■ 13

into A. quinni and a more northern population that
evolved into A. isonesus. This scenario is also sup-
ported by the observations of MacFadden and Hul-
bert (1988), Hulbert and MacFadden (1991), and
MacFadden (1992) that, during the late Heming-
fordian and early Barstovian of North America,
hypsodont horses underwent an explosive adaptive
radiation, in which many equid faunas exhibited
distinct regional endemism.

COMPARATIVE REEVALUATION OF
PLIOHIPPUS 5.5.

Pliohippus s.s. exhibits similarities in certain facial
and dental morphologies with Parapliohippus, As-
trohippus, Heteropliohippus, Acritohippus, and
“ Merychippus ” stylodontus (a monophyletic clade
of probable generic rank). In order to discuss the
relationships of Pliohippus s.s. to these taxa, a re-
view of Pliohippus is presented here. Marsh (1874,
p. 252) originally named Pliohippus based on Plio-
hippus pernix from the “Pliocene sands of the Ni-
obrara River, Nebraska.” In his diagnosis, he dif-
ferentiated Pliohippus from Protohippus and Equus
as follows: 1) Pliohippus differs from Protohippus
by the “absence of lateral digits, which are only
represented by slender splint bones”; and 2) Plio-
hippus differs from Equus by “the presence of a
large antorbital fossa, a functional upper first pre-
molar, and by a different composition of the crowns
of the upper molars.” Marsh’s only additional ref-
erence to the facial morphology of Pliohippus per-
nix was that it possesses a “deep irregular fossa in
front of the orbit.” Additional characters cited in
his description of Pliohippus pernix were as fol-
lows: “the molar teeth have very short crowns, the
folds of the enamel are very simple and there are
none in the inner lobes, the ungual phalanges are
broad, the femur has the fossa above its outer con-
dyle unusually deep, and the cuboid facet on the
astragalus is larger than most equines.” Gidley (1907)
noted that Marsh (1874) founded Pliohippus pri-
marily on the absence of lateral digits. Gidley (1907,
p. 868) regarded Pliohippus as closely related to
Protohippus, based on the shared character state
of having the “protocones and hypocones partially
or completely connected to the protolophs and
metalophs,” respectively. However, Gidley (1907,
pp. 868, 894) further noted that Pliohippus differs
from typical Protohippus by having “a large lach-
rimal fossa” and “a deep malar pit that is apparently
wanting in Protohippus .” Osborn (1918) revised
Pliohippus and listed 12 defining character states
for the genus. Although Osborn (1918) recognized
Pliohippus pernix as the genotype, he also assigned
species to Pliohippus that are now referred to Equus,
Protohippus, Dinohippus s.s., and “ Dinohippus ”
(e.g. Equus simplicidens [Cope, 1892], Equus cum-
minsii Cope, 1893, Protohippus supremus Leidy,
1869, Dinohippus leidy anus [Osborn, 1918], “Dz-
nohippus ” spectans [Cope, 1880], and “ Dinohip-

pus” interpolatus ). Nevertheless, the following de-
rived character states listed by Osborn (1918) still
are regarded by many investigators as typical for
Pliohippus: 1) a malar fossa is present; 2) the upper
cheek teeth are hypsodont and strongly curved; 3)
the protocones are oval in shape and connect with
the protolophs in very early wear; and 4) the fos-
settes are simple with few plications of the enamel
borders. However, each of these character states
also is present in other equid genera. For example,
Astrohippus (MacFadden 1984b) and Acritohippus
possess a malar fossa, and Dinohippus possesses
upper cheek teeth that have relatively strong cur-
vature, simple fossettes, and oval to elongated oval
protocones that connect with the protolophs in
very early wear (J.H. Quinn, 1955; MacFadden,
1984b). Therefore, rather than representing auta-
pomorphies of Pliohippus, the above character states
represent synapomorphies uniting Pliohippus with
certain other genera of the Equinae.

Hulbert (1989, 1993) and Hulbert and Mac-
Fadden (1991) regarded Pliohippus s.s. as being
characterized by a monophyletic lineage including
Pliohippus mirabilis (Leidy, 1858) (= Pliohippus
campestris [Gidley, 1907]), Pliohippus pernix (=
Pliohippus robustus Marsh, 1874, Pliohippus
pachyops [COPE, 1893], Pliohippus lullianus
Troxell, 1916), and Pliohippus nobilis Osborn, 1918.
They considered these species to represent a single
lineage, wherein the middle to late Barstovian Plio-
hippus mirabilis gave rise by anagenetic speciation
to the Clarendonian Pliohippus pernix, which then
gave rise by anagenetic speciation to the early Hem-
phillian Pliohippus nobilis. Additional species re-
ferable to Pliohippus s.s. are as follows: 1) Plio-
hippus fossulatus (Cope, 1893), a derived species
that exhibits a very deep compartmentalized malar
fossa (Stirton and Chamberlain, 1939; MacFadden
1984b; Hulbert, 1989); and 2) Pliohippus tantalus
(Merriam, 1913b) (= Pliohippus fairbanksi Mer-
riam, 1915), a West Coast species whose facial and
dental morphology are very similar to Pliohippus
pernix (Merriam, 1913b, 1915, 1919; Vanderhoof,
1933; Hulbert, 1988a). Pliohippus s.s., as typified
by Pliohippus pernix and characterized by the above
monophyletic lineage, exhibits the following com-
bination of derived character states relative to the
outgroup “ Parahippus ” leonensis and based on the
cladistic analyses presented below: 1) a DPOF that
is pocketed and distinctly rimmed posteriorly and
a malar fossa that is deep, pocketed posteriorly,
and well separated from the DPOF by a distinct
ridge of bone; 2) a posteriorly positioned infraor-
bital foramen, about below the middle of the P4;
3) a moderate relative PBL (ratio of PBL to UTRL
= 0.10-0.20); 4) thick cement on the deciduous
premolars; 5) moderate reduction of the dPl (ratio
of dPl APL to P2 APL about 0.30); 6) strongly
curved upper cheek teeth (ROC > 40 mm, sec-
ondarily derived); 7) hypsodont upper cheek teeth
(Ml unworn crown height > 50 mm); 8) hypoconal
lakes form with closure of the P2-4 hypoconal

14 ■ Contributions in Science, Number 455

Kelly: Miocene Horses

grooves; 9) hypoconal lakes form with closure of
the Ml-2 hypoconal grooves; 10) very simple in-
ternal fossette plications that commonly disappear
with wear; 11) reduced plis caballin; 12) connection
of the protocones to the protolophs occurs in very
early wear; 13) connection of the protocones to
the hypocones occurs prior to late wear; 14) closure
of the hypoconal grooves occurs in early wear; 15)
moderate depth of the p2-4 ectoflexids, only par-
tially penetrating the isthmuses between the me-
taconids and metastylids; and 16) the ml-3 metas-
tylids are notably smaller and positioned more la-
bially than the metaconids.

Parapliohippus, Astrohippus, Acritohippus,
Heteropliohippus, and “ Merychippus ” stylodontus
all possess a DPOF and malar fossa, but the fossa
morphology in each of these taxa differs from Plio-
hippus s.s. as follows (hypothesized polarity of each
character state included in parentheses): 1) Para-
pliohippus differs by having the DPOF extensively
developed posteriorly, resulting in the effective re-
moval of the lacrimal bone from the fossa, except
at the orbital rim (apomorphic), whereas in Plio-
hippus s.s. the lacrimal bone is not reduced ante-
riorly and extends well into the DPOF; 2) Astro-
hippus differs by having a large unpocketed DPOF
that is indistinctly separated from a large shallow
unpocketed malar fossa, which commonly contains
small concentric pits, and a very faint dorsoventral
bar that divides these fossae into anterior and pos-
terior portions (apomorphic); 3) Acritohippus dif-
fers by having a small, but distinct, shallow un-
pocketed malar fossa that is confluent with the
DPOF; that is, the fossae are only separated pos-
teriorly by a very low, indistinct ridge (apomorphic);
4) Heteropliohippus differs by having an unpock-
eted DPOF with a distinct anterior rim and a small,
shallow, unpocketed malar fossa (apomorphic re-
versal); and 5) “M.” stylodontus differs by having
an unpocketed DPOF and a shallow unpocketed
malar fossa (plesiomorphic).

In addition to the differences in facial fossa mor-
phology, each of these taxa exhibit a distinct com-
bination of character states that differs from the
combination of derived character states listed above
for Pliohippus s.s. Parapliohippus exhibits the fol-
lowing distribution of character states relative to
the 16 derived character states listed above for Plio-
hippus s.s. : 1) plesiomorphic for character state
numbers 2, 4, 5, 7, 9, 13, 14, and 15 and 2) syna-
pomorphic for character state numbers 1, 6, 8, 10,
12, and 16. Derived character states exhibited by
Parapliohippus relative to Pliohippus s.s. are as
follows: 1) extensive posterior development of the
DPOF with the anterior aspect of lacrimal bone
reduced; 2) a narrow relative PBL (ratio of PBL to
UTRL about 0.05); 3) the protolophs and metal-
ophs remain separate until the teeth are more than
50% worn; and 4) small size (UTRL = 90-100 mm).
Astrohippus exhibits the following distribution of
character states relative to the 16 derived character
states listed above for Pliohippus s.s.: 1) plesiom-

orphic for character states 8 and 9 and 2) synapo-
morphic for character states 4, 7, 10, 12, and 14.
Derived character states exhibited by Astrohippus
relative to Pliohippus s.s. are as follows: 1) a dis-
tinctive DPOF and malar fossa morphology (see
above); 2) the infraorbital foramen is positioned
farther posteriorly, about below posterior half of
P4 to Ml; 3) the relative PBL is narrow (ratio of
PBL to UTRL about 0.08); 4) the upper cheek teeth
are relatively straight (secondarily derived); 5) plis
caballin are usually absent; 6) the protocones and
hypocones connect only in late wear; 7) the P2-
M2 hypoconal grooves close in early wear; 8) the
p2-4 ectoflexids are shallow, not penetrating the
isthmuses between the metaconids and metastylids;
and 9) the ml-3 metaconids and metastylids are
about equal in size and position (secondarily de-
rived). Acritohippus exhibits the following distri-
bution of character states relative to the 16 derived
character states listed above for Pliohippus s.s.: 1)
plesiomorphic for character states 4, 5, 6, 7, 8, 9,

13, 14, 15, and 16 and 2) synapomorphic for char-
acter state 2. Derived character states exhibited by
Acritohippus relative to Pliohippus s.s. are as fol-
lows: 1) a distinctive DPOF and malar fossa mor-
phology (see above); 2) a narrow relative PBL (ratio
of PBL to UTRL about 0.05); 3) simple, but per-
sistent, internal fossette plications; 4) moderately
well-developed plis caballin; and 5) the connection
of the protocones to the protolophs occurs in early
moderate wear. Heteropliohippus exhibits the fol-
lowing distribution of character states relative to
the 16 derived character states listed above for Plio-
hippus s.s.: 1) plesiomorphic for character states 5,
6, 7, 8, 9, and 12 and 2) synapomorphic for char-
acter states 4, 10, 11, 14, and 15. Derived character
states exhibited by Heteropliohippus relative to
Pliohippus s.s. are as follows: 1) a distinctive DPOF
and malar fossa morphology (see above); 2) the
relative PBL is narrow (ratio of PBL to UTRL about
0.08); 3) the infraorbital foramen is positioned pos-
teriorly, about below Ml; 4) the upper cheek teeth
are moderately curved; 5) the p3-m3 protostylids
are moderately developed; and 6) the ml-3 meta-
conids and metastylids are about equal in size and
position (secondarily derived). “ Merychippus ” sty-
lodontus exhibits the following distribution of
character states relative to the 16 derived character
states listed above for Pliohippus s.s.: 1) plesiom-
orphic for character states 6, 7, 8, 9, and 15 and 2)
synapomorphic for character states 2, 5, 10, 11, 13,

14, and 16. Derived character states exhibited by
“M.” stylodontus relative to Pliohippus s.s. are as
follows: 1) an extremely narrow relative PBL (ratio
of PBL to UTRL about 0.035), and 2) the connec-
tion of the protocones to protolophs occurs in early
wear.

In summary, Pliohippus s.s. is derived relative to
Parapliohippus in at least 8 character states, derived
relative to Astrohippus in at least 5 character states,
derived relative to Acritohippus in at least 1 1 char-
acter states, derived relative to Heteropliohippus in

Contributions in Science, Number 455

Kelly: Miocene Horses! 15

at least 7 character states, and derived relative to
“ Merychippus ” stylodontus in at least 6 character
states. Parapliohippus is derived relative to Plio-
hippus s.s. in at least four character states and de-
rived relative to Astrohippus , Acritohippus, Het -
eropliohippus, and “M.” stylodontus in at least three
character states. Astrohippus is derived relative to
Pliohippus s.s. and Acritohippus in at least 9 char-
acter states, derived relative to Parapliohippus in
at least 11 character states, derived relative to “M.”
stylodontus in at least 8 character states, and de-
rived relative to Heteropliohippus in at least 3 char-
acter states. Acritohippus is derived relative to Plio-
hippus s.s., Parapliohippus, “M.” stylodontus, and
Heteropliohippus in at least five character states and
derived relative to Astrohippus in at least four char-
acter states. “ Merychippus ” stylodontus is derived
relative to Pliohippus s.s., Parapliohippus, Astro-
hippus, Acritohippus, and Heteropliohippus in at
least two character states. Heteropliohippus is de-
rived relative to Pliohippus s.s. in at least six char-
acter states, derived relative to Parapliohippus, Ac-
ritohippus, and “M.” stylodontus in at least four
character states, and derived relative to Astrohippus
in at least three character states. These comparisons
clearly demonstrate that paraphyly would result if
any of the other taxa ( Parapliohippus , Astrohippus,
Acritohippus, Heteropliohippus, “M.” stylodon-
tus) were assigned to Pliohippus s.s.

In conclusion, all of the above data and the cla-
distic analyses presented below and those of Hul-
bert (1989) and Hulbert and MacFadden (1991)
clearly justify recognizing Parapliohippus, Heter-
opliohippus, and Acritohippus as generically dis-
tinct from Pliohippus s.s. and all other genera of
the Equinae. “ Merychippus ” stylodontus also ap-
pears to represent a generically distinct clade. How-
ever, erecting a new genus for “M.” stylodontus
does not seem prudent because it exhibits only two
derived character states relative to Pliohippus s.s.,
Parapliohippus, Heteropliohippus, and Acritohip-
pus. Furthermore, as noted above, it is plesiom-
orphic for at least five character states relative to
Pliohippus s.s. If future discoveries result in the
identification of additional apomorphic character
states for “M.” stylodontus, then establishing a new
genus would be warranted. However, until such
time, I regard “M.” stylodontus as a plesion of
generic rank.

PHYLOGENETIC SYSTEMATICS OF
NEOGENE HYPSODONT HORSES

In recent years, several investigators have used cla-
distic analyses to clarify the systematics of the North
American Neogene hypsodont horses (e.g. Mac-
Fadden, 1984a; Webb and Hulbert, 1986; Hulbert,
1987a, 1988a, 1988b, 1989; MacFadden and Hul-
bert, 1988; Evander, 1989; Hulbert and Mac-
Fadden, 1991). In particular, the cladistic analyses
presented by Hulbert (1987a, 1989) and Hulbert
and MacFadden (1991) have provided many in-

sights regarding the phylogenetic relations of Neo-
gene hypsodont horses. Hulbert (1989) analyzed a
large number of late Neogene horse taxa, whereas
Hulbert and MacFadden (1991) restricted their
analysis to critical taxa involved in the basal Mio-
cene radiation of hypsodont horses. Based primarily
on these two analyses, Hulbert and MacFadden
(1991, figs. 13, 17) and Hulbert (1993, fig. 1) pro-
posed the most significant vertical phylogenetic re-
evaluation of these horses to date.

Most of the character states that Hulbert (1988b,
1989) and Hulbert and MacFadden (1991) listed
for the various equid genera appear valid, but a few
require additional discussion. Hulbert (1987b) re-
garded all hipparionine genera as being united based
on the following synapomorphies: 1) the metastyl-
ids are subequal or equal in size to the metaconids;
2) the entoflexids, metaflexids, and linguaflexids are
well developed and isolate the metaconids and me-
tastylids from each other; and 3) the deciduous
premolars have a thick coat of cement. Hulbert
(1988b) suggested that Acritohippus tertius, Acri-
tohippus isonesus, and “ Merychippus ” sejunctus
form a monophyletic group with the following syn-
apomorphies uniting them: 1) more isolated pro-
tocones; 2) increased size; and 3) well-developed
metastyles. Hulbert (1988b) listed five ancestral syn-
apomorphies that unite “M.” sejunctus with the
Hipparionini. However, according to his own listed
character states for “M.” sejunctus (Hulbert, 1988b,
tab. 9), only two are actually shared by “M.” se-
junctus and the Hipparionini: 1) strong plis caballin
on the upper molars and 2) metacarpal V articulates
primarily with metacarpal IV. In a much more com-
prehensive cladistic analysis of late Neogene horses,
Hulbert (1989) stated that the only apomorphy that
unites the Hipparionini is a well-separated meta-
conid and metastylid and that A. isonesus and “M.”
sejunctus possess this character state. However,
Hulbert (1988b) clearly stated that “M.” sejunctus
has the metaconid and metastylid well separated
only in early wear (Hulbert, 1988b, tab. 9, character
state 55.1), which is also the same plesiomorphic
character state he listed for the Equini and “Mcr-
ychippus ” primus. Furthermore, my examination
of lower dentitions referred to A. isonesus from
the Mascall Formation (Oregon), Sucker Creek For-
mation (Oregon), and High Rock Sequence (Ne-
vada) does not support A. isonesus as having the
derived state but, instead, the plesiomorphic state
of being well separated only in early wear. Hulbert
and MacFadden (1991) regarded “M.” sejunctus as
being anagenetically derived from the early Barsto-
vian “ Merychippus ” sp. near “M.” sejunctus of
Texas, and they clearly indicated that it exhibited
the hipparionine characters of well-developed plis
caballin, moderately complex fossette margins, and
well-separated metaconids and metastylids.

Hulbert (1988a) considered the synapomorphic
characters that unite the Equini to be the following:
1) the protocones connect to the protolophs in very
early wear stages; 2) the internal fossette plications

16 ■ Contributions in Science, Number 455

Kelly: Miocene Horses

have relatively simple margins; and 3) the lower
molar metastylids are positioned more labially than
the metaconids. Later, Hulbert (1989) regarded the
labially positioned lower molar metastylids as the
only apomorphy uniting the Equini. Hulbert (1989)
placed P arapliohippus carrizoensis and “Mery-
chippus ” stylodontus in plesions (plesiomorphs)
within the Equus genus group of the Equini, with
each taxon having the following synapomorphies:
1) the malar fossa is present and well separated from
the DPOF; and 2) the p3-m3 protostylids are ab-
sent. Hulbert (1989) also considered P. carrizoensis
as being united with all other members of the Equus
genus group by having the following synapomor-
phies: 1) the DPOF is pocketed posteriorly; and 2)
the protocones connect with the protolophs im-
mediately after the onset of wear. Hulbert (1989)
regarded P. carrizoensis as being derived relative
to other members of the Equus genus group by
having the following synapomorphies: 1) a very nar-
row PBL; and 2) small size. In my examination of
specimens referred to P. carrizoensis and “M.” sty-
lodontus, I have found the following: 1) P. carri-
zoensis and “M.” stylodontus occasionally have
the plesiomorphic character state of small anterior
cingulids or precingulids that extend only partially
up the anterior labial face of the lower cheek teeth;
and 2) “M.” stylodontus has lower molar metas-
tylids and metaconids that are well separated only
in early wear.

Hulbert and MacFadden (1991) and MacFadden
(1992) regard the character state of having the ml-
2 metastylids notably smaller and located more la-
bially than the metaconids as one of the synapo-
morphies that unite the Equini (Protohippina plus
Equina). Hulbert (1988a, tab. 18) noted that in early
species of Calippus and Protohippus the metacon-
ids and metastylids are only well separated in early
wear and that the evolution of increased protocone
length and a corresponding increased MML oc-
curred independently within these two lineages.
Furthermore, Hulbert (1988a, p. 285, tab. 17) also
noted that the Barstovian Protohippus perditus
(Leidy, 1858), and Protohippus supremus, exhibit
the character state of having the metastylids only
slightly smaller or equal in size to the metaconids,
respectively. Rensberger et al. (1984) concluded that
an increase in anteroposteriorly directed enamel
edges occurred in Neogene hypsodont horse teeth
as a functional response to diet, maximizing anter-
oposterior grinding, and this response can be ob-
served in such diverse genera as Neohipparion and
Equus. Hulbert (1988a) suggested that this func-
tional response may explain the observed trend in
protohippine and equine genera, wherein as the
protocone increases in length there is a correspond-
ing enlargement of the metaconid-metastylid com-
plex. This trend also is observed in the hipparion-
ines, Neohipparion, Cormohipparion, and Pseud-
hipparion. These facts indicate that the morphol-
ogy of the protocone and the corresponding
metaconid-metastylid complex may be prone to

homoplasy in response to the functional dietary
requirements of a taxon and, therefore, could be
independently derived within lineages depending
on the feeding strategy. This is not to say that the
morphology of the metaconid-metastylid complex
is not useful in establishing equid relations, but it
may be prone to convergence.

In recent years, the use of facial characters, in
particular the morphology of the DPOF and the
malar fossa, has been demonstrated to be significant
in elucidating the phylogenetic relations of Neo-
gene hypsodont horses (e.g. Skinner and Mac-
Fadden, 1977; Bernor et al., 1980; Woodburne,
1982, 1989; J.P. Quinn, 1984; MacFadden, 1984a,
1984b, 1985, 1992; Hulbert, 1988a, 1988b, 1989;
Kelly and Lander, 1988b; Alberdi, 1989; Hulbert
and MacFadden, 1991). The presence of a well-
developed malar fossa is regarded as a derived char-
acter state (Hulbert, 1989; Hulbert and Mac-
Fadden, 1991). Webb and Hulbert (1986) consid-
ered a shallow malar fossa to be a retained primitive
character and noted that a vestigial malar fossa may
occur occasionally in Pseudhipparion, Calippus,
Merychippus insignis Leidy, 1857 (Skinner and
Taylor, 1967), and Neohipparion (MacFadden,
1984a). However, Hulbert (1988a) and Hulbert and
MacFadden (1991, tab. 1) stated that a malar fossa
is absent in Calippus and Merychippus insignis,
respectively. Hulbert (pers. commun., 1994) con-
siders the very slight depression occasionally pres-
ent in Pseudhipparion, Calippus, Neohipparion,
and Merychippus insignis, which Webb and Hul-
bert (1986) referred to as a “malar fossa,” to be
presumably caused by a stronger than usual muscle
attachment on the malar crest. Although possibly
homologous with the malar fossa of the Equini,
this slight depression, which is highly variable in its
expression, cannot be equated with the derived
character state of a well-developed and consistently
present malar fossa, such as those of Pliohippus
s.s. and Acritohippus (Hulbert, pers. commun.,
1994).

In Hulbert’s (1989) analysis of late Neogene hyp-
sodont horses, he recognized the following three
character states for the malar fossa: 1) absent or
very shallow and variable (primitive); 2) present but
not confluent with the DPOF (derived); and 3) pres-
ent and confluent with the DPOF (derived). In Hul-
bert and MacFadden’s (1991) analysis of basal Neo-
gene hypsodont horses, they also recognized three
character states for the malar fossa, but these dif-
fered as follows: 1) no malar fossa present; 2) ru-
dimentary or shallow malar fossa present; and 3)
deep malar fossa present. Although the malar fossa
character states used in these two analyses overlap
somewhat, they emphasize slightly different polar-
ities. Except when character state reversal occurs,
such as in Equus and in Dinohippus s.s., the ab-
sence of a malar fossa is plesiomorphic, as indicated
by its absence in the outgroup “ Parahippus ” leo-
nensis (Sellards, 1916), and “ Merychippus ” primus,
and a very shallow, variably present malar fossa

Contributions in Science, Number 455

Kelly: Miocene Horses ■ 17

most probably is also plesiomorphic (Webb and
Hulbert, 1986). Furthermore, a distinct malar fossa
that is always present and confluent with the DPOF
is not equivalent to the plesiomorphic state of a
very shallow and variably present malar fossa. Hul-
bert (1989) recognized this fact and regarded this
character state as being derived. Similarly, a distinct
malar fossa that is confluent with the DPOF is not
equivalent to one that is well separated from the
DPOF. For these reasons, the malar fossa character
states used in the cladistic analyses presented herein
follow those of Hulbert (1989).

Hulbert and MacFadden (1991, p. 17, character
8) included the following character states for the
shape of the DPOF in their cladistic analysis: “1)
an elongate oval shape (the length is much greater
than the height); and 2) oval shape (the length is
about equal to the height).” The only taxa they
recognized as having an oval-shaped DPOF were
Parapliobippus carrizoensis and Acritohippus ter-
tius. However, the DPOF in P. carrizoensis and
the acritohippines (A. tertius, A. isonesus, and A.
quinni ) is about twice as long as it is high (Wood-
burne, pers. commun, 1994; this report). The only
way to regard these taxa as having an “oval-shaped”
DPOF, wherein “the length is about equal to the
height,” would be to include the malar fossa as
contributing to the “height” of the DPOF. There-
fore, I regard the DPOF shape as being an elongated
oval in P. carrizoensis and the acritohippines. In
the cladistic analyses presented herein, this char-
acter was excluded because all of the taxa analyzed
possess an elongate oval-shaped DPOF.

Sondaar (1968) first noted that certain hipparion-
ines possess the character state of metacarpal V
articulating primarily with metacarpal IV, differing
from certain equines in which metacarpal V pri-
marily articulates with the unciform carpal. Hulbert
and MacFadden (1991, p. 19, character 79) rec-
ognized the following character states for the ar-
ticulation of metacarpal V: 1) metacarpal V artic-
ulates primarily on the unciform carpal, wherein
the articulation facet for metacarpal IV is absent
or smaller than the articulation facet on the unci-
form carpal (plesiomorphic); and 2) metacarpal V
articulates primarily with metacarpal IV, wherein
the articulation facet on the unciform is smaller
than on metacarpal IV or absent (apomorphic).
Based on their cladistic analysis, Hulbert and
MacFadden (1991) regarded the derived character
state for this character as one of the synapomor-
phies uniting the Hipparionini. However, the char-
acter state distribution for this character is un-
known or has not been determined for many late
Neogene horses. For example, the character state
for this character is only known for 5 of the 13
taxa analyzed by Hulbert and MacFadden (1991)
and 12 of the 27 taxa analyzed herein. In Eocene
horses, such as Hyracotherium, Orohippus, and
Epihippus, metacarpal V, although slightly re-
duced, supported a functional digit with three pha-
langes (Matthew, 1926; Kitts, 1956, 1957). When

horses evolved a tridactyl manus during the early
Oligocene, metacarpal V was reduced to a very
small vestigial bone (Osborn, 1918; W.B. Scott, 1941;
Simpson, 1951). In modern horses there is no trace
of a metacarpal V (Evander, 1989). In terms of
morphological function, it is difficult to explain the
two different character states for the metacarpal V
articulation in late Neogene horses. It could be
argued that the articulation of metacarpal V should
not be included in a phylogenetic analysis of late
Neogene horses because metacarpal V is vestigial
and its functional morphology has not been deter-
mined. Nevertheless, there does seem to be a trend
in the known distribution of the articulation mor-
phology, wherein the hipparionines exhibit the de-
rived state and the equines exhibit the plesiom-
orphic state (Hulbert and MacFadden, 1991). How-
ever, this character state must be regarded as equiv-
ocal for establishing phylogenetic relationships until
the character state distribution for this character is
much better known.

To facilitate easy comparison of the analyses pre-
sented herein with those previously published, the
numbering scheme of the characters and character
states (Appendix A) generally corresponds to those
of Hulbert (1988b, 1989) and Hulbert and Mac-
Fadden (1991). Except as noted above, the polar-
ities and descriptions of all of the characters and
character states presented in Appendix A have been
discussed in detail by other investigators (e.g. Son-
daar, 1968; Webb, 1969; Hussain, 1975; Mac-
Fadden, 1984a, 1984b, 1992; Webb and Hulbert,
1986; Hulbert, 1987a, 1988a, 1988b, 1989; Hulbert
and MacFadden, 1991).

Hulbert (1989) and Hulbert and MacFadden
(1991) performed their cladistic computer analyses
using the PAUP algorithm (Swofford, 1985). In or-
der to compare the Hennig86 program used herein
and the PAUP program used by Hulbert (1989) and
Hulbert and MacFadden (1991), the character state
matrices used by Hulbert (1989) and Hulbert and
MacFadden (1991) in their respective analyses were
run on the Hennig86 program. The results of these
two analyses were identical to those of Hulbert
(1989) and Hulbert and MacFadden (1991).

Based on a synthesis of the analyses of Hulbert
(1989) and Hulbert and MacFadden (1991) plus
those of Hulbert (1987a, 1988a, 1988b) and Webb
and Hulbert (1986), Hulbert and MacFadden (1991,
fig. 13) hypothesized the phylogenetic relations of
the Neogene hypsodont horses but noted that these
proposed relations were very provisional. In order
to test these provisional hypothesized relations, cla-
distic analyses were performed on the character
state matrix presented in Appendix B, which com-
bines the taxa that Hulbert (1989) and Hulbert and
MacFadden (1991) used in their separate analyses.
Although the overall cladogram typologies and re-
sulting hypothesized relationships presented herein
can be compared with those presented by Hulbert
(1989) and Hulbert and MacFadden (1991), a direct
comparison of the cladogram lengths (number of

181 Contributions in Science, Number 455

Kelly: Miocene Horses

steps) cannot be made because each analysis was
performed using a different number of taxa, char-
acters, and character states.

Analysis of Appendix B with equally weighted
characters produced two equally most parsimoni-
ous cladograms of 246 steps with consistency in-
dices of 39 and retention indices of 66 (Figure 4).
The cladograms differ only in the positions of Ac-
ritohippus tertius and “ Merychippus ” sp. near “M.”
sejunctus; in one they are allied with the protohip-
pines and hipparionines (Figure 4A), whereas in the
other they are allied with the equines (Figure 4B).
One of the cladograms of the equally weighted
analysis of Appendix B (Figure 4A) is identical to
the hypothesized relationships proposed by Hul-
bert and MacFadden (1991, fig. 13) except for the
positions of the protohippines; in Hulbert and
MacFadden’s analysis, they are allied with the
equines, whereas, in this analysis, they are allied
with the hipparionines.

The analysis of Appendix B was then repeated
using successively weighted characters, a procedure
that has been shown to avoid the excessive weight-
ing of multistate characters relative to binary char-
acters and a means of basing outgroupings on more
dependable characters without making prior deci-
sions on weighting (Goldman, 1988; Farris, 1988;
Hulbert and MacFadden, 1991). This procedure
also reduces the ambiguity of complex data sets;
there may be multiple cladograms of minimal length
with equally weighted characters, but successive
weighting will often produce a single most parsi-
monious cladogram (Farris, 1988). Thus, clado-
grams produced using successively weighted char-
acters have higher consistency and retention indices
and are based on more reliable characters than those
produced using equally weighted characters. The
successively weighted analysis of Appendix B re-
sulted in a single most parsimonious cladogram of
556 steps with a consistency index of 53 and re-
tention index of 77 (Figure 5). The successively
weighted analysis confirms most of the hypothe-
sized relationships proposed by Hulbert and
MacFadden (1991, fig. 13) except for the following:
1) the protohippines are allied with the hipparion-
ines instead of the equines; 2) Acritohippus tertius
is allied with the equines instead of the hipparion-
ines; and 3) the positions of the Hipparion clade
and the Neohipparion clade (“ Merychippus ” co-
lor adense [Osborn, 1918], Pseudhipparion, Neo-
hipparion) are interchanged.

In order to determine the phylogenetic relations
of Acritohippus and Heteropliohippus to other
Neogene hypsodont horses, cladistic analyses were
performed using the character state matrix pre-
sented in Appendix C, which includes 24 of the
horse taxa analyzed by Hulbert (1989) and Hulbert
and MacFadden (1991) plus Acritohippus quinni,
Heteropliohippus hulberti, and “ Dinohippus ” in-
terpolates. Following Hulbert and MacFadden
(1991), “ Parahippus ” leonensis was selected as the
outgroup. Equus simplicidens was selected as the

representative for Equus because it is the oldest
known and least derived species of the genus (Win-
ans, 1989; Kelly, 1994; Downs and Miller, 1994).
Hulbert and MacFadden (1991) included Proto-
hippus vetus J.H. Quinn, 1955, in their cladistic
analysis because it is the oldest known species of
Protohippus. However, many of the character states
of Protohippus vetus are unknown. In the cladistic
analyses presented herein, the character states for
Protohippus were based not only on Protohippus
vetus but also on the better known Protohippus
perditus.

Analysis of Appendix C with equally weighted
characters resulted in two equally most parsimo-
nious cladograms of 260 steps with consistency in-
dices of 36 and retention indices of 65 (Figure 6).
These cladograms differ only in the positions of the
acritohippines and “ Merychippus ” sp. near “M.”
sejunctus ; in one they are allied with the equines
(Figure 6A), whereas in the other they are allied
with the protohippines and the hipparionines (Fig-
ure 6B). The equally weighted analysis of Appendix
C supports the following conclusions: 1) Paraplioh-
ippus carrizoensis is a generically distinct clade that
is the sister taxon of the higher equines ( Pliohippus ,
Heteropliohippus, Astrohippus, Dinohippus, and
Equus); 2) Heteropliohippus is the closest sister
taxon to Astrohippus and the pliohippine clade
{Heteropliohippus, Astrohippus, and Pliohippus) is
monophyletic and the closest sister group to the
Equus-Dinohippus clade; 3) the acritohippine clade
{Acritohippus tertius, Acritohippus isonesus, and
Acritohippus quinni) is monophyletic and its re-
lationships to the equines or the hipparionines are
unresolved; 4) the relationships of the “ Merychip-
pus” sp. near “M.” sejunctus clade to the acrito-
hippines, protohippines, equines, or hipparionines
are unresolved; and 5) the protohippine clade is
monophyletic and the closest sister group to the
hipparionines.

The analysis of Appendix C was then repeated
using successively weighted characters, which re-
sulted in a single most parsimonious cladogram of
556 steps with a consistency index of 53 and a
retention index of 77 (Figure 7). The cladogram
produced by the successively weighted analysis (Fig-
ure 7) is identical to one of the cladograms pro-
duced by the equally weighted analysis (Figure 6A)
except for the position of “ Merychippus ” sp. near
“M.” sejunctus. In the successively weighted anal-
ysis, “ Merychippus ” sp. near “M.” sejunctus rep-
resents a distinct clade that is more closely allied
to the hipparionines than to the acritohippines or
the equines. The successively weighted analysis of
Appendix C supports the following conclusions: 1)
Parapliohippus carrizoensis is a generically distinct
clade that is the sister taxon of the higher equines
{Pliohippus, Heteropliohippus, Astrohippus, Di-
nohippus, and Equus); 2) Heteropliohippus is the
closest sister taxon to Astrohippus and the plioh-
ippine clade {Heteropliohippus, Astrohippus, and
Pliohippus) is monophyletic and the closest sister

Contributions in Science, Number 455

Kelly: Miocene Horses ■ 19

Figure 4. Two equally most parsimonious cladograms of 246 steps with consistency indices of 39 and retention indices
of 66 using the character state matrix presented in Appendix B with characters equally weighted.

20 ■ Contributions in Science, Number 455

Kelly: Miocene Horses

Figure 5. Single most parsimonious cladogram of 560 steps with a consistency index of 54 and retention index of 77
using the character state matrix presented in Appendix B with characters successively weighted. The cladogram is
supported by the following list of ancestral synapomorphies. Number to left of period denotes character number and
to right of period character state of hypothesized ancestor. Node 1: 20.1; 21.1; 30.2; 31.1; 57.1; 62.1; 70.0; 71.1. Node
2: 23.1; 27.2; 28.2; 29.1; 33.2; 38.2; 39.2; 45.1; 55.1; 56.1. Node 3: 2.1; 5.1; 21.2; 45.2; 54.1; 70.2; 71.2. Node 4: 12.1;
30.1; 33.1; 67.1. Node 5: 3.2; 27.0; 28.0. Node 6: 20.2; 63.1; 70.3; 71.6. Node 7: 5.2; 23.2; 29.1; 43.1; 62.2; 63.2; 71.7.
Node 8: 1.2; 3.2; 7.1; 12.0; 38.2; 39.2; 70.4. Node 9: 5.0; 30.1; 33.2; 38.0; 39.0. Node 10: 79.1. Node 11: 52.1; 71.3.
Node 12: 7.1; 12.0; 23.2; 31.1; 43.1; 66.1; 67.1. Node 13: 16.0; 28.1; 48.2; 63.1. Node 14: 13.1; 40.0; 41.0. Node 15:
31.2; 33.3; 55.2. Node 16: 7.1; 12.0; 27.4; 28.4; 39.1. Node 17: 1.1; 6.1; 7.2; 23.2; 33.4; 35.0; 38.1; 48.1. Node 18:
35.2; 48.2. Node 19: 23.2; 33.3; 43.1; 62.2; 63.1; 71.4. Node 20: 3.0; 5.0; 16.1; 20.2; 22.1; 23.3; 31.1; 52.2; 63.2; 71.7.
Node 21: 5.2; 30.3; 32.1; 33.4; 38.1; 39.1. Node 22: 3.2; 5.3; 6.1; 27.5; 28.5; 33.5; 38.0; 39.0; 40.0; 52.2. Node 23: 3.3;
7.2; 20.2; 23.2; 29.0; 43.1; 62.2; 63.1; 71.5.

group of the Equus-Dinohippus clade; 3) the ac-
ritohippine clade ( Acritohippus tertius, Acritohip-
pus isonesus, and Acritohippus quinni) is mono-
phyletic and the closest sister group to the equines;
and 4) the protohippine clade is monophyletic and
the closest sister group of the hipparionines.

In the successively weighted analysis of Appendix
C (Figure 7), a single hypothesized ancestral syna-
pomorphy unites the acritohippines with the equines
(Figure 7, node 4); a malar fossa is present that is
well separated from the DPOF (character 12.1). In
this analysis, the acritohippines are assumed to be
derived relative to the equines by the character
transformation of an ancestral well-separated DPOF

and malar fossa to a confluent DPOF and malar
fossa. This scenario cannot be supported or refuted
by the geochronological distribution of these clades
because the earliest known acritohippine, Acrito-
hippus tertius, and the earliest known equine, Par-
apliohippus carrizoensis, first appeared at about
the same time in the late Hemingfordian. Consid-
ering the variability of the position of the acrito-
hippines within the cladograms presented herein
and those of Hulbert (1989) and Hulbert and
MacFadden (1991), I regard the relationships of the
acritohippines to the equines and hipparionines as
uncertain.

Hulbert and MacFadden (1991) listed the fol-

Contributions in Science, Number 455

Kelly: Miocene Horses ■ 21

Figure 6. Two equally most parsimonious cladograms of 260 steps with consistency indices of 36 and retention indices
of 65 using the character state matrix presented in Appendix C with characters equally weighted.

22 ■ Contributions in Science, Number 455

Kelly: Miocene Horses

Figure 7. Single most parsimonious cladogram of 556 steps with a consistency index of 53 and retention index of 77
using the character state matrix presented in Appendix C with characters successively weighted. The cladogram is
supported by the following list of ancestral synapomorphies. Number to left of period denotes character number and
to right of period character state of hypothesized ancestor. Node 1: 20.1; 21.1; 30.2; 31.1; 57.1; 62.1; 70.0; 71.1. Node
2: 23.1; 27.2; 28.2; 29.1; 33.2; 38.2; 39.2; 45.1; 55.1; 56.1. Node 3: 2.1; 3.1; 21.2; 45.2; 54.1; 70.2; 71.2. Node 4: 12.1.
Node 5: 12.2; 16.1; 27.3; 31.2. Node 6: 3.2; 5.2; 28.3. Node 7: 30.1; 33.1; 67.1. Node 8: 5.2; 16.2; 27.0; 28.0. Node
9: 20.2; 63.1; 70.3; 71.6. Node 10: 3.2; 38.3; 39.3. Node 11: 7.0; 43.1; 67.0. Node 12: 1.2; 3.1; 7.2; 33.2; 38.1; 39.1;
62.2; 63.2; 70.4; 71.6; 72.1. Node 13: 5.0; 12.0; 23.2; 38.0; 39.0; 71.7. Node 14: 48.2; 52.1; 71.3. Node 15: 7.1; 23.2;
43.1; 66.1; 67.1. Node 16: 16.0; 28.1; 63.1. Node 17: 31.2; 33.3; 55.2. Node 18: 27.4; 28.4; 39.1. Node 19: 23.2; 35.2;
43.1; 62.2; 63.1; 71.4. Node 20: 3.0; 5.0; 7.1; 16.1; 20.2; 22.1; 23.3; 31.1; 52.2; 63.2; 71.7. Node 21: 7.1; 33.4; 38.1.
Node 22: 1.1; 6.1; 7.2; 23.2; 35.0; 48.1. Node 23: 5.2; 30.3; 32.1; 35.2. Node 24: 3.2; 5.3; 6.1; 27.5; 28.5; 33.5; 38.0;
39.0; 40.0; 52.2. Node 25: 3.3; 7.2; 20.2; 23.2; 29.0; 43.1; 62.2; 63.1; 71.5.

lowing six hypothesized ancestral synapomorphies
that unite the protohippines and the equines in the
tribe Equini: 1) the DPOF depth is moderate; 2) the
DPOF has a shallow posterior pocket; 3) the P3-4
protocones connect to the protolophs in early wear;
4) hypoconal lakes are formed on P3-4 with closure
of the hypoconal grooves; 5) the p3-4 metastylids
are notably smaller and positioned more lingually
than the metaconids; and 6) the ml -3 metastylids
are notably smaller and positioned more labially
than the metaconids. The first four synapomorphies
listed above are not actually shared by all proto-
hippines. The DPOF is moderately deep in “Mer-
ychippus ” intermontanus (Merriam, 1915), shal-
low in Protohippus, and very shallow in Calippus.

The connection of the P3-4 protocones to the pro-
tolophs occurs in early wear for “M.” intermon-
tanus and Protohippus, but in Calippus it occurs
at the onset of wear, which may represent a reversal
or a further derived state. The formation of hy-
poconal lakes on the P3-4 occurs in “M.” inter-
montanus and Calippus, but not in Protohippus.
If the six hypothesized synapomorphies actually oc-
curred in the proposed ancestor of the protohip-
pines and equines, then at least three character state
reversals must have occurred in the protohippines.
Of these six synapomorphies, only the following
can be observed in all of the protohippines: 1) the
p3-4 metastylids are notably smaller and positioned
more lingually than the metaconids; and 2) the ml-

Contributions in Science, Number 455

Kelly: Miocene Horses ■ 23

2 metastylids are notably smaller and positioned
more labially than the metaconids. Furthermore,
certain hipparionines exhibit the same derived char-
acter states as numbers 1, 2, and 4 of the above six
hypothesized synapomorphies (Hulbert, 1987a,
1988b). If the protohippines shared a common an-
cestor with the equines, then it must be assumed
that certain hipparionines independently derived
these character states and, therefore, these char-
acter states are prone to convergence or parallelism.
Moreover, in Parapliohippus carrizoensis and
“ Merychippus ” stylodontus, the least derived and
oldest unequivocal equines, the p3-4 metastylids
and metaconids are subequal or equal in size and
position. This fact necessitates the assumption that,
in P. carrizoensis and “M.” stylodontus, the p3-4
metastylids underwent a character state reversal from
the synapomorphous character state of the hy-
pothesized ancestor of the protohippines and
equines. Another possible scenario is that later
equines, such as Pliohippus, and the protohippines
independently acquired this character state trans-
formation. As Szalay (1993) and K.M. Scott and
Janis (1993) have demonstrated, cladistic analyses
based on algorithms using parsimony will only pro-
duce a phylogenetic hypothesis with the least num-
ber of steps, which often necessitates the inclusion
of biologically unlikely character transformations.
K.M. Scott and Janis (1993) also noted that char-
acters utilized as synapomorphies should be well
corroborated and not prone to functional or eco-
logical homoplasy. Furthermore, in exclusive cla-
distic practice, the geochronologic distributions of
taxa are disregarded in determining a phylogenetic
hypothesis, which can also result in biologically
improbable character transformations (Szalay, 1993).
Thus, the only synapomorphy listed by Hulbert and
MacFadden (1991) that can be used with confi-
dence to unite the protohippines with the equines
is that the ml -2 metastylids are notably smaller
and positioned more labially than the metaconids.
However, as previously noted, the morphology of
the lower molar metaconid-metastylid complex may
be prone to homoplasy as a functional adaptive
response to feeding strategy.

In all the analyses presented herein, the proto-
hippines are united with the hipparionines by the
following hypothesized ancestral synapomorphies:
1) the dp3-4 protostylids are well developed; 2) the
p3-m3 protostylids are well developed; and 3) the
unworn molar crown height is about 35 mm. The
ancestral synapomorphy of well-developed dp3-4
protostylids is shared by most all hipparionines.
However, in certain equines, such as Astrohippus
and Dinohippus, moderately well-developed dp3-
4 protostylids were secondarily derived (Hulbert,
1987a, tab. 46), indicating that this character may
be prone to convergence. The ancestral synapo-
morphy of having moderately well-developed p3-
m3 protostylids is transformed to the more derived
state of very well-developed protostylids in certain
hipparionines, such as in Nannippus, Pseudhip-

parion, Neohipparion, and Cormohipparion. Ex-
cluding the protohippines, almost all equines, such
as Pliohippus, Dinohippus, Astrohippus, Onohip-
pidium, and most species of Equus, exhibit the
derived state of the loss of the p3-m3 protostylids
(Hulbert, 1987a, tab. 46). However, in very rare
instances certain equines, such as Heteropliohippus
hulberti and Equus parastylidens Mooser, 1959,
have secondarily acquired the derived character state
of moderately developed p3-m3 protostylids, in-
dicating that this character may be prone to a very
low degree of convergence. The ancestral syna-
pomorphy of an unworn crown height of 35 mm
is transformed to a more derived state of greater
hypsodonty in all hipparionines. The trend toward
greater hypsodonty in more derived taxa is also
observed in the equines and acritohippines and ap-
pears to be a functional adaptive transformation.
Considering the above facts, the only ancestral syn-
apomorphy that can be used with any confidence
to unite the protohippines with the hipparionines
is moderately to well-developed p3-m3 protostyl-
ids. However, it should be noted that the proto-
hippines do exhibit similarities in facial morphology
with certain hipparionines. For example, the oldest
protohippines for which the facial morphology is
known, the early to late Barstovian “ Merychippus ”
intermontanus and the late Barstovian Protohippus
perditus, exhibit the following similarities in facial
morphology with specimens of the hipparionine
“ Merychippus ” coloradense from the early Barsto-
vian of Nebraska: 1) the DPOF is elongate oval-
shaped, slightly pocketed posteriorly, moderately
well rimmed posteriorly, and positioned relatively
high and anterior on the face, resulting in a mod-
erately wide preorbital bar; and 2) a malar fossa is
lacking.

Because the character transformations in the
cheek teeth of Neogene hypsodont horses exhibit
high degrees of convergence, reversal, and paral-
lelism (Hulbert, 1987a, 1989; Hulbert and Mac-
Fadden, 1991; this paper), neither the proposed
phylogenetic hypothesis of Hulbert and Mac-
Fadden (1991), which unites the protohippines with
the equines, nor those presented herein, which unite
the protohippines with the hipparionines, appear
overly convincing. Based on the cladistic analyses
presented herein, I regard the protohippines as the
sister group of the hipparionines, but realizing that
this proposed relationship is very tentative. The
hypothesis that the protohippines are more closely
allied with the hipparionines than the equines (Fig-
ures 4-7) is very similar to an alternative, slightly
less parsimonious cladogram presented by Hulbert
and MacFadden (1991, fig. 11B), which places the
equines as the sister group of the hipparionines and
protohippines.

In the successively weighted cladistic analyses
presented herein, “ Merychippus ” sp. near “M.” se~
junctus is the least derived hipparionine and is the
closest sister taxon to all other hipparionines. This
is based on the assumption that “M.” sp. near “M.”

24 ■ Contributions in Science, Number 455

Kelly: Miocene Horses

sejunctus possesses the following ancestral syna-
pomorphies with other hipparionines: 1) well-de-
veloped Ml -3 plis caballin (character 31.2); 2) per-
sistent P2-M3 internal fossette plications (character
33.3); and 3) persistently well-separated metaconids
and metastylids (character 55.2). “ Merychippus ”
sejunctus, the presumed anagenetic descendant of
“M. ” sp. near “M. ” sejunctus (Hulbert and
MacFadden, 1991), is plesiomorphic for the above
character states. This fact is problematical because
it would require three character state reversals for
“M” sejunctus to be derived from “M.” sp. near
“M.” sejunctus. These character transformations
seem unlikely but are possible considering the high
degree of homoplasy exhibited in the cheek teeth
character states of Neogene hypsodont horses.
“ Merychippus ” sp. near “M.” sejunctus and the
acritohippines have a similar facial morphology,
wherein the DPOF and malar fossa are confluent.
The successively weighted cladistic analyses assume
that this apomorphic facial morphology was in-
dependently derived in “M.” sp. near “M.” se-
junctus and the acritohippines. This assumption
seems biologically unlikely because, in many other
Neogene hypsodont horses, a similar facial mor-
phology is generally a good indicator of a close
phylogenetic relationship (MacFadden, 1984a).
However, the oldest known acritohippine, Acri-
tohippus tertius, which first appeared in the late
Hemingfordian, is plesiomorphic for the three de-
rived cheek teeth character states exhibited by the
early Barstovain “M.” sp. near “M.” sejunctus. An
alternative scenario to explain a similar facial mor-
phology in “M.” sp. near “M.” sejunctus and the
acritohippines is that “M.” sp. near “M.” sejunctus
was derived from A. tertius or a similar ancestral
morphotype with a confluent DPOF and malar fos-
sa and, with this speciation event, it independently
developed stronger molar plis caballin, more per-
sistent P2-M3 internal fossettes, and better sepa-
rated metaconids and metastylids as a functional
dietary response. In this scenario, Acritohippus
would become the closest sister taxon to “M.” s
near “M.” sejunctus and the three character state
transformations observed in “M.” sp. near “M.”
sejunctus would represent convergence with the
hipparionines, not synapomorphies. Considering the
high degree of homoplasy observed in the cheek
teeth morphology of Neogene hypsodont horses
and the similarity in facial morphology of “M.” sp.
near “M.” sejunctus and Acritohippus, I regard the
position of “M.” sp. near “M.” sejunctus in the
successively weighted cladograms (Figures 5, 7) as
equivocal and its phylogenetic status as unresolved.
If “M.” sp. near “M.” sejunctus is removed from
the analyses, then the Hipparionini s.s. would in-
clude Neohipparion, Pseudhipparion, Hipparion,
Nannippus, Cormohipparion, Merychippus s.s.,
and “Merychippus” coloradense.

In summary, the analyses presented herein sup-
port many of the results of the cladistic analyses of
Hulbert (1989) and Hulbert and MacFadden (1991),

with the exception of the following: 1) the plioh-
ippines (Heteropliohippus, Astrohippus, and Plio-
hippus) form a monophyletic clade that is the clos-
est sister group of the Onohippidium-Hippidion
clade and Equus-Dinohippus clade; 2) the acrito-
hippines (Acritohippus tertius, A. isonesus, and A.
quinni) form a monophyletic clade whose relations
to the equines and hipparionines remain unre-
solved; and 3) the protohippines (“Merychippus”
intermontanus, Protohippus, and Calippus) form
a monophyletic clade that is the closest sister group
of the hipparionines.

Hulbert (1989) and Hulbert and MacFadden
(1991) argued for a basal dichotomy in the late
Neogene hypsodont horses that equated with the
tribal ranks of the Equini and Hipparionini, which
in the successively weighted analyses presented
herein would be equated with nodes 4 and 11 of
Figure 5 and nodes 4 and 14 of Figure 7, respec-
tively. According to this contention, the acritohip-
pines and protohippines represent monophyletic
clades of subtribal rank. However, as demonstrated
in the analyses presented herein and those of Hul-
bert (1989) and Hulbert and MacFadden (1991),
the position of the acritohippines in the cladograms
is highly labile, making any hypothesis of their re-
lationships to the equines or hipparionines equiv-
ocal. Furthermore, two of the three putative an-
cestral synapomorphies defining the nodes that unite
the protohippines with the hipparionines (e.g. node
11, Figure 5; node 14, Figure 7) are equivocal. Bas-
ing a tribal rank on a small number of putative
ancestral synapomorphies may not be prudent be-
cause it is well documented that many of the mor-
phological character states of the late Neogene hyp-
sodont horses are homoplasous; that is, they exhibit
a high degree of parallelism, convergence, and re-
versal in response to the functional requirements
of feeding and locomotion (Hulbert, 1987a, 1989;
MacFadden, 1992; this paper). According to this
contention, the four clades (acritohippines, proto-
hippines, equines, and hipparionines) could be
equated with tribal ranks and the nodes of the basal
dichotomy regarded as unnamed ranks.

The following is a conservative assessment of the
phylogenetic relationships of the late Neogene hyp-
sodont horses based on a consensus of all the anal-
yses presented herein. “Merychippus” gunteri
(Simpson, 1930) and “Merychippus” primus are the
inferred successive sister taxa to all other Neogene
hypsodont horses. Following the cladogenetic spe-
ciation events that produced “M.” gunteri and “M.”
primus, four additional basal monophyletic clades
of Neogene hypsodont horses evolved; the Equini
s.s., the Hipparionini s.s., the Protohippini s.s., and
the acritohippines. The Protohippini s.s. are pro-
visionally regarded as the closest sister group to the
Hipparionini s.s. The relationships of the acrito-
hippines to the Equini s.s. and the Hipparionini s.s.
are unresolved. The relationships of “Merychip-
pus” sp. near “M.” sejunctus to the Hipparionini
s.s. and the acritohippines remains unclear. The

Contributions in Science, Number 455

Kelly: Miocene Horses ■ 25

Figure 8. Hypothesized phylogenetic tree of late Neogene hypsodont horses. Thick vertical line indicates geochron-
ologic occurrence of taxon. Key to taxa: 1, “ Parahippus ” leonensis; 2, “ Merychippus ” gunteri ; 3, “ Merychippus ”
primus; 4, “ Merychippus ” stylodontus; 5, Parapliohippus carrizoensis; 6, Pliohippus; 7, Heteropliohippus; 8, Astro-
hippus; 9, “ Dinohippus ” interpolatus; 10, Dinohippus; 11, Equus; 12, Onohippidium ; 13, Hippidion; 14, Acritohippus ;
15, “ Merychippus ” sp. near “M.” sejunctus/“M .” sejunctus clade; 16, “ Merychippus ” intermontanus; 17, Protohippus;
18, Calippus; 19, “ Merychippus ” coloradense; 20, Neohipparion; 21, Pseudhipparion; 22, Hipparion; 23, Merychippus;
24, “Merychippus” goorisi; 25, Nannippus; 26, Cormohipparion.

Equini s.s. are united by the following ancestral
synapomorphies: 1) the DPOF depth is moderate;
2) the malar fossa is distinct, always present, and
well separated from the DPOF; 3) the P2-4 plis
caballin are commonly present, but small and non-
persistent (disappear in early to moderate wear); 4)
the plications of the internal fossette margins are
very simple and nonpersistent; and 5) the ml -3
metastylids are notably smaller in size and posi-
tioned more labially than the metaconids. The Hip-
parionini s.s. are united by the following ancestral
synapomorphies: 1) the Ml -3 plis caballin are well
developed; 2) the P3-M2 protocones connect with
the protolophs in late moderate wear; 3) the inter-
nal fossette margins are simple, but persistent (pres-
ent in moderate to late wear); 4) the Ml -2 hypo-

conal grooves close in late wear; 5) the p3-m3
metaconids and metastylids are persistently well
separated; and 6) metacarpal V articulates primarily
with metacarpal IV. The Protohippini s.s. are united
by the following ancestral synapomorphies: 1) the
relative PBL is moderate; 2) the P3-M2 protocones
have elongated oval occlusal outlines; 3) the Ml-
3 plis caballin are commonly present, but small and
nonpersistent; 4) the upper cheek teeth are mod-
erately curved; 5) the p3-4 metastylids are smaller
and positioned more lingually than the metaconids;
and 6) the ml-3 metastylids are notably smaller
and positioned more labially than the metaconids.
The acritohippines are united by the following an-
cestral synapomorphies: 1) the malar fossa is shal-
low, always present, and confluent with the DPOF;

26 ■ Contributions in Science, Number 455

Kelly: Miocene Horses

2) the relative muzzle length is moderate; 3) the
upper cheek teeth are moderately curved; 4) the
P3-4 protocones connect with the protolophs in
early moderate wear; and 5) the Ml -3 plis caballin
are moderately well developed and relatively per-
sistent. Based on this assessment, a hypothesized
phylogenetic tree was constructed and is presented
in Figure 8.

In conclusion, all the analyses presented herein
indicate that after the cladogenetic speciation events
that produced “ Merychippus ” gunteri and “Mcr-
ychippus" primus the Neogene hypsodont horses
underwent rapid cladogenesis, resulting in at least
four additional basal monophyletic lineages: the
Equini s.s., the Hipparionini s.s., the Protohippini
s.s., and the acritohippines. However, any phylo-
genetic hierarchical classification depicting the in-
terrelationships of these four additional lineages is
very tentative and will probably require modifica-
tion with further study or the discovery of addi-
tional diagnostic characters.

ACKNOWLEDGMENTS

I am indebted to Richard C. Hulbert, Jr., of the Georgia
Southern University, Michael O. Woodburne of the Uni-
versity of California at Riverside (UCR), Kenneth E.
Campbell, David P. Whistler, and Samuel A. McLeod of
the Natural History Museum of Los Angeles County
(LACM), James P. Quinn of Gorian and Associates, Inc.,
and a research associate of the LACM, and E. Bruce
Lander of Paleo Environmental Associates, Inc., and a
research associate of the LACM for their comments and
advice on the original drafts of this report. Collections at
the University of California Museum of Paleontology, San
Bernardino County Museum, University of California at
Riverside, and the Natural History Museum of Los An-
geles County were made available by J. Howard Hutch-
ison, Robert E. Reynolds, Michael O. Woodburne, and
Lawrence G. Barnes, respectively. I am especially grateful
to Samuel A. McLeod of the LACM for making specimens
available for loan, Michael O. Woodburne of the UCR
for providing critical information on specimens housed
at the American Museum of Natural History, and Alan
VanArsdale for his donation of equid specimens to the
LACM that he collected from the Caliente Formation,
including the referred skull of Heteropliohippus. Special
thanks is further given to James P. Quinn for allowing
me to describe the new species of horse from the Caliente
and Bopesta Formations, which he first recognized as rep-
resenting a distinct taxon.

LITERATURE CITED

Alberdi, M. 1989. A review of Old World hipparionine
horses. In The evolution of perissodactyls, ed. D.R.
Prothero and R.M. Schoch. Oxford Monographs on
Geology and Geophysics, no. 15, 234-261. New
York: Oxford University Press, xii -I- 537 pp.

Barbour, E.H. 1914. A new fossil horse, Hypohippus
matthewi. Nebraska Geological Bulletin 4: 169-173.

Bemor, R.L., M.O. Woodburne, and J.A. Van Covering.
1980. A contribution to the chronology of some
Old World Miocene faunas based on hipparionine
horses. Geobios 13(5):705-739.

Bode, F.D. 1934. Tooth characters of protohippine

horses with special reference to species from the
Merychippus zone, California. Carnegie Institution
of Washington Publication 453:39-63.

Buwalda, J.P., and G.E. Lewis. 1955. A new species of
Merychippus. United States Geological Survey Pro-
fessional Paper 264-G:143-152.

Cope, E.D. 1874. Report on the stratigraphy and Plio-
cene vertebrate paleontology of northern Colorado.
Bulletin of the United States Geological and Geo-
graphical Survey of the Territories 1(1):15.

— . 1879. Observations on the faunae of the Mio-

cene Tertiaries of Oregon. Bulletin of the United
States Geological and Geographical Survey of the
Territories 5(l):55-69.

. 1880. A new Hippidium. American Naturalist

14(3):223.

. 1889. A review of North American species of

Hippotherium. Proceedings of the American Philo-
sophical Society 26:429-458.

. 1892. A contribution to the vertebrate pale-
ontology of Texas. Proceedings of the American
Philosophical Society 30:123-131.

. 1893. A preliminary report on the vertebrate

paleontology of the Llano Estacado. Fourth Annual
Report of the Geological Survey of Texas, 1-136.

Dougherty, J.F. 1940. A new mammalian fauna from
Caliente Mountain, California. Carnegie Institution
of Washington Publication 514:109-143.

Downs, T. 1956. The Mascall Fauna from the Miocene
of Oregon. University of California Publications in
Geological Sciences 31(5):199-354.

Downs, T., and G.J. Miller. 1994. Late Cenozoic equids
from the Anza-Borrego Desert of California. Con-
tributions in Science 440:1-90.

Drescher, A.B. 1941. Later Tertiary Equidae from the
Tejon Hills, California. Carnegie Institution of
Washington Publication 530:1-23.

Evander, R.L. 1989. Phylogeny of the family Equidae.
In The evolution of perissodactyls, ed. D.R. Proth-
ero and R.M. Schoch. Oxford Monographs on Ge-
ology and Geophysics, no. 15, 107-127. New York:
Oxford University Press, xii + 537 pp.

. 1993. Astrohippus walked on faerie toes. Jour-
nal of Vertebrate Paleontology, Abstracts of Papers
13(3):34A.

Farris, J.S. 1988. Hennig86, users manual version 1.5.
Ann Arbor: Museum of Zoology, University of
Michigan, 21 pp.

Gazin, C.L. 1930. A Tertiary vertebrate fauna from the
upper Cuyama drainage basin, California. Carnegie
Institution of Washington Publication 404:55-7 6.

Gidley, J.W. 1907. Revision of the Miocene and Plio-
cene Equidae of North America. Bulletin of the
American Museum of Natural History 23(35):8 65-
934.

Goldman, N. 1988. Methods for discrete coding of
morphological characters for numerical analysis.
Cladistics 4(1):59-71.

Hulbert, R.C. 1987a. Phylogenetic systematics, bioch-
ronology, and paleobiology of Late Neogene horses
(family Equidae) from the Gulf Coastal Plain and
the Great Plains. Ph.D. dissertation, University of
Florida, Gainesville, 570 pp.

. 1987b. Late Neogene Neohipparion (Mam-
malia, Equidae) from the Gulf Coast Plain of Florida
and Texas. Journal of Paleontology 61(4):809-830.

. 1988a. Calippus and Protohippus (Mammalia,

Perissodactyla, Equidae) from the Miocene (Barsto-
vian-early Hemphillian) of the Gulf Coastal Plain.

Contributions in Science, Number 455

Kelly: Miocene Horses ■ 27

Bulletin of the Florida State Museum, Biological
Sciences 32(3):221-340.

. 1988b. Cormohipparion and Hipparion (Mam-
malia, Perissodactyla, Equidae) from the late Neo-
gene of Florida. Bulletin of the Florida State Mu-
seum, Biological Sciences 33(5):229-338.

. 1989. Phylogenetic interrelationships and evo-
lution of North American late Neogene Equinae. In
The evolution of perissodactyls, ed. D.R. Prothero
and R.M. Schoch. Oxford Monographs on Geology
and Geophysics, no. 15, 178-196. New York: Ox-
ford University Press, xii + 537 pp.

. 1993. Taxonomic evolution in North American

Neogene horses (subfamily Equinae): The rise and
fall of an adaptive radiation. Paleobiology 19(2):216-
234.

Hulbert, R.C., and B.J. MacFadden. 1991. Morpholog-
ical transformation and cladogenesis at the base of
the adaptive radiation of Miocene hypsodont horses.
American Museum Novitates 3000:1-61.

Hussain, S.T. 1975. Evolution and functional anatomy
of the pelvic limb in fossil and Recent Equidae (Per-
issodactyla, Mammalia). Anatomy, Histology, and
Embryology 4:179-222.

James, G.T. 1963. Paleontology and nonmarine stratig-
raphy of the Cuyama Valley Badlands, California.
Part 1. Geology, faunal interpretations, and system-
atic descriptions of Chiroptera, Insectivora, and Ro-
dentia. University of California Publications in
Geological Sciences 45:iv + 154 pp.

Kelly, T.S. 1992. New middle Miocene camels from the
Caliente Formation, Cuyama Valley Badlands, Cal-
ifornia. PaleoBios 13(52): 1-22.

— . 1994. Two Pliocene (Blancan) vertebrate faunas

from Douglas County, Nevada. PaleoBios 16(1):1-
23.

Kelly, T.S., and E.B. Lander. 1988a. Correlation of
Hemingfordian and Barstovian land mammal assem-
blages, lower part, Caliente Formation, Cuyama Val-
ley area, California. American Association of Petro-
leum Geologists Bulletin 72(3):384.

— . 1988b. Biostratigraphy and correlation of Hem-
ingfordian and Barstovian land mammal assem-
blages, Caliente Formation, Cuyama Valley area,
California. In Tertiary tectonics and sedimentation
in the Cuyama Basin, San Luis Obispo, Santa Bar-
bara, and Ventura Counties, California, ed. W.J.M.
Bazeley. Pacific Section, Society of Economic Pa-
leontologists and Mineralogists, Papers and Field
Guide from the Cuyama Symposium and Field Trip,
1-19.

-. 1992. Miocene land mammal faunas from the

Caliente Formation, Cuyama Valley Badlands, Cal-
ifornia. PaleoBios 14(l):3-8.

Kitts, D.B. 1956. American Hyracotherium (Perisso-
dactyla, Equidae). Bulletin of the American Museum
of Natural History 110:1-60.

. 1957. A revision of the genus Orohippus (Per-
issodactyla, Equidae). American Museum Novitates
1864:1-40.

Leidy,J. 1856. Notice of some remains of extinct Mam-
malia, recently discovered by Dr. F.V. Hayden in
the badlands of Nebraska. Proceedings of the Acad-
emy of Natural Sciences, Philadelphia 8:59-60.

. 1857. Notices of extinct Vertebrata discovered

by F.V. Hayden, during the expedition to the Sioux
country under the command of Lieutenant G.K.
Warren. Proceedings of the Academy of Natural
Sciences, Philadelphia 3:311-312.

. 1858. Extinct vertebrates from the valley of the

28 ■ Contributions in Science, Number 455

Niobrara River. Proceedings of the Academy of
Natural Sciences, Philadelphia 4:1-61.

— . 1869. The extinct mammalian fauna of Dakota

and Nebraska. Journal of the Academy of Natural
Sciences, Philadelphia 7:1-472.

Macdonald, J.R., M.L. Macdonald, and L.M. Toohey.
1992. The species, genera, and tribes of the living
and extinct horses of the world 1758-1966, from
the work of Morris F. Skinner. Dakoterra 4:1-429.

MacFadden, B.J. 1984a. Systematics and phylogeny of
Hipparion, Neohipparion, Nannippus, and Cor-
mohipparion (Mammalia, Equidae) from the Mio-
cene and Pliocene of the New World. Bulletin of
the American Museum of Natural History 179(1):
1-196.

. 1984b. Astrohippus and Dinohippus from the

Yepomera local fauna (Hemphillian, Mexico) and
implications for the phylogeny of one-toed horses.
Journal of Vertebrate Paleontology 4(2):273-283.

— — . 1985. Patterns of phylogeny and rates of evo-

lution in fossil horses: Hipparions from the Miocene
and Pliocene of North America. Paleobiology 11(3):
245-257.

— . 1992. Fossil horses. Systematics, paleobiology,

and evolution of the family Equidae. Cambridge,
New York: Cambridge University Press, xii + 369

pp.

MacFadden, B. J., and R.C. Hulbert, Jr. 1988. Explosive
speciation at the base of the adaptive radiation of
Miocene grazing horses. Nature 336:466-468.

MacFadden, B.J., and M.F. Skinner. 1981. Earliest Hol-
arctic hipparion, Cormohipparion goorisi n. sp.
(Mammalia, Equidae) from the Barstovian (medial
Miocene) Texas Gulf Coastal Plain. Journal of Pa-
leontology 55(3):61 9-627.

Madden, C.T. 1987. Primitive proboscidean in Califor-
nia borderlands during the late Miocene: First dwarf
species of Gomphotheriidae. Geological Society of
America, Abstracts with Programs 19(6):319.

Marsh, O.C. 1874. Notice of new equine mammals from
the Tertiary formation. American Journal of Science
7:247-258.

Matthew, W.D. 1926. The evolution of the horse: A
record and its interpretation. Quarterly Review of
Biology 1:139-185.

Merriam,J.C. 1913a. New anchitherine horses from the
Tertiary of the Great Basin area. University of Cal-
ifornia Publications, Bulletin of the Department of
Geology 7(22):41 9-434.

. 1913b. New protohippine horses from Tertiary

beds on the western border of the Mohave Desert.
University of California Publications, Bulletin of
the Department of Geology 7(23):435-441.

. 1915. New horses from the Miocene and Plio-
cene of California. University of California Publi-
cations, Bulletin of the Department of Geology 9(4):
49-58.

- — - — . 1916. Mammalian remains from the Chanac
Formation of the Tejon Hills, California. University
of California Publications, Bulletin of the Depart-
ment of Geology 10(8):1 11-127.

— . 1919. Tertiary mammalian faunas of the Mo-
have Desert. University of California Publications,
Bulletin of the Department of Geology 11(5):437-
585.

Miller, S.T. 1978. Geology and mammalian biostratig-
raphy of a portion of the northern Cady Mountains,
Mojave Desert, California. Master’s thesis, Univer-
sity of California, Riverside.

Mooser, O. 1959. La fauna “Cedazo” del Pleistocene

Kelly: Miocene Horses

en Aguacalientes. Anales del Instituto de Biologia
29:409-452.

Munthe, J. 1979. The Hemingfordian mammal fauna
of the Vedder locality, Branch Canyon Sandstone,
Santa Barbara County, California. Part III: Carniv-
ora, Perissodactyla, Artiodactyla and summary.
PaleoBios 29:1-22.

Osborn, H.F. 1918. Equidae of the Oligocene, Miocene,
and Pliocene of North America. American Museum
of Natural History Memoir 2:1-326.

Quinn, J.H. 1955. Miocene Equidae of the Texas Gulf
Coastal Plain. University of Texas, Bureau of Eco-
nomic Geology Publication 5516:1-102.

Quinn, J.P. 1984. Geology and biostratigraphy of the
Bopesta Formation, southern Sierra Nevada Moun-
tains, Kern County, California. Master’s thesis, Uni-
versity of California, Riverside.

. 1987. Stratigraphy of the middle Miocene Bo-
pesta Formation, souther Sierra Nevada, California.
Contributions in Science 393:1-31.

Rensberger, J.M., A. Forsten, and M. Fortelius. 1984.
Functional evolution of the cheek tooth pattern and
chewing direction in Tertiary horses. Paleobiology
10:439-452.

Savage, D.E. 1955. Nonmarine lower Pliocene sedi-
ments in California. University of California Pub-
lications in the Geological Sciences 31(l):l-26.

Savage, D.E., and D.E. Russell. 1983. Mammalian pa-
leofaunas of the world. Reading, Massachusetts: Ad-
dison-Wesley Publishing Company, xvii + 432 pp.

Scott, K.M., and C.M. Janis. 1993. Relationships of the
Ruminantia (Artiodactyla) and an analysis of the
characters used in ruminant taxonomy. In Mammal
phylogeny, placentals, ed. F.S. Szalay, M.J. Nova-
cek, and M.C. McKenna, 282-302. New York:
Sp ringer- Verlag, xi + 321 pp.

Scott, W.B. 1941. The mammalian fauna of the White
River Oligocene. Part V. Perissodactyla. Transac-
tions of the American Philosophical Society 28:747-
980.

Sellards, E.H. 1916. Fossil vertebrates from Florida: A
new Miocene fauna; new Pliocene species; the Pleis-
tocene fauna. Annual Report of the Florida Geo-
logical Survey 10:11-41.

Simpson, G.G. 1930. Tertiary land mammals from Flor-
ida. Bulletin of the American Museum of Natural
History 59(3):149-211.

. 1951. Horses: The story of the horse family in

the modern world and through sixty million years
of history. New York: Oxford University Press, 245

pp.

Skinner, M.F., and B.J. MacFadden. 1977. Cormohip-
parion n. gen. (Mammalia, Equidae) from the North
American Miocene (Barstovian-Clarendonian).
Journal of Paleontology 51(5):912-926.

Skinner, M.F., and B.E. Taylor. 1967. A revision of the
geology and paleontology of the Bijou Hills, South
Dakota. American Museum Novitates 2300:1-53.

Sondaar, P.Y. 1968. The osteology of the manus of fossil
and Recent Equidae. Nederlandse Akademie van
Wetenschappen, Natuurkunde, Amsterdam 25:1-
76.

Stirton, R. A., and W. Chamberlain. 1939. A cranium
of Pliohippus fossulatus from the Clarendon lower
Pliocene fauna of Texas. Journal of Paleontology
13(3):349-353.

Stock, C. 1935. Deep-well record of fossil mammal re-
mains in California. Bulletin of the American As-
sociation of Petroleum Geologists 19:1064-1068.

1947. A peculiar carnivore from the Miocene,

California. Bulletin of the Southern California
Academy of Sciences 46:84-89.

Swofford, D.L. 1985. PAUP, phylogenetic analysis us-
ing parsimony. User manual version 2.4. Cham-
paign: Illinois Natural History Survey, 70 pp.

Szalay, F.S. 1993. Metatherian taxon phylogeny: Evi-
dence and interpretation from the cranioskeletal sys-
tem. In Mammal phylogeny, Mesozoic differentia-
tion, multituber culates, monotremes, early therians,
and marsupials, ed. F.S. Szalay, M.J. Novacek, and
M.C. McKenna, 216-242. New York: Springer-Ver-
lag, x + 249 pp.

Tedford, R.H., T. Galusha, M.F. Skinner, B.E. Taylor,
R.W. Fields, J.R. Macdonald, J.M. Rensberger, S.D.
Webb, and D.P. Whistler. 1987. Faunal succession
and biochronology of the Arikareean through Hem-
phillian interval (late Oligocene through earliest
Pliocene epoch), North America. In Cenozoic mam-
mals of North America, geochronology and bio-
stratigraphy, ed. M. O. Woodburne, 153-210.
Berkeley: University of California Press, 336 pp.

Troxell, 1916. An early Pliocene one-toed horse, Plio-
hippus lullianus sp. nov. American Journal of Sci-
ence 42:335-348.

Vanderhoof, V.L. 1933. A skull of Pliohippus tantalus
from the later Tertiary of the Sierran foothills of
California. University of California Publications,
Bulletin of the Department of Geological Sciences
23(5):183-193.

Webb, S.D. 1969. The Burge and Minnechaduza Clar-
endonian mammalian faunas of north-central Ne-
braska. University of California Publications in
Geological Sciences 78:1-191.

Webb, S.D., and R.C. Hulbert. 1986. Systematics and
evolution of Pseudhipparion (Mammalia, Equidae)
from the late Neogene of the Gulf Coastal Plain and
the Great Plains. In Vertebrates, phylogeny, and phi-
losophy, ed. K.M. Flanagan and J.A. Lillegraven.
Contributions to Geology, University of Wyoming,
special paper 3; 237-285.

Wiley, E.O. 1981. Phylogenetics, the theory and practice
of phylogenetic systematics. New York: John Wiley
and Sons, 439 pp.

Winans, M.C. 1989. A quantitative study of North
American fossil species of the genus Equus. In The
evolution of perissodactyls, ed. D.R. Prothero and
R.M. Schoch. Oxford Monographs on Geology and
Geophysics, no. 15, 262-297. New York: Oxford
University Press, xii + 537 pp.

Wood, A.E. 1937. Additional material from the Tertiary
of the Cuyama Basin of California. American Jour-
nal of Science 33:29-43.

Woodburne, M.O. 1982. A reappraisal of the system-
atics, biogeography, and evolution of fossil horses.
Paleobiology 8(4):315-327.

. 1989. Hipparion horses: A pattern of endemic

evolution and intercontinental dispersal. In The evo-
lution of perissodactyls, ed. D.R. Prothero and R.M.
Schoch. Oxford Monographs on Geology and Geo-
physics, no. 15, 197-233. New York: Oxford Uni-
versity Press, xii + 537 pp.

Woodburne, M.O., S.T. Miller, and R.H. Tedford. 1982.
Stratigraphy and geochronology of Miocene strata
in the central Mojave Desert, California. In Geologic
excursions in the California desert. Geological So-
ciety of America, Cordilleran Section, Annual Meet-
ing, 78th, Anaheim, California, Volume and Guide-
book, 47-54.

Woodburne, M.O. and R.H. Tedford. 1982. Litho- and
biostratigraphy of the Barstow Formation, Mojave

Contributions in Science, Number 455

Kelly: Miocene Horses ■ 29

Desert, California. In Geologic excursions in the Cal-
ifornia desert. Geological Society of America, Cor-
dilleran Section, Annual Meeting, 78th, Anaheim,
California, Volume and Guidebook, 65-76.
Woodburne, M.O., R.H. Tedford, and C.C. Swisher III.
1990. Lithostratigraphy, biostratigraphy, and geo-

chronology of the Barstow Formation, Mojave Des-
ert, southern California. Bulletin of the Geological
Society of America 102:459-477.

Received 13 July 1994; accepted 3 February 1995.

APPENDICES
Appendix A

Characters and character states used in cladistic analyses presented herein.

Except as noted in text, numbering scheme and descrip-
tions of characters and character states correspond to
those of Hulbert (1988b, 1989) and Hulbert and Mac-
Fadden (1991).

1. Depth of nasal notch. Four states are recognized: 0,
posteriormost point dorsal to about the anterior three
quarters of C-P2 diastema or more anterior; 1, pos-
teriormost point dorsal to anterior half of P2 or just
anterior to P2; 2, posteriormost point dorsal to pos-
terior half of P2-3; 3, posteriormost point dorsal to
P4 or deeper.

2. Frontal bones. Two states are recognized: 0, frontal
bones notably domed; 1, frontal bones flat, not domed.

3. Depth of DPOF. The DPOF is a depression of varying
depth and morphology present in many fossil equids.
Four states are recognized: 0, depth of fossa relative
to the surrounding surface of the skull very shallow,
maximum depth less than 5 mm; 1, depth shallow,
about 5-10 mm; 2, depth moderate, 10-15 mm; 3,
depth deep, greater than 15 mm.

4. Anterior margin DPOF (character 4, Hulbert, 1988b;
character 7, Hulbert, 1989). Two states are recog-
nized: 0, anterior margin of DPOF confluent with face
without a rim; 1, anterior margin well defined with
a pronounced rim.

5. Posterior margin and pocket of DPOF. Four states
are recognized: 0, posterior margin of DPOF without
a pronounced rim, no pocket; 1, posterior margin
with a pronounced rim, but not pocketed; 2, posterior
margin with rim and shallow pocket, less than 5 mm
deep; 3, posterior margin with rim and pocket depth
greater than 5 mm. A pronounced rim means that the
margin of the fossa is very easily discernible because
of a distinct change in slope.

6. Distinct ventral rim on DPOF. Two states are rec-
ognized: 0, ventral rim of DPOF without distinct or
pronounced margin; 1, ventral rim distinctly rimmed.

7. Relative PBL. The PBL is the distance between the
orbit and the DPOF. Relative PBL is determined by
dividing it by UTRL. Three states are recognized: 0,
narrow (ratio < 0.10); 1, moderate (ratio 0.10-0.20);
2, long (ration > 0.20).

11. Zygomatic buckle (Hulbert, 1989; also see Webb,
1969). Two states are recognized: 0, present; 1, ab-
sent.

12. Malar fossa (character 10, Hulbert, 1989; character
12, Hulbert and MacFadden, 1991). In addition to a
DPOF, some equids have a depression in the ventral
preorbital region termed a malar fossa. Three states
are recognized: 0, absent or occasionally present as
a very slight depression; 1, distinct malar fossa always
present and well separated from DPOF; 2, distinct
malar fossa always present and confluent with DPOF.

13. Muzzle width relative to UTRL at moderate wear
stage. Two states are recognized; 0, moderate or nar-
row; 1, broad (> 36%).

14. Incisor arcade. Two states are recognized: 0, arcuate;
1, straight.

16. Relative muzzle length. Character state is determined
by comparison of upper I3-P2 diastema length (UDL)
and UTRL in middle-age adults. Four states are rec-
ognized: 0, short (UDL < 40% UTRL); 1, moderate
(UDL between 40 and 55% of UTRL); 2, elongated
(UDL between 56 and 70% UTRL); 3, very elongated
(UDL > 70% of UTRL).

20. Cement on deciduous premolars. Three states are
recognized: 0, no cement present; 1, cement layer
rudimentary and very thin (< 1 mm thick), commonly
only found on dP4 and dp4; 2, moderate to very thick
coating (> 1 mm thick) of cement on all deciduous
cheek teeth.

21. Cement on permanent cheek teeth. Three states are
recognized: 0, thin layer of cement present, < 1 mm
in thickness; 1, moderate layer of cement present,
about 1 mm thick; 2, thick (> 1.5 mm) layer of
cement present, as in Equus.

22. Orientation of long axis of the protocone of P2-4.
Two states are recognized: 0, approximately antero-
posteriorly; 1, markedly anterolabial-posterolinguai-

ly-

23. Protocone shape (P3-M2). Based on average value of
ratio of PRL to PRW, taken in moderate wear stages.
Four states are recognized: 0, round (ratio < 1.2); 1,
oval (ratio 1. 2-2.0); 2, elongate oval (ratio 2. 1-3.0);
3, elongate (ratio > 3.0).

26. P2 anterostyle. Two states are recognized: 0, large,
expanded; 1, reduced.

27. Timing of protocone connection to protoloph on the
P3 and P4. Seven states are recognized: 0, protocone
connected to protoloph immediately after onset of
occlusal wear; 1 , connected during the very early wear
stage; 2, connected during the early wear stage; 3,
connected during the early moderate wear stage; 4,
connected during the late moderate wear stage; 5,
connected during the late wear stage; 6, protocone
isolated from protoloph to base of crown.

28. Timing of protocone connection to protoloph on the
Ml and M2. Same seven states as character 27.

29. Protocone connection to hypocone on Ml and M2.
Three states are recognized: 0, protocone never con-
nects to hypocone; 1, connection occurs only in late
wear stage; 2, connection occurs prior to late wear
stage.

30. Pli caballin on P2-4. Four states are recognized: 0,
pli caballin absent or very rare; 1, pli caballin com-
mon, but small (< 2 mm) and nonpersistent; 2, pli
caballin well developed, relatively persistent, com-
monly single or unbranched; 3, pli caballin well de-

30 ■ Contributions in Science, Number 455

Kelly: Miocene Horses

veloped, persistent, commonly branched or multiple.

31. Pli caballin on Ml-3. Same four states as character
30.

32. External fossette plications. Three states are recog-
nized: 0, pli protoloph and pli hypostyle rare or, if
present, single and nonpersistent; 1, pli protoloph
and/or pli hypostyle common, persistent, but single;
2, multiple pli protoloph and/or pli hypostyle present
in early wear stages.

33. Internal fossette plications. This character is based
on the common fossette morphology observed in
early and moderate wear stages. Six states are rec-
ognized: 0, all internal fossette plications (pli prefos-
sette, pli postfossette, etc.) absent, or very rare (if
present simple, shallow, and nonpersistent); 1, very
simple internal fossette plications (pli prefossette and
pli postfossette single or absent, can be deep, pre-
fossette loop not prominent); 2, simple but nonper-
sistent internal fossette plications (pli prefossette and
pli postfossette multiple in early wear stages, single
in moderate wear stages, absent in late wear stages)
that are shallow and nonbranching; 3, simple but
persistent internal fossette plications (as in 2 but with
plications generally present in later wear stages); 4,
moderately complex internal fossette plications (two
to five plications present on each side in early and
moderate wear stages, with a limited amount of
branching); 5, complex internal fossette plications
(three to seven plications present on each side in early
and moderate wear stages, branching of plications
common).

35. Metastyle development. Three states are recognized:
0, metastyle generally absent or very weak; 1, meta-
style common but not strong; 2, metastyle very well
developed.

38. Timing of hypoconal groove closure on P2-4. Four
states are recognized: 0, hypoconal groove open to
near the base of the crown; 1, hypoconal groove
closed in late wear stages; 2, hypoconal groove closed
in moderate wear stages; 3, hypoconal groove closed
in early wear stages.

39. Timing of hypoconal groove closure on Ml-2. Same
four states as character 38.

40. Hypoconal lake on P3-4. Two states are recognized:
0, hypoconal groove does not form an isolated lake
when it closes; 1, hypoconal groove does form a lake
after closing.

41. Hypoconal lake on Ml-2. Same two states as char-
acter 40.

43. Curvature of upper cheek teeth (P3-M2). Based on
the ROC measured along the mesostyle (Skinner and
Taylor, 1967). Three states are recognized: 0, strongly
curved (ROC < 40 mm); 1, moderately curved (ROC
from 40 to 80 mm); 2, relatively straight (ROC > 80
mm).

45. Retention of the dpi. Three states are recognized: 0,
dpi relatively large, commonly retained with per-
manent dentition; 1, dpi very reduced (diameter <
2 mm), variable present with adult dentition; 2, dpi
very rarely present with permanent dentition; vestigial
if present.

48. Strength of protostylids on dp3-4. Three states are
recognized: 0, protostylids weak, may be present only
near base of crown as small anterior cingulids, and
do not appear on the occlusal surface until late wear
stages; 1, protostylids moderately well developed, po-
sitioned in anterolabial corner of the tooth, appear
in early wear stage; 2, protostylids very well devel-
oped, extend labially about as far as the protoconid,
straight.

52. Strength of protostylids on p3-m3. Same three states
as character 48.

54. Expansion of metaconid-metastylid complex. Three
states are recognized: 0, metaconid-metastylid rela-
tively small and unexpanded (MML of p3 or p4 av-
erages < 45% of APL); 1, metaconid-metastylid ex-
panded but not elongated (MML of p3 or p4 averages
between 45 and 50% of APL); 2, metaconid-meta-
stylid moderately elongated (MML of p3 or p4 >
50% of APL).

55. Separation of metaconid and metastylid on p3-m3.
Three states are recognized: 0, metaconid and meta-
stylid not well separated from one another even in
very early wear stages; 1, well separated from each
other only in very early and early wear stages; 2,
persistently well separated from each other. Well sep-
arated means that the areas of exposed dentine of
the metaconid and metastylid are distinctly separated
from each other by the linguaflexid, ectoflexid, me-
taflexid, and entoflexid, with only a narrow passage
of dentine connecting them.

56. Metaconid-metastylid on p2. Two states are recog-
nized: 0, single median lingual cuspid present, not
separated into metaconid and metastylid; 1, separate
metaconid and metastylid present on p2, at least in
early wear stage.

57. Development of pli entoflexid. Two states are rec-
ognized: 0, pli entoflexid absent or rudimentary; 1,
pli entoflexid commonly present, at least in early and
very early wear stages.

62. Ectoflexid depth on p2. Three states are recognized:
0, ectoflexid deep, completely penetrates isthmus; 1,
ectoflexid moderately deep, only partially penetrates
isthmus; 2, ectoflexid shallow, does not penetrate
isthmus.

63. Ectoflexid depth on p3-4. Same three states as char-
acter 62.

66. Relative size of metaconid and metastylid on p3-4.
Two states are recognized: 0, metaconid and meta-
stylid equal or subequal in size; 1, metastylid notably
smaller than metaconid and located more lingually.

67. Relative size of the metaconid and metastylid on ml-
m3. Two states are recognized: 0, metaconid and
metastylid equal or subequal in size and position of
their lingual borders; 1, metastylid notably smaller
than metaconid, lingual border located more labially
than that of metaconid especially in moderate and
late wear stages.

70. Tooth row length (character 43, Hulbert, 1989; char-
acter 70, Hulbert and MacFadden, 1991). This char-
acter is used as an indicator of overall size and reflects
the mean UTRL in moderate wear stage. Five states
are recognized: 0, less than 90 mm; 1, about 90-105
mm; 2, about 105-125 mm; 3, about 126-140 mm;

4, greater than 140 mm.

71. Unworn molar crown height. This character is de-
termined by unworn Ml mesostyle crown height or
ml metaconid crown height, ±2.5 mm. Eight states
are recognized: 0, less than 22 mm; 1, about 25 mm;
2, about 30 mm; 3, about 35 mm; 4, about 40 mm;

5, about 45 mm; 6, about 50 mm; 7, > 50 mm.

72. Number of digits (character 44, Hulbert, 1989). Two
states are recognized: 0, tridactyl; 1, monodactyl.

79. Articulation of metacarpal V. Two states are recog-
nized: 0, metacarpal V articulates primarily on the
unciform, articulation facet for metacarpal IV absent
or smaller than that for unciform; 1, metacarpal V
articulates primarily on metacarpal IV, articulation
facet for unciform absent or smaller than that for
metacarpal IV.

Contributions in Science, Number 455

Kelly: Miocene Horses ■ 31

Appendix B

Combined character state matrix for the two groups of taxa previously analyzed separately by Hulbert (1989) and
Hulbert and MacFadden (1991) and using characters and character states presented in Appendix A

t".

r»-.

rv.

-■

rv.

-■

rv.

N-

t-H

T— 1

r-H

t-H

r-H

«/■>

Hfr

rv.

-■

V— H

T— H

r-H

»n

<T>

<l>

N-

( J

t-h

T-H

t"H

N-

»o

N-

rj-

N-

fT)

-■

rv.

-•

rv.

•o

rv.

ffs

rv.

R £ |
R .S
R X v

»■ ''v) *V/>

% x x
a S-

cl .9- .9-
•9“ ~R -R

u u
1 £ £
J s; :=

Acritohippus tertius

Parapliohippus carrizoensis

Protohippus vetus

</>

Q vi
R §

2 5*

g s

& JO

s» o

•S u
"'</)

^ x a
2 Cl Cl

H.9* .9*

Cl -R -R
<o u

“R ^

O v. K
tS <L <L

2SS

&h S S

Merychippus insignis

</>

O Cf3
O <L>
bo C

% %
X X
■Cl CL

•S- .9*

■R -R
N u

L, ^

Pliohippus mirabilis

Hippariort shirley ae

Hipparion tehonense

"CS

<r>

"'vj

•9-

«R

Onohippidium

Equus simplicidens

Dinohippus

Astrohippus

Calippus

Nannippus

Pseudhipparion

Neobipparion

Cormohipparion

32 ■ Contributions in Science, Number 455

Kelly: Miocene Horses

Appendix C

Character state matrix for selected Neogene hypsodont horses using character and character states presented in Appendix A

-•

rv.

! f"

rv.

; so

\>-

”3-

P*-

rv.

•A)

<?\

rv.

rj"

rPt"

C\|

rv.

o..

rt.

rv.

"-1

•52

-22

“Mery chippies” gunteri

«X

’'vt

-X

Acritobippus tertius

Parapliohippus carrizoensis

Protohippus

■*»»

•S

”v,

«X

.xx

“ Merychippus ” coloradense

Merychippus insignis

O rt
O <U

oo e

% *v»
X X
XX XX
.XX

is is

u u

Pliohippus mirabilis

H ipparion shirley ae

H ipparion tehonense

Acritobippus quinni

Heteropliohippus hulberti

•XS

+->

.■Cv

Acritobippus isonesus

Equus simplicidens

Dinohippus

£-

•$-

•3

Astrohippus

Calippus

Nannippus

Pseudhipparion

Neohipparion

Cormohipparion

Contributions in Science, Number 455

Kelly: Miocene Horses ■ 33

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

L- SXX

Number 456
17 October 1995

Contributions
in Science

Additions to the Description of the
Fanworm Genus Pseudofabricia Cantone, 1972

(PoLYCHAETA: SABELLIDAE: FABRICIINAE)

Kirk Fitzhugh

&S&TS Natural History Museum of Los Angeles County

Serial

Publications

of THE
Natural History
Museum of
Los Angeles
County

• Contributions in Science, a miscellaneous series of tech-
nical papers describing original research in the life and
earth sciences.

• Science Series, long articles and collections of papers
on natural history topics.

Copies of the publications in these series are sold through
the Museum Book Shop. A catalog is available on request.

Scientific

Publications

Committee

James L. Powell, Museum President
Daniel M. Cohen, Committee
Chairman
Brian V. Brown
Kenneth E. Campbell
Kirk Fitzhugh
Robin A. Simpson, Managing Editor

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

Printed at Allen Press, Inc., Lawrence, Kansas
ISSN 0459-8113

Additions to the Description of the Fanworm
Genus Pseudofabricia Cantone, 1972

(PoLYCHAETA: SABELLIDAE: FABRICIINAE)

Kirk Fitzhugh1

ABSTRACT. The monotypic genus Pseudofabricia Cantone, 1972, with type species P. aberrans, was
originally described from two specimens, both of which lacked branchial crowns. Neither of these
specimens was, however, formally designated as a type. Only one of these specimens is extant. That
specimen is redescribed and designated as the lectotype and is compared to some of the complete specimens
used by Giangrande and Cantone (1990) in their redescription of the taxon. These latter specimens have
nonvascularized ventral filamentous appendages in the branchial crown, which appear to be homologous
to those in Fabriciola Friedrich, yet differ in that they are only slightly longer than the dorsal lips. Contrary
to what was described by Giangrande and Cantone, the manubria of abdominal uncini in P. aberrans are
about the same length as the dentate region. Based on the present redescription, the cladistic relationships
of Pseudofabricia to other fabriciin taxa are discussed.

INTRODUCTION

The monotypic genus Pseudofabricia Cantone,
1972, with type species P. aberrans , was recently
redescribed from complete specimens by Gian-
grande and Cantone (1990), substantially increasing
knowledge of this taxon. Cantone’s (1972) original
description was based on two living specimens, both
lacking the branchial crown. These specimens were
collected near Siracusa, Sicily, from among the rhi-
zomes of the sea grass Posidonia oceanica at a
depth of about 2 m. Giangrande and Cantone (1990)
based their redescription on specimens collected
from similar habitats at two localities: off the island
of Ponza, northwest of Naples, and from near the
city of Brindisi in the Adriatic Sea. Unfortunately,
no direct comparisons were made with the original
specimens.

Prior to Giangrande and Cantone’s (1990) rede-
scription, Cantone’s (1972: figs. 1=2, pi. 1) illustra-
tions of setal forms and the anterior end of the
body were too limited to fully assess the relation-
ship of the genus to other fabriciin taxa (e.g., Fitz-
hugh, 1989, 1991, 1992). Giangrande and Cantone
(1990) established the position of Pseudofabricia
in the Fabriciinae ( sensu Fitzhugh, 1989, 1991) and
suggested that the genus is most closely related to
Fabricia Blainville and Novafabricia Fitzhugh.
Based on Giangrande and Cantone’s description, I

1. Invertebrate Zoology Section, Research and Col-
lections Branch, Los Angeles County Museum of Natural
History, 900 Exposition Boulevard, Los Angeles, Cali-
fornia 90007.

was able to analyze the cladistic relationships of P.
aberrans to other fabriciin taxa (Fitzhugh, 1993).
The result of that analysis supported the continued
monophyly of all nonmonotypic genera and al-
lowed for the recognition of Pseudofabricia (Fig.
1) as part of a clade comprising Fabricia, Augener-
iella Banse, Parafabricia Fitzhugh, Fabricinuda
Fitzhugh, Novafabricia Fitzhugh, and Pseudofa-
briciola Fitzhugh.

I recently had the opportunity to examine several
of the specimens used in Giangrande and Cantone’s
(1990) redescription and discovered an attribute that
allows for a more precise definition of the genus.
As well, I have obtained from Dr. Cantone the only
remaining specimen of Pseudofabricia aberrans
upon which she based her original description. Un-
fortunately, Cantone (1972) neither designated nor
specifically referred to either of the original speci-
mens as types and made no mention of the final
disposition. With the growing awareness over the
past several years of the diversity and distribution
of fabriciins in the Mediterranean Sea (Giangrande
and Castelli, 1986; Martin and Giangrande, 1991;
Simboura, 1990; Fitzhugh et al, 1994; Fitzhugh and
Simboura, 1995), the need to minimize ambiguity
in the identification of small-bodied sabellids from
this region has become more acute. I will in the
present paper designate Cantone’s original speci-
men as the lectotype of P. aberrans as well as com-
pare this specimen with some of the nontype ma-
terial. All specimens examined have been deposited
in the Allan Hancock Foundation Polychaete Col-
lection of the Los Angeles County Museum of Nat-
ural History (LACM-AHF).

Contributions in Science, Number 456, pp. 1-6
Natural History Museum of Los Angeles County, 1995

Genus A
Monroika
Manayunkia
Fabriciola
Pseudofabriciola
Fabricia
Pseudofabricia
Parafabricia
Fabricinuda
Augeneriella
Novafabricia

Figure 1. Strict consensus cladogram, modified from
Fitzhugh (1993: fig. 3), indicating the relationship of Pseu-
dofabricia to other Fabriciinae genera.

SYSTEMATICS

Pseudofabricia aberrans
Cantone, 1972, emended
Figures 2-4

Pseudofabricia aberrans Cantone, 1972: 4, figs. 1-

2, pi. 1; Giangrande and Cantone, 1990: 363, figs.

1-3.

MATERIALS EXAMINED. Lectotype (LACM-
AHF 1708, slide mount), designated herein; Med-
iterranean Sea, Island of Sicily, near Siracusa, Pos-
idonia oceanica grass bed, depth about 2 m. Nine
nontype specimens (three complete, six lacking
branchial crown, LACM-AHF 1709), Tyrrhenian
Sea, Island of Fonza, near Lazio, among rhizomes
of Posidonia oceanica, depth 1 m, 1989.

DESCRIPTION OF LECTOTYPE. Specimen in

Figure 2. Lectotype of Pseudofabricia aberrans (LACM-AHF 1708), left side, entire specimen (figure split between
setigers 6 and 7). a, anterior half of body, b, posterior half of body; arrows indicate thorax-abdomen boundary.
Abbreviations: apr, anterior peristomial ring; ppr, posterior peristomial ring; vl, ventral lobe of anterior peristomial ring.

2 ■ Contributions in Science, Number 456

Fitzhugh: Redescription of Pseudofabricia

poor condition, mounted on slide, somewhat lat-
erally flattened (Fig. 2). Body with 8 thoracic and
3 abdominal setigers, branchial crown absent. Total
body length 1.30 mm, maximum width 0.25 mm
at setiger 3. Ventral margin of anterior peristomial
ring developed as broad lobe, about same length
as remainder of ring. Posterior peristomial ring
slightly shorter than anterior ring (exclusive of ven-
tral lobe). Annulation between anterior and pos-
terior rings only visible ventrally. Setigers 1-2 each
slightly longer than posterior peristomial ring. Ven-
trum of setiger 3 about same length as setiger 2,
dorsum of setiger 3 distinctly inflated, about 2 times
longer than ventrum. Setiger 4 about 2.5 times lon-
ger than setiger 1. Setigers 5-8 each about same
length as setiger 4 or slightly shorter. Anterior and
posterior regions of thorax slightly tapered. Ab-
dominal setigers of equal width, only slightly nar-
rower than posterior thoracic region. Setiger 9
slightly shorter than setiger 8, setiger 10 about one-
half length of setiger 9, setiger 11 and pygidium
each about one-half length of setiger 10. Posterior
margin of pygidium rounded. Anterior peristomial
ring and pygidium each with pair of dark brown
eyes. Superior thoracic notosetae elongate, narrow-
ly hooded; 3-5 per fascicle. Inferior thoracic no-
tosetae in setigers 2-8 short forms of elongate, nar-
rowly hooded form; 1-2 per fascicle. Abdominal
neurosetae modified, elongate, narrowly hooded;
1-3 per fascicle. Thoracic uncini acicular, in single
row of 4-7 per fascicle; large tooth above main
fang followed by series of smaller teeth. Abdominal
uncini with 10-11 rows of teeth, 3-5 teeth per row;
manubrium about same length as dentate region;
uncini not viewed in profile. Body pigmentation
absent.

REMARKS ON LECTOTYPE. In comparison
with Cantone’s (1972: fig. 1, pi. 1; Fig. 3) illustra-
tions of this species while still alive, the lectotype
is considerably contracted, the slide preparation
probably causing additional distortion by lateral
compression. The poorly preserved and incomplete
condition of the lectotype makes comparisons dif-
ficult with specimens described by Giangrande and
Cantone (1990). Comparing the lectotype and Pon-
za specimens, the latter being described below, with
Cantone’s (1972) original description points to a
misinterpretation of the anterior end. For instance,
Cantone’s (1972: fig. 1; Fig. 3) illustrations of the
whole animal, which were made from the two orig-
inal specimens while they were still alive, show the
anterior end with the ventral lobe of the anterior
peristomial ring distinctly demarcated by a very deep
gap from a large dorsal lobe (Fig. 3b). This dorsal
lobe probably represents at least a portion of the
narrow, median ridge situated just above the mouth,
which is present in most fabriciins (e.g., Fig. 4a). In
all specimens examined, this ridge is not nearly as
pronounced as indicated by Cantone, and the deep
gap between the ridge and the ventral lobe is an
artifact. It is not uncommon to find the ventral lobe
curved over the anterior end when the branchial

Contributions in Science, Number 456

a b c

Figure 3. Pseudo fabricia aberrans, modified from Can-
tone (1972: fig. 1). Abbreviations: apr, anterior peristomial
ring; m, mouth; ppr, posterior peristomial ring; vl, ventral
lobe of anterior peristomial ring.

crown is missing (e.g., Cantone, 1972: figs. 1-2, pi.
1; Giangrande and Cantone, 1990: fig. IB), which
at first sight can give the impression that the peri-
stomium margin is incised just above the lobe. This
interpretation of the anterior peristomial ring is also
seen in the illustrations of Giangrande and Cantone
(1990: figs. IB, 2B-D), discussed below.

ADDITIONS TO DESCRIPTION BASED ON
PONZA SPECIMENS. Inner margin of each bran-
chial lobe with a low, triangular dorsal lip (Fig. 3a,
b). One pair of nonvascularized ventral filamentous
appendages present just ventral to each dorsal lip;
appendages slightly longer than dorsal lips to up to
1.5 times longer than lips; surface of appendages
ranges from smooth (Fig. 4a) to wrinkled (Fig. 4b);
width relatively uniform except for slight widening
at base; distal end rounded. Ventral margin of an-
terior peristomial ring as a broad, flattened lobe;
proximal margin distinctly continuous with re-
mainder of anterior ring (Fig. 4a). Anterior peristo-

Fitzhugh: Redescription of Pseudofabricia ■ 3

Figure 4. Pseudofabricia aberrans from Ponza Island, Italy (LACM-AHF 1709). a, anterior end, right side (right half
of branchial crown removed, distal end of dorsal lip is folded over), b, inner margin of right half of branchial crown,
c, abdominal uncinus from setiger 9. Abbreviations: apr, anterior peristomial ring; bh, branchial heart; dl, dorsal lip;
mdr, middorsal ridge; ppr, posterior peristomial ring; vfa, ventral filamentous appendage; vl, ventral lobe of anterior
peristomial ring.

mial ring about one-half length of posterior ring.
Annulation between anterior and posterior rings
distinct dorsolaterally and ventrally. Abdominal un-
cini with 10-11 rows of teeth (Fig. 4c), 3-5 teeth
per row; manubrium slightly constricted distally
and about same length as dentate region.

REMARKS ON PONZA SPECIMENS. There
are two notable features presented here that differ
from those described by Giangrande and Cantone
(1990). First, ventral filamentous appendages were
overlooked in the redescription (e.g., Giangrande
and Cantone, 1990: fig. 2D). These structures are
extremely difficult to see without first completely
separating the branchial lobes and examining the
inner margin of a lobe with a compound micro-
scope. In terms of their position and structure rel-
ative to the dorsal lips, these appendages show a
striking resemblance to the nonvascularized ventral
filamentous appendages seen in all species of Fa-
briciola Friedrich (see Fitzhugh, 1990; Rouse, 1993).
The difference, however, is that the appendages in

Fabriciola are considerably longer, usually termi-
nating near the distal end of the crown.

Second, Giangrande and Cantone (1990: figs. 1,
2B-D) illustrated the ventral lobe of the anterior
peristomial ring as though it were distinctly sepa-
rated from the remainder of the ring, almost ap-
pearing to originate from the posterior peristomial
ring, a condition similar to what Cantone (1972:
fig. 1; Fig. 3c) illustrated. Giangrande and Cantone
stated, however, that the ventral lobe was a con-
tinuation of the anterior margin of the anterior ring.
It is likely that they mistook folding of the body
wall in the area of the ventral lobe to indicate the
lobe’s proximal boundary. Shown here in Figure
4a, the placement of the ventral lobe is identical
to what is seen in other fabriciins with this structure
(e.g., Fabricia, Parafabricia, Novafabricia).

Giangrande and Cantone (1990: 364, fig. 3E) stat-
ed that the manubria of abdominal uncini were
“about twice as long as the dentate region,” which
is not in agreement with what they illustrated. In

4 ■ Contributions in Science, Number 456

Fitzhugh: Redescription of Pseudofabricia

'Genus A'
Manayunkia

Monroika

Fabriciola

Pseudofabriciola

Pseudofabricia

Fabricia

Parafabricia

Augeneriella

Novafabricia

Fabricinuda

'Genus A'
Pseudofabricia

Fabriciola

Monroika

Manayunkia

Fabricia

Parafabricia

Augeneriella

Fabricinuda

Novafabricia

Pseudofabriciola

’Genus A'

Fabricia

Pseudofabricia

Fabriciola

Monroika

Manayunkia

Pseudofabriciola

Novafabricia

Parafabricia

Augeneriella

Fabricinuda

‘Genus A'
Manayunkia

Monroika

Pseudofabricia

Fabriciola

Fabricia

Parafabricia

Augeneriella

Novafabricia

Fabricinuda

Pseudofabriciola

Figure 5. Selected cladograms indicating topological variation among fabriciin genera. Pseudofabricia was coded with
nonvascularized ventral filamentous appendages present and homologous to those in Fabriciola. Analyses were performed
with all available species in each genus, but only genera are indicated here if monophyletic. Paraphyletic genera, and
included species, are delimited by a bar.

most instances, the manubrium is about the same
length as the dentate region. (Fig. 4c), though I have
seen some uncini with a slightly longer manubrium.
In no instance have I found manubria twice as long
as the dentate region.

I commented earlier (Fitzhugh, 1993: 10) on the
fact that Pseudofabricia is characterized by ple-
siomorphic characters. While this does not tech-
nically affect the integrity of the genus at this time,
as it is monotypic, the presence of very short, non-
vascularized ventral filamentous appendages easily
suffices as a synapomorphy in the event additional
Pseudofabricia species are described.

DISCUSSION

Relative to other fabriciin sabellids, Pseudofabricia
aberrans possesses an unusual combination of char-
acter states, certainly pointing to the need for a
reassessment of the relationship of this taxon to
other fabriciin genera. My earlier inclusion of Pseu-
dofabricia in a cladistic analysis of fabriciin genera
and species (Fitzhugh, 1993) placed the genus in a
clade including Pseudofabriciola, Parafabricia, Fa-
bricia, Novafabricia, Augeneriella, and Fabrici-

Contributions in Science, Number 456

nuda (Fig. 1), this clade being defined by the pres-
ence of a large tooth above the main fang in tho-
racic uncini. There was certainly nothing remark-
able about the placement of Pseudofabricia,
especially given the dentition pattern in thoracic
uncini, coupled with the presence of the ventral
lobe on the anterior peristomial ring, both of which
are common to most of these genera. In a subse-
quent series of unpublished cladistic analyses I have
performed with fabriciin genera and species, in
which nonvascularized ventral filamentous append-
ages are coded as present in Pseudofabricia (ho-
mologous to Fabriciola appendages), a substantially
greater number of tree topologies exist and far
greater instability occurs with regard to relation-
ships among genera. While some topologies are
consistent with those obtained in previous analyses
(e.g., Fig. 5a), I have observed two notable topo-
logical arrangements that have not been observed
previously: (1) Fabriciola, Novafabricia, and Au-
generiella are capable of being paraphyletic (e.g.,
Fig. 5b-d), and (2) Pseudofabricia can be the sister
group to Fabriciola by the presence of nonvascu-
larized filamentous appendages in both genera (e.g.,
Fig. 5b, c, e).

Fitzhugh: Redescription of Pseudofabricia ■ 5

The marked effects on tree topology and mono-
phyly due to the inclusion of filamentous append-
ages in Fseudofabricia readily precludes a straight-
forward reassessment of relationships among gen-
era at this time. The extreme sensitivity of tree
topologies to even minor modifications in character
data is a clear indication of the lack of sufficient
data relative to the ever increasing number of taxa
being described. The very small body sizes of fa-
briciin sabellids present distinct limits on the avail-
ability of gross morphological characters typically
used in sabellid systematics. This increasing dispar-
ity between the number of characters and taxa sim-
ply heightens the level of homoplasy, the principal
result being the more tenuous recognition of gen-
era. A suite of very promising characters is being
developed by Rouse (e.g., 1992, 1993, 1995; Rouse
and Fitzhugh, 1994), centering on reproductive and
developmental features. Indeed, based on his survey
of reproductive characters in Fabriciola, Rouse
(1993: 250) suggested that, once these characters
are taken into consideration cladistically, Fabri-
ciola might be paraphyletic. Such a view is sup-
ported by some of the topologies presented here
(Fig. 5b-d). The further integration of Rouse’s re-
productive data with characters currently available
will likely afford greater resolution to relationships
among fabriciin taxa.

ACKNOWLEDGMENTS

Sincere thanks are extended to Drs. Grazia Cantone (Uni-
versita di Catania, Italy) and Adriana Giangrande (Uni-
versita Degli Studi di Leece, Italy) for their answers to my
numerous questions and their willingness to make spec-
imens available for study. An earlier version of the manu-
script was greatly improved by comments from Thomas
H. Perkins and Jerry D. Kudenov.

LITERATURE CITED

Cantone, G. 1972. Fseudofabricia aberransn.gen.n.sp.,
un anellide polichete di incerta sede. Bollettino delle
Sedute della Accademia Gioenia di Scienze Naturali
in Catania, Serie IV 11:1-7.

Fitzhugh, K. 1989. A systematic revision of the Sabel-
lidae-Caobangiidae-Sabellongidae complex (Anne-
lida: Polychaeta). Bulletin of the American Museum
of Natural History 192:1-104.

— . 1990. Revision of the Fabriciinae genus Fabri-
ciola Friedrich, 1939 (Polychaeta: Sabellidae). Zool -
ogica Scripta 19:153-164.

— . 1991. Further revisions of the Sabellidae sub-

families and cladistic relationships among the Fa-
briciinae (Annelida: Polychaeta). Zoological Journal
of the Linnean Society 102:305-332.

. 1992. On the systematic position of Monroika

africana (Monro) (Polychaeta: Sabellidae: Fabrici-
inae) and a description of a new fabriciin genus and
species from Australia. Proceedings of the Biological
Society of Washington 105:116-131.

— . 1993. Novafabricia brunnea (Hartman, 1969),

new combination, with an update on relationships
among Fabriciinae taxa (Polychaeta: Sabellidae).
Contributions in Science 438:1-12.

Fitzhugh, K., A. Giangrande, and N. Simboura. 1994.
New species of Fseudofabriciola Fitzhugh, 1990
(Polychaeta: Sabellidae: Fabriciinae), from the Med-
iterranean Sea. Zoological Journal of the Linnean
Society 110:219-241.

Fitzhugh, K., and N. Simboura. 1995. An update on the
systematics and occurrence of the fan worm genus
Fseudofabricia Fitzhugh, 1990 (Polychaeta: Sabel-
lidae: Fabiciinae) in the Mediterranean. Contribu-
tions in Science (in press).

Giangrande, A., and G. Cantone. 1990. Redescription
and systematic position of Fseudofabricia aberrans
Cantone, 1972 (Polychaeta, Sabellidae, Fabriciinae).
Bollettino Zoologica 57:361-364.

Giangrande, A., and A. Castelli. 1986. Occurrence of
Fabricia filamentosa Day, 1963 (Polychaeta, Sabel-
lidae, Fabriciinae) in the Mediterranean Sea. Oeba-
lia, n.s. 13:119-122.

Martin, D., and A. Giangrande. 1991. Novafabricia bi-
lobata sp. nov. (Polychaeta, Sabellidae, Fabriciinae)
from the Mediterranean. Ophelia 33:113-120.

Rouse, G.W. 1992. Ultrastructure of the spermathecae
of Parafabricia ventricingulata and three species of
Oriopsis (Polychaeta: Sabellidae). Acta Zoologica
(Stockholm) 73:141-151.

— . 1993. New Fabriciola species (Polychaeta, Sa-
bellidae, Fabriciinae) from the eastern Atlantic, with
a description of sperm and spermathecal ultrastruc-
ture. Zoologica Scripta 22:249-261.

— — — . 1995. Is sperm ultrastructure useful in poly-
chaete systematics? An example using 20 species of
the Fabriciinae (Polychaeta: Sabellidae). Acta Zool-
ogica (Stockholm) (in press).

Rouse, G.W., and K. Fitzhugh. 1994. Broadcasting fa-
bles: Is external fertilization really primitive? Sex,
size, and larvae in sabellid polychaetes. Zoologica
Scripta 23:271-312.

Simboura, N. 1990. Fabricia filamentosa Day, 1963
(Polychaeta, Sabellidae, Fabriciinae) a lessepsian mi-
grant in Mediterranean Sea. Oebalia, n.s. 16:129-
133.

Received 6 July 1994; accepted 10 January 1995.

6 ■ Contributions in Science, Number 456

Fitzhugh: Redescription of Pseudo fa bricia

Number 457
17 October 1995

Contributions
in Science

An Update on the Systematics and
Occurrence of the Fanworm Genus
P SEUDOFABRICIOLA FlTZHUGH, 1990
(Polychaeta: Sabellidae: Fabriciinae)
in the Mediterranean

Kirk Fitzhugh and Nomiki Simboura

^5^ Natural History Museum of Los Angeles County

Serial
Publications
of THE
Natural History
Museum of
Los Angeles
County

iff

oct 2 6 1995

-pB RARitS_

• Contributions in Science, a miscellaneous series of tech-
nical papers describing original research in the life and
earth sciences.

• Science Series, long articles and collections of papers
on natural history topics.

Copies of the publications in these series are sold through
the Museum Book Shop. A catalog is available on request.

Scientific

Publications

Committee

James L. Powell, Museum President
Daniel M. Cohen,
Committee Chairman
Brian V. Brown
Kenneth E. Campbell
Kirk Fitzhugh
Robin A. Simpson, Managing Editor

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

Printed at Allen Press, Inc., Lawrence, Kansas
ISSN 0459-8113

An Update on the Systematics and

Occurrence of the Fanworm Genus
PSEUDOFABRICIOLA FITZHUGH, 1990
(POLYCHAETA: SABELLIDAE: FABRICIINAE)

in the Mediterranean

Kirk Fitzhugh1 and Nomiki Simboura2

ABSTRACT. Fitzhugh et al. (1994) originally described Pseudo fabriciola analis from the Adriatic Sea
off Italy at a depth of 22 m and P. longipyga off the island of Lesvos in the northeast Aegean Sea at 21
m. Range extensions for both species are reported. Pseudo fabriciola analis has been found in the Ionian
Sea off the west coast of Greece at a depth of 15 m, and P. longipyga appears to be distributed throughout
the Aegean Sea, having been found off the island of Rhodes (63-71 m), on the Cyclades Island plateau
(112 m), and off Greece in the Saronikos Gulf (85 m). The description of P. longipyga is emended based
on this additional material to include features of the branchial crown, which was originally unknown.
Specimens also display varying degrees of contraction of anterior and posterior ends, which can result in
body dimensions significantly differing from what was described for the types. Unlike the holotype and
paratype of P. longipyga, all nontype specimens have equally developed pygidial eyes. Cladistic relation-
ships of P. longipyga to other species in the genus are reassessed.

INTRODUCTION

Two fanworm species in the genus Pseudofabri-
ciola Fitzhugh, 1990, P. analis and P. longipyga ,
were recently described by Fitzhugh et al. (1994).
Both species are known only from the Mediterra-
nean, with P. analis originally described from the
Adriatic Sea off Brindisi, Italy, and P. longipyga
from the northeast Aegean Sea off the island of
Lesvos (Fig. 1). Additional specimens of both spe-
cies have been made available to us through the
National Centre for Marine Research, Greece, as
a result of benthic surveys carried out in the Ionian
and Aegean Seas. One of the objectives of the pres-
ent paper is to present range extensions for both
species based on this new material.

Pseudofabriciola longipyga was originally de-
scribed from two specimens, both lacking a bran-
chial crown. This species was first identified as Fa -
bricia filamentosa ( non Day, 1963) by Simboura
(1990), who reported its occurrence off Lesvos (the
type locality) and Rhodes. At the time they de-
scribed P. longipyga, Fitzhugh et al. (1994: 229)
considered the specimens from Rhodes to be a

1. Invertebrate Zoology Section, Research & Collec-
tions Branch, Natural History Museum of Los Angeles
County, 900 Exposition Boulevard, Los Angeles, Cali-
fornia 90007, USA.

2. National Centre for Marine Research, Gr. 16604
Hellenikon, Athens, Greece.

Contributions in Science, Number 457, pp. 1-10
Natural History Museum of Los Angeles County, 1995

separate, undescribed species. Subsequently, how-
ever, additional specimens of Pseudofabriciola have
been recovered from other localities in the Aegean
Sea, which exhibit a range of overlapping morpho-
logical variation sufficient to suggest that all are
members of P. longipyga. We here present addi-
tions to the description of P. longipyga as well as
a description of the branchial crown of this species.
Cladistic relationships between this species and oth-
ers in the genus are also reevaluated.

All specimens examined have been deposited in
the Allan Hancock Foundation Polychaete Collec-
tion of the Natural History Museum of Los Angeles
County (LACM-AHF).

SYSTEMATICS

Pseudofabriciola analis
Fitzhugh, Giangrande, and
Simboura, 1994

Fabricia filamentosa. Giangrande and Castelli, 1986:
119-122, fig. 1 ( non Day, 1963); Giangrande, 1989:
163 ( non Day, 1963).

Pseudofabriciola analis Fitzhugh, Giangrande, and
Simboura, 1994: 221-226, figs. 2-5.

MATERIAL EXAMINED. Western Ionian Sea:
west coast of Peloponissos Peninsula, depth 15 m,
sandy silt, 24 July 1982; 14 specimens (LACM-AHF
1710).

Figure 1. Known occurrences of Pseudofabriciola analis (circles) and P. longipyga (squares). Type localities are
indicated by an asterisk, other localities by diagonal lines.

REMARKS. All specimens are complete except
for the branchial crown. Total body length ranges
from 1.5 to 3.0 mm. The specimens agree well with
the description of the species given by Fitzhugh et
al. (1994). Pseudofabriciola analis is easily identi-
fied by the shallow, ventral anal depression on the
pygidium (cf. Fitzhugh et al., 1994: fig. 4A, C) and
especially the paired midventral lobes that border
the anterior margin of the depression. Another im-
portant diagnostic feature is the arrangement of
notches on the anterior peristomial ring collar.
Pseudofabriciola analis has two pairs of dorsolat-
eral notches but lacks a middorsal notch or incision
(Table 1; cf. Fitzhugh et al., 1994: fig. 3C). The
other five species with dorsolateral notches only
have one pair, whereas two species lack notches
altogether.

With respect to depth, the occurrence of Pseu-
dofabriciola analis in the Ionian Sea is comparable
to the depth (22 m) at which the types were col-
lected and is consistent with the pattern of subtidal
distribution recognized for most Pseudofabriciola
species by Fitzhugh et al. (1994: 238-240, fig. 14).
Similarly, the extension of the species’ range from
off southeastern Italy to the southwest coast of
Greece is not surprising.

Pseudofabriciola longipyga
Fitzhugh, Giangrande, and
Simboura, 1994, emended
Figures 2-7

Fabricia filamentosa. Simboura, 1990: 129-133 ( non
Day, 1963).

Pseudofabriciola longipyga Fitzhugh, Giangrande,
and Simboura, 1994: 226-230, figs. 6-7.

MATERIAL EXAMINED. Southeast Aegean
Sea: Rhodes Island, depth 71 m, corallagenous sub-
strate, August 1983, 7 specimens (LACM-AHF 1713,
only 1 specimen with branchial crown); Rhodes
Island, depth 63 m, muddy sand, February 1984, 7
specimens (LACM-AHF 1714, all lacking branchial
crown). South Aegean Sea: Cyclades Island plateau,
37°18'30"N, 25°23'45"E, depth 112 m, muddy sand
with detritus, September 1989, 2 specimens (LACM-
AHF 1712, all lacking branchial crown). Southwest
Aegean Sea: Saronikos Gulf, 37°48'00"N,
23°4T50"E, depth 85 m, muddy sand, 26 April 1990,
8 specimens (LACM-AHF 1711, all lacking bran-
chial crown).

ADDITIONS TO THE DESCRIPTION. All

specimens slender, elongate. Total body length (ex-

2 ■ Contributions in Science, Number 457

Fitzhugh and Simboura: Mediterranean Pseudofabriciola

Table 1. Comparisons of the principal features among described Pseudofabriciola species.

Species

Dorsal lips

Ventral

filamentous

appendages

Middorsal

collar

surface

Middorsal

collar

notch

Dorsolateral

notches

P. analis

Narrow ridge

Absent

Smooth

Absent

Two pairs

P. australiensis

Unknown

Smooth

Present

One pair

P. californica

Triangular lobe

Absent

Grooved

Absent

P. capensis

Triangular lobe

Absent

Grooved

Absent

One pair

P. filamentosa

Unknown

Absent

Smooth

Absent

One pair

P. incisura

Triangular lobe

Present

Smooth

Present

One pair

P. longa

Narrow ridge

Absent

Smooth

Absent

P. longipyga

Narrow ridge

Absent

Smooth

Present

One pair

elusive of branchial crown) from 2.1 to 6.3 mm
(Table 2). Specimen with branchial crown 4.0 mm
long, crown comprising 1.0 mm of this length.
Crown with 3 pairs of radicles, each terminating
to a filamentous tip; each radiole with 3-5 pairs of
pinnules, all terminating at same height as radioles.
Branchial lobes elongate, narrow (Fig. 5A), with
attachment point to peristomium as a peduncle-
like appendage. Dorsal lips low, broadly rounded
ridges; ventral lips absent; ventral filamentous ap-
pendages absent. Anterior peristomial ring collar as

described by Fitzhugh et al. (1994; Figs. 2, 6), with
narrow, V-shaped middorsal notch or incision and
pair of shallow dorsolateral notches (Fig. 3A); one
specimen with lateral collar margins distinctly un-
even (Fig. 7), presumed to be a malformation. Gen-
eral dimensions of the anterior end of some spec-
imens similar to the types, anterior end of other
specimens can appear contracted or inflated to vary-
ing degrees. In extremely contracted specimens (Fig.
2), posterior peristomial ring and setigers 1 to 3
with broad dorsum, giving lateral margins of seti-

Figure 2. Anterior end of Pseudofabriciola longipyga (LACM-AHF 1714) from Rhodes. Note inflated, or swollen,
state of setiger 1. A, dorsal view. B, lateral view, left side. C, ventral view. Abbreviations: c, anterior peristomial ring
collar; ppr, posterior peristomial ring.

Contributions in Science, Number 457

Fitzhugh and Simboura: Mediterranean Pseudofabriciola ■ 3

Figure 3. Pseudofabriciola longipyga (LACM-AHF 1714) from Rhodes. A, Head-on view of anterior end, branchial
crown removed. B, abdominal setigers and pygidium, dorsal view. C, posterior end, left side, of specimen shown in B.
Abbreviations: li, dorsolateral collar incision; m, mouth; mi, middorsal collar incision; s, swollen region on ventrum of
setiger 10.

gers a distinctly swollen appearance. Ventrum of
setiger 10 of most specimens with a low, rounded
swelling as described in the types (Figs. 3C, 4A-B;
e.g., Fitzhugh et ah, 1994: fig. 1A, C); midventral
surface of swelling with a single, or paired, opaque,
roughly triangular patch (Fig. 4A), which extends
into the body less than one-quarter of the width
of the setiger (Figs. 3C, 4B). Some specimens with
swelling and patch(es) indistinct, though area oc-

cupied by patch(es) more visible after specimen has
been stained with methyl green. Paired pygidial eyes
black or dark brown and equally developed (Figs.
3B-C, 4B, 5B). Pygidium shape variable, exhibiting
three basic shapes: (1) narrow, very elongate con-
dition seen in types (e.g., Fitzhugh et al., 1994: fig.
1A-C); (2) slightly longer than wide and more tri-
angular (Fig. 5B); and (3) nearly as wide as long,
posterior margin broadly rounded, with a small,

4 ■ Contributions in Science, Number 457

Fitzhugh and Simboura: Mediterranean Pseudofabriciola

Figure 4. Pseudo fabriciola longipyga (LACM-AHF 1714) from Rhodes. A, setiger 10, ventral view. B, posterior end,
left side, of specimen shown in A. C, thoracic uncinus from setiger 4. D, abdominal uncinus from setiger 9. Abbreviation:
s, swollen region on vent rum of setiger 10.

blunt caudal-like appendage between the pygidial
eyes (Fig. 3B-C). Thoracic uncini acicular (Fig. 4C),
teeth above main fang gradually decreasing in size,
hood present. Abdominal uncini (Fig. 4D) with 8-
9 rows of teeth, 2-4 teeth per row; manubrium
slightly constricted below dentate region and 1.5-
2 times longer than dentate region. Some specimens
from Rhodes with light brown pigmentation in
middorsal region of collar and anterior middorsal
margin of posterior peristomial ring. Tubes com-
posed of fine mud, some foraminiferan tests. No
brooding of young observed.

REMARKS. In comparison to the types, the
specimens described here show a range of variation
in the anterior end and pygidium. These differences
among specimens appear to be due to varying de-
grees of contraction as a result of fixation. The most

Table 2. Total body length (in millimeters), exclusive
of branchial crown, of complete Pseudofabriciola lon-
gipyga specimens examined, and mean body lengths from
each area.

Rhodes

Saronikos

Cyclades

Island

Sta. R1

Sta. R3

Gulf

plateau

2.1

3.5

3.4

2.3

4.0

3.6

2.5

2.3

4.0

3.6

2.4

5.0

4.2

2.5

5.3

4.3

3.0

5.5

4.6

3.2

6.3

x = 2.5

x = 4.8

x = 3.9

x = 2.4

notable divergence from the types that does not
appear to be a preservation artifact is in the pygidial
eyes. Fitzhugh et al. (1994: fig. 6B-C) reported that
both the holotype and paratype have a well-devel-
oped right eye and a left eye consisting of a group
of small, dispersed eyespots. All specimens exam-
ined here have pygidial eyes equally developed. We
have, however, chosen to place these specimens in
Pseudofabriciola longipyga primarily on the basis

Figure 5. Pseudofabriciola longipyga (LACM-AHF
1713) from Rhodes. A, inner margin of left half of bran-
chial crown. B, posterior end, dorsal view. Abbreviation:
dl, dorsal lip.

Contributions in Science, Number 457

Fitzhugh and Simboura: Mediterranean Pseudofabriciola ■ 5

Figure 6. Anterior end of Pseudo fabriciola longipyga (LACM-AHF 1713) from Rhodes. A, dorsal view. B, lateral
view, right side. Abbreviations: c, anterior peristomial ring collar; li, dorsolateral collar incision; mi, middorsal collar
incision; ppr, posterior peristomial ring.

of similarities in collar construction and setal types
and the occurrence of the midventral swelling on
setiger 10. Ideally, more material from the type
locality is required to determine the consistency of
pygidial eye development in this species.

The branchial crown of Pseudofabriciola Ion -
gipyga is similar to that of P. longa and P. analis
in that the dorsal lips are represented by low, nar-
row ridges (Table 1), in contrast to the plesiomor-
phic condition in which the lips are erect, triangular
processes. These species differ in that the collar
margin in P. analis has two pairs of dorsolateral
incisions, while P. longipyga has only one pair and
P. longa lacks incisions on the collar.

Pseudofabriciola longipyga appears to be a wide-
spread species in the Aegean Sea. The known depth
range for the species is now 21 (for the types) to
112 m. Simboura (1990) reported the occurrence
of P. longipyga (as Fabricia filamentosa Day) from
the vicinity of the type locality (Geras Gulf, Lesvos
Island) at depths ranging from 6 to 21 m. Unfor-
tunately, we cannot confirm these depth records

6 ■ Contributions in Science, Number 457

because the specimens no longer exist (Fitzhugh et
ah, 1994: 229).

RELATIONSHIP OF
PSEUDOFABRICIOLA LONGIPYGA
TO OTHER SPECIES
IN THE GENUS

Analyses of the cladistic relationships among Pseu-
dofabriciola species have been performed by Fitz-
hugh (1991a; see also Fitzhugh, 1991b, 1992, 1993)
and Fitzhugh et al. (1994). The latter study included
P. longipyga, but data on the branchial crown were
lacking because only the types were used. A further
consideration of the relationship of this species to
others in the genus is presented here.

CHARACTERS AND CLADISTIC
ANALYSIS

The nine characters used in the present analysis
(Table 3) are the same as those used by Fitzhugh
et al. (1994: table 1). In addition to the inclusion

Fitzhugh and Simboura: Mediterranean Pseudofabriciola

Table 3. Characters and states used to determine cla-

distic relationships among Pseudofabriciola species. State

0 is plesiomorphic based on outgroup comparisons.

1 . Dorsal lips: (0) well-developed, triangular lobe; (1) low,
narrow ridge.

2. Branchial lobe shape: (0) wide and short; (1) narrow
and elongate, and/or with a peduncle-like process.

3. Anterior peristomial ring collar: (0) low ridge or mem-
brane; (1) high, with even sides; (2) high, with flaring
sides.

4. Middorsal collar condition: (0) separate; (1) entire and
distinctly grooved; (2) entire and smooth.

5. Middorsal collar margin: (0) separate; (1) entire; (2)
notched or incised.

6. Dorsolateral incisions on collar margin: (0) absent; (1)
present.

7. Thoracic uncini main fang: (0) slender; (1) swollen.

8. Thoracic uncini dentition: (0) large tooth above main
fang; (1) series of subequal teeth above main fang.

9. Abdominal uncini manubrium: (0) about 1.5 times lon-
ger than dentate region; (1) about 2 times longer than
dentate region; (2) about same length as dentate region.

of branchial crown data for Pseudofabriciola Ion -
gipyga, several modifications were also made to
states of characters 8 and 9 in the outgroup and P.
longipyga (Table 4). Fitzhugh et al. considered the
outgroup condition for character 8 (thoracic uncini
dentition) to be state 0 (main fang surmounted by
a large tooth and series of smaller teeth) and char-
acter 9 (manubrium length of abdominal uncini) to
be state 0 (manubrium about 1.5 times longer than
dentate region). Fitzhugh et al. (1994: fig. 1) based
their coding of outgroup states on the relationship
of Pseudofabriciola to other fabriciin genera as
determined by Fitzhugh (1992; see also Fitzhugh,
1993). Subsequently, Fitzhugh (1995a: fig. 5) re-
ported a much higher degree of ambiguity with
regard to the relationship of Pseudofabriciola to
other fabriciins, leaving open the question of ple-
siomorphic conditions for characters 8 and 9. As a
result, these characters were coded as unknown
(“?”) for the outgroup.

Minimum-length cladograms were produced

Figure 7. Anterior end of Pseudofabriciola longipyga
(LACM-AHF 1712) from Cyclades Island plateau, dorsal
view. Note deformation of collar. Abbreviations: li, dor-
solateral collar incision; mi, middorsal collar incision.

from the data matrix (Table 4) using the ie* com-
mand in Hennig86 (Farris, 1988). Multistate char-
acters (3-5, 9) were treated as nonadditive, and no
differential weighting was applied. Character-state
changes were determined using FlennigSb’s xsteps
command with the h and c options.

Five trees were produced (Fig. 8), each with a
length of 16 steps and consistency (ci) and retention
(ri) indices of 0.81 and 0.76, respectively. Relation-

Table 4. Character-state matrix for Pseudofabriciola species based on character states presented in Table 3.

Characters

Outgroup

P. analis

P. australiensis

P. californica

P. capensis

P. filamentosa

P. incisura

P. longa

P. longipyga

Contributions in Science, Number 457

Fitzhugh and Simboura: Mediterranean Pseudofabriciola ■ 7

californica

capensis

longa

filamentosa

analis

longipyga

australiensis

incisura

1(0,1)

Figure 8. Minimum-length cladograms of Pseudo fabriciola species derived from the data matrix in Table 4. Length
of each tree is 16 steps, ci = 0.81, ri = 0.76. Slashes on stems indicate character-state changes, with characters and
states (in parentheses) indicated (cf. Tables 3 and 4). Homoplasious states are denoted by an asterisk. Ambiguous
character-state changes are indicated at the nodes; trees A^, B^, and Q.3 show allowable transformation series for
these characters. All character-state changes are shown in A; state changes in B-E are the same as in A except where
indicated.

8 ■ Contributions in Science, Number 457

Fitzhugh and Simboura: Mediterranean Pseudofabriciola

ships among four species, Pseudofabriciola inci-
sura Fitzhugh, 1990, P. australiensis (Hartmann-
Schroder, 1981), P. longipyga, and P. analis, were
constant among all trees and form an apomorphic
clade relative to other species; this clade is defined
by the peristomial ring collar being distally flared
[character state 3(2); cf. Fig. 8 A] as opposed to being
of even width. Pseudofabriciola longa Fitzhugh,
1990, and P. filamentosa (Day) are sister taxa to
this clade and display three possible patterns of
relationship (Fig. 8A-B, D), depending upon the
transformation series of character 6 (dorsolateral
collar incisions). Pseudofabriciola californica Fitz-
hugh, 1991a, and P. capensis (Monro, 1937) are
the most plesiomorphic species in the genus, with
three trees (Fig. 8A-B, D) showing P. californica
plesiomorphic to P. capensis, and other topologies
with these species either as exclusive sister taxa (Fig.
8C) or forming a trichotomy with a clade com-
prising all other species (Fig. 8E). These topologies
are mainly due to variations in transformation series
in characters 4 (middorsal collar condition), 6, and
8 (thoracic uncini dentition). The greatest ambiguity
in character-state transformation series within and
among trees was in characters 1 (dorsal lip devel-
opment), 6, and 9 (abdominal uncini manubrium
length). These were the only characters that showed
a ci and ri of less than 1.00; the ri of character 9
was 0.00.

DISCUSSION

Only two of the topologies (Fig. 8A-B) produced
with the current data set are the same as those found
by Fitzhugh et al. (1994: fig. 8A-B). Remaining trees
from this latter study differ from the present results
in that three of the eight trees contained the clade
( Pseudofabriciola longa, P. filamentosa, P. analis ),
and three trees had either the clade (P. longa, P.
filamentosa) or (P. longa, P. analis ). While a (P.
longa, P. filamentosa) clade is consistent, though
undefinable, with the topology in Figure 8D, an
additional step (17 total) would be required to allow
for the remaining clades.

The present study produced two topologies not
reported by Fitzhugh et al. (1994), wherein Pseu-
dofabriciola californica and P. capensis form ei-
ther an exclusive sister group (Fig. 8C) or a tri-
chotomy with remaining species (Fig. 8E). Depend-
ing on the transformation series, the (P. californica,
P. capensis) clade (Fig. 8C) can be defined by char-
acter 4 (middorsal collar condition; Fig. 8C1} C3) or
8 (thoracic uncini dentition; Fig. 8C2). Since, how-
ever, the outgroup condition for character 8 was
coded as unknown, either state 0 or 1 can be ple-
siomorphic for the genus. It is only when the ple-
siomorphic condition for the genus is state 1 (series
of small teeth above the main fang) that the (P.
californica, P. capensis) clade is defined by state 0
(presence of a large tooth above the main fang; Fig.
8C2).

There are now only two species, Pseudofabri-

australiensis :
filamentosa:

tree length:
i ci:

ri:

0.81

0.76

0.75

0.81

0.78

0.81

0.78

A W"

B W*

c W"

D W

e yir

Figure 9. Possible tree topologies that result from each
hypothetical combination of states for dorsal lip construc-
tion (character 1) in Pseudofabriciola australiensis and
P. filamentosa, as included in the data matrix in Table 4.
Topologies are labelled to correspond to those in Fig-
ure 8.

ciola australiensis and P. filamentosa, for which
information on the branchial crown and dorsal lips
(character 1) is lacking. Fitzhugh et al. (1994: fig.
10) presented the results of a series of cladistic
analyses in which all possible combinations of states
for character 1 might be encountered among P.
australiensis, P. filamentosa, and P. longipyga. In
those instances in which P. longipyga was coded
with state 1 (dorsal lips as low ridges), topologies
matched those presented here in Figure 8A-B and
D. This sort of comparison is limited, however,
given that the outgroup states for characters 8 and
9 have been modified for the present analysis. As
an update to the simulations performed by Fitzhugh
et al., we reanalyzed the possible resultant topol-
ogies that might occur once dorsal lip information
is obtained for P. australiensis and P. filamentosa.
The four data matrices derived from Table 4, which
account for each of the possible combinations of
states for character 1 that might occur in these two
species (cf. Fig. 9), were analyzed using the Hennig86
commands described earlier. Two data sets pro-
duced single trees, each with a topology as in Figure
8A, and the other two data sets produced four and
five trees, with topologies corresponding to Figure
8B-E and 8A-E, respectively.

Of the four possible combinations of dorsal lip
states among Pseudofabriciola australiensis and P.
filamentosa, we predict that P. australiensis has
state 0 (triangular lobe) and P. filamentosa has state
1 (low, narrow ridge). Under this prediction, the
current data set still yields the five topologies in
Figure 8A (cf. Fig. 9). Pseudofabriciola australien-
sis closely resembles P. incisura, the latter having

Contributions in Science, Number 457

Fitzhugh and Simboura: Mediterranean Pseudofabriciola ■ 9

triangular dorsal lips and ventral filamentous ap-
pendages (Table 1; Fitzhugh, 1990). All topologies
in Figure 8 place these two species as exclusive sister
taxa based on the main fang of thoracic uncini
(character 7). An additional species from Belize,
currently being described (Fitzhugh, in prep.), close-
ly resembles P. australiensis and P. filamentosa in
general collar construction and also has triangular
dorsal lips and ventral filamentous appendages. Our
prediction as to dorsal lip construction in P. fila-
mentosa follows from the general observations by
one of us (K.F.) that Pseudofabriciola species with
very slender branchial crowns, such as P. longipyga,
P. analis, and P. longa (Table 1), tend to have
reduced dorsal lips. This pattern is also exhibited
by another undescribed species from New Guinea
(Fitzhugh, 1995b).

ACKNOWLEDGMENTS

We thank the National Centre for Marine Research,
Greece, for making available to us the specimens used in
this study.

LITERATURE CITED

Day, J.H. 1963. The polychaete fauna of South Africa.
Part 8: New species and records from grab samples
and dredgings. Zoological Bulletin of the British Mu-
seum (Natural History) 10:384-445.

Farris, J.S. 1988. Hennig86 reference, version 1.5. Port
Jefferson Station, New York: published by the au-
thor, 18 pp.

Fitzhugh, K. 1990. Two new genera of the subfamily
Fabriciinae (Polychaeta: Sabellidae). American Mu-
seum Novitates 2967:1-19.

. 1991a. Systematics of several fabriciin fan worms

(Polychaeta: Sabellidae: Fabriciinae) previously re-
ferred to Fabricia or Fabriciola. Journal of Natural
History 25:1101-1120.

- — — . 1991b. Further revisions of the Sabellidae sub-
families and cladistic relationships among the Fa-
briciinae (Annelida: Polychaeta). Zoological Journal
of the Linnean Society 102:305-332.

— — — . 1992. On the systematic position of Monroika

africana (Monro) (Polychaeta: Sabellidae: Fabrici-
inae) and a description of a new fabriciin genus and
species from Australia. Proceedings of the Biological
Society of Washington 105:116-131.

— . 1993. Novafabricia brunnea (Hartman, 1969),

new combination, with an update on relationships
among Fabriciinae taxa (Polychaeta: Sabellidae).
Contributions in Science 438:1-12.

— . 1995a. Additions to the description of the fan-

worm genus Pseudofabricia Cantone, 1972 (Poly-
chaeta: Sabellidae: Fabriciinae). Contributions in
Science 456:1-6.

— . 1995b. New fanworm species (Polychaeta: Sa-
bellidae: Fabriciinae) in the genus Pseudofabriciola
Fitzhugh. Journal of Natural History (in press).

Fitzhugh, K., A. Giangrande, and N. Simboura. 1994.
New species of Pseudofabriciola Fitzhugh, 1990
(Polychaeta: Sabellidae: Fabriciinae), from the Med-
iterranean Sea. Zoological Journal of the Linnean
Society 110:219-241.

Giangrande, A. 1989. Censimento dei Policheti dei mari
Italiani: Sabellidae Malmgren, 1867. Atti della So -
cieta Toscana di Scienze Naturali Memorie, Serie
B 96: 153-189.

Giangrande, A., and A. Castelli. 1986. Occurrence of
Fabricia filamentosa Day, 1963 (Polychaeta, Sabel-
lidae, Fabriciinae) in the Mediterranean Sea. Oeba-
lia, n.s. 13:119-122.

Hartmann-Schroder, G. 1981. Teil 6. Die Polychaeten
der tropisch-subtropischen Westkiiste Australiens
(zwischen Exmouth im Norden und Cervantes im
Siiden). In Zur Kenntnis des Eulitorals der austral-
ischen Kusten unter besonderer Berucksichtigung der
Polychaeten und Ostracoden, ed. G. Hartmann-
Schroder and G. Hartmann. Mitteilungen aus dem
Hamburgischen zoologischen Museum und Institut
83:31-70.

Monro, C.C.A. 1937. Note on a collection of Poly-
chaeta from South Africa, with the description of a
new species belonging to the family Sabellidae. An-
nals and Magazine of Natural History, Series 10
19:366-370.

Simboura, N. 1990. Fabricia filamentosa Day, 1963
(Polychaeta, Sabellidae, Fabriciinae) a lessepsian mi-
grant in Mediterranean Sea. Oebalia, n.s. 16:139-
133.

Received 18 November 1994; accepted 7 April 1995.

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

Number 458
8 November 1995

Contributions
in Science

New Species of Brittle Stars from the
Western Atlantic, Ophionereis vittata,
Amphioplus sepultus, and Ophiostigma SIVA,
and the Designation of a Neotype for
Ophiostigma isocanthum (Say)
(Echinodermata: Ophiuroidea)

Gordon Hendler

Natural History Museum of Los Angeles County

•*V

Serial

Publications

of THE
Natural History
Museum of
Los Angeles
County

• Contributions in Science, a miscellaneous series of tech-
nical papers describing original research in the life and
earth sciences.

• Science Series, long articles and collections of papers
on natural history topics.

Copies of the publications in these series are sold through
the Museum Book Shop. A catalog is available on request.

Scientific

Publications

Committee

James L. Powell, Museum President
Daniel M. Cohen,
Committee Chairman
Brian V. Brown
Kenneth E. Campbell
Kirk Fitzhugh
Robin A. Simpson, Managing Editor

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

Printed at Allen Press, Inc., Lawrence, Kansas
ISSN 0459-8113

New Species of Brittle Stars from the

Western Atlantic, Ophionereis vittata,
Amphioplus sepultus, and Ophiostigma SIVA,
and the Designation of a Neotype for
Ophiostigma isocanthum (Say)
(Echinodermata: Ophiuroidea)

Gordon Hendler1

ABSTRACT. Ophionereis vittata, new species, heretofore misidentified as O. reticulata (Say) or O.
olivacea H.L. Clark, occurs from Florida and the Caribbean at depths from 23 to 126 m. Ophiostigma
siva, new species, which is fissiparous and usually six-armed, was previously considered to be conspecific
with five-armed Ophiostigma isocanthum (Say). The latter species is redescribed and a neotype designated.
Ophiostigma siva and O. isocanthum are sympatric, ranging from Bermuda to the Caribbean, from the
intertidal to depths over 40 m. Amphioplus sepultus, new species, has been reported to be the most
common intertidal amphiurid in southern Florida and is known only from Floridian waters. Amphioplus
abditus (Verrill), into which A. sepultus was formerly placed, is distributed from Maine to Georgia.

INTRODUCTION

This is an account of three species of brittle stars
that are widespread in warm waters of the western
Atlantic and consistently have been mistaken for
sympatric congeners. Of one, Ophionereis vittata,
few individuals have been found, possibly because
it belongs to the relatively inaccessible “deep-reef
fauna” (Hendler and Miller, 1984; Hendler and
Peck, 1988). In contrast, Amphioplus sepultus is
“the most common intertidal amphiurid of South
Florida” (Thomas, 1962:654, as Amphioplus ab-
ditus (Verrill)). Similarly, Ophiostigma siva can oc-
cur in profusion, with over 100 individuals per liter
of algal substratum (Hendler and Littman, 1986, as
Ophiostigma isocanthum ).

In this contribution the five-armed species
Ophiostigma isocanthum (Say), sister species of the
fissiparous, six-armed O. siva, is redescribed. A
neotype is designated because Say’s (1825) original
specimen is lost. The present contribution includes
diagnoses, full descriptions, and lists of designated
type material for the three new taxa. However,
abridged descriptions and color illustrations of these
species are provided in Hendler et al. (1995).

The following abbreviations for institutions and
programs are used in this paper: BLM (United States

1. Invertebrate Zoology Section, Natural History Mu-
seum of Los Angeles County, 900 Exposition Boulevard,
Los Angeles, California 90007.

Contributions in Science, Number 458, pp. 1-19
Natural History Museum of Los Angeles County, 1995

Bureau of Land Management), IRCZM (Indian Riv-
er Coastal Zone Museum, Harbor Branch Ocean-
ographic Institution), LACM (Natural History Mu-
seum of Los Angeles County), LMRS (South At-
lantic Outer Continental Shelf Area Living Marine
Resources Study), MCZ (Museum of Comparative
Zoology, Harvard University), MMS (United States
Minerals Management Service), SOFLA (Southwest
Florida Shelf Ecosystem Study), UMML (Rosenstiel
School of Marine and Atmospheric Sciences, Uni-
versity of Miami), USNM (National Museum of
Natural History, Smithsonian Institution), and UZM
(University Zoological Museum, Copenhagen).

SYSTEMATIC ACCOUNT
Family Ophionereididae
Genus Ophionereis Liitken, 1859

Ophionereis vittata, new species

Figure 1A-C

Ophionereis reticulata: Lyman, 1878:224 (not
Ophiura [= Ophionereis ] reticulata Say, 1825),
author notes “Station 11, 37 fathoms, 1 speci-
men,” corresponding to MCZ 1603, which is
Ophionereis vittata; 1883:253 (not Ophiura [=
Ophionereis] reticulata Say, 1825, in part), au-
thor notes “Station 278, Barbados, 69 fathoms,”
corresponding to USNM 6430, which is O. vit-
tata.

Figure 1. Ophionereis vittata, new species, holotype LACM 83-134.2: A, entire, dorsal view; B, disk, dorsal view; C,
disk, ventral view. Disk diameter — 7.0 mm.

Ophionereis olivacea: Thomas, 1973:590-593, fig.
3 (not Ophionereis olivacea H.L. Clark, 1901, in
part).

Ophionereis sp.: Hendler and Peck, 1988:413.

This species, first collected by the R/V Blake
during the 19th century, was misidentified as Ophi-
onereis reticulata (Say) by Lyman (see synonymy
above). Thomas (1973:592) mistakenly regarded it

2 ■ Contributions in Science, Number 458

Hendler: New Western Atlantic Ophiuroids

as O. olivacea H.L. Clark and mentioned that Clark’s
description of O. olivacea “. . . is accompanied by
a singularly poor illustration.” Indeed, most of
Clark’s description could apply to either O. oli-
vacea or O. vittata. However, Clark specified that
O. olivacea is marked by “color above olive green,
spotted on the disk with yellow; arms banded with
a darker shade of green . . . outside of oral shield
is a patch of dark brown, as in reticulata.” This
corresponds to the situation in the specimen of O.
olivacea (MCZ 4250) collected, identified, and dis-
cussed by Clark (1918). However it does not apply
to O. vittata, which lacks the dark brown patch
that is characteristic of O. olivacea. Furthermore,
O. vittata invariably possesses “golden reticula-
tion” (Thomas, 1973:593) on an otherwise grayish
disk. Moreover, Clark (1901:248) mentioned the
presence of more than one pair of accessory dorsal
arm plates on the “first few joints” of O. olivacea,
whereas O. vittata, as illustrated by Thomas (1973:
592), has “. . . accessory plates often appearing to
be composed of overlapping scales (Fig. 3A)” on
much of its arm.

ETYMOLOGY. Vittata, feminine form of the
Latin adjective meaning “decorated with ribbon,”
in reference to the colorful stripe, bands, and net-
work pattern adorning living individuals.

MATERIAL EXAMINED. Unless otherwise stated, all
specimens from Belize were collected using scuba gear
and the ichthyocide “Noxfish,” ESE of Carrie Bow Cay
on the seaward slope of the Belize Barrier Reef, 16°48.14'N,
88°04.50'W, by G. Hendler with the assistance of divers
noted in the Acknowledgments. Designated types origi-
nally cataloged as O. vittata unless otherwise noted.

Holotype. BELIZE: (LACM 83-134.2), holotype, dry,
Sta. Belize 83 No. 10, 7 Nov. 1983, 80 ft.

Paratypes. FLORIDA, ATLANTIC: (LACM 85-275.2),
1 ale, Sta. LK 55a, 17 Aug. 1985, 24°32.0'N, 81°24.3'W,
Looe Key National Marine Sanctuary, 24-26 m, scuba,
coll. G. Hendler et al. GULF OF MEXICO, OFF FLOR-
IDA: (USNM El 21 82), originally as O. olivacea, 3 ale,
VF Middle Grounds Cr., Sta. 6, 19 May 1973, 28°27'N,
84°19'W, coll. F.J.S. Maturo, 120 ft; (USNM E32313),
originally as O. olivacea, 1 ale, MMS/BLM SOFLA Sta.
23 II-23-c-TDS, 1 Aug. 1981, 25°16'53"N, 83°37'47"W,
off Florida, Gulf of Mexico, 70 m; (USNM E32314), orig-
inally as O. olivacea, 4 ale, MMS/BLM SOFLA Sta. 23
II-23-a-TDS, 1 Aug. 1981, 25°16'53"N, 83°37'47"W, off
Florida, Gulf of Mexico, 70 m. BELIZE: (LACM 83-
134.3), 2 ale, Sta. Belize 83 No. 10, 7 Nov. 1983, 80 ft,
coll. G. Hendler; (LACM 83-131.3), 1 ale, Sta. CBC 83-
9, 6 Nov. 1983, 80 ft; (LACM 85-461.1), 1 ale, Sta. CBC
85-6, 15 Jun. 1985, 75 ft, coll. G. Hendler.

Other Material Examined. FLORIDA, ATLANTIC:
UMML cat. no. 41.172, 1 ale. GULF OF MEXICO, OFF
FLORIDA: MCZ cat. no. 1603, 1 dry; USNM cat. nos.
E31841, 1 ale; E31842, 1 ale; E31843, 1 ale; E32312, 1
ale; E32316, 1 ale; E32317, 2 ale; E32318, 1 dry; E32319,
1 ale. BARBADOS: USNM cat. no. 6430, 1 dry. THE
GRENADINES: BEQUIA: LACM cat. no. 69-149.1, 1
ale. U. S. VIRGIN ISLANDS: UMML cat. no. 41.236, 1
ale. BELIZE: LACM cat. no. 83-132.4, 1 ale. PANAMA:
LACM cat. no. 66-318.1, 1 dry; 66-320.1, 2 ale; USNM
cat. nos. E45510, 1 ale; E45509, 1 dry.

Contributions in Science, Number 458

DIAGNOSIS. Disk of living individuals adorned
with yellow reticulation, arms dorsally with medial
red stripe and greenish brown bands. Disk scales
small, delicate; primary plates not discernable; gen-
ital papillae absent. Oral shield subcordate, nearly
equal in length and width. Arms long, slender
throughout; joints bear 3 compressed arm spines;
middle spine longest, equal in length to the width
of a joint, tip slightly expanded, surface micro-
scopically roughened. One or more supplementary
scales distal to most accessory dorsal arm plates.
Superficial ossicles thin and delicate; arm spines and
tentacle scales translucent in alcoholic specimens.

DESCRIPTION OF HOLOTYPE. Disk diame-
ter 7.0 mm; length of longest arm 60 mm. Disk
rounded pentagonal, somewhat inflated, delicate.
Arms long, very slender, narrow at edge of disk,
increasing in width to approximately 12th to 15th
arm joint, outer Vi of arm gradually tapering to
filiform tip. Arms gently rounded dorsally, flattened
ventrally; dorsal and ventral arm plates of successive
joints in contact, lateral arm plates connected by
soft tissue. Superficial ossicles delicate, thin; in al-
cohol, arm spines and tentacle scales appear trans-
lucent.

Disk covered with minute, thin, imbricating scales;
scales smallest at center of disk, largest near radial
shields at periphery of disk; primary plates not dis-
cernable; radial shields small, triangular, approxi-
mately twice as long as wide.

Each jaw with 1 apical oral papilla, 3 pairs of
oral papillae laterally; proximal lateral papillae
bluntly pointed, distalmost larger, broader, flat-
tened; tentacle scale arising from first ventral arm
plate. Teeth with squared tips, stereom of proximal
edge imperforate.

Oral shield subcordate, length and width nearly
equal, proximal edge bluntly pointed, distal edge
broadly rounded. Madreporite slightly longer than
wide, bearing 5 pores. Adoral shield covered prox-
imally by oral shield; radial edges rounded, abutting
first ventral arm plate in mouth angle.

Bursal slits long, extending to periphery of disk,
bordered by small, thin scales; genital papillae ab-
sent. Ventral interbrachial field covered with layer
of minute, thin scales.

Dorsal arm plates rounded triangular, longer than
wide at base of arm; rounded hexagonal through-
out wide central region of arm, only slightly wider
than long; markedly longer than wide on distal
region of arm. Accessory dorsal arm plates abutting
distal half of dorsal arm plate, less than half the
plate in length; present from base to tip of arm.
Supplementary scales originating at distal edge of
accessory dorsal arm plate, one or more scales on
each joint; present on proximal % of arm.

Lateral arm plates slightly flared, bearing 3 erect
arm spines. Dorsal and ventral spines laterally com-
pressed, markedly tapering to blunt tip. Dorsal spine
shorter and narrower than ventral spine, broader
on joints near disk edge than on distal joints. Mid-
dle spine longest, dorsoventrally compressed; shaft

Hendler: New Western Atlantic Ophiuroids ■ 3

tapering toward center, tip expanded; edges of spines
thin, microscopically rough. Length of middle spine
equal to width of arm joint in broad central region
of arm. Spines near arm tip slender, acutely pointed.

First ventral arm plate small, subtriangular; sec-
ond plate wider than long; third plate equidimen-
sional; succeeding plates longer than wide. Plates
shield-shaped, with narrow proximal edge convexly
rounded, distal edge broad, convex, with medial
notch; lateral edges deeply concave alongside ten-
tacle pore.

One tentacle scale covering each pore, large,
ovoidal; proximal end narrow, distal end broad.

Coloration. In holotype and other individuals
when alive, ground color of disk gray to white,
covered with an open network of thin, brilliant
yellow to brownish yellow lines extending onto the
ventral interradii; arms whitish with a thin medial
stripe of red originating between the radial shields,
extending to arm tip, every 3-8 joints with a black
and dark greenish brown band extending to ventral
surface; arm spines whitish with basal band of yel-
low, orange-brown, or brown. Dorsal arm plates
may be tinged with red, ventral arm plates with
yellow or brown, lateral arm plates with yellow,
red, or brown; oral shields, adoral shields, and jaws
may be blotched with yellow, orange, or brown.
Protracted tube feet, equal in length to the arm
joint, are transparent or whitish translucent. The
color of the ripe gonads is visible through the ven-
tral body wall; testes are white, ovaries are pinkish
red.

In alcoholic or dried specimens, pigmentation
lost except for the black and green bands and for
short sections of the arm stripe that become green
or brown; in some instances, a dusky or brown
portion of the netted pattern remains on the disk.

VARIATIONS. Disk diameters of available ma-
terial range from 2.9 to 9.5 mm; arm length ranges
from 23 to 89 mm; the ratio of arm length/disk
diameter ranges from approximately 7 to 13. The
smallest specimen (LACM 85-275.2) differs from
others examined in having dorsal arm plates longer
than wide and having middle arm spines of a length
less than the width of the arm segment. Differences
between some of the paratypes and the holotype
include adoral shields with proximal tips meeting
above the oral shield, teeth with proximal tip point-
ed or irregular in shape, middle arm spine with the
shaft very slightly to markedly narrowed before the
tip, and madreporite with more (or less) than 5
pores.

COMPARISONS. The association of supple-
mentary scales with the accessory dorsal arm plates
places O. vittata among the Ophionereis species
referred to Ophiocrasis H.L. Clark, 1911. The
group, according to A.M. Clark (1953), includes
Ophionereis dubia (Muller and Troschel), with
which she unites Ophiocrasis dictydisca H.L. Clark
and Ophiocrasis marktanneri Matsumoto and pos-
sibly Ophionereis squamulosa Koehler. Supple-
mentary scales are also present in Ophionereis oli-

vacea H.L. Clark and Ophionereis perplexa Zie-
senhenne (Hendler, pers. obs.).

Ophionereis vittata is distinguished from con-
geners by its unique pattern of arm stripe and bands
and its absence of a dark patch distal to the oral
shield. In contrast to O. squamulosa and O. oli-
vacea, it lacks oral papillae. Unlike O. dolabrifor -
mis and O. perplexa, its arm spines do not exceed
the length of an arm joint, and they are delicate
and translucent instead of dense. Its arms are rel-
atively longer than those of O. dubia, which attain
a length only 7-8 times the diameter of the disk,
and the dorsal arm plates of the latter species are
subtriangular rather than hexagonal as in O. vittata.

DISTRIBUTION. The range of O. vittata en-
compasses the perimeter of the Caribbean Region
and extends into the Gulf of Mexico. However,
sizeable gaps separate the collecting sites: the Great-
er and Lesser Antilles, Panama, Belize, and Florida.
In addition to the localities noted above, the species
has been photographed at Cayman Brae (by E. Fish,
Divers Alert Network 1991 Calendar). Although
one specimen from St. John was reportedly col-
lected from shallow water (Thomas, 1973), others
originated at depths between 23 and 126 m.

BIOLOGY. This species occupies hard-bottom
habitats near the edge of the continental shelf. This
zone has been referred to as the “mixed region” of
the Caribbean (Lyman, 1869:309) and its inhabi-
tants characterized as the “deep-reef fauna” (Hen-
dler and Peck, 1988). In recent years, newly de-
scribed ophiuroids have been reported from this
habitat, which starts near the lower limits of con-
ventional scuba activity and extends to depths sam-
pled with submersibles (Hendler and Miller, 1984;
Hendler and Turner, 1987; Hendler, 1988; Hendler
and Peck, 1988).

Near the upper limit of its bathymetric range, on
the Belize Barrier Reef, O. vittata was collected on
the forereef slope near massive shelves of Montas-
trea annularis (Ellis and Solander), large flat agar-
iciids, and other stony corals, in areas with a sig-
nificant cover of Halimeda spp., sponges, and gor-
gonians. Individuals were cryptic and encountered
when ichthyocide was applied to drive invertebrates
from the reef. On the intermediate-deep reef off
Looe Key, Florida, an individual was found within
a clump of Halimeda algae. Individuals from deep-
er water off Panama and Colombia were dredged
from hard bottoms characterized by coral plates,
coralline lumps, and shell fragments.

Several Ophionereis vittata were collected at the
same stations as O. reticulata and O. olivacea in
Panama and Belize; however, O. reticulata is typ-
ically found in shallow-water reef habitats and O.
olivacea in mangrove algae, seagrass beds, and less
commonly on the forereef slope (Hendler et al.,
1995).

Gonads are visible through the thin ventral body
wall of several specimens; there are 6-12 gonads
per interradius, and the sexes are separate. Two
females had relatively large gonads, and 10 of the

4 ■ Contributions in Science, Number 458

Hendler: New Western Atlantic Ophiuroids

largest oocytes were measured in each individual.
The mean oocyte diameter was 0.20 mm, but these
cells may not have been fully ripe. This is consid-
erably smaller than the 0.4-mm oocytes of Ophi-
onereis olivacea H.L. Clark, which broods its young
(Hendler and Littman, 1986; Byrne, 1991). None
of the specimens of O. vittata brooded embryos
in their bursae. The oocyte diameter of O. vittata
is slightly smaller than the 0.24-mm oocytes of O.
annulata (Le Conte) and similar to the 0.20-mm
oocytes of O. squamulosa Koehler (Hendler, 1982,
1991). All three species have oocyte sizes typical
for ophiuroids with abbreviated development, and
O. annulata and O. squamulosa have lecithotroph-
ic larvae (Hendler, 1982, 1991), suggesting that the
new species might have a similar mode of devel-
opment.

Family Amphiuridae
Genus Ophiostigma Liitken, 1856

Ophiostigma isocanthum (Say, 1825)

Figures 2A-C, 4A

Ophiura isocantha Say, 1825:150-151.
Ophiocoma isocantha: Muller and Troschel, 1842:
103.

Ophiostigma moniliforme: Liitken, 1856:13; 1859:
181, 186, 234 (in part).

Ophiostigma isocantha : Lyman, 1860:258.
Ophiostigma isocanthum: Lyman, 1865:12, 103-
104, 199, figs. 8, 9 (in part). Ljungman, 1866:317.
H.L. Clark, 1901:344. Hendler and Littman, 1986:
33-38 (in part). Hendler and Peck, 1988:413 (in
part).

Ophiostigma isacantha: Ljungman, 1871:636, 657.
Heilprin, 1888:316. H.L. Clark, 1898:412; 1899:
131.

Ophiostigma isacanthum: Lyman, 1875:5; 1878:
224; 1880:26; Lyman, 1882:165, 311, 314, 324,
380, pi. 42, fig. 16; 1883:229, 254. Rathbun, 1879:
155. Verrill, 1899:377; 1907:325; H.L. Clark,
1901:240, 249, 262 (in part); 1915:244; 1919:56,
58; 1933:36, 50-51 (in part); 1942:377 (in part).
A.H. Clark, 1922:210. Koehler, 1907:298; 1913:
352, 363-367, pi. 20, figs. 6, 7 (in part); 1914:2,
38, 154, 171 (in part). A.H. Clark, 1921:42; 1922:
210; 1939:446; 1954:377. Nielsen, 1932:308-309.
Pearson, 1937:71. Engel, 1939:4, 8. Fontaine,
1953:199, 201. A.M. Clark, 1955:38. Tabb and
Manning, 1961:566. McNulty et al., 1962a:229.
Thomas, 1962:689-692, fig. 23a, b (in part). Par-
slow and A.M. Clark, 1963:37, 44. Lewis, 1965:
1074. O’Gower and Wacasey, 1967:210. Tom-
masi, 1970:40-41. Singletary, 1971:940. Carrera,
1974:iii, 69-71, pi. 5, fig. 2a, b (in part). Alvarez
Larrauri, 1981:33. Abreu Perez, 1983:2. Emson
et al., 1985:87-100 (in part). Aronson and Harms,
1985:1483.

This was among the first brittle star species to be
described by an American scientist. The initial ac-

Contributions in Science, Number 458

count (Say, 1825) was based on a single specimen
collected from the Florida Keys. The types of sev-
eral species published coincidentally with Ophiura
(' Ophiostigma ) isocantha are deposited at the
Academy of Natural Sciences of Philadelphia (Spa-
mer and Bogan, 1992), but the Academy has no
record of a holotype of O. isocantha (E.E. Spamer,
pers. comm.). Neither is there a record of the orig-
inal specimen at The British Museum (Natural His-
tory) (G.L.J. Paterson, pers. comm.) to which Say
sent some material before his departure, around
1828, for the utopian colony of New Harmony,
Indiana (Spamer, pers. comm.). Other potential re-
positories, the USNM and MCZ, do not hold the
type (Downey, 1969).

Say’s description is sufficient to distinguish O.
isocanthum from any other shallow-water West
Indian brittle star. Since the 19th century, there has
been a clear concept of Say’s species, and specimens
have been illustrated several times (see synonymy).
However, as discussed below for Ophiostigma siva,
authors occasionally have remarked on six-armed
individuals in collections of five-armed O. isocan-
thum, either considering them variants or fissipa-
rously reproducing individuals. Formal recognition
of the fissiparous individuals as a distinct taxon in
the present publication necessitates the redescrip-
tion of O. isocanthum and the designation of a
neotype. In accordance with the International Code
of Zoological Nomenclature (ICZN) Article 75, the
specimen selected as the neotype originated from
the Florida Keys, the original collecting locality.

ETYMOLOGY. The original spelling isocantha
should have been written isacantha, but as it is not
“demonstrably incorrect” with regard to ICZN Ar-
ticle 32, it must be retained.

MATERIAL EXAMINED. Neotype. FLORIDA, AT-
LANTIC: (LACM 85-265.3), ale, LK 45, 15 May 1985,
24°33.7'N, 81°25.7'W, Looe Key National Marine Sanc-
tuary, Florida Keys, 11m, coll. G. Hendler et al.

Other Material Examined. NORTH CAROLINA:
USNM cat. nos. E27748, 4 ale; E27749, 5 ale; E27750, 3
ale; E27751, 10 ale; E27752, 2 ale; E28160, 1 ale; E28161,
1 ale; E28162, 2 ale; E28163, 4 ale; E28169, 5 ale; E28170,

6 ale; E28172, 1 ale; E29095, 5 ale; E29096, 2 ale; E29098,
5 ale; E29099, 9 ale; E29100, 1 ale; E29102, 1 ale; E29106,
5 ale; E29107, 3 ale; E29109, 1 ale; E29110, 11 ale; E29111,
1 ale; E29112, 2 ale; E29115, 1 ale; E29116, 1 ale; E29119,
1 ale; E29121, 1 ale; E29122, 29 ale; E29123, 2 ale; E29125,

7 ale; E29126, 9 ale; E29127, 4 ale; E29128, 4 ale; E29130,
3 ale; E29131, 2 ale; E29132, 9 ale; E29133, 1 ale; E29134,

1 ale; E29135, 8 ale; E29136, 29 ale; E29137, 3 ale; E29138,
10 ale; E29139, 4 ale; E29140, 6 ale; E29141, 54 ale;
E29142, 3 ale; E29143, 8 ale; E29144, 2 ale; E29146, 2
ale; E29147, 1 ale; E29149, 2 ale; E29150, 12 ale; E29155,

3 ale; E29274, 1 dry; E29275, 3 dry; E29276, 1 dry; E29277,

2 dry; E2941 7, 7 ale; E30494, 1 ale; E30508, 1 ale; E30524,
1 ale; E32211, 1 ale; E30496, 27 ale; E30525, 1 ale; E33143,

1 ale. GEORGIA: USNM cat. nos. E28165, 1 ale; E28167,

2 ale; E29097, 7 ale; E29101, 9 ale, E29103, 6 ale; E29108,

4 ale; E291 13, 1 ale; E291 14, 8 ale; E291 1 8, 2 ale; E29120,
1 ale; E29124, 3 ale; E29129, 1 ale; E29151, 1 ale; E29152,
1 ale; E29156, 4 ale; E29157, 1 ale; E29158, 1 ale; E29387,
1 ale; E30242, 19 ale; E30244, 3 ale; E30249, 9 ale. FLOR-

Hendler: New Western Atlantic Ophiuroids ■ 5

Figure 2. Ophiostigma isocanthum (Say), neotype LACM 85-265.3: A, entire, dorsal view; B, disk, dorsal view; C,
disk, ventral view. Disk diameter = 6.0 mm.

6 ■ Contributions in Science, Number 458

Hendler: New Western Atlantic Ophiuroids

IDA, ATLANTIC: LACM cat. nos. 70-291.5, 1 dry; 84-

230.5, 1 ale; 85-240.5, 1 ale; 85-242.6, 1 ale; 85-262.4, 2
ale; 85-275.4, 2 ale; 85-268.4, 1 ale; 85-277.4, 1 ale; 88-

194.5, 1 ale; 88-197.8, 23 ale; USNM cat. nos. E19742,
1 ale; E20290, 1 ale; E20291, 1 ale; E24331, 2 dry; E28158,
3 ale; E28159, 1 ale; E28164, 1 ale; E28166, 1 ale; E28168,

1 ale; E28171, 1 ale; E29148, 2 ale; E37731, 1 ale; IRCZM
cat. nos. 74:257, 1 dry; 74:258, 8 ale; 74:354, 2 dry; 74:
356, 7 ale; 74:384, 1 ale; 74:546, 3 ale; 74:549, 3 ale.
SOUTHWEST FLORIDA, GULF OF MEXICO: LACM
cat. nos. 89-237.5, 1 ale; 89-319.2, 2 ale; 89-322.4, 4 ale;
89-323.5, 1 ale; 89-325.1, 1 ale; 89-327.2, 1 ale; 89-329.5,

2 ale; 89-339.3, 1 ale; 89-348.1, 1 ale; 89-350.1, 1 ale; 89-
351.3, 1 ale; 89-352.1, 1 ale; 89-358.2, 1 ale; 89-359.2, 1
ale; 89-365.1, 3 ale; 89-371.1, 1 ale; USNM cat. nos. 6440,
1 ale; 12482, 2 dry; 14199, 3 ale; 14207, 1 dry; 14245, 3
ale; 15430, 2 dry; 15455, 1 dry; 33975, 5 ale; 33976, 1
ale; 33977, 1 ale; E22825, 1 dry; E25232, 1 dry; E28634,
1 ale; E31831, 1 ale; E31832, 1 ale; E31833, 2 ale; E31834,
1 ale; E31835, 1 ale; E31836, 1 ale; E31837, 2 ale; E32327,
1 ale; E32328, 3 ale; E32329, 1 ale; E32330, 1 ale; E32331,
1 ale; E32332, 1 ale; E32333, 1 ale; E32334, 1 ale; E32335,
1 ale; E32336, 1 ale; E32337, 1 ale; E32338, 1 ale; E39139,
1 ale; E39140, 1 ale; E39141, 1 dry; E39142, 3 ale; E39143,
1 dry; E39144, 1 ale; E39145, 1 dry; E39146, 1 ale; E39147,

1 dry; E39148, 3 ale; E39149, 2 dry; E39150, 2 ale; E39151,
5 ale; E39152, 1 dry; E39153, 1 ale; E40920, 1 ale; E40947,

2 ale. BAHAMA ISLANDS: LACM cat. no. 88-211.1, 4
ale; USNM cat. nos. E31328, 1 dry; E31330, 1 dry; E31332,
1 dry; E31340, 1 ale. VIRGIN ISLANDS: USNM cat. nos.
26660, 2 dry; E23014, 1 dry. CUBA: USNM cat. nos.
34733, 1 dry; 34734, 1 ale; 34782, 1 dry; E24183, 1 dry.
PUERTO RICO: USNM cat. no. E5448, 1 dry. MEXICO:
USNM cat. no. E27783, 1 dry. BELIZE: LACM cat. nos.
81-231.2, 1 ale; 82-124.2, 2 ale; 83-214.2, 2 ale; 83-216.1,
1 ale; 83-217.1, 2 ale; 83-219.2, 1 ale; 85-454.1, 1 ale; 85-
455.2, 2 ale; 85-455.4, 11 ale; 85-456.1, 9 ale; 86-55.1, 2
ale; 86-55.2, 1 ale; 86-57.1, 1 ale; 86-63.1, 1 ale; 86-65.1,
1 ale; 86-74.1, 1 ale; 86-77.1, 1 ale; 86-78.1, 1 ale; 86-

87.1, 1 dry; 86-489.1, 1 ale; USNM cat. no. E29792, 1
ale. COSTA RICA: LACM cat. nos. 86-98.1, 1 ale; 86-

143.1, 2 ale; 86-145.1, 3 ale. PANAMA: USNM cat. nos.
E24119, 1 dry; E26396, 1 ale; E26405, 3 ale; E26415, 1
dry; E28383, 4 ale; LACM cat. nos. 39-181.2 AHF, 1 dry;
39-221.2 AHF, 1 dry. COLOMBIA: LACM cat. nos. 39-
186.2 AHF, 19 dry; 39-187.8 AHF, 7 dry; 39-188.3 AHF,
19 dry. VENEZUELA: LACM cat. nos. 39-195.3 AHF,
1 dry; 39-213.3 AHF, 1 dry. NETHERLANDS ANTIL-
LES: LACM cat. no. 39-191.8 AHF, 9 dry; USNM cat.
nos. 15395, 1 dry; E606, 2 ale.

DIAGNOSIS. The species was described by Say
(1825:150-151) as follows: “O. isocantha. Disk
pentagonal, granulated; spines less than half the
length of the transverse diameter of the ray.
“Inhabits the coast of Florida.

“ Disk with the angles obtusely rounded; surface
with numerous elevated tubercles or granulations,
which are not crowded; edge not interrupted by
the rays: rays on the back with a single series of
transversely, angularly oval plates, on each side of
which are two very small spine like scales: spines
less than half the transverse diameter of the ray in
length, unarmed, prominent, equal, placed in three
series: mouth very regular, stellate: colour whitish,
rays annulate with greenish.

“Diameter of the disk less than lA of an inch.
“A single specimen was taken by Mr. Peale.

Contributions in Science, Number 458

“The granulations of the disk resemble those of
O. crassispina, but they are somewhat larger in
proportion.”

The diagnosis is revised as follows. Arms 5 in
number. Disk covered with short tubercles; several
tubercles seated at adradial corner of radial shield
especially prominent. Dorsal disk scales bumpy, ap-
pearing fused together. Three oral papillae, 2 distal
pairs thick, operculate, closing the oral slit. Arm
spines 3, peglike, slightly dorsoventrally flattened,
gradually tapering to blunt tip. Radial shield broad,
bearing tubercles; flat, thick plates joined to and
equal in width to narrow distal margin. Adoral plates
overlap first ventral arm plates, forming continuous
ring. Proximal dorsal arm plates ovoid diamond-
shaped, twice as wide as long, in contact on prox-
imal third of arm. Lateral arm plate stereom with
expanded peripheral trabeculae. Ventral arm plate
length equal to width.

DESCRIPTION OF NEOTYPE. Disk diameter
6.0 mm; length of longest arm 28 mm. Disk round
in outline, dorsal surface flattened, inflated inter-
radially, bearing short, blunt tubercles above and
below. Arm joints below disk narrower than those
beyond edge of disk. Arm gradually tapers from
edge of disk to tip, dorsal surface rounded, ventral
surface relatively flattened, arm tip dorsoventrally
flattened; joints set off by constrictions of lateral
arm plates. Major ossicles of disk and arms thick,
somewhat swollen, opaque.

Body wall pliable, bumpy, scale covered. Scale
edges not discernable; many scales with swollen
central region, bearing fixed tubercles, most 0.06-
0.14 mm tall, half as wide. One or two exception-
ally large, peglike tubercles at adradial edge of each
radial shield, approximately 0.3 mm tall, 0.16 mm
thick. Radial shield broad, length approximately V&
diameter of disk; shields 4-5 times longer than wide,
edges not clearly discernable, bearing scattered,
minute tubercles. Flat, thick plate borne on distal
end of radial shield, equal in width to distal edge
of shield. Primary plates not distinguishable.

Infradental papillae paired, blocklike, in contact;
middle oral papillae slightly larger than infradental,
pyramidal, broad edge facing oral slit; distal oral
papilla largest, spanning middle oral papilla to first
ventral arm plate, quadrilateral, much longer than
wide, long inner edge incised and ridged. Distal 2
pairs of oral papillae of adjacent jaws in contact,
sealing oral gap. Teeth chisel-shaped; proximal edge
concave, stereom imperforate; gap between tips of
opposing ventral teeth wider than gap between dor-
sal teeth.

Oral shields arrowhead-shaped, nearly as wide as
long, with sharply pointed proximal apex, proximal
edges markedly concave, lateral edges short, distal
edge bluntly rounded. Center of shield depressed.
Madreporite of similar shape, slightly larger than
oral shield.

Adoral shield quadrilateral, with part of convex
adradial edge adjoining incurved side of oral shield,
concave radial edge facing mouth. Adoral shields

Hendler: New Western Atlantic Ophiuroids ■ 7

with inner ends meeting proximal to oral shield,
outer ends meeting or nearly meeting distal to first
ventral arm plate, shields thereby forming nearly
continuous ring around the mouth.

Prominent genital scale bridging oral shield and
bursal slit. Bursal slit narrow, spanning approxi-
mately Vs length of interradius; slit inconspicuous,
obscured by arm.

Dorsal arm plates with markedly thickened distal
edge, extending above proximal surface of suc-
ceeding plate. Dorsal arm plates in contact on prox-
imal Vs of arm; distal plates separated by lateral arm
plates. Proximal dorsal arm plates approximately
twice as wide as long, diamond-shaped with round-
ed corners; distal plates becoming half-moon-
shaped, decidedly smaller than lateral arm plates
on distal portion of arm.

Lateral arm plates with stereom of expanded pe-
ripheral trabelculae, appearing as transparent, mi-
croscopic grains. Lateral plates constricted proxi-
mally; distal end broadened, forming thick spine-
bearing ridge.

Three arm spines, erect, with rounded base,
somewhat dorsoventrally flattened, gradually ta-
pering to blunt tip. Dorsal spine longest, shorter
than dorsal arm plate.

First ventral arm plate inserted at distal edge of
jaw slit, minute, diamond-shaped, surrounded by
distal oral papillae and adoral shields. Other ventral
arm plates distal to adoral shields, pentagonal;
proximal edges meeting at rounded apex, lateral
edges concave, distal edge convex. Ventral arm
plates on proximal portion of arm overlapping, plate
width exceeding length; ventral plates on central
portion of arm separated by lateral arm plates, plate
width equals length; ventral arm plates becoming
triangular in shape with rounded distal edge on
flattened tip of arm.

Tentacle scales paired, minute, ovoid, flattened,
only partially covering tentacle pore; scale on lateral
arm plate slightly smaller than more distal scale on
ventral arm plate.

Coloration. In alcohol, residual pigmentation is
pale yellow, some lateral arm plates brown above
and below arm, some dorsal arm plates with elon-
gate, brown, medial ring. Dark pigmentation pro-
duces appearance of irregular bands and discontin-
uous medial stripe. For a typical individual in life,
the disk above is brown, radial shields have a pale
distal tip, the oral shields are white, adoral shields
are gray and tan, jaws are brown and tan; the arms
above are cream and gray with a chainlike dark
brown pattern, ventral arm plates are gray with
white highlights, and the arm spines are white.

VARIATIONS. Specimens examined have disk
diameters from 0.9 to 6.6 mm and arm lengths from
1.7 to 42 mm. Maximum disk diameter/arm length
ratio is 1 /8. Primary plates are evident in individuals
smaller than 4.0 mm disk diameter but are not
discernable in large specimens. The prominent tu-
bercles associated with the radial shields are even
more conspicuous in small specimens than in large

ones. Interestingly, the enlarged tubercles are pres-
ent on disks that are regenerating subsequent to
autotomy. Some features that are distinct in adults
are less pronounced in small specimens, such as the
degree of overlap of dorsal arm plates and the in-
tegrity of the ring of adoral shields.

COMPARISONS. See following species ac-
count.

DISTRIBUTION. Previously reported from Ber-
muda, Florida, the Florida Keys, Dry Tortugas, Gulf
of Mexico, Belize, Cuba, Jamaica, Puerto Rico,
Tortola, St. John, St. Thomas, St. Croix, Anguilla,
St. Barthelemy, St. Christopher, Tobago, Curasao,
Aruba, and Brazil, from less than 1 m to 223 m.
Specimens from the USNM and LACM collections
extend the range to North Carolina, South Caro-
lina, Georgia, the Bahama Islands, Mexico, Costa
Rica, Panama, and Venezuela.

BIOLOGY. The species lives on soft-bottom
habitats, in sediments with seagrass, amidst rubble,
shell, stones, coral, and coralline algae and under
sponges. Individuals are cryptic and can conceal
themselves beneath a thin layer of sediment. When
freshly collected, individuals usually are coated with
fine grains of sediment, which appear to adhere to
mucus on the disk and arms.

Gonads are present in some individuals as small
as 1.0 mm in disk diameter. Several individuals of
moderate size were found to have up to 8 ovaries
per interradius, up to 76 oocytes per ovary, the
oocytes with a mean diameter of 0.15 mm.

Ophiostigma siva, new species

Figures 3A-C, 4B

Ophiostigma moniliforme: Liitken, 1856:13; 1859:
181, 186, 234 (in part).

Ophiostigma isacanthum: Lyman, 1865:12 (here
as Ophiostigma isocanthum ), 103, 199, figs. 8,
9 (not Ophiura [= Ophiostigma] isocantha Say,
1825, in part). H.L. Clark, 1901:240, 249, 262
(not Say, 1825, in part); 1933:36, 50-51 (not Say,
1825, in part); 1942:377 (not Say, 1825, in part).
Koehler, 1913:352, 363-367, pi. 20, figs. 6, 7 (not
Say, 1825, in part); 1914:2, 38, 154, 171 (not Say,
1825, in part). Thomas, 1962:689-692, fig. 23a,
b (not Say, 1825, in part). Parslow and A.M.
Clark, 1963:37 [in synonymy of Ophiocomella
ophiactoides (H.L. Clark) as “(?) Ophiostigma
isacanthum (pt), H. L. Clark, 1942, p. 377 (six-
armed specimen from Bermuda)”]. Carrera, 1974:
iii, 69-71, pi. 5, fig. 2a, b (in part). Emson et al.,
1985:87-100 (not Say, 1825, in part).
Ophiostigma sp. Hotchkiss, 1982:388, 392-393,
406-408, fig. 173a, b.

Ophiostigma isocanthum: Hendler and Littman,
1986:33-38 (not Say, 1825, in part). Hendler and
Peck, 1988:413 (not Say, 1825, in part).

Hotchkiss (1982) described a fissiparous speci-
men as Ophiostigma sp. from Belize and suggested
that six-armed individuals may be specifically dis-
tinct from O. isocanthum. Previously, six-armed

8 ■ Contributions in Science, Number 458

Hendler: New Western Atlantic Ophiuroids

Figure 3. Ophiostigma siva, new species, holotype LACM 83-226.1: A, entire, dorsal view; B, disk, dorsal view; C,
disk, ventral view. Disk diameter = 4.4 mm.

Contributions in Science, Number 458

Hendler: New Western Atlantic Ophiuroids ■ 9

Ophiostigma were regarded as a variant or as an
asexual reproductive stage of the five-armed O. is-
ocantbum.

Liitken (1859:234), in discussing five-armed
Ophiostigma moniliforme, recounted, “Of this spe-
cies I have had occasion to examine a six-armed
specimen, belonging to Apothecary Riise and taken
at St. Thomas at a depth of 4 fathoms. The Mu-
seum’s specimens have only 5 arms.” Liitken thought
that O. moniliforme might be identical to Ophiura
(= Ophiostigma ) isocantha Say, with which it was
synonymized by Lyman (1865) and subsequent au-
thors.

Additional six-armed individuals were reported
by H.L. Clark (1901:249), who found that one of
four Ophiostigma examined from Puerto Rico and
one of four from Bermuda had six arms (H.L. Clark,
1942). Koehler (1913, 1914) recorded a six-armed
specimen from the Tortugas and another from Key
West, Florida. More recent papers have noted that,
in Puerto Rico (Carrera, 1974), Jamaica (Emson et
al., 1985), and Belize (Hendler and Littman, 1986;
Hendler and Peck, 1988), six-armed individuals of
“ Ophiostigma isocanthum ” predominate.

H.L. Clark was the first author to suggest that
O. isocanthum exhibits fissiparity, which he re-
ferred to as “autotomous reproduction” (H.L. Clark,
1933:51) and “schizogenesis” (H.L. Clark, 1942:
377). The occurrence of fissiparity in O. isocan-
thum was accepted by Thomas (1962:692) but called
into question by Parslow and A.M. Clark (1963:
37), who suggested that H.L. Clark’s (1942) six-
armed specimen from Bermuda was a misidentified
Ophiocomella ophiactoides (H.L. Clark). Parslow
and A.M. Clark (1963:38) wrote that “H. L. Clark
followed Liitken, Lyman and other nineteenth cen-
tury workers in assuming that such six-armed
ophiuroids are conspecific with the sympatric five-
armed forms and represent the juvenile phase of
these.” In fact, fissiparous Ophiostigma are present
at Bermuda; Dennis Devaney identified an individ-
ual from the island (Pawson, pers. comm.). Six-
armed individuals are not a fissiparous growth stage
of O. isocanthum as suggested by H.L. Clark.
Hotchkiss’s (1982) proposal that the six-armed
Ophiostigma represent a separate species is con-
firmed in the present contribution.

The taxonomic status of fissiparous Ophiostigma
has been difficult to resolve because five-armed
specimens tend to be larger than those with six
arms. Therefore specimens of similar size are com-
pared in this study, and since some taxonomic char-
acteristics are altered by the process of fissiparity,
caution must be exercised in evaluating affected
structures.

ETYMOLOGY. A noun in apposition, from San-
skrit, for the Hindu god of fertility and destruction,
in reference to the fissiparous mode of reproduction
of this prolific, 6-armed species.

MATERIAL EXAMINED. Designated types originally
cataloged as Ophiostigma isocanthum unless otherwise
noted.

10 ■ Contributions in Science, Number 458

Holotype. BELIZE: (LACM 83-226.1), ale, Sta. CBC
83-14JAN, 16°48.12'N, 88°04.73'W, reef crest, Carrie Bow
Cay, Belize, 0.2-1. 5 m, coll. G. Hendler.

Paratypes. FLORIDA, ATLANTIC: (LACM 85-242.20),
18 ale, Sta. LK 22, 8 May 1985, 24°32.3'N, 81°24.1'W,
Looe Key National Marine Sanctuary, Florida, 6-8 m,
scuba, coll. G. Hendler et al.; (LACM 85-275.3), 13 ale,
Sta. LK 55, 17 Aug. 1985, 24°32.0'N, 81°24.3'W, Looe
Key National Marine Sanctuary, Florida, 24-26 m, scuba,
coll. G. Hendler et al.; (LACM 85-277.8), 12 ale, Sta. LK
57, 18 Aug. 1985, 24°31.9'N, 81°25.7'W, Looe Key Na-
tional Marine Sanctuary, Florida, 24 m, scuba, coll. G.
Hendler et al. BELIZE: (LACM 81-221.1), 7 ale, Sta. CBC
81-3, 17 Apr. 1981, 16°48.14'N, 88°04.50'W, ESE of Car-
rie Bow Cay, Belize, 15 m, scuba, coll. G. Hendler; (LACM
83-218.1), 13 ale, Sta. CBC 83-28MAR, 28 Mar. 1983,
16°48.14'N, 88°04.73'W, E of Carrie Bow Cay, Belize, 0-
1 m, coll. G. Hendler; (LACM 83-219.1), 41 ale, Sta. CBC
83-31MAR, 31 Mar. 1983, 16°48.12'N, 88°08.73'W, E of
Carrie Bow Cay, Belize, 0-1 m, coll. B. Littman; (LACM
83-226.2), 5 ale, CBC 83-14JAN, 16°48.12'N, 88°04.73'W,
reef crest, Carrie Bow Cay, Belize, 0.2-1. 5 m, coll. G.
Hendler; (LACM 85-455.3), 11 ale, Sta. CBC 85-2, 13 Jun.
1985, 16°48.70'N, 88°08.80'W, Blue Ground Range, Bar-
rier Reef Lagoon, Belize, 1 m, coll. G. Hendler; (LACM
86-32.1), 8 ale, Sta. CBC 86-1, 2 Apr. 1986, 16°48.14'N,
88°04.50'W, ESE of Carrie Bow Cay, Belize, 18 m, scuba;
(USNM E19910), originally as O. isacanthum, 4 ale, Sta.
CBC-66A, 29 Apr. 1974, Carrie Bow Cay, Belize, coll. F.
Hotchkiss; (USNM E45488), 12 ale, Sta. CBC 81-43, 26
Apr. 1981, 16°48.12'N, 88°04.73'W, E of Carrie Bow Cay,
Belize, 0-1 m, coll. G. Hendler; (USNM E45489), 13 ale,
Sta. CBC 83-15NOV, 15 Nov. 1983, 16°48.12'N,
88°04.73'W, E of Carrie Bow Cay, Belize, 0-1 m, coll. B.
Littman.

Other Material Examined. NORTH CAROLINA.
USNM cat. no. E29145, 1 ale. SOUTH CAROLINA:
USNM cat. nos. E29094, 2 ale; E29104, 1 ale (lot origi-
nally included 3 O. isocanthum ). FLORIDA, ATLAN-
TIC: LACM cat. nos. 84-230.13, 2 ale; 85-257.8, 1 ale;
85-262.10, 3 ale; 85-268.13, 6 ale; USNM cat. no. E37724,
3 ale (lot originally included 3 O. isocanthum ); IRCZM
cat. nos. 74:361, 2 dry; 74:412, 1 dry; 74:599, 2 ale (lot
originally included 1 O. isocanthum). FLORIDA, GULF
OF MEXICO: LACM cat. nos. 89-237.2, 3 ale; 89-320.3,
1 ale; 89-321.2, 1 ale; 89-323.6, 3 ale; 89-327.3, 1 ale; 89-
328.4, 2 ale; 89-329.6, 2 ale; 89-342.4, 3 ale; 89-343.4, 1
ale; 89-347.2, 1 ale; 89-360.2, 1 ale; 89-365.2, 1 ale; 89-

366.1, 1 ale; 89-367.5, 1 ale; 89-367.6, 1 ale; 89-376.2, 1
ale; 89-378.3, 4 ale; USNM cat. nos. 14115, 1 dry (lot
originally included 3 O. isocanthum ); 15400, 1 dry; E23744,
1 dry. VIRGIN ISLANDS: USNM cat. no. 15391, 1 dry
(lot originally included 2 O. isocanthum ). BELIZE: LACM
cat. nos. 80-155.1, 5 ale; 80-156.1, 1 ale; 80-157.1, 1 ale;
81-222.1, 4 ale; 81-223.1, 4 ale; 81-224.1, 2 ale; 81-225.1,

1 ale; 81-226.1, 2 ale; 81-227.1, 1 ale; 81-228.1, 2 ale; 81-

230.1, 7 ale; 81-231.1, 3 ale; 81-232.1, 1 ale; 81-233.1, 1
ale; 81-234.1, 1 ale; 81-235.1, 1 ale; 81-236.1, 1 ale; 82-

124.1, 3 ale; 83-214.1, 4 ale; 83-220.1, 2 ale; 83-221.1, 1
ale; 85-455.1, 3 ale; 85-164.3, 1 ale; 85-458.1, 2 ale; 86-

56.1, 1 ale; USNM cat. nos. E17688, 1 dry; E19913, 1
ale; E19916, 1 ale; E21164, 2 ale; E21240, 3 ale; E21253,

2 ale; E21261, 1 ale; E30601, 1 ale.

DIAGNOSIS. Arms 6 in number. Disk covered
with short tubercles, gradually increasing in size
toward radial shields. Dorsal scales bumpy, ap-
pearing fused together. Primary plates absent. Three
oral papillae, distalmost pair thick, operculate, clos-

Hendler: New Western Atlantic Ophiuroids

ing part of oral slit. Three arm spines, tapering to
narrow tips; middle spine curved, swollen at base,
drastically narrowed near midshaft, excavated dis-
tally. Radial shields nearly bare of tubercles, nar-
row; thick, bead-shaped ossicles joined to distal
margin. Adoral shields with distal lobes separated
by first ventral arm plates. Proximal dorsal arm
plates rounded triangular, slightly wider than long,
separated nearly entire length of arm. Lateral arm
plate stereom with expanded peripheral trabeculae.
Ventral arm plate length exceeds width.

DESCRIPTION OF HOLOTYPE. Disk diame-
ter 4.4 mm; longest arm 18 mm, regenerating. Disk
hexagonal in outline, indented interradially, flat-
tened dorsally, inflated near arms, bearing short
blunt spines above and below. Arm joints below
disk narrower than those beyond edge of disk. Arms
slender, tapering, dorsoventrally flattened only near
tip; joints set off by constrictions of lateral arm
plates. Major ossicles of the disk and arms are thick-
ened, dorsal arm plates more markedly so than
ventral arm plates.

Body wall pliable, bumpy, scale covered. Scale
edges not discernable; swollen central region on
many scales bearing blunt tubercles. Most tubercles
0.04-0.2 mm tall, size gradually increasing toward
edge of disk near radial shield; small tubercles
broader at base than tip, larger tubercles not ta-
pering. Radial shields clearly visible, bearing very
few tiny spines; diverging and broader proximally,
length approximately one-quarter diameter of disk.
Thick, bead-shaped plate borne on narrow distal
end of radial shield.

Infradental papillae paired, blocklike, in contact;
middle oral papillae slightly larger than infradental,
pyramidal, broad edge facing oral slit; distal oral
papilla largest, spanning middle papilla to first ven-
tral arm plate, quadrilateral, much longer than wide,
elongate inner edge deeply incised and ridged. Dis-
talmost papillae of adjacent jaws in contact, sealing
outer oral gap. Teeth chisel-shaped, narrowing
proximally such that tips of adjacent teeth not in
contact; proximal edge straight, stereom imperfo-
rate.

Oral shields small, arrowhead-shaped, length
nearly equal to width, approximately equal in size
to an adoral shield, with sharply pointed proximal
apex, proximal edges somewhat concave, lateral
edges short, distal edge bluntly rounded. Shield de-
pressed at center. Madreporite of similar shape,
slightly larger.

Adoral shield gently curving, quadrilateral, with
part of convex adradial edge adjoining incurved side
of oral shield, concave radial edge facing mouth.
Adoral shields with inner ends meeting broadly
proximal to oral shield, outer ends separated by
first ventral arm plate.

Prominent, swollen genital scale bridging oral
shield and genital slit; row of thin, smooth scales
sharply delineates edge of slit. Bursal slit narrow,
spanning approximately 2A length of interradius; slit
inconspicuous, obscured by arm.

Contributions in Science, Number 458

Dorsal arm plates rounded-triangular, slightly
wider than long; apex rounded, lateral edges con-
vex, distal edge nearly straight, with medial bulge
raised above proximal surface of following plate.
Proximal plates nearly touching, most separated by
lateral arm plates, gap increasing distally.

Lateral arm plates with stereom of expanded pe-
ripheral trabelculae appearing as transparent, mi-
croscopic grains. Lateral plates constricted proxi-
mally, distal end broadened, forming thick spine-
bearing ridge.

Three arm spines, erect, with rounded base,
somewhat dorsoventrally flattened, gradually ta-
pering to blunt tip. Dorsal spine longest, length less
than dorsal arm plate. Middle spine appearing
curved: bulging at base, much narrower near mid-
shaft, more excavated distally.

First ventral arm plate, small, pentagonal, straight-
sided, inserted at distal edge of jaw slit, adjoined
laterally by adoral shields. Other ventral arm plates
pentagonal, proximal edges meeting at rounded
apex, lateral edges concave, distal edge convex,
thickened. Length of second ventral arm plate equals
width, distal plates longer than wide. First several
ventral arm plates beyond edge of disk overlapping
other plates nearly in contact.

Tentacle scales paired, flattened, spinelike with
blunt tip, only partially covering tentacle pore;
proximal scale on lateral arm plate slightly smaller
than more distal scale on ventral arm plate.

Coloration. In alcohol yellowish, arm spines and
tips of radial shields white. Arms banded with pale
brown pattern on the dorsal arms plates. In life,
specimens typically have a pale disk, mottled with
rust-colored patches; the arms are tan with irregular
rusty marks and gray-brown blotches forming a
median chainlike pattern, the arm spines are pale
and may have reddish orange tips.

VARIATIONS. Specimens examined have disk
diameters from 1.0 to 4.4 mm and arm lengths from
6.8 to 19 mm. Maximum disk diameter/arm length
ratio is 1/7.4. The majority of individuals have 6
arms of nearly equal length, or 3 long arms and 3
short regenerating arms. Rarely, unmistakable spec-
imens of O. siva possess 5 arms. Of the specimens
examined, those with 5 arms were distributed as
follows: none from localities between North Car-
olina and Georgia (n = 5), 1.8% (3 of 165) from
Belize, 6.3% (3 of 47) from Looe Key, Florida, and
7.6% (2 of 26) from Florida Bay. Not even the
smallest specimens examined have discernable pri-
mary plates. Some Belizean specimens have aber-
rant swollen joints in the central region of the arm.

COMPARISON. Four Ophiostigma species are
now recognized, of which only O. siva is fissipa-
rous. The latter is readily distinguished from the
sympatric O. isocanthum by characteristics of its
dorsal arm plates (separate rather than in contact),
arm spines (abruptly narrowed rather than peglike),
and adoral shields (separate rather than united in a
ring). The disk lacks strikingly prominent tubercles
near the radial shields that characterize its congener.

Hendler: New Western Atlantic Ophiuroidsl 11

Figure 4. A portion of the inner surface of the disk
wall, showing radial shields and surrounding scales of A,
Ophiostigma isocanthum (Say), disk diameter = 2.8 mm
(USNM E29141); B, Ophiostigma siva, new species, disk
diameter = 2.8 mm (LACM 83-219.1). Scale bar = 0.5
mm.

Furthermore, by dissecting open the disks of both
species, a marked difference can be seen in the
shapes of the radial shields (Fig. 4A, B).

Ophiostigma abnorme (Lyman, 1878) is unique
within the genus in having clearly demarcated, im-
bricating disk scales. Unlike O. isocanthum it has
relatively few disk tubercles, which are spinelike,
adoral shields not in contact distally, and narrow
radial shields 3 times longer than wide (Madsen,
1970). Differences between Ophiostigma tenue
Liitken, 1856, and O. isocanthum include the shape
of oral and adoral shields and the pattern of disk
granules (Koehler, 1913; Nielsen, 1932). Ophiostig-
ma rugosum H.L. Clark, 1918, is distinguished by
the disk covering composed of widely spaced tu-
bercles and radial shields free of tubercles but with
edges concealed by a thick integument. Its lateral
arm plates, described as meeting above and below
except at the base of the arm, also set it apart.
Ophiostigma formosa Liitken, 1856, has been
transferred to Dougaloplus (A.M. Clark, 1970).

Ophiostigma abnorme is distributed from the
Gulf of Mexico (in deep water), to Ascension Is-
land, Cape Verde Islands, the Gulf of Guinea, and
Gold Coast of Africa, from 16 to 185 m (Madsen,
1970). O. tenue is known from the Pacific coasts
of Nicaragua and Panama, from 7 to 46 m (Liitken,
1859; Nielsen, 1932). O. rugosum is reported from
the Philippines (H.L. Clark, 1918).

DISTRIBUTION. Reliable records in the liter-
ature (noted in the synonymy) include Bermuda,
the Florida Keys and Tortugas, Puerto Rico, Ja-
maica, St. Thomas, and Belize, from less than 1 m
to 42 m. Material in USNM and LACM collections
shows that the species also occurs off North Car-
olina and South Carolina and in the Gulf of Mexico,
at depths to 99 m. O. siva may have been collected
at other localities and identified as O. isocanthum
by previous authors.

BIOLOGY. This species is fissiparous, as sug-
gested by H.L. Clark (1901, 1942) and Hotchkiss
(1982). The hundreds of specimens examined typ-
ically have 3 longer and 3 shorter arms, and every

stage between recent disk division and nearly com-
plete arm regeneration is in evidence. The incidence
of 5-armed individuals is very low, as noted under
Variations. Among the specimens examined, there
was no evidence of a development or transforma-
tion of 6-armed to 5-armed individuals. A small
number of specimens that were dissected lacked
discernable gonads, and swelling of the body wall
(indicative of ripe gonads) was not observed in any
specimens examined. This may indicate that sexual
reproduction of the species, if it occurs, is uncom-
mon.

It is notable that both O. isocanthum and O.
siva can occur in the same localities and habitats
and sometimes in the same clump of substrate
(Hendler et al., 1995). O. siva ranges from the back
reef to forereef slope zones and is most abundant
in calcareous algae, such as the Halimeda opuntia
(Hendler and Littman, 1986; Hendler and Peck,
1988).

Relative numbers of O. isocanthum and O. siva
vary widely among the available collections. Among
350 individuals of Ophiostigma from North Car-
olina to Georgia, from 20- to 30-m depths, only
1.4% were O. siva. In contrast, 69% of the indi-
viduals from Looe Key and Florida Bay, Florida,
and from Belize, all from less than 24 m, were O.
siva. Factors related to latitude may affect abun-
dance of the 2 species, and in a circumscribed region
other environmental factors may prevail. For ex-
ample, in Belize, O. isocanthum predominated in
collections from mangrove cays, and O. siva was
by far more abundant in the reef habitat, but exactly
comparable, quantitative collections were not made
in the 2 habitats.

As noted for O. isocanthum, individuals are col-
lected with particles of sediment stuck to the disk
and arms, and the particles continue to adhere to
preserved specimens. Evidently, these animals pro-
duce a mucus that binds sediment to the body wall.

Genus Amphioplus Verrill, 1899

Subgenus Amphioplus Verrill
(as restricted, A.M. Clark, 1970)

Amphioplus (Amphioplus) sepultus,
new species

Figures 5A-C, 6D-I

Amphioplus abdita: Koehler, 1914:71 (not Am -
phipholis [= Amphioplus] abdita Verrill, 1871,
in part).

Amphioplus abditus H.L. Clark, 1918:294 (not
Verrill, 1871); 1919:56, 59, 60, 66 (not Verrill,
1871, in part); 1933:38, 55-56, 141 (not Verrill,
1871, in part). A.H. Clark, 1921:42 (not Verrill,
1871); 1954:377 (not Verrill, 1871). Mortensen,
1933:110, 111 (not Verrill, 1871, in part). Pear-
son, 1937:70 (not Verrill, 1871). McNulty, 1961:
411, 414, 418, 422, 424, 429-431 (not Verrill,
1871); 1970:36, 40-43, 46, 49, 51, 54, 57, 59 (not
Verrill, 1871). Tabb and Manning, 1961:566 (not

12 ■ Contributions in Science, Number 458

Hendler: New Western Atlantic Ophiuroids

Figure 5, Amphioplus sepultus, new species: A, paratype LACM 85-363.4, entire, dorsal view, disk diameter = 7.0
mm, longest arm = 90 mm; B, holotype (USNM E14025), disk, dorsal view; C, disk, ventral view. Disk diameter =
8.7 mm.

Verrill, 1871). McNulty et al., 1962a:218, 223,
224, 229 (not Verrill, 1871). McNulty et al,
1962b:329 (not Verrill, 1871). Thomas, 1962:636,
651-654, 656, 657, 660 (not Verrill, 1871, in

part). Parslow and A.M. Clark, 1963:26, table I
(not Verrill, 1871). O’Gower and Wacasey, 1967:
197 (not Verrill, 1871). Halpern, 1970:630 (not
Verrill, 1871). Hudson et al., 1970:9 (not Verrill,

Contributions in Science, Number 458

Hendler: New Western Atlantic Ophiuroids ■ 13

1871). A.M. Clark, 1970:48, 56 (not Verrill, 1871,
in part). Singletary, 1971:940 (not Verrill, 1871).
Woodley, 1975:29, 37, 44 (not Verrill, 1871).
Amphioplus sepultus: Hendler, 1973 ( nomen nu-
dum):i-255; 1991 ( nomen nudum):366 , 377, 384,
402, 412, 414, 421, 424. Pettibone, 1993 ( nomen
nudum)'A\ , 42.

Amphioplus sp. Humes and Hendler, 1972:539, 541,
546, 551, 555.

The taxonomic status of the new species and two
congeners, Amphioplus abditus (Verrill, 1871) and
Amphioplus macilentus (Verrill, 1882), has been
chronically confused. The latter two species have
been lumped and split several times, and the new
species has, with few exceptions, gone unrecog-
nized (Hendler, 1973, 1991; Pettibone, 1993).
However, distinctions among the three taxa are
clearcut, based on external and internal morphol-
ogy, ontogeny, and physiology (Hendler, 1973).

A. abditus was originally described from Con-
necticut, off New Haven (Verrill, 1871). Its con-
firmed range extends from Grand Manan, Maine,
to Sapelo Island, Georgia, typically in shallow water
but at depths to 40 m (Hendler, 1973). Amphioplus
macilentus was first discovered off Martha’s Vine-
yard, Massachusetts and ranges at least to North
Carolina on the edge of the continental shelf be-
tween 97 and 210 m (Verrill, 1885).

Koehler (1914) was the first to record the new
species from Florida but, as reflected in the syn-
onymy above, he and other authors identified the
Floridian specimens as A. abditus. Under the name
A. abditus, Thomas (1962:654) considered A. se-
pultus “the most common intertidal amphiurid of
South Florida.”

ETYMOLOGY. Sepultus, the masculine form of
the Latin participle for “buried,” in reference to
the burrowing habit of the species.

MATERIAL EXAMINED. Designated types originally
cataloged as A. abditus unless otherwise noted.

Holotype. FLORIDA, ATLANTIC: (USNM E14025),
ale, 9 Jan. 1972, Virginia Key, Miami, Florida, 1.5 m, coll.
G. Hendler.

Paratypes. FLORID A, ATLANTIC: (LACM 68-456.1),
1+ ale, Sta. 15-68, 27 Mar. 1968, Biscayne Bay, Miami,
Florida, coll. R.L. Singletary and T. Borkowski; (LACM
85-265.4), originally as Amphioplus sp., 1 ale, Sta. LK 45,
15 May 1985, 24°33.7'N, 81°25.7'W, Looe Key National
Marine Sanctuary, Florida Keys, Florida, 11m, Hendler
et al.; (LACM 88-195.2), originally as A. sepultus, 7 ale,
Sta. MI-I/88-2, 17 Jan. 1988, 25°44'N, 80°10'W, Sea-
quarium flats, Rickenbacker Causeway, Virginia Key, Bis-
cayne Bay, Miami, Florida, 0-1 m, coll. J.E. Miller; (LACM
88-196.2), originally as A. sepultus, 1 ale, Sta. MI-1/88-
3, 17 Jan. 1988, 25°43.5'N, 80°09.5'W, NW Point, Key
Biscayne, Florida, 0-1 m, J.E. Miller et al.; (LACM 88-
196.3), originally as A. sepultus, 1 ale, Sta. MI-I/88-3, 17
Jan. 1988, 25°43'N, 80°09.5'W, NW Point, Key Biscayne,
Florida, 0-1 m, coll. J.E. Miller; (USNM E14026), 29 ale,
9 Jan. 1972, Virginia Key, Miami, Florida, 1.5 m, coll. G.
Hendler; SOUTHWEST FLORIDA, GULF OF MEXI-
CO: (LACM 85-361.1), originally as A. sepultus, 2 ale,

Sta. FK-04, 11 May 1985, 24°41'N, 81°13.5'W, W end of
Seven Mile Bridge, E end of Little Duck Key, 1 m, coll.
G. Hendler et al.; (LACM 85-362.2), originally as A. se-
pultus, 5 ale, Sta. LK 05, 11 May 1985, 24°39.1'N, 81°18'W,
E end of W Summerland Key, 0.3-0.9 m, coll. G. Hendler
et al.; (LACM 85-363.4), originally as A. sepultus, 5 ale,
Sta. FK-06, 11 May 1985, 24°42'N, 81°24.7'W, N end of
Middle Torch Key, Florida, coll. G. Hendler et al.

Other Material Examined. FLORIDA: cat. no. MCZ
6747, 1 dry. FLORIDA, ATLANTIC: MCZ cat. nos. 1457,

1 dry; 1533, 1 ale; 5505, 34 ale; 5506, 39 dry; 5571, 1
dry; uncat. lot, several. SOUTHWEST FLORIDA, GULF
OF MEXICO: USNM cat. nos. 6866, 3+ ale; 12674, 50+
alc; 14002, 1 ale; 33873, 1 dry; 38924, 5 dry; MCZ cat.
nos. 1458, 1 dry; 4055, 2 dry; 4056, 2 ale; 4239, 4 dry;
4240, 2 dry; 4282, 4 ale; 4283, 3 ale; 6647, 2 dry; 6660,

2 dry; UMML cat. nos. 41.99, 1 ale; 41.101, 14 ale; 41.136,
2 ale.

DIAGNOSIS. Entire disk covered with fine, im-
bricating scales; scales largest and thickest near ra-
dial shields and disk edge. Radial shields 2-3 times
longer than wide, separated by scales except dis-
tally. Five oral papillae. Oral shields twice as long
as wide, spearhead-shaped, with small, paired lat-
eral lobes. Two tentacle scales. Three arm spines;
dorsal shortest, laterally compressed, tapering, tip
rounded; middle spine broadest, dorsoventrally
flattened, tapering to blunt flat tip; ventral spine as
long as middle spine, curving ventrodistally, with
deep channel on dorsal-distal surface. Microscopic
spinules most prominent on middle spine.

DESCRIPTION OF HOLOTYPE. Disk diame-
ter 8.7 mm; arms broken, approximately 11 cm in
length. Disk thin, flexible, rounded pentagonal, out-
pouching between arms, but indented interradially.
Arms slender, narrowed near disk, distalmost third
of arm tapering to filiform tip. Arms gently rounded
dorsally, flattened ventrally; dorsal arm plates of
successive joints in contact, successive ventral arm
plates slightly separated by lateral arm plates, lateral
arm plates bridged by soft tissue.

Disk entirely covered with small, imbricating
scales; scales smallest at center of disk, largest around
radial shields; enlarged scales at disk edge demar-
cate dorsal and lateral disk surfaces. Primary plates
minute, widely separated. Radial shields with straight
adradial edge, broadly rounded abradial edge,
thickened medial ridge, thick distal tip directed ad-
radially; shields tapering proximally, 2-3 times lon-
ger than wide; separated by wedge of scales prox-
imally, joined distally.

Each jaw bears 5 pairs of oral papillae; infradental
largest, separated, thick, blocklike, longest axis dor-
soventral; buccal tentacle scale smallest, blunt spine-
shaped, high in jaw; 2 flattened scales on oral plate,
proximal scale smaller, distal scale larger, triangular
to quadrangular; outermost scale seated between
adoral shield and first ventral arm plate. Teeth with
concave tips, stereom of proximal edge imperfo-
rate.

Oral shield narrow, blunt spearhead-shaped with
small lateral lobes, approximately 1.5 times as long
as wide. Madreporite slightly larger than shields,

14 IS Contributions in Science, Number 458

Hendler: New Western Atlantic Ophiuroids

asymmetrical. Adoral shield 3-lobed; largest lobes
of adjacent adoral plates meet proximally to oral
plate; small radial lobe in contact with first ventral
arm plate; thin adradial lobe touching bursal slit.

Bursal slits spacious, extending to disk edge. Ven-
tral interbrachial field covered by fine, flakelike,
imbricating scales; scale size greatest near bursal
slit; density of scales diminishing near oral shield.

Dorsal arm plates ovoidal hexagonal, 1.5-1. 8
times broader than long; proximolateral edges
slightly concave; short lateral edges truncate; distal
edge broad, convex. Plates touching or nearly in
contact, distal edge contacting or overlapping prox-
imal edge of adjacent plate.

Lateral arm plates narrow in dorsal aspect, with
prominent spine-bearing ridge conspicuously pro-
truding from arm. Three arm spines, with broad
bases, tapering, equal in length to arm joint; dorsal
spine shortest, laterally compressed, tip rounded,
channel on lateral surface (Fig. 6D, G); middle spine
broadest, dorsoventrally flattened, with blunt, flat
tip, with channels on dorsal and ventral edges (Fig.
6E, H); \ entral spine as long as middle spine, flat-
tened dorsally, convex ventrally, curving ventro-
posteriorly, with deep channel on dorsal-distal sur-
face (Fig. 6F, I). All spines with microscopic spi-
nules, rugosity most prominent on middle spine.

Ventral arm plates pentagonal, edges slightly
concave, barely separated by lateral arm plates,
proximolateral edges shortest, distal edge longest.
Second plate longer than broad, center depressed,
lateral and distal edges raised; plate width increasing
to fifth or sixth plate, width/length ratio 0.9/1. 1.
Plates touching or nearly in contact.

Two tentacle scales proximally, in contact, re-
duced to 1 scale near arm tip; larger scale arising
on lateral and ventral arm plates, smaller scale on
ventral arm plate. Tube foot with smooth shaft and
bulbous tip.

Coloration. In life, pigmentation variable; gray
or brown overall, variegated with light gray or red-
dish tan and contrasting dark gray. Arms dorsally
mottled, sometimes having pale middorsal line.
Spines pale with dark basal spot internally. Disk
coloration uniform to the naked eye, at low mag-
nification appearing speckled, darker scales having
a pale border. Radial shield contrastingly dark or
mottled, with pale edges and distal tip. Ventral disk
scales light gray to tan. First ventral arm plate and
infradental papillae densely pigmented. Adoral and
genital shields often patterned with black. Ventral
arm plates darkest proximally, blotched distally;
sometimes with a midventral stripe. Color of testes
whitish. Ovaries off-white to gray, sometimes with
yellow or green tinge. In alcohol color fades to
white.

VARIATIONS. Disk diameters of specimens ex-
amined range from 0.3 to 0.9 cm, with arm lengths
from 3 to 11 cm. The distalmost (accessory) oral
papillae may number more than 1 in large individ-
uals. As individuals grow, the primary plates be-
come more widely separated from one another, and

Contributions in Science, Number 458

the central and radial plates display negative allom-
etry in diameter. The numbers of tentacle scales
range from 0 to 5 on the proximal arm joints. The
oral shields vary in shape, sometimes subelliptical,
pentagonal, or hexagonal, sometimes with the
proximal edge squared off. The madreporite usually
has 1 perforation, occasionally up to 4. Proximal
corners of the adoral shields may be touching or
separate and may be concealed by the proximal
edge of the oral shield. Individuals from reef en-
vironments tend to be smaller and more pale than
those from eutrophic Gulf Coast habitats.

Details of internal anatomy, including oral plate
shape, have been documented (Hendler, 1973). In
the place of peristomial plates there are groups of
small thin scales. Arm vertebrae are perforate.
Wedge-shaped dental plates possess up to 8, rarely
9, foramina. In ripe individuals, oocytes number
3,000-9,000, with a mean diameter of 0.17 mm.

COMPARISONS. A. ( Ampbioplus ) sepultus dif-
fers from the more than 30 species in the nomi-
notypical subgenus in having 3 blunt arm spines
with the morphology noted above. The character-
istic middle arm spine is similar in shape to those
in A. platy acanthus (Murakami, 1943), but the lat-
ter species has 4 arm spines.

Amphioplus sepultus is readily distinguished from
its Floridian congeners. A. thrombodes H.L. Clark,
1918, differs in having papillose scales on the dorsal
surface of the disk and only 1 tentacle scale. A.
coniortodes H.L. Clark, 1918, differs in having the
disk extremely fine-scaled dorsally and naked ven-
trally, with thin radial shields that are 4 times longer
than wide, and very elongate arms, 20 times the
disk diameter in length.

The distinctions between A. sepultus, its larger
northern congener A. abditus, and its smaller deep-
water congener A. macilentus, have been described
at length (Hendler, 1973). Distinguishing features
include the structure of the ventral arm plates and
oral plates and are most readily apparent for the
arm spines.

Individuals of the 3 species of equivalent size
have arm spines of contrasting length (Hendler,
1973) and shape. The 3 arm spines of A. macilentus
are all slender and acutely pointed (Fig. 6A-C). The
dorsal and ventral spines of A. abditus and A. se-
pultus are similarly shaped (e.g. Fig. 6G, I, M, O);
however, the middle spine of A. abditus is con-
stricted, with an expanded tip bearing large spinules
(Fig. 6K, N, Q), contrasting with the tapering, blunt-
tipped spine of A. sepultus (Fig. 6E, H).

DISTRIBUTION. Florida waters, from Biscayne
Bay to the Dry Tortugas on the Atlantic Coast, and
from Flamingo in Florida Bay to Destin on the Gulf
of Mexico coast; intertidal to 82-m depth.

BIOLOGY. A. sepultus is a burrowing deposit
feeder, ingesting sediment and plant material, algae,
pollen, fecal pellets, and microscopic invertebrates
(Hendler, 1973). Its arms undulate beneath the sed-
iment, circulating water through the burrow; a “re-
spiratory fringe” of mucus and sediment on the arm

Hendler: New Western Atlantic Ophiuroidsl 15

Figure 6. Light microscope preparations showing arm spine morphologies of Amphioplus macilentus (Verrill), A.
sepultus, new species, and A. abditus (Verrill); differences related to body size are illustrated for the latter two species.
A. macilentus: disk diameter = 23-4.2 mm— A, dorsal spine; B, middle spine; C, ventral spine. A. sepultus: disk diameter
= 2.4-4.4 mm — D, dorsal spine; E, middle spine; F, ventral spine; disk diameter = 4.4™ 6. 4 mm— G, dorsal spine; FI,
middle spine; I, ventral spine. A. abditus: disk diameter = 3.5-4. 9 mm— J, dorsal spine; K, middle spine; L, ventral
spine; disk diameter = 5. 2-9. 4 mm — M, dorsal spine; N, middle spine; O, ventral spine; disk diameter = 14.4-15.3
mm — P, dorsal spine; Q, middle spine; R, ventral spine. Scale bar = 0.5 mm.

16 ■ Contributions in Science, Number 458

Flendler: New Western Atlantic Ophiuroids

spines creates a gasket between the arm and the
burrow wall (Woodley, 1975).

Amphioplus sepultus (as A. abditus ) has been
designated a dominant in an “ Amphioplus-Dosinia
community,” where there may be several hundred
individuals per square meter (McNulty et ah, 1962a,
b), and has been found to occupy other benthic
communities (McNulty, 1970). The species has been
recorded from soft sticky sediment (Pearson, 1937),
carbonate mud (Hudson et al., 1970), and firm sandy
mud (H.L. Clark, 1918) and often associated with
marine spermatophytes (Pearson, 1937; McNulty,
1961; O’Gower and Wacasey, 1967) and sometimes
near sewage pollution (McNulty, 1961).

Halpern (1970) noted that this species is a food
item of the sea star Luidia sengalensis (Lamarck),
and Singletary (1971) determined its upper lethal
temperature limit, approximately 40°C. The annual
reproductive, feeding, respiratory, and growth cy-
cles of the species have been examined (Hendler,
1973). Two external copepod associates have been
reported from A. sepultus (Humes and Hendler,
1972); an internal copepod and nauplii (which cas-
trates the host) live in the coelom, a metacercaria
(Subfamily Allocreadioidea) occurs in the gonad, a
sessile rotifer (unidentified) lives on the arm spines,
and a polynoid polychaete ( Malmgreniella mac-
craryae Pettibone) and a bivalve ( Montacuta sp.)
are found on the disk (Hendler, 1973; Pettibone,
1993).

ACKNOWLEDGMENTS

Since this contribution incorporates research efforts span-
ning over 20 years, I am indebted to more colleagues for
more help with this work than I can relate here. Still, I
am grateful for access to museum specimens and data to
A. Clark and G. Paterson (British Museum, Natural His-
tory), P. Mikkelsen, J. Miller, and D. Vaughn (IRCZM),
A. Schoener, R. Woollacott, and the late H. Fell (MCZ),
N. Voss and the late G. Voss (UMML), C. Ahearn, D.
Pawson, and C. Walter (USNM), and M. Jensen (UZM);
for guidance and help at the University of Connecticut
to D. Ashton and D. Franz, at the University of Miami
to H. Moore, R. Singletary, and L. Thomas; for support
in Belize to K. Ruetzler, staff of the Carrie Bow Cay field
laboratory, and dive partners M. Byrne, K. Clark, D.
DeFreese, B. Littman, B. Spracklin, and B. Sullivan; for
collaboration in Florida to P. Kier, B. Littman, J. Miller,
P. Mikkelsen, D. Pawson, and A. Powell; for assistance
in Costa Rica to R. Brusca, R. Peck, R. Wetzer, and M.
Murillo and CIMAR staff; for translation of Danish pub-
lications to K. Friedmann; for information regarding the
specimen photographed at Cayman Brae to E. Fish and
J. Rorem (Divers Alert Network); for helpful comments
on the manuscript to D. Pawson, J. Dearborn (University
of Maine), and the publications committee of the Natural
History Museum of Los Angeles County.

LITERATURE CITED

Abreu Perez, M. 1983. Nuevos ofiuroideos (Echinoder-
mata: Ophiuroidea) del Golfo de Batabano, Cuba.
Poeyana 259:1-6.

Alvarez Larrauri, L.R. 1981. Listado preliminar de los

Contributions in Science, Number 458

equinodermos de la costa Atlantica Colombiana.
Boletin Museo del Mar Bogota 10:24-39.

Aronson, R.B., and C.A. Harms. 1985. Ophiuroids in a
Bahamian saltwater lake: The ecology of a Paleo-
zoic-like community. Ecology 66:1472-1483.

Byrne, M. 1991. Reproduction, development and pop-
ulation biology of the Caribbean ophiuroid Ophi-
onereis olivacea, a protandric hermaphrodite that
broods its young. Marine Biology 111:387-399.

Carrera, C.J. 1974. The shallow water amphiurid brittle
stars (Ophiuroidea: Echinodermata) of Puerto Rico.
M.Sc. Thesis, University of Puerto Rico, Mayaguez,
Puerto Rico, vii + 103 pp.

Clark, A.H. 1921. Report on the ophiurans collected
by the Barbados-Antigua Expedition from the Uni-
versity of Iowa in 1918. University of Iowa Studies
in Natural History 9:29-63.

Clark, A.H. 1922. Ophiurans of the island of Curasao.
Bijdragen tot de Dierkunde uitgegeven door het
Koninklijk Zoologisch Genootschap Natura Artis
Magistra te Amsterdam 22:209-213.

Clark, A.H. 1939. Echinoderms of the Smithsonian-
Hartford Expedition, 1937 with other West Indian
records. Proceedings of the United States National
Museum 86:441-456, pis. 53, 54.

Clark, A.H. 1954. Echinoderms (other than holothuri-
ans) of the Gulf of Mexico. Fishery Bulletin of the
Fish and Wildlife Service 55:373-379.

Clark, A.M. 1953. A revision of the genus Ophionereis
(Echinodermata, Ophiuroidea). Proceedings of the
Zoological Society of London 123:64-94, pis. 1-3.

Clark, A.M. 1955. Echinodermata of the Gold Coast.
Journal of the West African Science Association 1:
16-56, pi. 2.

Clark, A.M. 1970. Notes on the family Amphiuridae
(Ophiuroidea). Bulletin of the British Museum (Nat-
ural History) Zoology 19:1-81.

Clark, H.L. 1898. Notes on the echinoderms of Ber-
muda. Annals of the New York Academy of Sciences
11:407-413.

Clark, H.L. 1899. Further notes on the echinoderms of
Bermuda. Annals of the New York Academy of Sci-
ences 12:117 -138, pi. 4.

Clark, H.L. 1901. The echinoderms of Porto Rico. Bul-
letin of the United States Fisheries Commission 20:
231-263, pis. 14-17.

Clark, H.L. 1911. North Pacific ophiurans in the col-
lection of the U. S. National Museum. United States
National Museum Bulletin 75:i-xvi, 1-302.

Clark, H.L. 1915. Catalog of Recent ophiurans: Based
on the collection of the Museum of Comparative
Zoology. Memoirs of the Museum of Comparative
Zoology at Harvard College 25:165-376, 20 pis.

Clark, H.L. 1918. Brittle-stars, new and old. Bulletin of
the Museum of Comparative Zoology at Harvard
College 62:265-338, 8 pis.

Clark, H.L. 1919. The distribution of the littoral echi-
noderms of the West Indies. Carnegie Institution of
Washington Publication, no. 281, Papers from the
Department of Marine Biology of the Carnegie In-
stitution of Washington 13:49-74, pis. 1-3.

Clark, H.L. 1933. A handbook of the littoral echino-
derms of Porto Rico and the other West Indian
Islands. New York Academy of Sciences Scientific
Survey of Porto Rico and the Virgin Islands 16:1—
147, pis. 1-7.

Clark, H.L. 1942. The echinoderm fauna of Bermuda.
Bulletin of the Museum of Comparative Zoology at
Harvard College 89:367-391, 1 pi.

Hendler: New Western Atlantic Ophiuroids ■ 17

Downey, M.E. 1969. Catalog of Recent ophiuroid type
specimens in major collections in the United States.
United States National Museum Bulletin 293:i-vi,
1-239.

Emson, R.H., P.V. Mladenov, and I.C. Wilkie. 1985.
Patterns of reproduction in small Jamaican brittle
stars: Fission and brooding predominate. In The ecol-
ogy of coral reefs, ed. M.L. Reaka, 87-100. NOAA
Symposium Series of Undersea Research, vol. 3, 9-
15. NOAA Undersea Research Program, Rockville,
Maryland.

Engel, H. 1939. Echinoderms from Aruba, Curasao,
Bonaire and Northern Venezuela. Capita Zoologica
8:1-12.

Fontaine, A. 1953. The shallow-water echinoderms of
Jamaica. Natural History Notes. Natural History
Society of Jamaica 5:197-205.

Halpern, J.A. 1970. Growth rates of the tropical sea
star Luidia senegalensis (Lamarck). Bulletin of Ma-
rine Science 20:628-633.

Heilprin, A. 1888. Contributions to the natural history
of the Bermuda Islands. Proceedings of the Academy
of Natural Sciences of Philadelphia 3:309-318, pis.
14, 15.

Hendler, G. 1973. Northwest Atlantic amphiurid brit-
tlestars, Amphioplus abditus (Verrill), Amphioplus
macilentus (Verrill), and Amphioplus sepultus n. sp.
(Ophiuroidea: Echinodermata): Systematics, zooge-
ography, annual periodicities, and larval adaptations.
Ph.D. Thesis, The University of Connecticut, Storrs,
xii + 255 pp., 61 figs.

Hendler, G. 1982. An echinoderm vitellaria with a bi-
lateral larval skeleton: Evidence for the evolution of
ophiuroid vitellariae from ophioplutei. Biological
Bulletin 163:431-437.

Hendler, G. 1988. Western Atlantic Ophiolepis (Echi-
nodermata: Ophiuroidea): A description of O. paw-
soni new species, and a key to the species. Bulletin
of Marine Science 42:265-272.

Hendler, G. 1991. Ophiuroidea. In Reproduction of ma-
rine invertebrates: Vol. VI. Echinoderms and Lo-
phophorates, eds. A.C. Giese, J.S. Pearse, and V.B.
Pearse, Pacific Grove, California: The Boxwood Press.

Hendler, G., and B.S. Littman. 1986. The ploys of sex:
Relationships among the mode of reproduction, body
size and habitats of coral-reef brittlestars. Coral Reefs
5:31-42.

Hendler, G., and J.E. Miller. 1984. Ophioderma de-
vaneyi and Ophioderma ensiferum, new brittlestar
species from the western Atlantic (Echinodermata:
Ophiuroidea). Proceedings of the Biological Society
of Washington 97:442-461.

Hendler, G., J.E. Miller, D.L. Pawson, and P.M. Kier.
1995. Sea stars, sea urchins, and allies: Echino-
derms of the Florida Keys and Caribbean. Wash-
ington, D.C.: Smithsonian Institution Press.

Hendler, G., and R.W. Peck. 1988. Ophiuroids off the
deep end: Fauna of the Belizean fore-reef slope. In
Echinoderm biology: Proceedings of the Sixth In-
ternational Echinoderm Conference, Victoria, eds.
R.D. Burke, P.V. Mladenov, P. Lambert, and R.L.
Parsley, 411-419. Rotterdam: Balkema.

Hendler, G., and R.L. Turner. 1987. Two new species
of Ophiolepis (Echinodermata: Ophiuroidea) from
the Caribbean Sea and Gulf of Mexico: With notes
on ecology, reproduction, and morphology. Con-
tributions in Science, Natural History Museum of
Los Angeles County 395:1-14.

Hotchkiss, F.H.C. 1982. Ophiuroidea (Echinodermata)

from Carrie Bow Cay, Belize. In The Atlantic Barrier
Reef Ecosystem at Carrie Bow Cay, Belize, I: Struc-
ture and communities, eds. K. Riitzlerand I.G. Mac-
intyre, 387-412. Smithsonian Contributions to the
Marine Sciences, no. 12.

Hudson, J.H., D.M. Allen, and T.J. Costello. 1970. The
flora and fauna of a basin in central Florida Bay.
United States Fish and Wildlife Service Special Sci-
entific Report— Fisheries 604:1-14.

Humes, A.G., and G. Hendler. 1972. New cyclopoid
copepods associated with the ophiuroid genus Am-
phioplus on the eastern coast of the United States.
T ransactions of the American Microscopical Soci-
ety 91:539-555.

Koehler, R. 1907. Revision de la collection des ophiures
de museum d’histoire naturelle de Paris. Bulletin
Scientifique de la France et de la Belgique 41:279-
351, pis. 10-14.

Koehler, R. 1913. Ophiures. Zoologischen Jahrbiichern
supplement ll(3):351-380, pis. 20, 21.

Koehler, R. 1914. A contribution to the study of ophiur-
ans of the United States National Museum. Bulletin
of the United States National Museum 84:vii, 1-
173, 18 pis.

Lewis, J.B. 1965. A preliminary description of some
marine benthic communities from Barbados, West
Indies. Canadian Journal of Zoology 43:1049-1074.

Ljungman, A. 1866. Ophiuroidea viventia hue usque
cognita enumerat. Ofversigt af Kongl. V etenskaps-
Akademiens Forhlandlingar 9:303-336.

Ljungman, A. 1867. Om nagra nya arter af Ophiurider.
Ofversigt af Kongl. V etenskaps-Akademiens For-
handlingar 23:163-166.

Ljungman, A. V. 1871. Forteckning ofver uti Vestindien
af Dr A. Goes samt under korvetten Josefinas ex-
pedition i Atlantiska Oceanen samlade Ophiurider.
Ofversigt af Kongl. V etenskaps-Akademiens For-
handlingar 27:615-658.

Liitken, C.F. 1856. Bidrag til Kundskabom Slange-
stjerne. II. Oversigt over de Vestindiske Ophiurer.
III. Bidrag til Kundskab om Ophiurerne ved Central-
Amerikas Vestkyst. V idenskabelige Meddelelser fra
Dansk naturhistorisk Forening i Kjobenhavn 1856:
1-26.

Liitken, C.F. 1859. Additamenta ad historiam Ophiur-
idarum. Beskrivelser af nye eller hidtil kun ufuld-
staendigt kjendte Arter af Slangestjerner. Anden Af-
deling. Kongelige Danske V idenskabernes Selskabs
Skrifter 5(1861):177-271, pis. 1-5.

Lyman, T. 1860. Descriptions of new Ophiuridae, be-
longing to the Smithsonian Institution and to the
Museum of Comparative Zoology, at Cambridge.
Proceedings of the Boston Society of Natural His-
tory 7:193-205, 252-262, 424-425.

Lyman, T. 1865. Ophiuridae and Astrophytidae. Illus-
trated Catalog of the Museum of Comparative Zool-
ogy at Harvard College 1:1-200, pis. 1, 2.

Lyman, T. 1869. Preliminary report on the Ophiuridae
and Astrophytidae dredged in deep water between
Cuba and the Florida Reef, by L. F. de Pourtales,
Assist. U. S. Coast Survey. Bulletin of the Museum
of Comparative Zoology at Harvard College 1:309-
354.

Lyman, T. 1875. Zoological results of the Hassler Ex-
pedition. II. Ophiuridae and Astrophytidae, includ-
ing those dredged by the late Dr. William Stimpson.
Illustrated Catalog of the Museum of Comparative
Zoology at Harvard College 8:1-34, pis. 1-5.

Lyman, T. 1878. Reports on the results of dredging,

18 ■ Contributions in Science, Number 458

Hendler: New Western Atlantic Ophiuroids

under the supervision of Alexander Agassiz, in the
Gulf of Mexico, by the United States Coast Survey
Steamer “Blake,” Lieutenant-Commander C. D.
Sigsbee, U. S. N., Commanding. II. Ophiurans and
Astrophytons. Bulletin of the Museum of Compar-
ative Zoology at Harvard College 5: 217-238, pis.
1-3.

Lyman, T. 1880. A preliminary list of the known genera
and species of living Ophiuridae and Astrophytidae
with their localities, and the depths at which they
have been found; and references to the principal
synonymies and authorities. Cambridge, Massachu-
setts. v + 45 pp.

Lyman, T. 1882. Report on the Ophiuroidea dredged
by H. M. S. Challenger, during the years 1873-1876.
Report on the Scientific Results of the Voyage of H.
M. S. Challenger during the Years 1873-76. Zoology
5(14):l-386, pis. 1-48.

Lyman, T, 1883. Reports on the results of dredging,
under the supervision of Alexander Agassiz, in the
Caribbean Sea (1878-79), and on the East Coast of
the United States, during the summer of 1880, by
the United States Coast Survey Steamer “Blake,”
Commander j. R. Bartlett, U. S. N., Commanding.
XX. Report on the Ophiuroidea. Bulletin of the Mu-
seum of Comparative Zoology at Harvard College
10:227-287, pis. 1-7.

Madsen, F.J. 1970. West African ophiuroids. Atlantide
Report 11:151-243.

McNulty, J.K. 1961. Ecological effects of sewage pol-
lution in Biscayne Bay, Florida: Sediments and the
distribution of benthic and fouling macro-organ-
isms. Bulletin of Marine Science of the Gulf and
Caribbean 11:394-447.

McNulty, J.K. 1970. Effects of abatement of domestic
sewage pollution on the benthos, volumes of zoo-
plankton, and the fouling organisms of Biscayne Bay,
Florida. Studies in Tropical Oceanography (Miami)
9:1-107.

McNulty, J.K., R.C. Work, and H.B. Moore. 1962a.
Level sea bottom communities in Biscayne Bay and
neighboring areas. Bulletin of Marine Science of the
Gulf and Caribbean 12:204-233.

McNulty, J.K., R.C. Work, and H.B. Moore. 1962b.
Some relationships between the infauna of the level
bottom and the sediment in South Florida. Bulletin
of Marine Science of the Gulf and Caribbean 12:
322-332.

Mortensen, T. 1933. Ophiuroidea. The Danish Ingolf
Expedition 4:1-121, 3 pis., 1 chart, 1 table.

Muller, J., and F.H. Troschel. 1842. System der Aster-
iden. Braunschweig: Friedrich Vieweg und Sohn.

Murakami, S. 1943. Report on the ophiurans of Palao,
Caroline Islands. Journal of the Department of Ag-
riculture, Kyusyu Imperial University 7:159-204.

Nielsen, E. 1932. Papers from Dr. Th. Mortensen’s Pa-
cific Expedition 1914-1916. LIX. Ophiurans from
the Gulf of Panama, California, and the Strait of
Georgia. V idenskabelige Meddelelser fra Dansk na-
turhistorisk Forening i Kjobenhavn 91:241-346.

O’Gower, A.K., and J.W. Wacasey. 1967. Animal com-
munities associated with Thalassia, Diplanthera, and
sand beds in Biscayne Bay. I. Analysis of commu-
nities in relation to water movements. Bulletin of
Marine Science 17:175-210.

Parslow, R.E., and A.M. Clark. 1963. Ophiuroidea of
the Lesser Antilles. Studies on the Fauna of Curasao
and Other Caribbean Islands 15:24-50.

Pearson, J.F.W. 1937. Studies on the life zones of marine

Contributions in Science, Number 458

waters adjacent to Miami: I. The distribution of the
Ophiuroidea. Proceedings of the Florida Academy
of Sciences 1:66-72.

Pettibone, M.H. 1993. Scaled polychaetes (Polynoidae)
associated with ophiuroids and other invertebrates
and review of species referred to Malmgrenia Mc-
Intosh and replaced by Malmgreniella Hartman, with
descriptions of new taxa. Smithsonian Contribu-
tions to Zoology 538:vi + 92 pp.

Rathbun, R. 1879. A list of the Brazilian echinoderms,
with notes on their distribution, etc. Transactions
Connecticut Academy of Arts and Sciences 5:139-
158.

Say, T. 1825. On the species of the Linnaean genus
Asterias, inhabiting the coast of the United States.
Journal of the Academy of Natural Sciences of Phil-
adelphia 5:141-154.

Singletary, R.L. 1971. Thermal tolerance of ten shallow-
water ophiuroids in Biscayne Bay, Florida. Bulletin
of Marine Science 21:938-943.

Spamer, E.E., and A.E. Bogan. 1992. General inverte-
brate collection of the Academy of Natural Sciences
of Philadelphia. Tryonia 26:vi + 305 pp.

Tabb, D.C., and R.B. Manning. 1961. A checklist of
the flora and fauna of northern Florida Bay and
adjacent brackish waters of the Florida mainland
collected during the period July, 1957 through Sep-
tember, 1960. Bulletin of Marine Science of the Gulf
and Caribbean 11:552-649.

Thomas, L.P. 1962. The shallow water amphiurid brittle
stars (Echinodermata: Ophiuroidea) of Florida. Bul-
letin of Marine Science of the Gulf and Caribbean
12:623-694.

Thomas, L.P. 1973. Western Atlantic brittlestars of the
genus Ophionereis. Bulletin of Marine Science 23:
585-599.

Tommasi, L.R. 1970. Os ofiuroides recentes do Brasil
e de regioes vizinhas. Contribuiqoes avulsas do In-
stitute Oceanografico Sao Paulo, Series Oceano-
grafia Biologica 20:1-146.

Verrill, A.E. 1871. Brief contributions to zoology from
the museum of Yale College. XV. Descriptions of
starfishes and ophiurans from the Atlantic coasts of
America and Africa. American Journal of Science,
ser. 3, 2:130-133.

Verrill, A.E. 1882. Brief contributions to zoology from
the museum of Yale College. XLIX. Notice of the
remarkable marine fauna occupying the outer banks
off the southern coast of New England, Number 3.
American Journal of Science, ser. 3, 23:135-142.

Verrill, A.E. 1885. Results of the explorations made by
the Steamer Albatross, off the northern coast of the
United States, in 1883. Report of the Commissioner
for 1883. United States Commission of Fish and
Fisheries 11:503-699.

Verrill, A.E. 1899. North American Ophiuroidea. I. Re-
vision of certain families and genera of West Indian
ophiurans. II. A faunal catalogue of the known spe-
cies of West Indian ophiurans. Transactions of the
Connecticut Academy of Arts and Sciences 10:301-
386.

Verrill, A.E. 1907. The Bermuda Islands. Transactions
of the Connecticut Academy of Arts and Sciences
12:145-348, pis. 16-40.

Woodley, J.D. 1975. The behaviour of some amphiurid
brittle-stars. Journal of Experimental Marine Biol-
ogy and Ecology 18:29-46.

Submitted 16 January 1995; accepted 21 July 1995.

Hendler: New Western Atlantic Ophiuroids ■ 19

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

Number 459
8 November 1995

Contributions

in Science

Efficiency of Two Mass Sampling
Methods for Sampling Phorid Flies
(Diptera: Phoridae) in a Tropical
Biodiversity Survey

Brian V. Brown and Donald H. Feener, Jr.

Natural History Museum of Los Angeles County

Serial
Publications
of THE
Natural History
Museum of
Los Angeles
County

• Contributions in Science, a miscellaneous series of tech-
nical papers describing original research in the life and
earth sciences.

• Science Series, long articles and collections of papers
on natural history topics.

Copies of the publications in these series are sold through
the Museum Book Shop. A catalog is available on request.

Scientific

Publications

Committee

James L. Powell, Museum President
Daniel M. Cohen,
Committee Chairman
Brian V. Brown
Kenneth E. Campbell
Kirk Fitzhugh
Robin A. Simpson, Managing Editor

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

Printed at Allen Press, Inc., Lawrence, Kansas
ISSN 0459-8113

Efficiency of Two Mass Sampling

Methods for Sampling Phorid Flies
(Diptera: Phoridae) in a Trop
Biodiversity Survey

Brian V. Brown1 2 and Donald

ABSTRACT. The period of time necessary for a Malaise trap to collect a given percentage of the susceptible
fauna is calculated for females of Apocephalus, a genus of small phorid flies that parasitize ants. The 16
Malaise traps operated by the La Selva insect survey are expected to collect about 95% of the Apocephalus
fauna in 1 year. Comparison of pan traps with Malaise traps in Costa Rica shows that Malaise traps are
superior for collecting phorid flies; these results contrast with those from a study done in England. Small
pan traps catch relatively more phorids than do large ones. These data provide rough guidelines for
biodiversity surveys of phorid flies in tropical forests: presumably these guidelines will be more reliable
than those based on Northern Hemisphere studies.

INTRODUCTION

As the numerically dominant forms of terrestrial
animal life, insects play key roles in all ecosystems.
The loss of insect species through habitat destruc-
tion is expected to have profound consequences
for other members of natural communities (Wilson,
1987). Therefore, it is vital that we include studies
on insects in all major conservation and biodiversity
survey efforts. Indeed, some surveys are underway
that are dedicated exclusively to the study of insect
diversity (Erwin, 1990; Hammond, 1990; Lamas et
al., 1991; Longino, 1994).

To date, most information about insects has come
from groups that have been well studied in the past,
especially butterflies and larger beetles. Recently,
mass sampling methods have made the study of
many other, less visible groups of insects possible.
The use of canopy fogging, in particular, has been
found to sample beetles effectively (Erwin, 1990;
Stork, 1988). The collecting tools used in this study,
Malaise traps, are tent-like structures that intercept
large numbers of other groups of insects, notably
flies, bees, and wasps.

Currently, little is known about the effectiveness
of Malaise traps. Most collectors of our acquain-
tance use one to four Malaise traps during their
collecting trips and capture an unknown fraction
of the total diversity of a site. An ongoing insect
survey in Costa Rica, the Arthropod Survey of La

1. Entomology Section, Natural History Museum of
Los Angeles County, 900 Exposition Boulevard, Los An-
geles, California 90007 (correspondence).

2. Department of Biology, University of Utah, Salt Lake
City, Utah 84112.

Selva (ALAS; Longino, 1994), uses 16 Malaise traps.
It is not known whether this is an excess of effort
or this number of traps will still vastly undercollect
the total fauna during the period of the survey.

In this paper, we try to determine how much
Malaise trapping is enough. Of course, “enough”
can mean many things, and different goals will dic-
tate when this point is reached. We assume that an
intensive survey (sensu Castri et al., 1992) will re-
quire that 95% of the fauna be identified. Further,
we assume that 10 years is the maximum amount
of time anyone would reasonably sample a single
site. Using species accumulation curves for the group
of insects of interest to us (phorid flies, Diptera:
Phoridae), we predict roughly how long the 16 ALAS
Malaise traps must be operated to collect a given
percentage of the trappable species of Apocephalus,
a genus of ant-parasitizing flies that one of us cur-
rently is revising (Brown, 1993, 1994).

Another question we attempted to address was
whether or not Malaise traps could be replaced by
cheaper, reputedly more efficient collecting devices
like pan traps (=water traps). Disney (1986) pointed
out the need for sampling methods that produce
repeatable results and proposed that mean catches
for white pan traps had a lower variance than those
from Malaise traps, based on an earlier study (Dis-
ney et al., 1982). In this earlier study, however,
Disney et al. (1982) compared only two Malaise
traps, one of which had a highly dysfunctional de-
sign. No useful data are available on whether Mal-
aise trap catches are more or less “repeatable” than
those of pan traps, and the only measure of success
that was compared by Disney et al. (1982) was total
catch. At a site in England, white pan traps were
deemed to be more desirable collecting tools than
Malaise traps because they collected more speci-

Contributions in Science, Number 459, pp. 1-10
Natural History Museum of Los Angeles County, 1995

Sampling Period

Date

Figure 1. Ten-year average rainfall at La Selva (from
Sanford et al., 1994), superimposed on the nine trapping
periods sampled for Malaise traps 8 and 10.

mens and because sorting the catch required less
effort. It is not clear whether or not these results
can be relied upon for other areas, especially in the
tropics. The study of Disney et al. (1982) was a
pioneering effort in this type of analysis, and the
paucity of data from other areas makes it likely that
others will uncritically use these results for design-
ing biodiversity surveys. We replicated the methods
of Disney et al. (1982) in a tropical forest in Costa
Rica to see if the same results would be obtained.

A final question we addressed was one suggested
by Disney et al. (1982) as a further study: whether
a single large pan trap or several small pan traps
were better for sampling phorids. Their data sug-
gested that the number of specimens collected was
approximately proportional to the surface area of
traps. We tested this finding during the Malaise trap
versus pan trap experiment.

METHODS

The catch from 10 Malaise traps operated by the ALAS
project for one sampling period, 1-15 April 1993, was
examined. Also, the catch for two traps (traps 8 and 10)
was analyzed for nine trapping periods, each trapping
period being approximately 2 weeks in duration. This time
span, from 15 February to 1 July 1993, extended through
the dry and early rainy season at La Selva (Fig. 1). All
female Apochephalus specimens were removed from the
samples and identified to morphospecies. Records were
entered into a computer database. Specimens are currently
housed at the Natural History Museum of Los Angeles
County, but most will be returned to the Instituto Na-
tional de Biodiversidad, Costa Rica.

The number of Apocephalus species collected by all
possible combinations of the 10 traps was tabulated. All
two-arrangements, three-arrangements, . . . , to nine-ar-
rangements of the 10 traps were averaged, to find the
mean number of species collected per one trap, two traps,
three traps, . . . , 10 traps. These means were used as
separate observations to build a species accumulation curve

and to estimate the total “Malaise-trappable” fauna (La-
mas et al., 1991; Soberon and Llorente, 1993).

The catch from two individual Malaise traps operated
over 18 weeks was also identified. This allowed us to
examine the interplay between spatial and temporal het-
erogeneity; in other words, do 10 traps operated for 1
week collect the same number of species as one trap
operated for 10 weeks?

Soberon and Llorente (1993) recently reviewed and
justified three models for estimating the rate of species
accumulation and asymptotic species richness based on
repeated sampling of an area. According to these authors,
the exponential model is appropriate for small areas or
those with a well-known fauna that will eventually be
completely collected. This model was clearly inappro-
priate for our data, since La Selva, with its connection to
Braulio Carrillo National Park, certainly is not a small
area, and the phorid fauna is anything but well known.
The Clench model was the second model examined by
Soberon and Llorente (1993). It specifies that as more
time is spent in the field experience will increase the num-
ber of species collected. Our observations, however, are
based on collections made by Malaise traps, which do
not gain experience. We therefore rejected this model as
well. We selected the logarithmic model as the most ap-
propriate model for our Malaise trap data. Soberon and
Llorente (1993) described this model as being appropriate
for large areas with poorly known faunas, where the prob-
ability of adding new species decreases during sampling
but never entirely disappears. A major disadvantage of
the logarithmic model relative to the other two models
is that no asymptote, or estimate of the total size of the
fauna, can be generated. Instead, we only can calculate
the number of species expected at a given point in time.
Thus, we cannot state unequivocally when 95% of the
estimated fauna has been sampled. However, we can make
statements about how rapidly we can collect 95% of the
fauna estimate for a fixed period of sampling (e.g., 10
years).

The logarithmic model takes the form
S(f) = 1/z ln(l + zat),

where 5(f) is the predicted number of species at time t, z
is the slope of the species/sampling-effort curve (a straight
line when plotted on a log/log graph), and a is the list
increase rate at the beginning of the collection period
(Soberon and Llorente, 1993). We used non-linear re-
gression to estimate z and a for our Malaise trap samples.
The NONLIN procedure in SYSTAT Version 5.03 was
used to obtain these estimates (SYSTAT Inc., 1992). Mod-
el evaluation was based on least-squares with quasi-New-
ton optimization (SYSTAT Inc., 1992). Reliable estimates
of z and a were achieved after 7-12 interactions. Once
we obtained estimates of z and a, we used them to predict
5(f) at various times.

For the Malaise trap versus pan trap experiment, we
placed a Malaise trap in the La Selva forest (Clark, 1990;
Hartshorn, 1983) and placed 14 white pan traps in the
surrounding area. The Malaise trap was of the Townes
(1972) design, black, with a white roof, purchased from
D.A. Fochs Company (Gainesville, Fla). The killing and
preservation agent used was 70% ethanol. The pan traps
were white plastic, rectangular refrigerator compartments
and were of three sizes: (1) small: 7.6 cm x 7.6 cm =
58.1 cm2, (2) medium: 22.9 cm x 15.2 cm = 349 cm2,
and (3) large: 38.1 cm x 30.5 cm = 1,161.3 cm2 (large
pans were composed of two 38.1 cm x 15.2 cm pans
joined together). The pans were distributed about 1.5 m
apart in the vicinity of the Malaise trap; the total area

2 ■ Contributions in Science, Number 459

Brown and Feener: Insect Biodiversity Surveys

Figure 2. Cumulative catch of Apocephalus species at
La Selva by means of all combinations of 10 traps sampled
once (top line) and by Malaise traps 8 and 10 over nine
trapping periods.

covered by these pans was 4,765.2 cm2. The following
number of traps were operated: two large, six medium,
and six small. Pan trap fluid was water with several drops
of dishwashing detergent added to lower surface tension.
The traps were operated from 6 to 9 July 1993. An overall
difference in the number of specimens collected by pan
traps of different sizes was assessed by a one-way analysis
of variance. Post-hoc differences in means of small, me-
dium, and large pan traps were detected by the least-
square mean procedure with a Bonferroni adjustment to
ensure an experimentwise error rate of a = 0.05 (Sokal
and Rohlf, 1981).

RESULTS

SPATIOTEMPORAL VARIATION IN
MALAISE TRAP CATCHES

The ten ALAS Malaise traps collected a total of 52
Apocephalus species, ranging from 1 to 19 species
per trap (see the Appendix). The 1,023 possible
combinations of the ten traps produced the results
shown in Table 1. A species accumulation curve
was plotted (Fig. 2, “10 traps”), with the numbers
extrapolated out to over 5,000 days (approximately
15 years; Fig. 3, “10 traps combined”), using the
parameter estimates in Table 2. The extrapolated
numbers of species collected after 1, 2, 3, 5, and
10 years are given in Table 3. In 10 years of sam-
pling, these traps are predicted to catch a total of
164 species of Apocephalus.

The single Malaise traps, 8 and 10, collected
fewer species for the same trap effort than the the
traps together. In nine sampling periods, traps 8 and
10 accumulated 31 and 43 species, respectively,
substantially less than the mean number of species
(49.30 ± 2.54) collected in nine traps for one sam-
pling period. A mean of 16.4 ± 5.55 species per
sampling period was collected by a combination of
traps 8 and 10, compared to 17.24 ± 7.28 for the

Contributions in Science, Number 459

Trap Days

Figure 3. Extrapolated species accumulation curves for
Apocephalus species at La Selva.

mean of 2 of 10 traps sampled during a single period
(Table 1).

Extrapolation of the species accumulation curves
for traps 8 and 10 separately and together over a
duration of 10 years gave predicted species richness
values of 188-278 (Table 3). While these values
appear to be considerably higher than the 164 spe-
cies predicted by the 10 traps sampled simulta-
neously in time, it is perhaps surprising that the
estimates were not even more disparate, given the
fact that we projected the data up to 261 times
beyond the periods sampled.

The higher mean and predicted values of species
richness for the aggregate collecting effort of traps
8 and 10 is almost certainly a product of the in-
creased time span over which these traps operated.
Their trapping period encompassed both wet and
dry seasons, and the collections from these traps
strongly suggest that seasonal changes in abun-

Table 1. Number of Apocephalus species collected by
all combinations of ten Malaise traps.

Average

No. of
traps

No. of
combi-
nations

No. of
species
collected

Range

9.40

6.62

1-19

17.24

7.28

2-31

120

23.89

7.20

4-41

210

29.63

7.76

10-46

252

34.62

6.10

17-48

210

38.99

5.33

25-50

120

42.85

4.48

31-51

46.27

3.56

38-52

49.30

2.54

44-52

52.00

—

Brown and Feener: Insect Biodiversity Surveys ■ 3

Table 2. Parameter estimates (±SE) for the logarithmic model for the accumulation of species during repeated
sampling.

Trap 8

Nine sampling periods
Trap 10

Nine sampling periods
Traps 8 + 10
Nine sampling periods
Ten traps

One sampling period

0.016 ± 0.013
0.010 ± 0.006
0.014 ± 0.004
0.026 ± 0.001

0.331 ± 0.063
0.418 ± 0.050
0.308 ± 0.033
0.737 ± 0.010

dance, species richness, and activity of individual
species are common.

There is some evidence for seasonality in La Selva
Apocephalus , similar to strong seasonality reported
for Apocephalus activity at other tropical sites (Fee-
ner, 1988; Feener and Moss, 1990). Total catch of
Apocephalus specimens (Fig. 4A) and species (Fig.
4B) showed an increase at the beginning of the rainy
season (cf. Fig. 1). Totals for trap 8 increased sooner
than those from trap 10, probably because of dif-
ferences in the microhabitat sampled by each trap.
Similar variation among different habitats has been
shown for other phorids (Disney, 1994, p. 195).
Because most Apocephalus species were represent-
ed by few specimens, most of the variation in num-
bers of specimens can be attributed to a few species,
namely, species 21 and 130 in trap 8 and species
130 and 143 in trap 10 (Fig. 5).

COMPARISON OF WHITE PAN TRAPS
VERSUS MALAISE TRAPS

Over a 3-day sampling period, the Malaise trap
collected a total of 261 phorids, while the com-
bined total catch for all of the pan traps was 182
phorids (Table 4). The number of flies collected by
small, medium, and large pans varied significantly
(F = 9.50, df = 2, 11, P = 0.004). Small pan traps
collected significantly more specimens per unit area
(0.123 ± 0.053 flies/cm2) than did medium traps

(0.047 ± 0.013 flies/cm2; P = 0.012) and large traps
(0.018 ± 0.007 flies/cm2; P = 0.014). Medium pan
traps did not differ significantly from large pan traps
in number of specimens collected per unit area (P
> 0.5).

DISCUSSION

Our data contrast the effect of sampling with 10
traps over one time period versus sampling with
one trap for approximately 10 periods of time. Are
they equal, and can our estimates based on a single
time period be used with any justification?

Single traps, 8 and 10, collected only 63% and
87%, respectively, of the expected number of spe-
cies based on 10 traps sampled once (Fig. 2). We
see this lower number as a reflection of the initial
disadvantage of lower spatial heterogeneity, or few-
er microhabitats, sampled by the two traps, as op-
posed to the larger area sampled by 10 traps. Com-
bining the catches from the two traps gave an av-
erage of 16.3 species per period, a number that is
close to the mean value of 17.24 species collected
by two randomly selected traps in the 10-trap sam-
ple (Table 1).

The single traps predict a larger number of spe-
cies than the 10 traps sampled once (Fig. 3). Pre-
sumably, this is because they sampled a longer pe-
riod of time, including the period of the dry-wet
season rainfall change. Apparently, temporal het-

Table 3. Extrapolated Malaise trap catch (number of species) using the logarithmic model and parameter estimates
in Table 2.

Year

Trap(s)

Trap 8

Nine sampling periods

67.25

98.89

119.80

147.94

188.26

Trap 10

Nine sampling periods

92.65

139.91

171.87

215.51

278.85

Traps 8 + 10
Nine sampling periods

67.53

101.61

124.59

155.90

201.27

Ten traps

One sampling period

79.95

104.13

118.85

137.79

163.91

4 ■ Contributions in Science, Number 459

Brown and Feener: Insect Biodiversity Surveys

Trap 10

Trap Period

Figure 4. A. Number of specimens of Apocephalus collected by traps 8 and 10 during each trapping period. B.
Number of species of Apocephalus collected by traps 8 and 10 during each trapping period.

erogeneity is more important than spatial hetero-
geneity, at least over the time periods we sampled.

All of the projections predict that after 10 years
the number of species of Apocephalus present at
La Selva will exceed 150. This number, about 1.5
times the number of described species in the entire
genus, was much larger than we originally antici-
pated and reflects our true ignorance of the actual
size of the fauna. Obviously, collecting over short
periods of time, such as the 2-week length of our
typical visit to a site, with four Malaise traps will
not intensively sample the La Selva fauna, but what
about the 16 traps that the ALAS survey is currently
using? If we assume that 10 years is the maximum

Contributions in Science, Number 459

amount of time anyone would reasonably sample
a single site, we can predict the amount of time
necessary for an intensive survey. Different data sets
give different answers (Table 5), but generally we
predict that it takes 8-9 years for a single Malaise
trap to collect 95% of the 10-year total at La Selva.
Because there are 16 traps operating, however, we
have to divide the 8-9 years by some constant that
represents the increased catch of 16 traps over a
single trap. Using the accumulation curve for all
traps (Fig. 3), we calculate that 16 traps would catch
an average of 66.25 species, or 7.048 times the
average amount collected by a single trap. There-
fore, if we divide the amount of time predicted by

Brown and Feener: Insect Biodiversity Surveys ■ 5

Trap 8

Trap 10

4 -
2

species 21

i i i r

30 -
25 -

10 4

5 -

species 143

123456789

8
7 -
6 -
5

4 4

species 6

123456789

30
25
20 -
15 -
10
5 4

species 130

123456789

30
25

20 4
15

10 -

123456789

Trap Period

Figure 5. Number of specimens of some common Apocephalus species collected by traps 8 and 10 during each
trapping period.

6 ■ Contributions in Science, Number 459

Brown and Feener: Insect Biodiversity Surveys

Table 4. Number of Phoridae collected by fourteen pan
traps.

Pan trap
No.

Size

No. of
phorids
collected

Phorids
per cm2

Small

0.155

Small

0.189

Small

0.120

Small

0.103

Small

0.138

Small

0.034

Medium

0.029

Medium

0.034

Medium

0.052

Medium

0.063

Medium

0.052

Medium

0.049

Large

0.023

Large

0.013

Mean number of

Size

phorids per cm2

Small

0.123 (SD 0.053)

Medium

0.047 (SD 0.013)

Large

0.018 (SD 0.007)

the various data sets for the traps to collect 95%
of the 10-year total by 7.048, we calculate that an
intensive survey will be completed in 1.14-1.22
years. Naturally, these numbers are rough estimates
and are limited by the limited sampling used for
their calculation. They give some sort of time pe-
riod over which a Malaise trap can be operated for
the specific goal of collecting a given percentage of
Apocephalus females, or of a taxon with a similar
number of species and similar susceptibility to Mal-
aise trap capture.

The comparison of the Malaise and pan traps
showed that the Malaise trap collected 43% more
phorids than did the pan traps, but comparison with
the study of Disney et al. (1982) shows that the
Malaise trap was relatively even more superior. In
one experiment, Disney et al. (1982) found that
three pan traps collected 106 phorids, while the
Malaise trap caught only 15, but the total area of
pan traps was only about 1,845 cm2. Thus, although
we used more than 2.5 times the surface area of
pan traps used by Disney et al. (1982), our Malaise
trap was still superior in number of phorids col-
lected.

A further criticism of Malaise trap samples given
by Disney et al. (1982) was that they collect so many
other insects (non-phorids) that they are tiresome
to sort. This problem has been greatly reduced by
an idea of Dr. Lubomir Masner and the other hy-
menopterists at the Biosystematics Resources Di-
vision in Ottawa, Canada: samples are gently
screened by 0.25-inch and then 8-inch mesh in a
large tub of alcohol. Phorids are found almost ex-

Table 5. Years to collect 95% of the 10-year total at La
Selva, based on different data sets.

Data set

Years
(1 trap)

Years
(16 traps)

Trap 8

Nine sampling periods

8.53

1.21

Trap 10

Nine sampling periods

8.61

1.22

Traps 8 + 10
Nine sampling periods

8.60

1.22

Ten traps

One sampling period

8.05

1.14

clusively in the fraction that goes through the 8-inch
mesh, although the other fractions should be
scanned quickly to find odd specimens stuck to
larger insects. Screening eliminates at least 75% of
the bulk of the sample, making phorid sorting much
easier.

In the comparison of different-sized pan traps,
small pans were found to collect a relatively larger
number of phorid specimens than large and me-
dium pans. Based on these results, a collector should
take many small pans, rather than a few larger pans,
into the field.

SUMMARY

Biodiversity surveys in tropical countries eventually
will be more streamlined and efficient, when we
know more about the effectiveness of our sampling
methods. Guidelines based on studies in tropical
regions presumably will be more reliable for trop-
ical biodiversity studies than guidelines based on
studies from the temperate Northern Hemisphere.

Based on results obtained from Malaise trap sam-
ples at La Selva, sixteen Malaise traps operated for
about 1 year hypothetically should collect about
95% of a 10-year survey of the Apocephalus fauna,
probably satisfying the criteria for an intensive sur-
vey (Castri et al., 1992). Other taxa of similar species
richness and susceptibility to being caught in Mal-
aise traps should require a similar period of trapping
time.

Pan traps are distinctly inferior to Malaise traps
for collecting phorids at La Selva, in contrast to the
preference for pan traps found in a study in En-
gland. Additionally, Malaise traps are much more
convenient, because they require only weekly or
biweekly attention once deployed. Pan traps in the
warm tropics must be emptied daily to prevent rot-
ting of their contents, and new water must be car-
ried to the pans every few days. If pans are to be
used, however, many smaller pans should be used
rather than a few large ones.

ACKNOWLEDGMENTS

We are grateful to R. Colwell, C. Godoy, and J. Longino
of the ALAS project, who allowed us to sort Malaise trap

Contributions in Science, Number 459

Brown and Feener: Insect Biodiversity Surveys ■ 7

samples from their survey, and to Drs. Clark of the Or-
ganization for Tropical Studies (OTS), who permitted work
at La Selva. Long conversations with P. Hanson about
insect sampling using Malaise traps stimulated our interest
in this study. We thank B. Harris, B. Defibaugh, and V.
Brown for technical help. BVB’s trip to Costa Rica was
funded by the OTS Mellon Foundation Research Fund
and a grant to the Natural History Museum of Los An-
geles County from the Weiler Foundation. DHF’s trip
was funded in part by the OTS.

LITERATURE CITED

Brown, B.V. 1993. Taxonomy and preliminary phylog-
eny of the parasitic genus Apocephalus, subgenus
Mesophora (Diptera: Phoridae). Systematic Ento-
mology 18:191-230.

Brown, B.V. 1994. Revision and new species of the
Apocephalus ( Mesophora ) truncaticerus- infragroup
(Diptera: Phoridae). Contributions in Science 449:
1-7.

Castri, F., J.R. Vernhes, and T. Younes. 1992. Inven-
torying and monitoring biodiversity: A proposal for
an international network. Biology International
27(Special Issue):l-28.

Clark, D.B. 1990. La Selva Biological Station: A blue-
print for stimulating tropical research. In Four neo-
tropical rainforests, ed. A.H. Gentry, 9-27. New
Haven and London: Yale University Press.

Disney, R.H.L. 1986. Assessments using invertebrates:
Posing the problem. In Wildlife conservation eval-
uation, ed. M.B. Usher, 271-293. London: Chap-
man and Hall Ltd.

Disney, R.H.L. 1994. Scuttle flies: The Phoridae. Lon-
don: Chapman and Hall, xii + 467 pp.

Disney, R.H.L., Y.Z. Erzinglioglu, D.H.D.J. de C. Hen-
shaw, D.M. Unwin, P. Withers, and A. Woods. 1982.
Collecting methods and the adequacy of attempted
fauna surveys, with reference to the Diptera. Field
Studies 5:607-621.

Erwin, T.L. 1990. Natural history of the carabid beetles
at the BIOLAT Biological Station, Rio Manu, Pak-
itza, Peru. Revista Peruana de Entomologia 33:1—
85.

Feener, D.H., Jr. 1988. Effects of parasites on foraging
and defense behavior of a termitophagous ant, Phei-

dole titanis Wheeler (Hymenoptera:Formicidae).
Behavioral Ecology and Sociobiology 22:421-427.

Feener, D.H., Jr., and K.A.G. Moss. 1990. Defense against
parasites by hitchhikers in leaf-cutting ants: A quan-
titative assessment. Behavioral Ecology and Socio-
biology 26:17-29.

Hammond, P.M. 1990. Insect abundance and diversity
in the Dumoga-Bone National Park, N. Sulawesi,
with special reference to the beetle fauna of lowland
rain forest in the Toraut region. In Insects and the
rain forests of South-East Asia (Wallacea), ed. W.J.
Knight and J.D. Holloway, 197-254. London: Royal
Entomological Society of London.

Hartshorn, G.S. 1983. Plants. In Costa Rican Natural
History, ed. D.H. Janzen, 118-350. Chicago: Uni-
versity of Chicago Press.

Lamas, G., R.K. Robbins, and D.J. Harvey. 1991. A
preliminary survey of the butterfly fauna of Pakitza,
Parque Nacional del Manu, Peru, with an estimate
of its species richness. Publicaciones del Museo de
Historia Natural UNMSM (A) 40:1-19.

Longino,J.T. 1994. How to measure arthropod diversity
in a tropical rainforest. Biology International 28:3-
13.

Sanford, R.L.J., P. Paaby, J.C. Luvall, and E. Phillips.
1994. Climate, geomorphology and aquatic sys-
tems. In La Selva: Ecology and natural history of
a neotropical rain forest, ed. L.A. McDade, K.S.
Bawa, H.A. Hespenheide, and G.S. Hartshorn, 19-
33. Chicago and London: University of Chicago Press.

Soberon, J., and J. Llorente. 1993. The use of species
accumulation functions for the prediction of species
richness. Conservation Biology 7:480-488.

Sokal, R.R., and F.J. Rohlf. 1981. Biometry. New York:
W.H. Freeman & Co.

Stork, N.E. 1988. Insect diversity: Facts, fiction and
speculation. Biological Journal of the Linnean So-
ciety 35:321-337.

SYSTAT Inc. 1992. SYSTAT for Windows: statistics.
Version 5 edition. Evanston, 111. SYSTAT Inc.

Townes, H. 1972. A light-weight Malaise trap. Ento-
mological News 83:239-247.

Wilson, E.O. 1987. The little things that run the world
(The importance and conservation of invertebrates).
Conservation Biology 1:344-346.

Received 16 January 1995; accepted 20 July 1995.

8 ■ Contributions in Science, Number 459

Brown and Feener: Insect Biodiversity Surveys

APPENDIX

Species of Apocephalus collected by Malaise traps at La Selva.

Each species is represented by a number; Malaise trap numbers are those of the ALAS project.

Malaise trap 8

15.ii-l.iii.1993

3061

3073

3168

M5.iii.1993

3180

3194

3202

1 5 . iii- 1 . iv . 1 993
3022
3144
3180
3202
3211

1-15. iv. 1993
160
910
3022
3073
3105
3144

3148
3167
3177

3191

3192
3208
3211
3213
3222

15.iv-l.v.l993

910

2701

3022

3068

3105

3149
3162
3167
3180
3202

1-15. v. 1993
110
3068
3105
3172
3177
3180
3222

15.v-l.vi.1993

110

152

157

3022

3166

3167
3180
3194
3211

1-15. vi. 1993
76
160
3022
3068
3180
3211

15.vi-l .vii. 1993
76
110
152
3022
3102
3144

3168
3180

Malaise trap 10

15.ii-l.iii.1993

2701

3022

3148

3180

3227

l-15.iii.1993

152

3022

3103

3148

3180

3188

3202

15.iii-l.iv.1993

2701

3022

3061

3103

3147

3162

3167

3180

3198

3205

3210

1-15. iv. 1993
160
3022
3166
3192
3218
3222

15.iv-l.v.l993

152

2701

3105

3149

3167

3168
3174
3180
3192
3198
3203
3208

3211
3225
3227

1-15. v. 1993
110
121
152
3022
3061
3103
3105
3148
3160
3167

3169
3173
3180
3196
3198
3202
3208

3212
3222
3227

15.v-l.vi.1993

152

3022

3151

3166

3171

3180

3192

3196

3197

3198
3203
3208
3218
3225
3227

1-15. vi. 1993
152
2701
3022
3103
3105
3147
3171
3180
3192
3208
3211
3218
3227

15.vi-l .vii. 1993
152
160
2041
3022
3089
3166
3171
3173
3180
3190
3192
3198
3208
3211
3213
3227

Contributions in Science, Number 459

Brown and Feener: Insect Biodiversity Surveys ■ 9

Appendix continued

Malaise trap 3

1-15. iv. 1993
3105
3167

Malaise trap 4

1-15. iv. 1993
96
157
2701
3022
3151
3167
3173
3211
3226

Malaise trap 5

1-15. iv. 1993
110
121
157
2041

3089

3090
3151
3189
3194

3204

3205
3211

3217

3218

Malaise trap 7

1-15. iv. 1993
3171
3201

Malaise trap 11

1-15. iv. 1993
160
910
3068

3089
3177
3180

3189

3190
3194

Malaise trap 12

1-15. iv. 1993
160
910
2041
3022

3090
3103

3147

3148
3162
3167
3171
3173
3192
3198
3209
3220
3222

Malaise trap 15

1-15. iv. 1993
3022
3061
3090

3147

3148
3167
3169
3171
3180
3188
3193
3198

3202

3203
3205
3215
3222

3225

3226

Malaise trap 16

1-15. iv. 1993
3167

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

L-'oZX

A ) H Number 460

25 March 1996

Contributions
in Science

A New Species of Giant Anhinga
(Aves: Pelecaniformes: Anhingidae) from the
Upper Miocene (Huayquerian) of
Amazonian Peru

Kenneth E. Campbell, Jr.

Natural History Museum of Los Angeles County

Serial
Publications
of THE
Natural History
Museum of
Los Angeles
County

Scientific

Publications

Committee

Tod F. Stuessy, Deputy Director
for Research and Collections
John M. Harris, Committee Chairman
Brian V. Brown
Kenneth E. Campbell
Kirk Fitzhugh
Karen Wise

Robin A. Simpson, Managing Editor

• Contributions in Science, a miscellaneous series of tech-
nical papers describing original research in the life and
earth sciences.

• Science Series, long articles and collections of papers
on natural history topics.

Copies of the publications in these series are sold through
the Museum Book Shop. A catalog is available on request.

Printed at Allen Press, Inc., Lawrence, Kansas
ISSN 0459-8113

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

A New Species of Giant Anhinga

(Aves: Pelecaniformes: Anhingidae) from the
Upper Miocene (Huayquerian) of
Amazonian Peru

Kenneth E. Campbell, Jr.1

ABSTRACT. A new species of giant anhinga is described from Upper Miocene (Huayquerian) deposits
of southeastern Peru, the first species of fossil bird to be described from the Tertiary of lowland Amazonia.
In addition to the holotype tarsometatarsus, three partial humeri, the distal end of an ulna, a tibiotarsus,
and two cervical vertebrae are referred to the new species. The new species is slightly larger than
Meganhinga chilensis from Chile, but it is smaller than Macranhinga paranensis from Argentina, both
recently described Miocene paleospecies of giant anhingas. The hindlimb of the new species is approx-
imately 65 per cent larger than that of modern A. anhinga, but the wing appears to be only about 25
per cent larger. The size of the wing relative to that of the hindlimb appears to be a highly variable feature
of anhingas of the Americas.

INTRODUCTION

Anhingas are water birds of the suborder Sulae,
order Pelecaniformes. Primarily predators on fish,
they are excellent underwater swimmers, but they
are also good fliers and roost easily in trees. An-
hingas are commonly known as darters or snake-
birds, the latter in reference to the snake-like man-
ner in which they hold and move their head while
swimming partially submerged.

The family Anhingidae is one of four families
comprising the suborder Sulae, the other three be-
ing the Sulidae, the Phalacrocoracidae, and the ex-
tinct, marine family Plotopteridae (Olson, 1980,
1985). Occasionally, the family Anhingidae has been
reduced in rank to subfamilial status (e.g., Dorst
and Mougin, 1979), but this effort has had little
support (Olson, 1985; Becker, 1986). The fossil rec-
ord of anhingas has been reviewed by Olson (1985),
Becker (1986, 1987), Rasmussen and Kay (1992),
and Alvarenga (1995). Until recently, most paleo-
species of anhingas were known from Europe and
North America, but a recent series of finds, includ-
ing those reported by Wall et al. (1991), Rasmussen
and Kay (1992), Noriega (1992), and Alvarenga
(1995) and those described in this paper, have ex-
panded our knowledge of Tertiary anhingas of South
America considerably. These new discoveries do
not yet reveal much about the phylogenetic rela-
tionships of anhingas, but they do demonstrate a

1. Natural History Museum of Los Angeles County,
900 Exposition Boulevard, Los Angeles, California 90007.

Contributions in Science, Number 460, pp. 1-9
Natural History Museum of Los Angeles County, 1996

surprising diversity of large-bodied anhingas during
the Miocene in South America.

MATERIALS AND METHODS

Anatomical terminology is principally that of Baumel et
al. (1979). Measurements accurate to 0.01 mm were taken
with vernier dial calipers; they were then rounded one
decimal point. Measurements from other sources were
also rounded to one decimal point if originally given to
more than one. Osteological comparisons were made with
six modern specimens of Anhinga anhinga (Linnaeus
1766), two of A. novaehollandiae (Gould 1847), and one
of A. rufa (Daudin 1802). Comparisons were also made
with the holotype of the early Miocene Meganhinga chi-
lensis Alvarenga 1995 and a cast of the holotype of the
late Miocene Macranhinga paranensis Noriega 1992.

SYSTEMATICS

Order Pelecaniformes Sharpe 1891
Suborder Sulae Sharpe 1891
Family Anhingidae Ridgway 1887
Genus Anhinga Brisson 1760

DISCUSSION. I have chosen not to erect a new
genus for the species described below because I see
no characters that justify such action. In fact, I
doubt that the recently described genera Megan-
hinga Alvarenga 1995 and Macranhinga Noriega
1992 are sufficiently distinct from Anhinga to war-
rant recognition. Alvarenga (1995) based his genus
primarily on two features of the tarsometatarsus: a
prominent hypotarsis and a wide proximal me-

taphysis. Both of these characters would be ex-
pected in a larger version of a foot-propelled diving
bird and they do not, in themselves, justify the
recognition of a separate genus. Alvarenga (1995)
was also influenced in his decision by the fact that
the wing was very small relative to the rest of the
bird, possibly indicating flightlessness. As noted be-
low, however, such variation between fore- and
hindlimbs may be common in anhingas. Perhaps
this is a situation analogous to that seen in cor-
morants, where one species only of Phalacrocorax
is flightless, whereas the remaining species are vo-
lant.

Noriega (1992:220) erected Macranhinga pri-
marily on the basis of size, which is not a generic
character. The other characters he listed are more
readily interpreted as specific, not generic, charac-
ters. Unfortunately, other available elements of
Macranhinga paranensis were not described by
Noriega (1992, 1995), so we do not know if their
characters would support the establishment of the
genus. Until it can be documented that Meganhinga
and Macranhinga are valid genera, I consider them
to be junior synonyms of Anhinga.

Anhinga fraileyi new species
Figures 1, 2

HOLOTYPE. Right tarsometatarsus; proximal
end damaged, lacking hypotarsus; LACM 135356.

TYPE LOCALITY. LACM 4611; Acre VI,
southwest bank of Rio Acre, about 0.5 km down-
stream from the confluence of the Rio de Los Patos,
Departamento de Madre de Dios, Peru; approxi-
mately 69°55'41"W, 10°56T5"S.

TYPE HORIZON AND AGE. Acre Conglom-
erate of Campbell et al. (1985); Upper Miocene
(Huayquerian). This unit was originally described
as the Acre Conglomerate Member of the upper
Pleistocene Inapari Formation (ONERN, 1977) [=
Madre de Dios Formation of Oppenheim (1946)
(Campbell and Romero-P., 1989) = Iga Formation
of Maia et al. (1977)]. When this horizon was de-
scribed, the fossils in it were thought to have been
redeposited from Tertiary [Upper Miocene (Huay-
querian)] deposits in the Pleistocene, as also sug-
gested by Simpson and Paula Couto (1981). Recent
advances in our understanding of the geology of
the region now lead me to consider this unit and
its contained fossils to be in situ Upper Miocene
(Huayquerian) deposits (see also Kay and Frailey,
1992). Assignment of the fossils from the Acre Con-
glomerate to the Huayquerian South American Land
Mammal Age is based on the presence in the fauna
of such characteristic Huayquerian taxa as Kiyu-
therium orient alis Frances and Mones 1965, Tetra-
stylus sp. (Pascual et al., 1966), and possibly Po-
tamarchus murinus Burmeister 1885 (Frailey, 1986).

DIAGNOSIS. The holotypical tarsometatarsus
can be distinguished from all living species of An-
hinga by its large size. It differs from that of An-
hinga anhinga 1) by being approximately 65 per

cent larger and by having 2) eminentia intercon-
dylaris more prominent, 3) shaft with medial and
lateral sides much less excavated immediately distal
to cotylae, 4) shaft with medial side not expanded
mediad at the fossa metatarsal I, 5) shaft lacking
small, ridge-like projection on anterior face im-
mediately distal to cotyla medialis, and 6) shaft
lacking sharp corner or ridge on interno-medial
portion leading to base of trochlea metatarsi II.

The holotypical tarsometatarsus differs from that
of Anhinga chilensis (Alvarenga 1995) by having 1)
eminentia intercondylaris more pronounced (not
much elevated above hypotarsus, in external view,
in A. chilensis ); 2) cotyla medialis smaller, less pro-
nounced anteriad; 3) cotyla lateralis smaller, sloping
more steeply anteriad; 4) shaft wider in anterior
view, narrowing less distal to cotylae and widening
more as it approaches trochlea metatarsi II; 5) shaft
of similar depth, but wider; 6) shaft less excavated
between proximal lateral ridge and hypotarsus, in
medial view, with proximal lateral ridge meeting
distal end of crista plantaris mediana of hypotarsus
at low angle (much higher angle in A. chilensis ) and
proximal lateral ridge meeting distal end of crista
plantaris mediana about 45 per cent of the shaft
length downshaft (about 33 per cent in A. chilensis );
7) shaft with area between lateral ridge and postero-
external ridge less excavated, in external view, with
postero-external ridge less prominent and located
more mediad; 8) shaft, in medial view, not curving
as far craniad to meet lip of cotyla medialis; 9)
trochlea metatarsi III with central canal more pro-
nounced, sides smaller and less rounded, and ex-
tending less anteriad, i.e., shorter overall. In addi-
tion, in A. chilensis, in medial view, the distal lateral
ridge is rotated slightly posteriad, suggesting that
trochlea metatarsi III, which is missing from the
holotype, was positioned more posteriorly.

The holotypical tarsometatarsus differs from that
of Anhinga paranensis (Noriega 1992) by having
1) eminentia intercondylaris narrower, but more
prominent, with external side nearly vertical (does
not approach vertical in A. paranensis ); 2) both
cotyla medialis and cotyla lateralis smaller, with
latter sloping less steeply anteriad; 3) shaft, in me-
dial view, curves farther anteriad to meet anterior
lip of cotyla medialis; 4) hypotarsus missing, but
crista hypotarsi medialis a low ridge from level of
foramina vascularia proximalia medialis distad
(much more prominent ridge in A. paranensis ); 5)
shaft with crista lateralis of plantar side less prom-
inent; 6) trochlea metatarsi tertii turned slightly lat-
erad, in anterior view, but very nearly parallel to
axis of shaft (turned mediad in A. paranensis, at a
slight angle to axis of shaft); 7) trochlea metatarsi
II narrower, with medial side at low angle to axis
of shaft, and extending distad beyond trochlea
metatarsi III (broad, turned much more mediad at
greater angle to axis of shaft, and not extending
distad beyond trochlea metatarsi III in A. paranen-
sis ); 8) trochlea metatarsi II with “wing” missing,
but proximal portion of trochlea metatarsi II much

2 ■ Contributions in Science, Number 460

Campbell: Giant Anhinga from Amazonia

Figure 1. Specimens of Anhinga fraileyi new species, including the holotype tarsometatarsus (LACM 135356), in
cranial (A) and caudal (B) views; the referred left tibiotarsus (LACM 135357), in cranial (C) and caudal (D) views; the
referred 19th cervical vertebra (LACM 135359), in left lateral (E) and ventral (F) views; and the referred 18th vertebra
(LACM 135358), in left lateral (G) and ventral (H) views.

smaller than in A. paranensis, suggesting “wing”
much smaller than in latter; 9) trochlea metatarsi
IV narrower, more rounded distally in lateral view.
Although the trochleae of A. fraileyi appear to be
farther apart than in A. paranensis, this effect may
be a result of postmortem wear.

The holotypical tarsometatarsus differs from that
referred to Anhinga grandis by Becker (1987) by
its 1) larger size, 2) broader midshaft region, 3) more
prominent eminentia intercondylaris, and 4) a distal
crista plantaris mediana that extends farther distad.

Contributions in Science, Number 460

MEASUREMENTS. See Table 1.

REFERRED MATERIAL. From type locality
LACM 4611: distal end and shaft of right humerus
(LACM 135360), distal end of left ulna (LACM
135361), complete left tibiotarsus, condylus me-
dialis broken (LACM 135357); cervical vertebra #18
(LACM 135358); cervical vertebra #19 (LACM
135359). From locality LACM 5158: proximal end
and shaft of left humerus (LACM 135362), shaft
of right humerus (LACM 135363).

ETYMOLOGY. Patronymic, in honor of Dr. Carl

Campbell: Giant Anhinga from Amazonia ■ 3

Figure 2. Specimens referred to Anhinga fraileyi new species. Proximal end and shaft of left humerus (LACM 135362),
in caudal (A) and cranial (B) views, and the distal end of left ulna (LACM 135361), in dorsal (C) and caudal (D) views.

David Frailey, in recognition of his contributions
to Amazonian paleontology.

DESCRIPTION. Humerus. The humeri are re-
ferred to the family Anhingidae on the basis of
those characters listed by Becker (1986). These
specimens are referred to Anhinga fraileyi and dif-
fer from A. anhinga, by having 1) larger size; 2)
tuberculum dorsale projecting less prominently from
shaft; 3) margo caudalis a distinct ridge confluent
with attachment of coracohumeralis ligamentum
(not confluent in A. anhinga ); 4) fossa pneumotri-
cipitalis and impressio M. coracobrachialis cranialis
shallower; 5) attachment of M. pectoralis undivid-
ed (divided into proximal and distal portions in A.
anhinga ); 6) sulcus ligamentosus transversus pro-

portionately deeper ventrally, undercutting intu-
mescentia; and 7) distal end with processus flexorius
more prominently developed, bordering a more
deeply excavated olecranon fossa. The distal hu-
merus (LACM 135360) is too badly worn to reveal
any other clear characters. Interestingly, this spec-
imen carries paired tooth marks on both the shaft
and distal end. This may be taken as evidence of
predation, although it could also represent post-
mortem gnawing on the bone.

In spite of being badly worn, the distal humerus
can be seen to differ from the holotype of Anhinga
grandis Martin and Mengel 1975 from the Upper
Miocene of North America by having the condylus
dorsalis as deeply undercut, or more, as in A. an-

4 ■ Contributions in Science, Number 460

Campbell: Giant Anhinga from Amazonia

hinga ; the processus flexorius larger and more
prominent, projecting farther distad; and the epi-
condylus dorsalis less developed and extending less
proximad. The proximal humerus differs from that
referred to A. grandis by Becker (1987) by having
a more acute angle between the sulcus ligamentosus

transversus and the impressio M. coracobrachialis
cranialis, a shallower impressio M. coracobrachial-
is, and a less excavated fossa pneumotricipitalis.

For measurements, see Table 2.

Ulna. The ulna is placed to the family Anhingidae
and differs from those of cormorants by having the

Table 1. Measurements (mm) of the holotype tarsometatarsus of Anhinga fraileyi in comparison with those of A.
anhinga (n - 4), A. grandis, A. chilensis, and A. paranensis.

Measurements

Anhinga

anhinga

Anhinga
anhinga 1

Anhinga

fraileyi

Anhinga

grandis1

Anhinga
chilensis 1

Anhinga

paranensis

Total length

39.5-41.8
x = 41.2

41.5-43.5
x = 42.5

68.5

47.8

61.0

75.3 (75.5)

Distal width, excluding
“wing” of trochlea
metatarsi II

12.5-14.4
x = 13.5

22.0

16.5

25.7

Midshaft width

6. 1-6.9
x = 6.5

6.7-7 A
x - 7.1

11.9

7.8

10.0

12.5

Midshaft depth

3. 6-4.7
x = 4.0

4.5-5.0
x = 4.7

7.3

4.9

7.3

Proximal width

10.7-12.5
x = 11.2

11.8-12.7
x = 12.3

19.5 ± 0.5

12.8

18.0

21.1 (21.5)

Width trochlea metatarsi III

4.4-4. 9
x = 4.6

5.0-5.4
x - 5.2

8.0

—

8.0

9.6

1 This set of measurements for Anhinga anhinga (n = 5) and those for Anhinga chilensis from Alvarenga (1995).

2 Measurements for Anhinga grandis from Becker (1987).

3 Measurements from a cast. Figures in parentheses from Noriega (1992).

Table 2. Measurements (mm) of the humeri of Anhinga fraileyi, A. grandis, and A. anhinga.

Measurement

LACM

135362

Anhinga fraileyi

LACM

135360

LACM

135363

A. grandis 1

A. anhinga

A. anhinga 1

Proximal width

22.3

—

23.1

17.3-19.1
x = 18.2
(n = 4)

17.2-19.8
x = 18.0

Depth of caput humeri

7.5

8.0

62-7.2
x = 6.7
(n = 4)

6. 1-7.1
x = 6.7

Length of crista
deltopectoralis

44.6

42.3

32.6-37.6
x = 34.5
(n = 4)

31.7-37.8
x = 35.2

Midshaft width

8.3

9.1

8.7

5. 8-7.1
x = 6.5
(n = 5)

5. 7-7.1
x = 6.7

Midshaft depth

7.4

8.3

7.1

7.7

5. 3-6.2
x = 5.8
(n = 5)

5. 1-6.2
x = 5.8

Distal width

15.7 ± 2.0

13.6-16.2
x = 14.7
n = 5

1 Measurements from Becker (1986); n = 10 for Anhinga anhinga.

Contributions in Science, Number 460

Campbell: Giant Anhinga from Amazonia ■ 5

condylus ventralis less protruding distad, the tu-
berculum carpale more set off from the shaft prox-
imally, and the condylus dorsalis not elevated above
the shaft proximally, in ventral view. The ulna dif-
fers from that of Anhinga anhinga by having 1)
condylus ventralis less prominently protruding dis-
tad, 2) tuberculum carpale more massive, and 3)
condylus dorsalis extending farther proximad ven-
trally and by lacking 4) prominent papillae remi-
giales caudales.

This distal ulna is very close in size to the Middle
Miocene specimen from Colombia referred to An -
hinga sp. cf. A. grandis by Rasmussen and Kay
(1992). It differs from that specimen by having the
tuberculum carpale more set off from the shaft and
the ventral edge of the condylus ventralis projecting
distad less prominently.

For measurements, see Table 3.

Tibiotarsus. The tibiotarsus is placed to the fam-
ily Anhingidae and differs from those of cormorants
by its less prominent crista cnemialis lateralis and
its more centrally placed canalis extensorius. The
tibiotarsus differs from that of A. anhinga by having
1) crista patellaris less covered medially by area of
attachment of Lig. patellae; 2) crista cnemialis cra-
nialis not extending as far proximad as the crista
patellaris (the junction of the crista patellaris and
crista cnemialis cranialis is a prominent, proximad-
pointing projection in A. anhinga ); 3) crista cne-
mialis lateralis with cranial end roughly triangular
in proximo-lateral view, with apex pointing pos-
tered (linear in A. anhinga ); 4) shaft leading to
crista cnemialis lateralis much more massive; 5) fos-
sa flexoria less deeply excavated; posterior lip of
facies articularis medialis very thick, rounded, and
only slightly excavated distad; 6) groove for pero-
neus profundus not as distinct, especially proximad
(not as distinctly set off laterally in A. anhinga );
and 7) concavity at the antero-proximal end of con-
dylus lateralis linear and moderately deep, appear-
ing as a sharp cleft (a deep, oval pit in A. anhinga ).

A larger size distinguishes tibiotarsus LACM
135357 from that referred to Anhinga grandis by
Becker (1987). For measurements, see Table 4.

Cervical Vertebrae. Two water-worn cervical
vertebrae were found at the type locality. If it is
reasonable to assume that the same general verte-
bral form would have existed in Anhinga fraileyi
as in A. anhinga, these three vertebrae would fall
within numbers 17-19 of the cervical series. In A.
anhinga these cervical vertebrae (cv) have a broad,
relatively flat or slightly concave ventral surface with
a small central spine or ridge. In cv #17 the ventral
surface is significantly longer than it is wide, but in
cv #18 and cv #19 the ventral surface is approxi-
mately as wide as it is long, as is the case in the
fossil specimens.

In Anhinga anhinga, cv #18 is characterized by
a fairly sizable incisura in the side of the vertebral
wall immediately dorsal to the facies articularis cau-
dalis. The dorsal edge of this incisura leads in a
straight line to the base of the facies articularis of
the postzygapophysis. LACM 135358 has a com-
parable notch. Cv #19 also has an incisura in the
corresponding position, but it is much reduced in
size. This latter condition is observed in LACM
135359.

The ventral surface of cv #18 of Anhinga an-
hinga is slightly concave, but with two small lateral
ridges in addition to a central ridge with a caudal
spine. The lateral ridges are at an angle to the long
axis of the bone, approaching the midline more
craniad. The ventral surface is more excavated, or
concave, lateral to these ridges. A corresponding
condition is seen in LACM 135358.

The ventral surface of cv #19 of Anhinga an-
hinga lacks the two lateral ridges seen in cv #18,
and it has a narrower, more prominent central ridge.
Instead of the two lateral ridges, the ventral surfaces
lateral to the medial ridge have an excavated, or
concave, surface. Specimen LACM 135359 differs
by lacking a central ridge, but it does have a broad,
roughly triangular, elevated platform that narrows
craniad.

In Anhinga anhinga, cv #19 is the first in a series
of vertebrae extending caudad that has a foramen
through the vertebral wall into the foramen ver-
tebrale immediately craniad of the dorsal extension

Table 3. Measurements (mm) of ulnae of Anhinga fraileyi, A. anhinga (n = 4), and A. sp. cf. A. grandis.

Measurement

Anhinga

fraileyi

LACM 135361

Anhinga anhinga

A. anhinga 1

Anhinga sp. cf.
A. grandis 1

Maximum distal width

11.1

10.1-10.9
x = 10.5

x= 10.6

12.4

Depth condylus dorsalis ulnaris

8.9

7.8

x = 7.8

x = 7.7

10.0

Shaft width proximal to condylus
dorsalis ulnaris

6.7

63-6.8
x - 6.4

x = 6.2

7.7

Shaft depth proximal to condylus
dorsalis

6.7

4.7-5. 1
x = 5.0

x = 4.8

5.3

1 Measurements from Rasmussen and Kay (1992); n = 8 for Anhinga anhinga.

6 ■ Contributions in Science, Number 460

Campbell: Giant Anhinga from Amazonia

Table 4. Measurements (mm) of the tibiotarsus referred
to Anhinga fraileyi in comparison to those of A. anhinga

(n = 4).

Measurements

Anhinga

anhinga

Anhinga

fraileyi

LACM 135357

Total length

85.1-90.3
x = 87.0

143 ± 1.0

Total length to

81.7-85.5

136.2 ± 0.5

eminentia intercotylaris

x = 83.1

Proximal width

9.5-11.2
x = 10.4

17.0 ± 1.0

Proximal depth

11.7-18.9
x = 13.9

20.5 ± 0.5

Midshaft width

4.6-5.8
x - 5.4

9.5

Midshaft depth

4.0-4.6
x = 4.3

7.2

of the facies articularis caudalis. A corresponding
foramen is present in LACM 135359, but it is lack-
ing in LACM 135358. Cv #19 also has facets for
the articulation of a rib, as does the last cervical
vertebra (#20) (Garrod, 1876), but cv #18 does not.
Both of the fossil vertebrae lack articular facets for
ribs.

In Anhinga anhinga, the facies articularis cra-
nialis of cv #19 extends farther dorsad than it does
in cv #18, to partially extend around the foramen
vertebrale. In the former, there is also a small, but
prominent, protuberance immediately dorsal to the
tips of the facies articularis cranialis and caudal to
the small notch. These features also occur in LACM
135359.

The two fossil vertebrae appear to agree with the
18th and 19th cervical vertebrae (LACM 135358
and LACM 135359, respectively) of Anhinga an-
hinga in the characters that are determinable from
these well-worn specimens. One obvious differ-
ence, however, is the lack of articular facets for
ribs on LACM 135359. Nonetheless, I am inclined
to refer to LACM 135359 as a cv #19.

For measurements, see Table 5.

DISCUSSION

The two Old World Tertiary species of anhingas
not mentioned in the descriptions above, Anhinga

pannonica Lambrecht 1916 and A. hadarensis
Brodkorb and Mourer-Chauvire 1982, may be dis-
tinguished from A. fraileyi on the basis of their
smaller size. The only other purported Tertiary spe-
cies of anhinga, Protoplotus beauforti Lambrecht
1931, is quite small and may not even be an anhinga
(van Tets et al, 1989). The Tertiary species of an-
hingas were reviewed by Alvarenga (1995).

Measurements of the various fossil bones as-
signed to the several species of anhingas known
from the Americas suggest that limb proportions
among the fossil and living species were highly vari-
able. For example, the proximal humerus of An-
hinga fraileyi is approximately the same size as that
assigned to A. grandis by Becker (1986), whereas
the tarsometatarsus of the latter is only 70 per cent
the length of the former. Similarly, the distal ulna
referred to A. fraileyi is slightly smaller than that
referred to Anhinga sp. cf. A. grandis by Rasmus-
sen and Kay (1992) and only slightly larger than
that of A. anhinga. Anhinga chilensis was de-
scribed as a probably flightless anhinga because its
referred ulna and carpometacarpus were smaller
than those of A. anhinga, but the holotype tarso-
metatarsus was about 45 per cent larger than that
of A. anhinga (Alvarenga, 1995).

If the wing : leg proportions of Anhinga anhinga
are considered to be “normal” for anhingas, then
A. grandis had larger than normal wings relative to
its legs than the former. On the other hand, A.
chilensis had very small wings relative to its legs,
whereas the wings of A. fraileyi were smaller than
normal relative to the size of its legs. It can be
assumed that these differences among the fore- and
hindlimbs had some significance in the overall func-
tional attributes of these species, but more material
is needed of each species before any conclusions
pertaining to that subject can be drawn. It must
also be noted that although A. chilensis is known
from an associated skeleton, the wing elements of
A. grandis and A. fraileyi are only referred to those
species; they have yet to be found in association.

ASSOCIATED FAUNA. The paleofauna from
the type locality for Anhinga fraileyi, LACM 4611,
is the most diverse known from the Amazon Basin.
It contains both a sizable megafauna (Frailey, 1986)
and the first, and most diverse, vertebrate micro-
fauna yet known from Amazonia. The most abun-
dant fossils in terms of numbers of specimens are
those of fish, with representatives of at least nine

Table 5. Measurements (mm) of cervical vertebrae of Anhinga anhinga (n — 3) and those referred to A. fraileyi.

Anhinga anhinga

Anhinga fraileyi

Measurement

Cervical
Vertebra #18

Cervical
Vertebra #19

LACM

135358

LACM

135359

Width of facies articularis cranialis

9.7-10.9
x = 10.2

9.3-10.5
x = 9.9

15.2

16.5

Length along midline from facies articularis cranialis
to facies articularis caudalus

10.0-11.76
x = 10.6

10.2-11.9
x= 11.1

16.6

15.5

Contributions in Science, Number 460

Campbell: Giant Anhinga from Amazonia ■ 7

families included. Perhaps the most interesting
within this group are sharks, rays, piranhas, and
lungfish, the latter reaching sizes several times larger
than those of modern lungfish of Amazonia.

The largest vertebrate from the site is the giant
crocodylian Purussaurus brasiliensis Barbosa Ro-
drigues 1892 (Campbell and Frailey, 1992), which
is represented at this site by several vertebrae and
numerous isolated teeth. Other reptiles present in-
clude snakes, lizards, turtles, and other species of
crocodylians. In addition to the variety of mammals
described by Frailey (1986), dolphins, toxodonts,
marsupials, primates (Kay and Frailey, 1992), a fish-
eating bat (Czaplewski, 1996), and possibly as many
as a dozen species of several genera of micro-ro-
dents were found at this site.

The overwhelming majority of the vertebrate
species present at LACM 4611 indicates an aquatic
habitat, which is in keeping with the habitat of
modern anhingas. The only certainly terrestrial ver-
tebrates represented at the site are a few species of
mammals. The habitat preferences of the micro-
rodents are still unknown.

A similar but less diverse paleofauna is known
from Cachuela Bandeira, Bolivia, the second lo-
cality (LACM 5158) producing bones of Anhinga
fraileyi.

SUMMARY

A new species of anhinga, Anhinga fraileyi, whose
hindlimb was approximately 65 per cent larger than
the modern A. anhinga, is described from Upper
Miocene (Huayquerian) deposits of Amazonian
Peru. In addition to the holotype tarsometatarsus,
three partial humeri, a distal ulna, a tibiotarsus, and
a cv #18 and #19 are referred to the new species.
The referred wing elements suggest that A. fraileyi
had a smaller wing relative to its hindlimb than
does A. anhinga. Limb proportions of other pa-
leospecies of anhingas from the Americas also ap-
pear to be quite different from those of A. anhinga.

This is the third large paleospecies of anhinga to
be described from the Miocene of South America,
and still additional, undescribed species of anhingas
are known from the Miocene of Argentina (No-
riega, 1995). This is the first avian paleospecies to
be described from the Tertiary of Amazonia.

ACKNOWLEDGMENTS

I thank F. Hertel for his critical comments on earlier drafts
of this manuscript and for access to comparative materials
at the University of California, Los Angeles; H. Alvarenga
for access to and photographs of the specimens of An-
hinga chilensis’, N. Johnson for access to collections at
the University of California, Berkeley, Museum of Com-
parative Zoology; and L. Martin and S. Olson for their
critical reviews of the manuscript. I am grateful to J.
Noriega, Museo de La Plata, Argentina, for providing a
cast of the holotype of A. paranensis. Fieldwork in Ama-
zonia was funded by NSF DEB 78-05861 and NSF BSR
84-20012 and accomplished, in part, in collaboration with
the Universidade Federal do Acre, Brazil.

LITERATURE CITED

Alvarenga, H.M.F. 1995. A large and probably flightless
anhinga from the Miocene of Chile. Courier For-
schungsinstitut Senckenberg 181:149-161.

Barbosa Rodrigues, J. 1892. Les reptiles de la vallee de
I’Amazone. Vellosia, Contribuicoes do Museu Bo-
tanico do Amazonas 2:41-56.

Baumel, J.J., A.S. King, A.M. Lucas, J.E. Breazile, and
H.E. Evans. 1979. Nomina Anatomica Avium.
London: Academic Press, 637 pp.

Becker, J.J. 1986. Re-identification of “P^/acrocorax”
subvolans Brodkorb as the earliest record of An-
hingidae. The Auk 103:804-808.

Becker, J.J. 1987. Additional material of Anhinga gran-
dis Martin and Mengel (Aves: Anhingidae) from the
late Miocene of Florida. Proceedings of the Biolog-
ical Society of Washington 100(2):358-363.

Brodkorb, P., and C. Mourer-Chauvire. 1982. Fossil
anhingas (Aves: Anhingidae) from Early Man sites
of Hadar and Omo (Ethiopia) and Olduvai Gorge
(Tanzania). Geobios 15(4):505-515.

Burmeister, H.G. 1885. Examen critico de los mami-
feros y reptiles fosiles denominados por D. Augusto
BRAVARD y mencionados en su obra precedente.
Anales del Museo Nacional de Buenos Aires 3(14):
95-174.

Campbell, K., and C.D. Frailey. 1992. Uncovering the
mysteries of the Amazon. Vertebrate paleontology
reveals creatures great and small. Terra 29(3/4):36-
49.

Campbell, K.E., Jr., C.D. Frailey, and J. Arellano-L. 1985.
The geology of the Rio Beni: Further evidence for
Holocene flooding in Amazonia. Contributions in
Science 364:1-18.

Campbell, K.E., Jr., and L. Romero-P. 1989. Geologia
del cuaternario del Departamento de Madre de Dios
[Peru]. Sociedad Geologica del Peru, Boletin 79:53-
61.

Czaplewski, N.J. 1996. Opossums (Didelphidae) and bats
(Noctilionidae and Molossidae) from the late Mio-
cene of the Amazon Basin. Journal of Mammalogy
77(l):84-94.

Dorst, J., and J.L. Mougin. 1979. Order Pelecaniformes.
In Checklist of birds of the world, vol. 1, 2nd ed.,
ed. E. Mayr and G.W. Cottrell, 155-193. Cam-
bridge, Mass.: Museum of Comparative Zoology.

Frailey, C.D. 1986. Late Miocene and Holocene mam-
mals, exclusive of the Notoungulata, of the Rio Acre
region, western Amazonia. Contributions in Science
374:1-46.

Francis, J.C., and A. Mones. 1965. Sobre el hallazgo de
Cardiatherium talicei n. sp. (Rodentia, Hydrochoe-
ridae) en Playa Kiyu, Dept, de San Jose, Republica
Oriental de Uruguay. Kraglieviana l(l):3-44.

Garrod, A.H. 1876. Notes on the anatomy of Plotus
anhinga. Proceedings of the Zoological Society of
London 1876:335-345.

Kay, R., and C.D. Frailey. 1992. Large fossil platyrhines
from the Rio Acre local fauna, late Miocene, west-
ern Amazonia. Journal of Human Evolution 25:319-
327.

Lambrecht, K. 1916. Die Gattung Plotus im ungarischen
Neogen. Mitteilungen aus dem Jahrbuch Unga-
rischen Geologischen Anstalt 24:1-24.

Lambrecht, K. 1931. Protoplotus beauforti n.g.n.sp., en
Schlangenhalsvogel aus dem Tertiar von W. Suma-
tra. Wetenschappelijke Mededeelingen Dienst van
den Mijnbouw. in Nederlandsch-In-die 17:15-24.

8 ■ Contributions in Science, Number 460

Campbell: Giant Anhinga from Amazonia

Maia, R.G.N., H. Godoy, H. Yamaguti, P. Moura, F.
Costa, M. Holanda, and J. Costa. 1977. Projecto
Carvao no Alto Solimoes: Relatoria final. Manaus,
Brazil: Departmento Nacional da Produ^ao Mi-
neral/Companhia de Pesquisa de Recursos Miner-
als.

Martin, L., and R.M. Mengel. 1975. A new species of
anhinga (Anhingidae) from the upper Pliocene of
Nebraska. The Auk 92:137-140.

Noriega, J. 1992. Un nuevo genero de Anhingidae (Aves:
Pelecaniformes) de la formacion Ituzaingo (mioceno
superior) de Argentina. Notas del Museo de La Pla-
ta, Tomo XXI, Paleontologta, vol. 109, 217-223.

Noriega, J. 1995. The avifauna from the “Mesopota-
mian” (Ituzaingo Formation; Upper Miocene) of
Entre Rios Province, Argentina. Courier
Forschungsinstitut Senckenberg 181:141-148.

Olson, S.L. 1980. A new genus of penguin-like pele-
caniform bird from the Oligocene of Washington
(Pelecaniformes: Plotopteridae). Contributions in
Science 330:51-57.

Olson, S.L. 1985. The fossil record of birds. In Avian
biology, vol. 8, ed. D.S. Farner, J.R. King, and K.C.
Parkes, 79-252. New York: Academic Press.

ONERN (Oficina Nacional de Evaluacion de Recursos
Naturales). 1977. Inventario, evaluacion e inte-
gracion de los recursos naturales de la zona Iberia-
Ihapari. Lima, Peru, 334 pp.

Oppenheim, V. 1946. Geological reconnaissance in
southeastern Pern. American Association of Petro-
leum Geologists, Bulletin 30:254-264.

Pascual, R., N.V. Cattoi, J.C. Francis, O. Gondar, E. Or-
tega Hinojosa, E. Tonni, J.A. Pisano, A.G. de Rin-

guelet, and J. Zetti. 1966. Vertebrata. In Paleon-
tografta Bonaerense, ed. A.V. Borello, fasciculo 4,
202 pp. Provincia de Buenos Aires, Comision de
Investigation Cientifica, La Plata.

Rasmussen, D.T., and R.F. Kay. 1992. A Miocene an-
hinga from Colombia, and comments on the zoo-
geographic relationships of South America’s Tertiary
avifauna. In Papers in Avian Paleontology honoring
Pierce Brodkorb, ed. K.E. Campbell, Science Series,
Natural History Museum of Los Angeles County,
number 36, 225-230.

Simpson, G.G., and C. Paula Couto. 1981. Fossil mam-
mals from the Cenozic of Acre, Brazil, part 3, Pleis-
tocene Edentata Pilosa, Proboscidea, Sirenia, Pe-
rissodactyla and Artiodactyla. Iheringia, serie Geo-
logia 6:11-73.

van Tets, G.F., P.V. Rich, and H.R. Marino-Hadiwardojo.
1989. A reappraisal of Protoplotus beauforti from
the early Tertiary of Sumatra and the basis of a new
pelecaniform family. Publication of the Geological
Research and Development Centre, Ministry of Mines
and Energy, Republic of Indonesia, Paleontology
Series vol. 5, 57-75.

Wall, R., H.M.F. Alvarenga, L.G. Marshall, and P. Salinas.
1991. Hallazgo del primer ave fosil del terciario de
Chile: Un anade (Pelecaniformes: Anhingidae), pre-
servado en un ambiente deltaico-fluvial del mioceno
de Lonquimay, region de la Araucania, Chile. Re-
sumenes Expandidos del Congreso Geologico Chi-
leno 5:394-397.

Received 9 March 1995; accepted 5 December 1995.

Contributions in Science, Number 460

Campbell: Giant Anhinga from Amazonia ■ 9

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

A/ Number 461
8 July 1996

Contributions
in Science

A New Genus and Species of
Checkered Beetle From Honduras
with Additions to the
Honduran Fauna (Coleoptera: Cleridae)

Jacques Rifkind

Natural History Museum of Los Angeles County

Serial
Publications
ol THE
Natural History
Museum of
Los Angeles
County

Scientific

Publications

Committee

Tod F. Stuessy, Deputy Director
for Research and Collections
John M. Harris, Committee Chairman
Brian V. Brown
Kenneth E. Campbell
Kirk Fitzhugh
Karen Wise

Robin A. Simpson, Managing Editor

The scientific publications of the Natural History Museum
of Los Angeles County have been issued at irregular in-
tervals in three major series; the issues in each series are
numbered individually, and numbers run consecutively, re-
gardless of the subject matter.

# Contributions in Science, a miscellaneous series of tech-
nical papers describing original research in the life and
earth sciences.

# Science Series, long articles and collections of papers on
natural history topics.

Copies of the publications in these series are sold through
the Museum Book Shop. A catalog is available on request.

The Museum also publishes Technical Reports, a miscel-
laneous series containing information relative to scholarly
inquiry and collections but not reporting the results of
original research. Issue is authorized by the Museum’s Sci-
entific Publications Committee; however, manuscripts do
not receive anonymous peer review. Individual Technical
Reports may be obtained from the relevant Section of the
Museum.

Printed at Allen Press, Inc., Lawrence, Kansas
ISSN 0459-8113

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

A New Genus and Species of

Checkered Beetle From Honduras

with Additions to the

Honduran Fauna (Coleoptera: Cleridae)

Jacques Rifkind1

ABSTRACT. Barrotillus kropotkini, a new genus and species of Honduran tilline, is described and illus-
trated; a key and a comparative table of characters are provided to facilitate discrimination of the new
taxon from other members of its tribe in Honduras. New records for clerids in Honduras are presented.

RESUMEN. Se describe e illustra Barrotillus kropotkini, un genero y especie nuevos de Tillini hondurena;
se proveen una clave y una tabla comparativa de sus caracteristicas para facilitar la diferenciacion de la
nueva entidad taxonomica con respecto a los otros miembros de su tribu en Honduras. Se presentan nuevos
registros referentes a los clerids en Honduras.

INTRODUCTION

The clerid fauna of Honduras, the second largest
Central American country, is almost unknown.
Barr’s (1975) checklist recorded only two Hondur-
an clerid species: Cregya quadrisignata (Spinola
1844) and Lebasiella bisbinotata Gorham 1883.
Later, Ekis (1976, 1977) published Honduran re-
cords of Perilypus orthopleuridus (Thomson 1860)
and Colyphus cylindricus (Gorham 1878) in two
papers revising their respective genera. Most re-
cently, two new clerids have been described from
Honduras: the enopliine Parapelonides beckeri Barr
(1980:281) and the tilline Cymatoderella morula
Rifkind (1993:282).

Honduras is topographically varied, comprising
a diverse range of tropical, subtropical, and boreal
plant communities (Monroe, 1968; Campbell and
Lamar, 1989). It should therefore be expected to
host a correspondingly varied array of clerid spe-
cies, like its neighbor Guatemala. As in the case of
Belize, discussed elsewhere (Rifkind, 1995), Hon-
duras has suffered from neglect by coleopterists.
Distributions for many clerids are given as “Mexi-
co, Guatemala, Costa Rica, Panama” and it is
clearly not a lack of suitable habitat that excludes
Honduras from this north to south geographical se-
quence; these gaps in distribution are the result of
little or no collecting in the intermediate countries.
Selander and Vaurie (1962), compiling the known
collecting localities for insect specimens included in
the Biologia Centrali-Americana, were able to lo-
cate only six Honduran locations, as compared

1. Research Associate, Natural History Museum of Los
Angeles County, 900 Exposition Boulevard, Los Angeles,
California 90007.

Contributions in Science, Number 461, pp. 1-10
Natural History Museum of Los Angeles County, 1996

with 128 for Guatemala. They concluded that “on
an area basis, the least-collected of the Central
American countries is Honduras.”

Recent collecting has produced new records for
several clerid species in Honduras. The purpose of
this paper is to make these records available and to
describe a new genus and species also brought to
light by that collecting.

In addition to the taxa listed below, I examined
Honduran specimens of the genera Cymatodera
Gray 1832, Colyphus Spinola 1841, Aphelocerus
Kirsch 1870, Pbyllobaenus Dejean 1837, Phlogis-
tosternus Wolcott 1944, Ellipotoma Spinola 1844,
Epiphlceus Spinola 1841, Ichnea Castelnau 1836,
Cregya LeConte 1861, and Pelonium Spinola 1844
that were not determinable to species or that rep-
resent undescribed species.

METHODS AND MATERIALS

In describing pronotal characteristics, I use the term
“neck” to denote the broad area behind the basal constric-
tion. This is done to avoid confusion with the area known
as the “pronotal collar,” which properly refers to the usu-
ally narrow constriction at the pronotal basal margin in
clerids.

Specimens discussed below are deposited in the insti-
tutional and individual collections abbreviated as follows:
EAPZ (Escuela Agricola Panamericana Zamorano, Tegu-
cigalpa, Honduras); FSCA (Florida State Collection of Ar-
thropods, Gainesville, FL); LACM (Natural History Mu-
seum of Los Angeles County, Los Angeles, CA); FWSC
(Fred W. Skillman Jr., Deland, FL); JNRC (Jacques Rif-
kind, North Hollywood, CA); WFBC (William F. Barr,
Moscow, ID); RFMC (Roy F. Morris, Lakeland, FL); and
RHTC (Robert H. Turnbow Jr., Enterprise, AL).

Unless otherwise noted, the following account of Hon-
duran localities for clerids represents new country records
for the listed taxa. Numbers in parentheses indicate the

number of specimens collected at a given location. Data
presented in brackets are translated from the Spanish.

Suprageneric classification follows Barr (1975).

SYSTEMATICS
Subfamily CLERINAE
Tribe Tillini

A key to the tribe Tillini in Honduras is provided
to facilitate the identification of the new genus de-
scribed below. The Tillini are characterized by hav-
ing the first tarsomere distinctly visible from above,
the fourth tarsomere approximately equal in size to
the third, the prothorax not margined laterally, and
the anterior coxal cavities separated internally and
closed behind.

1. Antenna with fewer than 11 antennomeres

— Antenna with 11 antennomeres 3

2(1). Antenna of male with 8 or 10 antennomeres,

antenna of female with 9 or 10 antennomer-
es; terminal antennomere elongate, flattened,
and spatulate; elytra at most indistinctly fas-
ciate Monophylla Spinola 1841

— Antenna of both sexes with 10 antennomeres;

terminal antennomere not as above; elytra
with eburneous markings

Callotillus Wolcott 1911

3(1). Antennomere 11 of male double bean-like in
shape (bifabaceoid); pronotum campanulate;
elytra very finely punctate, with raised ebur-
neous markings Barrotillus n. gen.

Not as above 4

4(3). Eyes finely granulate; elytra uniformly black
or bluish black with surface coarsely punc-
tate; small sized and robust

Cymatoderella Barr 1962

— Eyes coarsely granulate; color not as above;

size, shape, and sculpturing variable

Cymatodera Gray 1832

Callotillus e. elegans

(Erichson 1847)

(1) Departamento Atlantida, La Ceiba, June 6,
1991, R. D. Cave, coll.

Barrotillus new genus

Figures 1-3

DESCRIPTION. Tillini; small sized; body elon-
gate, subparallel, moderately convex.

Head. Eyes medium sized, rather prominent,
moderately finely faceted, shallowly emarginate at
antennal insertion; clypeolabral area somewhat
narrowed and ventrally produced; labrum trans-
verse; antenna (Figs. 1,2) rather loosely composed
(8) or somewhat more compact ( 9 ), comprised of
1 1 antennomeres as follows: 1 elongate, more than
2X as long as 2, expanded at middle; 2 subglobu-

1 2

Figures 1-3. Barrotillus : 1, Male antenna. 2, Female an-
tenna. 3, Pronotum.

lar; 3-4 subcylindrical, 3 slightly longer than 4; 5-
10 broad, serrate, moderately flattened in cross sec-
tion, 5 a little longer than 6, 7 slightly broader than
6 and 8, 9-10 subequal in length and breadth; 11

(8) elongate, more than 3x the length of 10, mod-
erately compressed, feebly sinuately margined on
one side near middle, more deeply constricted on
the other at basal 2/s, giving the antennomere an
asymmetrical bifabaceoid (double bean-like) shape,
apex slightly tapered and obtusely rounded, or 11

(9) ovate-elongate, about 3X the length of 10,
moderately compressed, sides not constricted, apex
narrowly rounded; last maxillary palpomere elon-
gate-conical, apex acute; last labial palpomere ex-
panded, securiform.

Pronotum (Fig. 3). Campanulate, approximately
1.4X as long as wide, moderately convex; sides
weakly sinuate and subparallel to basal %, then
strongly, obliquely narrowed to basal neck; disk,
posteriorly, sloping acutely to basal neck; basal
neck elongate, convex behind, where it is divided
at middle by a shallow antescutellar impression;

2 ■ Contributions in Science, Number 461

Rifkind: Honduran Cleridae

Table t. Comparison of diagnostic characters of Barrotillus with those of similar genera of New World Tillini: Callo-
tillus Wolcott 1911, Cymatoderella Barr 1962, Onycbotillus Chapin 1945, and Bogcia Barr 1978.

Barrotillus n. gen. Callotillus Wolcott Cymatoderella Barr Onycbotillus Chapin Bogcia Barr

11 antennomeres
Antennomere 11 of 6
bifabaceoid

Pronotum campanu-
late; neck deeply
constricted and
elongate

Pronotal disk with
posterior slope
acute to neck

Elytra finely punctate;
punctures not ar-
ranged in distinct
striae

Small sized

10 antennomeres

Antennomere 10 of A
ovate-elongate

Pronotum scutiform;
neck moderately
narrowed and short

Pronotal disk more or
less plane to collar

Elytra finely punctate;
punctures not ar-
ranged in distinct
striae

Small sized

11 antennomeres
Antennomere 11 ovate

Pronotum expanded
laterally at middle

Pronotal disk with
posterior slope shal-
low and oblique to
collar

Elytra with deep,
coarse punctures ar-
ranged in striae

Small sized

11 antennomeres
Antennomere 11 of A
ovate-elongate; lat-
eral margins slightly
sinuate

Pronotum expanded
laterally at middle

Pronotal disk more or
less plane to collar

Elytra with deep,
coarse punctures ar-
ranged in striae

Small sized

11 antennomeres
Antennomere 11 of S
bifabaceoid (in
some undescribed
species)

Pronotum expanded
laterally at middle

Pronotal disk more or
less plane to collar

Elytra with deep,
coarse punctures ar-
ranged in striae

Medium to large sized

sides of neck weakly, obliquely expanded to elytral
base.

Scutellum. Moderate in size, rounded apically.

Elytra. Elongate (ratio of width to length 18:43),
convex posteriorly; disk with subbasal area some-
what swollen on either side of suture but without
tubercles, latitudinally depressed at middle, giving
the elytra a shallow saddle shape in lateral view;
sides subparallel from humeri to apical Vs (feebly
sinuate at basal Vs), then broadly, arcuately rounded
to separately rounded apices.

Prostereum. With coxal cavities closed behind.

Mesosternum. With anterior margin transverse
and costate, bordered behind by another, slightly
broader, transverse costa.

Metasternum. Strongly convex posteriorly.

Legs. Femora rather narrow; tibiae feebly ex-
panded apically; tarsomere 4 moderately expanded
laterally, approximately equal in size to 3; all tarsal
claws bearing on each side a narrow, elongate tooth
proximate to the primary claw and a short, trian-
gular basal tooth.

Abdomen. Moderately convex, with 6 visible
sternites.

TYPE SPECIES. Barrotillus kropotkini new spe-
cies.

ETYMOLOGY. This interesting genus is named
for William F. Barr, Emeritus Professor at the Uni-
versity of Idaho, in recognition of his lifelong con-
tributions to the systematics of New World Cleri-
dae.

DIAGNOSIS. Barrotillus is distinguished from
other clerid genera on the basis of a unique com-
bination of antennal composition (11 antenno-
meres), the shape of antennomere 11 in the male,
the shape of the pronotum, and the sculpturing of

the elytra. The new taxon’s small, shining, ant-like
form and convex, elongate-necked campanulate
pronotum recall Stenocylidrus Spinola 1844, a ge-
nus confined to East Africa, Madagascar, and a few
small Indian Ocean islands. Several other Old
World tilline genera have the pronotum campanu-
late, including the Indo Australian and Afro-Mal-
agasian genus Cladiscus Chevrolat 1843, the pri-
marily Malagasian Pseudopallenis Kuwert 1893,
and the South African Eburneocladiscus Pic 1954.
This last genus possesses, in addition, a raised ebur-
neous elytral fascia. Barrotillus is distinct from
these allopatric genera in antennal structure and
from both Cladiscus and Eburneocladiscus (whose
species are much larger), at least, in palpal mor-
phology as well.

Table 1 is presented to facilitate the separation
of Barrotillus from the other New World tilline
genera that it most closely resembles. Among these,
Callotillus Wolcott appears most similar to Barro-
tillus. Possible synapomorphies are: 1) small size;
2) raised eburneous elytral fascia; and 3) finely,
non-strially punctate elytral surface. Barrotillus is
distinct from Callotillus in having 11 antenno-
meres, in the shape of the terminal antennomere, in
having the pronotal slope acute, and in the shape
of the pronotum. It should be noted here that
Mawdsley (in lift.) points out the possibility that
the phylogenetic affinities of the new genus may lie
with some Malagasian til lines rather than with the
strictly New World Callotillus .

Barrotillus kropotkini new species

Figure 4

DESCRIPTION OF HOLOTYPE MALE. Color.
Deep pitchy black except antennomere 1, ocular

Contributions in Science, Number 461

Rifkind: Honduran Cleridae ■ 3

Figure 4. Barrotillus kropotkini. Habitus.

emargination, clypeus, labrum, mandibles, men-
tum, basal maxillary palpomere, sides of pronotum,
dorsum of pronotal neck (with the exception of a
black macula on the antescutellar impression), a
small oblong marking laterally behind umbones,
and most of the pro- and mesosterna brownish red;
antennomeres 5-11 dull reddish brown; each ely-
tron with a raised ivory subbasal macula and me-
dian fascia, arranged as in Figure 4; median fascia
attaining lateral margin.

Head. Measured across eyes, wider than prono-
tum; surface shining, deeply, coarsely, and subcon-
fluently punctate, sparsely clothed with long, erect
dark setae; antennomeres 1-4 shining, sparsely
punctate, sparsely setose, antennomeres 5-11 with
surface microgranularly roughened, rather densely
covered with minute grayish setae interspersed with
a few longer setae.

Pronotum. Surface shining, sculpturing as on
head, punctations becoming obsolete basally on
pronotal slope, pronotal neck smooth above; ves-
titure moderately sparse, consisting mostly of long,
erect, dark and pale, anteriorly directed setae inter-
spersed with fewer shorter, reclinate setae. Scutel-
lum thickly clothed with silvery setae.

Elytra. Surface shining, smooth, finely, sparsely
punctate, punctures somewhat coarser basally
where they form indistinct striae extending to basal
Vs; ivory postbasal maculae and midelytral fasciae
glabrous and slightly swollen above elytral surface;
vestiture sparsely, uniformly arranged, comprised
of mostly suberect, dark and pale setae with a
slightly denser concentration of longer, erect, dark
setae at base.

Legs. Profemora finely, moderately densely punc-
tate, meso- and metafemora shining, more coarsely

4 ■ Contributions in Science, Number 461

Rifkind: Honduran Cleridae

but sparsely punctate; tibiae transversely rugulose;
all surfaces rather sparsely vested with pale, erect
setae of varying lengths, setae more densely ar-
ranged on tibiae.

Metasternum. Surface shining, sparsely punctate
laterally, transversely rugulose anteriorly, otherwise
smooth.

Abdomen. Surface shining, sparsely, shallowly,
and indistinctly punctulate, very sparsely covered
with short, pale, mostly suberect setae; sternite 6
with hind angles arcuate, apex narrowly, rather
deeply notched at middle.

Body length 3.50 mm.

TYPE SERIES. Holotype <5 (LACM), HONDU-
RAS, (Department) Francisco Morazan, Tegucigal-
pa, El Rincon, October 5, 1988, R. D. Cave, coll.
Paratypes: 2 d, same data as holotype; 1 9 same
data as holotype except October 15, 1993, F. W.
Skillman Jr., coll. Paratypes in EAPZ, FSCA, and
JNRC.

VARIATION. Aside from the dimorphic anten-
nal characterisics delineated above under the ge-
neric description, the female differs from the male
by having abdominal sternite 6 with the hind mar-
gin complete, rather than notched.

Specimens vary in the extent of brownish red in-
tegumental coloration; the female, for example, has
the basal Vi of the elytra (except the maculae) and
the thoracic sternites uniformly this color. Male
specimens on hand range from 3.40-4.0 mm in
length. The female paratype is 4.10 mm long.

ETYMOLOGY. This species is named in honor
of Prince Peter Alekseyevich Kropotkin (1842-
1921), Russian social critic, economist, geographer,
theorist of the evolutionary basis of cooperation,
revolutionist, and anarchist philosopher.

DISTRIBUTION. The type series was collected
at a single locality (El Rincon) in the environs of
Tegucigalpa, Honduras, at an elevation of 1650 m.
According to Cave (pers. comm.), the habitat here
is disturbed second growth forest, dominated by
small leguminous trees ( Mimosa spp.) with small
broad-leafed trees and forbs intermixed. This area
probably supported a mixed oak forest before it
was altered by agriculture. El Rincon is situated in
the Honduran interior highlands, a geologically an-
cient area that has been above sea level since before
the Mesozoic (Monroe, 1968). It is possible that B.
kropotkini is narrowly endemic to one or more of
the ranges in the area; further systematic collecting
is needed to resolve this question.

HABITS. One example was collected by beating
a non-leguminous shrub or small tree (Skillman,
pers. comm.). Barrotillus kropotkini, with its shiny,
elongate facies and pale midelytral fascia, is most
probably an ant mimic. The presence of a pale ely-
tral fascia presumably imparts an impression of
myrmecoid segmentation to the beetle, thus deter-
ring some visually hunting predators. Although
many clerids exhibit this type of coloration, the el-
evation of the eburneous fascia onto a smooth ridge
is less common; among the New World Tillini, for

example, the only other apparent occurrence is in
Callotillus. This structure also appears in other
families, however, as in the cerambycid genus Eu-
derces LeConte 1850, which is often collected in
association with ants.

Monopbylla pallipes
Schaeffer 1908

(3) Departamento Francisco Morazan, Tegucigalpa
vie., El Sitio, 3100', May 24, 1993, beating burned
Acacia in Thorn-Scrub, J. Rifkind, P. Gum, colls.

Cymatodera depauperata grp.
Gorham 1882

(1) Departamento Francisco Morazan, San Antonio
de Oriente, Uyuca, July 14, 1993, R. Cordero, coll.

Cymatodera guatemalensis
Schenkling 1900

(1) Departamento Francisco Morazan, 32 km E Te-
gucigalpa, El Zamorano, June 14, 1986, D. Hidal-
go, coll.; (1) Departamento Francisco Morazan,
25.5 km SE Talanga, Finca Archaga, June 3, 1993,
beating burn, F. W. Skillman Jr., coll.; (1) Depar-
tamento Choluteca, 19.5 km E Choluteca, Villa
Guadelupe, June 5, 1993, beating roadside vegeta-
tion, F. W. Skillman Jr., coll.

Cymatodera prolixa
(Klug 1842)

(1) Departamento Cortes, Lago de Yojoa, Motel
Agua Azul, May 30, 1993, beating tree branches,
R. Turnbow, coll.; (1) Departamento Santa Bar-
bara, 3 km W Horconcitos, May 30, 1995, R.
Turnbow, coll.

Cymatodera sallei
Thomson 1860

(1) Departamento Santa Barbara, Coffee Institute,
La Fe, May 30, 1993, beating live vegetation, F. W.
Skillman Jr., coll.

Tribe Clerini

Priocera abdominalis
Blanchard 1844

(3) Departamento Francisco Morazan, 23 km S
Talanga, Finca Archaga, June 10, 1993, R. D.
Cave, coll.; (1) Departamento Francisco Morazan,
25.5 km SE Talanga, Finca Archaga, June 3, 1993,
F. W. Skillman Jr., coll.; (1) same data as last except
M. C. Thomas, coll.; (1) same data except June 5,
1993.

Priocera salamandra
Schenkling 1906

(3) Departamento Olancho, Parque Nacional La
Muralla, May 25-June 1, 1995, R. Turnbow, coll.

Contributions in Science, Number 461

Rifkind: Honduran Cleridae ■ 5

Priocera stictica

Gorham 1882

(1) Departamento El Paraiso, Danli, El Bordo, May
29, 1988, R. D. Cave, coll.; (6) Departamento At-
lantida, Tela, Lancetilla Botanical Gardens, May
28, 1993, M. C. Thomas, coll.

Colyphus cylindricus

New department records. (1) Departamento
Francisco Morazan, Parque Nacional La Tigra,
June 1, 1993, W. Morjan, coll.; (1) same data as
last except M. C. Thomas, coll.; (2) same data ex-
cept 5800', beating vegetation in cloud forest, F. W.
Skillman Jr., coll.

Perilypus distinctus

(Chevrolat 1874)

(1) Departamento Santa Barbara, Coffee Institute,
La Fe, May 30, 1993, M. C. Thomas, coll.

Perilypus frontalis

(Gorham 1886)

(3) Departamento Santa Barbara, Coffee Institute,
La Fe, May 30, 1993, beating live vegetation, F. W.
Skillman Jr., coll.; (1) Departamento Santa Barbara,
vie. Zacapa, May 30, 1993, M. C. Thomas, coll.;

(2) Departamento Comay agua, SW corner Lago de
Yojoa, 1.2 km W Pito Solo, 2000', May 30, 1993,
Broadleaf Hardwood Forest, beating shrubs, J. Rif-
kind, P. Gum, colls.; (1) Departamento Francisco
Morazan, Parque Nacional La Tigra, June 1, 1993,
F. W. Skillman Jr., coll.; (33) Departamento Olan-
cho, Parque Nacional La Muralla, May 25-27,
1995, R. Turnbow, coll.

Enoclerus (Coniferoclerus) arachnodes

(Klug 1842)

(1) Departamento Francisco Morazan, 32 km E Te-
gucigalpa, El Zamorano, M. Intriaso, coll.

Enoclerus (E.) ablusus

Barr 1978

(1) Departamento Lempira, Belen, October 6,
1993, F. W. Skillman Jr., coll.; (6) Departamento
Santa Barbara, vie. Lago de Yojoa, 10 km NW Pito
Solo, 2000', May 26, 1993, Pine-Oak Forest, beat-
ing dead pine, J. Rifkind, P. Gum, colls.

Enoclerus (E.) beatus

(Gorham 1882)

(3) Departamento Francisco Morazan, 1 km W Ha-
tilla, May 29, 1995, R. Turnbow, coll.; (1) Depar-
tamento Olancho, Parque Nacional La Muralla,
June 1, 1995, R. Turnbow, coll.

Enoclerus (£.) bicarinatus

(Gorham 1882)

(1) Departamento Olancho, Parque Nacional La
Muralla, May 26, 1995, R. Turnbow, coll.

Enoclerus (E.) fugitivus

Wolcott 1927

(2) Departamento El Paraiso, 7 km S Danli, Apa-
quis Mts., El Portillo, October 12, 1993, F. W. Skill-
man Jr., coll.

One of these specimens is an example of the
black, and the other of the reddish “morph,” both
commonly seen in this species.

Enoclerus (E.) gibbus

Ekis 1976

(1) Departamento Santa Barbara, vicinity Lago de
Yojoa, above El Mochito, El Cidral, September 9,
1984, C. W. O’Brien, coll.; (1) Departamento Fran-
cisco Morazan, San Antonio de Oriente, Uyuca,
May 20, 1993, R. Ortega, coll.; (1) same data as
last except March 5-12, 1990, R. Cave, coll.; (1)
Departamento El Paraiso, Yuscaran, Cerro Mont-
serrat, May 25, 1993, R. Ortega, coll.; (1) Depar-
tamento Francisco Morazan, Tegucigalpa vie., Ju-
tiapa (near Parque Nacional La Tigra), 5200', beat-
ing dead tree branches along stream, May 23,
1993, J. Rifkind, P. Gum, colls.; (46) Departamento
Olancho, Parque Nacional La Muralla, May 25-
June 1, 1995, R. Turnbow, coll.

Ekis described E. (£.) gibbus from a small series
of all black Costa Rican and Panamanian speci-
mens, remarking that these individuals “did not
show any noteworthy chromatic . . . variation”
(1976:161). It is thus of interest that some of the
Honduran examples of this species exhibit a largely
brownish-red integument; only the head, anten-
nomeres 4-11, midbasal tubercle, apical Vi of ely-
tra, abdomen, and legs (in part) remain darkened
or black. Structurally these specimens agree with
the original description. The typical black form also
occurs in Honduras.

Enoclerus (£.) boegei

(Gorham 1882)

(1) Departamento El Paraiso, vie. Yuscaran, Octo-
ber 16, 1993, F. W. Skillman Jr., coll.; (1) Depar-
tamento Francisco Morazan, San Antonio de Or-
iente, El Zamorano, August 21, 1991, M. Grena-
dino, coll.

This species has been collected from as far north
as Sinaloa, Mexico, south into El Salvador and now
Honduras. Although it appears to be rather uni-
form across its range in terms of punctation, sculp-
turing, and setation, it exhibits what seems to be a
clinal variation in the shape of its distinctive red
elytral marking. Honduran and Salvadoran speci-
mens have this midelytral marking in the form of a
broad fascia, somewhat obliquely narrowed toward
the suture at the anterior margin, more or less
transverse at the posterior margin, but never inter-
rupted at the suture. The red marking in most Mex-
ican specimens, on the other hand, is comparatively
narrower (with the posterior margin placed more
toward the middle of the elytra) and is interrupted

6 ■ Contributions in Science, Number 461

Rifkind: Honduran Cleridae

before the suture. One specimen from Chiapas
seems transitional between these forms, having the
fascia narrow as in typical Mexican specimens, but
not interrupted at the suture.

Enoclerus (E.) irregularis
Barr 1978

New department record. (6) Departamento Co-
pan, 19 km SW Santa Rosa de Copan, October 8,
1993, F. W. Skillman, coll.

These specimens differ from those of the type se-
ries in having the integumental color of the head,
pronotum, sterna, and forelegs reddish rather than
brownish. The elytral base is also more uniformly
darkened than in the type. In addition, Barr’s
(1978a) original description of E. (£.) irregularis
gives the metasternum as “finely, densely punctate.”
The Copan specimens, however, have the metaster-
num more or less distinctly tuberculate, especially
toward the midline. The reason for this discrepancy
is apparently that the types are paper-point mount-
ed with the contact point at, and thus obscuring,
part of the metasternum. The presence of a tuber-
culate metasternum in E. (E.) irregularis provides
another good character for distinguishing it from
its congeners, particularly those, like E. (£.) mexi-
canus (Castelnau 1836) which are similarly pat-
terned. This type of metasternal sculpturing is rare
in Enoclerus ; I have previously seen it only in spec-
imens of E. (£.) longipes (Schenkling 1907).

Enoclerus (E.) mexicanus

(1) Departamento Olancho, Juticalpa, Candeleros,
May 21, 1988, [on leafless Psidium guaiava ], R.
Cordero, coll.; (1) Departamento Francisco Mora-
zan, 32 km E Tegucigalpa, El Zamorano, July 7,
1986, M. Sanchez, coll.; (1) same data except May
4, 1989, R Monge, coll.; (1) Departamento Fran-
cisco Morazan, 5 km W El Zamorano, June 2,
1993, picking burn at night, F. W. Skillman Jr.,
coll.; (3) Departamento Francisco Morazan, 25.5
km SSW Talanga, June 3, 1993, M. C. Thomas,
coll.; (3) same data except R. Turnbow, coll.; (3)
Departamento El Paraiso, Yuscaran, Cerro Mont-
serrat, January-March 1993, R. Ortega, coll.; (2)
same data except November 16-31, 1992.

Enoclerus (£.) nigromaculatus
(Chevrolat 1843)

(3) Departamento Francisco Morazan, 32 km E Te-
gucigalpa, El Zamorano, June 2-July 7, 1990,
[malaise trap under Inga sp. in coffee plantation],
R. Cave, coll.; (1) Departamento Francisco Mora-
zan, San Antonio de Oriente, San Juan del Rancho,
July 8, 1992, [on Zea mays ], R. Cordero, coll.

Enoclerus (£.) opifex
(Gorham 1882)

(1) Departamento Francisco Morazan, 32 km E Te-
gucigalpa, El Zamorano, April 1986, D. Vivanco,

coll.; (2) Departamento Francisco Morazan, 5 km
W El Zamorano, May 25-June 2, 1993, F. W. Skill-
man Jr., coll.; (4) Departamento Francisco Mora-
zan, 5 km E Escuela Agricola Panamericana, June
2, 1993, M. C. Thomas, coll.

Enoclerus (£.) pilatei
(Chevrolat 1874)

(1) Departamento El Paraiso, Cerro Apaguis, Dan-
li, February 20, 1988, R. D. Cave, coll.

Enoclerus (£.) salvini
(Gorham 1876)

(3) Departamento Francisco Morazan, Tegucigalpa,
El Rincon, August 13, 1988, R. D. Cave, coll.; (4)
same data except October 15, 1993, F. W. Skillman
Jr., coll.; (1) Departamento El Paraiso, Montserrat,
5750', October 3, 1993, F. W. Skillman Jr., coll.

Enoclerus (£.) Venator
(Chevrolat 1843)

(1) Departamento Francisco Morazan, 30 km E Te-
gucigalpa, Escuela Agricola Panamericana, May
10, 1984, Galvis, coll.

Enoclerus (£.) zebra
(Chevrolat 1843)

(1) Departamento Francisco Morazan, 32 km E Te-
gucigalpa, El Zamorano, June 19-25, 1990, [taken
in malaise trap under Inga sp. in coffee plantation],
R. D. Cave, coll.; (1) Departamento Santa Barbara,
Coffee Institute, La Fe, May 30, 1993, M. C. Tho-
mas, coll.; (2) Departamento Olancho, Parque Na-
cional La Muralla, June 1-2, 1995, R. Turnbow,
coll.

Caestron concinnus
(Gorham 1878)

(4) Departamento Francisco Morazan, Tegucigalpa,
El Rincon, August 13, 1988, R. D. Cave, coll.; (3)
same data except October 5, 1988; (1) same data
except October 15, 1993, F. W. Skillman Jr., coll.;
(1) Departamento Francisco Morazan, Mottuas,
Tatumbla, November 19, 1988, R. D. Cave, coll.;
(1) Departamento Francisco Morazan, San Antonio
de Oriente, Cerro Uyuca, May 29-June 4, 1990,
[malaise trap at the edge of cloud forest], R. D.
Cave, coll.; (7) Departamento El Paraiso, Montser-
rat, 5750', October 3, 1993, F. W. Skillman Jr.,
coll.; (3) Departamento Francisco Morazan, Parque
Nacional La Tigra, June 1, 1993, M. C. Thomas,
coll.; (1) Departamento Francisco Morazan, Tegu-
cigalpa, vie., Jutiapa (near Parque Nacional La Ti-
gra), 5200', beating thorny vines, May 23, 1993,
J. Rifkind, P. Gum, colls.

Caestron sp. near contractus
(Gorham 1882)

These records are given for a new species, which
will be described elsewhere. (1) Departamento El

Contributions in Science, Number 461

Rifkind: Honduran Cleridae ■ 7

Paraiso, Cerro Apaguis, Danli, February 20, 1988,
R. D. Cave, coll.; (2) Departamento El Paraiso, Ja-
caleapa, June 20, 1989, R. D. Cave, coll.; (1) De-
partamento Francisco Morazan, 2.8 km NW Zam-
brano, July 21, 1989, R. D. Cave, coll.; (6) Depar-
tamento Copan, 16 km SW Santa Rosa de Copan,
October 8, 1993, F. W. Skillman Jr., coll.; (2) De-
partamento Comayagua, Monte Fresco de Taulabe,
2000', May 26, 1993, Broadleaf Forest, beating ri-
parian shrubs, J. Rifkind, P. Gum, colls.

Subfamily PHYLLOBAENINAE

Isohydnocera cryptocerina
(Gorham 1883)

(1) Departamento Comayagua, Monte Fresco de
Taulabe, 2000', May 26, 1993, Broadleaf Forest,
beating riparian shrubs, J. Rifkind, P. Gum, colls.

Subfamily EPIPHLOEINAE

P bio gist ost emus erythrocephalus
(Gorham 1882)

(1) Departamento El Paraiso, Yuscaran, Cerro
Montserrat, November 16-30, 1992, R. Ortega,
coll.

Epiphloeus setulosus
(Thomson 1860)

(1) Departamento El Paraiso, Yuscaran, Cerro
Montserrat, January 1-7, 1993, R. Ortega, coll.

Ichnea bistrica
Gorham 1883

(2) Departamento Atlantida, Tela, Lancetilla, May
27, 1993, R. D. Cave, coll.; (2) same data except
May 29, 1993.

Ichnea mexicana
Thomson 1860

(1) Departamento La Paz, old road from La Paz to
Tatule, September 13, 1987, R. D. Cave, coll.; (1)
Departamento Francisco Morazan, Tegucigalpa,
San Juan del Rancho, October 18, 1987, R. D.
Cave, coll.; (1) Departamento Francisco Morazan,
San Antonio de Oriente, El Zamorano, June 27,
1988, J. Magana, coll.

Subfamily KORYNETINAE

Tribe Enopliini

Neortbopleura cyanipennis
(Chapin 1920)

(1) Departamento Francisco Morazan, 25.5 km
SSW Talanga, June 3, 1993, R. Turnbow, coll.

Neortbopleura duplicata
Barr 1976

(1) Departamento El Paraiso, vie. Yuscaran, May
25, 1993, M. C. Thomas, coll.; (1) Departamento

Francisco Morazan, 25.5 km SSW Talanga, Finca
Archaga, June 3, 1993, beating burn, F. W. Skill-
man Jr., coll.; (3) Departamento Francisco Mora-
zan, 23 km SSW Talanga, Finca Archaga, April 7,
1994, R. D. Cave, coll.

Neortbopleura purpurea
(Gorham 1883)

(1) Departamento Francisco Morazan, Tegucigalpa,
El Rincon, July 13, 1989, R. D. Cave, coll.; (1)
Departamento Olancho, Parque Nacional La Mur-
alla, May 31, 1995, R. Turnbow, coll.

Platynoptera mexicana
Thomson 1860

(1) Departamento Francisco Morazan, 30 km SE
Tegucigalpa, El Zamorano, April 1, 1981, no coll.

Pelonides bumeralis
(Horn 1868)

(2) Departamento Francisco Morazan, Tegucigalpa
vie., El Sitio, 3100', May 24, 1993, beating burned
Acacia in Thorn-Scrub, J. Rifkind, P. Gum, colls.;
(1) Departamento Francisco Morazan, 30 km E Te-
gucigalpa, Horticultural Parcel #4, May 9, 1985,
[on weeds], M. Martinez, coll.; (1) Departamento
Francisco Morazan, San Antonio de Oriente, El Za-
morano, June 2, 1993, B. Castro, coll.; (4) Depar-
tamento Francisco Morazan, 5 km E Escuela Agri-
cola Panamericana, June 2, 1993, M. C. Thomas,
coll.; (1) Departamento Francisco Morazan, 6.2 km
W San Juancito, June 1, 1993, M. C. Thomas, coll.;
(1) Departamento El Paraiso, Yuscaran, Monte
Montserrat, May 25, 1993, B. Castro, coll.

Cbariessa vestita
(Chevrolat 1835)

(1) Departamento Francisco Morazan, San Antonio
de Oriente, El Zamorano, May 14, 1981, [potato
field], G. Cruz, coll.; (1) same data except May 18,
1989, B. Medina, coll.; (1) same data except June
13, 1989, Medrano, coll.; (1) Departamento Fran-
cisco Morazan, 25.5 km SSW Talanga, June 3,
1993, M. C. Thomas, coll.; (2) Departamento At-
lantida, Tela, Lancetilla, August 11, 1992, R. Cave,
coll.; (1) Departamento Olancho, Parque Nacional
La Muralla, May 24, 1995, R. Morris, coll.

Apolopba apicicornis
(Chevrolat 1876)

(1) Departamento Francisco Morazan, 25.5 km SE
Talanga, Finca Archaga, June 3, 1993, picking burn
at night, F. W. Skillman Jr., coll.; (1) Departamento
Copan, 19 km SW Santa Rosa de Copan, October
8, 1993, F. W. Skillman, coll.

Cregya lineolata
(Gorham 1883)

(1) Departamento El Paraiso, Agua Fria, Danli,
February 20, 1988, R. D. Cave, coll.

8 ■ Contributions in Science, Number 461

Rifkind: Honduran Cleridae

Pelonium nigroclavatum
Chevrolat 1874

(1) Departamento Cortes, Santa Cruz de Yojoa,
Azul Meambar, May 20, 1995, R. Morris, coll.

Tribe Korynetini

Lebasiella pallipes
(Klug 1842)

(1) Departamento Francisco Morazan, Distrito
Central, El Chaguite, June 18, 1991, on Zea mays ,
B. Castro, coll.

ACKNOWLEDGMENTS

I thank all the collectors who provided specimens exam-
ined for this paper, and especially Ronald D. Cave (EAPZ)
and Fred W. Skillman Jr., who collected and made avail-
able to me ail the known specimens of Barrotiiius and
provided information on its habitat. A pleasure increased
by repetition is that of gratefully acknowledging the assis-
tance of William F. Barr and my father, Melvyn S. Rifkind,
who kindly read the typescript and offered helpful criti-
cism and suggestions. I am particularly grateful for the
dose critical reading of the first draft of this article, and
valuable suggestions for its improvement, provided by Ro-
land Gerstmeier {Freising, Germany) and Jonathan Maw-
dsiey (Cornell University). I thank Chuck Bellamy for ar-
ranging the loan of a type specimen from the collection
of the Transvaal Museum. Richard E. White of the United
States National Museum allowed me to examine type ma-
terial in his care, and Brian V. Brown of the Natural His-
tory Museum of Los Angeles County facilitated the loan;
I thank them both for their efforts. Finally, I would like
to thank Cristina Fagin, who kindly provided the Spanish
abstract, and Sharon Belkin for her fine illustrations.

LITERATURE CITED

Barr, W.F. 1962. A key to the genera and a classification
of the North American Cleridae (Coleoptera). The
Coleopterists Bulletin 16:121-127.

— — . 1975. Family Cleridae. In Checklist of the beetles
of North and Central America and the West Indies,
voi. 4. Family 73, ed. R.H. Arnett, 1-18. Gainesville:
Flora and Fauna Publications.

— — — — . 1976. Taxonomy of the new clerid genus Neor-
thopleura (Coleoptera). Melanderia 24:1-14.

— — . 1978a. New species of Enoclems from Mexico,
Central America, and Venezuela (Coleoptera: Cleri-
dae). The Coleopterists Bulletin 32(4):269-278.

— — - — 1978b. Taxonomy of the new clerid genus Bo-
gota from Mexico. The Pan-Pacific Entomologist 54:
287-291.

— 1980. New genera and a new species of New

World Cleridae (Coleoptera). The Pan-Pacific Ento-
mologist 56( 4):277-282.

Blanchard, C.E. 1844. Part 2: Insectes. In Insectes de
VAmerique Meridionale recueillis par Alcide d’Or-
bigny, ed. Brulle, 89-104. Paris.

Campbell, J.A., and W.W. Lamar. 1989. The venomous
reptiles of Latin America. Ithaca and London: Cor-
nell University Press, 425 pp.

Castelnau, L. 1836. Etudes entomologiques, ou descrip-
tions d’insectes nouveaux et observations sur la syn-
onymic. Silbermann Revue Entomologique 4:5-60.
Chapin, E.A. 1920. New American Cleridae, with notes

on the synonymy of Micropterus Chevr. (Coleopt.).
Proceedings of the Entomological Society of Wash-
ington 22(3):50-54.

. 1945. A new genus and species of clerid beetle

from Jamaica. Transactions of the Connecticut
Academy of Arts and Sciences 36:595-598.

Chevrolat, A. 1835. Coleopteres du Mexique. Fascicle 7.
Strasbourg, 50 pp.

— ■ — -. 1843. Coleopteres du Mexique. Magasin de
Zoologie, pp. 1-37.

— — — . 1874. Catalogue des derides de la collection de
M. A. Chevrolat. Revue et Magasin de Zoologie
3{2):252-329.

— — — — . 1876. Memoirs sur la famille des Clerites. Paris,
51 pp.

Dejean, REM. A. 1837. Catalogue des coleopteres de la
collection de M. le comte Dejean 5:385-503.

Erichson, W.F. 1847. Conspectus insectorum coleopter-
orum quae in Republica Peruana observata sunt. Ar-
chiv fur Naturgeschichte 13:67-185.

Ekis, G. 1976. Neotropical checkered beetles of the genus
Enoclems (Coleoptera: Cleridae: Clerinae). Studies
on the Neotropical Fauna 11:151-172.

. 1977. Classification, phytogeny, and zoogeogra-
phy of the genus Peril y pus (Coleoptera: Cleridae) .
Smithsonian Contributions to Zoology, no. 227, 138
PP-

Gorham, H.S. 1876. Notes on the coleopterous family
Cleridae, with descriptions of new genera and spe-
cies. Cistula Entomologica 2:57-106.

— — — . 1878. Descriptions of new genera and species of
Cleridae, with notes on the genera and corrections
of synonymy. Transactions of the Entomological So-
ciety of London 2:153-167.

— -. 1882. Biologia Centrali- Americana, Insecta, Co-
leoptera, Cleridae 3(2):129— 168.

- — . 1883. Biologia Centrali- Americana, Insecta, Co-

leoptera, Cleridae 3(2):169— 193.

- — — . 1886. Biologia Centrali- Americana, Insecta, Co-
leoptera, Cleridae 3(2):335-337.

Gray, G.R. 1832. Notices of new genera and species. In
The animal kingdom arranged in conformity with its
organization by the Baron Cuvier, ed. E. Griffith, In-
secta 1:375. London.

Horn, G.H. 1868. New species of Coleoptera from the
Pacific District of the United States. Transactions of
the American Entomological Society 2:129-140.

Kirsch, T.F. 1870. Beitrage zur Kaferfauna von Bogota.
Berliner Entomologische Zeitschrift 14:337-378.

Klug, J.C.F. 1842. Versuch einer systematischen Bestim-
mung und Auseinandersetzung der Gattungen und
Arten der Clerii, einer Insecten familie aus der Ord-
nung der Coleopteren. Abhandlungen der Preussi-
sche Akademie der Wissenschaft, pp. 259-397.

Kuwert, A.F. 1893. Die epiphloinen gattungen der cleri-
den und einige neue arten derselben. Annates de la
Societe Entomologique de Belgique 37:492-497.

LeCoete, J.L. 1850. An attempt to classify the longicorn
Coleoptera of the part of America north of Mexico.
Journal of the Academy of Natural Sciences of Phil-
adelphia 2(2):5-38.

— — — . 1861. Classification of the Coleoptera of North
America. Smithsonian Institute Miscellaneous Col-
lection 136:1-208.

Monroe, B.L. Jr. 1968. A distributional study of the birds
of Honduras. American Ornithologists’ Union, Or-
nithological Monographs no. 7. 458 pp.

Pic, M. 1954. Nouveaux derides de L’Afrique australe

Contributions in Science, Number 461

Rifkind: Honduran Cleridae ■ 9

(Coleoptera). Annals of the Transvaal Museum
22(3):389-390.

Rifkind, J. 1993. A new species of Cymatoderella Barr
(Coleoptera: Cleridae) from Mexico and Central
America, with a key and distributional data for the
genus. The Coleopterists Bulletin 47(3):279-284.

. 1995. Additions to the checkered beetle fauna of

Belize with the description of a new species (Cole-
optera: Cleridae) and a nomenclatural change. In-
secta Mundi 9(1-2): 17-24.

Schaeffer, C. 1908. On new and known Coleoptera of
the families Coccinellidae and Cleridae. Journal of
the New York Entomological Society 16:125-135.

Schenkling, S. 1900. Neue amerikanische Cleriden nebst
Bemerkungen zu schon beschriebenen Arten. Deut-
sche Entomologische Zeitschrift, pp. 385-409.

— . 1906. Die Cleriden des Deutchen Entomolo-

gischen National-Museums, nebst Beschreibungen
neuer Arten. Deutsche Entomologische Zeitschrift,
pp. 241-320.

- . 1907. Neue Cleriden von Zentral-Amerika nebst

Bemerkungen iiber die Beziehungen der mittel-amer-

ikanischen Cleriden zu denen des iibrigen Amerika.
Deutsche Entomologische Zeitschrift, pp. 297-307.

Selander, R.B., and P. Vaurie. 1962. A gazetteer to ac-
company the “Insecta” volumes of the “Biologia
Centrali- Americana.” American Museum Novitates,
no. 2099, 70 pp.

Spinola, M. 1841. Monographic de terediles. Revue
Zoologique 4:70-76.

— — — 1844. Essai monographique sur les Clerites, In-
sectes Coleopteres. 2 vols., Genes, 602 pp.

Thomson, J. 1860. Materiaux pour servira a une mono-
graphs nouvelle de la famille des Clerides. Musee
Scientifique 1860:46-67.

Wolcott, A.B. 1911. New American Cleridae, with notes
on others. Entomological News 22:115-125.

— . 1927. A review of the Cleridae of Costa Rica.

Coleopterological Contributions 1(1): 1—104.

— . 1944. A generic review of the subfamily Phyllo-

baeninae (olim Hydnocerinae) (Col.). Journal of the
New York Entomological Society 52:121-152.

Received 10 November 1995; accepted 7 March 1996.

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

Number 462
3 October 1996

UsslX

a m

Contributions
in Science

Preliminary Analysis ol a Host Shift:
Revision oe the Neotropical Species of
Apocephalus, Subgenus Mesophora
(Diptera: Phoridae)

Brian V. Brown

Natural History Museum of Los Angeles County

Serial
Publications
ol THE
Natural History
Museum of
Los Angeles
County

Scientific

Publications

Committee

Robert J. Lavenberg, Acting Deputy Director
for Research and Collections
John M. Harris, Committee Chairman
Brian V. Brown
Kenneth E. Campbell
Kirk Fitzhugh
Karen Wise

Robin A. Simpson, Managing Editor

# Contributions in Science, a miscellaneous series of tech-
nical papers describing original research in the life and
earth sciences.

# Science Series, long articles and collections of papers on
natural history topics.

Copies of the publications in these series are sold through
the Museum Book Shop. A catalog is available on request.

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

Printed at Allen Press, Inc., Lawrence, Kansas
ISSN 0459-8113

Preliminary Analysis of a Host Shift:

Revision of the Neotropical Species of
Apocephalus, Subgenus Mesophora
(Diptera: Phoridae)

Brian V. Brown1

CONTENTS

ABSTRACT 2

INTRODUCTION 2

METHODS 2

SYSTEMATICS 3

Phylogenetic Reconstruction 3

Monophyly of Subgenus Mesophora 3

Recognition of Subgenus Mesophora 3

Relationships within Subgenus Mesophora 3

Apocephalus borealis- group 7

Apocephalus wheeleri- group 15

Apocephalus Coquillett, 1901; Subgenus Mesophora Borgmeier, 1937 16

Apocephalus apivorus new species 16

Apocephalus atavus new species 17

Apocephalus adustus Brown, 1993 17

Apocephalus borealis- group 18

Apocephalus anfr actus- subgroup 18

Apocephalus absentis Brown, 1993 18

Apocephalus bisetus Brown, 1993 19

Other Apocephalus borealis- group species 19

Apocephalus megalops new species 19

Apocephalus emphysemus new species 19

Apocephalus wheeleri- group .. 20

Apocephalus curtus- subgroup 20

Apocephalus lizanoi new species 20

Apocephalus curtus Brown, 1993 21

Apocephalus lemniscus new species 21

Apocephalus wheeleri- subgroup 22

Apocephalus niveus new species 22

Apocephalus antennatus- infragroup 23

Apocephalus antennatus Malloch, 1913 23

Apocephalus longistylus Brown, 1993 23

Apocephalus wheeleri- infragroup 23

Apocephalus mortifer Borgmeier, 1937 23

Apocephalus tritarsus Brown, 1993 23

Unplaced Species 23

Apocephalus micrepelis Brown, 1993 23

Apocephalus pilatus new species 23

Apocephalus crassus new species 24

Apocephalus prolixus new species 24

Apocephalus secundus new species 25

1. Natural History Museum of Los Angeles County, 900 Exposition Boulevard, Los Angeles, California, 90007. E-mail:
brianb@mizar.usc.edu.

Contributions in Science, Number 462, pp. 1-36
Natural History Museum of Los Angeles County, 1996

Apocephalus echinatus new species 25

Species Recognized but Not Named 26

Phorid Species 3251 26

Phorid Species 3223 26

Phorid Species 3246 . 27

Phorid Species 3247 27

Phorid Species 3250 27

Phorid Species 3252 28

Phorid Species 3253 29

IDENTIFICATION 29

Key to Males of Neotropical Region Mesophora Species 29

Key to Females of Neotropical Region Mesophora Species 31

EVOLUTION OF HOST SELECTION 32

ACKNOWLEDGMENTS 35

LITERATURE CITED 35

ABSTRACT. Twelve new species of Apocephalus, subgenus Mesophora, are described and named; an
additional six unassociated males and females are described but not named. The 12 new species are A.
apivorus, atavus, emphysemus, lemniscus, lizanoi, megalops, and niveus from Costa Rica and A. crassus,
echinatus, pilatus, prolixus, and secundus from the Dominican Republic. A. leptotarsus Brown is synon-
ymized with A. antennatus Malloch new synonymy. The previously unknown male of A. absentis Brown
is described, as are the previously unrecognized females of A. adustus and A. curtus. A new key to species
of the Neotropical Region is given. The relatively most primitive species, A. apivorus and adustus, are
parasitoids of stingless bees, whereas A. tritarsus is a parasitoid of lampyrid beetles, like most other Me-
sophora. The host shift within the subgenus Mesophora from ants to lampyrid beetles appears to have
been via parasitism of stingless bees.

INTRODUCTION

In a group of flies that are remarkably similar in
their life histories, species of the genus Apocepha-
lus, subgenus Mesophora, represent a major enig-
ma. Whereas all other species of this genus are par-
asitoids of ants (Hymenoptera: Formicidae), species
of Mesophora attack beetles, wasps, bees, and spi-
ders (Brown, 1993). The reasons for this host shift
are unknown; other genera hypothesized to be re-
lated to Apocephalus also are ant-parasitoids
(Brown, 1992), so attacking the new hosts seems
to represent a newly derived behavioral character
(Brown, 1993). Study of this phenomenon is an im-
portant reason for this revision.

Although only a short period of time has passed
since the first revision of Mesophora (Brown,
1993), already a significant amount of new material
has accumulated. Previously, I dealt with one small
subgroup (Brown, 1994b) and some new life his-
tory information (Brown, 1994a), but in general
there still seems to be an almost unlimited number
of new species to discover. Large collections still are
available from only two countries, Costa Rica and
the Dominican Republic, so it seems likely that as
other sites are surveyed still more species and re-
markable life histories will be uncovered.

In this paper, I describe newly discovered species
and life histories, reviewing previously described
taxa only when significant new information is
available. The key to species, however, includes all
species described previously (Brown, 1993, 1994b).

2 ■ Contributions in Science, Number 462

METHODS

Methods and terms are similar to those used in my pre-
vious revisions of this group (Brown, 1993, 1994b). One
important difference is that I no longer report on costal
sector ratios. The point that these measurements are of
little use in identification, because they vary so much, has
been adequately made (Disney, 1980). An overwhelming
reliance on them by previous workers is a thing of the
past.

Another change is that I no longer use potassium hy-
droxide to clear specimens. Following other dipterists
(Cumming, 1992), I now use lactic acid as an alternative
that is much safer for specimens.

For brevity, species names are cited as follows: Apo-
cephalus mortifer or A. mortifer, rather than Apocephalus
(Mesophora) mortifer and A. (M.) mortifer.

Geographical coordinates are quoted as decimal de-
grees, rather than degrees, minutes, and seconds.

In addition to the usual insect labels recording locality
information, specimens were labelled with barcoded insect
labels (Thompson, 1994) and data were recorded in a da-
tabase. All barcoded labels begin with the abbreviation
“LACM ENT,” indicating that the Natural History Mu-
seum of Los Angeles County is the institution where the
data are stored. To make later recognition of holotypes
easier, I list their individual barcode number in square
brackets.

Specimens were deposited in the following collections:

CMNH Section of Invertebrate Zoology, Carnegie Mu-
seum of Natural History, 4400 Forbes Avenue,
Pittsburgh, PA 15213-4080, U.S.A. (C.W.
Young)

INBIO Instituto Nacional de Biodiversidad, A.P.

22-3100, Santo Domingo, Heredia, Costa Rica
(M. Zumbado)

Brown: Neotropical Apocephalus, Subgenus Mesophora

LACM Entomology Section, Natural History Museum
of Los Angeles County, 900 Exposition Boule-
vard, Los Angeles, CA 90007, U.S.A. (B.V.
Brown)

MCZ Museum of Comparative Zoology, Harvard
University, Cambridge, MA 02138, U.S.A. (on
indefinite loan to B.V. Brown)

MICR Museo de Insectos, Universidad de Costa Rica,
San Pedro, San Jose, Costa Rica (P.S. Hanson)
USNM United States National Museum, Smithsonian
Institution, Washington, DC 20560, U.S.A. (on
indefinite loan to B.V. Brown)

As previously, I name species only based on male spec-
imens; an exception is the Apocephalus anfractus-sub-
group, which was based on females. Unfortunately, this
means that many forms known from one sex or another
only cannot be formally dealt with at this time, especially
the now known, but unassigned, males of A. anfractus-
subgroup taxa. Unnamed Costa Rican species are de-
scribed herein to facilitate their recognition in Costa Rica’s
ongoing biodiversity inventories; they are referred to by
their Phorid Species Number in my phorid names database
(e.g., Phorid species 3143). To make later recognition of
these specimens easier, I list their individual barcode num-
ber in square brackets in the Material Examined sections.

SYSTEMATICS

PHYLOGENETIC RECONSTRUCTION

There are numerous species known from a single
sex only. The lack of phylogenetic information
from the missing sexes makes phylogenetic analysis
using a computer program impossible; huge num-
bers of alternative trees are generated as the pro-
gram attempts to analyze all possibilities of the
missing character states. For that reason, the fol-
lowing analysis is based only on taxa known from
both sexes, including those from the Nearctic Re-
gion. Also excluded is the new species A. atavus,
for reasons stated in the discussion of the A. bo-
realis-group.

Monophyly of Subgenus Mesophora

The monophyly of subgenus Mesophora is no lon-
ger supported by all of the same character states I
proposed earlier (Brown, 1993). The discovery of
A. apivorus reduces the number of possible synapo-
morphies of the group to the two listed below. The
larval synapomorphies I proposed earlier are still
speculative. The larva of A. apivorus is similar to
that of other Mesophora, and divergent from the
only described larva of an Apocephalus { Apocepha-
lus) species (Brown and Feener, 1991), but our lack
of knowledge about the larvae of outgroup taxa
makes proposing larval synapomorphies of Meso-
phora premature.

1) Female with dark glands (plesiomorphic: ab-
dominal glands white)

Because dark glands are absent from A. ata-
vus and the next most primitive species, A. adus-
tus, it is also equally parsimonious to propose
that the dark glands arose separately in A. apiv-

Contributions in Science, Number 462

orus and the rest of Mesophora (excluding A.
atavus and adustus ).

2) Ovipositor notched (Brown, 1993, fig. 51, n)
(plesiomorphic: ovipositor not notched)

This character reverses later in the subgenus.

Recognition of Subgenus Mesophora

Because of the reversals of characters in the sub-
genus, identification of an Apocephalus species as
Mesophora is now complicated. The following key
will allow recognition of all species:

1 Male 2

Female 5

2 Flagellomere 1 greatly enlarged (Figs. 3-5);

lower and usually upper fronto-orbital setae
absent subgenus Mesophora

- Flagellomere 1 small, round; fronto-orbital

setae present (Fig. 1) 3

3 Wing vein CuAj short, not reaching wing

margin

.... A. ( Mesophora ) apivorus new species

- Wing vein CuAj reaching wing margin . . 4

4 Anteroventral row of setulae on hind basi-
tarsus enlarged basally (similar to Fig. 63);
flagellomere 1 pyriform; halter dark brown

A. ( Mesophora ) atavus new species

- Anteroventral row of setulae on hind basi-
tarsus not noticeably enlarged; other char-
acters various .... subgenus Apocephalus

5 Abdominal glands near segment 5 dark in
color (Brown, 1993, fig. 45, g); if not, ovi-
positor with ventral notch apically (Brown,
1993, fig. 51, n) ... subgenus Mesophora

- Abdominal glands white, invisible in cleared
specimens; ovipositor lacking ventral notch

6 Lower fronto-orbital setae absent; in some

species upper fronto-orbital setae also ab-
sent subgenus Mesophora

- Fronto-orbital setae present

subgenus Apocephalus

Relationships within Subgenus Mesophora

Within Mesophora, I propose the following hy-
pothesized synapomorphic characters:

3) Male flagellomere 1 enlarged, flattened (ple-
siomorphic: flagellomere 1 round)

4) Stylet with lateral barbs (plesiomorphic: barbs
absent)

5) Lower fronto-orbital seta of male absent (ple-
siomorphic: present)

6) Upper fronto-orbital seta of male absent (ple-
siomorphic: present)

7) Lower fronto-orbital seta of female 1) slightly
displaced medially, 2) in line with interfrontal
setae (Fig. 10) (plesiomorphic: lower fronto-or-
bital seta close to eye)

8) Ovipositor cylindrical (plesiomorphic: ovipos-
itor flat)

Brown: Neotropical Apocephalus, Subgenus Mesophora ■ 3

9) Dorsal apex of ovipositor rounded (plesio-
morphic: apex pointed)

10) Ventral apex of ovipositor rounded (plesio-
morphic: apex pointed)

1 1 ) Anteroventral row of setulae of hind tarsomere
with at least basal setulae enlarged (Figs. 63-
65) (plesiomorphic: setulae all short, subequal)

12) Anteroventral row of setulae of hind tarsomere
sinuate (Fig. 65) (plesiomorphic: row straight)

13) Ventral setae of female abdominal segment 6
1) reduced (thin, but still long), 2) greatly re-
duced (short, thin), 3) absent (plesiomorphic:
ventral setae large, thick, prominent)

14) Venter of mid tarsomere 2 with at least one
extra setula (plesiomorphic: with only one set-
ula)

15) Flagellomere 1 of male abruptly narrowed at
base (Fig. 9) (plesiomorphic: flagellomere 1
more gradually narrowed)

16) Anterior margin of ovipositor narrowly pro-
duced in an anterior strip (Brown, 1994b, fig.
5A) (plesiomorphic: ovipositor anteriorly
rounded)

17) Apex of male cercus truncate (plesiomorphic:
apex of cercus elongate)

18) Ovipositor darkly sclerotized over entire sur-
face (plesiomorphic: dark sclerotization con-
fined to margin of ovipositor)

19) Surstylus with three or fewer setulae (plesio-
morphic: surstylus with many setulae)

20) Palpus of male enlarged, with few short, stub-
by setulae (plesiomorphic: palpus small, with
normal, pointed setulae)

21) Venter of female abdomen with distinctive
combs of setulae (Figs. 66-67) (plesiomorphic:
abdomen without combs of setulae)

22) Female stylet distinctive, with medial sclerite
on venter (Figs. 60-61)

23) Male abdominal segment 6 with dark ventral
setae (Fig. 70) (plesiomorphic: setae absent)

24) Epandrium extremely short (Figs. 20-22) (ple-
siomorphic: epandrium longer)

25) Surstylus with 1) single carina, 2) short multi-
ple carinae, 3) long, complete carina (plesio-
morphic: carinae absent)

26) Mid tarsomere 1 of male enlarged (plesio-
morphic: tarsomeres slender)

27) Left side of epandrium with distinct ventral
ridge (plesiomorphic: ridge absent)

28) Right surstylus elongate (plesiomorphic: sur-
stylus short)

A number of hypothesized synapomorphies from
my previous analysis (Brown, 1993) had to be dis-
regarded. A list of them, and the reasons for no
longer considering them as phylogenetically useful,
is given below:

a) Lower fronto-orbital seta absent from female.

Lronto-orbital setae are extremely variable
throughout Mesophora. Lor instance, in the A.
borealis- group, all frontal setae are present in
females of species 3246 (Lig. 13), the lower fron-

4 ■ Contributions in Science, Number 462

to-orbital seta is absent from females of species
3223 (Fig. 12), and both the upper and lower
fronto-orbital setae are absent from females of
species 3251 (Fig. 11).

b) Series of long, thick ventral setae on left side of
epandrium.

Study of additional species has shown that
“thick” and “thin” epandrial setae grade into
each other. The distinction can no longer be
maintained.

c) Right side of epandrium with prominent, darkly
sclerotized longitudinal ridge.

This character is found widely, including in
the newly described, relatively primitive A. apiv-
orus.

Based on characters 1-28 listed above, I analyzed
a character matrix (Table 1) using HENNIG86.
Characters were polarized with reference to the
outgroup taxa Pseudacteon Coquillett and Neo-
dobrniphora Malloch; species of these genera have
the synapomorphic character of pointed horns on
the larval spiracle in common with Apocephalus
species (as in Brown, 1993, fig. 99s). All characters
were equally weighted, and multistate characters
unordered. This produced six equally parsimonious
trees (length 49, ci 67, ri 81). The number of trees
was unaffected by successive approximation
weighting.

Of these six trees, three treated character 25,
state 2, as a synapomorphy of A. unitarsus and
wheeleri. However, this state is also found in a rel-
atively primitive member of this group, A. hansoni
Brown, not included in this analysis. Because an-
other character supports the monophyly of a group
not including A. hansoni, character 25, state 2, is
a primitive character in reference to A. unitarsus
and A. wheeleri and cannot group these two spe-
cies.

The three remaining trees differ only by different
assignments of states of character 13 at various bas-
al nodes of the cladogram. In tree 1 (Lig. 71) there
is a branch including the species A. grandiflavus-
A. wheeleri. This branch is resolved by an arbitrary
assignment of the state of character 13 as 0, 1, or
3 and allows two further monophyletic groups: one
is composed of A. grandiflavus — A. truncaticercus
and is supported by character 13, state 3; the other
is A. longistylus-A. wheeleri and is supported by
character 13, state 1. Tree 2 differs by depicting a
sister-group relationship between A. niveus and the
other species, which are united by character 13,
state 1. This tree is similar to one that would stip-
ulate that character 13 is ordered (as 0 — » 1 — » 2
— > 3), an assumption not made in this analysis. Tree
3 has an unresolved polytomy including A. niveus,
the A. truncaticercus infragroup (see below), and
A. longistylus-wheeleri. The consensus tree of the
six equally parsimonious trees is given in Lig. 71.

All of the significant differences among these
trees are because of character 13. I prefer tree 2
because it has the most modest assumption about

Brown: Neotropical Apocephalus, Subgenus Mesophora

Figures 1—6. Heads. Abbreviations: 1 f-o — lower fronto-orbital seta; u f-o — upper fronto-orbital seta.

Contributions in Science, Number 462

Brown: Neotropical Apocephalus, Subgenus Mesophora ■ 5

Figures 7-12. Heads. Abbreviation: 1 f-o — lower fronto-orbital seta.

6 ■ Contributions in Science, Number 462

Brown: Neotropical Apocephalus, Subgenus Mesophora

15. ? A. antennatus 16. S A. longistylus 17. $Phorid species 3136

Figures 13-17. Heads.

this character: that the reduction of abdominal se-
tae is synapomorphic, regardless of later, further re-
ductions. Thus, it can define a monophyletic group
that does not include A. niveus.

Based on this analysis, the following informal
groups can be recognized.

Apocephalus borealis- group. This group is based
on the presence of enlarged ventral setulae on tar-
somere 1 of the hind leg. Within this group are two
subdivisions, one of which is the A. anfractus- sub-
group (represented by A. absentis in this analysis).
This assemblage of species is well defined by the
presence of a sinuate line of setulae on hind tarso-
mere 1. The relationships of the A. anfractus- sub-
group were discussed previously (Brown, 1993).

The relationships of the other species of the A.
borealis group are less clear. Based on the enlarged
costa of males, A. borealis Brues and A. emphyse-
mus are probably sister taxa. Species 3223 and
3246 are also clearly related to A. borealis based
on the broad apex of the ovipositor, and one of
these two species probably is the female of A. em-
physemas.

The remaining taxa are A. megalops, species
3251, species 3247, and possibly A. atavus. The
first three species are relatively similar, although A.
megalops and species 3247 have barbed stylets and
might be part of a group including A. borealis and
relatives. This character, however, has occurred
more than once in the subgenus Mesophora, and

Contributions in Science, Number 462

Brown: Neotropical Apocephalus, Subgenus Mesophora ■ 7

Figures 18-25. Male terminalia, right lateral and left lateral.

8 ■ Contributions in Science, Number 462

Brown: Neotropical Apocephalus, Subgenus Mesopbora

26.

27.

Figures 26-33. Male terminalia, right lateral and left lateral.

Contributions in Science, Number 462

Brown: Neotropical Apocephalus, Subgenus Mesophora I 9

35.

Figures 34-41. Male terminalia, right lateral and left lateral.

10 ■ Contributions in Science, Number 462

Brown: Neotropical Apocephalus, Subgenus Mesophora

A. prolixus

Figures 42-49. Male terminalia, right lateral and left lateral.

Contributions in Science, Number 462

Brown: Neotropical Apocephalus, Subgenus Mesophora ■ 11

50. A. adustus

51. A. apivorus

52. A. atavus

53. A. megalops

3223

3246

58. A. niveus

Figures 50-58. Ovipositors, dorsal.

its presence does not provide strong evidence for
grouping.

Apocephalus atavus is an extraordinarily difficult
species to place. On one hand, it lacks nearly all of

the apomorphies of Mesophora and is barely rec-
ognizable as a member of this subgenus with only
the anteroventral notch of the ovipositor indicating
its affinities. If we accept it as part of the A. bo-

ll ■ Contributions in Science, Number 462

Brown: Neotropical Apocephalus, Subgenus Mesophora

59. A. apivorus

60. A. curtus (ventral) 61. A. curtus (dorsal)

62. A. atavus

Figures 59-65.

63. A. borealis 64. Phorid species 3246 65. A. anfractus

59-62. Stylets, dorsal (except Fig. 60). 63-65. Tarsomere 1 of left hind leg, anterior (apex at bottom).

Contributions in Science, Number 462

Brown: Neotropical Apocephalus, Subgenus Mesophora ■ 13

69. 9 a. atavus

70. 3 A. lizanoi

Figures 66-70. 66-67. Venter of abdomen. 68. Wing. 69-70. Apex of venter of abdominal segment 6.

14 ■ Contributions in Science, Number 462

Brown: Neotropical Apocephalus, Subgenus Mesophora

Table 1. Data matrix for cladograms.

1111111111222222222
character # 1234567890123456789012345678

apivorus

adustus

borealis

megalops

absentis

curtus

lizanoi

niveus

brunnipes

antennatus

insulanus

qemursus

truncaticercus

satanus

brevicercus

grandiflavus

angustistyius

longistylus

unitarsus

mortifer

wheeleri

1101000000000000000000100000

0110110000000000000000000000

1101111000100000000100000000

1111111100100000000000100000

1110111100110000000000000000

1110111111000000000111110000

1110111111000000000011110000

1011111111000000000000000000

1011111101001000000000000000

0011111101002100000000000000

1011111101002101000000000000

1011111111001000000000000100

0011111111003011111000000000

0011111111003011111100000000

1011111111001000000000001001

1011111111001100000100000001

1011112111001000000000002110

1011112111001000000000003110

1011112111001000000000002110

realis- group, we must postulate the reversal or loss
of four important synapomorphies (characters 1, 3,
5, 6). On the other hand, it has enlarged setulae on
the hind basitarsus, a character that provides strong
support of the A. borealis- group because it is absent

from the many outgroups I have examined. Parsi-
mony demands that A. atavus be classified as a
primitive species of Mesophora, but further studies
on the larva and internal structure of this fly should
be undertaken to see if its peculiar mix of charac-
ters represents a single parallel development (of
synapomorphy 11) or a dramatic set of reversals.

Apocephalus wheeleri- group. This group in-
cludes most of the remaining Mesophora species
and is defined by characters 9 and 10, the rounded
dorsal and ventral apices of the ovipositor. Because
this diagnosis depends on having females present,
taxa known only from males cannot be definitely
placed in this group, unless they belong to one of
the well-established subgroupings.

Within the A. wheeleri- group are the A. curtus-
subgroup and the A. wheeleri- subgroup. The A.
curtus- subgroup is an extremely distinctive assem-
blage, based mainly on the abdominal structure of
females (Figs. 66-67). The A. wheeleri- subgroup is
defined by the presence of a barbed stylet, a char-
acter that occurs sporadically throughout this sub-
genus, and the loss of the notched ovipositor, which
was initially a defining character of Mesophora.
There are apparently four groupings in this sub-
group: the A. antennatus- infragroup, the A. whee-
/m-infragroup, the A. truncaticercus- infragroup,
and the A. longistylus-inhagroup.

tplot file 0 from bb 3 trees
tree 1

grandiflavus

brevicercus
satanus
truncaticercus
longistylus
—angustistyius
•brunnipes
^antennatus

insulanus
fi======^emursus

==j rj=unitarsus
1— mort i f er
^wheeleri

tree 3

grandiflavus

brevicercus

truncaticercus

longistylus

angustistyius

brunnipes

antennatus

insulanus

gemursus

unitarsus

mortifer

wheeleri

tree 2

^apivorus

=adustus

smegalops

-borealis

=absentis

=lizanoi

=curtus

=grandi£lavus

=brevicercus

=satanus

struncat icercus

-longistylus

sangustistylus

consensus tree

“1

=apivorus

=adustus

miegalops

=borealis

=absentis

=lizanoi

=curtus

=niveus

^grandiflavus

=brevicercus

=satanus

^truncaticercus

-longistylus

=angustistylus

^brunnipes

^antennatus

-insulanus

=gemursus

=unitarsus

^mortifer

=wheeleri

Figure 71. Three equally parsimonious resolutions of data from Table 1, plus consensus tree of six equally parsimonious
tress. On trees 1-3, the occurrence of character 13, states 1 and 3, are indicated.

Contributions in Science, Number 462

Brown: Neotropical Apocephalus, Subgenus Mesophora ■ 15

The A. antennatus- infragroup is a newly pro-
posed assemblage, based on an apparent reversal of
the rounded dorsal apex of the ovipositor. This in-
fragroup contains A. antennatus and A. insulanus,
a sister-group pair I recognized previously (Brown,
1993); it also includes A. brunnipes, a species that
was placed in this infragroup in some previous
analyses (Brown, 1993, fig. 110).

The relationships within the A. wheeleri- infra-
group were discussed previously (Brown, 1993), as
were those of the A. truncaticercus- infragroup
(Brown, 1994b). One change is that A. gemursus
Brown is now considered a basal member of the A.
wheeler i-'mhagroup .

In general, the relationships of the Costa Rican
Mesophora fauna are relatively well known. Even
taxa known only from one sex can be placed within
at least a subgroup. I am unable to classify the Do-
minican Republic species, however, as most are
known from males only. Many of these males are
relatively similar (A. pilatus, A. secundus, A. echi-
natus, and A. crassus ) and do not show any strong
divergence from a generalized Mesophora.

Apocephalus Coquillett, 1901

Subgenus Mesophora
Borgmeier, 1937

DIAGNOSIS. Abdominal glands of female dark.
Ovipositor primitively with anteroventral notch.
Larval mandible with lateral projection; larval spi-
racular region rounded, densely spinulose; terminal
segments of larva lacking large processes (Brown,
1993). One species (A. absentis ) lacks wing vein

R 2+3-

Apocephalus apivorus new species

(Figs. 1, 18-19, 51, 59)

SPECIES RECOGNITION. This species can be
distinguished from all other Apocephalus by the
short wing vein CuA1? which does not reach the
wing margin. It will not key to subgenus Mesopho-
ra in Borgmeier’s (1971) most recent key to Apo-
cephalus species.

DESCRIPTION.

General Characters. Body length 1.5-1. 8 mm.
Frons yellow to brown. Palpus yellow. Dorsum of
thorax yellow. Pleural regions yellow. Scutellum of
same color as dorsum of thorax. Scutellum with
long anterior and long posterior setae. Legs yellow-
ish-brown. Venter of tarsomere 2 of mid leg with
row of thin setulae and thick apical setula. Hind
femur yellowish-brown, evenly colored throughout.
Anteroventral row of enlarged setulae on hind ba-
sitarsus straight, with all setulae short, subequal to
those of other rows. Wing vein Rs slender, subequal
to or thinner than costa. Wing vein R2+3 well de-
veloped. Wing vein CuA2 attenuated, not reaching
wing margin. Halter yellow.

Male. Lower fronto-orbital seta present. Upper
fronto-orbital seta present. Flagellomere 1 yellow,

16 ■ Contributions in Science, Number 462

pyriform, apical two-thirds with narrowest part
greater than one-half basal width. Palpus small,
with setae normal, thick, moderately long, pointed.
Mid leg with tarsomeres are slender. Costal vein
not thickened. Abdominal tergites dark brown to
mostly dark, with yellow markings. Venter of ab-
domen yellow. Venter of segment 6 with distinct,
black setae, ventral setae in a single row. Venter of
segment 6 without sclerite. Epandrium globular,
approximately as long as high, with setae near cer-
cus not markedly larger than other epandrial setae.
Left side of epandrium with scattered, thin setae
(Fig. 19). Right side of epandrium with prominent,
darkly sclerotized, dorsolateral ridge (Fig. 18).
Right surstylus short, rounded, with ventral carina
absent. Cercus brown. Setae of cercus and proctiger
of normal size, longer and thicker than short setae
of epandrium.

Female. Lower fronto-orbital seta present, close
to eye (Fig. 1). Upper fronto-orbital seta present.
Flagellomere 1 yellow, pyriform. Palpus small, with
setae normal, thick, moderately long, pointed. Col-
or of abdominal tergites mostly brown, but tergite
6 yellow. Venter of abdomen yellow, but segment 6
dark gray. Venter of segment 3 with few dark setae.
Venter of abdominal segment 5 lacking dense setal
combs, without sclerite. Ventromedial setae of seg-
ment 6 present, long, thick, evenly spaced, arranged
in relatively straight line. Venter of segment 6 with-
out sclerite. In cleared specimens, dorsal abdominal
glands visible, dark. Intersegment 6-7 without dis-
tinct sclerite. Ovipositor dorsally broad (Fig. 51),
anterodorsal portion broad, rounded. Anteroven-
tral margin of ovipositor with notch. Posterodorsal
apex of ovipositor pointed. In lateral view, postero-
ventral apex of ovipositor straight. Posteroventral
apex of ovipositor pointed. Stylet with lateral
barbs. Dorsal sclerite consisting of two long pro-
cesses (Fig. 59).

GEOGRAPHICAL DISTRIBUTION. Manaus,
Brazil; Costa Rica.

PHYLOGENETIC RELATIONSHIPS. This is
hypothesized to be the most relatively primitive
species of Mesophora.

WAY OF LIFE. This species parasitizes male
stingless bees of the species Cephalotrigona (for-
merly Trigona ) capitata (F. Smith) (Apidae: Meli-
poninae). Detailed information about life history is
given in a separate publication (Brown, in press).
This is the first species of Apocephalus known to
attack meliponine bees.

DERIVATION OF SPECIFIC EPITHET. This
species is named for the larval habit of eating bees.

HOLOTYPE. A, COSTA RICA: Heredia, La
Selva Biological Station, 10.43°N, 84.02°W, 25-
26.vi.1993, B.V. Brown, reared from male Trigona
capitata (LACM) [LACM ENT 001075].

PARATYPES. BRAZIL: Para, Oriximina, 1.8°S,
53.83°W, Id, 13.X.1992, J. Rafael, Malaise trap
(INPA); COSTA RICA: Heredia, Chilamate,
10.45°N, 84.08°W, Id, v.1989, P. Hanson, Malaise
trap, primary forest (LACM), La Selva Biological

Brown: Neotropical Apocephalus, Subgenus Mesophora

Station, 10.43°N, 84.02°W, 15.ii-l.iii. 1993, ALAS,
1A, Malaise trap M/07/22, IS, Malaise trap
M/10/25 (INBIO), 369, 19-25.vi.1993, B.V.
Brown, attacking male Trigona capitata (LACM,
INBIO, MCZ, USNM, MICR, CMNH), 65 S,
66 9 , 25-26.vi.1993, B.V. Brown, reared from male
Trigona capitata (LACM, MCZ, USNM, MICR,
CMNH), 1 9, 26.vi.1993, attacking female Trigona
dorsalis (LACM), Puntarenas, 24 km W Piedras
Blancas, 8.77°N, 83.4°W, IS, ii-iii.1989, IS, ix-
xi.1989, P. Hanson, Malaise trap, 200 m (LACM).

Apocephalus atavus new species

(Figs. 28-29, 52, 62, 69)

SPECIES RECOGNITION. This species does not
key to Mesophora in the key of Borgmeier (1971).
It can be recognized, with difficulty, as belonging
to this group by the enlarged basal setulae on the
hind basitarsus and by the notched ovipositor in the
female.

DESCRIPTION.

General Characters. Body length 1.6-2. 2 mm.
Frons yellow. Palpus yellow. Dorsum of thorax yel-
low. Pleural regions yellow. Scutellum of same color
as dorsum of thorax. Scutellum with short anterior
and long posterior seta. Legs yellow. Venter of tar-
somere 2 of mid leg with row of thin setulae and
thick apical setula. Hind femur yellowish-brown,
with abrupt, distinctive darkening at apex. Antero-
ventral row of enlarged setulae on hind basitarsus
straight, with some setulae markedly longer than
those of other rows. Wing vein Ra slender, subequal
to or thinner than costa. Wing vein R2+3 well de-
veloped. Wing vein CuAa reaches wing margin.
Halter brown.

Male. Lower fronto-orbital seta present. Upper
fronto-orbital seta present. Flagellomere 1 yellow,
pyriform, apical two-thirds with narrowest part
greater than one-half basal width. Palpus small,
with setae normal, thick, moderately long, pointed.
Mid leg with tarsomeres all slender. Costal vein not
thickened. Abdominal tergites mostly dark, with
yellow markings to mostly yellow, with dark mark-
ings. Venter of abdomen yellow. Venter of segment
6 bare. Venter of segment 6 without sclerite. Epan-
drium globular, approximately as long as high, with
setae near cercus not markedly larger than other
epandrial setae. Left side of epandrium with scat-
tered, thin setae (Fig. 29). Right side of epandrium
without ridge (Fig. 28). Right surstylus short,
rounded, lacking ventral carina. Cercus yellow. Se-
tae of cercus and proctiger of normal size, longer
and thicker than short setae of epandrium.

Female. Lower fronto-orbital seta present, close
to eye. Upper fronto-orbital seta present. Flagello-
mere 1 yellow, round. Palpus small, with setae nor-
mal, thick, moderately long, pointed. Color of ab-
dominal tergites mostly yellow, with some darker
markings. Venter of abdomen yellow. Venter of seg-
ment 3 bare. Venter of abdominal segment 5 lack-
ing dense setal combs, without sclerite. Ventrome-

dial setae of segment 6 present, long, thick, evenly
spaced, arranged in relatively straight line (Fig. 69).
Venter of segment 6 without sclerite. In cleared
specimens, dorsal abdominal glands invisible. In-
tersegment 6-7 with dark sclerite. Ovipositor dor-
sally narrow, anterodorsal portion broad, rounded
(Fig. 52). Anteroventral margin of ovipositor with
notch. Posterodorsal apex of ovipositor pointed. In
lateral view, posteroventral apex of ovipositor
straight. Posteroventral apex of ovipositor pointed.
Stylet without lateral barbs (Fig. 62). Dorsal sclerite
forked, on long stalk (Fig. 62).

GEOGRAPHICAL DISTRIBUTION. This spe-
cies is known from three middle elevation sites in
Costa Rica.

PHYLOGENETIC RELATIONSHIPS. Un-
known. This species fits most parsimoniously at the
base of the tree, with A. apivorus, but shares with
the A. borealis- group enlarged setulae on the hind
basitarsus.

DERIVATION OF SPECIFIC EPITHET. This
name is based on a Latin word for ancestor, reflect-
ing the large number of (possibly homoplastic)
primitive characters of this species.

HOLOTYPE. S, COSTA RICA: San Jose, Zur-
qui de Moravia, 10.05°N, 84.02°W, ix.1991, P.
Hanson, Malaise trap, 1600 m (LACM) [LACM
ENT 000689].

PARATYPES. COSTA RICA: Puntarenas, Las
Alturas, 8.95°N, 82.83°W, 3S, i.1992, P. Hanson,
Malaise trap, 1500 m (LACM), Monteverde,
10.1°N, 83.43°W, IS, l-10.iii.1992, D.M. Wood,
Malaise trap, 1500 m (LACM), San Jose, Zurqui
de Moravia, 10.05°N, 84.02°W, 19, ix-x.1990,
IS, x-xii.1990, IS, ii.1991, 3 9, iv.1991, 3 A, 49,
v.1992, 5S, vi.1992, 3 A, vii.1992, 1A, 19, iv-
v.1993, 1A, 19, l-15.vi.1993, 39, ix-x.1993, P.
Hanson, 19, 7-9.iii.1995, I. Bohorquez, Malaise
trap, 1600 m (LACM, MCZ, USNM, MICR, IN-
BIO, CMNH).

Apocephalus adustus Brown, 1993

(Figs. 2, 50)

SPECIES RECOGNITION. This species was pre-
viously known only from males; it is redescribed
here, including the hitherto unknown females.
These females will not key to Mesophora in Borg-
meier’s (1971) key to Apocephalus and lack nearly
all diagnostic characters of the subgenus. They can
be recognized by a combination of their notched
ovipositor (visible in cleared specimens only), some-
what flattened flagellomere 1 (Fig. 2), short, simple
ovipositor (Fig. 50), and generally dark color with
contrasting light halter.

NOTES ON VARIATION. An noted previously
(Brown, 1993), this species is extremely variable in
color. The length of frontal setae in females also is
highly variable. Some individuals have short frontal
setae and are lacking upper (but not lower) fronto-
orbital setae, whereas others have longer, complete
setation.

Contributions in Science, Number 462

Brown: Neotropical Apocephalus, Subgenus Mesophora ■ 17

DESCRIPTION.

General Characters. Body length 1. 4-2.1 mm.
Frons brown. Palpus yellow. Dorsum of thorax
dark brown. Pleural regions yellow, or light brown,
or dark brown. Scutellum of same color as dorsum
of thorax. Scutellum with short anterior and long
posterior seta. Legs yellowish-brown. Venter of tar-
somere 2 of mid leg with row of thin setulae and
thick apical setula. Hind femur yellowish-brown,
with abrupt, distinctive darkening at apex. Antero-
ventral row of enlarged setulae on hind basitarsus
straight, with all setulae short, subequal to those of
other rows. Wing vein Rs slender, subequal to or
thinner than costa. Wing vein R2+3 well developed.
Wing vein CuA3 reaches wing margin. Halter yel-
low.

Male. Lower fronto-orbital seta absent. Upper
fronto-orbital seta absent. Flagellomere 1 yellow to
brown, greatly enlarged, apically flattened, apical
two-thirds with narrowest part greater than one-
half basal width. Palpus small, with setae normal,
thick, moderately long, pointed. Mid leg with tar-
someres all slender. Costal vein not thickened. Ab-
dominal tergites dark brown. Venter of abdomen
yellow, or light brown, or dark brown. Venter of
segment 6 bare. Venter of segment 6 without scler-
ite. Epandrium globular, approximately as long as
high, with setae near cercus not markedly larger
than other epandrial setae. Left side of epandrium
with short setae confined to posterior margin. Right
side of epandrium with faint, unsclerotized dorso-
lateral ridge. Right surstylus short, rounded, lack-
ing ventral carina. Cercus yellow. Setae of cercus
and proctiger of normal size, longer and thicker
than short setae of epandrium.

Female. Lower fronto-orbital seta present, close
to eye (Fig. 2). Upper fronto-orbital seta present.
Flagellomere 1 yellow to brown, pyriform, flat-
tened. Palpus small, with setae normal, thick, mod-
erately long, pointed. Color of abdominal tergites
black, tergite 6 at least partly brown. Venter of ab-
domen white to gray. Venter of segment 3 bare.
Venter of abdominal segment 5 lacking dense setal
combs, without sclerite. Ventromedial setae of seg-
ment 6 present, short, thin, evenly spaced, arranged
in relatively straight line. Venter of segment 6 with-
out sclerite. In cleared specimens, dorsal abdominal
glands invisible. Intersegment 6-7 without distinct
sclerite. Ovipositor dorsally broad, anterodorsal
portion broad, rounded (Fig. 50). Anteroventral
margin of ovipositor with notch. Posterodorsal
apex of ovipositor pointed. In lateral view, postero-
ventral apex of ovipositor straight. Posteroventral
apex of ovipositor pointed. Stylet without lateral
barbs. Dorsal sclerite consisting of unforked, me-
dial process.

WAY OF LIFE. I reared this species from an un-
described species of stingless bee, which will be de-
scribed in a new genus (Paul Hanson, personal
communication), at Zurqui de Moravia, Costa
Rica. One bee was collected because of its abnor-
mal behavior: it was walking on the ground near

18 1 Contributions in Science, Number 462

the nest entrance and seemed unable to fly. A fur-
ther sample of 21 bees was collected, and of these
one was parasitized. I reared two and four flies,
respectively, from the two parasitized bees.

Apocephalus borealis- group
Apocephalus anfractus-sxibgroup

Apocephalus absentis Brown, 1993
(Figs. 4, 24-25)

NOTES. This species was previously known only
from females; below the species is redescribed, in-
cluding the hitherto unknown males.

SPECIES RECOGNITION. Both sexes of this
species can be recognized by the sinuate row of en-
larged setulae on the basal tarsomere of the hind
leg, the absence of wing vein R2+3, and the absence
of the lower fronto-orbital seta.

DESCRIPTION.

General Characters. Body length 0.8-1.05 mm.
Frons yellow to brown. Palpus yellow. Dorsum of
thorax yellow to light brown. Pleural regions yel-
low to light brown. Scutellum of same color as dor-
sum of thorax. Scutellum with short anterior and
long posterior seta. Legs yellowish-brown. Venter
of tarsomere 2 of mid leg with row of thin setulae
and thick apical setula. Hind femur yellowish-
brown, evenly colored throughout. Anteroventral
row of enlarged setulae on hind basitarsus sinuate,
with some setulae markedly longer than those of
other rows. Wing vein Rs slender, subequal to or
thinner than costa. Wing vein R2+3 absent. Wing
vein CuAa reaches wing margin. Halter brown.

Male. Lower fronto-orbital setal absent (Fig. 4).
Upper fronto-orbital seta present. Flagellomere 1
brown, slightly enlarged, apically flattened, apical
two-thirds with narrowest part greater than one-
half basal width. Palpus small, with setae normal,
thick, moderately long, pointed. Midleg with tar-
someres all slender. Costal vein not thickened. Ab-
dominal tergites dark brown. Venter of abdomen
light brown. Venter of segment 6 bare. Venter of
segment 6 without sclerite. Epandrium globular,
approximately as long as high, with large, promi-
nent seta near cercus. Left side of epandrium with
scattered, thin setae (Fig. 25). Right side of epan-
drium without ridge (Fig. 24). Right surstylus
short, rounded, lacking ventral carina. Cercus yel-
low. Setae of cercus and proctiger of normal size,
longer and thicker than short setae of epandrium.

Female. Lower fronto-orbital seta absent. Upper
fronto-orbital seta present. Flagellomere 1 brown,
pyriform. Palpus small, with setae normal, thick,
moderately long, pointed. Color of abdominal ter-
gites mostly brown, but tergite 6 yellow. Venter of
abdomen white. Venter of segment 3 bare. Venter
of abdominal segment 5 lacking dense setal combs,
without sclerite. Ventromedial setae of segment 6
present, long, thick, evenly spaced, arranged in rel-
atively straight line. Venter of segment 6 without
sclerite. In cleared specimens, dorsal abdominal

Brown: Neotropical Apocephalus, Subgenus Mesophora

glands visible, dark. Intersegment 6-7 without dis-
tinct sclerite. Ovipositor dorsally broad, anterodor-
sal portion broad, rounded. Anteroventral margin
of ovipositor with notch. Posterodorsal apex of ovi-
positor pointed. In lateral view, posteroventral apex
of ovipositor straight. Posteroventral apex of ovi-
positor pointed. Stylet without lateral barbs. Dorsal
sclerite consisting of two small processes.

GEOGRAPHICAL DISTRIBUTION. Besides
being known from several sites in Costa Rica, at
both middle and lower elevations, this species has
also been collected at Pakitza, Peru.

Apocepbalus bisetus Brown, 1993

NOTES ON SPECIES RECOGNITION. This
species, known only from a single female, was sep-
arated from A. trisetus Brown (Brown, 1993) based
on the presence of two lateral setae on the venter
of segment 6 and by a different stylet shape. Some
more recently collected specimens, however, have
three lateral setae like females of A. trisetus. A sum-
mary of the differences between the species, as I
recognize them, is as follows:

A. bisetus: 1) flagellomere 1 yellow throughout

2) 2-3 lateral setae on venter of seg-
ment 6

3) medial section of stylet short

A. trisetus: 1) flagellomere 1 yellow, with dark

apex

2) 3 lateral setae on venter of seg-
ment 6

3) medial section of stylet long

GEOGRAPHICAL DISTRIBUTION. This spe-
cies is known only from Costa Rica.

Other Apocepbalus borealis- group
Species

Apocepbalus megalops new species

(Figs. 5-6, 26-27, 53)

SPECIES RECOGNITION. This species can be
recognized most easily by the large eyes and narrow
frons.

DESCRIPTION.

General Characters. Body length 1.7-2 mm,
Frons yellow, narrow (Figs. 5-6). Palpus yellow.
Dorsum of thorax yellow. Pleural regions white.
Scutellum of same color as dorsum of thorax. Scu-
tellum with short anterior and long posterior seta.
Legs yellow. Venter of tarsomere 2 of mid leg with
row of thin setulae and thick apical setula. Hind
femur yellowish-brown, with abrupt, distinctive
darkening at apex. Anteroventral row of enlarged
setulae on hind basitarsus straight, with some setu-
lae markedly longer than those of other rows. Wing
vein Rs slender, subequal to or thinner than costa.
Wing vein R2+3 well developed. Wing vein CuA3
reaches wing margin. Halter yellow.

Male. Lower fronto-orbital seta absent (Fig. 5).

Upper fronto-orbital seta absent. Flagellomere 1
yellow, greatly enlarged, apically flattened, apical
two-thirds with narrowest part greater than one-
half basal width. Palpus small, with setae normal,
thick, moderately long, pointed. Midleg with tar-
someres all slender. Costal vein not thickened. Ab-
dominal tergites mostly yellow, with dark mark-
ings. Venter of abdomen yellow. Venter of segment
6 with distinct, black setae, ventral setae in a single
row. Venter of segment 6 without sclerite. Epan-
drium globular, approximately as long as high, with
setae near cercus not markedly larger than other
epandrial setae. Left side of epandrium with scat-
tered, thin setae (Fig. 27). Right side of epandrium
without ridge (Fig. 26). Right surstylus elongate,
rounded at apex, lacking ventral carina. Cercus yel-
low. Setae of cercus and proctiger of normal size,
longer and thicker than short setae of epandrium.

Female. Lower fronto-orbital seta absent (Fig. 6).
Upper fronto-orbital seta absent. Flagellomere 1
yellow, pyriform. Palpus small, with setae normal,
thick, moderately long, pointed. Color of abdomi-
nal tergites mostly yellow, with some darker mark-
ings. Venter of abdomen yellow. Venter of segment
3 bare. Venter of abdominal segment 5 lacking
dense setal combs, without sclerite. Ventromedial
setae of segment 6 present, long, thick, evenly
spaced, arranged in anteriorly pointed “V”. Venter
of segment 6 with sclerite present. In cleared spec-
imens, dorsal abdominal glands visible, dark. Inter-
segment 6-7 without distinct sclerite. Ovipositor
dorsally broad, anterodorsal portion broad, round-
ed (Fig. 53). Anteroventral margin of ovipositor
with notch. Posterodorsal apex of ovipositor point-
ed. In lateral view, posteroventral apex of oviposi-
tor straight. Posteroventral apex of ovipositor
pointed. Stylet with lateral barbs. Dorsal sclerite
forked, on long stalk.

GEOGRAPHICAL DISTRIBUTION. This spe-
cies is known from a single, middle elevation site
in Costa Rica.

PHYLOGENETIC RELATIONSHIPS. This spe-
cies is part of the A. borealis- group.

DERIVATION OF SPECIFIC EPITHET. The
name is based on Greek words for large eyes.

HOLOTYPE. A, COSTA RICA: San Jose, 26 km
N San Isidro, 9.5°N, 83.72°W, viii-ix.1991, P. Han-
son, Malaise trap, 2100 m (LACM) [LACM ENT
000609].

PARATYPES. COSTA RICA: San Jose, 26 km N
San Isidro, 9.5°N, 83.72°W, 2$, viii-ix.1991, 1A,
ii-v.1992, P. Hanson, Malaise trap, 2100 m
(LACM).

Apocepbalus emphysemas
new species

(Figs. 30-31, 68)

SPECIES RECOGNITION. The male of this spe-
cies is similar to the Nearctic Region species A. bo-
realis but differs by the dark brown coloring, the
enlarged first flagellomere, and the normal-sized

Contributions in Science, Number 462

Brown: Neotropical Apocepbalus, Subgenus Mesophora ■ 19

palpus. It can be separated from other brown Me-
sophora by the swollen costal vein (Fig. 68). The
female is unknown.

DESCRIPTION.

General Characters. Body length 1.5 mm. Frons
brown. Palpus brown. Dorsum of thorax light
brown. Pleural regions light brown. Scutellum of
same color as dorsum of thorax. Scutellum with
short anterior and long posterior seta. Legs yellow-
ish-brown. Venter of tarsomere 2 of mid leg with
row of thin setulae and thick apical setula. Hind
femur dark brown, evenly colored throughout, but
gradually darkened at apex. Anteroventral row of
enlarged setulae on hind basitarsus straight, with
some setulae markedly longer than those of other
rows. Wing vein Rs slender, thinner than costa.
Wing vein R2+3 well developed. Wing vein CuA!
reaches wing margin. Halter brown.

Male. Lower fronto-orbital seta absent. Upper
fronto-orbital seta absent. Flagellomere 1 yellowish
at base, apically darker, greatly enlarged, apically
flattened, apical two thirds with narrowest part
greater than one-half basal width. Palpus small,
with setae normal, thick, moderately long, pointed.
Mid leg with tarsomeres all slender. Costal vein
thickened between humeral crossvein and Rl. Ab-
dominal tergites dark brown. Venter of abdomen
light brown. Venter of segment 6 bare. Venter of
segment 6 without sclerite. Epandrium globular,
approximately as long as high, with setae near cer-
cus not markedly larger than other epandrial setae.
Left side of epandrium with thick, ventral setae
(Fig. 31). Right side of epandrium with prominent,
darkly sclerotized, dorsolateral ridge (Fig. 30).
Right surstylus short, rounded, lacking ventral Ca-
rina. Cercus brown. Setae of cercus reduced in size.

GEOGRAPHICAL DISTRIBUTION. Known
from a single site in Costa Rica.

PHYLOGENETIC RELATIONSHIPS. This spe-
cies is probably the sister-species of A. borealis.

DERIVATION OF SPECIFIC EPITHET. This
name is based on a Greek word for swelling, refer-
ring to the enlarged costa.

HOLOTYPE. <3, COSTA RICA: San Jose, Zur-
qui de Moravia, 10.05°N, 84.02°W, ix.1991, P.
Hanson, Malaise trap, 1600 m (LACM) [LACM
ENT 000646].

Apocephalus wheeleri- group
Apocephalus curtus- subgroup

Apocephalus lizanoi new species

(Figs. 20-21, 57, 67, 70)

SPECIES RECOGNITION. The male of this spe-
cies is extremely similar to that of A. curtus but
differs by the shape of the epandrium, which in A.
lizanoi has a well-defined lateral ridge (Figs. 20-
21), and by the less flared right surstylus in A. li-
zanoi (Fig. 20). The presumed female of this species
differs from that of A. curtus by the pattern of ven-

20 ■ Contributions in Science, Number 462

tral setation (Fig. 67) and by the structure of the
ovipositor (Fig. 57).

There is no direct evidence linking the female
specimens with the one known male, other than
that they were collected together. This species along
with A. curtus and lemniscus new species (see be-
low) form a species complex that requires further
attention. Rearing a series of males and females
would be helpful, but the hosts are still unknown.
DESCRIPTION.

General Characters. Body length 1.7-2. 3 mm.
Frons yellow. Palpus yellow. Dorsum of thorax yel-
low. Pleural regions yellow. Scutellum of same color
as dorsum of thorax. Scutellum with short anterior
and long posterior seta. Legs yellow. Venter of tar-
somere 2 of mid leg with row of thin setulae and
thick apical setula. Hind femur yellowish-brown,
evenly colored throughout. Anteroventral row of
enlarged setulae on hind basitarsus straight, with
all setulae short, subequal to those of other rows.
Wing vein Rs slender, subequal to or thinner than
costa. Wing vein R2+3 well developed. Wing vein
CuAj reaches wing margin. Halter brown.

Male. Lower fronto-orbital seta absent. Upper
fronto-orbital seta absent. Flagellomere 1 yellow,
greatly enlarged, apically flattened, apical two-
thirds with narrowest part greater than one-half
basal width. Palpus small, with setae normal, thick,
moderately long, pointed. Mid leg with tarsomeres
all slender. Costal vein not thickened. Abdominal
tergites mostly dark, with yellow markings. Venter
of abdomen white. Venter of segment 6 with dis-
tinct, black setae, ventral setae in a single row (Fig.
70). Venter of segment 6 without sclerite. Epan-
drium globular, band-shaped, shorter than height,
with setae near cercus not markedly larger than
other epandrial setae. Left side of epandrium with
scattered, thin setae (Fig. 21). Right side of epan-
drium with faint, unsclerotized dorsolateral ridge
(Fig. 20). Right surstylus short, rounded, lacking
ventral carina. Cercus brown. Setae of cercus and
proctiger of normal size, longer and thicker than
short setae of epandrium.

Female. Lower fronto-orbital seta absent. Upper
fronto-orbital seta present. Flagellomere 1 yellow,
greatly enlarged, subequal to eye height, apically
flattened. Palpus small, with setae normal, thick,
moderately long, pointed. Color of abdominal ter-
gites mostly yellow, with some darker markings.
Venter of abdomen yellow. Venter of segment 3
with few dark setae (Fig. 67). Venter of abdominal
segment 4 with dense combs of setae. Venter of ab-
dominal segment 5 with dense combs of setae, with
sclerite present. Venter of abdominal segment 6
with short median spine, without lateroventral
patch of setae. Ventromedial setae of segment 6
present, short, thick, evenly spaced, arranged in rel-
atively straight line. Venter of segment 6 with scler-
ite present. In cleared specimens, dorsal abdominal
glands visible, dark. Intersegment 6-7 with dark
sclerite. Ovipositor dorsally broad, anterodorsal
portion broad, rounded (Fig. 57). Anteroventral

Brown: Neotropical Apocephalus, Subgenus Mesophora

margin of ovipositor with deep, posterior projec-
tion. Posterodorsal apex of ovipositor rounded. In
lateral view, posteroventral apex of ovipositor
straight. Posteroventral apex of ovipositor round-
ed. Stylet without lateral barbs. Dorsal sclerite con-
sisting of unforked, medial process.

GEOGRAPHICAL DISTRIBUTION. Known
from three middle elevation sites in Costa Rica.

PHYLOGENETIC RELATIONSHIPS. This spe-
cies is closely related to A. curtus and A. lemniscus
new species (see below).

DERIVATION OF SPECIFIC EPITHET. This
species is dedicated to Sefior Jorge Arturo Lizano,
owner of the property at the incredibly diverse Zur-
qui de Moravia site.

HOLOTYPE. A, COSTA RICA: Puntarenas, Las
Alturas, 8.95°N, 82.83°W, x.1991, P. Hanson, Mal-
aise trap, 1500 m (LACM) [LACM ENT 000859].

PARATYPES. COSTA RICA: Puntarenas, Las
Alturas, 8.95°N, 82.83°W, 1 9, x.1991, 7$, i.1992,
5 9 , v.1992, 3 9 , vi. 1992, P. Hanson, Malaise trap,
1500 m (LACM, MCZ, USNM, MICR, INBIO),
San Jose, Braulio Carrillo National Park, 10.12°N,
83.97°W, 29, vi-xi.1990, P. Hanson, Malaise trap,
1000 m (LACM), Zurqui de Moravia, 10.05°N,
84.02°W, 19, v.1992, P. Hanson, Malaise trap,
1600 m (LACM).

Apocepbalus curtus Brown, 1993
(Figs. 56, 60-61, 66)

SPECIES RECOGNITION. Further coincident
collections, along with other subtle character agree-
ments with male specimens, confirmed my suspi-
cion that Apocephalus species female C (of Brown,
1993) is the unnamed female of A. curtus. Females
of this species are recognized by the dense patches
of thick, black setae on the venter of the abdomen
(Fig. 66), which are patterned differently than those
of A. lizanoi (Fig. 67).

NOTES ON VARIATION. One male specimen
has the ventral setae of the abdomen, which are
usually elongate (as Fig. 70), reduced to tiny stubs.

DESCRIPTION.

General Characters. Body length 1.9-2. 2 mm.
Frons yellow. Palpus yellow. Dorsum of thorax
light brown. Pleural regions yellow. Scutellum of
same color as dorsum of thorax to darker than and
contrasting with dorsum of thorax. Scutellum with
short anterior and long posterior seta. Legs yellow.
Venter of tarsomere 2 of mid leg with row of thin
setulae and thick apical setula. Hind femur yellow-
ish-brown, evenly colored throughout. Anteroven-
tral row of enlarged setulae on hind basitarsus
straight, with all setulae short, subequal to those of
other rows. Wing vein Rs slender, subequal to or
thinner than costa. Wing vein R2+3 well developed.
Wing vein CuAa reaches wing margin. Halter
brown.

Male. Lower fronto-orbital seta absent. Upper
fronto-orbital seta absent. Flagellomere 1 yellow,
greatly enlarged, apically flattened, apical two-

thirds with narrowest part greater than one-half
basal width. Palpus enlarged, with setae reduced,
thin, short, stubby. Mid leg with tarsomeres all
slender. Costal vein not thickened. Abdominal ter-
gites mostly dark, with yellow markings. Venter of
abdomen white. Venter of segment 6 with distinct,
black setae, ventral setae in a single row. Venter of
segment 6 without sclerite. Epandrium globular,
band-shaped, shorter than height, with setae near
cercus not markedly larger than other epandrial se-
tae. Left side of epandrium with scattered, thin se-
tae. Right side of epandrium with faint, unsclero-
tized dorsolateral ridge. Right surstylus truncate,
apically flared, lacking ventral carina. Cercus
brown. Setae of cercus and proctiger of normal size,
longer and thicker than short setae of epandrium.

Female. Lower fronto-orbital seta absent. Upper
fronto-orbital seta present. Flagellomere 1 yellow,
pyriform. Palpus small, with setae normal, thick,
moderately long, pointed. Color of abdominal ter-
gites mostly yellow, with some darker markings.
Venter of abdomen yellow. Venter of segment 3
bare. Venter of abdominal segment 4 lacking dense
setal combs. Venter of abdominal segment 5 with
dense combs of setae, without sclerite (Fig. 66).
Venter of abdominal segment 6 without short me-
dian spine, with lateroventral patch of setae. Ven-
tromedial setae of segment 6 present, short, thick,
in medial group, arranged in relatively straight line.
Venter of segment 6 with sclerite present. In cleared
specimens, dorsal abdominal glands visible, dark.
Intersegment 6-7 with dark sclerite. Ovipositor
dorsally broad, anterodorsal portion broad, round-
ed (Fig. 56). Antero ventral margin of ovipositor
with deep, posterior projection. Posterodorsal apex
of ovipositor rounded. In lateral view, posteroven-
tral apex of ovipositor straight. Posteroventral apex
of ovipositor extremely widely rounded. Stylet
without lateral barbs (Figs. 60-61). Dorsal sclerite
consisting of unforked, medial process (Fig. 61).

GEOGRAPF1ICAL DISTRIBUTION. This spe-
cies is known only from Zurqui de Moravia in San
Jose Province, Costa Rica.

PHYLOGENETIC RELATIONSHIPS. This spe-
cies and A. lemniscus might be most closely related,
based on the flared right surstylus. Discovery of the
female of A. lemniscus is necessary to allow further
resolution of these relationships.

Apocephalus lemniscus
new species

(Figs. 3, 22-23)

SPECIES RECOGNITION. The head of this spe-
cies (Fig. 3) is strikingly different from A. curtus
and A. lizanoi: flagellomere 1 is relatively small and
dark brown in color, and the upper fronto-orbital
seta is present. In A. lemniscus, the venter of seg-
ment 6 has two rows of large setae, rather than a
single row as found in A. curtus and lizanoi (Fig.
70). The pattern of epandrial setation and shape of

Contributions in Science, Number 462

Brown: Neotropical Apocepbalus, Subgenus Mesophora 121

the surstylus also is different from those of the oth-
er two species. The female is unknown.
DESCRIPTION.

General Characters. Body length 2 mm. Frons
yellow. Palpus yellow. Dorsum of thorax light
brown. Pleural regions light brown. Scutellum of
same color as dorsum of thorax. Scutellum with
short anterior and long posterior seta. Legs yellow-
ish-brown. Venter of tarsomere 2 of mid leg with
row of thin setulae and thick apical setula. Hind
femur yellowish-brown, evenly colored throughout.
Anteroventral row of enlarged setulae on hind ba-
sitarsus straight, with all setulae short, subequal to
those of other rows. Wing vein Rs slender, subequal
to or thinner than costa. Wing vein R2+3 well de-
veloped. Wing vein CuAj reaches wing margin.
Halter brown.

Male. Lower fronto-orbital seta absent (Fig. 3).
Upper fronto-orbital seta present. Flagellomere 1
brown, pyriform, apical two-thirds with narrowest
part greater than one-half basal width. Palpus
small, with setae normal, thick, moderately long,
pointed. Mid leg with tarsomeres all slender. Costal
vein not thickened. Abdominal tergites mostly
dark, with yellow markings. Venter of abdomen
yellow. Venter of segment 6 with distinct, black se-
tae, ventral setae in two rows. Venter of segment 6
without sclerite. Epandrium globular, band-shaped,
shorter than height, with setae near cercus not
markedly larger than other epandrial setae. Left
side of epandrium with scattered, thin setae. Right
side of epandrium without ridge. Right surstylus
truncate, apically flared, lacking ventral carina.
Cercus brown. Setae of cercus and proctiger of nor-
mal size, longer and thicker than short setae of
epandrium.

GEOGRAPHICAL DISTRIBUTION. Known
from a single site in Costa Rica.

PHYLOGENETIC RELATIONSHIPS. This spe-
cies possibly is the sister-taxon of A. curtus.

DERIVATION OF SPECIFIC EPITHET. This
name is derived from a Latin word for band or
ribbon, referring to the short, narrow epandrium.

HOLOTYPE. c3, COSTA RICA: San Jose, 6 km
N San Gerardo, 2800 m, 9.95°N, 84.05°W,
viii.1993, P. Hanson, Malaise trap (LACM)
[LACM ENT 001433].

Apocephalus wheeleri- subgroup

Apocephalus niveus new species

(Figs. 32-33, 58)

SPECIES RECOGNITION. This species can be
recognized in part by the distinctive white color of
flagellomere 1 (in females only), the legs, and pleu-
ral region.

DESCRIPTION.

General Characters. Body length 1.4-1. 7 mm.
Frons yellow. Palpus yellow. Dorsum of thorax
light brown. Pleural regions white. Scutellum of
same color as dorsum of thorax. Scutellum with
short anterior and long posterior seta. Legs white.

22 ■ Contributions in Science, Number 462

Venter of tarsomere 2 of mid leg with row of thin
setulae and thick apical setula, or with four en-
larged setulae besides apical, or with two enlarged
setulae, one apical, one at midpoint. Hind femur
white, evenly colored throughout. Anteroventral
row of enlarged setulae on hind basitarsus straight,
with all setulae short, subequal to those of other
rows. Wing vein Rs slender, subequal to or thinner
than costa. Wing vein R2+3 well developed. Wing
vein CuAj reaches wing margin. Halter brown.

Male. Lower fronto-orbital seta absent. Upper
fronto-orbital seta absent. Flagellomere 1 yellowish
at base, apically darker, greatly enlarged, apically
flattened, apical two-thirds with narrowest part
greater than one-half basal width. Palpus small,
with setae normal, thick, moderately long, pointed.
Mid leg with tarsomeres all slender. Costal vein not
thickened. Abdominal tergites dark gray, with ter-
gite 6 yellow. Venter of abdomen white. Venter of
segment 6 bare. Venter of segment 6 without scler-
ite. Epandrium globular, approximately as long as
high, with setae near cercus not markedly larger
than other epandrial setae. Left side of epandrium
with scattered, thin setae (Fig. 33). Right side of
epandrium with faint, unsclerotized dorsolateral
ridge (Fig. 32). Right surstylus short, rounded,
lacking ventral carina. Cercus brown. Setae of cer-
cus and proctiger of normal size, longer and thicker
than short setae of epandrium.

Female. Lower fronto-orbital seta absent. Upper
fronto-orbital seta present. Flagellomere 1 white,
pyriform. Palpus small, with setae normal, thick,
moderately long, pointed. Color of abdominal ter-
gites mostly brown, but tergite 6 yellow. Venter of
abdomen white. Venter of segment 3 bare. Venter
of abdominal segment 5 lacking dense setal combs,
without sclerite. Ventromedial setae of segment 6
present, long, thick, in lateral group of three, ar-
ranged in anteriorly pointed “V”. Venter of segment
6 with sclerite present. In cleared specimens, dorsal
abdominal glands visible, dark. Intersegment 6-7
without distinct sclerite. Ovipositor dorsally nar-
row (Fig. 58), anterodorsal portion broad, round-
ed. Anteroventral margin of ovipositor smooth,
without notch. Posterodorsal apex of ovipositor
rounded. In lateral view, posteroventral apex of
ovipositor straight. Posteroventral apex of ovipos-
itor rounded. Stylet without lateral barbs. Dorsal
sclerite forked, on long stalk.

GEOGRAPHICAL DISTRIBUTION. This spe-
cies is known from low and middle elevation sites
in Costa Rica.

PHYLOGENETIC RELATIONSHIPS. Based on
the presence of thick, dark seta on the venter of
segment 6 in females, this species is hypothesized
to be the sister taxon of the rest of the A. wbeeleri-
subgroup.

DERIVATION OF SPECIFIC EPITHET. This
name is based on a Latin word for snow-white, re-
ferring to the white-colored body parts.

HOLOTYPE. A, COSTA RICA: Limon, 7 km W
Bribri, 9.58°N, 82.88°W, ix-xi.1989, P. Hanson,

Brown: Neotropical Apocephalus, Subgenus Mesophora

Malaise trap, 50 m (LACM) [LACM ENT
000802].

PARATYPES. COSTA RICA: Limon, 7 km W
Bribri, 9.58°N, 82.88°W, 3<J, 3?, ix-xi.1989, P.
Hanson, Malaise trap, 50 m, altered forest
(LACM), 16 km W Guapiles, 10.15°N, 83.92°W,
1<5, iii-v. 1990, 19, i-iv.1991, P. Hanson, Malaise
trap, 400 m (LACM), San Jose, Zurqui de Mora-
via, 10.05°N, 84.02°W, Id, v.1992 (LACM).

Apocepbalus antennatus-'mirngroup

Apocephalus antennatus
Malloch, 1913
(Figs. 7, 15)

Apocephalus leptotarsus Brown, 1993 new synon-

ymy

NOTES ABOUT SYNONYMY. My previous
description of A. leptotarsus as a separate species
was based on an error: the surstylus of this species
does not bear a medial carina.

NOTES ABOUT VARIATION. Some specimens
from Costa Rica assigned to this species have the
tip of the hind femur darkened, whereas North
American specimens, and those from Colombia,
lack this darkening. All other species of Mesophora
are consistently one or the other, so this difference
might represent some real distinction between the
Costa Rican and other species. I can find no struc-
tural characters to support this differentiation,
however.

Also, some specimens have no extra setulae on
the venter of mid tarsomere 2. This is not correlat-
ed with any other variation, but it adds weight to
the idea that A. antennatus is a highly variable spe-
cies or a cryptic species complex.

NOTES ABOUT DISTRIBUTION. Previously, I
tentatively concluded that this species was not lim-
ited to the Nearctic Region. Further collections
show this to be true; there are now many specimens
from the Neotropical Region: from Costa Rica, the
Dominican Republic, and Mexico.

WAY OF LIFE. Nearctic Region hosts of this spe-
cies, several species of fireflies (Coleoptera: Lam-
pyridae), were listed previously (Brown, 1994a).
Recently, I also reared A. antennatus from a diur-
nal, non-luminous species of Lucidota (Coleoptera:
Lampyridae) at Monteverde, Costa Rica. This was
a different firefly species from L. atra, a host I re-
ported previously.

Apocephalus longistylus
Brown, 1993
(Fig. 16)

NOTES ABOUT DESCRIPTION. The female of
A. longistylus lacks the lower fronto-orbital seta,
not the “middle” (“upper) fronto-orbital seta, as I
originally stated (Brown, 1993).

WAY OF LIFE. I reared adult males of this spe-
cies from a species of Bicellonycha (Lampyridae) at
Zurqui de Moravia, Costa Rica.

Apocephalus wheeleri- infragroup

Apocephalus mortifer
Borgmeier, 1937

NEW MATERIAL EXAMINED. This species,
previously known only from Brazil, was collected
in Central America: Id, COSTA RICA: San Jose,
Zurqui de Moravia, 10.05°N, 84.02°W, ix.1991, P.
Hanson, Malaise trap, 1600 m (LACM).

Apocephalus tritarsus
Brown, 1993

NOTES ON VARIATION. Newly collected Cos-
ta Rican specimens of this species differ from the
original specimens from Mexico by the dark brown
color of the halter and cercus.

WAY OF LIFE. I reared this species from a di-
urnal, non-luminous lampyrid of the genus Luci-
dota at Monteverde, Costa Rica. The beetle was a
different species than that from which A. antenna-
tus was reared (see above).

Unplaced Species

Apocephalus micrepelis
Brown, 1993

NOTES ABOUT DESCRIPTION. This species
was previously known only from the holotype
male. Additional specimens (CMNH, LACM) have
brought to light a previously unnoticed character:
males have a few ventral setae on abdominal seg-
ment 6.

Apocephalus pilatus new species
(Figs. 34-35)

SPECIES RECOGNITION. Males of this species
are similar to those of A. antennatus, but the palpal
setae are not shortened, the surstylus is more
rounded, and the large setae near the cercus are
more posteroventrally placed. The female is un-
known.

DESCRIPTION.

General Characters. Body length 1.5-1. 8 mm.
Frons yellow. Palpus yellow. Dorsum of thorax yel-
low. Pleural regions yellow. Scutellurn of same color
as dorsum of thorax. Scutellurn with short anterior
and long posterior seta. Legs yellow. Venter of tar-
somere 2 of mid leg with row of thin setulae and
thick apical setula. Hind femur yellowish-brown,
with abrupt, distinctive darkening at apex. Antero-
ventral row of enlarged setulae on hind basitarsus
straight, with all setulae short, subequal to those of
other rows. Wing vein Rs slender, subequal to or
thinner than costa. Wing vein R2+3 well developed.
Wing vein CuA, reaches wing margin. Halter yel-
low.

Male. Lower fronto-orbital seta absent. Upper
fronto-orbital seta absent. Flagellomere 1 yellow,
greatly enlarged, apical! y flattened, apical two-
thirds with narrowest part greater than one-half
basal width. Palpus small, with setae normal, thick,

Contributions in Science, Number 462

Brown: Neotropical Apocepbalus, Subgenus Mesophora ■ 23

moderately long, pointed. Mid leg with tarsomeres
all slender. Costal vein not thickened. Abdominal
tergites mostly yellow, with dark markings. Venter
of abdomen yellow. Venter of segment 6 bare. Ven-
ter of segment 6 without sclerite. Epandrium glob-
ular, approximately as long as high, with large,
prominent seta near cercus. Left side of epandrium
with scattered, thin setae (Fig. 35). Right side of
epandrium with prominent, darkly sclerotized, dor-
solateral ridge (Fig. 34). Right surstylus short,
rounded, lacking ventral carina. Cercus yellow. Se-
tae of cercus and proctiger of normal size, longer
and thicker than short setae of epandrium.

GEOGRAPHICAL DISTRIBUTION. This spe-
cies is known only from the Dominican Republic.

PHYLOGENETIC RELATIONSHIPS. Un-
known.

DERIVATION OF SPECIFIC EPITHET. The
name of this species is derived from a Latin word
for dense, referring to the number of setae on the
left side of the epandrium.

HOLOTYPE. <3, DOMINICAN REPUBLIC:
Pedernales, Las Abejas, 38 km NNW Cabo Rojo,
18.13°N, 72.63°W, 15.vii. 1987, J. Rawlins, R. Da-
vidson, Malaise trap, 1250 m (CMNH) [LACM
ENT 000897].

PARATYPES. DOMINICAN REPUBLIC: Ped-
ernales, Las Abejas, 38 km NNW Cabo Rojo,
18.13°N, 71.63°W, 2A, 15.vii.1987, J. Rawlins, R.
Davidson, Malaise trap, 1250 m (CMNH, LACM),
3.3 km NE Los Arroyos, 18.25°N, 71,75°W, Id,
1 6-1 8.vii. 1990, L. Masner et ai, sweep samples,
wet montane forest, 1450 m (CMNH).

Apocephalus crassus new species

(Figs. 40-41)

SPECIES RECOGNITION. This is another spe-
cies whose male is similar to that of A. antennatus,
but it differs by the sparse, thicker setae on the left
side of the epandrium and the absence of a large
seta near the cercus. The female is unknown.

DESCRIPTION.

General Characters. Body length 2.4 mm. Frons
yellow. Palpus yellow. Dorsum of thorax yellow.
Pleural regions yellow. Scutellum of same color as
dorsum of thorax. Scutellum with short anterior
and long posterior seta. Legs yellow. Venter of tar-
somere 2 of mid leg with row of thin setulae and
thick apical setula. Hind femur yellowish-brown,
with abrupt, distinctive darkening at apex. Antero-
ventral row of enlarged setulae on hind basitarsus
straight, with all setulae short, subequal to those of
other rows. Wing vein Rs slender, subequal to or
thinner than costa. Wing vein R2+3 well developed.
Wing vein CuAj reaches wing margin. Halter yel-
low.

24 ■ Contributions in Science, Number 462

moderately long, pointed. Midleg with tarsomeres
all slender. Costal vein not thickened. Abdominal
tergites mostly yellow, with dark markings. Venter
of abdomen yellow. Venter of segment 6 bare. Ven-
ter of segment 6 without sclerite. Epandrium glob-
ular, approximately as long as high, with setae near
cercus not markedly larger than other epandrial se-
tae. Left side of epandrium with thick, ventral setae
(Fig. 41). Right side of epandrium with prominent,
darkly sclerotized, dorsolateral ridge (Fig. 40).
Right surstylus short, rounded, lacking ventral ca-
rina. Cercus yellow. Setae of cercus and proctiger
of normal size, longer and thicker than short setae
of epandrium.

GEOGRAPHICAL DISTRIBUTION. This spe-
cies is known only from the Dominican Republic.

PHYLOGENETIC RELATIONSHIPS. Un-
known.

DERIVATION OF SPECIFIC EPITHET. The
name is from a Latin word for thick, referring to
the enlarged setae of the left side of the epandrium.

HOLOTYPE A, DOMINICAN REPUBLIC:
Pedernales, 3.3 km NE Los Arroyos, 18.25°N,
71.75°W, 16-18.vii.1990, L. Masner et ai, sweep
samples, wet montane forest, 1450 m (CMNH)
[LACM ENT 000895].

Apocephalus prolixus
new species
(Figs. 42-43)

SPECIES RECOGNITION. Males of this species
resemble those of A. longistylus and A. prolatus by
their elongate right surstylus but differ by having a
small palpus and no ventral abdominal setae. The
epandrium is greatly depressed, like that of A. in-
sulanus (Brown, 1993, figs. 5, 26), but unlike A.
insulanus there are large setae on both sides of the
epandrium (Figs. 42-43). The female is unknown.

DESCRIPTION.

General Characters. Body length 1.5-1. 8 mm.
Frons yellow. Palpus yellow. Dorsum of thorax
light brown. Pleural regions yellow. Scutellum of
same color as dorsum of thorax. Scutellum with
short anterior and long posterior seta. Legs yellow.
Venter of tarsomere 2 of mid leg with row of thin
setulae and thick apical setula. Hind femur yellow-
ish-brown, evenly colored throughout. Anteroven-
tral row of enlarged setulae on hind basitarsus
straight, with all setulae short, subequal to those of
other rows. Wing vein Rs slender, subequal to or
thinner than costa. Wing vein R2+3 well developed.
Wing vein CuA! reaches wing margin. Halter yel-
low to brown.

Brown: Neotropical Apocephalus, Subgenus Mesophora

gites mostly yellow, with dark markings. Venter of
abdomen white. Venter of segment 6 bare. Venter
of segment 6 without sclerite. Epandrium markedly
depressed, approximately as long as high, with
large, prominent seta near cercus. Left side of epan-
drium with scattered, thin setae (Fig. 43). Right side
of epandrium with prominent, darkly sclerotized,
dorsolateral ridge (Fig. 42). Right surstylus elon-
gate, rounded at apex, lacking ventral carina. Cer-
cus yellow. Setae of cercus and proctiger of normal
size, longer and thicker than short setae of epan-
drium.

GEOGRAPHICAL DISTRIBUTION. This spe-
cies is known only from the Dominican Republic.

PHYLOGENETIC RELATIONSHIPS. Un-
known.

DERIVATION OF SPECIFIC EPITHET. The
name is from a Latin word for elongate, referring
to the long epandrium.

HOLOTYPE. S, DOMINICAN REPUBLIC:
Pedernales, Las Abejas, 38 km NNW Cabo Rojo,
18.13°N, 71.63°W, Id, 15.vii.1987, J. Rawlins, R.
Davidson, Malaise trap, 1250 m (CMNH) [LACM
ENT 000888].

PARATYPE. DOMINICAN REPUBLIC: Peder-
nales, 3.3 km NE Los Arroyos, 18.25°N, 71.75°W,
lc3, 16-18.vii.1990, L. Masner et al., sweep sam-
ples, wet montane forest, 1450 m (LACM).

Apocephalus secundus new species

(Figs. 36-37)

SPECIES RECOGNITION. The male of this spe-
cies can be recognized by the enlarged palpus and
the enlarged setula on the second tarsomere of the
mid leg. The female is unknown.

DESCRIPTION.

General Characters. Body length 1.7 mm. Frons
yellow. Palpus yellow. Dorsum of thorax yellow.
Pleural regions yellow. Scutellum of same color as
dorsum of thorax. Scutellum with short anterior
and long posterior seta. Legs yellow. Venter of tar-
somere 2 of mid leg with row of thin setulae and
thick, elongate, apical setula. Hind femur yellow-
ish-brown, with abrupt, distinctive darkening at
apex. Anteroventral row of enlarged setulae on
hind basitarsus straight, with all setulae short, sub-
equal to those of other rows. Wing vein Rs slender,
subequal to or thinner than costa. Wing vein R2+3
well developed. Wing vein CuAj reaches wing mar-
gin. Halter yellow.

cus not markedly larger than other epandrial setae.
Left side of epandrium with scattered, thin setae
(Fig. 37). Right side of epandrium with prominent,
darkly sclerotized, dorsolateral ridge (Fig. 36).
Right surstylus short, rounded, lacking ventral ca-
rina. Cercus yellow. Setae of cercus and proctiger
of normal size, longer and thicker than short setae
of epandrium.

GEOGRAPHICAL DISTRIBUTION. This spe-
cies is known only from the Dominican Republic.

PHYLOGENETIC RELATIONSHIPS. Un-
known.

DERIVATION OF SPECIFIC EPITHET. The
name is a Latin word for second, referring to the
second tarsomere of the mid leg, which has an elon-
gate apical seta.

HOLOTYPE. d, DOMINICAN REPUBLIC:
Pedernales, Las Abejas, 38 km NNW Cabo Rojo,
18.13°N, 71.63°W, 1(3, 15.vii.1987, J. Rawlins, R.
Davidson, Malaise trap, 1250 m (CMNH) [LACM
ENT 000893].

Apocephalus echinatus
new species

(Figs. 38-39)

SPECIES RECOGNITION. Males of this species
are recognized by the extremely dense setae on the
epandrium. The female is unknown.
DESCRIPTION.

General Characters. Body length 1.8-2. 2 mm.
Frons yellow. Palpus yellow. Dorsum of thorax yel-
low. Pleural regions yellow. Scutellum of same color
as dorsum of thorax. Scutellum with short anterior
and long posterior seta. Legs yellow. Venter of tar-
somere 2 of mid leg bare. Hind femur yellowish-
brown, evenly colored throughout. Anteroventral
row of enlarged setulae on hind basitarsus straight,
with all setulae short, subequal to those of other
rows. Wing vein Rs slender, subequal to or thinner
than costa. Wing vein R2+3 well developed. Wing
vein CuA3 reaches wing margin. Halter yellow.

Male. Lower fronto-orbital seta absent. Upper
fronto-orbital seta absent. Flagellomere 1 yellow,
greatly enlarged, apically flattened, apical two-
thirds with narrowest part greater than one-half
basal width. Palpus small, with setae normal, thick,
moderately long, pointed. Mid leg with tarsomeres
all slender. Costal vein not thickened. Abdominal
tergites mostly dark, with yellow markings. Venter
of abdomen yellow. Venter of segment 6 bare. Ven-
ter of segment 6 without sclerite. Epandrium glob-
ular, approximately as long as high, with setae near
cercus not markedly larger than other epandrial se-
tae. Left side of epandrium with dense, thicker setae
(Fig. 39). Right side of epandrium without ridge
(Fig. 38). Right surstylus short, rounded, lacking
ventral carina. Cercus yellow. Setae of cercus and
proctiger of normal size, longer and thicker than
short setae of epandrium.

GEOGRAPHICAL DISTRIBUTION. This spe-
cies is known only from the Dominican Republic.

Contributions in Science, Number 462

Brown: Neotropical Apocephalus, Subgenus Mesophora ■ 25

PHYLOGENETIC RELATIONSHIPS. Un-
known.

DERIVATION OF SPECIFIC EPITHET. This
name is based on a Latin word for spiny, referring
to the dense setae on the left side of the epandrium.

HOLOTYPE. <J, DOMINICAN REPUBLIC:
Pedernales, 5 km NE Los Arroyos, 18.25°N,
71.75°W, 33d, 17~18.vii.1990, C. Young et al.
Malaise trap (CMNH) [LACM ENT 001155].

PARATYPES. DOMINICAN REPUBLIC: Inde-
pendence, Sierra de Neiba, 5.5 km NNW Angel Fel-
iz, 18.68°N, 71.78°W, Id, 21-22.vii.1992, J. Rawlins
et al. Malaise trap, dense cloud forest, 1750 m
(CMNH), Pedernales, 3.3 km NE Los Arroyos,
18.25°N, 71.75°W, Id, 16-18.vii.1990, L. Masner et
al, sweep samples, wet montane forest, 1450 m
(CMNH, LACM), 5 km NE Los Arroyos, 18.25°N,
71.75°W, 32d, 17-18.vii.1990, C. Young et al. Mal-
aise trap (CMNH, LACM, USNM, MCZ).

Species Recognized but Not Named

The following species are represented by a single
sex only: males of the A. anfractus-group or fe-
males of the other groups.

Phorid Species 3251 9
(Fig. 11)

A. limai Prado; Brown, 1993 (misidentification)

SPECIES RECOGNITION. The two specimens
of this unnamed species have a different ovipositor
structure than that of A. limai. My previous use of
the name was in part because both taxa lack the
upper fronto-orbital setae, but this character state
is now known to be fairly widespread. This species
belongs in the A. borealis- group.

DESCRIPTION.

General Characters. Body length 2.9-3. 1 mm.
Frons yellow. Palpus yellow. Dorsum of thorax yel-
low. Pleural regions yellow. Scutellum of same color
as dorsum of thorax. Scutellum with short anterior
and long posterior seta. Legs yellow. Venter of tar-
somere 2 of mid leg with row of thin setulae and
thick apical setula. Hind femur yellowish-brown,
with abrupt, distinctive darkening at apex. Antero-
ventral row of enlarged setulae on hind basitarsus
straight, with some setulae markedly longer than
those of other rows. Wing vein Rs slender, subequal
to or thinner than costa. Wing vein R2+3 well de-
veloped. Wing vein CuAj reaches wing margin.
Halter yellow.

Female. Lower fronto-orbital seta absent (Fig.
11). Upper fronto-orbital seta absent. Flagellomere
1 yellow, greatly enlarged, subequal to eye height,
apically flattened. Palpus small, with setae normal,
thick, moderately long, pointed. Color of abdomi-
nal tergites yellow. Venter of abdomen yellow. Ven-
ter of segment 3 bare. Venter of abdominal segment
5 lacking dense setal combs, without sclerite. Ven-
tromedial setae of segment 6 present, long, thick,
evenly spaced, arranged in relatively straight line.

26 ■ Contributions in Science, Number 462

Venter of segment 6 without sclerite. In cleared
specimens, dorsal abdominal glands visible, dark.
Intersegment 6-7 with dark sclerite. Ovipositor
dorsally broad, anterodorsal portion broad, round-
ed. Anteroventral margin of ovipositor with notch.
Posterodorsal apex of ovipositor pointed. In lateral
view, posterovental apex of ovipositor straight.
Posteroventral apex of ovipositor pointed. Stylet
without lateral barbs. Dorsal sclerite forked, on
long stalk.

MATERIAL EXAMINED. PANAMA: Chiriqui,
Potrerillos, 19, 25.vii.1964, A. Broce, light trap,
975 m (USNM) [LACM ENT 000638]; VENE-
ZUELA: Yacambu, 19, lO.v.1981, H.K. Townes,
365 m (LACM) [LACM ENT 000629].

Phorid Species 3223 9

(Figs. 12, 54)

Species recognition. This is one of two species
that have ovipositors closely resembling those of A.
borealis. Unlike females of A. borealis, species 3223
is darker in color and has flagellomere 1 enlarged
and apically flattened.

DESCRIPTION.

General Characters. Body length 2.2 mm. Frons
yellow. Palpus yellow. Dorsum of thorax yellow.
Pleural regions yellow. Scutellum of same color as
dorsum of thorax. Scutellum with short anterior
and long posterior seta. Legs yellowish-brown.
Venter of tarsomere 2 of mid leg with row of thin
setulae and thick apical setula. Hind femur yellow-
ish-brown, with abrupt, distinctive darkening at
apex. Anteroventral row of enlarged setulae on
hind basitarsus straight, with basal setulae mark-
edly longer than those of other rows. Wing vein Rs
slightly swollen, thicker than costa. Wing vein R2+3
thin. Wing vein CuA2 reaches wing margin. Halter
brown.

Female. Lower fronto-orbital seta absent (Fig.
12). Upper fronto-orbital seta present. Flagellomere
1 yellow, greatly enlarged, subequal to eye height,
apically flattened. Palpus small, with setae normal,
thick, moderately long, pointed. Color of abdomi-
nal tergites mostly yellow, with some darker mark-
ings. Venter of abdomen yellow. Venter of segment
3 bare. Venter of abdominal segment 5 lacking
dense setal combs, without sclerite. Ventromedial
setae of segment 6 present, long, thick, evenly
spaced, arranged in relatively straight line. Venter
of segment 6 without sclerite. In cleared specimens,
dorsal abdominal glands visible, dark. Intersegment
6-7 without distinct sclerite. Ovipositor dorsally
broad, anterodorsal portion broad, rounded (Fig.
54). Anteroventral margin of ovipositor with notch.
Posterodorsal apex of ovipositor pointed, but
broadly rounded before apex. In lateral view, pos-
teroventral apex of ovipositor straight. Posteroven-
tral apex of ovipositor pointed. Stylet with lateral
barbs. Dorsal sclerite consisting of two long pro-
cesses.

Brown: Neotropical Apocephalus, Subgenus Mesophora

GEOGRAPHICAL DISTRIBUTION. Known
from a single middle elevation site in Costa Rica.

PHYLOGENETIC RELATIONSHIPS. This spe-
cies is clearly part of the A. borealis-subgroup , pos-
sibly a female of A. emphysemus.

MATERIAL EXAMINED. COSTA RICA: San
Jose, Zurqui de Moravia, 10.05°N, 84.02°W, 19,
v.1991 [LACM ENT 000640], 29, ix-x.1993
[LACM ENT 001496, 001492], P. Hanson, Mal-
aise trap (LACM).

Phorid Species 3246 9
(Figs. 13, 55, 64)

SPECIES RECOGNITION. This female is also
similar to that of A. borealis but has the lower fron-
to-orbital seta present and is darker in color. Also,
the extremely enlarged setulae of the hind tarso-
mere of species 3246 is diagnostic (Fig. 64).

DESCRIPTION.

General Characters. Body length 1.9 mm. Frons
yellow. Palpus yellow. Dorsum of thorax yellow.
Pleural regions yellow. Scutellum of same color as
dorsum of thorax. Scutellum with short anterior
and long posterior seta. Legs yellowish-brown.
Venter of tarsomere 2 of mid leg with row of thin
setulae and thick apical setula. Hind femur yellow-
ish-brown, with abrupt, distinctive darkening at
apex. Anteroventral row of enlarged setulae on
hind basitarsus straight, markedly longer than
those of other rows (Fig. 64). Wing vein Rs slender,
subequal to or thinner than costa. Wing vein R2+3
well developed. Wing vein CuAx reaches wing mar-
gin. Halter brown.

Female. Lower fronto-orbital seta present, slight-
ly medially displaced (Fig. 13). Upper fronto-orbital
seta present. Flagellomere 1 yellow, pyriform. Pal-
pus small, with setae normal, thick, moderately
long, pointed. Color of abdominal tergites mostly
yellow, with some darker markings. Venter of ab-
domen yellow. Venter of segment 3 bare. Venter of
abdominal segment 5 lacking dense setal combs,
without sclerite. Ventromedial setae of segment 6
present, long, thick, evenly spaced, arranged in rel-
atively straight line. Venter of segment 6 without
sclerite. In cleared specimens, dorsal abdominal
glands visible, dark. Intersegment 6-7 with dark
sclerite. Ovipositor dorsally broad, anterodorsal
portion broad, rounded (Fig. 55). Anteroventral
margin of ovipositor with notch. Posterodorsal
apex of ovipositor pointed, but broadly rounded
before apex. In lateral view, posteroventral apex of
ovipositor straight. Posteroventral apex of ovipos-
itor pointed. Stylet with lateral barbs. Dorsal scler-
ite consisting of unforked, medial process.

GEOGRAPHICAL DISTRIBUTION. Known
from a single middle elevation site in Costa Rica.

PHYLOGENETIC RELATIONSHIPS. This spe-
cies also is clearly part of the A. borealis- subgroup;
it might also be the female of A. emphysemus.

MATERIAL EXAMINED. COSTA RICA: San
Jose, Zurqui de Moravia, 10.05°N, 84.02°W, 19,

x-xii.1990, P. Hanson, Malaise trap (LACM)
[LACM ENT 00614].

Phorid Species 3247 9

SPECIES RECOGNITION. This species is simi-
lar to A. megalops and species 3251 but differs by
having a large patch of ventral setae on a sclero-
tized area of segment 6. The ovipositor is similar
to that of species 3251 (see Brown, 1993, fig. 52).

DESCRIPTION.

General Characters. Body length 1.5 mm. Frons
yellow. Palpus yellow. Dorsum of thorax yellow.
Pleural regions yellow. Scutellum of same color as
dorsum of thorax. Scutellum with short anterior
and long posterior seta. Legs yellowish-brown.
Venter of tarsomere 2 of mid leg with row of thin
setulae and thick apical setula. Hind femur yellow-
ish-brown, with abrupt, distinctive darkening at
apex. Anteroventral row of enlarged setulae on
hind basitarsus straight, with some setulae mark-
edly longer than those of other rows. Wing vein Rs
slender, subequal to or thinner than costa. Wing
vein R2+3 well developed. Wing vein CuAj reaches
wing margin. Halter yellow.

Female. Lower fronto-orbital seta absent. Upper
fronto-orbital seta present. Flagellomere 1 yellow,
greatly enlarged, subequal to eye height, apically
flattened. Palpus small, with setae normal, thick,
moderately long, pointed. Color of abdominal ter-
gites mostly yellow, with some darker markings.
Venter of abdomen yellow. Venter of segment 3
bare. Venter of abdominal segment 5 lacking dense
setal combs, without sclerite. Ventromedial setae of
segment 6 present, long, thick, evenly spaced, scat-
tered. Venter of segment 6 with sclerite present. In
cleared specimens, dorsal abdominal glands visible,
dark. Intersegment 6-7 with dark sclerite. Ovipos-
itor dorsally broad, anterodorsal portion broad,
rounded. Anteroventral margin of ovipositor with
notch. Posterodorsal apex of ovipositor pointed. In
lateral view, posteroventral apex of ovipositor
straight. Posteroventral apex of ovipositor pointed.
Stylet with lateral barbs. Dorsal sclerite consisting
of unforked, medial process.

GEOGRAPFIICAL DISTRIBUTION. This spe-
cies is known only from the Dominican Republic.

PHYLOGENETIC RELATIONSHIPS. This spe-
cies is part of the A. borealis- group, although not
closely related to A. borealis.

MATERIAL EXAMINED. DOMINICAN RE-
PUBLIC: Pedernales, Las Abejas, 38 km NNW
Cabo Rojo, 18.13°N, 71.63°W, 19, 15.vii.1987, J.
Rawlins, R. Davidson, Malaise trap, 1250 m
(CMNH) [LACM ENT 000628].

Phorid Species 3250 S
(Figs. 44-45)

SPECIES RECOGNITION. The males of this
species are superficially similar to those of A. an-
tennatus but differ by the dark brown halter and
the different genitalia.

Contributions in Science, Number 462

Brown: Neotropical Apocephalus, Subgenus M esophora ■ 27

DESCRIPTION.

General Characters. Body length 1.5-2. 1 mm.
Frons yellow. Palpus yellow. Dorsum of thorax yel-
low. Pleural regions white. Scutellum of same color
as dorsum of thorax. Scutellum with short anterior
and long posterior seta. Legs yellowish-brown.
Venter of tarsomere 2 of mid leg with row of thin
setulae and thick apical setula. Hind femur yellow-
ish-brown, with abrupt, distinctive darkening at
apex. Anteroventral row of enlarged setulae on
hind basitarsus slightly sinuate, with all setulae
short, subequal to those of other rows. Wing vein
Rs slender, subequal to or thinner than costa. Wing
vein R2+3 well developed. Wing vein CuAx reaches
wing margin. Halter brown.

Male. Lower fronto-orbital seta absent. Upper
fronto-orbital seta absent. Flagellomere 1 yellow,
greatly enlarged, apically flattened, apical two-
thirds with narrowest part greater than one-half
basal width. Palpus small, with setae normal, thick,
moderately long, pointed. Mid leg with tarsomeres
all slender. Costal vein not thickened. Abdominal
tergites mostly dark, with yellow markings (tergite
6 yellow). Venter of abdomen white. Venter of seg-
ment 6 bare. Venter of segment 6 without sclerite.
Epandrium globular, approximately as long as
high, with setae near cercus not markedly larger
than other epandrial setae. Left side of epandrium
with scattered, thin setae (Fig. 45). Right side of
epandrium with prominent, darkly sclerotized, dor-
solateral ridge (Fig. 44). Right surstylus short,
rounded, lacking ventral carina. Cercus brown. Se-
tae of cercus and proctiger of normal size, longer
and thicker than short setae of epandrium.

GEOGRAPHICAL DISTRIBUTION. This spe-
cies has been collected at two middle elevation sites
in Costa Rica.

PHYLOGENETIC RELATIONSHIPS. These
males probably belong to the A. anfractus-sub-
group, previously known only from females. They
could be males of A. bisetus, A. gracilis , A. mora-
viensis, or A. trisetus.

MATERIAL EXAMINED. COSTA RICA: Li-
mon, 16 km W Guapiles, 10.15°N, 83.92°W, Id,
iii— v. 1990 [LACM ENT 000856], P. Hanson, Mal-
aise trap, 400 m (LACM), 7 km W Bribri, 9.58°N,
82.88°W, 1A, ix-xi.1989 [LACM ENT 000848]
(LACM), San Jose, Zurqui de Moravia, 10.05°N,
84.02°W, 2d, iv.1991 [LACM ENT 000845,
000858], Id, v.1991 [LACM ENT 000851], 4d,
vi.1991 [LACM ENT 000786, 000790, 000791,
000792], 8d, v.1992 [LACM ENT 000833,
000834, 000835, 000839, 000840, 000841,
000843, 000852], 2d, vi.1992 [LACM ENT
000832, 000850], Id, l-15.vi.1993 [LACM ENT
000844], P. Hanson, Malaise trap, 1600 m
(LACM).

Phorid Species 3252 d
(Figs. 14, 46-47)

SPECIES RECOGNITION. Males of this species
can be separated from other Mesophora by the dis-

28 ■ Contributions in Science, Number 462

tinctly narrowed flagellomere 1 (Fig. 14) and the
presence of ventral setae on segment 6.

DESCRIPTION.

General Characters. Body length 1.2-1 .4 mm.
Frons brown. Palpus yellow. Dorsum of thorax
light brown. Pleural regions light brown. Scutellum
of same color as dorsum of thorax. Scutellum with
short anterior and long posterior seta. Legs yellow-
ish-brown. Venter of tarsomere 2 of mid leg with
row of thin setulae and thick apical setula. Hind
femur yellowish-brown, with abrupt, distinctive
darkening at apex. Anteroventral row of enlarged
setulae on hind basitarsus sinuate, with some setu-
lae markedly longer than those of other rows. Wing
vein Rs slender, subequal to or thinner than costa.
Wing vein R2+3 well developed. Wing vein CuA3
reaches wing margin. Halter brown.

Male. Frons opaque. Lower fronto-orbital seta
absent. Upper fronto-orbital seta absent. Flagello-
mere 1 yellow, greatly enlarged, apically flattened,
apical two-thirds narrowed to one-third basal
width. Palpus small, with setae normal, thick, mod-
erately long, pointed. Mid leg with tarsomeres all
slender. Costal vein not thickened. Abdominal ter-
gites mostly dark, with yellow markings. Venter of
abdomen white to yellow. Venter of segment 6 with
distinct, black setae (but few and scattered), ventral
setae in a single row. Venter of segment 6 without
sclerite. Epandrium globular, approximately as long
as high, with anterior part short, straight, with se-
tae near cercus not markedly larger than other
epandrial setae. Left side of epandrium with scat-
tered, thin setae (Fig. 47). Right side of epandrium
with faint, unsclerotized dorsolateral ridge (Fig.
46). Right surstylus short, rounded, lacking ventral
carina. Number of setae on right surstylus many.
Cercus brown. Setae of cercus and proctiger of nor-
mal size, longer and thicker than short setae of
epandrium.

GEOGRAPHICAL DISTRIBUTION. This spe-
cies is known from a single middle elevation site in
Costa Rica.

PHYLOGENETIC RELATIONSHIPS. This spe-
cies is a male of the A. anfractus-subgroup and thus
possibly of the Costa Rican species A. bisetus, A.
gracilis, A. moraviensis, or A. trisetus. Apparently
this species and species 3253 are sister taxa, based
on the following putative synapomorphies: 1) fla-
gellomere 1 narrowed apically (Fig. 14), and 2) ap-
icoventral setula of each tarsomere of mid leg en-
larged.

MATERIAL EXAMINED. COSTA RICA: San
Jose, Zurqui de Moravia, 10.05°N, 84.02°W, 2d,
x-xii.1990 [LACM ENT 000880, 000882], 2d,

iii. 1991 [LACM ENT 000876, 000877], 2d,

iv. 1991 [LACM ENT 000885, 000887], Id,

v. 1991 [LACM ENT 000878], Id, vii.1991
[LACM ENT 000883], Id, v.1992 [LACM ENT
000879], 2d, vii.1992 [LACM ENT 000881,
000884], P. Hanson, Malaise trap, 1600 m
(LACM).

Brown: Neotropical Apocephalus, Subgenus Mesophora

Phorid Species 3253 8
(Figs. 48-49)

SPECIES RECOGNITION. Males of this species
can be separated from the similar males of species
3253 by the enlarged palpus and absence of ventral
setae on abdominal segment 6.

DESCRIPTION.

General Characters. Body length 1.3-1 .4 mm.
Frons yellow. Palpus yellow. Dorsum of thorax yel-
low. Pleural regions yellow. Scutellum of same color
as dorsum of thorax. Scutellum with short anterior
and long posterior seta. Legs yellow. Venter of tar-
somere 2 of mid leg with row of thin setulae and thick
apical setula. Hind femur yellowish-brown, with
abrupt, distinctive darkening at apex. Anteroventral
row of enlarged setulae on hind basitarsus sinuate,
with some setulae markedly longer than those of oth-
er rows. Wing vein Rs slender, subequal to or thinner
than costa. Wing vein R2+3 well developed. Wing vein
CuA, reaches wing margin. Halter yellow.

Male. Frons opaque. Lower fronto-orbital seta ab-
sent. Upper fronto-orbital seta absent. Flagellomere 1
yellowish at base, apically darker, greatly enlarged,
apically flattened, apical two-thirds narrowed to one-
third basal width. Palpus enlarged, with setae re-
duced, thin, short, stubby. Mid leg with tarsomeres all
slender. Costal vein not thickened. Abdominal tergites
mostly dark, with yellow markings. Venter of abdo-
men yellow. Venter of segment 6 bare. Venter of seg-
ment 6 without sclerite. Epandrium globular, approx-
imately as long as high, with anterior part short,
straight, with setae near cercus not markedly larger
than other epandrial setae. Left side of epandrium
with scattered, thin setae (Fig. 49). Right side of epan-
drium with prominent, darkly sclerotized, dorsolateral
ridge (Fig. 48). Right surstylus short, rounded, lacking
ventral carina. Number of setae on right surstylus
many. Cercus yellow. Setae of cercus and proctiger of
normal size, longer and thicker than short setae of
epandrium.

GEOGRAPHICAL DISTRIBUTION. This spe-
cies is known from two middle elevation sites in
Costa Rica.

PHYLOGENETIC RELATIONSHIPS. See Apo-
cephalus 3252, above.

MATERIAL EXAMINED. COSTA RICA: San
Jose, 26 km N San Isidro, 9.5°N, 83.72°W, Id, ii-
v.1992, P. Hanson, Malaise trap, 2100 m (LACM)
[LACM ENT 000875], Zurqui de Moravia,
10.05°N, 84.02°W, Id, iv.1991, P. Hanson, Mal-
aise trap, 1600 m (LACM) [LACM ENT 000886].

IDENTIFICATION

Key to Males of Neotropical Region
Mesophora Species

1 Anterior face of hind femur with abrupt,

distinctive darkening at apex 2

- Anterior face of hind femur evenly colored

throughout (posterior face may have dis-
tinctive darkening) 17

Contributions in Science, Number 462

2 (1) Palpus enlarged, inflated, subequal to eye

height (Figs. 16-17) 3

- Palpus small, normal in appearance, short-
er than eye height (Fig. 15) 7

3 (2) Apical two-thirds of flagellomere 1 nar-

rowed abruptly to one-third basal width
(Figs. 9, 14); flagellomere 1 yellowish at

base, apically darker 4

Flagellomere 1 broader, yellow 5

4 (3) Setae of cercus and proctiger of normal

size, longer and thicker than short setae of
epandrium (Figs. 48-49); anteroventral
row of enlarged setulae on hind basitarsus
sinuate (Fig. 65); left side of epandrium
with scattered, thin setae (Fig. 48); num-
ber of setae on right surstylus many (Fig.

49) Phorid species 3253

Setae of cercus and proctiger markedly re-
duced, subequal to short setae of epan-
drium (Brown, 1993, figs. 8, 29); antero-
ventral row of enlarged setulae on hind
basitarsus straight; left side of epandrium
with thick, ventral setae (Brown, 1993, fig.
8); number of setae on right surstylus two,

or three

Apocephalus brevicercus Brown

5 (3) Venter of segment 6 with distinct, black

setae .... Apocephalus prolatus Brown

- Venter of segment 6 bare 6

6 (5) Right surstylus short, rounded (Fig. 36);

left side of epandrium with scattered, thin

setae (Fig. 37)

.... Apocephalus secundus new species
Right surstylus elongate (similar to Fig.
42); left side of epandrium with thick, ven-
tral setae (Brown, 1993, fig. 14)

Apocephalus longistylus Brown

7 (2) Halter brown 8

- Halter yellow 11

8 (7) Venter of segment 6 with distinct, black

setae; frons brown; anteroventral row of
enlarged setulae on hind basitarsus sinu-
ate; flagellomere 1 apical two-thirds nar-
rowed to one-third basal width (Fig. 14)
Phorid species 3252

- Venter of segment 6 bare; frons yellow;
anteroventral row of enlarged setulae on
hind basitarsus straight; flagellomere 1
apical two-thirds with narrowest part
greater than one-half basal width .... 9

9 (8) Upper and lower fronto-orbital seta pres-

ent (similar to Fig. 1); anteroventral row
of enlarged setulae on hind basitarsus with
some setulae markedly longer than those
of other rows (similar to Fig. 63); flagel-
lomere 1 pyriform; cercus yellow

Apocephalus atavus new species

- Upper and lower fronto-orbital seta ab-
sent (similar to Fig. 5); anteroventral row
of enlarged setulae on hind basitarsus with
all setulae short, subequal to those of oth-
er rows; flagellomere 1 greatly enlarged,

Brown: Neotropical Apocephalus, Subgenus Mesophora ■ 29

apically flattened (similar to Fig. 5); cercus
brown 10

10 (9) Epandrium globular, approximately as

long as high (Figs. 44-45); pleural regions

white Phorid species 3250

- Epandrium markedly depressed, longer

than height (Brown, 1993, figs. 5, 26);
pleural regions yellow or light brown . . .
Apocephalus insulanus Borgmeier

11 (7) Venter of segment 6 with distinct, black

setae

.... Apocephalus megalops new species

Venter of segment 6 bare 12

12 (11) Epandrium with large, prominent seta
near cercus (Figs. 34-35) 13

Epandrium with setae near cercus not
markedly larger than other epandrial setae
14

13 (12) Palpus with setae normal, thick, moder-

ately long, pointed (similar to Fig. 5) ...

Apocephalus pilatus new species

Palpus with setae reduced, thin, short,

stubby (Fig. 15)

Apocephalus antennatus Malloch

14 (12) Epandrium markedly depressed (Brown,

1993, figs. 5, 26); left side of epandrium
with scattered, thin setae; cercus brown

Apocephalus insulanus Borgmeier

Epandrium globular; left side of epan-
drium with short setae confined to poste-
rior margin (Brown, 1993, fig. 20), or
with thick, ventral setae (Fig. 41); cercus
yellow 15

15 (14) Frons brown; dorsum of thorax dark

brown; left side of epandrium with short

setae confined to posterior margin

Apocephalus adustus Brown

- Frons yellow; dorsum of thorax yellow;

left side of epandrium with thick, ventral
setae 16

16 (15) Right surstylus short, rounded, lacking

ventral carina (Fig. 40)

Apocephalus crassus new species

- Right surstylus elongate, pointed (Brown,

1993, fig. 33); ventral carina short, ex-
tended partially across surstylus

Apocephalus angustistylus Brown

17 (1) Venter of segment 6 with distinct, black

setae 18

- Venter of segment 6 bare 22

18 (17) Halter yellow; epandrium not band-

shaped 19

- Halter brown; if lighter-colored, then

epandrium short, band-shaped (Figs. 20-
23) 20

19 (18) Flagellomere 1 pyriform; scutellum with

long anterior and long posterior setae;
wing vein CuAl attenuated, not reaching
wing margin; upper and lower fronto-or-
bital seta present

Apocephalus apivorus new species

Flagellomere 1 greatly enlarged, apically

30 ■ Contributions in Science, Number 462

flattened; scutellum with short anterior
and long posterior seta; wing vein CuAj
reaches wing margin; upper and lower

fronto-orbital seta absent

Apocephalus micrepelis Brown

20 (18) Flagellomere 1 pyriform (Fig. 3), brown;
upper fronto-orbital seta present; ventral

setae in two rows

.... Apocephalus lemniscus new species
- Flagellomere 1 greatly enlarged, apically

flattened, yellow; upper fronto-orbital seta
absent; ventral setae in a single row (Fig.
21

70)

Apical two-thirds of flagellomere 1 broad-
er (similar to Fig. 5) 26

23 (22) Palpus brown; dorsum of thorax and pleu-

ral regions dark brown; cercus brown 24
Palpus, dorsum of thorax, pleural regions
and cercus yellow 25

24 (23) Frons opaque; legs, including hind femur

dark brown; flagellomere 1 brown; epan-
drium with anterior part bulging anteri-
orly; number of setae on right surstylus
three (see figures in Brown, 1994b) . . . .
Apocephalus satanus Brown

- Frons glossy; legs, including hind femur
yellowish-brown; flagellomere 1 yellowish
at base, apically darker; epandrium with
anterior part straight; number of setae on
right surstylus one (see figures in Brown,
1994b) . . Apocephalus nitifrons Brown

25 (23) Palpus enlarged, with setae reduced, thin,

short, stubby (similar to Fig. 16); epan-
drium with anterior part elongate; right
side of epandrium, beside ventral setae,
with numerous lateral setae; number of se-
tae on right surstylus two (see figures in

Brown, 1994b)

Apocephalus grandiflavus Brown

- Palpus small, with setae normal, thick,
moderately long, pointed; epandrium with
anterior part short, straight; right side of
epandrium, beside ventral setae, with one
to three lateral setae; number of setae on

right surstylus one to many

.... Apocephalus truncaticercus Brown

26 (22) Mid leg with tarsomere 1 expanded; right

surstylus with ventral carina present 27

- Mid leg with tarsomeres all slender; right

surstylus lacking ventral carina 29

21 (20) Palpus small, with setae normal, thick,

moderately long, pointed; right surstylus

short, rounded (Fig. 20)

Apocephalus lizanoi new species

Palpus enlarged, with setae reduced, thin,
short, stubby (similar to Fig. 16); right
surstylus truncate, apically flared (Brown,

1993, fig. 42)

Apocephalus curtus Brown

22 (17) Apical two-thirds of flagellomere 1 nar-

rowed to one-third basal width (Fig. 9)
23

Brown: Neotropical Apocephalus, Subgenus Mesophora

27 (26) Mid leg with tarsomeres 2 and 3 expand-

ed; right surstylus with ventral carina
short, extended partially across surstylus

(Brown, 1993, fig. 40)

Apocephalus tritarsus Brown

- Mid leg with tarsomeres 2 and 3 slender;

right surstylus with ventral carina long,
with many short side branches, or with
ventral carina long, unbroken 28

28 (27) Halter and flagellomere 1 yellow; venter

of tarsomere 1 of midleg with scattered
setulae; right surstylus with ventral carina
long, unbroken (Brown, 1993, fig. 37) . .
Apocephalus mortifer Borgmeier

- Halter brown; flagellomere 1 yellowish at

base, apically darker; venter of tarsomere
1 of midleg with setulae confined to pos-
terior margin; right surstylus with ventral
carina long, with many short side branch-
es (Brown, 1993, fig. 36)

Apocephalus hansoni Brown

29 (26) Halter yellow 30

- Halter brown 32

30 (29) Left side of epandrium with dense, thicker

setae (Fig. 39); epandrium with setae near
cercus not markedly larger than other

epandrial setae

.... Apocephalus echinatus new species

- Left side of epandrium with scattered, thin

setae; epandrium with large, prominent
seta near cercus (Figs. 42-43) 31

31 (30) Epandrium markedly depressed (Figs. 42-

43); right surstylus elongate

Apocephalus prolixus new species

Epandrium globular (similar to Figs. 36-

41); right surstylus short, rounded

Apocephalus antennatus Malloch

32 (29) Wing vein R2+3 absent; anteroventral row

of enlarged setulae on hind basitarsus sin-
uate (similar to Fig. 65); upper fronto-or-
bital seta present (Fig. 4); flagellomere 1
slightly enlarged, apically flattened (Fig. 4)
Apocephalus absentis Brown

- Wing vein R2+3 well developed; anteroven-

tral row of enlarged setulae on hind basi-
tarsus straight; upper fronto-orbital seta
absent; flagellomere 1 greatly enlarged,
apically flattened 33

33 (32) Costal vein thickened between humeral

crossvein and R! (Fig. 68); frons and pal-
pus brown; hind femur dark brown; left
side of epandrium with thick, ventral setae
(Fig. 31); setae of cercus markedly re-
duced, subequal to short setae of epan-
drium (Figs. 30-31)

. . Apocephalus emphysemus new species
Costal vein not thickened; frons and pal-
pus yellow; hind femur yellowish-brown,
or white; left side of epandrium with scat-
tered, thin setae; setae of cercus of normal
size, longer and thicker than short setae of
epandrium 34

Contributions in Science, Number 462

34 (33) Flagellomere 1, pleural regions and legs
yellow; epandrium markedly depressed,
with large, prominent seta near cercus
(Figs. 42-43); right surstylus elongate (Fig.

42); cercus yellow

Apocephalus prolixus new species

- Flagellomere 1 yellowish at base, apically
darker; pleural regions and legs white;
epandrium globular, with setae near cercus
not markedly larger than other epandrial
setae (Figs. 32-33); right surstylus short,

rounded (Fig. 32); cercus brown

Apocephalus niveus new species

Key to Females of Neotropical
Region Mesophora Species

(Note: the female of A. limai Prado, which I have
not examined, is not included in this key.)

1 Anteroventral row of enlarged setulae on
hind basitarsus straight (Fig. 63) .... 2

- Anteroventral row of enlarged setulae on
hind basitarsus sinuate (Fig. 65) .... 20

2 (1) Lower fronto-orbital seta present (Fig. 1),

in some species displaced medially (as in

Fig. 10) 3

Lower fronto-orbital seta absent (as in Fig.
12) 7

3 (2) Flagellomere 1 round

Apocephalus atavus new species

Flagellomere 1 pyriform (Figs. 1, 8) . . 4

- Flagellomere 1 greatly enlarged, subequal
to eye height, apically flattened (Fig. 10)
Apocephalus mortifer Borgmeier

(presumably the undescribed females
of A. hansoni and A. tritarsus
would key here also)

4 (3) Hind femur evenly colored throughout;

wing vein CuAj attenuated, not reaching
wing margin; venter of segment 3 with few

dark setae

Apocephalus apivorus new species

- Hind femur with abrupt, distinctive dark-
ening at apex; wing vein CuAj reaches
wing margin; venter of segment 3 bare 5

5 (4) Anteroventral row of enlarged setulae on

hind basitarsus with all setulae short, sub-
equal to those of other rows; halter yellow
6

- Anteroventral row of enlarged setulae on

hind basitarsus with setulae markedly lon-
ger than those of other rows (Fig. 64); hal-
ter brown Phorid species 3246

6 (5) Frons and dorsum of thorax yellow; in

cleared specimens, dorsal abdominal

glands visible, dark

Apocephalus angustistylus Brown

Frons and dorsum of thorax dark brown;
in cleared specimens, dorsal abdominal

glands invisible

Apocephalus adustus Brown

Brown: Neotropical Apocephalus, Subgenus Mesophora ■ 31

7 (2) Upper fronto-orbital seta present (e.g.,

Fig. 12) 8

Upper fronto-orbital seta absent (e.g., Fig.
7) 16

8 (7) Flagellomere 1 pyriform (similar to Figs.

8, 17) 9

Flagellomere 1 greatly enlarged, subequal
to eye height, apically flattened (e.g., Fig.
11) . 11

9 (8) Hind femur evenly colored throughout;

halter brown; dorsum of thorax light
brown; palpus small 10

- Hind femur with abrupt, distinctive dark-
ening at apex; halter yellow; dorsum of
thorax yellow; palpus elongate (Fig. 17)
Phorid species 3136

(—Apocephalus unnamed
female B of Brown, 1993)

10 (9) Venter of abdomen with large setal combs

(Fig. 66); hind femur yellowish-brown;

pleural regions yellow

Apocephalus curtus Brown

- Venter of abdomen lacking setal combs;
hind femur white; pleural regions white
Apocephalus niveus new species

11 (8) Either venter of segment 6 with extremely

short, thin setae (A. antennatus, insulanus
and longistylus) or venter of abdomen
with large setal combs (A. lizanoi; Fig.
67); anteroventral row of enlarged setulae
on hind basitarsus with all setulae short,
subequal to those of other rows .... 12
Venter of segment 6 with large setae (sim-
ilar to Fig. 69); venter of abdomen with-
out setal combs; anteroventral row of en-
larged setulae on hind basitarsus with
some setulae markedly longer than those
of other rows (Fig. 63) 15

12 (11) Palpus small (similar to Fig. 15) .... 13

Palpus elongate (similar to Fig. 16) ....
....... Apocephalus longistylus Brown

13 (12) Venter of abdomen with large setal combs

(Fig. 67); venter of segment 3 with few

dark setae

Apocephalus lizanoi new species

- Venter of abdomen lacking setal combs;

venter of segment 3 bare 14

14 (13) In cleared specimens, dorsal abdominal

glands invisible; anterodorsal portion of
ovipositor broad, rounded (see figures in

Brown, 1993)

Apocephalus antennatus Malloch

- In cleared specimens, dorsal abdominal

glands visible, dark; anterodorsal portion
of ovipositor narrowed, elongate (see fig-
ures in Brown, 1993)

Apocephalus insulanus Borgmeier

15 (11) Ventral setae of segment 6 in several rows

on crescent-shaped, anteriorly pointed

sternite; halter yellow

Phorid species 3247

- Ventral setae of segment 6 in single

32 ■ Contributions in Science, Number 462

straight row; sternite absent; halter brown
Phorid species 3223

16 (7) Hind femur evenly colored throughout

Hind femur with abrupt, distinctive dark-
ening at apex 18

17 (16) Palpus, dorsum of thorax and pleural

regions yellow; hind femur yellowish-

brown

Apocephalus grandiflavus Brown

Apocephalus truncaticercus Brown
(females of these two species cannot
be separated at this time)

- Palpus, dorsum of thorax, hind femur and

pleural regions dark brown

Apocephalus satanus Brown

18 (16) Segment 6 without ventral setae; antero-

ventral row of enlarged setulae on hind
basitarsus with all setulae short, subequal
to those of other rows; in cleared speci-
mens, dorsal abdominal glands invisible;
ovipositor anterodorsal portion narrowed,
elongate (Brown, 1994b, fig. 5) .......

Apocephalus hrevicercus Brown

Segment 6 with ventral setae present; an-
teroventral row of enlarged setulae on
hind basitarsus with some setulae mark-
edly longer than those of other rows; in
cleared specimens, dorsal abdominal
glands visible, dark; ovipositor anterodor-
sal portion broad, rounded 19

19 (18) Flagellomere 1 greatly enlarged, subequal

to eye height, apically flattened; frons
broad (Fig. 11) .... Phorid species 3251

- Flagellomere pyriform; frons narrow (Fig.
6) . . Apocephalus megalops new species

20 (1) Lower fronto-orbital seta absent; R2+3 ab-

sent; hind femur evenly colored through-
out; halter brown

Apocephalus absentis Brown

- Lower fronto-orbital seta present; R2+3
present; hind femur with abrupt, distinc-
tive darkening at apex; halter yellow . . .
.... other A. anfractus-subgroup species

(see Brown, 1993)

EVOLUTION OF HOST SELECTION

To analyze the host shift in Mesophora, the known
hosts of the various species are plotted on the
cladogram of relationships (Fig. 73). Although
some progress has been made in identifying hosts
of the various species, there are still large gaps in
our knowledge. Only 8 of the 42 named species of
Mesophora have been reared, and most of them
only once. Still, we have enough information to
speculate about the observed host shift.

All outgroups to Mesophora are parasites of
ants, including all other species of Apocephalus.
Therefore, ants must be considered the primitive
hosts of the group. The most derived clades of Me-
sophora, those of the A. wheeleri-subgroup, appar-

Brown: Neotropical Apocephalus, Subgenus Mesophora

Figure 72. Preferred cladogram based on data in Table 1.

Contributions in Science, Number 462

Brown: Neotropical Apocephalus, Subgenus Mesophora

stingless bees

-l stingless bees

borealis -
group

1 bees, wasps,

__ spiders

anfractus-

subgroup

curtus - ^ ?

subgroup

A. niveus ?

AMngl stylus , lampyrids

A. angustistylus

antennatus •
infragroup

A. antennatus .

lampyrids

A. insulanus

-| lampyrids

wheeleri-

infragroup

cantharid,

lampyrid

truncaticerus-

infragroup

Figure 73. Hosts mapped on parasite cladogram.

ently are parasites of cantharoid beetles, including
lampyrids and cantharids. Only one species, A.
mortifer, has been reared from a cantharid, Chau-
liognathus fallax, but I also reared larvae from a
different cantharid in Costa Rica; unfortunately
they died soon after pupation.

Cantharoid beetles are extremely divergent hosts
from ants, and the interesting question to answer is
how did the flies move from ants to beetles. Ac-
cording to the information presented here, they first
shifted from ants to stingless bees. This shift is fair-
ly easy to understand: the physiological and behav-
ioral adaptations necessary to shift from one social
hymenopteran to another are probably relatively
minor.

The next shift, from bees to beetles, is still unex-
34 ■ Contributions in Science, Number 462

plained. The only tantalizing piece of evidence is
the host information for A. borealis, a species that
has been reared from such divergent hosts as vespid
wasps (Hymenoptera: Vespidae), bumble bees (Hy-
menoptera: Apidae), and a black widow spider (Ar-
aneida: Theridiidae) (Borgmeier, 1963; Brown,
1993; Disney, 1994; Ennik, 1973). Perhaps within
the A. borealis- group, or its ancestors, there was
behavioral or physiological plasticity that allowed
species to expand the range of acceptable hosts.
Some of these attacked cantharoid beetles and were
able to diversify to form the A. wbeeleri- subgroup.

More field work is necessary to explore this pos-
sibility and especially to find hosts of the A anfrac-
tas-subgroup and the A. cwrtas-subgroup. Also,
laboratory experiments to establish the range of

Brown: Neotropical Apocephalus, Subgenus Mesopbora

A. gemursus

A. hansoni

A. mortifer

A. unitarsus

Chauliognathus fallax
"® (Cantharidae)

A. tritarsus Lucidota sp.

(Lampyridae)

1 ?

A. wheeleri

Figure 74. Host taxa mapped on cladogram of A. wheeleri- infragroup.

hosts acceptable to A. borealis would be of great
interest.

Within the A. antennatus- and A. wheeleri- infra-
groups, there is not enough information to say
whether specific host-parasite coevolution has
taken place (Figs. 73-74), although such an hy-
pothesis seems unlikely. Apocephalus antennatus is
a generalized firefly parasite, attacking at least 5
genera throughout its extensive New World range.
Although A. insulanus has been reared only once,
from a species of Photinus in Peru (Brown, 1994a),
it has a similarly large range and probably attacks
more than one species of firefly. Less is known
about the A. wheeleri-'miragroup, and the two spe-
cies with known hosts, A. mortifer and A. tritarsus,
have each been reared from only a single host.

ACKNOWLEDGMENTS

The illustrations in this paper were skillfully executed by
Jesse Cantley. I thank Drs. Paul Hanson and Chen Young
for sending me the material upon which this revision is
based. For technical help I am grateful to Vladimir Bere-
zovskiy, Vicky Brown, Jesse Cantley, Betty Defibaugh, and
Brian Harris. I thank Isabel Borhorquez and Dr. Jim Lloyd
for identifying fireflies, and Roy Sneliing for identifying
stingless bees. For hospitality during field work in Costa
Rica I am grateful to the personnel of La Selva Biological
Station, Dr. Monty Wood (Monteverde Biological Sta-
tion), Senor Jorge Arturo Lizano (Zurqui de Moravia),
and Dr. Paul Hanson. This work was funded by the Or-
ganization for Tropical Studies Mellon Foundation Re-
search Fund, by a grant to the Natural History Museum
of Los Angeles County from the Weiler Foundation, and
by the National Science Foundation Grant DEB-9407190.

LITERATURE CITED

Borgmeier, T. 1937. Um nova especie de Apocephalus
(Dipt. Phoridae), endoparasita de Chauliognathus

fallax Germ. (Col. Cantharidae). Revista de Ento-
mologia, Rio de Janeiro 7:207-216.

. 1963. Revision of the North American phorid

flies. Part I. The Phorinae, Aenigmatiinae and Me-
topininae, except Megaselia. Studia Entomologica 6:
1-256.

— . 1971. Further studies on phorid flies, mainly of

the Neotropical Region (Diptera, Phoridae). Studia
Entomologica 14:1-172.

Brown, B.V. 1992. Generic revision of Phoridae of the
Nearctic Region and phylogenetic classification of
Phoridae, Sciadoceridae and Ironomyiidae (Diptera:

Phoridea). Memoirs of the Entomological Society of

Canada 164:1-144.

. 1993. Taxonomy and preliminary phylogeny of

the parasitic genus Apocephalus, subgenus Me-
sophora (Diptera: Phoridae). Systematic Entomology
18:191-230.

. 1994a. Life history parameters and new host rec-
ords of phorid parasites of fireflies. Coleopterists
Bulletin 48:145-147.

. 1994b. Revision and new species of the Apo-
cephalus ( Mesophora ) truncaticercus- infragroup
(Diptera: Phoridae). Contributions in Science 449:1-
7.

. in press. Parasitic phorid flies: A previously un-
recognized cost to aggregation behavior of male
stingless bees. Biotropica.

Brown, B.V., and D.H. Feener, Jr. 1991. Life history pa-
rameters and immature stages of Apocephalus par-
aponerae (Diptera: Phoridae), a parasitoid of the gi-
ant tropical ant Paraponera clavata (Hymenoptera:
Formicidae). Journal of Natural History 25:221-
231.

Coquillett, D.W. 1901. Apocephalus Coquillett, nov. gen.
Proceedings of the Entomological Society of Wash-
ington 4:501.

Cumming, J.M. 1992. Lactic acid as an agent for mac-
erating Diptera specimens. Fly Times 8:7.

Disney, R.H.L. 1980. Variation in Megaselia pulicaria
(Fall.) (Dipt., Phoridae) with the recognition of new

Contributions in Science, Number 462

Brown: Neotropical Apocephalus, Subgenus Mesophora ■ 35

synonymies. Entomologist’s Monthly Magazine 115:
97-103.

. 1994. Scuttle flies: The Phoridae. London: Chap-
man and Hall, xii + 467 pp.

Ennik, F. 1973. Apocephalus borealis Brues parasitic
upon Vespula spp. (Diptera: Phoridae; Hymenop-
tera: Vespidae). Pan-Pacific Entomologist 49:403-
404.

Malloch, J.R. 1913. Three new North American Diptera.

Canadian Entomologist 45:273-275.

Thompson, F.C. 1994. Bar codes for specimen data man-
agement. Insect Collection News 9:2-4.

Received 19 October 1995; accepted 16 April 1996.

36 ■ Contributions in Science, Number 462

Brown: Neotropical Apocephalus, Subgenus Mesophora

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

•S'Ss*

Number 463
12 December 1996

Contributions
in Science

A New Species of Batrachoseps
(Amphibia: Plethodontidae) from the
San Gabriel Mountains,

Southern California

David B. Wake

Natural History Museum of Los Angeles County

Serial
Publications
of THE
Natural History
Museum of
Los Angeles
County

Scientific

Publications

Committee

Robert J. Lavenberg, Acting Deputy Director
for Research and Collections
John M. Harris, Committee Chairman
Brian V. Brown
Kenneth E. Campbell
Kirk Fitzhugh
Karen Wise

Robin A. Simpson, Managing Editor

# Contributions in Science, a miscellaneous series of tech-
nical papers describing original research in the life and
earth sciences.

# Science Series, long articles and collections of papers on
natural history topics.

Copies of the publications in these series are sold through
the Museum Book Shop. A catalog is available on request.

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

Printed at Allen Press, Inc., Lawrence, Kansas
ISSN 0459-8113

A New Species of Batrachoseps

(Amphibia: Plethodontidae) from the
San Gabriel Mountains,

Southern California

David B. Wake1

ABSTRACT. A new species of slender salamander, Batrachoseps gabrieli, is described based on specimens
from two populations that occur in mixed coniferous forest at elevations of about 1,100-1,500 m in the
San Gabriel Mountains, Los Angeles County, California. The new taxon is an attenuate species belonging
to the morphologically more derived of the two clades in the genus. B. gabrieli is distinguished from other
attenuate members of the genus by the combination of its relatively broad head, long limbs, wide feet,
tapering tail of moderate length, and distinctive color pattern, including markings of bright pigmentation.
Within the attenuate clade of Batrachoseps the species has no close relatives. In addition to its morpho-
logical and ecological distinctiveness, the species differs substantially from all other members of the genus
in allozymes and mitochondrial DNA (cytochrome B) sequences. The species is apparently restricted to a
small geographic area, and special attention should be given to its preservation.

INTRODUCTION

In the course of a continuing study of geographic
variation and systematics of the slender salaman-
ders, genus Batrachoseps Bonaparte, 1841, I un-
expectedly discovered a strikingly distinct, unde-
scribed species in the San Gabriel Mountains, Los
Angeles County, California. The discovery at this
late date of a new vertebrate that is both morpho-
logically and ecologically distinctive in such a
heavily populated and well-explored region is sur-
prising.

DESCRIPTION OF NEW SPECIES

Batrachoseps gabrieli, new species
San Gabriel Mountain Slender Salamander
Figure 1

HOLOTYPE. MVZ 196449, an adult female
collected from under cover on a steep talus slope
above Soldier Creek in the upper San Gabriel River
drainage, approximately 1 km ESE Crystal Lake,
San Gabriel Mountains, Los Angeles County, Cal-
ifornia. SW Section 28, R9W T3N. 34°18'47" N,
117°49'57" w Approximately 1,550 m elevation.
Collected on 28 March 1985 by David B. Wake,
Nancy Staub, Samuel S. Sweet, Adonis Tate, Ste-
phen G. Tilley, and Jennifer Tilley.

PARATYPES. MVZ 178631-178646, 195577-
195583, 196450-196463, 215938, 215940-

1. Museum of Vertebrate Zoology and Department of
Integrative Biology, University of California, Berkeley,
California 94720-3160; Research Associate in Herpetol-
ogy, Natural History Museum of Los Angeles County, Los
Angeles, California 90007.

Contributions in Science, Number 463, pp. 1-12
Natural History Museum of Los Angeles County, 1996

215946, 215948, 222957-222961, LACM

143239-143240, (total 53), ail from vicinity of
type locality, collected between 1982 and 1995.
MVZ 178632, 178634, 178639, 178642, 195582-
195583 are cleared and stained skeletal prepara-
tions.

REFERRED SPECIMENS. MVZ 223570-
223571, from Rockbound Canyon, above highway
39, Los Angeles County, California. 34°18'02" N,
117°49'57" W. 1,158 elevation. Collected by R.H.
Goodman, Jr. and S. Teh, 29 March 1996.

DIAGNOSIS. This member of the attenuate
clade of Batrachoseps is a slender salamander of
moderate size with a relatively broad head, long
limbs, large hands and feet, and a markedly tapered
tail, distinguished from its geographically nearest
neighbors in the genus by these traits and by its
color pattern of bright coppery to orange-colored
diffuse stripes over the shoulders and in the pelvic
region, becoming patches on the tail. It is distin-
guished from B. nigriventris Cope, 1869, by its
larger size, its much broader and longer head, its
much longer limbs and larger hands and feet, and
its tapering tail. It is similar in size to B. pacificus
major. Camp, 1915, but differs in having a more
flattened head that is more sharply differentiated
from the neck, much longer limbs, broader hands
and feet, and a tapering tail.

DESCRIPTION. Batrachoseps gabrieli is a grac-
ile, slender species that is relatively generalized in
morphology, being of moderate size ( 8 adult males
range from 39.8-46.3, mean 42.4 mm standard
length; 16 adult females from 41.0-50.0, mean
46.1) and having a relatively broad, flattened head
that is well demarcated from the neck (standard

I cm 2 cm 3 cm 4cm 5 cm

Figure 1. Batrachoseps gabrieli, new species. A paratype specimen (MVZ 196454) collected at the type locality on 28
March 1985.

length is 7.5-8. 1, mean 7.7 times head width in
males; 7.3-8. 3, mean 7.8 in females). The species
has a relatively large facial region, with a broadly
rounded and somewhat flattened snout. Nostrils
are small, and there are small nasolabial protuber-
ances associated with the prominent nasolabial
groove. Sexually mature males have engorged tissue
around the anterior parts of the upper and lower
jaws and snout, including somewhat enlarged na-
solabial protuberances. No mental hedonic glands
are observed in males. Grooving patterns of the
head, throat, and neck are typical of the genus.
Eyes are relatively large and prominent. Vomerine
teeth are in short series, or, more commonly, in
patches (19-35, mean 25.7 total in males; 13-29,
mean 20.4 in females). Small maxillary teeth are
borne in a short, single series that ends under the
eye (25-69, mean 46.3 total in males; 30-51, mean
36 in females). Premaxillary teeth are small and rel-
atively numerous (7-13, mean 9.9 in females; 4-
10, mean 7.3 in males); those of males are slightly
enlarged and far forward in position, piercing the
upper lip. Numbers of trunk vertebrae vary from
19-20 (mean 19.5), with one individual having an
asymmetrical pelvic articulation and 20.5 trunk
vertebrae. Accordingly there are either 18 or 19
costal grooves between the limbs (counting one
each in the axilla and the groin). The tail is mod-
erately short compared with other members of this
clade (1.14-1.37, mean 1.18 times standard length
in females; 1.08-1.36, mean 1.24 in males), and it
is narrow at its base; the tail tapers rather strongly
to a fine tip. The tail has no discernible basal con-
striction and is broader than deep basally, becom-
ing round posteriorly. No postiliac gland is evident.
Limbs are relatively long and well developed. Limb
interval, the number of costal interspaces between
adpressed fore- and hind limbs, ranges from 5.5-
7.5 (mean 6.5) in females, and from 4-6.5 (mean
5.5) in males. Hind limb length ranges from 4.8-
5.8 (mean 5.2) times standard length in females,
and from 4. 7-5.0 (mean 4.9) in males. The hands
and feet are relatively large and well developed for

2 ■ Contributions in Science, Number 463

this clade, with well demarcated, stoutly rounded
digits and expanded digital tips that bear well-de-
veloped subterminal pads. Webbing is insignificant.
Fingers and toes in order of decreasing length are
3-2-4-1.

MEASUREMENTS OF THE HOLOTYPE (in
mm). Maximum head width 5.9; snout to gular
fold (head length) 9.5; head depth at posterior an-
gle of jaw 2.8; eyelid length 2.4, eyelid width 1.5;
anterior rim of orbit to snout 1.7; horizontal orbital
diameter 2.0; interorbital distance 3.4; snout to
forelimb 12.7; distance separating external nares
1.7; snout projection beyond mandible 0.9; snout
to posterior angle of vent (standard length) 46.7;
snout to anterior angle of vent 42.6; axilla to groin
length 26.3; tail length 58.8; tail width at base 3.7;
tail depth at base 2.8; forelimb length 7.7; hind
limb length 8.0; limb interval 6; width of right hand
2.1; width of right foot 2.8; length of third toe 1.1;
body width behind forelimbs 4.4. There are 11 pre-
maxillary, 44 maxillary, and 26 vomerine teeth, and
20 trunk, 3 caudosacral, and 39 tail vertebrae.

COLORATION (in alcohol). This is a dark
brownish black salamander that is marked with
bright highlights over the shoulder and pelvic re-
gion and on the tail. There is not a complete mid-
dorsal stripe in any of the specimens; most have a
pair of stripes over the shoulder and another pair
over the pelvic region, with streaks or patches of
color on the dorsum and especially on anterior and
middle parts of the dorsal surface of the tail. The
venter is a more or less uniform gray-black with
some small guanophore patches that are most
prominent on the gular region.

COLORATION (in life). Coloration was studied
in the holotype. The ground color was black with
a complete reticulum of the melanophores. The
dorsum and lateral surfaces of the trunk were
darker than the ventral surfaces, but the venter was
very dark as well. The dorsum was overlain by a
layer of dense, coppery iridophores, which intensi-
fied and accumulated to form a large spot on the
snout in front of the eyes and two short, broad

Wake: New Species of Batrachoseps

stripes over the shoulders. The stripes reappeared
in the pelvic region and extended as broken stripes,
then as blotches, onto the tail. The stripes were in-
tensely orange (when viewed in bright light) in col-
or. The overall impression of the dorsum was of a
broad but irregularly bounded coppery-bronze
stripe or band, intensified over the shoulders and in
the pelvic region. The dorsal proximal parts of the
limbs (limb insertions) were an intense coppery-or-
ange. There were a few superficial whitish irido-
phores evident dorsally, with a few more laterally
that extended ventrolaterally beyond the underly-
ing whitish iridophores. The iridophores extended
onto the venter where they were found as single
cells and small groups of cells on the gular region.
They became thinly scattered more posteriorly, and
were almost absent in the middle of the abdomen.
A scattering of whitish iridophores reappeared in
the vent region and extended posteriorly onto the
tail. White spots were strongly evident in the throat
region. The venter of the tail was mainly black. The
iris was dark brown.

There was, in general, great consistency in col-
oration of the specimens observed, with the pattern
being almost constant and differences mainly relat-
ing to hue and intensity of color. The animals from
Rockbound Canyon appear to be more vividly col-
ored than those from the type locality, with the
ground color being a deeper black and the light
coloration being brighter and more evident.

OSTEOLOGY. Information concerning osteolo-
gy is derived from six cleared and stained speci-
mens, and from radiographs of many of the type
series. The skull (Fig. 2) is typical of members of
the attenuate clade (see below), with a single pre-
maxillary bone, no prefrontal bone, and a large
fontanelle. The frontal processes of the premaxil-
lary are long and slender and may either remain
separate or fuse to each other between the expand-
ed and anteriorly protruding cartilaginous olfactory
capsules, which approach each other closely and
apparently squeeze the processes together. Distally
the processes separate and broaden to extensively
overlap the expanded anterior portions of the fron-
tals. The nasals are large bones that overlap the
well-developed septomaxillaries and frontals but
fall slightly short of contacting the maxillary fron-
tal processes. The latter do not contact any bones.
The frontals are in contact between the eyes and
for some distance posteriorly, so the fontanelle,
while very large, is not so large as in some of the
other attenuate Batrachoseps (illustrated in Mar-
low et al., 1979). The maxillaries are long and slen-
der and bear small teeth in a single series that ex-
tends well posterior to the center of the eyeball. The
parietals are well separated from each other, and
they are narrow, short bones that fall short of the
synotic tectum. There are no crests on the otic cap-
sules, but there is a small ridge-like ledge above the
articulation zone of the elongate, thin, and slender
squamosals. The quadrate is also a relatively small
bone, and the suspensorium is weak. The opercu-

lum either has no columellar rod, or it is reduced
to a barely perceptible projection. The vomers are
well articulated to each other and they bear a small
preorbital process that varies among individuals.
The process supports a row or more frequently a
patch of vomerine teeth. The process is less well
developed than in members of the robust clade of
Batrachoseps but more prominent than in other
members of the attenuate clade. Paired patches of
teeth underlie the large parasphenoid bone. Each
patch contains more than 100 tiny teeth. The lower
jaw is slender and consists of the elongate dentaries
and prearticulars, the latter having only a broad,
low coronoid process.

The hyolingual skeleton is typical of the genus in
being entirely cartilaginous, lacking the urohyal,
having elongated radii, and having long epibran-
chials that extend under the skin of the neck and
shoulder region as far posteriorly as the forelimb
insertion.

There are 19 or 20 trunk vertebrae, the last one
lacking ribs. There are two or three caudosacral
vertebrae. In other plethodontid genera the last
caudosacral is the first one to have a complete hae-
mal arch, but in Batrachoseps the arch may be
nearly complete on the second as well as the third
postsacral, or be complete on one or the other. Tails
of adults contain from 37 to 47 vertebrae, but sev-
eral show evidence of regeneration.

The limbs are long, and the tibial spur is usually
well developed and free, but it may be partly fused
to the tibia. Phalangeal formulas are 1-2-3 -2 for
both hands and feet. The hands and feet are rela-
tively broad and have well-developed terminal pha-
langes that are expanded and flattened (Fig. 3). The
mesopodials have the standard arrangement for the
genus (illustrated in Marlow et al., 1979). In most
individuals the ulnare and intermedium of the car-
pus are fused or joined, but they may be separate.
In one individual the intermedium and centrale are
fused in both carpi and both tarsi.

HABITAT AND DISTRIBUTION. Batracho-
seps gabrieli is known only from the immediate vi-
cinity of the type locality and another nearby lo-
cality in the upper San Gabriel River drainage on
the southern versant of the San Gabriel Mountain
range. The type locality is on the southeast margins
of a local flat (Pine Flats), about 1 km SE of a nat-
ural lake, Crystal Lake. The type locality is a talus
slope on a spur of South Mt. Hawkins (2,372 m
maximum elevation), which lies near the southern
end of a ridge extending south about 4 km from
the crest of the San Gabriel Mountains at Mt. Haw-
kins. The surrounding mountains form a large
semicircular half-ring, open to the southwest, with
a crest between 2,200 and 2,600 m. The talus slope
faces northwest and is shaded by large Canyon Live
Oak ( Quercus chrysolepis) and Big Cone Spruce
( Pseudotsuga macro carp a). Other vegetation on the
flat in the immediate vicinity includes Ponderosa
Pine ( Pinus ponderosa ), Coulter Pine ( Pinus coul-
ter /), Jeffrey Pine ( Pinus jeffreyi ), Sugar Pine ( Pinus

Contributions in Science, Number 463

Wake: New Species of Batrachoseps ■ 3

1 mm

Figure 2. Skull of a specimen (MVZ 178634) of Batrachoseps gabrieli. Cartilage not shown. The entire dorsal surface
of the skull is shown (below), and a ventral view of the premaxillary, maxillaries, and vomers is shown above. The dorsal
fontanelle is shown in stipple.

4 ■ Contributions in Science, Number 463

Wake: New Species of Batrachoseps

Figure 3. Right forelimb and right hind limb of a skeletal preparation of a specimen (MVZ 195583) of Batracboseps
gabrieli. Cartilage is stippled.

iambertiana ), and White Fir ( Abies concolor). In-
cense Cedar ( Calocedrus decurrens ), Black Oak
(Quercus kelloggii ), California Laurel { Umbellular-
ia californica), and Oregon Big-Leaf Maple ( Acer
macrophylium), as well as Western Sword Fern
{Polystichum munitum) occur on the talus slope.
Although this is a rich and diverse woodland, there
are many openings in the forest where native bunch
grasses and Yucca whipplei are common. Soldier
Creek rises from springs at the base of the talus
slope and has a strong winter flow but is dry at the
type locality throughout most of the year.

All but a few salamanders were found on the
steep talus slope, under large (ca 0.5 m long) rocks,
rotting logs, or downed tree limbs and bark (Fig.
4). The salamanders were occasionally found under
dried old fronds at the base of large ferns. A few
were found on soil along Soldier Creek at the base
of the talus slope, in sympatry with Ensatina escb-
scholtzii escbscholtzii, the only other salamander
known to occur in the area. Batracboseps nigriven -
tris occurs at lower elevations (1,200 m) less than
1 km to the SW at Falling Springs, along Soldier
Creek. Aneides iugubris has been taken a short dis-
tance below the type locality (2.6 km by air) at

Coldbrook Camp (1,000 m), where both Ensatina
and B. nigriventris are also found.

The activity of these salamanders near the sur-
face is probably limited to a few winter and early
spring months. At this elevation snow is present
nearly every winter and may persist for one or two
months. Salamanders have been observed in Feb-
ruary and March when scattered snow banks were
present. On 28 March 1985, when salamanders
were abundant, the air temperature was 3.5° C.
One small rock (10 X 20 cm) well set into the slope
but resting on a surface of smaller talus rocks had
a cavity beneath it that sheltered three adult sala-
manders. The soil temperature at this spot was 4.2°
C. Summer and fall drought probably drives the
animals deep into the talus slope.

Two specimens were collected on 29 March
1996, at a second locality, Rockbound Canyon,
about a km to the S of the type locality, but at a
lower elevation (1,158 m). One animal was found
under a rock about 10 m from a small flowing
stream, while the other was found under a log
about 15 m from the stream. The species occurs in
sympatry with B. nigriventris at this locality, and

Contributions in Science, Number 463

Wake: New Species of Batracboseps M 5

Figure 4. The habitat at the type locality of Batracboseps gabrieli, in the San Gabriel Mountains, Los Angeles County,
California. The steep talus slope is covered with rocks, leaf litter, and fallen branches and is deeply shaded for most of
the year. Photograph taken 19 February 1989. The collector in the photograph is Todd R. Jackman.

both species have been taken under the same log,
but separated by an interval of one week.

BEHAVIOR. No salamanders have been ob-
served active on the surface, but they have been
found under some rather superficial cover. When
salamanders are first uncovered they form a tight
coil that then slowly relaxes. When coiled the ani-
mals appear to be stout and brightly colored dor-
sally, but shiny black when the coil is turned over.
The species is alert and active; it rests in an s-
shaped posture with its limbs outspread and it
springs or leaps readily. The head is held above the
substrate, and the eyes in life are strongly protu-
berant.

ETYMOLOGY. The species is named for the
mountain range in which it is found. I recommend
that the species name be pronounced gha-bree-el-
ee.

DISCUSSION

There are two distinctive clades within Batracbo-
seps. Because of the relatively robust morphology
of the body and limbs of the members, I refer to
one of these as the robust clade; it includes only B.
wrighti (Bishop, 1937) (central and northern Cas-
cade Range, Oregon) and B. campi Marlow, Brode
and Wake, 1979 (Inyo Mountains, California). I re-
fer to the other as the attenuate clade; it includes
the remaining species (seven currently recognized,

including the new species, with approximately 8-
10 to be described; Marlow et al., 1979; Yanev,
1980). B. gabrieli is a member of the more speciose,
attenuate clade based on the osteological data re-
ported herein.

Members of the attenuate clade of Batracboseps
are similar in osteology, and study of six skeletons
of the new species reveals no especially distinctive
features, although the feet are broad and well de-
veloped and the digits are stout, with especially
well-developed and distally expanded terminal
digits. The vomers of the new species have short
but distinct, tooth-bearing preorbital processes.
There is a single premaxillary bone, and prefrontal
bones are absent. As is typical in this genus, there
are 2 or 3 (usual condition in this species) caudo-
sacral vertebrae. The dorsal fontanelle of the skull
is broadly open, again a characteristic of the genus.
Most species of Batracboseps have a columellar rod
of the operculum of moderate size, but none has
been found in B. gabrieli ; this may prove to be a
useful derived trait, but its distribution has yet to
be determined and it is difficult to observe.

The genus Batracboseps is currently under study
in my laboratory, and several manuscripts are in
preparation that will extensively revise current tax-
onomy. The point of departure for this revision is
the allozyme study by Yanev (1978, 1980), which
compared 105 populations. Yanev (1980) presented

6 ■ Contributions in Science, Number 463

Wake: New Species of Batracboseps

Figure 5. Photograph of three species of Batrachoseps that occur in close proximity in the mountains and adjacent
lowlands of southern California, to show relative proportions. Top, Batrachoseps gabrieli (from type locality); middle,
B. nigriventris (from Coldspring Camp, Los Angeles County); bottom, B. major (from lower Ortega Highway, Orange
County).

a phenogram of genetic distances for 18 “taxonom-
ic units” but recognized only seven species, one of
which was indicated as “sp. nov.” in her figures 1
and 2, but which was named Batrachoseps campi
(Marlow et al., 1979) by the time Yanev’s paper
was published. Yanev had no samples of an eighth
species, Batrachoseps aridus Brame, 1970, which
she predicted (correctly, based on new but unpub-
lished allozyme and mtDNA data from my labo-
ratory) would be found to be a close genetic relative
of her superspecies complex Batrachoseps pacificus
(Cope, 1865) (and not close to B. stebbinsi Brame
and Murray, 1968, which Brame, 1970, thought
was the closest relative of B. aridus ). Yanev showed
that the coastal attenuate taxa replace one another
in a series of parapatrically distributed units having
almost non-overlapping ranges that extend from
southwestern Oregon to northwestern Baja Cali-
fornia. Work in progress in my laboratory (using
data from allozymes, mtDNA sequences, and mor-
phology) will document even more extensive taxo-
nomic diversity than Yanev (1980) reported.

Several allopatric and parapatric populations of
varying degrees of morphological distinctiveness
were considered by Yanev (1980) to be semispecies;
she chose to recognize these taxonomically as sub-
species of a widespread superspecies, B. pacificus. I
recognize four species of Batrachoseps in southern
California below the region of the Santa Paula Riv-
er: B. aridus, B. gabrieli, B. nigriventris, and B. pa-
cificus. There are two morphologically and bio-
chemically distinct taxa of the pacificus superspe-

cies in southern California, B. p. pacificus and B.
p. major. Three taxa occur in the San Gabriel River
drainage of the San Gabriel Mountains: B. gabrieli,
B. p. major, and B. nigriventris, and the latter two
occur in sympatry at several localities on the south-
ern California mainland (Brame, 1970). The other
taxa are more restricted in distribution. B. aridus
has been found in two canyons on the north and
east margins of the Santa Rosa Mountains in the
Colorado Desert, and B. p. pacificus occurs on the
northern Channel Islands. B. p. major is found on
the southern Channel Islands as well as on the
mainland, and B. nigriventris occurs in sympatry
with B. p. pacificus on Santa Cruz Island. Distri-
bution of southern California Batrachoseps has
been summarized and mapped by Brame and Mur-
ray (1968), Brame (1970) (their B. attenuatus
(Eschscholtz, 1833) in southern California is pres-
ent-day B. nigriventris ), and Stebbins (1985).

The most slender, narrow-headed, and short-
limbed of these five taxa is B. nigriventris, a species
associated with upland habitats, mesic habitats, or
both, in southern California. This species occurs in
sympatry with B. gabrieli. The two species differ
sharply in morphology, with B. gabrieli having a
much broader head, longer limbs, larger hands and
feet, a shorter, more tapering tail, and a distinctive
color pattern (Fig. 5). The species overlap in num-
bers of trunk vertebrae, but B. gabrieli usually has
one or two fewer than B. nigriventris.

While B. p. major has a relatively broad head
and gets larger than B. gabrieli, it has shorter limbs

Contributions in Science, Number 463

Wake: New Species of Batrachoseps ■ 7

and smaller feet, has a pale, almost patternless
adult coloration (especially ventrally), and has a
very long, cylindrical tail (Fig. 5). B. gabrieli usu-
ally has one to two fewer trunk vertebrae. In gen-
eral, B. p. major is restricted to flatlands and open
country below 700 m elevation.

Both B. aridus and B. p. pacificus have broad
heads, but the former is a smaller species than B.
gabrieli (the holotype of B. aridus, at 48.4 mm SL,
is 18% larger than the next largest known speci-
men) and differs in coloration. Both species have
relatively short tails, shorter than in B. gabrieli. B.
p. pacificus is considerably larger and more robust
than B. gabrieli, has a head that is less flattened
and less distinct from the neck, and lacks the dis-
tinctive coloration of B. gabrieli.

There are two other species from the southern
Sierra Nevada and Tehachapi Mountains that re-
semble B. gabrieli in some morphological features,
the closest locality being about 100 km to the
northwest of the type locality. The species that oc-
curs at this locality (Ft. Tejon, where it occurs in
sympatry with topotypic B. nigriventris) is assigned
tentatively to B. stebbinsi. It is a talus-dweller and
resembles B. gabrieli in proportions and even in
some aspects of its color pattern (patches of bright
pigmentation on the tail). It is larger and more ro-
bust, however, and has longer limbs and a larger,
more rounded head. The two species have equiva-
lent numbers of trunk vertebrae, but B. stebbinsi
has a shorter tail. The other species, B. simatus
Brame and Murray, 1968, from the lower Kern
River Canyon, attains larger size than B. gabrieli
and has 21-22 trunk vertebrae. It has a broad, flat-
tened head, like B. gabrieli, but has longer limbs
and a more gracile appearance. While it has a ta-
pering tail of similar relative length to that of B.
gabrieli, it lacks the distinctive color pattern of the
latter.

Much unpublished allozyme and mtDNA se-
quence data exist for Batrachoseps, and they will
be published in detail elsewhere by me and others.
Here only some results relevant to the new taxon
are presented. Batrachoseps gabrieli is highly dif-
ferentiated genetically from all other taxa, such that
it is not possible to infer what its sister taxon might
be. Nei’s genetic distances (DN) derived from allo-
zyme studies (for details of methods see Yanev and
Wake, 1981; Nei, 1972) are all large. The lowest
value of Dn to any other taxon is DN = 0.65, which
is measured to different populations of B. p. major
and to the Ft. Tejon population tentatively assigned
to B. stebbinsi. There are fixed differences separat-
ing the taxa at 12 of the 26 allozymic loci sampled.
Smallest DN to other taxa are: B. nigriventris 0.70,
B. p. pacificus 0.71, B. aridus 0.86, B. simatus
0.94.

With respect to mtDNA sequences, differentia-
tion is also great, with no obvious close relatives,
although it falls within a pacificus complex (includ-
ing the B. pacificus superspecies of Yanev, 1980,
and B. aridus ; Jockusch, 1996). Unpublished se-

quence data are available for the cytochrome B
gene from more than 80 populations representing
all taxa in the genus (more than 350 base pairs for
all individuals, more than 700 base pairs for most
individuals; Jockusch, 1996). The sequence of B.
gabrieli is again distinctive, and differs by more
than 10% (corrected for multiple hits; Kimura,
1980) sequence divergence from all other popula-
tions. The lowest level of differentiation is 10.1%,
observed between B. gabrieli and an unnamed pop-
ulation of the pacificus complex from Monterey
County. A difference of 11.1% is measured be-
tween B. gabrieli and a Santa Cruz Island popula-
tion of B. p. pacificus, and a difference of 12.0%
is recorded to a southern California population of
B. p. major. The new species is more than 15%
different from B. stebbinsi. B. gabrieli is about
equally distant to many populations of different
species in the attenuate clade, and the range extends
out to about 20% divergent to different popula-
tions of the most remote member of the clade, B.
attenuatus. Divergence from members of the robust
clade is even greater and ranges from 23.6-36.5%.

The conclusion from the genetic data is that B.
gabrieli is strikingly distinct relative to all conge-
neric taxa and populations. Various methods have
been used to calibrate a “molecular clock” for al-
lozymes. Yanev (1980), following Sarich (1977),
used a calibration of DN = 1 being equivalent to
about 20 million years of divergence. Calibrations
are most appropriate when based on independent
geological events and inferred vicariant events (e.g.,
as in Good and Wake, 1992). Yanev showed a rel-
atively good correspondence between inferred vi-
cariant events (based on geological dating) and her
molecular clock estimates. If we use her calibration
we obtain a divergence time of B. gabrieli from oth-
er extant members of the genus of about 13 million
years. If we use a more conservative estimate based
on comparisons from several plethodontid genera
of allozymic data with albumin immunological dis-
tances (Maxson and Maxson, 1979), we obtain a
time of about 9 million years; this is approximately
what would be obtained from use of the indepen-
dent calibration for a non-plethodontid salaman-
der, Rhyacotriton (Good and Wake, 1992). Anoth-
er calibration for the salamandrid genus Taricha
(Tan, 1993) yields a slightly shorter time. So, under
any time estimation scheme currently available, B.
gabrieli represents a lineage that has been separated
from other lineages within the genus for a very long
time, perhaps for 8 to 13 million years.

TAXONOMIC COMMENTS. The Batracho-
seps of southern California have long confused tax-
onomists, and the resultant taxonomic history is
complex. Cope (1865) described B. pacificus (type
locality listed as Santa Barbara, considered to be an
error by Van Denburgh, 1905, who assumed the
type specimen was obtained from the northern
Channel Islands). Later Cope (1869) described B.
nigriventris (type locality Ft. Tejon, Kern County).
The California species of the genus were ignored

8 ■ Contributions in Science, Number 463

Wake: New Species of Batrachoseps

for many years, until the description of B. major
(Camp, 1915; type locality Sierra Madre, Los An-
geles County). The Grinnell and Camp (1917) dis-
tributional list recognized three species in southern
California: a wide-spread slender, mainly northern
form B. attenuatus (type locality “Umgebung der
Bai St. Francisco auf Californien” Eschscholtz
1833), B. major, and B. pacificus. Grinnell and
Camp listed B. nigriventris as a synonym of B. at-
tenuatus, which was considered to be the most
widespread taxon, extending from the region of
San Francisco Bay and the western slopes of the
central and southern Sierra Nevada through south-
ern California and into northwestern Baja Califor-
nia (although they carefully noted “south at least
to mountains immediately north of Claremont, Los
Angeles County”), including Santa Catalina Island.
B. pacificus was known only from the three largest
of the northern Channel Islands, and B. major only
from Pasadena and Sierra Madre. Fowler and Dunn
(1917) also synonymized B. nigriventris with B. at-
tenuatus. Dunn (1922) described two more insular
species, B. leucopus Dunn, 1922 (type locality Los
Coronados, North Island) and B. catalinae Dunn,
1922 (type locality Santa Catalina Island). Storer
(1925) recognized four species: B. pacificus from
the northern Channel Islands, B. catalinae from
Santa Catalina Island (which he believed was a val-
id species), B. major from Los Angeles and River-
side counties, and B. attenuatus, which was listed
from Ft. Tejon, the Santa Monica Mountains, and
the mountains north of Claremont. Storer also in-
cluded specimens from the Sierra San Pedro Martir
in northern Baja California, Mexico, in B. attenu-
atus, and noted that the species had been recorded,
but not confirmed, from La Paz, Baja California.
Storer made no comment concerning B. leucopus,
presumably because it occurred only in Mexico and
was not part of the California fauna.

In his famous monograph on plethodontid sala-
manders, Dunn (1926) reduced all taxa of Batra-
choseps to subspecies of B. attenuatus, commenting
that “the forms of Batrachoseps present rather a
problem to the systematist, the question being how
many forms to recognize and on what characters.
There is apparently no overlapping of ranges and
the species is one with a number of more or less
emphasized local races, and with very indefinite
and variable characters” (p. 230). Dunn placed sal-
amanders from San Diego Co. and the Sierra San
Pedro Martir Mountains of northern Baja Califor-
nia in B. a. leucopus. His B. a. attenuatus occurred
as far south as the Santa Monica and San Gabriel
mountains, while B. a. major occurred in Los An-
geles, Riverside, and Orange counties. Finally, B. a.
catalinae was restricted to Santa Catalina Island.

Slevin (1928) followed Dunn (1926) in most re-
spects, but he synonymized B. a. catalinae with B.
a. attenuatus, extended the range of the latter
southward to Laguna Beach, Orange Co., and east-
ward to San Bernardino, San Bernardino Co., and
considered B. a. major to be restricted to Los An-

geles and Riverside counties. At the time few spec-
imens had been examined, and field work was in-
sufficient to resolve the taxonomic problems. This
situation was corrected by Campbell (1931), whose
revised taxonomy recognized two species, B. paci-
ficus (with three subspecies: pacificus, major, catal-
inae) and B. attenuatus (with two southern Cali-
fornia subspecies: attenuatus, leucopus). Campbell
was familiar with the organisms in the field, and he
studied large samples of preserved specimens. He
reported two instances of sympatry: 1) between B.
a. attenuatus and B. pacificus on Santa Cruz Island,
and 2) between B. p. major and B. a. attenuatus in
South Pasadena and the Monterey Hills, Los An-
geles Co. His B. a. leucopus was said to express
characters of both species, and the range was that
recorded by Dunn (1926).

Hilton (1945) overlooked Campbell’s important
work and regressed to the taxonomy of Slevin
(1928). Hilton’s work is characteristically erratic,
adding some useful observations, but in some re-
spects confusing matters. For example, he reports
that B. a. major occurs at higher elevations than his
B. a. attenuatus, while the reverse is true.

Lowe and Zweifel (1951) and Stebbins (1951)
agreed that there were two species in southern Cal-
ifornia and reported three sites at which they occur
in sympatry, the two reported by Campbell (1931)
and a third near Irvine Park, Orange Co. However,
these authors thought they detected evidence of an
intermixing or introgression southeast of Redlands
in San Bernardino Co. Stebbins (1951) treated ma-
jor and catalinae as subspecies of pacificus, and leu-
copus as a subspecies of attenuatus.

A monographic study by Hendrickson (1954)
was regressive. He simplified taxonomy by recog-
nizing a single species in southern California, B.
attenuatus, with two subspecies, pacificus (includ-
ing major) and attenuatus. While he accepted that
there was some sympatry, he felt that there was
substantial intergradation and treated the popula-
tion reported by Lowe and Zweifel (1951) and
Stebbins (1951) from near Redlands as a hybrid
swarm. Hendrickson referred to a taxonomically
awkward category of “intermediate” populations,
which encompassed leucopus, catalinae, and, in
part, major.

Savage and Brame undertook a detailed study of
southern California Batrachoseps in the 1950s and
produced a still unpublished manuscript that has
influenced subsequent taxonomy. Peabody and Sav-
age (1958), citing the above manuscript, returned
to the Campbell (1931) taxonomy, except that leu-
copus was considered to be a subspecies of pacificus
instead of attenuatus, and catalinae was not rec-
ognized as distinct and was placed in the species B.
pacificus. Brame and Murray (1968) showed that
major and pacificus were morphologically distinct
and recognized them as separate species. Their B.
pacificus was restricted to the northern Channel Is-
lands, while B. major (including leucopus and ca-
talinae) was said to occur on the mainland south

Contributions in Science, Number 463

Wake: New Species of Batrachoseps ■ 9

of the Santa Monica and San Gabriel mountains,
and on the southern islands: Santa Catalina, Los
Coronados, and Todos Santos. They also reaf-
firmed the distinctiveness of B. attenuatus and B .
major, but referred the Santa Cruz Island popula-
tion formerly assigned to B. attenuatus to a new
species, B. relictus Brame and Murray, 1968 (type
locality in Kern River Canyon, Kern County). The
Batracboseps in the Sierra San Pedro Martir also
were assigned to B. relictus. Later Brame (1970)
described B. aridus (type locality Hidden Palm
Canyon, Riverside County).

Starting in the 1970s, genetic (allozymic) studies
of the genus were undertaken by Yanev, and these
remain largely unpublished. However, a general
summary of her findings, published in 1980, has
influenced present taxonomy (see especially Steb-
bins, 1985). She showed that B. nigriventris is dis-
tinct from B. attenuatus (which does not extend
below central California), and she assigned the
Santa Cruz Island population considered by Brame
and Murray as B. relictus to B. nigriventris. She
reduced relictus and major (of Brame and Murray,
1968) to subspecies of B. pacificus.

At present I recognize the following taxa of Ba-
trachoseps in southern California and adjacent
Mexico:

1. B. p. pacificus occurs on San Miguel, Santa
Rosa, Santa Cruz, and the Anacapa Islands (but not
on Santa Barbara Island, as sometimes indicated).

2. B. p. major (including catalinae and leucopus,
which I consider to be subjective junior synonyms)
occurs on the mainland and on the following south-
ern islands: Islas Todos Santos, Islas Coronados,
and Santa Catalina. I tentatively include the pop-
ulations in the Sierra San Pedro Martir (assigned to
B. relictus by Brame and Murray, 1968, and iden-
tified as an unnamed new subspecies by Yanev,
1980) in this taxon (as did Stebbins, 1985). B. p.
major is widespread in the open lowland areas of
southern California. These areas are now heavily
populated, with accompanying habitat destruction,
and the once common salamanders are difficult to
find in much of the region. B. p. major occurs well
to the east, near Cabazon and in Palm Springs, rel-
atively near to the type locality of B. aridus (Cor-
nett, 1981). As shown by Brame and Murray
(1968), the two subspecies of B. pacificus are mor-
phologically distinct (B. p. pacificus has a broader
head, longer limbs, a shorter tail, and is more sim-
ilar to B. gabrieli in appearance than is the more
cylindrical, elongate, and paler B. p. major). How-
ever, pending completion of studies in progress, I
continue to use the taxonomy of Yanev (1980).

3. B. nigriventris occurs on Santa Cruz Island
and in upland and more mesic coastal zones, and
in the San Gabriel Mountains. The map presented
by Brame (1970) for B. attenuatus shows most of
the currently known range of B. nigriventris in
southern California. It is the slender, dark-bellied,
short-limbed species of upland regions, and it oc-
curs in sympatry with B. p. major at various places

in Los Angeles and Orange counties, including the
Baldwin Hills, the Palos Verdes region, the Mon-
terey Hills, the Puente Hills, and the Chino Hills.
It also occurs in the San Joaquin Hills and along
the coast as far south as Aliso Creek, Orange Co.,
and we recently found it in Tenaja Canyon, extreme
southwestern Riverside Co., the southeastern limit
of its known distribution.

4. B. aridus is found only in two isolated desert
canyons, the type locality in Hidden Palm Canyon
and in Guadelupe Canyon. This species is distin-
guished from B. pacificus major, which is the closest
geographical neighbor, by its smaller size, broader
head, and much shorter tail, and by its distinctive
color pattern (brightly colored flecks which form
poorly bounded patches, well displayed in the il-
lustration in Brame and Hansen, 1994).

5. B. gabrieli is known only from the two local-
ities reported herein in the San Gabriel Mountains.

CONCLUDING REMARKS. I was surprised to
discover that the population on which this species
is based was highly differentiated genetically from
all other populations with which it was compared.
Apparently this species is restricted to the general
vicinity of the type locality. A number of apparently
suitable microhabitats in the San Gabriel Moun-
tains have been searched unsuccessfully. Given this
limited distribution and deep differentiation, I infer
that the species is a relic that requires special pro-
tection. The habitat is restricted and fragile, and
special efforts should be taken to protect the area.
The known habitat, two small areas of about one
hectare in extent, are vulnerable. The type locality
lies on the margins of a once developed recreation
area in the vicinity of several large springs. The old
development has been totally removed except for
some nearby stone foundation remnants, but the
area has been subject to disturbance from casual
human use. The type locality is in the Crystal Lake
Recreation Area of Angeles National Forest, and I
urge administrators of the forest and the California
Resources Agency to give special protection to the
region.

I believe that the genus Batracboseps is tropical
in origin and that it has moved to the northwest
from its origins in association with the complex
movements of the crust of this planet over a period
of many years (Hendrickson, 1986; Wake, 1987).
This new population was discovered by chance,
and it is possible that more undiscovered popula-
tions exist in southern California. I urge that special
attention be given to careful exploration of south-
ern California and adjacent Mexico for additional
populations of this genus.

Batracboseps is comprised of a large number of
genetically differentiated groups, and the current
taxonomy is inadequate to express the extent of the
evolutionary diversification of the lineage. The new
species described herein adds another entity to this
complex group. Work in progress in my laboratory
will explore the evolutionary history of this lineage.

10 ■ Contributions in Science, Number 463

Wake: New Species of Batracboseps

ACKNOWLEDGMENTS

I first collected this species in 1982, in the company of
Marvalee H. Wake and Ronald W. Marlow, who were
helping me search for blotched populations of Ensatina in
this region of “Bob’s Gap,” where there were possible sight
records of the species (Jackman and Wake, 1994). I did
not consider the Batrachoseps to be special, although I
wrote in my field notes that they were surprisingly large
for B. nigriventris. Monica Frelow convinced me they
were special, based on her extensive technical work for
me on allozymes of members of the genus. The mtDNA
data were gathered by Elizabeth Jockusch and Geoff
Applebaum. Among those who have accompanied me on
searches to the area are A. Graybeal, T. Jackman, S.
Marks, M. Mahoney, N. Staub, S. Sweet, A. Tate, S. Tilley,
J. Tilley, and T.A. Wake. I thank Robert H. Goodman, Jr.,
for sharing his discovery of the new population of B. ga-
brieli with me, for providing relevant information, and for
sending the specimens for my examination and for depo-
sition in MVZ. I thank R. Bezy, S. Deban, M. Garcia Par-
is, R. Highton, E. Jockusch, M. Mahoney, S. Minsuk, B.
Stein, and K. Zamudio for discussion and helpful com-
ments on the manuscript. My work has been supported
by grants from the National Science Foundation and U.
S. Forest Service (through the Biodiversity Research Con-
sortium).

LITERATURE CITED

Brame, A.H., Jr. 1970. A new species of Batrachoseps
(slender salamander) from the desert of southern
California. Contributions in Science 200:1-11.
Brame, A.H., Jr., and R.W. Hansen. 1994. Desert slender
salamander. In Life on the edge, ed. C.G. Thelander,
248-249. Santa Cruz, Calif.: Biosystems Books.
Brame, A.H., Jr., and K.F. Murray. 1968. Three new slen-
der salamanders ( Batrachoseps ) with a discussion of
relationships and speciation within the genus. Sci-
ence Bulletin, Museum of Natural History, Los An-
geles County 4:1-35.

Camp, C.L. 1915. Batrachoseps major and Bufo cognatus
californicus, new amphibia from southern Califor-
nia. University of Calif ornia Publications in Zoology
12:327-329.

Campbell, B. 1931. Notes on Batrachoseps. Copeia 1931:
131-134.

Cope, E.D. 1865. Third contribution to the herpetology
of tropical America. Proceedings of the Academy of
Natural Sciences, Philadelphia 17:185-198.

Cope, E.D. 1869. A review of the species of Plethodonti-
dae and Desmognathidae. Proceedings of the Acad-
emy of Natural Sciences, Philadelphia 21:93-118.
Cornett, J.W. 1981. Batrachoseps major (Amphibia: Cau-
data: Plethodontidae) from the Colorado Desert.
Bulletin of the Southern California Academy of Sci-
ences 80:94-9 5.

Dunn, E.R. 1922. Two new insular Batrachoseps. Copeia
109:60-63.

Dunn, E.R. 1926. The salamanders of the family Pletho-
dontidae. Northampton, Mass.: Smith College, 441

pp.

Eschscholtz, F. 1833. Zoologischer Atlas, enthaltend Ab-
bildungen und Beschreibungen neuer Thierarten
wdhrend des Flottcapitains von Kotzebue zweiter
Reise um die Welt, auf der Russisch-Kaiserlichen
Kriegsschlupp Predpriaetie in den Jahren 1823-
1826. Berlin, part V, viii + 28 pp., plates XXI-XXII.
Fowler, H.W., and E.R. Dunn. 1917. Notes on salaman-

ders. Proceedings of the Academy of Natural Sci-
ences, Philadelphia 1917:7-28.

Good, D.A., and D.B. Wake. 1992. Geographic variation
and speciation in the torrent salamanders of the ge-
nus Rhyacotriton. University of California Publica-
tions in Zoology 126:1-91.

Grinnell, J., and C.L. Camp. 1917. A distributional list of
the amphibians and reptiles of California. University
of California Publications in Zoology 17:127-208.

Hendrickson, D.A. 1986. Congruence of bolitoglossine
biogeography and phylogeny with geologic history:
Paleotransport on displaced suspect terranes? Cla-
distics 2:113-129.

Hendrickson, J.R. 1954. Ecology and systematics of sal-
amanders of the genus Batrachoseps. University of
California Publications in Zoology 54:1-46.

Hilton, W.A. 1945. Distribution of the genus Batracho-
seps, especially on the coastal islands of southern
California. Bulletin of the Southern California Acad-
emy of Sciences 44:101-129.

Jackman, T.R., and D.B. Wake. 1994. Evolutionary and
historical analysis of protein variation in the
blotched forms of salamanders of the Ensatina com-
plex (Amphibia: Plethodontidae). Evolution 48:876-
897.

Jockusch, E.L. 1996. Evolutionary studies in Batrachoseps
and other plethodontid salamanders: Correlated
character evolution, molecular phylogenetics, and re-
action norm evolution. Ph.D. thesis in Integrative Bi-
ology, University of California, Berkeley.

Kimura, M. 1980. A simple method for estimating evo-
lutionary rate of base substitutions through compar-
ative studies of nucleotide sequences. Journal of Mo-
lecular Evolution 16:11 1—120.

Lowe, C.H., Jr., and R.G. Zweifel. 1951. Sympatric pop-
ulations of Batrachoseps attenuatus and Batracho-
seps pacificus in southern California. Bulletin of the
Southern California Academy of Sciences 50:128-
135.

Marlow, W.R., J.M. Brode, and D.B. Wake. 1979. A new
salamander, genus Batrachoseps, from the Inyo
Mountains of California, with a discussion of rela-
tionships in the genus. Contributions in Science 308:
1-17.

Maxson, L.R., and R.D. Maxson. 1979. Rates of molec-
ular and chromosomal evolution in salamanders.
Evolution 33:734-740.

Nei, M. 1972. Genetic distance between populations.
American Naturalist 106:283-292.

Peabody, F.E., and J.M. Savage. 1958. Evolution of a
Coast Range corridor in California and its effect on
the origin and dispersal of living amphibians and
reptiles. In Zoogeography, ed. C.L. Hubbs, 159-186.
Washington, D.C.: American Association for the Ad-
vancement of Science.

Sarich, V.M. 1977. Rates, sample sizes, and the neutrality
hypothesis for electrophoresis in evolutionary stud-
ies. Nature (London) 265:24-28.

Slevin, J.R. 1928. The amphibians of western North
America. Occasional Papers of the California Acad-
emy of Sciences 16:1-152.

Stebbins, R.C. 1951. Amphibians of western North Amer-
ica. Berkeley: University of California Press, ix +
539 pp.

Stebbins, R.C. 1985. A field guide to western reptiles and
amphibians, second edition, revised. Boston: Hough-
ton Mifflin, xiv + 336 pp.

Storer, T.I. 1925. A synopsis of the Amphibia of Califor-

Contributions in Science, Number 463

Wake: New Species of Batrachoseps ■ 11

nia. University of Calif ornia Publications in Zoology
27:1-308, pis. 1-18.

Tan, A.-M. 1993. Systematics, phytogeny and biogeogra-
phy of the northwest American newts of the genus
Taricha (Caudata: Salamandridae). Ph.D. disserta-
tion, University of California, Berkeley, 296 pp.

Van Denburgh, J. 1905. The reptiles and amphibians of
the Pacific Caost of North America from the Faral-
lons to Cape San Lucas and the Revilla Gigedos.
Proceedings of the California Academy of Sciences
(ser. 3) 4:1-40.

Wake, D.B. 1987. Adaptive radiation of salamanders in
Middle American cloud forests. Annals of the Mis-
souri Botanical Garden 74:242-264.

Yanev, K.P. 1978. Evolutionary studies of the plethodon-
tid salamander genus Batrachoseps. Ph.D. thesis in
Zoology, University of California, Berkeley.

Yanev, K.P. 1980. Biogeography and distribution of three
parapatric salamander species in coastal and border-
land California. In The California Islands: Proceed-
ings of a multidisciplinary symposium, ed. D.M.
Power, 531-550. Santa Barbara, Calif.: Santa Bar-
bara Museum of Natural History.

Yanev, K.P., and D.B. Wake. 1981. Genic differentiation
in a relict desert salamander, Batrachoseps campi.
Herpetologica 37:16-28.

Received 16 April 1996; accepted 22 August 1996.

12 ■ Contributions in Science, Number 463

Wake: New Species of Batrachoseps

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

.{A*

Number 464
12 December 1996

Contributions
in Science

Suction Feeding in Beaked Whales:
Morphological and Observational
Evidence

John E. Heyning and James G. Mead

^9^ Natural History Museum of Los Angeles County

Serial
Publications
ol THE
Natural History
Museum ol
Los Angeles
County

Scientific

Publications

Committee

Robert J. Lavenberg, Acting Deputy Director
for Research and Collections
John M. Harris, Committee Chairman
Brian V. Brown
Kenneth E. Campbell
Kirk Fitzhugh
Karen Wise

Robin A. Simpson, Managing Editor

# Contributions in Science, a miscellaneous series of tech-
nical papers describing original research in the life and
earth sciences.

# Science Series, long articles and collections of papers on
natural history topics.

Copies of the publications in these series are sold through
the Museum Book Shop. A catalog is available on request.

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

Printed at Allen Press, Inc., Lawrence, Kansas
ISSN 0459-8113

Suction Feeding in Beaked Whales:

Morphological and Observational
Evidence

John E. Heyning1 and James G. Mead2

ABSTRACT. Beaked whales possess several unique morphological features of the head. In most species,
females and juveniles lack erupted teeth; the erupted teeth of adult males are used primarily for intraspecific
fighting. All beaked whales possess one pair of throat grooves. Several authors have hypothesized that
some toothed whales may capture prey primarily by suction. To test this hypothesis, we dissected a series
of beaked whales and other cetaceans to ascertain the functional morphology related to this mode of
feeding. The throat grooves of beaked whales allow for expansion of the gular region. A layer of loose
connective tissue between muscle groups in the floor of the mouth allows the tongue to move freely as it
is retracted posteriorly by the contractions of the hyoglossus and styloglossus muscles. The combined effects
of tongue retraction and gular floor distention appear to function to create a sudden drop in intraoral
pressure allowing beaked whales to suck in prey items. The relatively large hyoid bones provide the origin
for several large gular muscles that control the tongue and the floor of the mouth. Observations of live
animals corroborate the anatomical findings that beaked whales use suction to acquire prey.

INTRODUCTION

Toothed whales differ from mysticetes in that they
typically pursue, capture, and swallow single prey
rather than filter out numerous prey items simul-
taneously. The morphological adaptations and for-
aging behavior that have evolved relating to feeding
vary greatly among odontocetes. Most species have
supernumerary teeth and a secondarily homodont
dentition that is used to pierce and hold the prey
of odontocetes: fish and/or squid (Norris and Mohl,
1983; ITeyning, 1989a). Oceanic dolphins (Del-
phinidae) of the genera Stenella Gray, 1866 and
Delphinus Linnaeus 1758 may have as many as
200 teeth (Leatherwood et al., 1988:216) within
long, slender snouts, which have been referred to
as pincer-type jaws by Norris and Mohl (1983).

The beaked whales (Family Ziphiidae) are char-
acterized by the numerical reduction of erupted
teeth. Only in one extant species, Shepherd’s
beaked whale ( Tasmacetus shepherdi Oliver, 1937),
does a full complement of teeth erupt in both jaws
(Mead, 1989a). In all other extant species, only one
or two pairs of mandibular teeth erupt at sexual
maturity. In both Arnoux’s and Baird’s beaked
whales (genus Berardius Duvernoy, 1851), two
pairs of apical teeth erupt in both sexes. In Cuvier’s
beaked whale ( Ziphius cavirostris Cuvier, 1823),
both species of bottlenose whales (genus Hyperoo-
don Lacepede, 1804), and all 13 species of Meso-

1. Section of Vertebrates, Natural History Museum of
Los Angeles County, 900 Exposition Blvd., Los Angeles,
California 90007.

2. Division of Mammals, National Museum of Natural
History, Washington, D.C. 20560.

Contributions in Science, Number 464, pp. 1-12
Natural History Museum of Los Angeles County, 1996

plodon Gervais, 1850, only one pair of teeth erupts
in sexually mature males.

All extant species of ziphiids possess at least one
pair of anteriorly converging throat grooves (Heyn-
ing, 1989a). Some specimens of Baird’s beaked
whale ( Berardius bairdii Stejneger, 1883) may have
accessory throat grooves (Omura et al., 1955).

Ziphiids are often described as primarily squid-
eaters (e.g., Gaskin, 1982; Clarke, 1986), however,
they also may consume large quantities of fish and/
or other invertebrates (Mead, 1989a,b,c; Reyes et
al., 1991; Heyning, 1989b; Balcomb, 1989).

Thus, how do these functionally edentulous
odontocetes capture their prey? We tested the hy-
pothesis that ziphiids capture their prey primarily
by means of suction.

It has been suggested that many other groups of
odontocetes use suction as a means to acquire food
items (Caldwell et al., 1966; Norris and Mohl,
1983; Brodie, 1989). Several authors have suggest-
ed that the tongue is employed in a piston-like man-
ner to produce suction in order to capture prey
items (Caldwell et al., 1966; Norris and Mohl,
1983), although none provided evidence that this
occurs. Lawrence and Schevill (1965) provided de-
tailed anatomical descriptions of delphinid gular
musculature but limited their functional interpre-
tations to the laryngeal region. An analysis of suc-
tion feeding within certain delphinoid cetaceans is
in progress by Werth (1989, 1991).

For suction feeding to occur, some mechanism
must allow the intraoral pressure to drop relative
to that of the surrounding water (Lauder, 1985).
This would require an increase in volume of the
oral and/or pharyngeal cavities. Therefore, we

Table 1. Specimens examined for this study. Institutional acronyms are: LACM, Natural History Museum of Los Angeles
County; USNM, National Museum of Natural History; CAS, California Academy of Sciences.

Species

Sex

Length

Specimen number

SPECIMENS DISSECTED

Ziphiidae

Berardius baridii

10.30

LACM 86029

10.10

LACM 86033

10.55

LACM 86030

9.50

LACM 86031

field #91-21

Mesoplodon densirostris

2.72

USNM 571325

4.20

USNM 550754

Mesoplodon europaeus

3.44

USNM 571354

Mesoplodon mirus

4.50

USNM 571357

Mesoplodon stejnegeri

2.32

LACM 84299

Delphinidae

Orcinus orca

6.65

LACM 84249

4.50

LACM 84291

Tursiops truncatus

3.17

LACM 84269

Grampus griseus

3.06

LACM 84298

Delpbinus delpbis

1.91

LACM 86000

Balaenopteridae

Balaenoptera acutorostrata

USNM 571487

Balaenoptera physalus

? p

LIVE SPECIMENS OBSERVED

USNM 571762

Ziphiidae

Mesoplodon carlbubbsi

2.87

CAS 23751

2.99

CAS 23122

Ziphius cavirostris

3.25

released alive

asked the following series of questions: Are the
throat grooves distensible? Does the tongue func-
tion as a piston, and if so, what muscles enable it
to do so? Is the arrangement of the muscles and/or
hyoid apparatus different in suction versus nonsuc-
tion feeders? We also noted any observational data
that support the hypothesis of suction feeding.

MATERIALS AND METHODS

We dissected a series of nonpreserved heads from ziphiids
(Table 1) to ascertain both gross and functional morphol-
ogy. Dissections varied in technique, with some progress-
ing from superficial to deep and others from deep to su-
perficial. These dissections focused on the morphology of
the musculature, connective tissue, and bone. A series of
delphinid and baleen whale (Balaenopteridae) heads was
also dissected for comparative purposes. Dissections were
documented by photographs, illustrations, and video re-
cordings. One specimen of the Bering Sea beaked whale
(Mesoplodon stejnegeri True, 1885, LACM 84299) was
serially scanned by Computer Assisted Tomography
(CAT) to provide three-dimensional orientations of ana-
tomical structures. For anatomical structures, we followed
the nomenclature of Lawrence and Schevill (1965).

To test whether or not the hyoid apparatus differs be-
tween ziphiids and delphinids, we measured the maximum
length, width, and thickness of the thyrohyoid bones from
a series of specimens. The thyrohyoid was chosen because

it is the site of attachment for the sternohyoideus and hy-
oglossus muscles (Lawrence and Schevill, 1965). We in-
cluded only those species of delphinids that we suspected
do not use suction feeding to any great extent because they
possess either numerous teeth and have pincer-type jaws
(bottlenose dolphin, Tursiops truncatus (Montagu, 1821)
and short-beaked common dolphin, Delpbinus delphis
Linnaeus, 1758) or consume relatively large prey that can-
not be easily sucked in (e.g., killer whale, Orcinus orca
(Linnaeus, 1758) and false killer whale, Pseudorca eras-
sidens (Owen, 1846)). In addition, the total lengths for
the specimens of Orcinus and Pseudorca approximated or
exceeded the total lengths for the beaked whale specimens
of Ziphius and Mesoplodon. This overlap in total length
compensates for any allometric scaling in the hyoid bones.

To test whether intraoral volume can be increased, we
measured the volume of the oropharynx for an adult An-
tillean beaked whale (M. mirus True, 1913, USNM
571357). Volume measurements were made by suturing
closed the esophagus at the level of the occipital condyles
and filling the oral cavity with water. We assumed that at
rest, the oral cavity and esophagus are a potential space.
Thus, measurements should approximate a minimum vol-
ume of fluid the animal could potentially draw in during
suction feeding. We also manipulated anatomical struc-
tures, such as the tongue, in planes parallel to different
muscle groups to determine what effects contractions of
these muscles might evoke.

Measurements of the maximum gape were collected

2 ■ Contributions in Science, Number 464

Heyning and Mead: Suction Feeding in Ziphiids

Figure 1. Right lateral view of an adult male Mesoplodon densirostris (USNM 550754) showing the relaxed state (left)
and the distended state (right) of the gular region by expansion (arrow) of throat grooves. Distension was done manually
and probably does not represent the maximum possible in a living ziphiid.

from ziphiids and delphinids, including the maximum ex-
tent of jaw opening on a series of specimens of Baird’s
beaked whale, B. bairdii. The gape was measured from
the tip of the upper jaw to the tip of the lower jaw as the
jaws were opened by an electrical winch. The amount of
force exerted in this manner would presumably represent
the maximum opening of the mouth without possible tis-
sue damage.

In two instances we studied live-stranded beaked whale
calves that were temporarily held in captivity. In the first
instance, the gular region on two calves of Hubb’s beaked
whales, M. carlhubbsi Moore, 1963, was palpated to de-
termine if the throat grooves are distensible; we also pal-
pated the oral cavity of the calves to determine if suction
could be detected. In the second instance, a series of video
recordings were taken of a young stranded female Cuvier’s
beaked whale, Z. cavirostris. The tips of the heads of
whole fishes were placed into the gape of this beaked
whale, and feeding was recorded. Using stop-frame anal-
ysis, the elapsed time from when the fish began to move
into the oral cavity until it disappeared completely into
the mouth could be documented.

RESULTS

FUNCTIONAL MORPHOLOGY AND
OBSERVATIONS

We found that in the species of ziphiids examined
the throat grooves are situated between the man-
dibular symphysis and the hyoid apparatus. In spe-
cies with a long mandibular symphysis, such as the
strap-toothed whale (M. layardii (Gray, 1865)), the
anterior termini of the throat grooves are situated
relatively farther posterior to the tip of the lower
jaw than in those species that have a short sym-
physis, such as M. mirus (see Mead, 1989c: figs.
11,12).

Both live-stranded M. carlhubbsi calves exhibited
a strong sucking response when either feeding upon
fish or suckling on a finger. When this sucking ac-
tion was being produced, most of the movement
detectable by external palpation was in the region
of the throat grooves. This region was actively ex-
panded and contracted by muscular activity. Only
during respiration was movement detected caudal

Contributions in Science, Number 464

to the hyoid apparatus. When suckling on a finger,
these calves used their tongues to maneuver the fin-
ger to the roof of the mouth and hold it there. The
frayed anterior and lateral margins of the tongue
assisted in forming a tight seal around the finger
for suckling.

During dissections, the tongue could easily be re-
tracted posteriorly toward the hyoid apparatus by
manipulation. The tongue could move fore and aft
smoothly over a loose connective tissue interface
between the geniohyoideus muscle and the extrinsic
tongue muscles ( genioglossus , styloglossus, hy-
oglossus) (Fig. 2). The region of the throat grooves
was distended when the tongue was retracted. This
distension increased when moderate pressure was
applied by hand to the floor of the mouth (Fig. 1).

Anterior and lateral to the tongue, the floor of
the mouth possesses numerous folds when the
tongue is situated in its relaxed, forward position
(Fig. 3). When the tongue is retracted, these folds
unfurl like pleats. When relaxed, the dorsal surface
of the tongue is slightly concave. This shape ap-
proximately corresponds with that of the convex
palate. Retracting the tongue by manipulation
causes the medial surface of the tongue to depress
and the lateral edges to turn dorsally: the tongue
becomes U-shaped in transverse section.

The mouths of ziphiids can only be opened to a
limited extent (Fig. 4). For example, we measured
a maximum gape of 7 cm for the adult M. mirus
we examined. In the sample of 19 B. bairdii whales
(9.10 to 10.75 m total lengths), the jaws could be
opened from 17 to 34 cm with an average of 28
cm.

The oral cavity and esophagus of the adult male
M. mirus held 1.71 liters. This measure did not in-
clude any tongue retraction.

The fish placed into the mouth of the young, live
Z. cavirostris would remain motionless within the
mouth for several seconds. The fish was then drawn
into the mouth almost instantly, without any no-
ticeable movement of the whale’s head from a dor-
solateral view. The elapsed time from when the fish

Heyning and Mead: Suction Feeding in Ziphiids ■ 3

Figure 2. Left lateral diagram of a Mesoplodon stejnegeri calf (LACM 84299) showing the orientation of selected gular
muscles (e = esophagus, gg = genioglossus, gh = geniohyoideus, hg = byoglossus, ih = interhyoideus, sg = styloglossus,
sh = sternohyoideus, t = tongue).

began to move until it was completely within the
oral cavity ranged between one and two frames on
the video recording (less than 0.08 seconds).

ANATOMICAL DESCRIPTIONS OF
BEAKED WHALE GULAR REGION

The sphincter colli is external to the mylohyoideus
and is clearly separated from this underlying muscle
posteriorly as these two muscles are divided by the
digastricus. In the region of the throat grooves, the
sphincter colli has transversely oriented fibers that
originate from connective tissue just lateral to the
throat grooves.

The mylohyoideus is found entirely anterior to
the hyoid bones. The mylohyoideus originates pos-
teriorly from the pterygoid hamuli and/or from
along a membrane extending from the hamuli that
form the outer surface of the pterygoid sinus. An-
teriorly, the mylohyoideus adheres to the genio-
hyoideus. The mylohyoideus is about 1 cm thick.

The geniohyoideus originates as a thin, flat ten-
don on the rostral surface of basihyoid just ventral
to the origin of the hyoglossus. The origin of this
muscle in ziphiids extends along the medial one
third of the fused basihyoid/thyrohyoid bones,
whereas in the common dolphin the geniohyoideus
originates only from the anterior horn of the basi-
hyoid. The fleshy body of this muscle has a wide
insertion on the mandible posterior to the symphy-
sis beginning approximately one third the ramus

length and continues caudally to just posterior to
the gape.

The genioglossus inserts on the posterior-lateral
surface of the tongue and extends along the lateral
surface of the tongue. The insertion is primarily
along the ventral median raphe of the tongue, but
farther posteriorly some fibers extend dorsally over
the esophagus and were first thought by us to be a
remnant of the palatoglossus.

As in delphinids (Lawrence and Schevill, 1965),
the styloglossus originates from the ventrolateral
surface of the stylohyoid (Fig. 3), not from the sty-
loid process. It extends anteriorly and medially as
a rather cylindrical muscle until it fuses with the
lateral head of the hyoglossus.

The hyoglossus is a thin, broad muscle which
originates from the anterior surface of the medial
one third of the thyrohyoid. Its origin is deep to the
digastricus and superficial to the interhyoideus
muscle. The lateral head of the hyoglossus extended
about 5 cm anteriorly before fusing with the sty-
loglossus in the neonate M. stejnegeri we dissected.

The intrinsic muscles of the tongue were quite
complex, as is typical for mammals. We did not
dissect these muscles in detail.

The interhyoideus is a thick, short muscle that is
similar to that found in delphinids (Lawrence and
Schevill, 1965), except for being relatively larger in
ziphiids correlated with the larger hyoid bones.

The sternohyoideus is a massive muscle in most
cetaceans. In an adult specimen of M. mirus the

4 ■ Contributions in Science, Number 464

Heyning and Mead: Suction Feeding in Ziphiids

Figure 3. Dorsal diagram of lower jaw and gular muscles of a Mesoplodon mirus. The left figure depicts the muscles at
rest and figure on right shows tongue retracted by contraction of the styloglossus and hyoglossus muscles (e = esophagus,
hg = hyoglossus, sg = styloglossus, t = tongue).

midmuscle cross-sectional diameters were 16 by 6
cm. The sternohyoideus originates from the ante-
rior margin of the sternum. Near the origin, it is
difficult to separate into right and left sides. At ap-
proximately midlength, this muscles separates into
a bilateral muscle pair. Along the medial surface,
about one third the distance to the sternum, there
is a connective tissue band that extends medially
from the sternum and forms the anterior surface of
the thoracic cavity.

The digastricus originates as a broad, thin tendon
from the anterior margin of the fused basihyoid/
thyrohyoid. The belly of this muscle is fleshy near
the origin and then becomes more tendinous as it
inserts onto the intramandibular fat pad and ven-
tral posterior half of the mandible.

We were unable to locate a palatoglossus muscle
in any of the specimens of Mesoplodon or Ziphius
dissected.

The fused thyrohyoid and basihyoid bones of
beaked whales (Ziphiidae) are relatively wider and
thicker than those of dolphins (Delphinidae) (Table
2, Fig. 5). The thyrohyoid/basihyoid complex is the
attachment for the sternohyoideus as well as most
of the extrinsic gular musculature.

Contributions in Science, Number 464

DISCUSSION

Many aquatic vertebrates capture prey by suction.
Examples include sharks (Tanaka, 1973), numer-
ous groups of actinopterygians, some salamanders
and turtles (Lauder, 1985; Lauder and Shaffer,
1986) and walruses (Fay, 1981; Kastelein et al.,
1994). Suction is an effective technique of prey ac-
quisition for aquatic predators because of the in-
compressibility of water.

Several authors have previously alluded to the
possibility that several species of odontocetes may
use suction to capture food. For example, suction
feeding has been suggested for sperm whales (Phy-
seteridae) (Caldwell et al., 1966; Berzin, 1971), be-
luga whales (Monodontidae) (Ray, 1966:671; Bro-
die, 1989), and for most extant odontocetes (Norris
and Mohl, 1983). Ray (1966:671) observed that
captive belugas ( Delphinapterus leucas (Pallas,
1776)) could suck a 50-cent coin off the bottom
from about 4 inches distance. Tomilin and Moro-
zov (1968) noted that captive harbor porpoises
(. Phocoena phocoena (Linnaeus, 1758)) could suck
a fish “with some force” from the hand of a trainer
at a distance of 3 to 5 cm; they also stated that the

Heyning and Mead: Suction Feeding in Ziphiids ■ 5

Figure 4. Maximum jaw opening for several species of delphinids (top) and ziphiids (bottom). Clockwise from upper
left: Tursiops truncatus, Orcinus orca, Delphinus delphis, Berardius bairdii, Mesoplodon mirus, and Zipbius cavirostris.

harbor porpoises could suck in dead, drifting fish
from at least 10 cm.

THROAT GROOVES

As noted previously, at least one pair of anteriorly
converging throat grooves is present in all species
of extant ziphiids. Throat grooves of varying num-
ber are also found on the sperm whale ( Physeter
catodon Linnaeus, 1758) (Clarke, 1956), on the
pygmy sperm whale ( Kogia breviceps Blainville,
1838), and on some specimens of the dwarf sperm
whale ( Kogia simus (Owen, 1866)) (Caldwell and
Caldwell, 1989; unpublished data, LACM and
USNM). Several authors (Beddard, 1900; Boschma,
1938; Clarke, 1956) have suggested that the throat
grooves of sperm whales and beaked whales func-
tion to allow the throat to expand when the ani-
mals swallowed large prey. The term “throat
grooves” is somewhat of a misnomer as these
grooves are situated ventral to the oral cavity and
anterior to the hyoid bones, not the esophageal re-
gion. Thus, it is unlikely that the distension of the
throat grooves could directly facilitate the passing
of food down the esophagus in these species.

In the ziphiids we examined, the throat grooves
allowed the floor of the oral cavity to be distended
ventrally, thus increasing the intraoral volume. The
throat grooves of ziphiids are always located pos-
terior to the mandibular symphysis, the region that
allows gular distention. We suspect that this may

also be the case for sperm whales, as this anatom-
ical region seems to be distended in photographs of
some living individuals (e.g., Cousteau and Diole,
1972:135).

The function of throat grooves is better under-
stood for rorqual whales (Balaenopteridae) in
which these grooves open like pleats allowing the
distension of the oral cavity by the filling of the
cavum ventrale (Pivorunas, 1977; Lambertsen,
1983; Orton and Brodie, 1987).

Gray whales ( Eschrichtius robustus (Lilljeborg,
1861)) possess from two to five throat grooves (An-
drews, 1914). Andrews (1914) suggested that these
grooves may function to allow the gular region to
expand during feeding. Observations of a captive
juvenile gray whale (Gigi) demonstrated that this
species can also generate suction as a method of
capturing food prey (Ray and Schevill, 1974). This
captive whale was able to suck squid off the bottom
in a swath about 30 to 50 cm wide while swimming
10 to 20 cm above the bottom. These authors also
noted that the throat grooves allowed for the dis-
tention of the floor of the mouth while the whale
was drawing in food. This distention is clearly vis-
ible in photographs of bottom-feeding wild gray
whales (e.g., Simon, 1989:23).

The method by which the gular region is distended
in gray whales, sperm whales, and beaked whales
probably differs significantly from that of rorqual
whales. In the former groups it is likely that active

6 ■ Contributions in Science, Number 464

Heyning and Mead: Suction Feeding in Ziphiids

Table 2. Hyoid bone measurement ratios for beaked whales and selected delphinids.

Species

Sex

Length

LACM

number

Hyoid

Thickness/

length

Width/

length

Ziphiidae

Berardius bairdii

9.50

86031

0.20

0.41

9.50

86032

0.22

0.40

10.10

86033

0.22

0.37

Ziphius cavirostris

5.37

88971

0.17

0.43

5.49

84111

0.15

0.44

Mesoplodon carlhubbsi

4.96

52437

0.17

0.43

Delphinidae

Orcinus orca

6.65

84249

0.08

0.32

Pseudorca crassidens

4.80

84047

0.12

0.33

3.78

84289

0.07

0.30

Tursiops truncatus

3.17

84269

0.10

0.35

3.04

84194

0.08

0.35

2.88

88918

0.08

0.25

Delphinus delphis

1.81

88926

0.08

0.34

muscular contraction increases the volume of the oral
cavity. In rorquals, the mouth is opened while the
whale is swimming and thus water pressure passively
fills the oral cavity (Orton and Brodie, 1987).

We postulate that after distention, the gular re-
gion returns to its resting position primarily as a
result of the elastic nature of most of the tissues,
especially the blubber and adjacent fascia. In dead
specimens, the throat grooves return to their closed
state after being manually distended. This type of
elastic recoil has been suggested to function simi-
larly in constricting the cavum ventrale of rorqual
whales (Lambertsen, 1983; Orton and Brodie,
1987). In addition, the muscle fibers of the sphinc-
ter colli profundus and the mylohyoideus are ori-
ented to assist in reducing the intraoral volume by
actively constricting the gular region.

TONGUE

The tongue of beaked whales can be retracted in a
piston-like manner, as previously suggested for oth-
er odontocetes. Because of their anatomical orien-
tation, the hyoglossus and styloglossus appear to
be the primary retractors of the tongue. The in-
creased width and thickness of the hyoid bones in
ziphiids provides a greater surface of attachment
for these relatively large tongue muscles. This in-
creased surface area of the basihyoid/thyrohyoid
bones in ziphiids also serves as the attachment of
the relatively massive sternohyoideus ( contra Rei-
denberg and Laitman, 1994). Only sperm whales
(both Physeter and Kogia) have relatively larger hy-
oids. The hyoid apparatus is an important func-
tional component for suction feeding in vertebrates
(Lauder, 1985). On a dissected head, when the
tongue is retracted manually, it inherently becomes
shorter longitudinally but thicker in cross-section.

Contributions in Science, Number 464

If this is not a postmortem artefact, this may allow
the tongue to form a tight seal against the walls
and roof of the mouth as the tongue is retracted,
thereby creating a more efficient piston. The inser-
tion of the genioglossus facilitates tongue retraction
because it inserts along the posterolateral aspect of
the tongue. By not attaching to the rostral aspect
of the tongue, the genioglossus does not impede
tongue retraction or result in the over-stretching of
the fibers of this muscle.

Other odontocetes may be able to retract their
tongues to varying degrees. Donaldson (1977:194)
stated that in Tursiops the tongue could be retract-
ed by the hyoglossus and styloglossus. He also not-
ed that the tongue became shorter longitudinally
and thicker in cross-section. He suggested that the
thickened, retracted tongue functioned at least par-
tially to occlude the esophagus, thus preventing wa-
ter from entering the esophagus during prey acqui-
sition. Sonntag (1922) wrote that the tongue of the
blunt-headed delphinid Orcella Gray, 1866, was
quite mobile, whereas the tongue of the extremely
long-beaked river dolphin Platanista Wigler, 1830,
exhibited very restricted movement. This observa-
tion corroborates the idea that blunt-headed del-
phinids may employ suction feeding whereas long-
snouted species with a very high tooth count do not.

All neonate and young juvenile ziphiids we ex-
amined possessed marginal papillae on their
tongues. These marginal papillae have been dem-
onstrated to regress in size with age in delphinids
and phocoenids (Donaldson, 1977; Kastelein and
Dubbeldam, 1990). These latter authors speculated
that these papillae function to assist the tongue in
forming a tighter seal while nursing, and later in
life may assist in holding prey against the palate
while letting water escape between the papillae.

Heyning and Mead: Suction Feeding in Ziphiids ■ 7

Figure 5. Dorsal views of the fused basihyoid and thyrohyoids of a beaked whale ( Ziphius cavirostris, LACM 84111),
top, and a presumed nonsuction-feeding dolphin ( Pseudorca crassidens, LACM 84047).

8 ■ Contributions in Science, Number 464

Heyning and Mead: Suction Feeding in Ziphiids

The floor of the mouth anterior and lateral to the
tongue is deeply pleated in beaked whales. Shallow
folds have been recorded in several taxa of odon-
tocetes (Sonntag, 1922; Arvy and Pilleri, 1972;
Donaldson, 1977). Arvy and Pilleri (1972) termed
these folds the “linguopharygeal grooves.” These
authors examined these folds for ducts and glands,
but did not comment on their function. We found
that these pleats unfurl during tongue retraction.

Anteriorly, the geniobyoideus consists of a broad,
flat tendon that may allow the tongue to slide over
it smoothly. In addition, there is a layer of loose
connective tissue between the tongue with its ex-
trinsic muscles and the more superficial geniobyoi-
deus. It is along this plane that the tongue moves
internally. This loose connective tissue layer is sim-
ilar in construction and function to that of the ca-
vum ventrale of rorqual whales (see Lambertsen,
1983). However, the cavum ventrale of rorquals al-
lows for significant lateral movement of the gular
floor as the tongue inverts during feeding.

In order to enhance gular distention, a ventrally
directed force is advantageous. The enlarged pter-
ygoid hamuli found on the skulls of ziphiids extend
ventrally quite far compared to those in other ce-
taceans (Fig. 6). As the tongue is drawn caudally,
it also must move ventrally along the pterygoid
hamuli. This may create the downward vector
needed to expand the throat grooves and thus con-
tribute to increasing the intraoral volume, thereby
decreasing the intraoral pressure.

GAPE

Ziphiids are unable to open their mouths very wide.
Most species possess a great deal of superficial tissue
around the corner of the mouth (Heyning, 1989a).
This is particularly true for adult males of the genus
Mesoplodon in which the teeth are situated more
posteriorly along the mandible (Mead et al., 1982).
This combination of a small gape and tissue around
the corner of the mouth results in a relatively small
and somewhat restricted oral opening.

A restricted aperture of the mouth is most efficient
for suction feeding, as all the water is drawn in
through a relatively small orifice (Alexander, 1967).
Belugas are able to purse their lips in order to form
a small, round, terminal opening to the mouth for
suction feeding (see Brodie, 1989: fig. 5). Tomilin
and Morozov (1968) observed that harbor porpoises
only opened their mouths about 0.5 to 1 cm when
engaging in suction feeding. Therefore, we propose
that the limited extent to which ziphiids can open
their mouths is an adaptation for suction feeding.

The ziphiid with the most restricted gape is the
strap-toothed whale (M. layardii). It has been ques-
tioned how adult male M. layardii feed as their erupt-
ed mandibular teeth curve above the snout, thereby
limiting the extent to which they can open their
mouths (e.g., Ellis, 1980). The measured gapes of an
adult female and an immature male strap-toothed
whale were markedly wider (6.5 cm) than the gapes

Contributions in Science, Number 464

of two adult males (3.2 and 4.0 cm), indicating that
the erupted teeth do in fact limit the maximum extent
to which the mouth can be opened in this species
(Sekiguchi et al., 1996). Leatherwood et al. (1983)
stated that the erupted teeth of adult males may func-
tion as “guide rails” funneling prey into the mouth.
Because females and immature males lack such guide
rails, it is more reasonable to assume that these teeth
function in intraspecific fighting (Heyning, 1984).
Both male and female M. layardii typically feed on
squid significantly smaller (less than 100 g and mantle
lengths less than 16 cm) than those fed upon by Zi-
pbius cavirostris in spite of the fact that both of these
ziphiids are approximately the same size (Sekiguchi et
al., 1996). The restricted gape of strap-toothed whales
may be an adaptation for sucking in small squid. By
specializing on small squid, M. layardii may avoid
competition for food with sympatric teuthophagous
odontocetes such as Z. cavirostris, Hyperoodon plan-
ifrons, and Pbyseter catodon.

REDUCED TOOTH COUNT

There is a trend for primarily squid eating (teutho-
phagous) species to exhibit a reduced tooth count
(Gaskin, 1982; Clarke, 1986). For those that retain
teeth, such as the sperm whale and beaked whales,
it has been postulated that these teeth serve more
for social interactions than for prey acquisition
(Heyning, 1984).

The supernumerary teeth typical of most odon-
tocetes may not be efficient for suction feeding.
This is because there is a significant region where
water could enter the oral cavity between the teeth,
but the sucking in of prey items would be hindered
by this dental barrier. The broad-headed delphinoid
species with reduced dentition, such as Grampus
Gray 1828, Globicepbala Lesson, 1828, and Del-
pbinapterus, which are teuthophagous, lack teeth
at the extreme terminus of the mouth where prey
would be sucked in. It has been previously noted
that most species of ziphiids lack erupted teeth. We
hypothesize that a reduction in dentition would be
selected for as an advantage in the evolution of suc-
tion feeding.

The palates of the ziphiids we examined were ru-
gose with small cornified papillae. We found a sim-
ilar condition on the palate of the delphinid Gram-
pus griseus, a teuthophagous species of delphinid
with a reduced tooth count. In Dali’s porpoise
( Pbocoenoides dalli ), there are small dermal ridges
interspersed between the extremely reduced teeth,
and the palate is coarsely wrinkled with broad ridg-
es (Miller, 1929). Gaskin (1982) commented that
such a rugose palate would facilitate the holding of
slippery squid. Such a palate would provide a non-
skid surface which does not offer projections (i.e.,
teeth) upon which the squid’s tentacles can grasp
and thus hinder the ingestion of the still live prey.

WATER FLOW

Vertebrates with gills have a unidirectional flow of
water into the mouth and out from the gills. Higher

Heyning and Mead: Suction Feeding in Ziphiids ■ 9

Figure 6. Right lateral diagrammatic view of the skulls and body outline of a beaked whale (A) Mesoplodon carlhubbsi,
and a dolphin (B) Tursiops truncatus. Note the enlarged pterygoid (PT) of the beaked whale compared to the small
pterygoid of the dolphin. Illustrations are not to scale.

vertebrates, including adult amphibians, lack gills
and thus during suction feeding water flow is bi-
directional, both in and out through the mouth. In
a study that compared larval and adult tiger sala-
manders, Lauder and Shaffer (1986) suggested that
unidirectional flow was more efficient than the bi-
directional flow. Species that employ a bidirectional
flow system must hold water while they secure the
prey prior to expressing the water out of the
mouth. In an unrelated study, Harrison et al.
(1970) noted that the forestomach of dolphins can

10 ■ Contributions in Science, Number 464

hold up to several liters of water. These authors
pointed out that there is no muscular sphincter be-
tween the esophagus and forestomach. They also
observed dolphins ejecting water from the mouth
in such a volume as to suggest that some of this
water was coming from the forestomach. It may be
possible that one function of the forestomach of
odontocetes may be to act as a reservoir for sea
water that has been sucked in. We suggest that it is
plausible that these odontocetes may hold water in
their forestomach while they secure the prey against

Heyning and Mead: Suction Feeding in Ziphiids

the rugose palate and then release the water past
the secured prey.

PIGMENTATION

Many of the beaked whale species we dissected had
white or lightly pigmented epidermis on the anterior
floor of the mouth. We hypothesize that this may be
an attractant for bioluminescent squid. Previously,
Gaskin (1967) hypothesized that the white lower
jaws and teeth of sperm whales may retain the bio-
luminescent bacteria from consumed squid and this
in turn may attract additional squid. Other odon-
tocetes (e.g., Pepenocephala Nishiwaki and Norris,
1966, Feresa Gray, 1870, and Steno Gray 1846) also
have strikingly delimited white lip patches. However,
the white on the tip of the snout of some adult male
beaked whales has been suggested to visually en-
hance the size of the erupted teeth (Heyning, 1984).
This particular visual cue may relate more to intra-
specific communication than to feeding.

CONCLUSIONS

The throat grooves of beaked whales provide for the
distension of the floor of the mouth. The styloglossus
and hyoglossus muscles are the principle tongue re-
tractors. This extensive movement of the tongue is
facilitated by a smooth, loose connective tissue layer
between the tongue and genioglossus muscles.
Tongue retraction and gular floor distention results
in the expansion of the oral cavity, thereby creating
low pressure within the oral cavity and allowing wa-
ter to be sucked into the mouth. Limited observa-
tions show that beaked whales are able to suck in
their prey. The ability to echolocate coupled with
suction feeding imparts a tremendous advantage to
odontocetes foraging in the aphotic zone.

Suction feeding could explain the observations of
intact squid found in the stomachs of sperm whales
and ziphiids. Previously these prey items, which
lack teeth marks, were used as evidence to support
the hypothesis that the prey had been stunned by
high intensity echolocation sounds (Berzin, 1971;
Norris and Mohl, 1983). As noted previously, the
primary function of the teeth of sperm whales and
beaked whales, as well as some delphinids, is relat-
ed to intraspecific aggression, not to prey acquisi-
tion and processing. Suction feeding by these ceta-
ceans would explain why squid found in the stom-
ach have no teeth marks.

ACKNOWLEDGMENTS

This work was supported in part by a grant to JEH from
the Taylor Life Sciences Fund of the Natural History Mu-
seum of Los Angeles County. A. Pabst, W. McLellen, A.
Watson, and L. Barnes reviewed the manuscript and sig-
nificantly improved the text. We thank Lori Gage of Af-
rica/Marine World USA and Dan Odell of Sea World Flor-
ida for data and access to live stranded beaked whales
under their care. R.L. Brownell, Jr. assisted with data col-
lection from Berardius bairdii. Kirsten Oleson’s discus-
sions with JGM helped clarify our approach to this pro-
ject. All illustrations were rendered by Elizabeth Dyni.

Contributions in Science, Number 464

LITERATURE CITED

Alexander, R. McN. 1967. Functional design in fishes.
London: Hutchinson and Co. Ltd., 160 pp.

Andrews, R.C. 1914. The California gray whale ( Rhachi -
anectes glaucus Cope): Its history, habits, external anat-
omy, osteology and relationship. Memoirs of the Amer-
ican Museum of Natural History, New Series, vol. 1,
part V Monographs of the Pacific Cetacea, 227-287.

Arvy, L., and G. Pilleri. 1972. Comparison of the tongues of
some odontocetes (Pontoporia, Neomeris, and Delphi-
nus). Investigations on Cetacea 4:191-200 + 6 pis.

Balcomb, K.C., III. 1989. Baird’s beaked whale Berardius
bairdii Stejneger, 1883; Arnoux’s beaked whale Ber-
ardius arnuxii Duvernoy, 1851. In Handbook of ma-
rine mammals, vol. 4, ed. S.H. Ridgway and R. Har-
rison, 261-288. London: Academic Press.

Beddard, F.E. 1900. A book of whales. London: John
Murray, xv + 320 pp.

Berzin, A. A. 1971. The sperm whale. Jerusalem: Israel
Program for Scientific Translation (English transla-
tion, 1972), v + 394 pp.

Boschma, H. 1938. On the teeth and some other partic-
ulars of the sperm whale (Physeter macrocephalus
L.). Temminckia 3:151-278.

Brodie, P.F. 1989. The white whale — Delphinapterus leu-
cas (Pallas, 1776). In Handbook of marine mam-
mals, vol. 4, ed. S.H. Ridgway and R. Harrison,
119-144. London: Academic Press.

Caldwell, D.K., and M.C. Caldwell. 1989. Pygmy sperm
whale Kogia breviceps (de Blainville, 1838): Dwarf
sperm whale Kogia simus Owen, 1866. In Hand-
book of marine mammals, vol. 4, ed. S.H. Ridgway
and R. Harrison, 235-260. London: Academic Press.

Caldwell, D.K., M.C. Caldwell, and D.W. Rice. 1966. Be-
havior of the sperm whale, Physeter catodon L. In
Whales, dolphins, and porpoises, ed. K.S. Norris,
677-717. Berkeley: University of California Press.

Clarke, M.C. 1986. Cephalopods in the diets of odonto-
cetes. In Research on dolphins, ed. M.M. Bryden and
R. Harrison, 281-321. Oxford: Clarendon Press.

Clarke, R. 1956. Sperm whales of the Azores. Discovery
Reports 28:237-298.

Cousteau, J-Y., and P. Diole. 1972. The whale: Mighty
monarch of the sea. Garden City, New York: Dou-
bleday and Co., 304 pp.

Donaldson, B.J. 1977. The tongue of the bottle-nosed dol-
phin ( Tursiops truncatus). In Functional anatomy of
marine mammals, vol. 3., ed. R. Harrison, 175-197.
London: Academic Press.

Ellis, R. 1980. The book of whales. New York: Alfred A.
Knopf, xvii + 202 pp.

Fay, F. 1981. Walrus — Odobenus rosmarus. In Handbook
of marine mammals, vol. 1, ed. S.H. Ridgway and
R. Harrison, 1-23. London: Academic Press.

Gaskin, D.E. 1967. Luminescence in a squid Moroteuthis
sp. (probably ingens Smith), and a possible feeding
mechanism in the sperm whale Physeter catodon L.
Tuatara 15(2):86-88.

Gaskin, D.E. 1982. The ecology of whales and dolphins.
London: Heineman Education Books Ltd., pagination.

Harrison, R.J., F.R. Johnson, and B.A. Young. 1970. The
oesophagus and stomach of dolphins ( Tursiops , Del-
phinus, Stenella). Journal of Zoology, London 160:
377-390.

Heyning, J.E. 1984. Functional morphology involved in
intraspecific fighting of the beaked whale, Mesoplo-
don carlhubbsi. Canadian Journal of Zoology 62:
1645-1654.

Heyning and Mead: Suction Feeding in Ziphiids 111

Heyning, J.E. 1989a. Comparative facial anatomy of
beaked whales (Ziphiidae) and a systematic revision
among the families of extant Odontoceti. Contri-
butions in Science 406:1-65.

Heyning, J.E. 1989b. Cuvier’s beaked whale, Ziphius ca-
virostris G. Cuvier, 1823. In Handbook of marine
mammals, vol. 4, ed. S.H. Ridgway and R. Harrison,
289-308. London: Academic Press.

Kastelein, R.A., and J.L. Dubbeldam. 1990. Marginal pa-
pillae on the tongue of the harbour porpoise ( Pho -
coena phocoena ), bottlenose dolphin (Tursiops trun-
catus) and Commerson’s dolphin (Cephalorhynchus
commersonii). Aquatic Mammals 15:158-170.

Kastelein, R.A., M. Muller, and A. Terlouw. 1994. Oral
suction of a Pacific walrus ( Odobenus rosmarus di-
vergens) in air and under water. Zeitschrift fur Sau-
getietrkunde 59:105-115.

Lambertsen, R.H. 1983. Internal mechanisms of rorqual
feeding. Journal of Mammalogy 64:76-88.

Lauder, G.V. 1985. Aquatic feeding in lower vertebrates.
In Functional vertebrate morphology, ed. M. Hilde-
brand, D.M. Bramble, K.F. Liem, and D.B. Wake,
210-229. Cambridge, Massachusetts: Harvard Uni-
versity Press.

Lauder, G.V., and H.B. Shaffer. 1986. Functional design
of feeding mechanism in lower vertebrates: Unidirec-
tional and bidirectional flow systems in the tiger sal-
amander. Zoological Journal of the Linnean Society
88:277-290.

Lawrence, B., and W.E. Schevill. 1965. Gular muscula-
ture in delphinids. Bulletin of the Museum of Com-
parative Zoology 133(1 ):l-65.

Leatherwood, S., R.R. Reeves, and L. Foster. 1983. The
Sierra Club handbook of whales and dolphins. San
Francisco: Sierra Club Books, xviii + 302 pp.

Leatherwood, S., R.R. Reeves, W.F. Perrin, and W.E.
Evans. 1988. Whales, dolphins, and porpoises of the
eastern North Pacific and adjacent Arctic waters: A
guide to their identification. New York: Dover Pub-
lications, Inc., ix + 245 pp.

Mead, J.G. 1989a. Shepherd’s beaked whale Tasmacetus
shepherdi Oliver, 1937. In Handbook of marine
mammals, vol. 4, ed. S.H. Ridgway and R. Harrison,
309-320. London: Academic Press.

Mead, J.G. 1989b. Bottlenose whales Hyperoodon am-
pullatus (Forster, 1770) and Hyperoodon planifrons
Flower, 1882. In Handbook of marine mammals,
vol. 4, ed. S.H. Ridgway and R. Harrison, 321-348.
London: Academic Press.

Mead, J.G. 1989c. Beaked whales of the genus Mesoplo-
don. In Handbook of marine mammals, vol. 4, ed.
S.H. Ridgway and R. Harrison, 349-430. London:
Academic Press.

Mead, J.G., W.A. Walker, and W.J. Houck. 1982. Biolog-
ical observations on Mesoplodon carlhubbsi (Ceta-

12 ■ Contributions in Science, Number 464

cea: Ziphiidae). Smithsonian Contributions to Zo-
ology 344:iii + 1-25.

Miller, G.S. 1929. The gums of the porpoise Phocoenoides
dalli (True). Proceedings of the U.S. National Mu-
seum 74:1-4 + pis. 1-4.

Norris, K.S., and B. Mohl. 1983. Can odontocetes debilitate
prey with sound? American Naturalist 122:85-104.

Omura, H., K. Fujino, and S. Kimura. 1955. Beaked
whale Berardius bairdi of Japan with notes on Zi-
phius cavirostris. Scientific Reports of the Whales
Research Institute 10:89-132.

Orton, L.S., and P.F. Brodie. 1987. Engulfing mechanics
of fin whales. Canadian Journal of Zoology 65:
2898-2907.

Pivorunas, A. 1977. The fibrocartilage skeleton and relat-
ed structures of the ventral pouch of balaenopterid
whales. Journal of Morphology 151:299-314.

Ray, G.C. 1966. Comment in Round Table: Practical
Problems. In Whales, dolphins, and porpoises, ed.
K.S. Norris, 649-673. Berkeley: University of Cali-
fornia Press.

Ray, G.C., and W.E. Schevill. 1974. Feeding of a captive
gray whale, Eschrichtius robustus. Marine Fisheries
Review 36(4):31-38.

Reidenberg, J.S., and J.T. Laitman. 1994. Anatomy of the
hyoid apparatus in Odontoceti (toothed whales):
Specializations of their skeleton and musculature
compared with those of terrestrial mammals. The
Anatomical Record 240:598-624.

Reyes, J., J.G. Mead, and K. van Waerebeek. 1991. A new
species of beaked whale Mesoplodon peruvianus sp.
n. (Cetacea: Ziphiidae) from Peru. Marine Mammal
Science 7:1-24.

Sekiguchi, K, N.T.W. Klages, and P.B. Best. 1996. The
diet of strap-toothed whales (Mesoplodon layardii).
Journal of Zoology 239:453-463.

Simon, S. 1989. Whales. Trophy (Harper Collins Chil-
dren’s Division) New York, v. + 40 pp. (unpagin-
ated).

Sonntag, C.F. 1922. The comparative anatomy of the
tongues of the Mammalia, VII Cetacea, Sirenia and
Ungulata. Proceedings of the Zoological Society of
London 1922: 639-657.

Tanaka, S.K. 1973. Suction feeding by the nurse shark.
Copeia 1973:606-608.

Tomilin, A.G., and D.A. Morozov. 1968. Sucking in of
food — A previously unknown method of Phocoena
phocoena feeding. Trudy Vsesoyuznogo Sel’skokhoz-
yaystvennogo Instituta Zaochnogo Obrazovaniya
31:201-202 (in Russian).

Werth, A.J. 1989. Suction feeding in odontocetes: Water
flow and head shape. American Zoologist 29(4):92A.

Werth, A.J. 1991. Suction feeding in pilot whales: Kine-
matic and anatomical evidence. American Zoologist
31(5):17A.

Received 21 January 1996; accepted 16 August 1996.

Heyning and Mead: Suction Feeding in Ziphiids

Natural History Museum
of Los Angeles County
900 Exposition Boulevard
Los Angeles, California 90007

HECKMAN

BINDERY INC.

AUG 97

Bound -To -Pleas!5 N; MANCHESTER.

INDIANA 46962

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Contributions in science"

See other formats