I

i

3(o

THE CARNEGIE MUSEUM OF NATURAL HISTORY

4400 FORBES AVENUE # PTTTSBURGiU PENNSYLVANIA 15213

VOLUME 69 r 18 FEBRUARY 2000 ' NUMBER 1

CONTENTS

ARTICLES

New record of Ctenodus (Osteichthyes: Dipnoi) from the Carboniferous of

Montana A. Kemp and R. Lund 1

Snake fauna associated with the “Earliest Recent” mammalian fauna in north-
eastern North America J. Alan Holman 5

New Atokan productoid brachiopods from the Upper Carboniferous Ladrones
Limestone of southeastern Alaska, with a preliminary note on the
phytogeny and classification of the Tribe Retariini

Stanislav S. Lazarev and John L. Carter 1 1

Eocene decapod crustaceans from Pulali Point, Washington

Carrie E. Schweitzer, Rodney M. Feldmann,

Annette B. Tucker, and Ross E. Berglund 23

Editors, ANNALS, BULLETIN and SPECIAL PUBLICATIONS:

D. S. Berman, Ph.D.

D. R. Watters, Ph.D.

J. R. Wible, Ph.D.

M. A. Schmidt, ELS, Managing Editor

Manuscripts, subscriptions, orders for individual numbers, and changes of address ,
should be sent to:

Office of Scientific Publications
Carnegie Museum of Natural History
4400 Forbes Avenue
Pittsburgh, PA 15213-4080
Phone (412) 622-3287
Fax (412) 622-8837
E-mail: scipubs@clpgh.org

ANNALS OF CARNEGIE MUSEUM is published quarterly by The Carnegie Museum of Natural
History, 4400 Forbes Avenue, Pittsburgh, Pennsylvania 15213-4080, by the authority of the Board of
Trustees of Carnegie Institute.

© 2000 Carnegie Institute.

@ This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).

ANNALS OF CARNEGIE MUSEUM

VoL. 69, Number 1, Pp. 1-4

18 February 2000

NEW RECORD OF CTENODUS (OSTEICHTHYES: DIPNOI)

FROM THE CARBONIFEROUS OF MONTANA

A. Kemp^

R. Lund2

Research Associate, Section of Vertebrate Paleontology

Abstract

A partial tooth plate of Ctenodus cf. C. interruptus from the uppermost Heath Formation (Chester-
ian, equivalent to Namurian E 2b), Montana, is described. This record extends the occurrence of
Ctenodus to the western part of North America. Ctenodus species are common in Europe and have
also been recorded from the eastern side of North America and from Australia.

Key Words: fossil lungfish, Ctenodus, Carboniferous, Montana

Introduction

Few Paleozoic dipnoans are more widespread in occurrence than the Carbon-
iferous genus Ctenodus (Agassiz, 1838). Once considered to be characteristic of
European deposits (Woodward, 1891; Romer and Smith, 1934), recent records of
Ctenodus and related genera have suggested that in fact the genus has a world-
wide, if sporadic, distribution (Thomson, 1965; Baird, 1978; Long and Campbell,
1985).

A number of species of Ctenodus from Europe were described late in the 19th
century (Agassiz, 1838; Barkas, 1869; Fritsch, 1885-1889). Although Woodward
(1891) applied some judicious synonymies to this plethora of species, recent de-
scriptions or redescriptions of Ctenodus and related taxa (Thomson, 1965; Baird,
1978; Long and Campbell, 1985) have increased the number of valid taxa refer-
rable to the group. Revision of the genus is, however, beyond the scope of this
report of the incidence of a tooth plate of Ctenodus from Carboniferous deposits
in Montana. This specimen is conspecific with the tooth plate described by Baird
(1978) as Ctenodus interruptus, from Carboniferous deposits at Grand Etang,
Nova Scotia. Neither specimen conforms completely to the characters of Ctenodus
interruptus (Barkas, 1869) because both are broader than European specimens of
the species, but reassignment of the material is deferred pending a complete re-
view of the family Ctenodontidae.

Abbreviations are as follows: CMNH, Carnegie Museum of Natural History,
Pittsburgh, Pennsylvania; YPM (PU), Princeton University Collection at Peabody
Museum of Natural History, Yale University, New Haven, Connecticut.

^ Centre for Microscopy and Microanalysis, University of Queensland, Saint Lucia, Queensland, Aus-
tralia, 4072.

^Department of Environmental Sciences, Adelphi University, Garden City, New York 11530.
Submitted 21 October 1997.

1

2

Annals of Carnegie Museum

VOL. 69

Systematic Paleontology

Subclass Dipnoi Muller 1845
Family Ctenodontidae Woodward 1891
Genus Ctenodus cf. C. interruptus Barkas 1869
(Fig. 1)

Schultze (1992) gives a complete synonymy list.

Geological Setting. — Surenuff Sand Channel, uppermost Heath Formation
(Chesterian, equivalent to Namurian E 2b), Big Snowy Group of Forest Grove,
Fergus County, Montana. There is some doubt regarding the precise biostrati-
graphic determination of the horizon from which the specimen was obtained,
although a Chesterian age is the most plausible. The Surenuff beds lie above and
at the western edge of the Bear Gulch Limestone beds, within the Heath For-
mation, according to Homer (1985). The Heath Formation is entirely contained
within the Namurian E 2b, and is bounded above by an unconformity believed
to mark the Mississippian-Pennsylvanian boundary in this area. The channel at
this locality could also be Morrowan (Lower Pennsylvanian) or Lower Jurassic.
The latter age is impossible because a cochliodont tooth plate is found in the
same deposit, and the former is also unlikely because the lower Morrowan de-
posits at this locality are terrestrial. With the Ctenodus specimen are found a
cochliodont tooth plate that does not occur in the Bear Gulch Limestone and has
not been found elsewhere in the Heath Formation, as well as linguloid brachio-
pods, unidentified large ganoid scales, and other fish bones. It is found in a strong-
ly crossbedded coarse oolitic sand channel deposit. This lentil is traceable south-
ward over approximately 2 km distance along the outcrop into laminar Surenuff
silty limestone beds (Horner, 1985). These beds have a sparse but typical Ches-
terian articulate brachiopod fauna and contain scales of the crossopterygian Strep-
sodus sp. and skeletal elements of Acanthodes that typify the fish fauna of the
western edge of the Bear Gulch beds. The sedimentology and atypical faunal
association suggest that the Ctenodus specimen was found in a delta or estuarine
deposit on the edge of a bay or an estuary. The Ctenodus specimen, therefore,
could have come from either a fresh-water fish or from an animal that lived in
brackish water.

Description. — The specimen, CMNH 62780A + B, part and counterpart, consists of the posterior
three-quarters of a broad upper tooth plate (Fig. 1). Ridges bearing short conical cusps radiate slightly
from the curved mediolingual junction. The labial extremities of the ridges are missing. All cusps
present are heavily abraded, although still distinct. The inter-ridge furrows carry grooves formed by
occlusion with the cusps of the matching lower tooth plate. This indicates that the jaw movements
involved subterminal rotational grinding, as found in more derived dipnoans (Kemp, 1991).

Ten ridges are present in CMNH 62780 A. The first ridge and most of the second are missing but
the remaining eight are complete except for the labial margin. Cusps in the second ridge show traces
of doubling, and the eighth ridge is bifid (Fig. 1). Ridges seven and nine are consequently displaced.
This anomaly, bifid instead of single ridges, is common in Ctenodus interruptus, and also found in
other species of Ctenodus. Apart from the displaced portions, the ridges are subparallel.

A similar specimen, YPM (PU) 21741, described by Baird in 1978, comes from the Pomquet For-
mation, Mabou Group, latest Mississippian (Namurian A) at Grand Etang, Inverness County, Cape
Breton Island, Nova Scotia. This is a complete upper left tooth plate, almost as broad as it is long,
with 1 1 ridges and a curved mediolingual margin. The specimen has undivided ridges, each with a
single row of cusps. The ridges radiate from a point situated mediolingually, and have no trace of the
parallel arrangement found in other specimens of this genus.

Both CMNH 62780 A and YPM (PU) 21741 have an unusual double curvature of the occlusal sur-
face. Most of the occlusal surface is concave, but the concavity rises to an irregular crest close to the
mediolingual aspect of the tooth plate and then shelves downwards towards the mediolingual junction.

2000

Kemp and Lund — Carboniferous Ctenodus from Montana

3

Fig. 1. — Tooth plate of Ctenodus cf. C. interruptus, CMNH 62780A, from the uppermost Heath For-
mation (Chesterian, equivalent to Namurian E 2b), Montana. Scale bar = 2 cm.

Remarks. — The appearance of the tooth plate from the Surenuff Sand Channel
is consistent with that of an early species of Ctenodus, similar to Ctenodus inter-
ruptus. It differs from this species in the greater breadth of the tooth plate. The
Grand Etang specimen is also unusually broad, but otherwise similar to Ctenodus
interruptus. Unfortunately, most of the preserved characters of these specimens
are heavily influenced by growth and wear of the structure in life, and have no
taxonomic validity (Kemp, 1993, 1991a, \991b).

Species of Ctenodus are not confined to the Northern Hemisphere. In addition
to the extensive records of the genus from northern England and Scotland (Schultze,
1992), the genus, or genera within the same family, are represented in eastern
Europe (Fritsch, 1885-1889) as well as both sides of the North American continent,
and in Australia (Woodward, 1906; Long and Campbell 1985). Although the
description by Long and Campbell (1985) includes a skull roof, originally de-
scribed by Woodward (1906), that is consistent with other specimens of the genus
Ctenodus (Watson and Gill, 1923), the presence of dipnoan tooth plates in the
same locality that are not classically ctenodontid in form led Long and Campbell
(1985) to redescribe the material in a related genus, Delatitia. The recent find of
a perfect upper tooth plate with typical Ctenodus characteristics from the mid-
Visean Ducabrook Formation at Middle Paddock, Drummond Basin, east-central
Queensland, indicates that this genus is truly a part of the Australian Carbonif-
erous fauna. This deposit is fresh-water, and includes early tetrapods (Thulbom
et al., 1996).

Most records of Ctenodus belong in the Lower Carboniferous, but they are
widely distributed geographically, and seem to occur in fresh-water as well as in
brackish or marine deposits.

Acknowledgments

Dr. D. Baird, Carnegie Museum of Natural History, Pittsburgh, Pennsylvania, provided access to a
cast of YPM (PU) 21741 and assisted with discussions on the occurrence of dipnoans in the eastern
United States. Comparative material was borrowed from Dr. S. Turner, Queensland Museum, Brisbane,
Australia; the Hancock Museum in Newcastle, England; and the National Museum of Scotland, Ed-
inburgh, Scotland. Assistance from all sources is acknowledged with many thanks.

4

Annals of Carnegie Museum

VOL. 69

Literature Cited

Agassiz, L. 1838. Recherches sur les Poissons fossiles. Volume III. Pp. 73-140. Petitpierre; Neuchatel
et Soleure, France.

Baird, D. 1978. Studies on Carboniferous freshwater fishes. American Museum Novitates, 2641:1-22.

Barkas, T. 1869. Ctenodus interruptus. Scientific Opinion, 2:113.

Fritsch, a. 1885-1889. Fauna der Gaskohle und der Kalksteine der Permformation Bbhmens. 2.
Stegocephali (Schluss) — Dipnoi, Selachii (Anfang): 1-1 14. Selbstverlag, Prague, Czechslovakia.

Horner, J. R. 1985. The stratigraphic position of the Bear Gulch Limestone (Lower Carboniferous)
of central Montana. Compte Rendu, Neuvieme Congres International de Stratigraphie et de Geo-
logie du Carbonifere, 5:427-436.

Kemp, A. 1991. Australian Mesozoic and Cainozoic lungfish. Pp. 465-498, in Vertebrate Palaeon-
tology of Australasia (P. Vickers-Rich, J. M. Monaghan, R. F. Baird, and T. Rich, eds.). Pioneer
Design Studio, Melbourne, Australia.

. 1993. Ceratodus diutinus, a new fossil ceratodont from Cretaceous and Tertiary deposits in

Australia. Journal of Paleontology, 67:883-886.

. 1997«. Four species of Metaceratodus (Osteichthyes: Dipnoi, Family Ceratodontidae) from

Australian Mesozoic and Cenozoic deposits. Journal of Vertebrate Paleontology, 17:26-33.

. 1997/?. A revision of Australian Mesozoic and Cenozoic lungfish of the family Neocerato-

dontidae (Osteichthyes: Dipnoi) with a description of four new species. Journal of Paleontology,
71:713-733.

Long, J., and K. S. W. Campbell. 1985. A new lungfish from the Lower Carboniferous of Victoria,
Australia. Proceedings of the Royal Society of Victoria, 97:87-93.

Muller, J. 1845. Uber den Bau und die Grenzen der Ganoiden und fiber das natfirliche System der
Fische. Abhandlungen der Akademie der Wissenschaften zu Berlin, 1844:117-216.

Romer, a. S., and H. Smith. 1934. American Carboniferous dipnoans. Journal of Geology, 42:700-
719.

ScHULTZE, H.-P. 1992. Fossilium Catalogus 1: Animalia, pars 131, Dipnoi (F. Westphal, ed.). Kugler
Publications, Amsterdam, The Netherlands.

Thomson, K. S. 1965. On the relationships of certain Carboniferous Dipnoi; with a description of
four new forms. Proceedings of the Royal Society of Edinburgh, 69B:22 1-245.

Thulborn, T, a. Warren, S. Turner, and T. Hamley. 1996. Early Carboniferous tetrapods in
Australia. Nature, 381:777-780.

Watson, D. M. S., and E. L. Gill. 1923. The structure of certain Palaeozoic Dipnoi. Journal of the
Linnean Society, 35:163-216.

Woodward, A. S. 1891. Catalogue of the Eossil Pishes in the British Museum (Natural History).
Part II. Containing the Elasmobranchii (Acanthodii), Holocephali, Ichthyodorulites, Ostracodermi,
Dipnoi and Teleostomi. British Museum of Natural History, London, Great Britain.

. 1906. On a Carboniferous fish fauna from the Mansfield District, Victoria. Memoirs of the

National Museum, Melbourne, 1:1-32.

ANNALS OF CARNEGIE MUSEUM

VoL. 69, Number 1, Pp. 5-9

1 8 February 2000

SNAKE FAUNA ASSOCIATED WITH THE ‘‘EARLIEST RECENT’’
MAMMALIAN FAUNA IN NORTHEASTERN NORTH AMERICA

J. Alan Holman ‘

Abstract

The Hosterman’s Pit Local Fauna, Centre County, Pennsylvania, produced what was considered to
be the earliest known date for a “Recent” mammalian fauna in northeastern North America (Guilday,
1967). The few snake bones from the site represent five species (Diadophis punctatus. Coluber con-
strictor, Elaphe obsoleta, Thamnophis sirtalis, and Crotalus horridus), all of which presently occur
in Centre County. The snakes could have lived in a woodland situation with rocky outcrops and
ledges.

The Hosterman’s Pit Local Fauna

This local fauna accumulated in a limestone cave in Centre County, PennsyL
vania, latitude 40°53'34" N, longitude 77°26'22" W at an elevation of 378 m.
Approximately 150 m southwest of the 22.5~m”deep pit opening, a deposit of
mixed talus, vertebrate bones, and charcoal was found beneath dome pits that
were sealed and no longer open to the surface. The bones are presumed to be the
remains of animals that fell down these former sinkhole openings (Guilday, 1967).

Charcoal fragments, presumably derived from forest fires near the cave, were
radiocarbon dated at 7290 bc ± 1000. The modem nature of the Hosterman’s Pit
mammalian local fauna is reflected by the presence of temperate mammals such
as the southern flying squirrel (Glaucomys volans), southern bog lemming {Syn~
aptomys cooperi), pine mouse (Microtus pinetorum), cottontail rabbit (Sylvilagus
floridanus), and white-tailed deer {Odocoileus virginianus) and the lack of any
boreal species such as the northern flying squirrel {Glaucomys sabrinus), northern
bog lemming {Synaptomys borealis), collared lemming {Dicrostonyx hudsonius),
yellow-cheeked vole {Microtus xanthognathus), and caribou {Rangifer tarandus)
that occurred in the area in early posUWisconsinan times (Guilday, 1967). A
mammalian fauna with boreal species was found in the New Paris 4 cave deposit
105 km southwest of Hosterman’s Pit. The New Paris site had a radiocarbon date
of 9300 BC ± 1000 and Guilday (1967) suggested that the change to a “Recent”
mammalian fauna in the area took place between about 9300 to 7290 bc, a time
span of about 2,000 years.

The snake remains were collected from the Hosterman’s Pit local fauna by
Allen D. McCrady, Research Associate, Carnegie Museum of Natural History, on
18 September 1961, and subsequently donated to the Section of Paleontology and
Geology of the State Museum of Pennsylvania, Harrisburg. Following is an an-
notated taxonomic list of those snake remains. Numbers are those of the Vertebrate
Paleontological Collections of the State Museum of Pennsylvania (SMP VP-).

' Michigan State University Museum, East Lansing, Michigan 48824-1045.
Submitted 17 December 1997.

5

6

Annals of Carnegie Museum

VOL. 69

Annotated List

Family Colubridae
Subfamily Xenodontinae
Diadophis punctatus (Linnaeus, 1766)

Ring-necked Snake

Material. — A single trunk vertebra SMP VP- 1047.

Remarks. — The trunk vertebrae of the xenodontine genera Diadophis and Car-
phophis (worm snakes) are very similar although these taxa are strikingly different
in their external appearance and other aspects of their morphology. The trunk
vertebrae of both genera have the platyspondylous (very depressed neural arch)
condition, with very low neural spines and wide, indistinct hemal keels.

Differences between Diadophis punctatus and Carphophis amoenus are very
subtle and hinge on the fact that Diadophis punctatus has a somewhat longer
vertebral form and more laterally truncated postzygapophyseal buttresses than in
Carphophis amoenus. The Hosterman’s Pit vertebra was identified as Diadophis
punctatus on the basis of its relatively elongate vertebral form, laterally truncated
postzygapophyseal buttresses, and relatively large size. The fossil specimen was
larger than a trunk vertebra (see Holman, 1995:94-98) from the middle of the
postcervical- precaudal region of a Diadophis punctatus specimen with a total
length of 31 cm. The longest Carphophis amoenus recorded is 33.7 cm (Conant
and Collins, 1991) and most specimens are much smaller.

Presently, the Northern Ringneck Snake, Diadophis punctatus edwardsii, occurs
in Centre County, Pennsylvania (McCoy, 1982:map 64). This taxon is a secretive
woodland species that presently requires logs, stumps, fallen bark, rocks, or hu-
man debris under which to hide. Earthworms, salamanders, and small insects are
the most frequent food in eastern Diadophis punctatus (Ernst and Barbour, 1989).

Subfamily Colubrinae
Coluber constrictor Linnaeus, 1758
Racer

Material. — Two trunk vertebrae SMP VP- 1048.

Remarks. — Auffenberg (1963) discussed the identification of fossil Coluber
constrictor on the basis of individual trunk vertebrae. The two above specimens
represent small racers. Presently the Northern Black Racer, Coluber constrictor
constrictor, occurs in Centre County, Pennsylvania (McCoy, 1982:map 66).

Eastern Coluber constrictor prefers woodlands and woodland edges where wa-
ter is available and sometimes hibernates with Crotalus horridus (Ernst and Bar-
bour, 1989). The racer has been found hibernating in a cave (Sexton and Hunt,
1980). In eastern North America this species eats a very wide variety of food
including insects, salamanders, frogs, lizards, snakes, small mammals, and bird
nestlings and eggs (Ernst and Barbour, 1989).

Elaphe obsoleta (Say, 1823)

Rat Snake

Material. — Eour trunk vertebrae SMP VP-1049.

Remarks. — Auffenberg (1963) discussed the identification of fossil Elaphe ob-
soleta on the basis of individual trunk vertebrae. Presently the Black Rat Snake,
Elaphe obsoleta obsoleta, occurs in Centre County, Pennsylvania (McCoy, 1982:

2000

Holman — Snakes with “Earliest Recent” Mammals

7

3mm

Fig. 1. — Right pterygoid in lateral view of Crotalus horridus (SMPVP-1052) from Hosterman’s Pit
Local Fauna, Centre County, Pennsylvania.

map 69). The natural habitat of Black Rat Snakes consists of a variety of wood-
land situations. They often hibernate with Coluber constrictor and Crotalus hor~
ridus (Ernst and Barbour, 1989). Elaphe obsoleta that overwinter in caves are
active and may change their position several times in the winter (Sexton and
Hunt, 1980). Adults eat a wide variety of small endothermic vertebrates as well
as frogs.

Subfamily Natricinae
Thamnophis sirtalis Linnaeus, 1758
Conunon Garter Snake

Material. — Four trunk vertebrae SMP VP- 1050.

Remarks. — Holman (1984) discussed the identification of fossil Thamnophis
sirtalis on the basis of individual trunk vertebrae. Presently the Eastern Garter
Snake, Thamnophis sirtalis sirtalis, occurs in Centre County, Pennsylvania (Mc-
Coy, 1982:map 61). Eastern Garter Snakes occupy many kinds of habitats and
are generalistic feeders, eating earthworms, small fishes, small snakes, salaman-
ders, frogs, toads, small fishes, mice, and even nestling birds. They sometimes
hibernate with Diadophis punctatus (Ernst and Barbour, 1989).

Indeterminate Colubidae

Material. — One broken vertebra SMP VP- 1051.

Remarks. — One rather elongate vertebra belongs to a colubrid, but it is not
complete enough to assign to subfamily or genus.

Family Viperidae
Subfamily Crotalinae
Crotalus horridus Linnaeus, 1758
Timber Rattlesnake

Material. — One right pterygoid (Fig. 1) and eight trunk vertebrae SMPVP-
1052.

Remarks. — Holman (1967) discussed the identification of Crotalus horridus on
the basis of individual vertebrae. The pterygoid of Crotalus horridus (Fig. 1) is
rather similar to that of Agkistrodon contortrix, but differs in that a prominent,
dorsally projecting, notched, ectopterygoidal process is absent in Crotalus hor-
ridus and present in Agkistrodon contortrix. This process is absent in the fossil
(Fig. 1). Presently Crotalus horridus occurs in Centre County, Pennsylvania (Me-

8

Annals of Carnegie Museum

VOL. 69

Coy, 1982:map 73). Timber rattlesnakes are commonly found in wooded areas
with rocky outcrops and ledges. Congregations of Crotalus horridus near rock
outcrops and hibernating dens in spring and autumn were historically common in
upland, wooded regions in Pennsylvania and Hartwig (1966) saw as many as 17
at one time near a 6 m rock. Adult timber rattlesnakes feed mainly on small,
endothermic vertebrates (Ernst and Barbour, 1989).

Indeterminate Viperidae

Material. — Seventeen vertebrae SMP VP- 1053.

Remarks. — These vertebrae are either fragmentary or come from an undiagnos-
tic portion of the vertebral column and I am unable to identify them to the generic
level.

Indeterminate Snake

Material. — Eleven vertebrae SMP VP- 1054.

Remarks. — These are fragmentary vertebrae and caudal vertebrae that I am
unable to identify to the familial level.

Comments

All of the Hosterman’s Pit snake species are extant and presently occur in
Centre County, Pennsylvania. Like the mammalian fauna of the site which was
reported as the earliest “Recent” mammalian fauna in the northeastern United
States (Guilday, 1967), the snake fauna has a completely modem aspect.

Usually, Appalachian Quaternary cave faunas (Holman, 1995) yield hundreds
of snake vertebrae representing relatively few species. Thus, one of the most
interesting aspects of the Hosterman’s Pit snake fauna is the relatively large num-
ber of species represented by a very small number of snake bones. Only 48
vertebrae and one pterygoid represent five species in two families and three sub-
families.

The snakes represent an assemblage that could have occurred in woodland or
woodland-edge situations with rocky outcrops and ledges. Presently, the three
large snakes {Coluber constrictor, Elaphe obsoleta, and Crotalus horridus) often
occur in the same hibernaculum and the same is tme of the two small snakes
Diadophis punctatus and Thamnophis sirtalis (Ernst and Barbour, 1989). Both
Coluber constrictor and Elaphe obsoleta have been recorded hibernating in caves
(Sexton and Hunt, 1980). Thus, it could be suggested that at least some of the
Hosterman’s Pit fossil snakes were hibernating in or near the cave. With the
exception of Elaphe obsoleta, which can climb in an out of sinkhole caves (per-
sonal observations), it is possible that, over the years, other species were trapped
by falling into the cave.

Acknowledgments

I wish to thank Robert Sullivan of the State Museum of Pennsylvania for allowing me to study the
Hosterman’s Pit local fauna snakes. Two anonymous reviewers made helpful suggestions.

Literature Cited

Auffenberg, W. 1963. The fossil snakes of Florida. Tulane Studies in Zoology, 10:131-216.
CoNANT, R., AND J. T. CoLLiNS. 1991. A Field Guide to Reptiles and Amphibians, Eastern and Central
North America. Houghton Mifflin Company, Boston, Massachusetts.

2000

Holman — Snakes with “Earliest Recent” Mammals

9

Ernst, C. H., and R. W. Barbour. 1989. Snakes of Eastern North America. George Mason University
Press, Fairfax, Virginia.

Guilday, J. E. 1967. The climatic significance of Hosterman’s Pit local fauna, Centre County, Penn-
sylvania. American Antiquity, 32:231-232.

Hartwig, S. H. 1966. Rattlesnakes are where and when you find them. Journal of the Ohio Heipe-
tological Society, 5:163.

Holman, J. A. 1967. A Pleistocene herpetofauna from Ladds, Georgia. Bulletin of the Georgia Acad-
emy of Science, 25:154-166.

— — . 1984. Herpetofaunas of the Duck Creek and Williams local faunas (Pleistocene: Illinoian)

of Kansas. Pp. 20-38, in Contributions in Quaternary Vertebrate Paleontology: A Volume in
Memorial to John E. Guilday (H. H. Genoways and M. R. Dawson, eds). Carnegie Museum of
Natural History Special Publication 8.

— — — . 1995. Amphibians and Reptiles in the Pleistocene of North America. Oxford University
Press, New York, New York.

McCoy, C. J. 1982. Amphibians and Reptiles in Pennsylvania. Carnegie Museum of Natural History
Special Publication 6.

Sexton, O. J., and S. R. Hunt. 1980. Temperature relationships of snakes (Elaphe obsoleta and
Coluber constrictor) in a cave hibernaculum. Herpetologica, 36:20-26.

am

-. j- c

>1 <\)r

»* VI

iw ;Vii(l^ y^ *!?<■. .‘i

> - .♦• ..-.i’,Dir ■ . mite|f^>*A..;»^->....v ..jyi ^

7’ *•■ ■ WS

■T,';? V,-®|ff" •' ^ ■' *■ •>: -r

I' f --«epcVll#| J'l' ..t*.s);||PLH . frj.v ■ ^

■^*’/ 4^ 'n>*/t ' • . ■--''iVV- '. ^,. ■- '

| .y \ .••iii?il ■’■) \ \'‘v^--- AW •■-

/ I . ■'' ’■ '

f,' . r|/..^#Jri;' V *

-' W.'l

( iH AMftil Ml,

JB- . •"'pt

’■'• li -.-'Pi

^ "i-^. vi

i:S

l;

U.,.

^ I *»«!

. ‘ » .t» 4*4rs.^

* »i'

■ ,i.’ A-'w:.:;;# '■.

. * I If ■ |’.» ■* ^ ■ ' ■

^ ^ 'r« >»Mu ,

sr V fjifn'. '■ ‘ '*’ ** i'**'*:**,

*v * ■' »**<'’ '^' ‘ *1 J ■ ’*»»,» iJvr

%^ »_. * j ?.*;■'. v»i vy ’v

•. -v/ iij

.*% r4Mvrt

^ 5^^■ 'i.i ■

, hix • P •«■ - ' '^* *4 '<rf

r*¥«« ‘'I

‘■it '»".‘s. ^.f-;..

0 ^f*y'

«v V ‘^*'!V

BE,

■;( ’

, '-'V ■

■. * i
•?• '.'^ > vuf *■

.^.. .'•ifiti,^ i4gir«%»

n . 't£

wr

• n '‘"i; -' JH.'* " '■

'■V.^ ■' ■ *■« ' ii*^'' Wf«

*.' ;^i

.IS

', - ' . * . ■'' • ■>^- ■ 5 ,* / >■ ' ^ -vr

■'I

7 - '''^"

1

)f. f^%*.\y:h^ .
.-.-.r-w- . * « 4»*C

'/:' i:|#;vT*' ■ ■, ■ :.'■ ■' i-mv-

ANNALS OF CARNEGIE MUSEUM

VoL. 69, Number 1, Pp. 11-21

18 February 2()0()

NEW ATOKAN PRODUCTOID BRACHIOPODS FROM THE UPPER
CARBONIFEROUS LADRONES LIMESTONE
OF SOUTHEASTERN ALASKA, WITH A PRELIMINARY NOTE ON THE
PHYLOGENY AND CLASSIFICATION OF
THE TRIBE RETARIINI

Stanislav S. Lazarev *

John L. Carter

Curator, Section of Invertebrate Paleontology

Abstract

One new genus and two new species of productoid brachiopods of the Subfamily Productininae are
described from the Carboniferous Ladrones Limestone (Atokan, late Bashkirian or early Moscovian)
of Prince of Wales Island, southeastern Alaska. The new genus Caruthia, type species Caruthia bo-
realis n. sp., is assigned to the Tribe Productinini. A new species of the genus Rugivestis Muir- Wood
and Cooper, 1960, of the Tribe Paramarginiferini, R. girtyi, is similar and probably closely related to
Rugivestis pristina Carter and Poletaev, 1998, of approximately the same age, from Ellesmere Island,
Canadian Arctic Archipelago.

Discovery of a shagreen texture within the ventral beak region of the genera Keokukia Carter, 1991,
Tesuquea Sutherland and Harlow, 1973, and at least two species of Spinocarinifera Roberts, 1971,
necessitates a new interpretation of the phylogenetic relationships and derivation of the Tribe Retariini
(Subfamily Productinae). We suggest that 'EuvdLsidLW Antiquatonia Miloradovich, 1945, and North Amer-
ican Tesuquea Sutherland and Harlow, 1973, were sister genera, both derived from Keokukia Carter,
1991, common to both continents, and probably derived from the Tournaisian Spinocarinifera nigra—
arcuata group.

Key Words; brachiopods, productoids. Carboniferous, Atokan, Alaska, Ladrones Limestone, Caruthia

Introduction

The diverse brachiopod fauna of the Ladrones Limestone of southeastern Alas-
ka is undescribed. G. H. Girty, of the U. S. Geological Survey, made sizable
collections from this formation in 1918 and initiated identification of the fauna
but did not publish on it although he apparently recognized its unusual nature;
several of Girty’s sorted taxa are labeled as new species. Several nonbrachiopod
faunal elements of the Ladrones Limestone fauna have been described. Savage
and Barkeley (1985) described the conodonts, Hahn and Hahn (1991, 1992) the
trilobites, and Douglas (1971) the fusulinids.

The Ladrones Limestone was named by Eberlein and Churkin (1970:59) for
about 300 m of thick or indistinctly bedded gray sublithographic limestone ex-
posed on the Ladrones Islands in Trocadero Bay near Prince of Wales Island,
southeastern Alaska. The fossils described here are from two very small islands
composed entirely of Ladrones Limestone with no good indication of relative
stratigraphic position.

The precise age of these collections is not certain. The age of the Ladrones

' Paleontological Institute, Russian Academy of Sciences, Profsoyuznaya 123, Moscow 117647, Rus-
sia.

Submitted 31 July 1998.

11

12

Annals of Carnegie Museum

VOL. 69

Limestone fusulinids is Middle Pennsylvanian according to Douglas (1971). Ac-
cording to Savage and Barkeley (1985) the Ladrones Limestone is Lower to
Middle Pennsylvanian as based on several conodont samples. The brachiopods
suggest an Atokan age. The generic composition of the brachiopod fauna is close-
ly similar to that of the Hare Fiord Formation of Ellesmere Island which is known
to be of Atokan age (Nassichuk, 1975).

This paper contains descriptions of two new productoid species in the Ladrones
Limestone fauna and represents the first description of Alaskan Carboniferous
brachiopods.

All of the specimens upon which these new species are based were collected
by G. H. Girty of the U. S. Geological Survey in 1918 from small unnamed
islands of the Ladrones group in Trocadero Bay, about six and one-half miles
south-southeast of Craig (Craig Quadrangle), Prince of Wales Island, Alaska. In
May 1997, the junior author collected more specimens from these localities, al-
though none of the specimens described here was collected at that time. The three
collections used here — USGS 3708-PC, 3762-PC, and 3763-PC — are from the
south coast of a small island in the middle of the Ladrones Islands, near the
mouth of Trocadero Bay, Prince of Wales Island, Alaska. Girty gave the coordi-
nates of locality 3708-PC as latitude 55°23'00"N, longitude 133°04'52"W as
based on an old coastal chart. Coordinates of latitude 55°23.095' N, longitude
133°05.136' W were obtained using a GPS device in 1997 for a locality probably
slightly to the west of Girty’s 3708-PC locality. USGS 3762-PC includes speci-
mens from various localities on the south coast of this island. USGS 3763-PC is
located just east of locality 3708-PC.

Systematic Paleontology

A revised classification of the Order Productida was recently published by
Brunton et al. (1995). In general, that classification is followed here.

All primary types are deposited in the National Museum of Natural History,
Washington, D.C. (acronym USNM). The specimens in Figure 3 are in the col-
lections of the Carnegie Museum of Natural History (CM), Pittsburgh, Pennsyl-
vania.

Suborder Productidina Waagen, 1883
Superfamily Productoidea Gray, 1840
Family Productellidae Schuchert, 1929
Subfamily Productininae Schuchert, 1929
Tribe Productinini Muir- Wood and Cooper, 1960
[= Chonetellini Likharev, 1960]

Caruthia, new genus

Type Species. — Caruthia borealis, n. sp.

Derivation of Name. — This genus is named in honor of Ruth C. Carter and is derived from her
surname and first name.

Diagnosis. — Small, outline subtriangular but not nasute; disc concavoconvex
with shallow corpus; lateral profile nearly semicircular or moderately asymmet-
rical but without clear geniculation; venter near midlength weakly convex, flat-
tened or with weak sulcus, dorsum occasionally with low fold; ornament con-
sisting of fine weak ribs on both valves, excluding posterior part of visceral disc.

2000

Lazarev and Carter — Alaskan Atokan Productoid Brachiopods

13

and weak rugae on dorsal valve; rare thick spines only on ventral valve, including
row of spines at base of flanks; interiors of both valves with strong lateral ridges
bordering ears and extending anteriorly as weak marginal ridges; muscle scars
weakly impressed in both valves.

Comments. — The discovery of this genus permits us to suggest a new phylog=
eny and systematic composition for the Subfamily Productininae, as submitted
for the forthcoming revised edition of the Treatise on Invertebrate Paleontology
(see also the revised classification proposed by Brunton et aL, 1995). The latter
was based on the assumption that the Tribe Chonetellini appeared in the Permian,
being derived from the Paramarginiferini by the loss of ribbing. Now we have
evidence to suggest two parallel lineages during the Carboniferous and Permian.
One of them, the Productinini [= Chonetellini] was more conservative, that is,
without pronounced ribbing, always with a shallow corpus cavity, lacking a cinc=
ture, and with a row of spines on the flanks. The second lineage consisted of the
Tribe Paramarginiferini, forms with a shallow or moderately deep corpus cavity,
stronger ribbing, sometimes with a cincture, and lacking a row of spines on the
flanks. Both lineages are comprised in the Subfamily Productininae, characterized
by a subtrigonal outline (sometimes nasute anteriorly), a few coarse spines on the
ventral valve only, the development of marginal structures inside both valves, and
the peculiar orientation of the brachial ridges (with the anterior lobe axes directed
anteromediad).

Comparisons.- — Caruthia differs from all other members of the Tribe Produc-
tinini in its weak, obscure ribbing and strong lateral ridges inside each valve. In
additioti, it differs from Productina Sutton, 1938, in its costate nonlamellose dor-
sal valve. Argentiproductus Cooper and Muir- Wood, 1951, is much more trans-
verse and less convex and has a lamellose dorsal exterior. Similarly, Dorsirugatia
Lazarev, 1992, from the Late Devonian of Mongolia, is more transverse in outline
and has a lamellose, more weakly costate dorsal exterior. Productellina Reed,
1943, from the very early Carboniferous of England, is similar in outline to
Caruthia but differs in having a strongly lamellose, noncostate dorsal exterior, in
addition to the general differences noted above.

Age and Distribution. — Atokan (late Bashkirian or early Moscovian) of south-
eastern Alaska.

Species Assigned.~~~Type species only.

Caruthia borealis, new species
(Fig. lA-GG)

Holotype.—-A ventral valve, USNM 498809 (Fig. 1A--D), from USGS locality
3708-PC.

Paratypes. — Four ventral valves, USNM 498810- 498813 (Fig. lE-T), from USGS locality 3708-
PC; a natural mold of the dorsal exterior, USNM 498817 (Fig. lEE-GG), from USGS locality 3708-
PC; two ventral valves, USNM 498814 and 498815 (Fig. lU-BB), from USGS locality 3763-PC; a
dorsal interior, USNM 498816 (Fig. ICC, DD), from USGS locality 3763-PC.

Description. — Shell small (length of hinge up to 16 mm), length of visceral disc 6-8 mm; outline
subtriangular with acute cardinal extremities forming cardinal angle of about 45°; corpus cavity shallow
(about 2 mm deep); shell material moderately thick; trails present in both valves but not well separated,
being approximately the same length or a little longer than visceral disc.

Ventral valve strongly inflated, hemispherical in transverse profile; venter generally without sulcus
but occasionally flattened; beak overhanging hingeline, wide; umboeal angle a little less than 90°; ears
subtriangular, flattened, sharply delimited from umbonal region by concave flexures; weak costellae
present on anterior part of visceral disc and trail; irregular fluting present on trail of some specimens;

14

Annals of Carnegie Museum

VOL. 69

2000

Lazarev and Carter — Alaskan Atokan Productoid Brachiopods

15

rare thick spines (diameter 0.6— 0.8 mm, sometimes up to 1 mm) scattered on trail; hinge spines absent;
rare spines scattered on venter and trail; curving row of up to four spines wraps around ears and down
flanks.

Dorsal valve moderately concave; dorsum commonly arched as low fold; ornament consisting of
weak concentric rugae, weak ribs, and rounded pits complementary to ventral spines.

Ventral interior with strong, but not crenulated, lateral ridges becoming weaker anteriorly; muscle
scars generally not seen, or more rarely, short adductor scars situated medially near posterior ends of
larger diductor impressions.

Dorsal interior with lateral and marginal ridges bordering corpus and peripheral cavities as in ventral
valve, with ridges following closely those of ventral valve but being placed slightly outside those of
opposite valve; cardinal process wide, quadrilobed internally; dorsal muscle scars almost indistinguish-
able; brachial ridges not impressed; surface finely tuberculate excepting posteromedianly in both
valves.

Measurements. — See Table 1.

Distribution. — USGS Locality 3708-PC (more than 60 specimens) and USGS
locality 3763“PC (about 30 specimens).

Tribe Paramarginiferini Lazarev, 1986
Genus Rugivestis Muir- Wood and Cooper, 1960
Rugivestis girtyi, new species

(Fig. 2A--R)

Holotype.—A ventral valve, USNM 498817 (Fig. 2E--H), from USGS locality
3708-PC.

Paratypes. — Three partial ventral valves missing the ventral visceral disks, USNM 498818, 498821,
and 498822 (Fig. 2A-D, M-R); a ventral valve, USNM 498820 (Fig. 2I-L); all from USGS locality
3708-PC.

Diagnosis. — This species is characterized by relatively weak concentric rugae
on the visceral disc and coarse costae on the trail.

Description. — Shell small (width up to 2 cm), length of visceral disc about 7-9 mm; outline trans-
versely subtrigonal, variably nasute anteriorly but commonly with well- defined, incomplete siphon;
both valves geniculate, angle between visceral disc and trail 90°; corpus cavity shallow (about 2 mm).

Ventral visceral disc weakly convex, without sulcus but medially flattened or with very weak de-
pression; beak small, obtuse, slightly overhanging hingeline; ears subtriangular, flattened, well delin-
eated from flanks and trail; cardinal angle about 60°; trail approximately twice as long as visceral
disc, almost straight in longitudinal profile but with small concave flexure posterior to nasute extension;
posterior part of trail flat in transverse profile or with weak sinus; lateral portions of trail sloping
steeply to commissures near ears; outline of nasute extension variable in ventral view but sharply
delimited from remainder of trail; radial ribs covering most of surface, excluding beak and ears; weak
rugae present on visceral disc, where ribs are better developed than rugae, 6—9 ribs per 5 mm (com-
monly 7-9) near point of geniculation; width of ribs increases anteriorly but becoming more variable
(3-7 per 5 mm); nasute region generally with weaker ribbing; rugae variably developed but commonly
with less relief than ribs, or when well developed they can cover the point of geniculation; spines rare
and of moderate diameter (0.3-0.4 mm), often difficult to detect; interior with strong lateral ridges

Fig. 1. — Caruthia borealis, n. gen. n. sp. A-D. Ventral, anterior, posterior, and lateral views of the
holotype from USGS locality 3708-PC, USNM 498809. E-T. Ventral, anterior, posterior, and lateral
views of four ventral valve paratypes from USGS locality 3708-PC, USNM 498810-498813. U-BB.
Ventral, anterior, posterior, and lateral views of two ventral valve paratypes from USGS locality 3763-
PC, USNM 498821, 498822. CC, DD. Dorsal and posterior views of a dorsal valve interior paratype
with cardinal process from USGS locality 3763-PC, USNM 498816. EE-GG. Dorsal, anterior, and
posterior views of a natural mold of a dorsal valve exterior paratype from USGS locality 3708-PC,
USNM 498817. All X 2.

16

Annals of Carnegie Museum

VOL. 69

Table 1. — Measurements in mm of the type specimens o/Caruthia borealis, n. gen. n. sp. DV = dorsal

valve.

USNM #

Locality

Length

Width

Height

Surface

measure

498809

3708-PC

11.7

+ 11.8

7.2

21.0

498810

3708-PC

11.7

12.2

7.1

19.5

498811

3708-PC

11.2

12.6

6.9

20.2

498812

3708-PC

10.0

11.4

7.8

17.1

498813

3708-PC

10.1

12.1

6.0

14.8

498817 (DV)

3708-PC

8.6

+ 10.2

4.3

11.3

498814

3763-PC

11.2

+ 11.5

7.2

19.1

498815

3763-PC

10.1

+ 11.0

6.7

18.0

498816 (DV)

3763-PC

7.9

9.7

3.3

10.5

Fig. 2. — Rugivestis girtyi, n. sp. A-D. Ventral, anterior, posterior, and lateral views of a large paratype
with the visceral disc removed leaving the ventral trail and mold of the dorsal exterior, USNM 498818.
E-H. Ventral, anterior, posterior, and lateral views of the nearly complete ventral valve holotype,
USNM 498819. 1-L. Ventral, anterior, posterior, and lateral views of a ventral valve paratype, USNM
498820. M-P. Ventral, anterior, posterior, and lateral views of a paratype with the visceral disc removed
leaving the ventral trail and mold of the dorsal exterior, USNM 498821. Q, R. Anterior and ventral
views of a partial specimen with the corpus missing of an unusually elongated nasute paratype, USNM
498822. All X 1.5; all from USGS locality 3708-PC.

2000

Lazarev and Carter — Alaskan Atokan Productoid Brachiopods

17

Table 2. — Measurements in mm of the type specimens of Rugivestis girtyi, n. sp. from USGS locality

370H-PC.

USNM #

Length

Width

Height

Surface

measure

498818

17.0

+ 19.7

6.8

23.4

498819

18.5

16.7

8.3

26.7

498820

16.1

+ 17.2

8.6

27.0

498821

15.8

+ 15.9

6.7

21.2

498822

16.4

+ 14.0

6.1

—

which merge into thick marginal ridge anteriorly; cincture sometimes formed externally, marking
internal thickened marginal ridge and producing weakly concave resupination in longitudinal profile;
muscle scars not impressed.

Visceral disk of dorsal valve flatter and more sharply geniculate than opposite valve; ears delimited
from flanks by narrow ridges; trail slightly longer than visceral disc but anterior portion of trail not
forming nasute extension; ribs weaker than those of ventral valve; concentric rugae dominate over
ribs in visceral disk, unlike ventral valve; spines and pits lacking; internally with vague, short, lateral
ridges which delimit ears in juvenile stages; cardinal process short and wide; adductor muscle scars
forming obscure triangular thickening.

Measurements. — See Table 2.

Comparisons. — This new species differs from R. pristina Carter and Poletaev,
1998, from similar-- aged strata of Ellesmere Island in having broader ribs, much
weaker rugae and a narrower, better defined anterior nasute extension. The weak
ornamentation of the visceral disc also differentiates R. girtyi from other species
of the genus Rugivestis.

Comments.— A well-preserved dorsal interior of this distinctive genus is still
not known.

Distribution. — USGS Locality 3708-PC (more than 60 specimens); USGS lo-
cality 3763-PC (about 25 specimens); USGS locality 3762-PC (3 specimens).

A Preliminary Note on the Phytogeny and
Classification of the Tribe Retariini

During the past several years, ideas about the phytogeny and systematics of
the group of productoids with a deep corpus cavity, commonly referred to the
productoid Family Dictyoclostidae of Muir- Wood and Cooper (1960), have
changed fundamentally. This is due to reevaluation of the systematic importance
of the internal characters of productidines as a whole. This reevaluation is based
on the sequence of the appearance of these morphological characters during on-
togeny (Brunton et ah, 1995). In particular, the appearance of the deep corpus
cavity (formerly termed the body cavity) and the marginal or peripheral cavities
separated from this deep corpus, which functioned as a defense against penetration
of the mantle cavity by undesirable particulate matter, are of great evolutionary
significance (Lazarev, 1985). This note is a preliminary report on the ongoing
investigation of the nature of the ventral portion of the corpus cavity.

Here we emphasize the systematic importance of a relatively recently appre-
ciated morphological character, the presence of a shagreen (rough or pitted) tex-
ture on the inner umbonal surface of the ventral valve beak. Lazarev (1985, 1990)
discussed the nature of the coelomic cavity of productoids at some length.

The ventral beak region posterior to the muscle scars within most productoids
is smooth, indicating that a normal visceral or coelomic cavity was present. How-

18

Annals of Carnegie Museum

VOL. 69

ever, some productoids have fine pits or irregular grooves posterior to the muscle
scars that may indicate gonadal attachment. Small tubercles (or endospines) that
are the sites of papillae may be present. This rough shagreen texture indicates the
extension of mantle cavity posterior to the ventral muscle field. In other words
we infer that the mantle cavity, with concomitant gonads and papillae, has oc-
cupied the ventral beak region. Thus, the visceral or coelomic cavity hung sus-
pended freely between the dorsal valve cardinal process and the adductor field,
except for the distal portion of the ventral beak. The position and size of the
suspended portion of the visceral cavity was associated with migration of the
dorsal adductor field anteriorly from the hingeline during ontogeny. For example,
in the early Visean genus Keokukia Carter, 1991, the distance between the cardinal
process and the adductor scars is relatively great (Carter, 1991 :fig. 6.1, 6.2). This
anterior position of the adductor scars is additional, indirect evidence of a pene-
tration by the mantle cavity of the middle portion of the ventral valve beak region.

The Tribe Retariini of the Subfamily Productinae, as perceived in this paper,
consists of genera now known to bear a shagreen texture in the ventral beak. It
would include most of the genera assigned by Lazarev (1990) to his Subfamily
Retariinae, namely Retaria Muir- Wood and Cooper, 1960, Kutorginella Ivanova,
1951, Antiquatonia Miloradovich, 1945, Tesuquea Sutherland and Harlow, 1973,
Thamnosia Cooper and Grant, 1969, Thuleproductus Sarycheva and Waterhouse,
1972, and Tubaria Muir- Wood and Cooper, 1960. A shagreen texture has been
observed in the ventral umbo in all of the preceeding genera by the senior author.
Until recently we have not known of true retariins older than the Upper Visean,
but it is now clear that the Lower Visean genus Keokukia belongs in the Tribe
Retariini because the inner beak of its ventral valve has the shagreen texture (see
Fig. 3A of Keokukia rotunda Carter) discussed above. This suggests that Keokukia
might be the ancestor of the Upper Visean retariin genus Antiquatonia and the
later retariins. In Keokukia sulcata Carter there is a short row of spines on the
flanks anterior to the point of geniculation, a feature that is also suggestive of the
genus Antiquatonia.

The ancestor of the earliest retariin Keokukia probably most closely resembled
some species of the genus Spinocarinifera Roberts, 1971. We have detected the
shagreen texture inside the ventral beak, characteristic of the retariins, in Spino-
carinifera nigra (Gosselet, 1888) from northeastern France (Fig. 3D) and S. ar-
cuata (Hall, 1858) from northeastern Missouri (Fig. 3B). These Spinocarinifera
species differ from typical Keokukia in their smaller size, more posteriorly posi-
tioned dorsal adductor scars, and the usual absence of a row of spines on the
flanks in front of the ears.

Therefore, we regard the Spinocarinifera nigra-arcuata group to be the prob-
able ancestor for all of the deep corpus productoids with the shagreen-textured
inner surface of the ventral beak. These include the genus Dictyoclostus Muir-
Wood, 1930 (in the strict sense) which belongs in the Subfamily Dictyoclostinae
(the senior author has observed a shagreen texture in the ventral beak of this
entire subfamily). It seems likely that the genus Keokukia, which occurs in both
North America and Eurasia, was the ancestor of the Eurasian genus Antiquatonia
Miloradovich, 1945, because of its morphology and stratigraphic age.

No authentic species of the genus Antiquatonia (species with shagreen texture
in the ventral beak) is known from North America. We have examined the inner
surfaces of the ventral umbones of several North American species formerly as-
signed to Antiquatonia such as “A. hermosana (Girty, 1903), “A. colora-

2000

Lazarev and Carter — Alaskan Atokan Productoid Brachiopods

19

Fig. 3. — Posterior views of four taxa showing the shagreen texture within the ventral beak. A. Keokukia
rotunda Carter, 1991, from the Keokuk Limestone of eastern Missouri (SL468), CM 45656. B. Spi-
nocarinifera arcuata (Hall, 1858) from the Chouteau Limestone of northeastern Missouri (SL610),
CM 45657. C. Tesuquea formosa Sutherland and Harlow, 1973, from the lower Gobbler Formation
of southwestern New Mexico (SL4758), CM 45658. D. Spinocarinifera nigra (Gosselet, 1888) from
the Calcaire d’Avesnelles Formation (lower Hastarian?) of northeastern France (M. Legrand-Blain
Collection, SL 1244), CM 45659. All X 3.

doensis (Girty, 1915), “A.” crassicostata (Dunbar and Condra, 1932), and “A/’
portlockiana (Norwood and Pratten, 1855). None of these species has the shagreen
texture inside the ventral umbones, nor do any of the Permian species assigned
to the genus Antiquatonia by Cooper and Grant (1975). The North American
species assigned to Antiquatonia may represent an undescribed homeomorph of
considerably later origin than true Eurasian Antiquatonia of Visean and Serpu-
khovian age. This unnamed North American genus is closely related to the genus
Reticulatia Muir- Wood and Cooper, 1960, of Upper Carboniferous and Lower
Permian age, which lacks the shagreen texture characteristic of the retariins. In
fact, Reticulatia has a row of spines delimiting the ears and essentially differs
externally from the so-called North American Antiquatonia only in lacking the
ear ridge bearing the row of spines.

Following the Lower Visean Keokukia, there are no other retariins in North
America until the appearance of the Morrowan genus Tesuquea Sutherland and
Harlow, 1973, which also bears a shagreen texture in the ventral beak (Fig. 3C).
Sutherland and Harlow (1973) suggested close affinity of their new genus with
Antiquatonia and noted the lack of an external ridge delimiting the ears and
supporting the row of spines. However, a dorsal internal ridge delimiting the ears
is absent in Antiquatonia but appears in the retariin genus Kutorginella Ivanova,
1951. Thus, similar morphologies and stratigraphic distributions suggest to us that
Tesuquea and Kutorginella are sister genera, not Tesuquea and Antiquatonia as
suggested by Sutherland and Harlow (1973). In North America, Kutorginella ap-

20

Annals of Carnegie Museum

VOL. 69

pears in the late Desmoinesian as K. lasallensis (Worthen, 1873) and ranges
through the Missourian and Virgilian of the midcontinent; in Eurasia, the genus
appears much earlier. The shagreen texture within the ventral umbo of K. lasal-
lensis attests to its assignment to the genus Kutorginella.

The genera of the lineage Keokukia—Antiquatonia— Kutorginella are widely dis-
tributed in Eurasia. In North America, only Keokukia and Kutorginella are pres-
ent, but the endemic genus Tesuquea replaces Antiquatonia. If this line of rea-
soning is correct, it permits us to draw the following conclusions:

1) The genus Keokukia probably appeared first in Eurasia in the late Tournaisian
and migrated to North America by the late Osagean or early Visean. In Eurasia,
Keokukia gave rise to true Antiquatonia, whereas in North America, much later,
it gave rise to Tesuquea.

2) The marine connection between North America and Eurasia during the Mos-
covian may have been temporary and short- lived. Kutorginella appeared in North
America only near the end of the Desmoinesian (Moscovian). The North Amer-
ican sister genus of Eurasian Antiquatonia, Tesuquea, also was derived from Keo-
kukia and appeared earlier than Kutorginella, sometime in the Morrowan (Bash-
kirian). It was endemic to North America and appeared allopatrically, becoming
extinct later in the Morrowan.

Acknowledgments

We thank Bob Blodgett for stimulating our interest in the Ladrones Limestone fauna. We are grateful
to Bruce Wardlaw for the loan of the Girty collections of the USGS from the Ladrones Limestone of
southeastern Alaska. We also thank our anonymous reviewers for their most useful suggestions for
improvement of the paper.

Literature Cited

Brunton, C. H. C., S. S. Lazarev, and R. E. Grant. 1995. A review and classification of the
brachiopod Order Productida. Palaeontology, 38:915-936.

Carter, J. L. 1991. New brachiopods (Brachiopoda: Articulata) from the Late Osagean of the Upper
Mississippi Valley. Annals of Carnegie Museum, 59:219-247.

Carter, J. L., and V. I. Poletaev. 1998. Atokan (late Bashkirian or early Moscovian) brachiopods
from the Hare Fiord Formation of Ellesmere Island, Canadian Arctic Archipelago. Annals of
Carnegie Museum, 67:105-180.

Cooper, G. A., and R. E. Grant. 1969. New Permian brachiopods from West Texas. Smithsonian
Contributions to Paleobiology, 1:1-20.

. 1975. Permian brachiopods of west Texas. III. Smithsonian Contributions to Paleobiology,

19:795-1298.

Cooper, G. A., and H. M. Muir-Wood. 1951. Brachiopod homonyms. Washington Academy of
Sciences Journal, 41:195-196.

Douglas, R. C. 1971. Pennsylvanian fusulinids from southeastern Alaska. U. S. Geological Survey
Professional Paper, 706:1-21.

Dunbar, C. O., and G. E. Condra. 1932. Brachiopods of the Pennsylvanian System in Nebraska.
Nebraska Geological Survey Bulletin, 5:1-377.

Eberlein, G. D., and M. Churkin, Jr. 1970. Paleozoic stratigraphy in the northwest coastal area of
Prince of Wales Island, southeastern Alaska. U. S. Geological Survey Bulletin, 1284:1-64.

Girty, G. H. 1903. The Carboniferous formations and faunas of Colorado. U. S. Geological Survey,
Professional Paper, 16:1-546.

. 1915. Fauna of the Wewoka Formation of Oklahoma. U. S. Geological Survey Bulletin, 544:

1-353.

Gosselet, j. 1888. L’Ardenne. Memoir pour servir a Texplication de la Carte geologique detaillee
de la France. Baudry, Paris, France.

Hahn, G., and R. Hahn. 1991. Trilobiten aus dem Karbon von SF-Alaska, Teil 1. Geologica et
Palaeontologica, 25:147-191.

2000

Lazarev and Carter — Alaskan Atokan Productoid Brachiopods

21

. 1992. Trilobiten aus dem Karbon von SE- Alaska, Teil 2. Geologica et Palaeontologica, 26;

99-133.

Hall, J. 1858. Part 2. Paleontology. Pp. 473-724, in Report on the Geological Survey of the State
of Iowa, Volume 1 (J. Hall and J. D. Whitney).

Ivanova, E. A. 1951. [New data on productid systematics (genus Kutorginella)]. Akademiia Nauk
SSSR, Doklady, 77:329-331. [In Russian]

Lazarev, S. S. 1985. [The visceral cavity, median septum, and buttress plates in productids.] Pa-
leontologicheskii Zhurnal, 1985:49-58. [In Russian]

. 1990. [Evolution and systematics of the Productida.] Akademiia Nauk SSSR, Paleontologi-

cheskogo Instituta Trudy, 242:1-173. [In Russian]

Lazarev, S. S., and S. Suursuren. 1992. [New productids (Brachiopoda) in the Carboniferous of
Mongolia.] In [New taxa of fossil invertebrates of Mongolia] (T. A. Grunt, ed.). Trudy Sovmest-
naya Rossiisko- MongoFskaya Paleontologichkaya Ekspeditsiya, Moscow, 41:61-69. [In Russian]

Miloradovich, B. V. 1945. [Some data on the morphology of the shells of Productidae.] Bulletin de
L’Acad’aaemie des Sciences de Union des Republiques Sovietiques Socialistes (Biologique), 4:
485-500. [In Russian]

Muir-Wood, H. M. 1930. The classification of the British Carboniferous brachiopod subfamily Prod-
uctinae. Annals and Magazine of Natural History, Series 10, 5:100-108.

Muir-Wood, H. M., and G. A. Cooper. 1960. Morphology, classification and life habits of the
Productoidea (Brachiopoda). Geological Society of America Memoir, 81:1-447.

Nassichuk, W. W. 1975. Carboniferous ammonoids and stratigraphy in the Canadian Arctic Archi-
pelago. Geological Survey of Canada Bulletin, 237:1-240.

Norwood, J. G., and H. Pratten. 1855. Notice of producti found in the western states and territories
with descriptions of twelve new species. Academy of Natural Sciences of Philadelphia Journal,
3:5-22.

Reed, E R. C. 1943. Notes on certain Upper Devonian brachiopods figured by Whidborne, Part 1-

2. Geological Magazine of London, 80:69-78.

Roberts, J. 1971. Devonian and Carboniferous brachiopods from the Bonaparte Gulf Basin, north-
western Australia (two volumes). Bureau of Mineral Resources, Geology and Geophysics Bulletin,
122:1-319.

Sarycheva, T G., and J. B. Waterhouse. 1972. Nekotor’ye brakhiopody semeistva Retariidae iz
permi severnoi kanady. Paleontologicheskii Zhurnal, 1972:62-74.

Savage, N. M., and S. J. Barkeley. 1985. Early to Middle Pennsylvanian conodonts from the Klawak
Formation and the Ladrones Limestone, southeastern Alaska. Journal of Paleontology, 59:1451-
1475.

Sutherland, P. K., and E H. Harlow. 1973. Pennsylvanian brachiopods and biostratigraphy in
southern Sangre de Cristo Mountains, New Mexico. New Mexico Bureau of Mines and Mineral
Resources Memoir, 27:1-173.

Sutton, A. H. 1938. Taxonomy of Mississippian Productidae. Journal of Paleontology, 12:537—569.

WoRTHEN A. H. 1873. Productus lasallensis. Pp. 569, 570, in Paleontology. Descriptions of Inver-
tebrates from the Carboniferous System (E B. Meek and A. H. Worthen). Geological Survey of
Illinois, 5:323-619.

■< y*i: ’- ■ K <'Ut,4#, ' ! '*■■ ^ I#. »

*, *•. * . tj *y^.

»■-;*. '» • 5 '>♦•'' ■/i'4

■%

♦ ' *”1
'- » ■

. . -I

T.-. ■ ■^4;

/.I

i

u», •■ T

T#>;#yi5?' * •’: ■ ' i - '■ -••'*•'■ * * *' -iv^f ‘(rA c

W\':'

i •*'=* ! 4; * JiS\ii.n .'' •4|*Mi.W'**l ■•J.rs ,>, ..t*». :. , rsrriir fJM?i .'^

A . ;J .;:

1 -**»r^-f tBii? -*•■♦»' •“J ‘Jl* l ‘ I .4 \4'*-t. ,.}*^ ,Vl

»♦*.»/ Ifl^W * ' # M’ l !»■ i ' -'*» i *. j tm^ 'te

■■ ., I - I ..■ ‘V -,*,1

-h ♦•!« il-«.4»'>-.,

’ ' -► ■•' - ■ ‘ rM *

LM .M*'.^ 4ij**‘;»V , . ^ A4i3a'J*'>^ . >4

... .-(irfi fcl fJlWJV'W' A

;^i

Oa)

i^f'^

':t.J

'.Li A .^j-j'

U;

** «>rvV. • 'V I'K * IP* *»•

7 . ■ V ► -w. >a'.^

-M . «i,

I V

v'*vA

^’4^. I.,',

j: .. -'Y,.. ii

.,4.

{.z.-

f- •.•<»*

■,/ diiC ♦.I'V? . - ^

,r.

|t . *. ■ ;»

i .!Md|

. -^*4* t l‘Sj

fi*t .M»vrc ta 4 ttr%Y<A^-“=' ii’^V fclf

.... .r* f' ^ •'

K'-: ■-■■■■.■...■• ,. ■^.■; ■ - ^■-2^,'^ ■" :;'-^Sl"kl

ANNALS OF CARNEGIE MUSEUM

VoL. 69, Number 1, Pp. 23-67

1 8 February 2000

EOCENE DECAPOD CRUSTACEANS FROM
PULALI POINT, WASHINGTON

Carrie E. Schweitzer ^

Rodney M. Feldmann^

Research Associate, Section of Invertebrate Paleontology

Annette B. Tucker^

Ross E. Berglund^

Abstract

A robust decapod fauna, including ten species in nine genera, has been recovered from the Eocene
?Aldwell Formation near Pulali Point, Washington. New taxa include Portunites macrospinus n. sp.,
Carpilius occidentalis n. sp., Pulalius dunhamorum n. gen. and sp. Pulalius vulgaris (Rathbun, 1926)
new combination has been removed from Zanthopsis McCoy. Palaeopinnixa rathbunae new name is
offered as a replacement name for Pinnixa eocenica Rathbun which is herein referred to Palaeopinnixa
Via. Palaeopinnixa is placed within the Hexapodidae. This is the first notice of Carpilius Leach in
Desmarest in the eastern Pacific region. The decapod fauna is typical of those previously reported
from deep-water, continental slope settings of the Pacific Northwest with the exception of the occur-
rence of Carpilius, which has a mainly tropical distribution, and Palaeopinnixa which has previously
been reported primarily from lower-latitude settings.

Key Words: Decapoda, Brachyura, Eocene, Olympic Peninsula, Washington

Introduction

A previously undescribed decapod fauna from Eocene rocks at Pulali Point,
Washington, has yielded ten species in nine genera, including one new genus and
four new species. The decapod fauna was mentioned by Squires et al. (1992) and
illustrated by Tucker et al. (1994); however, neither report provided a detailed
systematic treatment of the fauna. The fauna is composed of several decapod taxa
common in continental slope deposits of the Pacific Northwest including Ma-
croacaena Tucker, Laeviranina Lorenthey and Beurlen, Raninoides H. Milne Ed-
wards, Portunites Bell, Branchioplax Rathbun, Pulalius n. gen., and Neopilum-
noplax Serene in Guinot. This report marks the first notice of the genus Carpilius
Leach in Desmarest in the fossil record, based upon material other than a manus,
and on the west coast of North America. This decapod fauna is significant bio-
geographically because it represents a mix of taxa typical of the west coast of
North America (Rathbun, 1926; Tucker and Feldmann, 1990; Feldmann et al.,
1991; Schweitzer and Feldmann, 1999) and taxa not previously reported from the
eastern Pacific region.

Institutional abbreviations used throughout this paper are: CM, Carnegie Mu-
seum of Natural History, Pittsburgh, Pennsylvania; GHUNLPam, Geology Col-

* Department of Geology, Kent State University, Kent, Ohio 44242.

2 11689 NE Sunset Loop, Bainbridge Island, Washington 98110.

Submitted 3 December 1998.

23

24

Annals of Carnegie Museum

VOL. 69

lection, National University of La Pampa, Santa Rosa, Argentina; In., The Natural
History Museum, London, United Kingdom; SM, Sedgwick Museum, Cambridge
University, Cambridge, United Kingdom; USNM, United States National Museum
of Natural History, Washington, D.C.; and JSHC, private collection of J. S. H.
Collins, Forest Hill, London, England.

Geologic Setting

Decapod fossils were collected from the northeast coast of Pulali Point which
extends into the Hood Canal of Puget Sound, south of Quilcene, Washington, in
the Nl/2, Sec. 18, T26N, RIW, of the Seabeck 7.5 minute quadrangle (Fig. 1).
The section at Pulali Point (Fig. 2) includes two major units, the Crescent For-
mation and an overlying sedimentary sequence that has been questionably as-
signed to the Eocene Aldwell Formation based upon lithology, paleontology, age,
and geographic location (Squires et ah, 1992). Squires et al. (1992) determined
that the unit should be referred either to the Aldwell or the Humptulips Formation,
because the two units are similar in both age and lithology. They assigned the
sedimentary rocks at Pulali Point to the ?Aldwell Formation because that unit
crops out within 15 km to the north of the site. The Humptulips Formation crops
out about 80 km to the southwest of Pulali Point (Squires et al., 1992). Because
our study is concerned primarily with decapod systematics, we have no basis for
offering new information concerning the identity of the rocks and concur with
their decision to assign the rocks to the ?Aldwell Formation.

The Crescent Formation, which at Pulali Point occurs as basalts alternating
with basaltic sandstone, basalt pebble conglomerate, and siltstone, lies uncon-
formably beneath the sedimentary sequence referred to as the ?Aldwell Formation
(Squires et al., 1992). They reported one decapod fossil, a specimen of Glyphi-
thyreus weaveri (Rathbun) from the Crescent Formation, but no decapods in this
report were recovered from that unit. The Crescent Formation was reported to be
of Penutian to Ulatisian age based upon calcareous nannofossils, benthic fora-
minifera, and macrofossils (Squires et al., 1992), Babcock et al., (1994:145) re-
ported that the Crescent Formation in the vicinity of Pulali Point had an age of
50.5-1.6 ma based upon ‘^^Ar-^^Ar geochronometry.

The sedimentary sequence at the Pulali Point locality, which is assigned to the
?Aldwell Formation, unconformably overlies the Crescent Formation. It consists
of boulder conglomerate, sandy siltstone, and siltstone with gradational contacts
between each (Squires et al., 1992). The boulder conglomerate, located at the base
of the section, contains crab fossils in siltstone concretions occurring with a pre-
dominance of basalt boulders (Squires et al., 1992). The concretions were con-
sidered by Squires et al. (1992) to have been reworked. Decapods were also
recovered from the silty sandstone facies and were contained within reworked,
well-indurated siltstone concretions (Squires et ah, 1992). Squires et al. (1992)
reported a middle Eocene (early Narizian) age for the formation based upon ben-
thic foraminifera and macrofossils. Babcock et al. (1994) reported that the ?Ald-

Fig. 1. — A. Locality map of the Olympic Peninsula, Washington, showing the Pulali Point collecting
locality (modified from Squires et al., 1992). B. Map of the Pulali Point area, indicating California
State University, Northridge (CSUN) collecting localities, and the location at which the stratigraphic
section in Figure 2 was measured (modified from Squires et al., 1992).

2000

Schweitzer et al. — Eocene Decapod Crustaceans

25

26

Annals of Carnegie Museum

VOL. 69

Fig. 2. — Stratigraphic section of the rocks exposed at Pulali Point. Decapods were reported by Squires
et al. (1992) from CSUN macrofossil localities 1515 and 1516 (modified from Squires et ah, 1992).

2000

Schweitzer et al. — Eocene Decapod Crustaceans

27

well rocks at Pulali Point interfinger with the Crescent basalts, suggesting that
they were syndepositional and placing the age of the ?Aldwell Formation at Pulali
Point as Penutian-Narizian age, the age of the Crescent Formation near Pulali
Point. Squires et al. (1992) had not noted interfingering of these two units and
instead described an angular unconformity between the two units. Decapod oc-
currences support an Eocene age for this unit.

Babcock et al. (1994) summarized the depositional environment of the Aldwell
Formation sensu stricto as being cold, deep, and predominated by submarine
landslides. The depositional environment of the ?Aldwell Formation at Pulali
Point is reported to have been part of a deep-water, submarine fan complex
(Squires et al., 1992). The basalt boulder conglomerates were considered to have
been proximal turbidite deposits, and the silty sandstones were interpreted to have
been derived from multiple provenances, based upon the occurrence of reworked
concretions in that facies (Squires et al., 1992). Both the boulder conglomerate
and the silty sandstone contain decapods, and these fossils were interpreted to
have been derived from both deep- and shallow- water environments (Squires et
al., 1992), also supporting the theory of multiple provenances for the sediments.
Some of the decapods reported here have modem congeners that are typical of
deep-water environments, including Raninoides (Tucker, 1995) and Neopilumno-
plax (Feldmann et al., 1991). The fossil genus Branchioplax was reported from
the Eocene Orca Group, suggested to have been deposited at bathyal depths (Eeld-
mann et al., 1991). The two fossil genera Portunites and Pulalius have been
reported from rocks on the west coast of North America that are interpreted to
have been deep-water, continental-slope deposits. Conversely, the modem species
of Carpilius have been reported from relatively shallow, sublittoral depths of
approximately 5-35 m (Sakai, 1976). The mixed depth preferences of these deca-
pods supports the interpretation that these rocks were deposited in a deep-water,
continental- slope setting with some downslope mixing of the fauna, possibly ex-
plaining the reworked concretions.

Systematic Paleontology

Order Decapoda Latreille, 1803
Infraorder Brachyura Latreille, 1803
Section Podotremata Guinot, 1977
Family Raninidae de Haan, 1841
Subfamily Lyreidinae Guinot, 1993
Genus Macroacaena Tucker, 1998

Type Species. — Lyreidus succedanus Collins and Rasmussen, 1992.

Remarks. — Macroacaena Tucker is characterized by possession of a tridentate
front, two orbital fissures, an anterolateral margin with a hypertrophied spine at
the anterolateral comer and usually a small protuberence just posterior to the
midlength of the anterolateral margin, and a moderately distinct median ridge
(Tucker, 1998). The fronto-orbital width of members of this genus is typically
greater than the posterior margin (Tucker, 1998). Specimens from the Pulali Point
locality are referred to the genus based upon possession of a moderately distinct
ridge, an extremely reduced anterolateral spine, and a hypertrophied spine at the
anterolateral comer. This combination of characters is unique to Macroacaena
(Table 1). Characteristics of the rostmm or orbits are not available because the
front is broken. The posterior width exceeds the fronto-orbital width, occupying

28

Annals of Carnegie Museum

VOL. 69

Table 1. — Distinguishing characters of the dorsal carapace useful in differentiating among the genera
Lyreidus, Lysirude, and Macroaceana. FOW = Fronto-orhital width, PW — posterior width.

Character

Lyreidus

Lysirude

Macracaena

Carapace shape

Fusiform, much longer

Carapace moderately

Carapace widest of the

than wide

wide

three genera

Orbital fissures

One

One

Two

Orbital teeth

One

One

Two

Dorsal ridge

Absent or indistinct

Absent or indistinct

Present

FOW vs. PW

FOW always narrower

FOW equal to or wider

FOW wider

Rostrum

Typically less pro-

Typically more pro-

Typically more pro-

duced

duced

duced

Front

Tridentate

Tridentate

Tridentate

Anterolateral spines

Short or absent

One hypertrophied;
one much reduced

One hypertrophied;
one much reduced

Anterolateral spines

One

Two or less often, one

Two

approximately 0.45 and 0.41 the maximum carapace width respectively. This falls
outside the range typically ascribed to the genus by Tucker (1998), who described
Macroacaena as having a fronto-orbital margin that is usually wider than the
posterior margin (p. 325).

Distinguishing members of the genus Macracoaena from members of the two
closely related genera Lyreidus de Haan and Lysirude Goeke can be difficult, as
has been addressed by Tucker (1998) and Feldmann (1989, 1992). The three
genera possess many overlapping characteristics that indicate that a morphological
gradation exists with Lyreidus and Macroacaena as end members and Lysirude
occupying an intermediate morphologic position (Table 1). However, the combi-
nation of characters of the dorsal carapace for each genus is unique, and distin-
guishing members of each genus is relatively straightforward if specimens are
well preserved. The relative width of the front and the posterior margin should
be used with care. For example, the two genera Macroacaena and Lysirude were
characterized by Tucker (1998) as possessing a posterior width that is often less
than the fronto-orbital width. Lyreidus was described as possessing a fronto-orbital
width that is usually less than the posterior width and is always much less than
one-half the maximum carapace width (p. 324). The specimen from Pulali Point,
which has a combination of other characters that clearly place it in Macroacaena,
has a posterior margin that is slightly wider than the fronto-orbital width, such as
is often seen in Lyreidus. Additionally, Goeke {in Tucker, 1998) noted that the
relative fronto-orbital width with respect to the total width in some species of
Lysirude changes during growth, so that younger individuals typically possess
relatively wider fronto-orbital margins. Therefore, it is possible that this character
may not be useful in disinguishing among Lyreidus, Lysirude, and Macroacaena.

Macroacaena alseanus (Rathbun, 1932)

(Fig. 3B)

Lyreidus alseanus Rathbun, 1932:239, 240, 242, fig. 3, 4. Glaessner, 1960:17; Bennett, 1964:24;

Feldmann, 1989:63-69, fig. 1.1, 1.2, 3. 1-3.8, text fig. 4. 1-4.3.

Lyreidus {Lysirude) alseanus Rathbun. Feldmann, 1992: 951, fig. 4.10, 4.1 1.

Material Examined. — CM 45822, 45823, deposited in the Section of Invertebrate Paleontology,
Carnegie Museum of Natural History, Pittsburgh, Pennsylvania.

Remarks. — Only one of the two specimens from Pulali Point is well preserved.

2000

Fig. 3. — A. Laeviranina vaderensis (Rathbun), CM 45829. B. Macroacaena alseamis (Rathbun),
CM 45822. C. Raninoides fulgidus Rathbun, CM 45824. D. Laeviranina goedertorum Tucker,
CM 45827. Scale bars = 1 cm.

and unfortunately, the entire rostrum and orbital region are missing. Nevertheless,
the specimens can be confidently referred to this species based on several char-
acters. The length/width ratio is less than 0.65, the front is narrow and occupies
about 0.41 the maximum carapace width, the anterolateral margin has an extreme-
ly reduced spine and a hypertrophied spine at the anterolateral angle, the lateral

30

Annals of Carnegie Museum

VOL. 69

Table 2. — Number of species in the genera Laeviranina and Raninoides with wide orbital fissures,

narrow orbital fissures, or grooves.

Character of orbit

Laeviranina

Raninoides

Well-developed grooves

1

8

Narrow grooves

10

4

Fissures

1

1

margins are nearly straight, the posterolateral margins are slightly concave and
converge posteriorly, and there is a moderately welhdeveloped axial ridge on the
dorsal carapace. All of these characteristics conform to the description of the
species provided by Feldmann (1989).

Because the front is missing on this specimen, making it impossible to observe
characteristics of the rostrum and orbits, it is possible to confuse it with members
of another genus known from the Pacific Coast of North America, Carinaranina
Tucker. Members of Carinaranina and Macroacaena look superficially similar if
characters of the front are discounted. Species of Carinaranina typically possess
a hypertrophied spine at the anterolateral corner but do not possess the reduced
spine on the anterolateral margin (Tucker, 1998). This appears to be the best
method of distinguishing Macroacaena alseanus (Rathbun) from members of
Carinaranina, especially C. naselensis (Rathbun), which has a very narrow front
as seen in species of Macroacaena. Other characteristics that distinguish members
of Carinaranina from members of Macroacaena are the more well-developed
axial ridge of Carinaranina that extends the entire length of the dorsal carapace
onto the rostrum, the coarsely punctate ornamentation of Carinaranina, and the
wider orbital grooves and better-developed orbital teeth of Carinaranina. Distin-
guishing members of the two genera is relatively straightforward if the orbits and
rostrum are preserved since those areas are distinctly different in each genus.

Subfamily Raninoidinae de Haan, 1841

Remarks. — The problem of differentiating Laeviranina Ldrenthey and Beurlen
from Raninoides H. Milne Edwards has been addressed on numerous occasions.
Tucker (1998) summarized previous work on this problem, noting that several
criteria have been proposed to differentiate the two genera including the nature
of the orbital fissures, the placement of the anterolateral spine with respect to the
total length of the carapace, the presence or absence of a postfrontal ridge, and
several characteristics of the sternum.

Laeviranina has been considered to possess orbital fissures, while Raninoides
has been considered to possess orbital grooves; however. Tucker (1998) indicated
that both character states occur in each genus. Data from 25 species analyzed in
this study indicates that this character does indeed occur in both species (Table
2), but that there is a general tendency for members of the genus Laeviranina to
possess narrow orbital grooves or fissures and for members of the genus Rani-
noides to possess relatively more open orbital grooves. Feldmann (1991) listed
several features of the sternum that he believed could be used to differentiate the
two genera, including the shape of sternal element four and the nature of the cleft
in sternal element five. Tucker (1998) indicated that these sternal characteristics
are mixed between the two genera as well. According to the key to the genera of
the Raninoidinae provided by Tucker (1998), in species of Raninoides, the ster-

2000

Schweitzer et al. — Eocene Decapod Crustaceans

31

30 40

Specimen No.

▲ Laeviranina

■ Carinaranina

O Quasilaeviranina I

X Raninoides i

- - - Linear (Laeviranina) .

Linear (Quasilaeviranina)

— - Linear (Raninoides)

50 60 70

Fig. 4. — Specimens of Laeviranina, Raninoides, Quasilaeviranina, and Carinaranina plotted by the
ratio of the width between the bases of the anterolateral spines (W2) to the maximum carapace width
(Wl). Linear regression lines show trends in ratios within taxa.

num between the third pereiopods is “usually quite wide” (p. 347), where it is
known, and this same area, where known, is “moderately narrow” (p. 347) in
species of Laeviranina. However, she did not list this as a useful character in
differentiating the two genera (Tucker, 1998). It is necessary to bear in mind that
the generic characteristics of the sternum of Laeviranina are not well known due
to poor or lack of preservation (Tucker, 1998), so this character needs further
investigation as more material becomes available. Tucker (1998) also noted that
the ischium of the first pereiopod of species of Raninoides bears a spine; however,
this element of the pereiopods of Laeviranina is unknown. This character could
prove useful if more and better fossil material is discovered.

Members of the two genera have also been differentiated based on the relative
width of the carapace with respect to the total length, the position of the antero-
lateral spines with respect to the total length, and the width of the fronto-orbital
margin with respect to the maximum width. To test this, specimens and illustra-
tions of specimens referred to nearly all species of both genera, as assigned by
Tucker (1998), were measured, and relevant ratios were calculated. Scatter dia-
grams of the results of this work indicate that there is absolutely no difference
between the two genera with respect to the ratios calculated (Fig. 4-6). Therefore,
the relative width of the carapace, the position of the anterolateral spine, and the
position of maximum width with respect to the total length of the carapace cannot

32

Annals of Carnegie Museum

VOL. 69

Fig. 5. — Specimens of Laeviranina, Raninoides, Qiiasilaevircmina, and Carinaranina plotted by the
ratio of the fronto-orbital width (W3) to maximum carapace width (Wl). Linear regression lines show
trends in ratios within taxa.

be used reliably to differentiate between the two genera. Ratios for members of
another closely related genus, Quasilaeviranina Tucker and the sole specimen of
Carinaranina from Pulali Point, were included on the scatter diagrams for com-
parative purposes.

Tucker (1998) indicated that one characteristic that can reliably be used to
differentiate between the two genera, especially in fossil taxa, is the presence or
absence of a postfrontal ridge. Members of species currently assigned to Laevir-
anina possess such a ridge while members of most species of Raninoides do not.
However, there are several problems that arise when using this single character
to assign specimens to either Laeviranina or Raninoides. Use of the ridge to
differentiate between the two genera may be hampered by differential preservation
of the ridge, because in some specimens, the ridge is very poorly preserved. More
importantly, members of several extant species of Raninoides possess weakly
developed postfrontal ridges in some specimens. For example, specimens of Ran-
inoides louisianensis Rathbun (USNM 12002; each of the USNM specimens con-
tains a lot of several specimens, ranging from one to approximately 50 individ-
uals) have a weakly developed postfrontal ridge marginally but the ridge is not
developed axially. Members of Raninoides bouvieri Capart (USNM 121423) pos-
sess a very weak postfrontal ridge. Finally, members of the species R. personatus
Henderson (USNM 216741) and R. lamarcki Milne Edwards and Bouvier

2000

Schweitzer et al. — Eocene Decapod Crustaceans

33

0.7

0.6

0.5

a 0.4

0.3

0.2

0.1

X

X

X X

i —

A A

O

^ X X

■X" -X

X

X

X

X

X

▲ Laeviranina

■ Carinaranina

O Quasilaeviranina

X Raninoides

1

fU&s \r\Gi fit

- - - Linear (Laeviranina)

Linear (Quasilaeviranina)

10

20

30 40

Specimen No.

60

60

70

Fig. 6 — Specimens of Laeviranina, Raninoides, Quasilaeviranina, and Carinaranina plotted by the
ratio of the length to the position of maximum carapace width (L2) to the total length of the carapace
(LI). Linear regression lines show trends in ratios within taxa.

(USNM 121660) possess weak postfrontal ridges which are best developed on
larger specimens. Additionally, the degree of development of the ridges appears
to vary within each species, sometimes, but apparently not always, associated
with overall carapace size. Specimens of the extant species R. benedicti Rathbun
(USNM 268507) and R. loevis (Latreille) (USNM 273193) exhibit the smooth,
ridgeless, postfrontal regions that have traditionally been ascribed to members of
that genus.

Generic assignment of taxa to Raninoides or Laeviranina, therefore, is difficult
at best, especially since there are no unequivocal characters with which to accom-
plish this. While generic assignment of specimens with well- developed postfron-
tal ridges or smooth postfrontal regions is relatively straightforward, some taxa
possess ridges that are poorly or variably developed. Development of a postfrontal
ridge on several extant species of Raninoides casts further doubt upon the use of
the postfrontal ridge as the only reliable means by which to distinguish between
the two genera Laeviranina and Raninoides. It appears, therefore, that all of the
characters that have been suggested as being useful for differentiating between
the two genera are problematic, unreliable, and mixed between members of the
two genera. While it is premature to assert that these two genera are synonymous,
it is suggested that this possibility be further investigated, especially with regard
to features of the sternum as first noted by Feldmann (1991).

34

Annals of Carnegie Museum

VOL. 69

Genus Laeviranina Lorenthey and Beurlen, 1929

Type Species. — Ranina budapestinensis [sic] Lorenthey, 1897:23.

Remarks.- — Tucker (1998) provided a summary of all species that have been
assigned to the genus Laeviranina^ including several species from the Pacific
Coast of North America that were originally assigned to the genus Raninoides.
North American species originally assigned to Raninoides include L. lewisanus
(Rathbun) and L. vaderensis (Rathbun); Tucker (1998) assigned them to Laevir-
anina based upon their possession of a postfrontal ridge. Upon examination of
these two species, it is apparent that they are synonymous. Rathbun (1926) orig-
inally differentiated between the two species by observing that L. lewisana had
a wider fronto-orbital width, a more anteriorly placed anterolateral spine, a less
convex lateral margin, a straighter posterolateral margin, and a more convex dor-
sal surface. However, examination of illustrations of both L. lewisana and L.
vaderensis in Rathbun’s work (1926) shows that the fronto-orbital width to total
width ratio is 0.70 and 0.68 respectively, not a significant difference. Further,
examination of illustrations of specimens subsequently referred to each species
by Tucker and Feldmann (1990) and Tucker (1998) shows that the two species
do not differ substantially in the range of fronto-orbital width to total width ratios,
averaging 0.70 in L. lewisana and 0.67 in L. vaderensis. Rathbun (1926) indicated
that the anterolateral spine was placed more anteriorly in L. lewisana, and ex-
amination of her plates confirms this (plate 22, fig. 4, 5). However, examination
of figures of specimens referred to each species by Tucker and Feldmann (1990)
and Tucker (1998) indicates that the two species overlap in range in this ratio
also. The length to the base of the anterolateral spine occupies an average of 0.20
of the total length in L. lewisana and 0.21 in L. vaderensis. Tucker (1995) reported
that the orbital spines were slightly shorter than the rostrum in L. lewisana but
she did not quantify the length of the orbital spines with respect to the rostrum
for L. vaderensis. Rathbun’s (1926) illustrations indicate that the orbital spines do
not extend as far as the rostrum in L. vaderensis, but she did not comment on
the length of the spines in her descriptions of either L. vaderensis or L. lewisana.
Examination of figures of specimens referred to each species by Tucker and Feld-
mann (1990) and Tucker (1998) indicates that there is a range in orbital spine
length with respect to the rostrum in both species, ultimately not providing a
means of differentiating the two species. The orbital spines in specimens assigned
to L. lewisana are typically longer than those of specimens of L. vaderensis, but
there is a range in the length of the spines within each species.

Tucker (1998) provided several means by which the two species could be dif-
ferentiated, including the greater width of the fronto-orbital margin in L. lewisana
which has already been shown to be virtually the same in each species. She also
suggested that the more egg-shaped, ovate carapace of L. lewisana might be used
to distinguish the two genera. However, there appears to be a range of variation
with regard to this character in both species, and in fact, Rathbun (1926) reported
that the lateral margins of L. lewisana were “not so strongly outcurved behind
the [anterolateral] spine” and that the posterolateral margin was straighter than
that of L. vaderensis (p. 94), seemingly contradicting Tucker (1998). Tucker
(1998) suggested that the posterior width of L. lewisana was slightly wider than
that of L. vaderensis, but there is a large degree of overlap in the ratio of the
posterior width to the maximum width in the two species in this character. Based
upon measurements in Tucker (1995), the posterior width to maximum width ratio

2000

Schweitzer et al. — Eocene Decapod Crustaceans

35

averages 0.46 in L. vaderensis with a range of 0.44-0.46 and that ratio averages
0.41 in L. lewisana with a range of 0.37-0.45. She also suggested that the an-
terolateral spines form a V-shaped angle with the anterolateral margin in L. lew-
isana, while in L. vaderensis, they form a U- shaped angle. Examination of plates
of specimens referred to these two taxa by Rathbun (1926), Tucker and Feldmann
(1990), and Tucker (1998) indicates that there is a range of variation in the shape
of the angle in both species and that both U-shaped and V-shaped angles occur
in each species. Laeviranina vaderensis was distinguished from L. lewisana by
Tucker (1998) based on its less distinct postfrontal ridge that is steeply and tightly
arched. However, it has been demonstrated above that use of the postfrontal ridge
may not be useful in differentiating between Raninoides and Laeviranina, because
the ridge can be variable in degree and nature of development within genera and
even within species. Therefore, it seems unlikely that this particular character will
be useful for differentiating taxa at the species level, especially when used as the
sole distinguishing characteristic. Based upon this evidence, therefore, the two
taxa are herein synonymized, with L. lewisana being the junior synonym.

Laeviranina vaderensis has previously been reported from the middle Eocene
of Alaska and Oregon and the upper Eocene of Washington, and L. lewisana was
previously reported from the Eocene of Washington. The range of L. vaderensis
is therefore not extended either geographically or chronologically by the synon-
ymy.

Laeviranina vaderensis (Rathbun, 1926)

(Fig. 3A)

Raninoides vaderensis Rathbun, 1926:93, pi. 22, fig. 5. Glaessner, 1929:372; Tucker and Feldmann,
1990:412, fig. 3.1, 3.2; Karasawa, 1992:1252.

Laeviranina vaderensis (Rathbun): Tucker, 1998:353, pi. 17, 18.

Raninoides lewisana Rathbun, 1926:94, pi. 22, fig. 4. Glaessner, 1929:372; Forster and Mundlos, 1982:

158.

Laeviranina lewisana (Rathbun): Glaessner and Withers, 1931:490, 491; Via, 1965:263; Via, 1969:
125; Tucker, 1998:351, fig. 15, 16.

Material Examined. — CM 45829-45832, deposited in the Section of Invertebrate Paleontology, Car-
negie Museum of Natural History, Pittsburgh, Pennsylvania.

Remarks. — -The specimens here referred to L. vaderensis possess several char-
acters diagnostic of the species and are therefore confidently referred to this spe-
cies. They possess an oblong carapace that ranges from being oblong-ovate to
ovate. The fronto-orbital width occupies about 0.70 the maximum carapace width,
and the orbital spines range from extending about half the length of the rostrum
to extending to about the same length as the rostrum. The postfrontal ridge is
subtle, arcuate, and slightly sinuous. The anterolateral spines extend more forward
than laterally and form a U-shaped or V-shaped angle with the anterolateral mar-
gins. The posterolateral margins taper markedly to their intersection with the
posterior margin. Some of the specimens here referred to Laeviranina vaderensis
possess long orbital spines that extend to the same length as the rostrum, which
is at the long end of the range of variation for the species. However, some spec-
imens possess orbital spines that extend only about half the distance of the ros-
trum, clearly illustrating the range of variation in the species.

Laeviranina vaderensis has previously been reported from the middle Eocene
Orca Group of Alaska, the middle Eocene of Oregon, and the upper Eocene Tejon
Formation and Hoko River Formation of Washington (Tucker, 1998). This oc-

36

Annals of Carnegie Museum

VOL. 69

currence extends the geographic range of the species to the eastern Olympic Pen-
insula, Washington.

Laeviranina goedertorum Tucker, 1998
(Fig. 3D)

Laeviranina goedertorum Tucker, 1998:348, fig. 13, 14.

Material Examined. — CM 45827, 45828 deposited in the Section of Invertebrate Paleontology, Car-
negie Museum of Natural History, Pittsburgh, Pennsylvania.

Remarks. — The Pulali Point specimens referred to Laeviranina goedertorum
Tucker possess a suite of characters that place them in this species as described
by Tucker (1998). The specimens possess an elongate carapace that achieves its
maximum width at about the midpoint and converges slightly posteriorly, as seen
in L. goedertorum. The front on the Pulali specimens and L. goedertorum con-
verges anteriorly, producing a relatively narrow fronto-orbital width of around
0.60 the maximum carapace width. The postfrontal ridge on L. goedertorum and
the Pulali specimens is sinuous and moderately well developed. The orbital teeth
are shorter than the rostrum in both the new specimens and L. goedertorum. The
specimens are therefore confidently referred to this species.

The maximum carapace width of Laeviranina goedertorum has been described
as being located about one-third the distance posteriorly on the carapace. Dimen-
sions of the new specimens indicate that there is a range from about one-third
the distance to just under half the distance posteriorly on the carapace. The pos-
terolateral margins of the carapace do not converge as acutely in this species as
they do in other species of the genus. Laeviranina goedertorum is therefore wider
along its entire length than other species of the genus and possesses a relatively
wide posterior margin that is markedly wider than that seen in L. vaderensis. The
two species may also be distinguished based upon several other criteria as well.
The front of Laeviranina goedertorum converges anteriorly, resulting in an overall
narrower fronto-orbital width to maximum width ratio in L. goedertorum as com-
pared to L. vaderensis. The postfrontal ridge of L. goedertorum is better developed
and more sinuous than that of L. vaderensis.

Laeviranina goedertorum has previously been reported from the Eocene Hoko
River Formation (Tucker, 1998). This occurrence in the Eocene Aldwell Forma-
tion extends the geographic range to the eastern Olympic Peninsula, Washington.

Genus Raninoides H. Milne Edwards, 1837

Type Species. — Ranina laevis Latreille, 1825.

Remarks. — The specimens here referred to Raninoides H. Milne Edwards lack
a postfrontal ridge, the main characteristic that can be reliably used to differentiate
between the two genera Laeviranina and Raninoides. These specimens can there-
fore be unequivocally assigned to Raninoides because they possess a completely
smooth postfrontal region.

Raninoides fulgidus Rathbun, 1926
(Fig. 3C)

Raninoides fulgidus Rathbun, 1926:96-97.

Material Examined. — CM 45824-45826 deposited in the Section of Invertebrate Paleontology, Car-
negie Museum of Natural History, Pittsburgh, Pennsylvania.

2000

Schweitzer et al. — Eocene Decapod Crustaceans

37

Remarks. — Two specimens from the Pulali Point locality exhibit a combination
of characteristics that place them in the species Raninoides fulgidus Rathbun. The
specimens possess relatively long orbital and anterolateral spines and wide orbital
fissures, characteristic of the species R. fulgidus. The rostrum does not extend as
far as the orbital spines as described for this species by Tucker (1998); however,
she described the inner orbital fissure as being deeper than the outer. In the Pulali
Point specimens, the fissures are the same length.

Section Heterotremata Guinot, 1977
Superfamily Portunoidea Rafinesque, 1815
Family Portunidae Rafinesque, 1815
Subfamily Polybiinae Ortmann, 1893
Genus Portunites Bell, 1858

Type Species. — Portunites incerta Bell, 1858.

Remarks. — The diagnosis for the genus Portunites Bell given by Glaessner
(1969) describes the genus as having a hexagonal carapace that is not much wider
than long, a four-toothed front, an anterolateral margin with four or five teeth
with the last being the longest, an arcuate ridge extending from the lateral tooth
to the medial area, a straight or concave posterolateral margin, well-defined gas-
trocardiac regions, and fifth pereiopods with dactyli that are not flattened (p.
R513). Tucker and Feldmann (1990) emended the generic diagnosis to include a
paddlelike dactylus on the fifth pereiopod based on its presence in Portunites
alaskensis Rathbun.

Glaessner (1969) did not mention the width of the orbits in this genus, each of
which occupies a little over one-third the fronto-orbital width in most members
of the genus and about one-fifth to one-quarter the maximum carapace width. The
fronto-orbital width of most members of the genus occupies greater than two-
thirds the maximum carapace width. Bell (1858:21) described the orbits in the
type species, P. incerta Bell, as extending to the middle of the hepatic region,
and measurements of individuals of this species indicate that the fronto-orbital
width occupies approximately two-thirds the maximum width of the carapace. A
cast of a previously unillustrated specimen of P. incerta (JSHC 1732d) (Fig. 7B)
documents the great width of the orbits in the type species of the genus. The front
is not clearly evident in previously published plates. The subsequently named
British species including P. stintoni Quayle and P. sylviae Quayle and Collins
have orbits that are similar in relative size to those of the type species. The orbits
of P. alaskensis Rathbun are extremely large; the fronto-orbital width in this
species occupies more than 0.70 the maximum width of the carapace. The fronto-
orbital width of the new species to be described below, Portunites macrospinus,
occupies 0.67 the maximum carapace width.

Two species of Portunites possess relatively narrow fronto-orbital width-to-
width ratios but can be assigned to the genus based upon a combination of other
important characters. The orbits of Portunites insculpta Rathbun occupy approx-
imately 0.50 the maximum carapace width, but this taxon can be assigned to
Portunites based upon its possession of a well-developed epibranchial ridge, car-
apace shape and regions typical of the genus, and four anterolateral spines ex-
cluding the outer orbital spines. Portunites kattachiensis Karasawa possesses a
fronto-orbital width which occupies about 0.53 the maximum carapace width.
However, Portunites kattachiensis has several other generic characters that clearly

38

Annals of Carnegie Museum

VOL. 69

Fig. 7. — A. Portunites macrospinus, new species, holotype, CM 45833. B. Cast of Portunites incerta
Bell, JSHC 1732d. Scale bar = 1 cm.

place it in Portunites, including a well-developed epibranchial spine, a relatively
straight front with very small teeth, four anterolateral spines excluding the outer
orbital spine, two orbital fissures, an ovate-hexagonal carapace, and carapace re-
gions typical of species of Portunites. It is therefore suggested that the diagnosis
of the genus be expanded to include a statement pertaining to the relatively great
width of the orbits in members of this genus, occupying more than one-half and
generally two-thirds or greater the maximum carapace width. The size of the orbits
cannot be determined for P. eocenica Lorenthey and Beurlen, for which the type
specimen has been lost (Quayle and Collins, 1981).

Portunites triangulum Rathbun differs from members of the genus Portunites
in several important aspects. Portunites triangulum possesses rather small orbits;
they occupy approximately 0.52 of the maximum width of the carapace, which
does fall within the range for the genus. Schweitzer and Feldmann (1999) listed
several differences between Portunites triangulum and other members of the ge-
nus, including overall carapace dimensions, the relative length of the anterolateral
spines, definition of the carapace regions, and carapace shape. At that time those
differences were considered to be specific; however, additional observations in-
dicate that there are further differences between Portunites triangulum and other
members of the genus. Portunites triangulum possesses four sharp spines on the

2000

Schweitzer et al. — Eocene Decapod Crustaceans

39

front in addition to small inner-orbital spines; species of Portunites have only
four spines or bosses including the inner-orbital spine. The epibranchial ridge on
Portunites triangulum is not as well developed as on the other species of Por-
tunites and it lacks the transverse ridges on the hepatic and protogastric regions
typical of Portunites. Because of these important differences between this species
and other species of Portunites, reevaluation of its generic assignment is neces-
sary. Reassignment at this time is premature, but the problem currently is being
addressed by two of the authors (Feldmann and Schweitzer) and other workers
(Karasawa, personal communication, 1998). Because of the strong probability that
Portunites triangulum will be removed from the genus, interpretations about the
genus presented in this paper therefore will exclude this taxon.

It is also suggested herein that the diagnosis of the genus be emended to de-
scribe the frontal margin as ranging from relatively straight to possessing four
blunt knobs or small spines. Portunites incerta, P. stintoni, P. sylviae, and P.
kattachiensis all possess four very small, blunt knobs or spines on the frontal
margin. The species P. alaskensis possesses four small but distinct and sharp
spines, and P. insculpta Rathbun appears to possess four small spines. The new
species P. macrospinus possesses two very small, blunt spines axially and blunt
swellings on the inner-orbital margin which are not developed as spines. Because
of this, it is necessary to state that the front in members of this genus may possess
very small spines or blunt knobs, since the current diagnosis (Glaessner, 1969)
indicates that the front must possess four spines. Portunites triangulum possesses
rather large, well-developed spines on the frontal margin, again suggesting that it
may be referrable to a different genus.

Portunites macrospinus, new species
(Fig. 7A, 8)

Diagnosis. — ^Carapace rounded-hexagonal in outline; postorbital spine extreme-
ly long, sharp, and attenuated; anterolateral spines sharp, last anterolateral spine
extremely long, attenuated and sharp; orbits very broad and deeply excavated;
carapace moderately vaulted; frontal spines very weakly developed.

Description. — Carapace rounded-hexagonal in outline; carapace length 64% maximum width; wid-
est at position of last anterolateral spine; position of maximum width situated approximately one-third
the total distance posteriorly; carapace moderately vaulted both transversely and longitudinally.

Front about 0.21 maximum carapace width; relatively straight; extending only slightly beyond orbits;
with blunt bilobed projection axially; axially sulcate; very poorly developed blunt projections on inner
orbital margin. Orbits large; fronto-orbital width about 0.67 maximum carapace width; each orbit
about 0.35 fronto-orbital width; orbits sinuous, curving posteriorly, then slightly anteriorly, and finally
posteriorly before intersecting anterolateral margin; directed anterolaterally; rimmed; two closed fis-
sures; deeply excavated laterally and shallowly excavated axially.

Anterolateral margins short, bearing four or five spines including extra-orbital spine; extra-orbital
spine extremely long, projecting anteriorly and well beyond the rostrum; acute and attenuated, some-
times with small triangular spine at base; next two anterolateral spines smaller, triangular, acute,
projecting anterolaterally; last anterolateral spine longest, attenuated, acute, projecting laterally.

Posterolateral margins weakly concave, sinuous, with pronounced, rimmed reentrant at posterolateral
comer. Posterior margin weakly concave; rimmed; about 0.31 maximum carapace width.

Carapace regions moderately well defined by grooves. Frontal region with two weakly inflated areas
on either side of the axis; protogastric regions weakly ovate, moderately inflated, with very weak
transverse ridge; hepatic regions with transverse ridge extending from second anterolateral spine across
region; mesogastric region extremely narrow anteriorly and widening posteriorly, straight-sided ante-
riorly, convex posteriorly, posterior margin nearly straight; meso- and metagastric regions not differ-
entiated; urogastric region weakly differentiated, narrow, depressed, with concave lateral margins;
cardiac region rounded-triangular in shape, apex directed posteriorly, with two large swollen bosses

40

Annals of Carnegie Museum

VOL. 69

Table 3. — Measurements (in mm) taken on the dorsal carapace of specimens o/ Portunites macros-
pinus. Orientation of measurements is illustrated in Figure H. LI = maximum length of carapace, W1
= maximum width of carapace, W2 — fronto-orbital width, W3 = frontal width, W4 = posterior width,

WJ = orbital width.

Specimen number

Ll

Wl

W2

W3

W5

W4

CM 45834

20.5

30.0

19.3

9.4

15.3

6.4

CM 45836

16.8

24.6

17.6

—

—

6.0

CM 45833

15.2

26.2

16.9

7.0

7.4

6.0

CM 45835

15.9

26.0

18.4

—

9.0

6.7

anteriorly placed on either side of the axis, very weakly developed boss at apex of triangle; intestinal
region flattened.

Branchial regions not well differentiated; well-developed epibranchial ridge, extending from last
anterolateral spine, curving first anteriorly, then posteriorly, terminating at lateral margin of mesogas-
tric region; branchial region inflated axially, sloping to posterolateral margin, flattening along posterior
margin.

Chelipeds poorly known, manus appearing to be rather short; venter unknown.

Measurements. — Measurements, in millimeters, taken on specimens of Portun-
ites macrospinus are found in Table 3. The orientation of measurements taken is
indicated in Figure 8.

Etymology. — The trivial name is a combination of “macro,” meaning long, with the word “spine,”
to indicate the extremely long postorbital and anterolateral spines which are diagnostic for the species.

Types. — The holotype, CM 45833, and paratypes, CM 45834-45836, are de=
posted in the Section of Invertebrate Paleontology, Carnegie Museum of Natural
History, Pittsburgh, Pennsylvania.

Remarks. — Four specimens from the Pulali Point locality can be referred to
this species. All four specimens exhibit most of the dorsal carapace, two of which
are quite well preserved. Portions of the cheliped and fragments of pereiopods
are present in all four specimens; however, none of these elements is well-enough
preserved to permit description. No aspects of the venter were preserved.

j- W2—

p- W5

W1

Fig. 8. — Line drawing of Portunites macrospinus showing position and orientation of measurements
taken.

2000

Schweitzer et al. — Eocene Decapod Crustaceans

41

The new species can be distinguished immediately from all other members of
the genus because of its extremely long, attenuated postorbital spine; sinuous,
deeply excavated orbits; and sharp, attenuated anterolateral spines. The new spe=
cies is most like the three species Portunites alaskensis, P. kattachiensis, and P.
hexagonalis Nagao, but differs from those species in several ways. Portunites
alaskensis has a well-developed ridge on the hepatic region that extends onto the
protogastric region; this ridge is weakly developed in P. macrospinus. The car-
apace of Portunites macrospinus is more ovate and less hexagonal than that of
P. alaskensis, and the frontal teeth, especially the axial pair, of P, alaskensis are
much better developed than in P. macrospinus. Portunites alaskensis possesses
three well-developed bosses on the cardiac region, while only two well-developed
bosses are present on that region in P. macrospinus. Furthermore, the carapace
in P. alaskensis is more flattened than that of P. macrospinus, which has a mod-
erately vaulted carapace. The spines on P. hexagonalis are shorter and more tri-
angular than those of P. macrospinus. The last anterolateral spine is directed
slightly posterolaterally in P. hexagonalis, while that of P. macrospinus is directed
laterally. The frontal spines of P. hexagonalis are much better developed than in
P. macrospinus.

Portunites macrospinus can be differentiated from P. kattachiensis because P.
macrospinus has wider, more widely spaced orbits, longer and more attenuated
anterolateral spines, and a less tightly arched epibranchial ridge than P. katta-
chiensis. Furthermore, the epibranchial ridge in P. kattachiensis terminates in a
tubercle, and there is a tubercle on the protogastric region and two tubercles on
the branchial region in P. kattachiensis; all of these are missing in P. macrospinus.
Additionally, P. kattachiensis has a much better developed urogastric region than
does P. macrospinus.

The description of this new species brings the number of species to nine that
are referrable to the genus Portunites. This occurrence brings the number of spe-
cies of Portunites in North America to three and the number of North Pacific
species to five. The five species from the North Pacific region (P. alaskensis, P.
hexagonalis, P. insculpta, P. kattachiensis, and P. macrospinus) are more similar
morphologically and therefore are probably more closely related to each other
than to the other species of the genus, which are known from the Eocene of
Europe. Species from the North Pacific range in age from Eocene to Oligocene.
Portunites macrospinus is currently known only from the Eocene Aldwell For-
mation and the Pulali Point localities, not extending the geologic range of the
genus.

Superfamily Xanthoidea Dana, 1851
Family Xanthidae Dana, 1851
Genus Pulalius, new genus

Type Species.— Zanthopsis vulgaris Rathbun, 1926.

Diagnosis. — ^Carapace ovate or hexagonal, highly vaulted longitudinally; car-
apace regions inflated; front four-lobed; orbits oval, with one orbital fissure; epi-
branchial region forming linear transverse ridge across dorsal carapace; antero-
lateral margin with three small, blunt spines, last spine longest, last spine at end
of transverse epibranchial ridge; branchial region large and inflated; posterolateral
margins convex; abdominal somites free.

42

Annals of Carnegie Museum

VOL. 69

Etymology. — The genus takes its name from Pulali Point, a prominent geographic feature near the
collecting locality for the new species, Pulalius dunhamorum, to be described below.

Remarks. — The collection of previously undescribed fossils from the Aldwell
Formation contains a group of specimens that are clearly congeneric with speci-
mens assigned to Zanthopsis vulgaris Rathbun and which have been assigned to
the new species described below. The new species and Z. vulgaris share many
significant features. The two species possess a four-lobed front which occupies
about 0.25 the maximum carapace width; oval orbits with one orbital fissure; an
anterolateral margin with three blunt spines with the last being the longest; broad,
inflated protogastric regions with well-defined lateral margins and poorly defined
axial margins; inflated carapace regions; a carapace that is highly vaulted longi-
tudinally; an inflated epigastric region that forms a transverse ridge; large, inflated
metabranchial regions; flattened intestinal regions; convex anterolateral and pos-
terolateral margins; a similar arrangement and shape of the mesogastric, urogas-
tric, and cardiac regions; an abdomen with somites 3-5 free; stemites of similar
size and shape; and major chelae with similar shape and ornamentation. Based
on these numerous similarities, it is apparent that the two species must be con-
generic.

Inspection of all known species of the genus Zanthopsis McCoy indicates that
Z. vulgaris and the new species differ significantly from the type species, Z. leachi
Desmarest, and cannot be retained in that genus. According to Glaessner’s (1969)
diagnosis for the genus, Zanthopsis sensu stricto and Zanthopsis vulgaris share
some characters, including a four-lobed front and a vaulted carapace; however,
species of Zanthopsis sensu stricto possess large, nodular bosses or tubercles on
the branchial regions of the carapace and exhibit fusion of abdominal somites 3-
5. Zanthopsis vulgaris does not possess either characteristic.

McCoy (1849) erected the genus Zanthopsis and designated Z. leachi as the
type species. He also described several other species of Zanthopsis, all of which
possess nodular bosses on the dorsal carapace. Bell (1858) illustrated the type
species and two of McCoy’s species, Z. bispinosa McCoy and Z. unispinosa
McCoy. Inspection of McCoy’s (1849) descriptions and Bell’s (1858) plates in-
dicates that Zanthopsis sensu stricto and Zanthopsis vulgaris are not at all closely
related. The carapace shape in Zanthopsis sensu stricto is rounded or oval; in
Zanthopsis vulgaris the carapace is rounded-hexagonal. The carapace regions of
Zanthopsis sensu stricto are indistinct and faintly marked by wide, shallow
grooves. Zanthopsis vulgaris has well-defined carapace regions that are marked
by narrow grooves. The carapace regions of the two taxa are completely different
in terms of shape, ornamentation, and degree of inflation. The protogastric region
of Zanthopsis sensu stricto is narrow and poorly delimited. In Z. vulgaris, those
regions are broad, inflated, and well delimited by grooves on the lateral margins.
The hepatic region is better developed, more inflated, and of an overall broader
shape than in Zanthopsis sensu stricto. Zanthopsis vulgaris has one orbital fissure;
members of Zanthopsis sensu stricto have none. The epibranchial region of Z.
vulgaris is swollen and forms a transverse ridge; in Zanthopsis sensu stricto, this
region does not appear to be differentiated at all. The branchial region of Zan-
thopsis sensu stricto is ornamented by large tubercles or bosses; there are no
bosses on the branchial region of Z. vulgaris, which also has an overall larger
and more inflated branchial region. The shape of the urogastric and cardiac regions
are completely different. The urogastric region in Z. vulgaris is extremely con-

2000

Schweitzer et al. — Eocene Decapod Crustaceans

43

stricted and flattened and the cardiac region is broadly triangular; both regions
are very well defined. In Zanthopsis sensu stricto, these regions are not well
defined and possess tubercles or bosses. The male abdomen of Zanthopsis sensu
stricto exhibits fusion of the third through fifth somites, while in Z. vulgaris, these
somites are free. For all of these reasons, Zanthopsis vulgaris is herein referred
to the new genus Pulalius and is designated as the type species for that genus.

Three other species from the Tertiary of the west coast of North America have
been referred to the genus Zanthopsis. Zanthopsis stembergi Rathbun, from the
Cretaceous of California (Rathbun, 1926), is known only from a claw, making
confirmation of its placement in the genus difficult. Zanthopsis hendersonianus
Rathbun from the Oligocene of Oregon and the Eocene of California (Rathbun,
1926) displays characteristics that place it within Zanthopsis sensu stricto. Zan-
thopsis rathbunae Kooser and Orr {not sensu Maury) resembles members of the
Hepatinae Stimpson of the Calappidae de Haan and is currently being reevaluated
by two of us (Schweitzer and Feldmann) to determine its generic and familial
placement.

Some other species referred to Zanthopsis sensu stricto have been removed to
other genera or families or should be reevaluated. Zanthopsis peytoni Stenzel and
Z. carolinensis Rathbun may be compared with members of the genus Harpac-
tocarcinus A. Milne Edwards. Zanthopsis terryi Rathbun and Zanthopsis rath-
bunae Maury were removed to the genus Eriosachila Blow and Manning by Blow
and Manning (1996; Blow, personal communication, 1998). Both Z. cretacea
Rathbun and Z. kressenbergensis Meyer possess a long lateral spine, which is not
characteristic of members of either Zanthopsis sensu stricto or Pulalius, suggest-
ing that their generic placement should be reconsidered.

Possession of a vaulted carapace with inflated regions; very small, blunt an-
terolateral teeth; a relatively simple arrangement of regions; and a distinctive
transverse epibranchial ridge set Pulalius apart from all other members of the
Xanthidae. Some other xanthid genera possess a more or less well-developed
transverse ridge, and these were assigned to the subfamily Menippinae Ortmann
by Sakai (1976). These genera include Ozius H. Milne Edwards, Pseudozius Dana,
Sphaerozius Stimpson, and Lydia Gistel. However, members of this subfamily
possess several characteristics that cannot accommodate Pulalius, including a car-
apace that is significantly wider than long or nearly circular instead of ovate-
hexagonal as seen in Pulalius, carapace regions that are much less inflated than
Pulalius, carapaces that are much less vaulted than Pulalius, more and larger
anterolateral teeth than Pulalius, a branchial ridge that ends in the penultimate
tooth instead of the last anterolateral tooth as in Pulalius, and better developed
anterolateral and frontal teeth than in Pulalius.

Other families contain genera with transverse ridges on the dorsal carapace, but
none can accommodate Pulalius. Members of the Portunidae typically possess a
transverse ridge on the dorsal carapace, but portunids usually have a flattened
dorsal carapace, a dentate front, and a dentate anterolateral margin with at least
three well-developed spines; these characteristics are not seen in Pulalius. Further,
in all but one subfamily of the Portunidae, the dactylus of the fifth pereiopod is
paddlelike. In the new species to be described below, the dactylus of the fifth
pereiopod is long and lanceolate.

The Retroplumidae Gill have transverse carapace ridges, but they are of an
altogether different placement and conformation. Retroplumids have extremely
wide orbits, a narrow front, and a flattened carapace, none of which are charac-

44

Annals of Carnegie Museum

VOL. 69

teristic of Puialius, Additionally, the carapace regions of Pulalius have an overall
different conformation than members of the Retroplumidae.

Pulalius vulgaris appears to be superficially similar to some members of the
Goneplacidae Dana. Goneplacids typically possess a quadrate carapace, a straight
front, and a flattened dorsal carapace (Williams, 1984; Tucker and Feldmann,
1990), which are not characteristic of Pulalius. However, distinguishing between
the Xanthidae and the Goneplacidae can be extremely difficult. Tucker et al.
(1994) described sternal characteristics that can sometimes be used to distinguish
between xanthids and goneplacids. In typical xanthids and some goneplacids, the
sternum is narrow and the eighth thoracic stemite is not visible because the ab-
domen obscures it. In other goneplacids, the sternum is wider and a small part of
the eighth stemite is therefore visible in ventral view (Tucker et al., 1994). The
goneplacid species Branchioplax washingtoniana Rathbun, also known from Ter-
tiary deposits of Washington, displays a conformation of the carapace regions that
is very similar to that seen in Pulalius vulgaris. The sternum of Branchioplax
washingtoniana is also xanthidlike because the eighth stemite is not visible in
ventral view, and therefore cannot be used to distinguish between the two taxa.
However, Pulalius vulgaris possesses a distinctly four-lobed front, inflated cara-
pace regions, a rounded carapace, and a highly vaulted carapace, distinguishing
it from Branchioplax washingtoniana, which has a straight front, flattened cara-
pace regions, and a carapace that is rounded-hexagonal in shape. Pulalius can be
differentiated from members of the goneplacid genus Speocarcinus Stimpson be-
cause the latter has a much less vaulted carapace, less inflated carapace regions,
and a more square carapace. Additionally, Speocarcinus exhibits a distinctly gone-
placidlike sternum in which the eighth stemite is clearly visible on either side of
the abdomen, while Pulalius exhibits a typically xanthidlike abdomen and ster-
num. In summary, use of the sternum to differentiate between members of the
Xanthidae and Goneplacidae must be used with extreme caution because the meth-
od does not work in all cases.

The new genus ranges from late Eocene to Oligocene rocks in British Colum-
bia, Washington, and Oregon (Rathbun, 1926; Berglund, personal records). Ac-
cording to Berglund, members of the genus have been found in several rock units
including the late Eocene Keasey Formation (Armentrout, 1987; Snavely, 1987),
the late Eocene to early Miocene Lincoln Formation (Rau, 1964; Prothero and
Armentrout, 1985; Armentrout, 1987; Babcock et al., 1994), the Oligocene Blak-
ely and Marrowstone formations (Armentrout and Berta, 1977), the Eocene to
Oligocene Makah Formation (Snavely et al., 1980), the Eocene ?Aldwell For-
mation at Pulali Point, Washington (Squires et al., 1992), and the Oligocene of
Nootka Island, British Columbia (Jeletzky, 1973).

Pulalius dunhamorum, new species
(Fig. 9^11)

Diagnosis. — Carapace ovate; anterolateral and posterolateral margins convex,
posterolateral margins especially so; regions inflated, especially the branchial re-
gions; epibranchial region with prominent inflated ridge extending from last an-
terolateral spine to axial area; anterolateral teeth small, blunt, rounded.

Description. — Carapace ovate in outline, wider than long, maximum width about 1.2 times length,
widest between last anterolateral spine, position of maximum width approximately 50% total length.
Carapace strongly vaulted longitudinally, moderately vaulted transversely, vaulting most extreme at
posterior part of mesogastric region.

2000

Schweitzer et al. — Eocene Decapod Crustaceans

45

Fig. 9. — Pulalius dunhamorum new species, holotype, CM 45837. A. Anterior view showing cheli-
peds. B. Oblique anterior view showing orbits. C. Dorsal carapace. Scale bar = 1 cm.

Front approximately 0.25 maximum carapace width, produced beyond orbital teeth, sinuous in
antero-oblique view, margin downturned; bearing four small, blunt teeth, outer teeth rounded, inner
teeth small, rounded, closely spaced, separated by narrow, V-shaped medial notch, base of notch
extending posteriorly into narrow sulcus. Orbits small, deep, slightly ovate in anterior view, directed
forward and slightly upward; eyestalks appear to fill orbits; supraorbital margins raised with one closed
fissure near midpoint.

Anterolateral margins with flanks turned under, very thick, sides nearly vertical at maximum width
of carapace, becoming less so toward anterior; convergent anteriorly. Anterolateral margin bearing

46

Annals of Carnegie Museum

VOL. 69

Fig. 10. — Pulalius dunhamorum new species. A. Dorsal carapace, CM 45840. B. Ventral surface of
carapace, CM 45844. C. Dorsal carapace, CM 45839. D. Anterior view showing orbits, CM 45839.
E. Venter and appendages, CM 45838. Scale bar = 1 cm.

2000

Schweitzer et al. — Eocene Decapod Crustaceans

47

I 1

Fig. 11. — Line drawing of Pulalius dunhamorum showing position and orientation of measurements
taken.

three blunt nodes; largest node at anterolateral angle located at termination of broad epibranchial ridge;
nodes 1 and 2 smaller than third; nodes 1 and 2 separated by small triangular notch; nodes 2 and 3
separated by broad, shallow notch; nodes separated from extra-orbital angle by sinuous area, sometimes
bearing one small, very blunt node. Posterolateral margins convex; posterior margin nearly straight,
approximately 0.50 maximum carapace width.

Regions well defined by grooves; carapace surface finely granular except in grooves. Epigastric
region small, weakly inflated; protogastric region longer than wide, separated from front by shallow
postorbital furrow, narrowing distally into triangular point; hepatic region inflated, possessing bulbous
swelling medially and flattening towards anterolateral margin; mesogastric region narrow anteriorly,
widening posteriorly into a triangular area, lateral margins concave, posterior margin convex; urogas-
tric region deeply depressed, narrowing distally, lateral margins very concave, highly restricted cen-
trally; cardiac region moderately inflated, defined laterally by deep branchiocardiac grooves, narrowing
distally, anterior portion of region most inflated and bearing two transverse pits; intestinal region poorly
defined, flattened, broad, extending to posterolateral corner.

Cervical groove deep, well defined, interrupted axially by urogastric region; broad and shallow
along anterior margin of epibranchial ridge, ending at base of second anterolateral node; branching at
distal end into two short grooves anteriorly and posteriorly around second anterolateral node; posterior
branch terminating at carapace margin between second and third anterolateral nodes; anterior branch
extending through notch separating first and second anterolateral node onto subhepatic region.

Epibranchial region marked by prominent, inflated, arcuate, transverse ridge; extending from third
anterolateral node anteriorly and then sharply posteriorly, terminating at lateral margin of urogastric
region in sharp triangle. Branchial region strongly inflated, pentagonal, accentuating broadly rounded
posterolateral margin of carapace.

Abdominal somites 1 and 2 approximately equal in width, somite 2 about twice as long as somite
1, lower margin of somite 2 convex, lower margin slightly sinuous; somite 3 widest of all somites,
about 0.3 maximum carapace width, lateral comers rounded, upper margin slightly sinuous, posterior
margin weakly concave; somites 4 and 5 about twice as wide as long; somite 4 slightly wider than
somite 5, somite 4 widening posteriorly, lateral margins of somites 4 and 5 weakly concave, upper
and lower margins of somite 4 and 5 nearly straight; somite 6 1.5 times as wide as long, margins
nearly straight; telson equilaterally triangular, apices rounded; weak axial ridge extending from somites
1-6.

Sternum slightly longer than wide, widest at stemite 6, ovate; stemites 1 and 2 form equilateral
triangle, surface weakly inflated, lateral margins slightly concave, apex sharp; stemite 3 with sinuous
upper margin, lateral margins rounded, posterior margin converging distally, broad sulcus axially;
stemite 4 with very concave anterior and posterior margins, with central and lateral depressions, blunt
epistemal projections; stemite 5 wider than stemite 4 and widest of all stemites, directed anterolat-
erally, upper and lower margins nearly straight, possessing sharp epistemal projections; stemite 6
widening distally, directed slightly anterolaterally, upper and lower margins straight, with sharp epi-
sternal projections; stemite 7 directed slightly posterolaterally, upper and lower margins nearly straight,
widening distally, with sharp epistemal projections; stemite 8 about as wide as long, appearing to
possess weak epistemal projections, not well known; surface of all stemites granular.

48

Annals of Carnegie Museum

VOL. 69

Table 4. — Measurements (in mm) taken on the dorsal carapace of specimens o/Pulalius dunhamomm.
Orientation of measurements is illustrated in Figure 11. LI = maximum length, W1 — maximum
width, W2 = fronto-orhital width, W3 = frontal width, W4 = posterior width.

Specimen number

wi

W2

W3

W4

Ll

CM 45837

36.5

16.0

9.2

17.0

29.7

CM 45845

36.0

17.6

—

15.6

30.3

CM 45839

32.6

14.8

8.2

14.1

27.1

CM 45844

26.0

12.5

7.6

12.6

21.6

CM 45843

36.1

17.0

9.4

8.6

30.0

CD 45838

46.8

—

—

17.3

38.8

CM 45842

31.7

14.8

8.6

14.0

28.3

CM 45840

27.9

13.0

7.0

12.8

23.2

CM 45841

36.7

17.7

9.3

13.2

31.1

Pterygostomial region triangular, incised by broad shallow groove extending from notch separating
first and second anterolateral nodes; exopod of third maxilliped much longer than wide, outer margins
nearly straight, widening slightly posteriorly, then converging to triangular point at posteriormost end;
ischium of endopod somewhat longer than wide, upper margin nearly straight, outer margin slightly
concave, inner margin weakly convex, lower margin weakly concave, narrow sulcus located about
one-third the distance distally from inner margin; basis triangular, widening anteriorly.

Chelipeds unequal, both robust. Carpus of major cheliped longer than high, widening distally, distal
margin nearly straight, upper margin very convex, lower margin concave, tapering to very narrow
proximal margin, bulbous. Merus of major cheliped longer than high, lower margin weakly convex,
appearing to be rather bulbous. Manus of major cheliped widening distally; lower margin extending
nearly straight, then extending obliquely at propodus; outer surface highly vaulted transversely and
longitudinally; upper margin weakly rounded, small blunt tooth at proximal corner; proximal margin
sinuous, with broad inflated area centrally and blunt tooth at intersection with lower margin; distal
margin sinuous, with trapezoidal projection centrally, distal margin extending obliquely between upper
and lower margins.

Fixed finger of major cheliped slightly deflected; occlusal surface with four large, blunt teeth, first
and third approximately equal in size, second largest and broadly triangular, fourth smallest, remainder
of occlusal surface with small denticles; widest proximally and narrowing distally, tip upturned slight-
ly, outer surface with setal pits arranged linearly below denticles, denticles also with setal pits.

Movable finger of major cheliped highly arched; narrowing distally; occlusal surface bearing at
least three teeth, first tooth wide, blunt; second and third bluntly triangular, second larger than third,
remainder of occlusal surface smooth or with small denticles. Each finger more darkly pigmented than
remainder of carapace.

Minor chela similar in shape to major chela except not as bulbous and with more slender fingers;
finger more weakly arched.

Basis of pereiopod 2 tubular; higher than long; merus much longer than high; carpus appearing to
be longer than high. Basis of pereiopod 3 tubular, higher than long, distal margin concave; merus
much longer than high, surface granular; ischium slightly longer than high; carpus longer than high,
widening distally; manus longer than high. Basis of pereiopod 4 tubular, higher than long, distal margin
concave; ischium not well known; merus much longer than high; carpus longer than high, widening
distally. Basis and ischium of pereiopod 5 not known; merus much longer than high; carpus somewhat
longer than high; propodus longer than high; dactylus much longer than high, lanceolate, tapering to
sharp point distally.

Measurements. — Measurements of the dorsal carapace are given in Table 4,
and orientation of measurements is illustrated on Figure 1 1 . Measurements of the
venter are given in Table 5, and measurements of the major and minor chelae are
given in Table 6.

Types. — The holotype, CM 45837, and paratypes, CM 45838-45846, are de-
posited in the Section of Invertebrate Paleontology, Carnegie Museum of Natural
History, Pittsburgh, Pennsylvania.

Etymology. — The specific name honors Mr. and Mrs. George Dunham of Brinnon, Washington, who

2000

Schweitzer et al. — Eocene Decapod Crustaceans

49

Table 5. — Measurements (in mm) taken on the sternum of specimens o/ Pulalius dunhamorum. L =

length, W - width.

Specimen number

Max. L
of sternum

Max. W
of sternum

Max. W
of abdomen

Max. L
of abdomen

W-telson

CM 45844

15.2

13.7

—

—

—

CM 45840

>14.2

14.5

8.7

9.1

5.0

CM 45838

27.8

22.8

14.2

13.1

8.1

kindly allowed collecting on their property and access to the cobble beaches of Pulali Point, Wash-
ington.

Remarks. — Ten specimens are herein referred to this species. The specimens
are extremely well preserved, with some retaining cuticular material. At least three
of the specimens appear to be corpses, based upon their possession of the venter,
but others may also be corpses based upon retention of the chelae. All of the
specimens were preserved in concretions. The nine specimens of Pulalius dun-
hamorum exhibit a range of variation in some characteristics. For example, some
specimens, CM 45838 and CM 45842, possess a somewhat more hexagonal car-
apace than the other specimens. Additionally, the development of the anterolateral
teeth ranges from weak to extremely reduced or almost nonexistent.

Pulalius dunhamorum can be differentiated from the only other species in the
genus, P. vulgaris, in several ways. Pulalius dunhamorum possesses a more linear,
well-developed epibranchial ridge while that of P. vulgaris is arcuate and discon-
tinuous. The posterolateral margins of P. dunhamorum are shorter and much more
convex and the anterolateral margins are also more convex than those of P. vul-
garis. The carapace of P. dunhamorum is not as highly vaulted longitudinally and
the anterolateral spines are not as well developed as in P. vulgaris. The urogastric
region is more depressed in P. dunhamorum than in P. vulgaris. The overall shape
of P. dunhamorum is rounder and more ovate than that of P. vulgaris, which has
a more hexagonal carapace.

The new species has been reported only from the Pulali Point locality.

Family Carpiliidae Ortmann, 1893

Remarks.- — -Traditionally, Carpilius Leach, 1823, has been referred to the Xan-
thidae (Rathbun, 1930; Glaessner, 1969; Sakai, 1976; Manning and Holthuis,
1981; Dai and Yang, 1991). However, Guinot (1978) presented compelling ar-
guments for separation of Carpilius and the related genera, Palaeocarpilius A.
Milne Edwards, 1862, and Ocalina Rathbun, 1929, into a separate family, the
Carpiliidae. We concur. In that same year, Collins and Morris (1978) also referred
to the Carpiliidae and assigned two new species from Pakistan to their new genus

Table 6. — Measurements ( in mm) taken on the major and minor chelae of Pulalius dunhamorum. L

= length.

Specimen

Height

Length

L-Movable finger

L-Palm

CM 45838

19.0

40.4

26.2

29.2

CM 45837

>11.0

>30.0

19.7

23.1

CM 45845

12.3

27.7

16.8

20.8

CM 45838 (minor)

13.0

32.4

19.0

22.6

CM 45837 (minor)

10.4

25.7

15.4

17.7

50

Annals of Carnegie Museum

VOL. 69

Proxicarpilius Collins and Morris, 1978. To these four genera must be added
Eocarpilius Blow and Manning, 1996 and Harpactoxanthopsis Via, 1959.

The characters of the dorsal carapace which unite these genera include a gen-
erally oval outline which is wider than long. The front is downturned and is either
simple and triangular or undulatory, with a pair of inner orbital projections and
a pair of generally smaller projections on either side of a shallow axial depression.
The orbits are small, entire, and circular. The anterolateral margin may be smooth
or may bear one or more blunt projections. The anterolateral comer is unmodified
in some and bears a blunt projection extending onto the carapace as a subtle ridge
on others. The carapace surface is generally smooth, the carapace regions are
weakly defined or not discemable, and the carapace is strongly vaulted longitu-
dinally. The characters of the appendages, sternum, and other features that define
the family are given by Guinot (1978).

Genus Carpilius Leach in Desmarest, 1823

Type Species. — Cancer maculatus Linnaeus, 1758.

Remarks. — The Carpiliidae includes four genera known only from the fossil
record and a single genus, Carpilius, that ranges into the Recent. The genera are
distinguished from one another on the basis of the configuration of the front and
oil the nature of the anterolateral margin. In these regards, the specimens from
Pulali Point must be assigned to Carpilius. The front bears four projections that
extend forward and downward, very much like those of the type species, C.
maculatus. The anterolateral margin is generally unmodified, although there is a
very subtle angulation just in advance of the projection defining the anterolateral
comer. The form of the anterolateral comer and the ridge projecting onto the
carapace surface is nearly identical to those developed on extant species.

Recognition of the Pulali Point specimens as representatives of Carpilius pro-
vides the first authentic documentation of that genus in the fossil record. Although
specimens from the Miocene of central Europe have previously been assigned to
Carpilius (Muller, 1984), that species, C. antiquus Glaessner, 1928, has subse-
quently been assigned to Eocarpilius (Feldmann et al., 1998). Karasawa (1993)
reported Carpilius from Miocene rocks of southwestern Japan, but that occurrence
was based only upon a portion of a manus. Thus, the geologic range of Carpilius
has been extended from Eocene to Recent.

Carpilius occidentalis, new species
(Eig. 12, 13)

Diagnosis. — Typical-sized Carpilius with subtle swelling on anterolateral mar-
gin just in advance of well-defined prominence on anterolateral comer; straight
posterolateral margin as long as anterolateral margin. Gastric region weakly de-
fined, branchiocardiac groove well defined for genus.

Description. — Moderate-sized carpiliid; carapace width approximately 1.4 times length, ovoid out-
line, flattened transversely, strongly arched longitudinally, regions weakly defined.

Front about 0.24 maximum width, downturned, bilobed axially, bounded by rounded inner orbital
projections. Orbits smoothly rounded, complete, with subtle orbital rim. Fronto-orbital margin about
equal in width to posterior margin, approximately 0.43 maximum width. Anterolateral margin smooth,
with subtle rim, curvature increasing toward anterolateral comer which is marked by a swollen prom-
inence extending onto carapace as a faint ridge. Less distinct swelling developed on anterolateral
margin in advance of anterolateral corner. Maximum width developed at anterolateral comer situated

Fig. 12. — Carpilius occidentalis new species. A. Anterior view, holotype, CM 45847. B. Dorsal car-
apace, holotype, CM 45847. C. Dorsal carapace, CM 45848. Scale bar 1 for A and B and scale bar 2
for C. Scale bars = 1 cm.

2000

Schweitzer et al. — Eocene Decapod Crustaceans

52

Annals of Carnegie Museum

VOL. 69

-W2 ]

Fig. 13. — Line drawing of Carpilius occidentalis showing position and orientation of measurements
taken.

at point about 0.66 maximum length. Posterolateral margin straight, intercepting straight posterior
margin at about 140° angle. Posterior margin with narrow, distinct rim.

Carapace regions poorly defined as broadly domed surfaces and shallow grooves. Gastric region
longer than wide, bounded by very shallow parabolic groove. Urogastric region about 0.13 maximum
width, well defined by narrow, arcuate grooves laterally. Cardiac region about 0.24 maximum width,
widest near anterior, tapering to indistinct intestinal region. Hepatic region separated from smooth
branchial region by depression paralleling anterolateral margin.

Carapace surface with fine setal pits, more numerous and larger along anterior and anterolateral
margins.

Measurements. — Measurements taken on specimens of Carpilius occidentalis
are given in Table 7, and the orientation of those measurements is shown in Figure
13.

Types. — The holotype, CM 45847, and paratype, CM 45848, are deposited in
the Section of Invertebrate Paleontology, Carnegie Museum of Natural History,
Pittsburgh, Pennsylvania.

Etymology. — The species takes its trivial name from its occurrence on the eastern Pacific rim.

Remarks. — Two specimens from the Pulali Point region can be referred to this
species. Both are extremely well preserved; however, the front is broken away on
the paratype. Fragments of the proximal elements of the pereiopods are preserved
as is a portion of the right cheliped on the holotype. Unfortunately, preservation
of this element is not sufficient to permit a good description. The ventral surface
of this crab remains unknown.

Comparison with extant species within the genus permits distinguishing the
Eocene form as a new species. The anterolateral comer of C. occidentalis is

Table 7. — Measurements (in mm) taken on the dorsal carapace of specimens o/ Carpilius occidentalis.
Position and orientation of measurements is illustrated in Figure 13. LI = maximum length, L2 =
length to maximum width, W1 = maximum width, W2 = fronto-orhital width, W3 = frontal width,

W4 — posterior width.

Specimen number

Ll

L2

Wl

W2

W3

W4

CM 45847

31.5

20.5

44.2

18.0

10.7

19.0

CM 45848

36.0

21.0

55.7

—

—

22.0

2000

Schweitzer et al, — Eocene Decapod Crustaceans

53

situated at about the midlength, and the posterolateral margins converge posteri=
orly to form an angle of about 98°. In all of the living species, the anterolateral
comer is positioned in the posterior half of the carapace, and the posterolateral
margins converge at angles between 110-118°. The curvature of the anterolateral
margin of C. occidentalis is very much like that of C. convexus (Forskal); how-
ever, the point of inflection on the former species is a moderately well-defined
prominence whereas none is present on C. convexus. The front on C. occidentalis
bears four lobes, the axial pair being only slightly smaller than the inner orbital
lobes. The axial region of C. corallinus (Herbst) is not bilobed; that region on
the other two species is bilobed but the axial lobes are markedly smaller than the
inner orbital lobes. The overall carapace surface is very smooth on all living
species except C. convexus. On that species, the branchiocardiac groove is evident
as a subtle demarkation that can be discerned between the gastric regions and the
branchial regions. That same level of development is present on C. occidentalis.

In summary, the carapace morphology within the Carpiliidae has remained
relatively conservative since the appearance of the family in the Eocene. The
number of genera assigned to the family was greatest in the Eocene and has been
reduced, possibly to a single genus, Carpilius, in modem environments. Sakai
(1976) placed Liagore de Haan, 1833, with Carpilius in the Alliance Carpilioida.
Guinot (1978), however, rejected that association and Liagore will be retained in
the Xanthidae sensu lato until specimens can be examined in the light of the
familial characters designated by Guinot.

Family Goneplacidae Macleay, 1838
Genus Branchioplax Rathbun, 1916

Branchioplax washingtoniana Rathbun, 1916
(Fig. 14A)

Branchioplax washingtoniana Rathbun, 1916:344. Rathbun, 1926:42, plate 9, fig. 6; Tucker and Feld-
mann, 1990:415, fig. 6, 7.1, 7.3.

Material Examined. — CM 45849-45851, deposited in the Section of Invertebrate Paleontology, Car-
negie Museum of Natural History, Pittsburgh, Pennsylvania.

Remarks. — The Pulali Point specimens are referred to this taxon based on their
possession of a relatively square, equidimensional carapace; anterolateral teeth;
well-defined regions, inflated branchial regions; and a straight front. The speci-
mens conform well to the descriptions of the species provided by Rathbun (1916,
1926) and Tucker and Feldmann (1990).

Genus Neopilumnoplax Serene in Guinot, 1969
Neopilumnoplax hannibalanus (Rathbun, 1926)

(Fig. 14C)

Pilumnoplax hannibalanus Rathbun, 1926:37, 39, plate 10, fig. 1-4.

Neopilumnoplax hannibalanus (Rathbun): Tucker and Feldmann, 1990:418, fig. 7.2, 7.4, 8.

Material Examined. — CM 45855-45869, CM 45901, deposited in the Section of Invertebrate Pa-
leontology, Carnegie Museum of Natural History, Pittsburgh, Pennsylvania.

Remarks. — The Pulali Point specimens are referred to Neopilumnoplax hanni-
balanus based upon their possession of a straight front, flattened carapace, well-
developed anterolateral teeth, prominent midorbital suture, arcuate posterolateral
margins, and coalesced first and second anterolateral teeth. The specimens con-

54

Annals of Carnegie Museum

VOL. 69

form very well to both Rathbun’s (1926) description and the emended description
of the species provided by Tucker and Feldmann (1990).

Differentiating between Branchioplax washingtoniana and Neopilumnoplax
hannibalanus can be extremely difficult, as has been addressed by Tucker and
Feldmann (1990). They suggested that N. hannibalanus has more posteriorly con-
vergent posterolateral margins and more poorly marked regions than B. washing-
toniana. Additionally, B. washingtoniana possesses much more inflated branchial
regions than does N. hannibalanus, which provides the best means of differenti-
ating the two taxa. Neopilumnoplax hannibalanus appears to have sharper, better
developed anterolateral teeth than does B. washingtoniana, but this characteristic
does not always hold and should be used with caution.

As has been discussed above. Tucker et al. (1994) provided a means by which
some goneplacids could be distinguished from xanthids. Three of the goneplacid
specimens in this study, CM 45852-45854, possess well-preserved sterna, and on

Fig. 14. — K. Branchioplax washingtoniana Rathbun, CM 45850. B. Venter of unknown goneplacid,
CM 45852. C. Neopilumnoplax hannibalanus (Rathbun), CM 45855. Scale bar 1 for A and B and
Scale bar 2 for C. Scale bars = 1 cm.

2000

Schweitzer et al. — Eocene Decapod Crustaceans

55

these specimens, the eighth sternite is clearly obscured by the abdomen (Fig. 14B).
Unfortunately, none of the three specimens exhibits dorsal carapace material, mak-
ing them impossible to assign to a genus or species. However, the sterna corre-
spond very closely to those illustrated by Tucker and Feldmann (1990) as be-
longing to the two species Branchioplax washingtoniana and Neopilumnoplax
hannibalanus . Tucker and Feldmann (1990) suggested that the two taxa differ in
the conformation of the sixth abdominal somite which has convex sides in Bran-
chioplax washingtoniana and straight or concave sides in Neopilumnoplax han-
nibalanus. This character was not useful in assigning the three venters to a taxon,
because the lateral margins are intermediate between the two conditions.

Section Thoracotremata Guinot, 1977
Superfamily Hexapodoidea Miers, 1886
Family Hexapodidae Miers, 1886

Remarks. — The Hexapodidae was raised to family level by Manning and Hol-
thuis (1981) based upon the possession of three pairs of walking legs in this group
instead of four as in other decapod families. They considered the Hexapodidae
to be closely related to the Goneplacidae (Manning and Holthuis, 1981). In ad-
dition to loss of the fifth pereiopod, the eighth sternite is much reduced and is
hidden by the seventh sternite and the posterior margin of the dorsal carapace
(Gordon, 1971). Genera within the Hexapodidae are distinguished by the shape
and degree of fusion of the abdomen, form of the third maxillipeds, shape of the
eye, development of the sternal grooves of the abdomen, and the structure of the
male pleopod (Manning and Holthuis, 1981). The specimens from Pulali Point
are referred to the Hexapodidae because they possess four pairs of pereiopods
and have no evidence of an eighth sternite.

Extant genera of the Hexapodidae include Hexapinus Manning and Holthuis,
Hexaplax Doflein, Hexapus de Haan, Lambdophallus Alcock, Paeduma Rathbun,
Parahexapus Balss, Pseudohexapus Monod, Spiroplax Manning and Holthuis,
Stevea Manning and Holthuis, Thaumastoplax Miers, and Tritoplax Manning and
Holthuis. Genera described from the fossil record include Goniocypoda Wood-
ward, Thaumastoplax, Hexapus, Stevea, and Prepaeduma Morris and Collins. Pre-
paeduma possesses five pereiopods; Morris and Collins (1991) considered it to
be an ancestor to Paeduma in which the fifth pereiopod was not yet fully sup-
pressed. Beschin et al. (1994) doubted its placement in the Hexapodidae based
upon its possession of five pairs of pereiopods; reevaluation of that genus by two
of the authors (Schweitzer and Feldmann) is in progress.

Fossil hexapods have not previously been reported from North America. Fossil
species of Palaeopinnixa, with the referral of several species to that genus to be
discussed below, have been reported from Spain, Panama, Peru, Trinidad, and
Argentina, (Rathbun, 1918; Woods, 1922; Via, 1966; Collins and Morris, 1976;
Feldmann et al., 1995). Hexapus nakajimai Imaizumi has been described from
the Miocene of Japan (Imaizumi, 1959), and Hexapus pinfoldi Collins and Morris
has been reported from Pakistan (Collins and Morris, 1978). At least eight species
of the hexapod genus Goniocypoda have been reported from England, Europe,
Africa, and India (Woodward, 1867; Bittner, 1893; Carter, 1898; Glaessner, 1933;
Remy and Tessier, 1954; Secretan, 1971; Crane, 1981; Crane and Quayle, 1986).
The genus Stevea has been reported from the Eocene of Italy (Beschin et al.,
1994).

56

Annals of Carnegie Museum

VOL. 69

Genus Palaeopinnixa Via, 1966

Type Species. — Palaeopinnixa rathbunae new name for Pinnixa eocenica Rath-
bun, 1926, by original designation. Via (1966) designated Pinnixa eocenica Rath-
bun, 1926 as the type species of his new subgenus Pinnixa {Palaeopinnixa).
However, the referral herein of both Thaumastoplax eocenica Woods, 1922 and
Pinnixa eocenica Rathbun, 1926 to Palaeopinnixa makes Pinnixa eocenica Rath-
bun a junior homonym. Therefore, the substitute name, Palaeopinnixa rathbunae
is provided for Rathbun’s (1926) species, which remains the type species for the
genus as designated by Via (1966).

Included Species. — Palaeopinnixa rathbunae Rathbun, 1926 (formerly Pinnixa
eocenica); P. eocenica Woods, 1922 as Thaumastoplax; P. intermedia (Collins
and Morris, 1976), as Thaumastoplax; P. mytilicola Via, 1966; P. perornata Col-
lins and Morris, 1976 (In the title for the original species description, the trivial
name was erroneously spelled porornata. Derivation of the name as well as con-
sistent spelling as perornata throughout the text indicate that perornata was the
intended trivial name.); P. prima (Rathbun, 1918), as Thaumastoplax; P. rocaensis
(Feldmann et al., 1995), as Thaumastoplax.

Diagnosis. — Carapace wider than long, length to width radio about 0.67, car-
apace widest just anterior to posterolateral reentrants; carapace rounded rectan-
gular to ovoid, narrowing weakly anteriorly; carapace regions distinct; front wid-
ened distally, extending well beyond orbits, axially sulcate, frontal width to fronto-
orbital width ratio about 0.42; orbits wider than high, with sinuous upper margins,
moderately deeply excavated, fronto-orbital width to width ratio about 0.55; lat-
eral rim absent or weakly developed; posterolateral reentrant well-developed; pos-
terolateral width to maximum width ratio about 0.80; fronto-orbital width to pos-
terior width ratio about 0.55; abdominal somites 3-5 fused in males; fourth ster-
nite with anterior projections; third pereiopod longest.

Material Examined. — Thaumastoplax intermedia. In. 60008 (holotype); T. prima, USNM 324227
(holotype); USNM 324228 (paratype); T. eocenica Woods, SMC 1394 (holotype); Pinnixa (Palaeo-
pinnixa) perornata. In. 61361 (holotype); T. rocaensis, GHUNLPam 7006 (holotype); GHUNLP-
am 7007-7009, 7026, 7027.

Discussion. — Via (1966) erected the subgenus Palaeopinnixa to embrace the
new species Pinnixa {Palaeopinnixa) mytilicola as well as Pinnixa eocenica Rath-
bun which was designated as the type species for the subgenus. Newly collected
specimens referrable to Pinnixa eocenica sensu Rathbun, described herein, possess
four pairs of pereiopods and seven sternites; therefore, the subgenus Palaeopin-
nixa is removed from Pinnixa and the Pinnotheridae and is elevated to generic
status. Because of the homonymy discussed above, the type species for the genus
is Palaeopinnixa rathbunae new name.

Four fossil species previously assigned to Thaumastoplax, T. intermedia, T.
prima, T. eocenica Woods, and T. rocaensis, are clearly congeneric with Palaeo-
pinnixa rathbunae and Pinnixa {Palaeopinnixa) mytilicola and are therefore as-
signed to Palaeopinnixa. These fossil species possess all of the diagnostic char-
acters of Palaeopinnixa including a rounded to ovoid carapace, distinct carapace
regions, a front that is flared and axially sulcate, a well-developed posterolateral
reentrant, and a posterior width-to- width ratio of about 0.80.

The new specimens cannot be referred to any of the genera previously known
from the fossil record. Members of the genus Goniocypoda possess an extremely

2000

Schweitzer et al. — Eocene Decapod Crustaceans

57

wide fronto-orbital margin which occupies more than 0.60 the maximum carapace
width. Members of that genus also possess an extra-orbital tooth that can range
from small to large; the fossils described here do not possess an extra-orbital
tooth and have a fronto-orbital width to maximum width ratio of approximately
0.44. In the genus Hexapus, individuals possess deep sternal grooves, which the
fossils here referred to Palaeopinnixa do not possess. Members of the genus
Stevea have stridulating ridges on the pterygostomial region and exhibit fusion
of somites 2-6; neither condition is exhibited on the new specimens.

The new specimens can be easily differentiated from most extant genera of the
Hexapodidae. The sole extant species of Thaumastoplax has much smaller orbits
and rostrum, a rectangular carapace, a convex posterolateral margin, and poorly
defined carapace regions which clearly distinguish it from species of Palaeopin-
nixa. Members of Lambdophallus possess a well-developed sternal groove, which
species of Palaeopinnixa do not possess. In the genus Paeduma, individuals ex-
hibit fusion of somites 3 and 4 and somites 5 and 6, a pattern not seen in the
Pulali Point specimens. Members of the two genera Parahexapus and Pseudo-
hexapus possess a ridge that parallels the lateral margin of the carapace, a char-
acteristic not seen in the specimens here referred to Palaeopinnixa. In the genus
Tritoplax, the telson is distinctly trilobed; the telson in the specimens here referred
to Palaeopinnixa forms an equilateral triangle. The male abdomen in members
of the genus Spiroplax is very broad and exhibits somewhat convex lateral mar-
gins (Manning and Holthuis, 1981:177), while the male abdomen in the specimens
described herein is narrow and has concave lateral margins. The sole species of
Spiroplax has a much more rounded carapace than the new specimens. Addition-
ally, the carapace is equidimensional in Spiroplax, while in the new specimens it
is about 1.5 times as wide as long. Members of the two genera Hexapinus and
Hexaplax are distinguishable from members of Palaeopinnixa because Palaeo-
pinnixa has better developed carapace regions and a more rounded carapace.

Palaeopinnixa rathbunae new name
(Fig. 15-17)

Pinnixa eocenica Rathbun, 1926:34, plate 1, fig. 3, 4.

Pinnixa {Palaeopinnixa) eocenica Rathbun, 1926. Via, 1966:2, fig. 1.

Diagnosis. — Carapace subrectangular, with rounded anterolateral comers; bran-
chial regions moderately well defined; orbits well developed, with sinuous upper
margin; rostrum widening anteriorly.

Emendation to Description. — Carapace wider than long (LAV = 68.4%); surface finely granular,
granules better developed on posterior of carapace; lateral sides steep; carapace convex longitudinally
and weakly vaulted transversely; carapace regions weakly inflated; carapace grooves moderately to
weakly developed.

Frontal margin about 0.18 maximum carapace width. Orbits well developed, directed forward,
subrectangular in shape; upper margin sinuous, subtly rimmed. Rostrum about as wide as an orbit,
subrectangular, widening anteriorly, sulcate medially, anteriormost edge flattened, lateral margins form-
ing inner margins of orbits. Fronto-orbital width about 0.44 maximum carapace width.

Anterolateral margin rounded, continuous with lateral margin, weak ridge sometimes developed
along edge of anterolateral and lateral margins, lateral sides steeply rounded, well developed. Pos-
terolateral corners forming concave reentrants into carapace margin. Posterior margin about 0.84 max-
imum carapace width, very slightly sinuous.

Carapace grooves ranging from moderately defined to poorly defined. Cervical groove extending
in broad sinuous U-shape from anterolateral margin posteriorly to axial region, moderately well de-
veloped posteriorly and weakly developed anteriorly.

Protogastric region moderately inflated, two small swellings located at base of rostrum; hepatic

58

Annals of Carnegie Museum

VOL. 69

Fig. 15. — Palaeopinnixa rathhiinae new name, CM 45870. A. Dorsal carapace and appendages.
B. Anterior view showing orbits. Scale bar = 1 cm.

region bounded by weakly developed grooves, somewhat flattened; branchial region moderately in-
flated, surface uneven with localized bulbous swellings, two swellings located on either side of the
urogastric and cardiac regions, ornamented with granules, granules especially well developed poste-
riorly; metabranchial region flattened, not well ornamented, separated from branchial region by groove,
groove well defined laterally and disappearing axially, groove extending from lateral margin axially
and posteriorly.

Mesogastric region triangular, very narrow anteriorly, widening posteriorly, lateral margins concave,
posterior margin convex, bounded by poorly developed grooves, grooves best developed posteriorly,
two axial pits located on groove defining posterior margin. Urogastric region wider than long, con-
stricted axially, upper margin concave, lower margin weakly concave, lateral edges bounded by rather
deep, broad pits. Cardiac region subtriangular in shape, apex directed posteriorly; possessing three
broad, weakly developed granules arranged in a triangular pattern, apex directed posteriorly; region
weakly inflated.

Buccal frame rectangular; ischium of endopod of third maxilliped slightly longer than wide, nar-
rowing anteriorly, lateral margins straight, anterior margin slightly convex, posterior margin sinuous;
remainder of third maxilliped unknown. Subhepatic and sub-branchial regions finely granular; ptery-
gostomial region finely granular, arcuate ridge paralleling entire length of lower margin, short ridge
paralleling adaxial half of upper margin.

Sternum of male wider than long (LAV = 46.7%), semicircular in shape, lateral margins convex,
widest at about midlength, narrowing anteriorly and posteriorly. Sternites 1-4 fused; broadly trian-
gular; faint evidence of suture lines, suture between sternites 1 and 2 most distinct; surface finely
granular; projection on anterior portion directed anteriorly, appearing to be associated with stemite 2,
rounded and approximately equidimensional in shape; lower margin of stemite 4 with epistemal pro-
jection, first pereiopod associated with stemite 4.

Stemite 5 wider than long, widest of sternites, finely granular, with epistemal projections, length-
ening laterally, associated with base of pereiopod 2. Stemite 6 wider than long but not as wide as
stemite 5, lengthening laterally, with epistemal projections, finely granular, associated with base of

2000

Schweitzer et al. — Eocene Decapod Crustaceans

59

Fig. 16. — Palaeopinnixa rathbunae, new name. A. Venter of female, CM 45881. B. Venter of male,
CM 45876. Scale bar = 1 cm.

pereiopod 3. Stemite 7 wider than long but not as wide as sternite 6, surface granular, lengthening
laterally, associated with base of pereiopod 4. Sternite 8 and pereiopod 5 unknown.

Sternum of female wider than long (LAV = 45.3%); semicircular in shape, lateral margins convex;
widening posteriorly, widest at posterior margin of carapace. Sternites 1-4 fused; broadly triangular;
faint evidence of suture lines, suture between sternites 1 and 2 most distinct; surface finely granular;
two anterior projections, directed anterolaterally, associated with sternites 2 and 3, projection on ster-
nite 2 rounded and equidimensional, projection on stemite 3 longer and narrower; lower margin of
stemite 4 with episternal projection, pereiopod 1 associated with stemite 4.

Sternite 5 of female wider than long, finely granular, with episternal projections, lengthening lat-
erally, widest of all sternites, associated with base of pereiopod 2. Stemite 6 wider than long, almost
as wide as stemite 5, with episternal projections, surface finely granular, associated with base of
pereiopod 3. Stemite 7 wider than long but not as wide as sternite 6, surface finely granular, length-
ening laterally, associated with base of pereiopod 4. Stemite 8 and pereiopod 5 unknown.

Abdomen an isosceles triangle in females, 1.2 times as long as wide, about 0.36 maximum width
of sternum, straight-sided; telson isosceles triangular, apex appearing to be rounded, telson extending

60

Annals of Carnegie Museum

VOL. 69

i — W3 1

L

Fig. 17. — Line drawing of Palaeopinnixa rathhunae, new name, showing position and orientation of
measurements taken.

slightly beyond suture of sternites 1 and 2; somites 5 and 6 fused, widening posteriorly, lateral margins
slightly convex, lower margin convex, upper margin concave, about equidimensional when measured
at maximum dimensions, possessing broadly rounded medial ridge; somite 4 wider than long, upper
margin concave, lower margin convex, lateral margins slightly convex, widening posteriorly; somites
2 and 3 widest of somites; somite 3 much wider than long, upper margin slightly concave, lateral
margins nearly straight, lower margin sinuous, widening posteriorly; somite 2 much wider than long,
upper and lower margins sinuous, lateral margins weakly convex; somite 1 wider than long, narrowing
posteriorly, upper margin sinuous, lower margin sinuous, lateral margins convex, medial swelling and
weak lateral swellings; all somites except telson with rounded medial projection on lower margins,
also with broad medial swellings and marginal swellings.

Abdomen of male narrowly triangular, approximately 1.65 times as long as wide, about 0.27 max-
imum sternal width, surface of somites granular, abdomen with weakly concave sides, widening pos-
teriorly. Telson longer than wide, isosceles triangular in shape, apex rounded, extending beyond suture
of sternites 1 and 2; somite 6 trapezoidal in shape, longer than wide, upper and lower margins nearly
straight, lateral margins slightly concave; somites 3, 4, and 5 fused, fusion lines weakly developed,
trapezoidal in shape, upper margin straight, lateral margins sinuous, weak medial swelling posteriorly
and somewhat more well- developed lateral swellings posteriorly, lower margin weakly convex; somite
2 wider than long, upper margin weakly concave, lower margin weakly convex, lateral margins round-
ed, broad, gently swollen area medially; somite 1 poorly known.

Pereiopods ornamented with fine granules. Coxae of pereiopods 2-4 longer than wide, cylindrical.
Merus of pereiopod 4 much longer than high, triangular in crosssection, lateral margins slightly sin-
uous; carpus about as long as high, widening distally, convex upper margin, concave lower margin,
distal margin concave; propodus slightly longer than high. Merus of pereiopod 3 much longer than
high, triangular in cross section, lateral margins slightly sinuous; carpus about as long as high, wid-
ening distally, convex upper margin, concave lower margin, concave distal margin, very short proximal
margin. Merus of pereiopod 2 much longer than high, triangular in cross section, lateral margins
slightly sinuous; carpus about as long as high, widening distally, convex upper margin, concave lower
margin, distal margin concave. Chelipeds appearing to be unequal. Manus of major cheliped stout,
longer than high, widening distally, surface finely granular, with sharp spine on lower proximal corner,
spine extending posteriorly; carpus of cheliped about equidimensional, stout, bulbous; fingers arched,
widest proximally and becoming narrow distally, appearing to possess moderately large denticles on
occlusal surface. Manus of minor cheliped about equidimensional; fingers longer than manus, ap-
pearing to be finely granular, possessing moderately large denticles on occlusal surfaces; carpus equi-
dimensional, stout, bulbous.

Measurements. — See Table 8 for measurements (in millimeters) taken on the
dorsal carapace and Table 9 for measurements taken on the venter of specimens
of Palaeopinnixa rathbunae. Position and orientation of measurements taken on
the dorsal carapace are shown in Figure 17.

Material Examined. — CM 45870-45900, 45903, deposited in the Section of Invertebrate Paleon-
tology, Carnegie Museum of Natural History, Pittsburgh, Pennsylvania, are referrable to this species.

2000

Schweitzer et al. — Eocene Decapod Crustaceans

61

Table 8. — Measurements taken (in mm) on the dorsal carapace d/ Palaeopinnixa rathbunae. Position
and orientation of measurements taken are illustrated in Figure 17. LI = maximum length, W1 =
maximum width, W2 = fronto-orbital width, W3 = frontal width, W4 = posterior width.

Specimen number

Ll

Wl

W2

W3

W4

L2

CM 45897

11.1

16

9.3

3.6

13.3

10

CM 45899

>7.5

>12.2

—

—

10.2

>7.5

CM 45900

9.8

13.8

6

2.2

—

9

CM 45896

>10.3

15.9

—

—

—

>10.3

CM 45886

>9.0

14.3

—

—

—

>9.0

CM 45874

>11.0

16

—

—

13.9

>11.0

CM 45898

>11.0

15.7

—

—

—

11

CM 45872

8.5

12.7

5.7

—

9.9

8

CM 45888

—

16.4

7

3.2

—

—

CM 45891

>9.0

13.6

—

—

—

>9.0

CM 45870

12.2

16.9

7

3.3

13.8

>10.5

CM 45873

>9.8

>13.5

—

—

10.8

<9.8

CM 45885

8.6

12.9

6.2

1.8

10.4

8.3

CM 45895

>8.3

13.1

5.5

—

—

>8.3

CM 45887

>7.9

11.3

5.8

—

9.5

>7.7

CM 45883

10.3

16

6.7

1.9

12.3

9.7

CM 45884

>9.0

16

7.7

2.6

—

>7.7

CM 45889

10.6

15.9

7

2.5

—

9.7

CM 45871

11.1

16.5

7.5

3

13.6

10

CM 45892

>9.2

14.2

5.3

2.6

—

8.5

CM 45875

9.9

16

6.7

—

12.5

—

CM 45880

>10.0

16

7

—

13.6

>10.0

CM 45881

—

16.8

6.9

3.1

—

—

CM 45894

6.5

10

4.5

—

7.6

—

CM 45903

7.1

10.4

5

2.8

8

—

CSH 37

6.7

9.4

4.7

—

—

—

CSH 38

6.3

8.9

—

—

—

—

CSH 39

4

5.4

—

—

—

—

Two specimens, CM 45872 and CM 45873 were collected at the Pulali Point locality. Several of the
specimens, CM 45871, 45876, 45877, 45879-81, 45883, 45887-89, 45891-93, 45896-99, and 45900
were collected from the Eocene Hoko River Formation at the RB32 locality of Berglund, located in
the SWl/4, NWl/4, Sec. 4, T33N, R15W, Cape Flattery Quadrangle, Clallam County, Washington,
7.5 minute series, near Neah Bay. CM 45870, 45874, 45875, and 45878 were collected from the RB33
locality of Berglund, located in the Wl/2, Nl/4, Sec. 4, T33N, R15W, Cape Flattery Quadrangle, 15
minute series, near west Kydikabbit. CM 45884 and 45885 were collected from the RB31 locality of
Berglund located in the El/2, Sec. 4, T33N, R15W, Cape Flattery Quadrangle, Clallam County, Wash-
ington, in the intertidal zone west of Kydikabbit, Washington. CM 45882, 45894, and 45895 were
collected from an unknown locality. CM 45890 and 45886 were reported as having been collected
from the RB29 locality of Berglund; however, details are not available.

Table 9. — Measurements (in mm) taken on the venters o/ Palaeopinnixa rathbunae. W1 = maximum
width of sternum, LI = maximum length of sternum, W2 = maximum width of abdomen, L2 =

maximum length of abdomen.

Specimen number

Wl

Ll

W2

L2

Sex

CSH 27

14.9

6.7

Female?

CSH 29

13.5

5.8

3.6

5.4

Male

CSH 30

14.7

7.4

3.9

7.0

Male

CSH 31

18.1

8.0

6.4

7.5

Female

CSH 28

16.1

7.5

5.9

7.3

Female

62

Annals of Carnegie Museum

VOL, 69

Remarks. — Palaeopinnixa rathbunae may be distinguished from other members
of the genus in several ways. Palaeopinnixa eocenica possesses small orbits that
are directly slightly axially; the orbits of P. rathbunae are larger and are directed
forward. The front in P. eocenica is more narrow and less flared than that of P.
rathbunae. The carapace regions of P. eocenica are poorly marked; in P. rath-
bunae the carapace regions are weakly but noticeably defined. Palaeopinnixa
eocenica has a much more rounded carapace and more rounded lateral margins
than does P. rathbunae, which has a more rectangular carapace. Palaeopinnixa
intermedia Collins and Morris from the Miocene of Trinidad possesses a more
rectangular carapace, straighter lateral margins, and narrower mesogastric, uro-
gastric, and cardiac regions than does P. rathbunae. Palaeopinnixa prima Rathbun
from the Oligocene of Panama possesses less clearly defined carapace regions
than does P. rathbunae. Palaeopinnixa prima has a very sharp anterolateral mar-
gin and steep to slightly concave lateral sides, while P. rathbunae lacks a sharp
anterolateral margin and has more rounded, slightly convex lateral sides. The
metabranchial region in P. rathbunae is much more depressed and more clearly
defined than that of P. prima. The lateral margins of the carapace of Palaeopin-
nixa rocaensis Feldmann et al., from the earliest Paleocene of Argentina, are much
less rounded than those of P. rathbunae, which are markedly convex. Addition-
ally, P. rocaensis achieves its maximum width at approximately the midlength;
P. rathbunae reaches its maximum width about three-quarters the distance pos-
teriorly on the carapace.

Although all of the specimens may be assigned to Palaeopinnixa rathbunae,
there is a range of morphological variation in several aspects of the dorsal cara-
pace. Development of the carapace grooves varies among specimens. Those spec-
imens that appear to have well-developed grooves also are molds of the interior,
suggesting that this apparent difference in development may be related to manner
of preservation. The ridge paralleling the anterolateral and lateral margins is well
developed in some specimens such as CM 45875 and absent in others such as
CM 45870. This may be due to abrasion before or during burial, weathering of
the specimen at surface conditions, or variation within the population. In the case
of CM 45870, the carapace is badly weathered, perhaps accounting for the absence
of the ridge. Ornamentation of the dorsal carapace is also variable, again probably
due to both variation in the population and to weathering and abrasion of the
specimens. Most of the specimens appear to be highly weathered and retain no
ornamentation. Those that retain cuticle range from being finely granular to punc-
tate. CM 45872 is finely granular on the posterior portion of the carapace, and
CM 45870 possesses fine granules on the branchial region. While these ranges in
variation exist, there seems to be no pattern in the variation and the magnitude
of the variations is not sufficient to warrant removal of any of the specimens from
this taxon.

Many of the specimens possess asymmetrical bulbous swellings on the dorsal
carapace. In some specimens, the bulbous areas of the epibranchial regions are
more well developed than in others. These swellings could be attributed to several
factors. One is infestation by bopyrid isopods (Glaessner, 1969; Hessler, 1969;
Overstreet, 1983), which could explain some of the asymmetries. Other cases can
probably be explained by deformation or weak crushing of the carapace so as to
slightly deform it.

The overwhelming majority of specimens possess some portions of the ap-
pendages suggesting that most of the specimens are corpses and that the animals

2000

Schweitzer et al. — Eocene Decapod Crustaceans

63

were living near the site of deposition or were buried rapidly. The excellent pres-
ervation of some of the carapace material and appendages also supports this in-
terpretation. However, the deformation of some of the carapaces suggests at the
very least the material was crushed by compaction of the sediment.

Interestingly, there is a difference in the shape of the sterna of males and
females as well as the typical differences in the shape of the abdomina. The
sternum of the male is semicircular in shape and widest at the midlength while
that of the female is widest at the posterior margin of the carapace. The stemites
also appear to be wider in the females than in the males because more of the
sternites are visible in ventral view in females than in males. In both males and
females stemite 5 is widest and the sternum extends to the base of the buccal
frame.

Discussion

While an in-depth analysis of the biogeography of these decapods is premature,
some overall patterns are clear. Most of the genera herein reported from the Pulali
Point locality are well known and commonly reported from Tertiary rocks of the
west coast of North America, including Macroacaena, Laeviranina, Raninoides,
Portunites, Pulalius, Branchioplax, and Neopilumnoplax. Carpilius and Palaeo-
pinnixa are the only genera that have not previously been reported from the west
coast of North America.

Several members of the decapod fauna recovered from Pulali Point are known
only from the Northern Hemisphere. Members of the genus Macroacaena have
been reported from the Cretaceous and the Paleocene of Greenland and Japan,
and the Eocene and Oligocene of Oregon and Washington (Tucker, 1995). Por-
tunites has been reported from England, Europe, Japan, and the west coast of
North America (Schweitzer and Feldmann, 1999), and Pulalius is known only
from the west coast of North America. Branchioplax has been reported from
western North America, Japan, and Europe (Karasawa, 1992) with a (very) ques-
tionable occurrence from Senegal (Remy and Tessier, 1954). These genera may
have either a Tethyan or a north Polar distribution; Karasawa (1992) suggested
that both Portunites and Branchioplax have a Tethyan distribution.

The closely related genera Laeviranina and Raninoides each have widespread
distributions. Laeviranina is known only from fossil occurrences and has been
reported from Europe, South America, North America, New Zealand, and Pakistan
(Tucker, 1995). Raninoides, which ranges from the Eocene to the Recent, has a
cosmopolitan range having been reported from the Atlantic, Pacific, and Indian
oceans and the Caribbean and Central America. The oldest occurrence of Laevir-
anina is from the Paleocene of Alabama and Greenland, and the earliest report
of Raninoides is from Eocene rocks of Japan (Tucker, 1995).

Carpilius has been reported from the fossil record in Japan and from the Pulali
Point locality. However, the occurrence in Japan is based only upon a portion of
a manus (Karasawa, 1993), so its use in biogeographic analysis awaits more com-
plete material. The genus is currently known from the Indo-Pacific region and
from the western Atlantic and Caribbean. Neopilumnoplax is known from the
fossil record only on the west coast of North America and is currently found in
the Indo- Pacific region, a similar biogeographic pattern to that of Carpilius.

The genus Palaeopinnixa has previously been reported from Eocene rocks of
Panama, Peru, Trinidad, and the Paleocene of Argentina, and from Miocene rocks

64

Annals of Carnegie Museum

VOL. 69

of Spain (Rathbun, 1918; Woods, 1922; Rathbun, 1926; Via, 1966; Collins and
Morris, 1976; Feldmann et al., 1995). The first occurrence of the genus is in
Paleocene rocks of Argentina, and the genus appears to have subsequently dis^
persed northward into the western Pacific Ocean during the Eocene and into the
northeastern Atlantic Ocean by the Miocene.

Acknowledgments

The specimens that formed the basis of this study were collected by R. and M. Berglund, J. and
G. Goedert, G. Dunham, B. Buchanan, and A. Tucker. H. Schasse, of the Washington Division of
Geology and Earth Sciences, originally discovered the crab concretions at the Pulali Point, Washington,
locality and notified K. Kaler, who referred the material to R. Berglund. Mr. and Mrs. G. Dunham of
Brinnon, Washington, kindly allowed access to the Pulali Point locality via their private property. W.
Rau provided micropaleontological data. Type material was examined at The Natural History Museum,
London, with the assistance of A. Ross and at the Sedgwick Museum in Cambridge with the assistance
of M. Dorling and R. Long. M. Dorling of the Sedgwick Museum, Cambridge University; W. Blow,
United States National Museum, Washington, D.C.; W. Orr, Condon Museum of Geology, University
of Oregon; and J. Carter, Carnegie Museum of Natural History, Pittsburgh, Pennsylvania, all kindly
loaned comparative material for this study. R. B. Manning provided access to specimens at the United
States National Museum and offered numerous useful comments on the systematics of the taxa dis-
cussed in this paper. H. Karasawa, Mizunami Fossil Museum, Mizunami, Japan, and H. Kato, Natural
History Museum and Institute, Chiba, Japan, provided access to material at their respective institutions.
Travel for examination of Japanese fossil material and specimens deposited in museums in the United
Kingdom was funded by National Geographic Society Grant 6265-98 to Feldmann and Schweitzer.

Literature Cited

Armentrout, j. M. 1987. Cenozoic stratigraphy, unconformity-bounded sequences, and tectonic his-
tory of southwestern Washington. Pp. 291-320, in Selected Papers on the Geology of Washington
(J. E. Schuster, ed.). Washington Division of Geology and Earth Resources Bulletin 77.
Armentrout, J. M., and A. Berta. 1977. Eocene-Oligocene foraminiferal sequence from the north-
east Olympic Peninsula, Washington. Journal of Foraminiferal Research, 7:216-233.

Babcock, R. S., C. A. Suczek, and D. C. Engebretson. 1994. The Crescent “terrane,” Olympic
Peninsula and southern Vancouver Island. Washington Division of Geology and Earth Resources,
80:141-157.

Bell, T. 1858. A Monograph of the Fossil Malacostracous Crustacea of Great Britain, Part 1, Crus-
tacea of the London Clay. Palaeontological Society, London, United Kingdom.

Bennett, E. W. 1964. The marine fauna of New Zealand: Crustacea Brachyura. (New Zealand Ocean-
ographic Institute, Memoir 22.) New Zealand Department of Scientific and Industrial Research
Bulletin, 153:1-119.

Beschin, C., a. Busulini, A. De Angeli, and G. Tessier. 1994. I crostacei Eocenici della cava
“Boschetto” di nogarole Vicentino. Lavori, Societa Venziana de Scienze Natural!, 19:159-215.
Bittner, A. 1893. Decapoden des pannonischen Tertiars. Sitzungberichte der Kaiserlichen Akademie
der Wissenschaften Wien, Abteilung I, 102:10-37.

Blow, W. C., And R. B. Manning. 1996. Preliminary descriptions of 25 new decapod crustaceans
from the middle Eocene of the Carolinas, U.S.A. Tulane Studies in Geology and Paleontology,
29:1-25.

Carter, J. 1898. A contribution to the palaeontology of the decapod Crustacea of England. Quarterly
Journal of the Geological Society of London, 54:15-44.

Collins, J. S. H., and S. E Morris. 1976. Tertiary and Pleistocene crabs from Barbados and Trinidad.
Palaeontology, 19:107-131.

— . 1978. New lower Tertiary crabs from Pakistan. Palaeontology, 21:957-982.

Collins, J. S. H., and H. W. Rasmussen. 1992. Upper Cretaceous-Lower Tertiary decapod crusta-
ceans from west Greenland. Grpnlands Geologiske Unders gelse Bulletin, 162:1-46.

Crane, M. D. 1981. Hexapod crabs of the genus Goniocypoda H. Woodward from the upper Eocene
of Hampshire. Zoological Journal of the Linnean Society, 72:1-19.

Crane, M. D., and J. Quayle. 1986. Two new hexapod crabs of the genus Goniocypoda Woodward
(Crustacea, Decapoda) from the Hampshire Basin. Tertiary Research, 7:101-105.

Dai, a., and S. Yang. 1991. Crabs of the China Seas. China Ocean Press, Beijing, Peoples' Republic
of China.

2000

Schweitzer et al. — Eocene Decapod Crustaceans

65

Dana, J. D. 1851. Conspectus Crustaceorum quae in Orbis Terrarum Circumnavigatione, Carolo
Eilkes e Classe Riepublicae faederate Duce, Lexit et Descripsit. Proceedings of the Academy of
Natural Sciences of Philadelphia, 5:247-254.

DE Haan, W. 1833-1849. Crustacea. Pp. 109-164, in Fauna Japonica 4, XVII, XXXI (P. E de
Siebold). J. Muller, Amsterdam, The Netherlands.

Desmarest, a. G. 1823. Maiacostraces. Dictionnaire des sciences naturelles, 28:138-425.

Feldmann, R. M. 1989. Lyreidus alseanus Rathbun from the Paleogene of Washington and Oregon,
USA. Annals of Carnegie Museum, 58:61-70.

. 1991. Decapod Crustacea from the Tapui Glauconitic Sandstone (Bartonian: middle Eocene)

in the Waitaki Valley, South Island, New Zealand. New Zealand Journal of Geology and Geo-
physics, 34:17-22.

— . 1992. The genus Lyreidus de Haan, 1839 (Crustacea, Decapoda, Raninidae): Systematics

and biogeography. Journal of Paleontology, 66:943-957.

Feldmann, R. M., A. B. Tucker, and R. E. Berglund. 1991. Paleobathymetry of decapod crusta-
ceans, Washington. National Geographic Research and Exploration, 7:352-363.

Feldmann, R. M., S. Casadio, L. Chirino-Galvez, and M. Aguirre-Urreta. 1995. Fossil decapod
crustaceans from the Jaguel and Roca formations (Maastrichtian-Danian) of the Neuquen Basin,
Argentina. Paleontological Society Memoir 43.

Feldmann, R. M., K. L. Bice, C. S. Hopkins, E. W. Salva, and K. Pickford. 1998. Decapod
crustaceans from the Eocene Castle Hayne Limestone, North Carolina; Paleoceanographic impli-
cations. The Paleontological Society Memoir, 48:1-28.

Eorster, R., and R. Mundlos. 1982. Krebse aus dem Alttertiar von Helmstedt und Handorf (Nied-
ersachsen). Palaeontographica, Abteilung A, 179:148-184.

Glaessner, M. F. 1928. Die dekapodenfauna des Osterreichischen Jungtertiars. Jahrbuch Geologisch-
en Bundesanstalt Wien, 78:161-219.

— . 1929. Crustacea Decapoda. Pp. 1-464, in Fossilium Catalogus I: Animalia, Part 41 (J. E

Pompeckj, ed.). W. Junk, Berlin, Germany.

— — . 1933. New Tertiary crabs in the collection of the British Museum. Annals and Magazine of

Natural History, Series 10, 12:1-28.

— . 1960. The fossil decapod Crustacea of New Zealand and the evolution of the Order Deca-

poda. New Zealand Departent of Scientific and Industrial Research, New Zealand Geological
Survey, Palaeontological Bulletin, 31:5-79.

■ . 1969. Decapoda. Pp. R400-R566, in Treatise on Invertebrate Paleontology, Part R, Arthrop-

oda 4 (R. C. Moore, ed.). Geological Society of America and University of Kansas Press,
Lawrence, Kansas.

Glaessner, M. F, and T. H. Withers. 1931. On London Clay crabs of the family Raninidae. Annals
and Magazine of Natural History, London, 8:484-493.

Gordon, I. 1971. On the thoracic sternum in the subfamily Hexapodinae (Brachyura, Goneplacidae).
Crustaceana, 21:106-110.

Guinot, D. 1969. Recherches preliminaires sur les groupements naturels chez les crustaces decapodes
brachyoures. VIL Les Goneplacidae. Bulletin du Museum National d’Histoire Naturelle, 2 Serie,
41:507-528.

. 1977. Propositions pour une nouvelle classification des Crustaces Decapodes Brachyoures.

Compte Rendu Academie des Science de Paris, Serie D, 285:1049-1052.

—. 1978. Principes d’une classification evolutive des crustace decapodes brachyoures. Bulletin

Biologique de la France et de la Belgique, 112:209-292.

. 1993. Donnees nouvelles sur les Raninoidea de Haan, 1841 (Crustacea Decapoda Brachyura

Podotremata). Compte Rendu, Academie Science Paris, Sciences de la vie/Life Sciences, 316:
1324-1331.

Hessler, R. R. 1969. Peracarida. Pp. R360-R393, in Treatise on Invertebrate Paleontology, Part R,
Arthropoda 4 (R. C. Moore, ed.). Geological Society of America and University of Kansas Press,
Lawrence, Kansas.

Imaizumi, R. 1959. A fossil crab, Hexapus nakajimai n. sp. from Joban coal field. Recent Progress
of Natural Science (Japan), 30:25-30.

Jeletzky, j. a. 1973. Age and depositional environment of the lower part of the Escalante Formation,
western Vancouver Island, British Columbia (92-E): Mollusks, versus foraminifers. Canadian Jour-
nal of Earth Sciences, 10:331-365.

Karasawa, H. 1992. Fossil decapod crustaceans from the Manda Group (middle Eocene), Kyushu,
Japan. Transactions and Proceedings of the Palaeontological Society of Japan, New Series, 167:
1247-1258.

66

Annals of Carnegie Museum

VOL. 69

. 1993. Cenozoic decapod Crustacea from southwest Japan. Bulletin of the Mizunami Fossil

Museum, 20:1-92.

Latreille, P. a. 1802-1803. Histoire naturelle, generale et particuliere, des crustaces et des insectes,
Volume 3. F. Dufart, Paris, France.

. 1825. Histoire naturelle, generale et particuliere, des Crustaces et des Insectes. Ouvrage

faisant suite aux ouvres de Leclerc de Buffon, et partie du cours complet d’ Histoire naturelle
redige par C. S. Sonnini, membre de plusiers societes savants. Volume VI. F. Dufart, Paris, France.

Leach, W. E. 1823. Pp. 139-425, in Malacostraces: Dictionnaire des Sciences Naturelles, Volume
28 (A. G. Desmarest, ed.). F. G. Levreault, Paris, France.

Linnaeus, C. von. 1758. Systema naturae per regna tria naturae, secundum classes, ordines, genera,
species, cum characteribus, differentiis, synonymis, locis: edition 10, volume 1, Laurentii Salvii
(Homiae).

Lorenthey, E. 1897. Beitrage zur Decapodenfauna des ungarischen Tertian Termeszetrajzi-Fiizetek,
21:1-133.

Lorenthey, E., and K. Beurlen. 1929. Die fossilen decapoden der lander der Ungarischen Krone.
Geologica Hungarica, Series Palaeontologica, Fasciculus 3.

Macleay, W. S. 1838. Illustrations of the Annulosa of South Africa. On the brachyurous decapod
Crustacea brought from the Cape by Dr. Smith, Pp. 53-71, in Illustrations of the Zoology of
South Africa, Invertebrates (A. Smith, ed.). Smith, Elder, and Company, London, United Kingdom.

Manning, R. B., and L. B. Holthuis. 1981. West African brachyuran crabs (Crustacea: Decapoda).
Smithsonian Contributions to Zoology, 306:1-379.

McCoy, F. 1849. On the classification of some British fossil Crustacea, with notices of the new forms
in the University Collection at Cambridge. Annals and Magazine of Natural History, Second
Series, 4:161-179.

Miers, E. J. 1886. Report of the Brachyura collected by H. M. S. Challenger during the years 1873-
76. Pp. 1-362, in Report of the Scientific Results of the Voyage of H. M. S. Challenger during
the Years 1873-76 (C. W. Thomson and J. Murray, eds.). Zoology, Johnson Reprints, New York,
New York.

Milne Edwards, A. 1862-1865. Monographie des crustaces de la famille canceriens. Annals des
Sciences Naturelles, Zoologie, Series 4, 18(1862):3 1-85; 20( 1863):273-324; Series 5, 1(1864):
31-88; 3(1865):297-35L

Milne Edwards, H. 1837. Histoire naturelle des crustaces comprenant I’anatomie, la physiologie, et
la classification des animaux, 2:1-532. Paris, France.

Morris, S. E, and J. S. H. Collins. 1991. Neogene crabs from Brunei, Sabah, and Sarawak. Bulletin
of the British Museum of Natural History (Geology), 47:1-33.

Muller, P. 1984. Decapod Crustacea of the Badenian. Institutum Geologicicum Hungaricum, Geo-
logica Hungarica, Series Palaeontologica, Fasciculus 42.

Ortmann, a. E. 1893. Die decapoden-krebse des Strassburger Museums. 7. Theil. Abtheilung: Brach-
yura (Brachyura genuina Boas) 2. Unterabtheilung: Cancroidea, 2. Section: Cancrinea, 1. Gruppe:
Cyclometopa. Zoologischen Jahrbiicher, Abtheilung fiir Systematik, Geographic und Biologic der
Thiere, 7:411-495.

Overstreet, R. M. 1983. Metazoan symbionts of crustaceans. Pp. 156-250, in The Biology of Crus-
tacea, Volume 6, Pathobiology (A. J. Provenzano, Jr., ed.). Academic Press, New York, New
York.

Prothero, D. R., and j. M. Armentrout. 1985. Magnetostratigraphic correlation of the Lincoln
Creek Formation, Washington: Implications for the age of the Eocene/Oligocene Boundary. Ge-
ology, 13:208-211.

Quayle, W. j., and j. S. H. Collins. 1981. New Eocene crabs from the Hampshire Basin. Palaeon-
tology, 24:733-758.

Rafinesque, C. S. 1815. Analyse de la nature, ou tableau de I’Universe et des corps organisee.
Palermo, Italy.

Rathbun, M. j. 1916. Description of a new genus and species of fossil crab from Port Townsend,
Washington. American Journal of Science, 41:344-346.

. 1918. Contributions to the geology and paleontology of the Canal Zone, Panama, and geo-
logically related areas in Central America and the West Indies. Decapod crustaceans from the
Panama region. United States National Museum Bulletin 103:123-184.

. 1926. The fossil stalk-eyed Crustacea of the Pacific slope of North America. United States

National Museum Bulletin 138.

. 1929. A new crab from the Eocene of Florida. Proceedings of the United States National

Museum, 73:1-6.

2000

Schweitzer et al. — Eocene Decapod Crustaceans

67

. 1930. The cancroid crabs of America of the families Euryalidae, Portunidae, Atelecyclidae,

Cancridae, and Xanthidae. United State National Museum Bulletin 152.

. 1932. New species of fossil Raninidae from Oregon. Journal of the Washington Academy

of Science, 22:239-242.

Rau, W. W. 1964. Foraminifera from the Olympic Peninsula, Washington. Geological Survey Pro-
fessional Paper 374-G.

Remy, J. M., and F. Tessier. 1954. Decapodes nouveaux de la partie ouest du Senegal. Bulletin de
la Societe Geologique de France, Series 6, 4:185-191.

Sakai, T. 1976. Crabs of Japan and the Adjacent Seas. Kodansha Ltd., Tokyo, Japan.

Schweitzer, C. E., and R. M. Feldmann. 1999. Fossil decapod crustaceans of the late Oligocene to
early Miocene Pysht Formation and the late Eocene Quimper Sandstone, Olympic Peninsula,
Washington. Annals of Carnegie Museum, 68:215-273.

Secret an, S. 1971. La fauna de Crustaces decapodes des calcaires a Alveolines Ypresiens des Cor-
bieres septentrionales (Aude). Geobios, 4:117-141.

Snavely, P. D., Jr. 1987. Tertiary geologic framework, neotectonics, and petroleum potential of the
Oregon-Washington margin. Pp. 305-335, in Geology and Resources Potential of the Continental
Margin of Western North America and Adjacent Ocean Basins — Beaufort Sea to Baja California
(D. W. Scholl, A. Grantz, and J. G. Vedder, eds.). Circum-Pacific Council for Energy and Mineral
Resources, Earth Science Series Volume 6.

Snavely, R D., Jr., A. R. Niem, N. S. MacLeod, J. E. Pearl, and W W. Rau. 1980. Makah
Formation — A deep marginal basin sequence of late Eocene and Oligocene age in the north-
western Olympic Peninsula, Washington. Geological Survey Professional Paper 1162-B.

Squires, R. L., J. L. Goedert, and K. L. Kaler. 1992. Paleontology and stratigraphy of Eocene
rocks at Pulali Point, Jefferson County, eastern Olympic Peninsula, Washington. Washington Di-
vision of Geology and Earth Resources Report of Investigations 3 1 .

Tucker, A. B. 1995. A systematic evaluation of fossil Raninidae from the Twin River Group, Olympic
Peninsula, Washington, and a re-examination of the Raninidae. Unpublished Ph.D. Dissert., Kent
State University, Kent, Ohio.

— . 1998. Systematics of the Raninidae (Crustacea: Decapoda: Brachyura), with accounts of

three new genera and two new species. Proceedings of the Biological Society of Washington,
111:320-371.

Tucker, A. B., and R. M. Feldmann. 1990. Fossil decapod crustaceans from the lower Tertiary of
the Prince William Sound region. Gulf of Alaska. Journal of Paleontology, 64:409-427.

Tucker, A. B., R. E. Berglund, and R. M. Feldmann. 1994. Deep-water decapod crustaceans from
the Aldwell Formation (middle Eocene) at Pulali Point, Jefferson County, Washington. Abstracts
with Program of the Geological Society of America Annual Meeting, 7A-56.

Tucker, A. B., R. M. Feldmann, and C. L. Powell II. 1994. Speocarcinus berglundi n. sp. (Deca-
poda: Brachyura), a new crab from the Imperial Formation (late Miocene-late Pliocene) of south-
ern California. Journal of Paleontology, 68:800-807.

Via, L. 1959. Decapodos fosiles del Eoceno espanol. Boletin Instituto Geologico y Minero de Espafia,
70:333-395.

— — — . 1965. Raninidos fosiles del Eoceno. Contribution al estudio paleontologico de la familia
“Raninidae” (Crustaceos decapodos). Boletin Instituto Geologico y Minero de Espana, 76:233-
275.

. 1966. Pinnixa {Palaeopinnixa) mytilicola, nuevo braquiuro fosil, en el mioceno marino del

Valles (Barcelona). Acta Geologica Hispanica, 1:1-4.

— . 1969. Crustaceos decapodos del Eoceno Espanol. Pirineos, 91-94:1-479.

Williams, A. B. 1984, Shrimps, Lobsters, and Crabs of the Atlantic Coast. Smithsonian Institution
Press, Washington, D.C.

Woods, H. 1922. Crustacea from the Eocene deposits of Peru. Pp. 114-118, in Geology of the
Tertiary and Quaternary Periods in the North-west Part of Peru (T. O. Bosworth, ed.). Macmillan
and Co., London, United Kingdom.

Woodward, H. 1867. On a new genus of shore-crab, Goniocypoda edwardsi from the lower Eocene
of Hampshire. Geological Magazine, 4:529-531.

L <

a ' * *>xnyihn

'rt^ ■»’ » » .V

:^u ' ‘ ’ . ivh*/ /n«t /•*Ui»3^*>

i^.wk.iSr‘5(K%*- ' - ■,*«.■-■ ■JJ'

Vat. J .ItL <vO« ' •' h

•n*’ vV*

3'- Tils' • ; , . '. *>.-; ^o-

My*'* *' /•/

~ — ^ ^ BWjHI ' ’ ♦j(4*»’jiii<\..^r>*(v

’i4H;i*'t! . : "■‘■■^^^^ BMHI : »3|m ..Ti^Wtfr

/Vi '

■ ■» ••- - _ ...

’ ,, .in: ' .. QriM^nNiT'-^ -v...

i. ^ 1.4i.^

> wms-^^ I’H*-* ^ Vr* . • A

• *'••*>.. JKjiP

’teiit.jt Vjt^v'ViJi' .' r-*’ md0y

■ .^^irxSi

• ?* " ' *♦? ■ * ’ .•■ ^ ■ rf.'r,,:*r^'^

V. ■ «!»■- ‘^43^^819^ '" <’'■ ii „•• ■' j’ ’»!>■■•’.■■ i;, V- !• _1 ‘ -■ rt'illSy® ■

I** ' I ifatlllllil^^ * <]!(-'.>.- ;.V'V/n ! . <»; 4ltl^

'■'V'K!^^ U '■ - ' :. ...t». ; .-'-s - . /■.v,0'.;}.

A V* '‘'*' *' * *■ V •■■* i ■ -iV flrjfc-iliFm'. •

*lK-;t'r

H*V-.

, ,v^y

•r;Vi:

'< . ' ... '- ■-^P’ '-

/-VI

...-'i^ <!ll&(P;.*''^'''!''- Kuf' V'liT y^-»<VW«».?

n-:-ja^,''‘^>rtH; ,v ••injkir ;/ .v /,f, ‘-.

•' v-'fc'H-'- .--vft’-', -4^ -^ ”* 3* if-*-- '* ■■'■ -- ' r-^^^BI

■ ■•. •/ ffl.i5%,..'‘-,? * “t ' ^\u-).: ■■•. 4 , ^ /• ; laiaHiaji ;.

, > T/.;,

.. ftf".' '1^/

',v * ;

'‘^ip

t ru ■: rir

WK^: ... . ,1.-

. . ., . . . *4- - i ' •• i/ .. ^r

V ' ■ -

■ -4^:* '

,ir’-

INSTRUCTIONS FOR AUTHORS

ANNALS OF CARNEGIE MUSEUM consist of
contributions to the earth seiences (including paleontol-
ogy), life sciences, and anthropology, in 30 by 46 picas
format (127 by 195 mm or 5 by 7% inches). Submit all
manuscripts to the Offiee of Scientific Publications. Au-
thors should give particular attention to scientific con-
tent, format, and general style for the ANNALS. Manu-
scripts that do not conform to the style of the AN-
NALS will be returned to the author immediately.
Every manuscript will be subjected to external peer re-
view. Authors should submit the names, addresses,
phone and fax numbers, and e-mail addresses of at least
four qualified potential reviewers for each manuscript.
Authors will be asked to subsidize, if funds are available,
any or all costs of publication (approximately $ 1 00/page
printed). A detailed set of instructions to authors may be
requested from the Office of Scientific Publications.

Manuscript Style. — Articles should include the fol-
lowing items in this order: title page, abstract, text (with
desired headings), acknowledgments, literature cited, ta-
bles, figure captions, and copies of illustrations. All
manuscripts must be typed double-spaced on standard
814 by 11 -inch white bond paper, with at least one-inch
margins all around, and submitted in quadruplicate — an
original for the editors, and three review copies. All
pages should be numbered, including tables, literature
cited, and the list of figure captions. Only correspon-
dence quality or better dot matrix printouts will be ac-
cepted; draft mode is unacceptable. A monospaced font,
(e.g.. Courier) is required. A disk in WordPerfect x.x or
ASCII is helpful.

Title Page. The title should be brief, include the ani-
mal or plant group involved, be placed on a separate
page, and appear two to three inches below the top mar-
gin. Include the author’s name(s) and the affiliations of
non-CMNH authors. In the case of multiple authorship,
indicate the address to whieh proofs should be sent.

Abstracts. Abstracts should be short, but substantive,
and included at the head of the first page of text.

Text. Do not right justify text or break (hyphenate) a
word at the end of a line. Footnotes and acknowledg-
ments as footnotes are unacceptable. All text is double-
spaced in a monospaced font (e.g.. Courier).

Tables and Figure Legends. The list of figure legends
and each table should be typed consecutively on indi-
vidual pages separate from the text. Tables must be dou-
ble-spaced throughout with no vertical lines.

Measurements. Metric units should be used, if possi-
ble.

Literature Cited. All references to literature in text
must appear in the Literature Cited seetion. The data
(author, date, and page) in both citations must agree. Do
not abbreviate the titles of periodicals or serials. The
following style, double-spaced, should be used in Lit-
erature Cited:

1) Two authors in a journal series;

Soltis, D. E., and P. S. Soltis. 1992. The distri-
bution of selling rates in homosporous ferns.
American Journal of Botany, 79:97—100.

2) Same authors repeated — ^use three-em dash:

. 1923. The fauna of the Ardyn Obo For-
mation. American Museum of Natural History
Novitates, 98:1-5.

3) Same authors plus a third author — ^repeat all authors:

Knutson, L. V., R. E. Orth, and W. L. Murphy.
1986. Catalog of Sciomyzidae (Diptera) of
America north of Mexico. Entomography, 4;
1-53.

4) Chapter in an edited volume:

Rausch, R. L. 1963. A review of the distribution
of Holarctic mammals. Pp. 29^3, in Pacific
Basin Biogeography (J. L. Gressitt, ed.).
Bishop Museum Press, Honolulu, Hawaii.

5) Unpublished dissertation;

Smith, J. P. 1976. Review of Eocene Mammals.
Unpublished Ph.D. Dissert., University of
California, Berkeley, California.

6) Book:

White, M.J.D. 1961. The Chromosomes. Meth-
euen and Co., Ltd., London, United Kingdom.

7) Journal articles with usual volume and issue number:

Anderson, W. 1. 1969. Lower Mississippian con-
odonts from northern Iowa. Journal of Pale-
ontology, 43:916-928.

Figures. Including all illustrative materials (line art,
halftones, photographs), figures are to be numbered in
Arabic numerals. Four sets of figures are required, one
(original artwork) for reproduetion, three for reviewers.
Photocopies of photographs for reviewers are usually not
acceptable but are adequate for line drawing review cop-
ies. Figures may not be larger than 17 by 12 inches.
Reducing figures is the responsibility of the author. All
figures must be reducible to a maximum of 127 by 195
mm (30 by 46 picas) without loss of clarity. Line copy
should be designed for reduction to % or Vi or actual size.
Typewritten figure copy will not be accepted. Pho-
tographic figures should be submitted at actual repro-
duction size, if possible.

Rectangular halftone figures should be abutted, with-
out intervening spaces. The printer will insert narrov/
white spaces during the reproduction process. All figures
must have minimally one-inch borders all around. Each
figure should be given a protective cover and identified
on the back side.

Lettering and/or a magnification scale (linear metric
scale) for rectangular halftone figures should be placed
directly on the photo, not in a blank space between pho-
tos. The scale or lettering for closely cropped photos can
be placed in blank areas close to the figure.

Proof. — The author should answer all queried proof
marks and check the entire proof copy. Return corrected
page proof with the edited manuscript promptly to the
Office of Scientific Publications.

If an author chooses to make extensive alterations to
a paper in proof stage, the author will bear the cost.
Original manuscripts will not be returned unless request-
ed. Figures will be returned to the author only if re-
quested prior to publication.

ISSN 0097-4463

AS

3Co

?'7X

ANNALS

of CARNEGIE MUSEUM

THE CARNEGIE MUSEUM OF NATURAL HISTORY

4400 FORBES AVENUE • PITTSBURGH, PENNSYLVANIA 15213

VOLUME 69

23 MAY 2000

NUMBER 2

CONTENTS

ARTICLES

Rhysodine beetles (Insecta: Coleoptera: Carabidae): New species, new data. II

Ross T. Bell and Joyce R. Bell 69

The microtine rodents from the Pit locality in Porcupine Cave, Park County,

Colorado Christopher J. Bell and Anthony D. Barnosky 93

New Lower Mississippian trilobites from the Chouteau Group of Missouri . .

David K. Brezinski 135

■SWliTHSO?^^

Editors, ANNALS, BULLETIN and SPECIAL PUBLICATIONS:

D. S. Berman, Ph.D.

D. R. Watters, Ph.D.

J. R. Wible, Ph.D.

M. A. Schmidt, ELS, Managing Editor

Manuscripts, subscriptions, orders for individual numbers, and changes of address
should be sent to:

Office of Scientific Publications
Carnegie Museum of Natural History
4400 Forbes Avenue
Pittsburgh, PA 15213-4080
Phone (412) 622-3287
Fax (412) 622-8837
E-mail: scipubs@clpgh.org

ANNALS OF CARNEGIE MUSELFM is published quarterly by The Carnegie Museum of Natural
History, 4400 Forbes Avenue, Pittsburgh, Pennsylvania 15213-4080, by the authority of the Board of
Trustees of Carnegie Institute.

© 2000 Carnegie Institute.

@ This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).

ANNALS OF CARNEGIE MUSEUM

VoL. 69, Number 2, Pp. 69-91

23 May 2000

RHYSODINE BEETLES (INSECTA: COLEOPTERA: CARABIDAE):
NEW SPECIES, NEW DATA. II

Ross T Bell’

Research Associate, Section of Invertebrate Zoology

Joyce R. Bell’

Abstract

Seven new species of Rhysodini (Coleoptera: Carabidae) are described, Kaveinga (Kaveinga) waai
(Moluccas), Plesioglymmius {Juxtaglymmius) negara (Malay Peninsula), Omoglymmius {O.) emdomani
(New Guinea), Rhyzodiastes (Temoana) riedeli (New Guinea), Rhyzodiastes (Temoana) mindoro (Phil-
ippines), Clinidium (Clinidium) onorei (Ecuador), Clinidium (C.) gilloglyi (Panama). Descriptions are
given for either males or females of the following five species, previously known from only one sex;
Omoglymmius (O.) pulvinatus (Grouvelle), Clinidium (C.) howdenorum Bell and Bell, Clinidium (C.)
dormans Bell and Bell, Clinidium (C.) crater Bell and Bell, Clinidium (C.) spatulatum Bell and Bell.
An error in the description of Clinidium (C.) boroquense Bell is corrected. Range extensions or
clarifications are given for the following 20 species: Arrowina anguliceps (Arrow), Yamatosa nipo-
nensis (Lewis), Yamatosa draco (Bell), Yamatosa sinensis Bell and Bell, Plesioglymmius {Ameroglym-
mius) reichardti, Omoglymmius (Orthoglymmius) coomani (Grouvelle), Omoglymmius (O.) sakuraii
Nakane, Omoglymmius (O.) semperi Bell and Bell, Omoglymmius (O.) hiekei Bell and Bell, Omo-
glymmius {O.) bucculatus (Arrow), Omoglymmius (O.) patens Bell and Bell, Omoglymmius(0.) pul-
vinatus (Grouvelle), Omoglymmius (O.) sedlaceki Bell and Bell, Rhyzodiastes (Temoana) convergens
Bell and Bell, Clinidium (Arctoclinidium) rosenbergi Bell, Clinidium (C.) insigne Grouvelle, Clinidium
(C.) oberthueri Grouvelle, Clinidium (C.) rossi Bell, Clinidium (C.) moldenkei Bell and Bell, Clinidium
(C.) sulcigaster Bell.

Key Words; Rhysodini, Rhysodina, Clinidiina, Omoglymmiina, new species, distributions

Introduction

Rhysodini is a taxon of about 350 beetles which live inside dead wood. They
have been long interpreted as a separate family in Suborder Adephaga. It was
recently demonstrated (Bell, 1998) that they are highly modified ground beetles
(Carabidae) and belong in Subfamily Scaritinae. The sister group is Genus So-
lenogenys Westwood.

Adults are 4-10 mm long. All known species look red brown in bright light
and piceous to black when seen in dimmer light. The general shape is like the
smaller scaritine genera, especially Clivina Latreille but Rhysodini have many
distinctive features, including a deep median pit communicating with an inner
cavity in the head, moniliform antennae, and a condylelike neck. The pronotum
has deep pits or longitudinal grooves. Males have calcars (anteriorly directed
processes on middle and hind tibiae).

The mouthparts are highly unusual (Bell, 1994). The mentum covers the other
mouthparts in ventral view and extends so far forward as to prevent any objects
from coming between the mandibles. The leading edge of the mentum acts as a
cutting edge as the insect thrusts itself forward between layers of wood fibers.

' Biology Department, University of Vermont, Burlington, Vermont 05405-0086. rtbell@zoo.uvm.edu
Submitted 24 November 1997.

69

70

Annals of Carnegie Museum

VOL. 69

The mandibles serve only as covers for the other mouthparts. The galea and
laciniae of the maxillae form two pairs of stylets. The palpi are completely re-
tractile.

Rhysodine beetles are limited to moist forest areas where their food source is
thought to be the amoeboid stages of slime molds (Myxomycetes), About a third
of the species have vestigial wings. There is no evidence that the fully winged
species can fly very far. Nevertheless, rhysodines, including flightless groups, have
been very effective in colonizing islands.

This is a second paper in the pattern of Bell and Bell (1993), extending our
revision of the Rhysodini of the world. It contains descriptions of seven new
species, descriptions of sexes not previously described for five species, significant
locality data for additional species, and changes in keys to accommodate the new
species. An error in the description of Clinidium boroquense is rectified, while
that of Plesioglymmius (Juxtagiymmius) is altered to accommodate a second spe-
cies, and a modified key to the subgenera of Plesioglymmius is presented.

Abbreviations used in the text are: BMNH, Natural History Museum, London,
United Kingdom; CAS, California Academy of Sciences, San Francisco, Califor-
nia; CMNH, Carnegie Museum of Natural History, Pittsburgh, Pennsylvania;
CMNO, Canadian Museum of Nature, Ottawa, Ontario, Canada; CNCO, Canadian
National Collection, Ottawa, Ontario, Canada; CUIC, Cornell University, Ithaca,
New York; MNHB, Museum fur Naturkunde der Humboldt-Universitat zu Berlin,
Germany; NMW, Naturhistorisches Museum Wien, Vienna, Austria; SMNS, Staa-
tliches Museum fiir Naturkunde in Stuttgart, Germany; TAMU, Texas A & M
University, College Station, Texas; UVM, University of Vermont, Burlington,
Vermont. L/GW represents the ratio of pronotal length divided by its greatest
width.

Systematic Entomology
Subtribe Rhysodina

Kaveinga (Kaveinga) waai, new species
(Fig. 1A~C)

Type Specimens. — Holotype female, labelled “AMBON: Waai, Gg. Salahutu;
6.2: 300-600 m., Indonesia 1989; leg. Jach” (NMW). “Ambon” refers to a po-
litical division, and not to the island, Waai is at 3°33'S, 128°18'E. Paratypes: two
females, same data as holotype (NMW) (CMNH),

Etymology. — The specific epithet is derived from the name of the type locality.

Diagnosis. — This is the only species of Kaveinga sensu strict© which has sev-
eral precoxal setae, and which has the outer pronotal carina much narrower than
the inner one.

Description. — Length 6. 7-7.0 mm. Antennal segment I pollinose above; segments II-V each with
basal band of pollinosity; segments VII-XI with basal setae; segments VI-X with minor setae.

Head (Fig. lA) as long as wide; clypeus broadly separated from median lobe by band of pollinosity;
parafrontal bosses entirely pollinose; sides of median lobe deeply sinuate opposite temporal lobe;
orbital groove dilated, as long as eye; temporal lobes slightly wider than long, their anteriomedial
margins oblique, nearly straight, convergent posteriorly; medial angle rounded, scarcely overlapped
by median lobe; two or three temporal setae, anteriormost seta at or in contact with orbital groove;
pollinosity of postorbit well developed, with dorsal boundary opposite upper margin of eye; temporal
lobe without overhang; no suborbital tubercle or gular ridge.

Pronotum (Fig. lA) short; L/GW 1.05; widest near middle, sides curved, convergent to narrow
apex; sides oblique, curved in posterior half, shallowly sinuate near posterior angle; latter obtuse; edge

2000

Bell and Bell — New Rhysodine Beetles

71

Fig. 1. — Kaveinga (sensu stricto) waai, n. sp. A. Head, pronotum, dorsal aspect. B. Prothorax, left
ventrolateral aspect. C. Metasternum, abdomen, ventral aspect; female.

emarginate between hind angle and basal knob; latter rather prominent; paramedian grooves deep,
poliinose; marginal grooves very broad; outer carina very narrow, about one-third of inner carina,
much narrower than grooves on either side of it; anterior ends of inner carinae not abbreviated ante-
riorly by pollinosity; angular seta and about five marginals present; prosternum (Fig. IB) with strong
precoxal carinae, extended about 65% of distance to anterior margin; three to five precoxal setae; no
transverse groove between carinae.

Elytra moderately broad; elytral intervals narrow, convex, outer ones subcarinate; strial punctures
coarse; striae poliinose between punctures; posterior parts of outer striae deeply impressed; stria IV
joined to III posteriorly; V and VI joined posteriorly; VII apparently joined to V + VI posteriorly;
detached tip of VII marginal at apex of elytron; stria I with one seta at apex; stria II with one seta at
apex, in some specimens also with one at base; stria IV with several (average seven) setae along its
length, including one basal seta; detached apical part of stria VII with about seven setae. Abdominal
sterna (Fig. 1C) III and IV with transverse grooves widely separated in middle; V with space between
grooves occupied by a single line of punctures; VI with similar arrangement anteriorly, plus scattered,
coarse punctures posteriorly; lateral pits in female not enlarged; front and middle femora partly pol-
linose; tibia of middle leg apparently minutely serrulate on lateral margin in profile view. Male un-
known.

Distribution. — Known only from the type locality.

Remarks. — If a specimen were taken through our key to Kaveinga sensu stricto
(Bell and Bell, 1979), it would trace to K. abbreviata (Lea), but the narrow outer
carinae, the longer inner carinae, excavate marginal groove, and more numerous
setae would easily separate it. The short pronotum and extensive pollinosity sug-
gest that these two species are sister species. The key to Kaveinga sensu stricto
(Bell and Bell, 1979:399) can be amended as follows:

3. remove K. abbreviata; substitute 3.1'

3.1(3). Outer carina as wide as inner; marginal groove narrow; pronotum with no lateral, one

angular seta; parafrontal boss shining, glabrous . Kaveinga abbreviata (Lea)

3.1'. Outer carina less than half as wide as inner; marginal groove very wide, with angular and

about five marginal setae; parafrontal boss entirely poliinose Kaveinga waai, n. sp.

72

Annals of Carnegie Museum

VOL. 69

The presence of many pronotal and temporal setae would make this species
trace to subgenus Vakeinga Bell and Bell. The presence of serrulations on the
middle tibia, the absence of enlarged pits on the abdomen of the female, both
diagnostic of Kaveinga sensu stricto, as well as the characters linking K. waai
with K. abbreviata, suggest that the new species really belongs to the latter sub-
genus. A modified key to the subgenera of Kaveinga is therefore in order.

In the key to subgenera (Bell and Bell, 1979:390), substitute the following for
couplet 3:

3(2'). Middle tibia not apparently serrulate in profile view of lateral surface; female with enlarged
lateral pits on abdominal segments III and IV; prothorax without precoxal setae, but with

postcoxal setae Vakeinga

3'. Middle tibia apparently serrulate in profile view of lateral surface; female without enlarged
lateral pits on abdomen; prosternum with or without precoxal setae, but without postcoxal
setae Kaveinga sensu stricto

Subtribe Omoglymmiina
Yamatosa niponensis (Lewis, 1888)

Additional Locality. — TAIWAN: Taichung, Pilushi, 2200-m series of 16 spec-
imens, coll. Davidson, Young, Rawlins. May 22-23, 1988. (A second locality
from Taiwan.) (CMNH).

Yamatosa draco (Bell, 1977)

Additional Locality. — CHINA: Sichuan, Gongga Shan, Hailuogou above camp
3, 3050-3200 m., 7-VIL1996. J. Farkac, B. Kabatek, A. Smetana. 29°35'N,
102°00'E. (A. Smetana personal collection.) This is the first record for China;
previous records are from Bhutan and Pakistan. This is also the highest altitude
from which rhysodines have been collected.

Yamatosa sinensis Bell and Bell, 1987

Additional Locality. — CHINA: Yunnan, Heishui — 35-50 km N of Lijian,
27°I3'N, 100°19'E. Five males, 15 females. Coll. E. Jendak and O. Sausa (NMW).
Previously known only from Szechuan Province.

Plesioglymmius (Ameroglymmius) reichardti Bell and Bell, 1979

Additional Locality. — SURINAM: Saramacca, Kabo Agric. Sta., 3-7 VIII,
1980, M. I. Russell, under bark, dead tree. The first record from Surinam. Pre-
viously known from Venezuela (near the Orinoco River) and Brazil (Rio Madeira)
(BMNH).

The discovery of a new species related to the subgenus Juxtaglymmius, but not
conforming to the original description of that subgenus, necessitates a new defi-
nition of it, as well as a new key to the subgenera.

Subgenus Juxtaglymmius

Redescription. — Antennal segment XI longer than wide, with pointed apex (in some species the
point represents a very reduced conical stylet); antennal basal setae very sparse, beginning on segments
V or VII; segment I pollinose dorsally; segment II with basal pollinose band; segments III-XI without
pollinosity; antennal lobes close together, either separated by one-half width of basal condyle of
segment I or else in contact in midline, in form of median suture between them; clypeal setae absent;

2000

Bell and Bell — New Rhysodine Beetles

73

median lobe of head small, narrow, elongate, or oval, strongly convex; medial angles of temporal
lobes obtuse, narrowly separated at midline, margins oblique from medial angles to posterior angles;
orbital groove, temporal setae absent; temporal lobe in form of long overhang in lateral view, separated
by deep notch from suborbital tubercle; eye with posterior margin clearly anterior to middle of tem-
poral lobe; ventral surface of head with gular shelf, extended between the two suborbital tubercles;
pollinosity limited to margins of shelf; one pair of postlabial setae.

Paramedian groove limited to posterior two-thirds of pronotum, varied in form; base of elytron
without tooth opposite interval I; striae shallow, finely punctured, pollinosity limited to punctures;
apical impression of elytron small, limited to apices of striae I and II; apical tubercle not sinuate
medially; metasternum with lateral band of very coarse punctures; punctures in midline present or
absent; punctures of abdominal sterna III-V coarse, tending to form transverse line in each sternum;
sternum VI with row of very coarse punctures parallel to posterior margin, with scattered punctures
anterior to it; female known from only one species, with deepened lateral pits on sternum IV, and also
moderately deepened on sternum V. Male with ventral tooth on anterior femur.

Key to Subgenera

1. Head with prominent suborbital tubercle on each side; in ventral view, these tubercles

apparently prominent lateral angles on shelf extended across base of mentum; in lateral

view, eye extended far posterior to middle of temporal lobe 2

1'. Head without suborbital tubercle; eye anterior to middle of temporal lobe . . . Ameroglymmius
2(1). Antennal lobes either contiguous in midline, or else narrowly separated by less than half

width of median lobe; base of elytral interval not in form of tooth; temporal lobe in lateral

view with strong overhang posteriorly Juxtaglymmius

2' . Antennal lobes in dorsal view separated by at least width of median lobe; base of elytral
interval I in form of tooth; temporal lobe in lateral view not in form of overhang
Plesioglymmius sensu stricto

Key to Species oe Juxtaglymmius

1. Antennal lobes slightly separated, not extended to midline; basal impressions of pronotum

preceded by row of coarse punctures Plesioglymmius negara, n. sp.

1'. Antennal lobes in contact at midline; basal impressions of pronotum preceded by linear groove

Plesioglymmius jugatus Bell and Bell

Plesioglymmius {Juxtaglymmius) negara, new species
(Fig. 2A~D)

Type Specimen. — Holotype male, labelled “MALAYSIA:PAHANG, Taman
Negara N.R, 12-14, 7, 1993, leg. H. Forster” (NMW).

Etymology. — The specific epithet is derived from the name of the type locality.

Diagnosis. — The separation of the antennal lobes and the coarse punctures
forming the anterior end of the paramedian groove are diagnostic.

Description. — Length 4.9 mm. Antennal segments VII-X with basal setae; antennal lobes separated
from one another by about half width of basal condyle of antenna (Fig. 2A); head (Fig. 2B) with
median lobe elongate, narrow, nearly parallel-sided; medial angle of temporal lobe obtuse, margin
posterior to it slightly emarginate; temporal lobe very finely punctate; orbital groove absent (Fig. 2C);
temporal setae absent.

Pronotum (Fig. 2B) relatively elongate, L/GW 1.50, widest at middle; sides only slightly curved;
median groove fine, linear except for anterior and posterior median pits; each paramedian groove
represented by deep basal impression preceded by five or six coarse punctures in a line, ended ante-
riorly at anterior third of pronotum; marginal groove linear, fine; prosternum impunctate in midline,
punctate laterally; pronotum sparsely punctate.

Elytron elongate; striae shallow, coarsely punctate, pollinosity limited to punctures; one seta in
apical pit at apex of stria II; one seta at posterior end of stria IV; about nine setae in apex of stria
VII. Metasternum impunctate in midline, with coarse lateral punctures. Abdominal sterna (Fig. 2D)
with transverse row of punctures; sternum II with a shorter, more iiTegular row; sterna IV and V with

74

Annals of Carnegie Museum

VOL. 69

Fig. 2. — Plesioglymmius {Juxtaglymmius) negara, n. sp. A. Head, anterior aspect; AL, antennal lobe;
BC, basal condyle. B. Head, pronotum, dorsal aspect. C. Head, left lateral aspect. D. Metasternum,
abdomen, ventral aspect; male.

scattered punctures; sternum VI with coarse row parallel to posterior margin, with scattered punctures
anterior to it; lateral pits very shallow in male; female unknown. Male with small ventral tooth on
anterior femur; calcar of hind tibia with acute angle, raised well above tibial apex (more strongly so
than in P. jugatus).

Distribution. — Known only from the type locality.

Arrowina anguliceps (Arrow, 1901)

Additional Locality. — INDIA: Tamil Nadu, Ootacamund Pykara, 11 km from
Mysore. A. Riedel (Hendrich colL). Previously known only from the Cardamum
Mountains, further south.

Omoglymmius {Ortho glymmius) coomani (Arrow, 1942)

Additional Localities. — CHINA: Yunnan, 14-21 VI, 1993. 100 km W. Baoshan,
Gaoligongshan Nat. Res. coll. E. Jendak and O. Sausa (NMW). THAILAND:
(Northwest), 19°18'N, 97°59'E, Hae Hong Son, 1600-2000 m, Ban Huai Po. 9-
16, 1991. S. L. Dembicky (NMW). This species was previously known only from
Vietnam.

Omoglymmius (sensu stricto) sakuraii (Nakane, 1978)

Additional Locality. — TAIWAN: Taichung (same data as for Y. niponensis
above). The type locality is Amami-O-Shima in the Ryukyus. The species has
also been recorded from southern Japan (Kagoshima Prefecture, Kyushu Island)
and Vietnam. The precoxal carina and other characters vary in this species, and
we provisionally treat it as one variable species. The Taiwan specimens were in
a very large (50 m) spruce {Picea morrisonicola Hayata) log (CMNH).

2000

Bell and Bell — New Rhysodine Beetles

75

Omoglymmius (sensu stricto) semperi Bell and Bell, 1982

Additional Locality. — PHILIPPINES: Luzon, Laguna, Mt. Makiling, 3000 ft,
26 April, 1931, E C. Hadden coll. (CAS). This is the first definite locality for the
species.

Omoglymmius (sensu stricto) hiekei Bell and Bell, 1982

Additional Locality. — MALAYSIA: Sabah 60 km E of Kota Kinabalu, Crocker
Mtns., Gunung Emas, 16-27 IV, coll. 1. Janis (NMW). This puts the species on
the island of Borneo. Previous records are from Luzon, Philippines.

Omoglymmius (sensu stricto) bucculatus (Arrow, 1901)

Additional Locality. — INDONESIA: Lombok: SapiCSembalun Bumbung, 14-
16 Feb 1994. Bolm leg., 900-1500 m (SMNS). This is the first record of a
rhysodine from Lombok. This species has previously been recorded only from
Sumbawa.

Omoglymmius (sensu stricto) patens Bell and Bell, 1982

Additional Locality. — IRIAN JAYA: Manokwari, Ransiki, Mayuby, ca. 300 m,
A. Riedel (Hendrich coll.). The known range of this species is on the north side
of New Guinea, from Maffin Bay in the east to the east side of the Vogelkop
Peninsula.

Omoglymmius {Omoglymmius) pulvinatus (Grouvelle, 1903)

(Fig. 3A-C)

Description of Female. — 7.2 mm, labeled “W. Neuguinea/Pariai, Nabire Strass v. Nabire nach Ilasa,
km 50, unter Rinde 29.9.90, leg. M. Balke, L. Hendrich” (collection of L. Hendrich, Berlin); head
and pronotum (Fig. 3A and B) similar to male, anterior trochanter rounded, abdominal sterna (Fig.
3C) with lateral pit of sternum IV deep, but not much wider than that of male.

Range Extension. — IRIAN JAYA: Testega, Manokwari Province, 1100-1300
m elevation. A. Riedel (Riedel coll.); Nabire (Hendrich coll.). This species is
perhaps confined to the Vogelkop Peninsula and south coast of Geelvink Bay.

Omoglymmius {Omoglymmius) emdomani, new species
(Fig. 4A-C)

Type Specimens. — Holotype female, labelled “Irian Jay a, Jayawijaya Prov. leg.
A. Riedel 1993, Emdoman 900-1200 m., 29“IX-1993'’ (CMNH). The locality is
in the mountainous interior of New Guinea, at about 4°S, 140°E. Paratype: female,
mounted on same card as holotype. The holotype is the larger and darker speci-
men.

Etymology. — The specific epithet is derived from the name of the type locality.

Diagnosis.— An Omoglymmius sensu stricto with suborbital tubercles, the only
such species with the antennal lobe entirely pollinose and with the median lobe
shallowly emarginate posteriorly.

Description. — Length 6. 5-8. 2 mm. Antennal segments I-X punctate, XI impunctate; head (Fig. 4A)
with length and width subequal; median lobe impunctate, truncate, slightly emarginate at apex, tip
pollinose; frontal space broader than long; temporal lobe with anteriomedial margins slightly curved.

76

Annals of Carnegie Museum

VOL. 69

B

Fig. 3. — Omoglymmius (sensu stricto) pulvinatus, (Grouvelle). A. Head, pronotum, dorsal aspect. B.
Head, left lateral aspect. C. Metasternum, abdomen, ventral aspect; female.

medial angles narrowly separated, slightly lobate; posteriomedial margins concave to posteriomedial
angle, posterior margin transverse, temporal lobes margined posteriorly with pollinosity; medial slope
of antennal lobes entirely pollinose, as are broad antennal grooves; orbital grooves broad, extended
posteriorly nearly to posterior margin of eye; temporal lobe impunctate; temporal seta absent; eye
large, round, prominent; suborbital tubercle rather large, about one-third as long as eye (Fig. 4B).

Pronotum (Fig. 4 A) only moderately elongate, L/GW 1.17; pronotum almost quadrate but narrowed
to anterior angles; lateral margins nearly straight; outer carina about three-fifths as wide as inner one
at middle; medial margin of outer carina sinuate near base; marginal groove dilated, anterior half
punctate; pronotum without setae; pronotal carinae impunctate; prosternum microsculptured in anterior
half; precoxal carinae absent.

Elytra with striae shallowly impressed, intervals flat; strial punctures coarse, sparse; base of stria
IV with pollinose scarp; transverse basal scarp pollinose from stria II-IV; stria IV with one seta at
apex; subapical striole with one seta; stria VII with several setae near apex. Metasternum (Fig. 4C)
with single row of punctures along each lateral margin, but without median row. Abdominal sterna
III-V with coarse punctures in form of transverse row widely interrupted in midline; sternum VI with

2000

Bell and Bell — New Rhysodine Beetles

77

Fig. 4. — Omoglymmius (sensu stricto) emdomani, n. sp. A. Head, pronotum, dorsal aspect. B. Head,
left lateral aspect. C. Metasternum, abdomen, ventral aspect; female.

scattered, very coarse punctures on posterior two-thirds; female with small but deep lateral pit on each
side of sternum IV, without tooth on anterior femur; male unknown.

Distribution. — Known only from the type locality.

Remarks. — In the revised key to Omoglymmius sensu stricto, this species traces
to couplet 74 (Bell and Bell, 1993). The latter couplet should be altered as follows:

74(73). Marginal groove of pronotum not dilated 75

74'. Marginal groove of pronotum dilated ................................. 74.1

74.1(74'). Inner carinae of pronotum impunctate but outer carinae coarsely punctate; one temporal

seta present; pronotal margins curved O. pulvinatus (Grouvelle)

74.1'. Both carinae impunctate; temporal seta absent; pronotal margins nearly straight . . .

O. emdomani, n. sp.

Omoglymmius (sensu stricto) sedlaceki Bell and Bell, 1982
Additional Locality. — IRIAN JAYA: Manokwari Province, Anggi, Gunung Dis-
behey, 29.8.1991, 2000-2150 m, leg. A. Riedel (SMNS). Previous records of this

78

Annals of Carnegie Museum

VOL. 69

species are from the highlands of northeastern New Guinea, and this is the first
record from the western half of the island. The unity of this species is still in
doubt. We found what appeared to be five distinct local forms in northeastern
New Guinea which we designated by letters. The Anggi specimen is similar to
our form E, from Sepalakambang, on the Huon Peninsula, from which it might
differ only in having the median lobe of the head even narrower and more pointed.
The status of these local forms remains in doubt, and they should be reexamined
when more specimens are available.

Subtribe Clinidiina

Rhyzodiastes (Temoana) riedeli, new species
(Fig. 5A and B)

Type Specimen. — Holotype male, labelled “Irian Jaya: Baliem Dist., Ilugwa,
Melanggama, Pass- Valley, 2100-2300 m., leg. A. Riedel, 9-10, IX, 1990” (Hen-

drich Colin.) (MNHB).

Etymology. — The specific epithet honors the collector, Alexander Riedel of Friedberg, Germany,
who has sent us many interesting rhysodids from his collecting expeditions to Irian Jaya, New Guinea.

Diagnosis. — A species of Rhyzodiastes {Temoana) with antennal tufts begin-
ning on segment V; marginal stria reduced to a row of punctures except near
humerus and apex; sutural stria two-thirds length of elytra, pronotum rather short
and subquadrate, and elytra without setae except near apex of marginal stria.

Description. — Length 7.0 mm. Tufts of minor setae on antennal segments V-X; basal setae of
antennal segments absent; segments I-V with apical pollinose bands; apical stylet small, acute; head
(Fig. 5A) about as long as wide; median lobe broad, triangular, its margins straight, apex slightly
acute, opposite anterior third of eye; antennal lobe pollinose except for narrow lateral margin; temporal
lobes widely separated from one another; orbital groove almost absent, represented only by very short
narrow pollinose area opposite anterior end of eye; eye very narrow; temporal lobe with one seta in
pollinose depression.

Pronotum (Fig. 5 A) rather short, L/GW 1.21; subquadrate, sides only slightly curved, widest near
middle, median groove deep, moderately narrow (but wider than in R. rajfrayi), anterior pit slightly
widened; median groove impressed beyond posterior pit to pronotal base; paramedian grooves with
medial margins indistinct, lateral margins abrupt; outer carina broad, its width about half that of
distance from paramedian groove to midline; outer carina of even width, truncate anteriorly and
posteriorly; pronotal setae absent; marginal groove limited to posterior fourth of pronotum, visible
only in lateral view.

Elytra only moderately elongate (shorter than in R. rajfrayi), scarcely narrowed anteriorly; sutural
stria impressed, impunctate, its apical third effaced; parasutural stria deep, entire, impunctate; intra-
tubercular stria deep, entire, impunctate; marginal stria impressed only near humerus and apex, oth-
erwise represented by widely spaced punctures; apical, subapical tubercles inflated, former extended
to suture; elytra without setae except for four or five in apex of marginal stria. Metasternum not
sulcate, its surface microsculptured, with bluish sheen. Abdominal sterna (Fig. 5B) with sterna III-VI
with deep, pollinose transverse sulci broadly interrupted at midline; sternum VI with submarginal
groove pollinose, joined at anterior ends to transverse sulci; lateral end of sulcus on sternum IV in
form of a slightly enlarged pit in male; male with ventral surface of anterior femur with many small
tubercles, with anterior trochanter acutely pointed ventrally; tibial spurs of middle and hind legs equal;
calcars nearly triangular, but abruptly truncate apically; trochanter of middle leg with dorsal sinuation.
Female unknown.

Distribution. — Known only from the type locality.

Remarks. — This species is closest to R. rajfrayi (Grouvelle) and keys to that
species in Bell and Bell (1985:12), except for the lack of setae in the parasutural
stria. The two species are separated by the lack of most elytral setae in R. riedeli,
and the quite different shapes of the pronota of the two species.

2000

Bell and Bell — New Rhysodine Beetles

79

Fig. 5. — Rhyzodiastes (Temoana) riedeli, n. sp, A. Head, pronotum, dorsal aspect. B. Metasternum,
abdomen, ventral aspect; male.

Rhyzodiastes riedeli is the second species of the genus found in New Guinea.
It is a high-altitude species, while R. guineensis (Grouvelle) is a lowland species.
Rhyzodiastes guineensis differs conspicuously in having a long, oval pronotum
with a linear median groove, a complete orbital groove, the sutural stria com-
pletely lacking, and the tibiae incrassate. The two species do not appear to be
closely related. The type locality of R. riedeli is high in the main central mountain
range of New Guinea.

The key to the subgenus (Bell and Bell, 1985:12) should be altered as follows:

6(3'). Orbital grooves very much abbreviated or absent; temporal setae one or none 7

6'. Orbital groove complete at least to posterior margin of eye; temporal setae one or two

9

Lateral margins of pronotum nearly straight; outer carinae as far apart at apex of pronotum

as at base; one temporal seta . Rhyzodiastes {Temoana) riedeli, n. sp.

Lateral margins of pronotum clearly convergent anteriorly; outer carinae much closer

together anteriorly than posteriorly; temporal seta absent 7.1

Orbital groove present, abbreviated opposite middle of eye; one temporal seta ......

. ... R. mishmicum (Arrow)

Orbital groove absent 8

Rhyzodiastes {Temoana) mindoro, new species
(Fig. 6A and B)

Type Specimen. — Holotype female, labeled “Philippinen-Mindoro, 28 km. S.
Calapan 1992, Balete 100-700 m. (19) leg. Jach 27-29.11.” (NMW).

7(6).

7'.

7.1(7').

7.1'.

80

Annals of Carnegie Museum

VOL. 69

Etymology. — This specific epithet is derived from the name of the native island of the species.

Diagnosis. — A Rhyzodiastes (Temoana) with the characters of the singularis
group (marginal stria entire, orbital groove complete, temporal setae present), with
sutural stria impressed, and with two temporal setae, both in the pollinosity of
the orbital groove. It differs from R. {T.) bipunctatus, since the latter species has
the second temporal seta arising from a pit in the temporal lobe, well away from
the margin. It resembles R. rimoganensis in having tubercles on sternum VI but
differs from it by the presence of two temporal setae and by the linear median
pronotal groove.

Description. — Length 7.5 mm. Antennal stylet short, conical; antennal segments short subcylindri-
cal; tufts of minor setae on segments V-X; segments I and II partly pollinose, remaining segments
glabrous; head (Fig. 6A) slightly longer than wide; median lobe short, triangular; antennal lobe gla-
brous, shining, well separated from median lobe; frontal grooves rather narrow pollinose; temporal
lobe about twice as long as wide; medial margins curved evenly; orbital groove complete, with two
temporal setae, anterior one at posterior end of eye, posterior near posteriolateral angle of head; eye
crescentic, about three-fourths length of temporal lobe; posterior face of temporal lobe pilose.

Pronotum (Fig. 6A) elongate, L/GW 1.46, widest at posterior fourth, strongly narrowed anteriorly,
rounded and narrowed posteriorly; median groove narrow in middle, dilated and tapered into anterior
and posterior pits; paramedian groove with medial margin gradually sloped into disc, but lateral margin
abrupt, pollinose scarp; medial margin of outer carina evenly curved; outer carina tapered, attenuate
anteriorly, slightly narrowed posteriorly, rounded at base; marginal groove entire, visible in dorsal
view; submarginal groove absent; pronotal setae absent; sternopleural groove absent.

2000

Bell and Bell — New Rhysodine Beetles

81

Elytra moderately elongate, cauda absent; base of elytron deeply concave medial to base of para-
sutural stria, concavity densely pilose except narrow strip along suture; large scutellar pits opened into
this depression; sutural interval nearly flat; sutural stria fine, impressed, finely punctate, apex outcur-
ved, nearly extended to parasutural stria; second interval depressed; parasutural stria in form of me-
dially facing scarp; intratubercular stria deep, entire; marginal deep, entire; submarginal impressed,
ended opposite base of sternum VI; parasutural stria with one seta each near base and apex; marginal
stria with four setae near apex; apical tubercle without setae. Metasternum not sulcate. Abdominal
sterna (Fig. 6B) with transverse sulci of sterna III-VI broadly interrupted at midline, sulci pollinose,
with distinct circular pit at medial ends; sternum IV with very large lateral pits (much larger than in
R. (T.) rimoganensis); sternum VI with submarginal groove dilated, well separated from transverse
grooves; posterior half of sternum VI impressed, bounded anteriorly by pair of tubercles; tibial spurs
of middle and hind legs equal.

Distribution. — Known only from Mindoro Island in the Philippines.

Remarks. — The discovery of a member of this subgenus from the Philippines
is not unexpected, as it has been found in all surrounding island groups. Other
species probably await discovery in the Philippines.

To accommodate R. mindoro, and to make a better separation of R. convergens,
the key to the subgenus (Bell and Bell, 1985:12) should be altered as follows:

11(9')^

ir.

12(11').

12'.

13(12').

13'.

13.1(13).

13.1'.

Metasternum with median sulcus; temporal lobe with one seta in orbital groove and

one discal seta R. bipunctatus Bell and Bell

Metasternum without median groove; temporal lobe without discal seta but with one

or two in orbital groove 12

Portion of median groove of pronotum between pits linear; orbital groove with two

setae R. mindoro, n. sp.

Median groove not as narrow; orbital groove with one seta .................. 13

Outer carinae of pronotum strongly tapered anteriorly 13.1

Outer carinae not tapered anteriorly R. mirabilis

Median groove of pronotum narrow, although not linear, about one-tenth of width of

pronotum . R. convergens Bell and Bell

Median groove broad 14

Rhyzodiastes (Temoana) convergens Bell and Bell, 1985

Additional Locality.— VAVVK NEW GUINEA: Manus L, 18-11, 1981, Bowar,
23 km W of Lorengau, 230 m, rainforest fragment, 21, coll. W. L. Brown (CUIC).
The species was previously known from New Britain, in the Bismarck Archipel-
ago. This locality is in the Admiralty Islands, nearly 400 km distant, across deep
waters. Although we are not naming the Manus specimen, we have some doubts
about its identity with the New Britain species. The marginal groove of the pron-
otum is finer and fades out completely before the apex. The orbital groove is
much finer and disappears opposite the posterior margin of the eye. In most of
the type series of convergens, this groove is complete. In one paratype, however,
it is as short as in the Manus specimen, although it is as broad as in the other
New Britain specimens. The Manus specimen has a complete row of setae in the
parasutural stria. We did not list such setae in the convergens description; however,
on close examination, two of the types have a few setae in such a series, and they
may have broken off in other specimens. There is no trace of the sinuation at the
posteriomedial angle of the temporal lobe, but this is very reduced in some of
the type series of convergens. We need a series of specimens from Manus before
deciding whether its population merits a name. Whether the two populations are
separate species or only incipient subspecies, they give the impression of being
closely related populations just beginning to develop morphological differences
after being isolated after a dispersal event. The key can be altered so as to separate

82

Annals of Carnegie Museum

VOL. 69

Fig. 7. — Clinidium (sensu stricto) howdenorum. Bell and Bell. A. Head, pronotum, dorsal aspect. B.
Metasternum, abdomen, ventral aspect; female. C. Sternum VI, abdomen, left lateral aspect; female.

convergens (both populations) from related species. (See above under Rhyzodias-
tes mindoro.)

Clinidium (Arctoclinidium) rosenbergi Bell, 1970

Additional Locality. — ARKANSAS: Cross County, Village Street State Park,
29 VI, 1987, coll. Kovarik, one male, two females (TAMU). A new state record.

Clinidium {Clinidium) insigne Grouvelle, 1903

Additional Locality. — This species has previously been known only from the
type locality, Cali, Colombia. Through the collecting efforts of Onore and asso-
ciates, it is now revealed to occur in the northern half of the Andean part of
Ecuador, south at least to Bolivar Province. Ecuadorean localities are Pichincha
Province: Chiriboga, Est. Forestal “La Favorita,” many specimens, several col-
lectors; Palmeros, 26- VI-87, leg. Bustamente; Santo Domingo, 23-V-86, coll. P.
Vega. Cotopaxi Province: Calupina, VII-87, coll. Onore; Sigchos, VII-87, coll.
Onore. Bolivar Province: Totoras, XII-87, coll. P. Mendoza (all specimens
CMNH).

Clinidium {Clinidium) howdenorum Bell and Bell, 1985
(Fig. 7A-C)

Description of Female. — Three specimens, length 4. 9-6. 4 mm, labeled “Trinidad: Arima, Blan-
chisseuse Rd. 13.1 km n. of Arima, 29 Mar 1987, elev. 500 m. Trinidad Field Party, 1987, M. E.
Carter, E. R. Hoebeke, J. K. Liebherr” (CUIC). Head and pronotum similar to male (Fig. 7A), but
abdominal sterna (Fig. 7B) with transverse sulci of sterna III-VI narrowly interrupted in midline;

2000

Bell and Bell — New Rhysodine Beetles

83

sternum IV with lateral pit targe, deep, triangular; sternum VI (Fig. 1C) impressed posteriorly with
slight transverse carina helping to define rudimentary median tubercle.

Remarks. — Hovorka (1997) recently described another new species of Clini=
dium (sensu stricto) from Ecuador, Clinidium mareki. The female of this species
differs from C. howdenorum in having a pair of grooves on sternum VI which
join together at the submarginal groove forming a characteristic V-shape.

Clinidium {Clinidium) boroquense Bell, 1970

Remarks. — The tufts of minor setae on the antenna are present on segments
V-X, not VI-X as previously reported. Additional material has shown that pre-
coxal setae are present, at least in some specimens. The antennal tufts take this
species out of the Insigne group, as defined by Bell and Bell (1985). It belongs
to the Guildingii group. The complete intercalary and intratubercular striae and
absence of a longitudinal metastemal sulcus would put it in the Jamaicense sec-
tion, but this is only doubtfully distinct from the Oberthueri section, in which the
sulcus is very shallow in some species. Within the Jamaicense section, this species
agrees with C. (C.) trionyx in the presence of precoxal setae, but the latter species
has “false spurs” on the middle and hind tibiae, lacks discal strioles, and has the
eye modified. Clinidium boroquense agrees with C. (C.) jamaicense and related
species of Jamaica and Hispaniola in having narrow, elongate eyes and welb
developed discal strioles, but the latter group of species lack precoxal setae and
do not have the temporal lobes convergent posteriorly.

In the key in Bell and Bell (1985), boroquense should be removed from couplet
8, so couplet 7 leads directly to couplet 9. Clinidium boroquense would trace to
couplet 18. At this point, those specimens with a precoxal seta would key to C.
(C.) humboldti Bell and Bell, and those without a precoxal seta would probably
trace to C. (C.) jolyi Bell and Bell. The key can be corrected as follows, using a
new couplet 18, and renumbering the old one as 18.1.

18(15')- Temporal lobes convergent posteriorly . C. boroquense Bell

18'. Temporal lobes not convergent 18.1

18.1(18'). Precoxal setae present 19

18.1'. Precoxal setae absent 21

Clinidium {Clinidium) spatulatum Bell and Bell, 1985
(Fig. 8A-C)

Description of Male. — Length 6.4 mm. Head, pronotum, and abdomen (Fig. 8A and B) similar to
female, but anterior tibia (Fig. 8C) with proximal tooth; anterior femur with raised carina; middle tibia
with small acute calcars; hind tibia with short, rather blunt calcars.

Similar to C. validum Grouvelle male, except that the latter lacks the femoral carina and hind tibia
has acutely pointed calcar. This specimen labeled “Panama: Darien, Estacion Ambiental Cana,
07°45.32'N, 77°41.07'W, Cerro Pirre, 1450 m., 6-VI-1996, R. S. Anderson, 96-1 12C, cloud forest
litter” (CMNO).

The eye area of this specimen is heavily pigmented and almost invisible. This was true also of the
holotype. In other members of this subgenus, paler (younger?) individuals may have unpigmented,
obviously functional eyes, while the darker specimens have the area pigmented even more heavily
than the rest of the exoskeleton.

Clinidium {Clinidium) oberthueri Grouvelle, 1903

Remarks. — The type locality of this species was simply “Ecuador.” Bell and
Bell (1985) listed three specimens from Papallacta, Napo-Pastaza Province. The

84

Annals of Carnegie Museum

VOL. 69

Fig. 8. — Clinidium (sensu stricto) spatulatum Bell and Bell. A. Head, pronotum, dorsal aspect. B.
Metasternum, abdomen, ventral aspect; male. C. Anterior leg, tibia with proximal tooth, femur with
raised ventral carina; male.

additional records below indicate that the range in Ecuador is similar to that of
C. insigne, and that the species coexist in at least two localities.

Further locality records are: Napo-Pastaza Province: Cosanga, IP89, coll. G.
Onore; km 36 via La Alegria La Bonita, 2100 m. Apr. 1-86, D. Bastidas (colL P.
Moret). Pichincha Province: Ayuriqui, 12-VIL87, M. Ferro; Palmeros, 26-VL87,
leg. Bustamente, Cotopaxi Province: Calupina, IIL87, G. Onore. Tungurahua
Province: Tungurahua nord, R Moret, lO^IILSS, 3400 m. (coll. P Moret). (All
specimens CMNH except those cited as R Moret.)

Clinidium {Clinidium) rossi Bell, 1970

Remarks. — This species was previously known from the type locality, Golfito,
on the Pacific coast of Costa Rica. Additional records show that it occurs near
the Atlantic coast, as well as in Panama. Costa Rica: Limon Province: Valle de
la Estrella, Pandora, 17^20 Feb. 1984, H. & A. Howden (CMNO). Puerto Viejo
Province: Sarapiqui, VIIL4-65, Raske (CNCO). Panama: Chiriqui Province: Dst.
Renacimiento, Sta. Clara, 4400-4200 ft, July 5, 1976 (TAMU).

Clinidium {Clinidium) dormans Bell and Bell, 1985
(Fig. 9 A and B)

Description of Female. — Length 5.8 mm, labeled “Panama; Chiriqui Prov., Reserva la Fortuna,
Continent Divide Trail, 19-20 Apr., 1993 A. Gillogiy” (TAMU). Head and pronotum (Fig. 9A) similar
to male, but abdominal sterna (Fig. 9B) with lateral pit of sternum IV slightly deeper; sternum VI
with five pairs of setae, two in submarginal groove, two on disc, and one in transverse sulci.

2000

Bell and Bell — New Rhysodine Beetles

85

Fig. 9. — Clinidiurn (sensu stricto) dormans Bell and Bell. A. Head, pronotum, dorsal aspect. B. Me-
tasternum, abdomen, ventral aspect; female.

Clinidiurn {Clinidiurn) crater Bell and Bell, 1985
(Fig. 10A~D)

Description of Male. — Series of seven; lengths 5. 6-6. 5 mm. Six females in the same series measure
6.0-6. 8 mm; labeled “Panama, Pan. Pr., km 8, El Llano-Carti Rd., VI and VII (various dates), 1994,
elev. 400 m., A. R. Gillogly” (UVM). Head and pronotum (Fig. lOA) similar to female; abdominal
sterna (Fig. lOB) with transverse sulci of sterna III-V shallow, all equally narrowly interrupted in
midline; lateral pits of sterna III-V equal, very shallow; sternum VI with transverse sulcus represented
only by pair of coarse, shallow punctures on each side; some specimens with short, thin, curved,
pollinose line transversely arranged in midline, anterior part of disc separated from posterior, sloped
part; anterior femur (Fig. IOC) with ventral tooth; calcar of middle leg minute; calcar of hind leg
small, acutely pointed (Fig. lOD).

Clinidiurn (sensu stricto) onorei, new species
(Fig. IIA and B)

Type Specimen. — Fiolotype male, labeled “Ecuador, Pichin (cha), Ayuriqui 12-
VIF82. legit. M. Ferro ex copal, Hora RM., haciendo galeria” (CMNH). Paratype:
one female, same label as holotype (UVM).

Etymology. — The specific epithet is derived from the surname of Dr. Giovanni Onore, a Research
Associate of Carnegie Museum of Natural History, whose energetic collecting and generosity with

specimens has added much to our knowledge of Ecuadorian Rhysodini.

Diagnosis. — A Clinidiurn sensu stricto in the rossi section, combining a narrow
median pronotal groove, a deep median metastemal groove, with a lack of setae
in the sutural stria.

Description. — Length 5. 6-6.0 mm; antennal stylet very slender, acute, about three-tenths of length

86

Annals of Carnegie Museum

VOL. 69

Fig. 10. — Clinidium (sensu stricto) crater Bell and Bell. A. Head, pronotum, dorsal aspect. B. Metas-
ternum, abdomen, ventral aspect; male. C. Anterior leg, ventral tooth on femur; male. D. Hind leg,
femur and tibia; male.

of antennal segment XI; tufts of minor setae on segments V-X; segments I-VII each with subapical
pollinose ring; basal setae present on segments VI-XI or VII-XI. Head (Fig. 1 1 A) slightly longer than
wide; frontal grooves deep, narrow; median lobe short, triangular, its tip opposite anterior part of eye;
medial margins of temporal lobes closest together posterior to tip of median lobe, divergent posteriorly;
posteriomedial margins of temporal lobes very broadly bordered by pollinosity, pollinosity along
lateral border of lobes also broad, irregular, curved around bases of temporal setae; four temporal
setae (one anterior to eye, two opposite eye, one posterior to eye); two pairs of postlabial setae.

Pronotum (Fig. 11 A) moderately elongate; L/GW 1.44, widest posterior to middle; pronotum more
narrowed anteriorly than in segue or oberthueri; median groove very narrow, its margins parallel,
slightly widened between median pits, extremely narrow behind posterior median pit; basal impres-
sions oblong; discal striole linear, nearly straight, about 0.45 of length of pronotum, marginal groove
narrow, visible in dorsal view, eight to nine lateral setae; one seta posterior to each basal impression,
one pair of discal setae; precoxal seta present; sternopleural suture faintly suggested.

Elytra moderately elongate; sutural, parasutural, intercalary striae complete, impressed; intratuber-
cular stria complete apically, but not impressed, apical and subapical tubercles thus not differentiated;
marginal stria impressed; sutural striae without setae; parasutural stria with one seta at base; intercalary
stria with complete series of about 12 setae; series of five setae near apex of intratubercular seta and
continued onto apical tubercle; marginal stria with complete series of about 16 setae. Metasternum
(Fig. IIB) with deep, linear median groove. Abdominal sterna (Fig. IIB) with transverse sulci of
sterna III-V represented by rows of coarse punctures, widely separated at middle; sternum VI with
submarginal sulcus widely separated from transverse sulci; latter short but deep; lateral pit of sternum
IV moderately large, shallower in male; Sternum VI with eight setae in two transverse rows, four
anterior to submarginal groove, and four posterior to it; each tibia with two equal spurs, and an apical
cusp (homologous to the “false spur” in some related species); male with both mesotibial and me-
tatibial calcars cultrate.

Distribution. — Known only from the type locality.

Remarks. — The species is startlingly similar in general shape to the sympatric

2000

Bell and Bell — New Rhysodine Beetles

87

Fig. 11. — Clinidium (sensu stricto) onorei, n. sp. A. Head, pronotum, dorsal aspect. B. Metasternum,
abdomen, ventral aspect; female.

C. oberthueri Grouvelle. The latter is easily distinguished by the separation of
the subapical and apical tubercles, and by the setae in the sutural stria.

To accommodate C. onorei, substitute the following for couplets 26 and 27 in
the key of Bell and Bell (1985:96):

26(25').

26'.

27(26').

27'.

27.1(27').

27.1'.

Sutural stria with five setae; female with brash of dense hairs at base of hind femur

C. segne Bell and Bell

Sutural stria without setae or with one basal seta . 27

Frontal grooves very shallow; antennal stylet minute, about one-tenth of length of

antennal segment XI; pronotum without discal setae C. dormans Bell and Bell

Frontal grooves deep; antennal stylet about one-third of length of antennal segment

XI; pronotum with one pair of discal setae 27.1

Median groove of pronotum dilated, its greatest width subequal to about half of dis-
tance to discal striole at this point; transverse sulci of abdominal sterna III-V im-
pressed, continuously pollinose; female with anterior margin of subapical sulcus of
sternum VI angulate at midline ..................... C. kochalkai Bell and Bell

Median groove very narrow, at greatest width less than one-fifth of distance to striole;
transverse sulci of abdomen formed of single rows of very coarse punctures, pollinosity
limited to punctures; female with anterior margin of subapical sulcus unmodified . . .
C onorei, n. sp.

Clinidium (sensu stricto) moldenkei Bell and Bell, 1985

Additional Locality. — PANAMA: Chiriqui Province, 2200 m. Las Nubes,
Parque Amistad, 20 Dec. 1992, A. R. and T Gillogly (TAMU). Bocas del Toro
Province, Fortuna area, N. continental divide, 750 m, Aug. 6, 1993. A. R. Gillogly
(TAMU). Previously known from Oso Peninsula of Costa Rica.

88

Annals of Carnegie Museum

VOL. 69

Clinidium (sensu stricto) sulcigaster Bell, 1973

Additional Locality. — PANAMA: Chiriqui Province: 1 km. n. of Jurutungo, 20-
VI- 1994, el. 1900 m, A. R. Gillogly (UVM). Previously known from Guatemala.
This at present is the only species known to extend through the entire length of
Central America.

Clinidium {Clinidium) gilloglyi, new species
(Fig. 12A-C)

Type Specimens. — Holotype male, labeled “Panama: Chiriqui Pr., 1 km SE
Hornito 23-VII-1994, 1000 m. A. R. Gillogly” (CMNH). Paratypes: one female,
one male same data as holotype (UVM).

Etymology. — The specific epithet is derived from the surname of the collector, Alan R. Gillogly, in
appreciation of his efforts in sending us fine rhysodine material from Panama.

Diagnosis. — A Clinidium sensu stricto with the compound eye divided into two
ocelluslike organs, and with tufts of minor hairs on antennal segments VII-X.

Description. — Length 6. 0-6. 6 mm. Antennal stylet compressed, chisel-like, obliquely truncate,
length about one-third of length of last antennal segment. Minor setae present on segments VII-X;
basal setae absent; segments I-X each with subapical pollinose ring; head (Fig. 12A) longer than
broad; median lobe small, shield-shaped; frontal grooves broad, pollinose; antennal lobe small, sepa-
rated from temporal lobe by broad, pilose postantennal area; frontal space moderately broad; medial
margins of temporal lobes diverging slightly posteriorly; posterior margin of temporal lobe broadly
pilose; orbital groove complete; three to five temporal setae in orbital groove; compound eye divided
into two ocelluslike organs (Fig. 12B); two pairs of postlabial setae.

Pronotum (Fig. 12A) elongate, L/GW 1.48; widest near middle, sides curved; apex strongly nar-
rowed, truncate, base slightly narrowed, curved; median groove open posteriorly, strongly dilated,
margins of groove straight, slightly convergent posteriorly; anterior pit only slightly wider than median
groove; discal striole straight, extended to middle of pronotum; marginal groove slightly dilated, visible
in dorsal view; six to eight marginal setae, none medial to basal impression (such setae present in C.
argus); sternopleural groove intact anteriorly; posterior half formed by line of three pits.

Elytra relatively long, narrow; sutural, parasutural striae deep, narrow, conspicuously punctate; in-
tercalary stria wider, deeper than others; intratubercular fine, entire; marginal stria entire, strongly
dilated posteriorly; preapical tubercle scarcely inflated; apical tubercle inflated; sutural, parasutural
striae without setae; intercalary stria with series of eight setae; intratubercular stria with three setae
opposite apical tubercle; apical tubercle with three or four setae; marginal stria with nine or ten setae.
Metasternum (Fig. 12C) widely, shallowly impressed in midline, but without distinct median sulcus.
Abdominal sterna (Fig. 12C) with transverse sulci rather broadly interrupted in midline, interruption
about half length of sulcus on either side of it; transverse sulcus of sternum VI broadly separated from
submarginal sulcus; lateral pit of sternum IV moderately large in female, shallower in male; sternum
VI with one pair of apical setae; male abdomen without median pollinose areas, without paired tu-
bercles; anterior femur of male without ventral tooth, but with slight ventral ridge; lateral pit of sternum
IV in female large, nearly as long as sternum, that of male small; calcars triangular, that of hind tibia
sinuate on dorsal margin.

Distribution. — Known only from the type locality. One other specimen is
known from Veraguas Province, Panama, near Hato Chami, Cerro Colorado, 1400
m (TAMU).

Remarks. — The combination of divided eyes and minor setae on antennal seg-
ments VII-X has not been seen before in Clinidium sensu stricto, and prevents
this species from keying out in Bell and Bell (1985:93-97). The minor setae send
it ultimately to couplet 5, where it would match neither alternative. The eyes
would send it to couplet 43, the bee earn group. The latter group appears to be
its real kin; the head and pronotum are almost a perfect match for C. argus Bell
and Bell. The Clinidium key should be altered as follows:

2000

Bell and Bell — New Rhysodine Beetles

89

B

Fig. 12. — Clinidium (sensu stricto) gilloglyi, n. sp. A. Head, pronotum, dorsal aspect. B. Head, left
lateral aspect. C. Metasternum, abdomen, ventral aspect; female.

2(1'). Eye not constricted in the middle nor divided into anterior, posterior halves; tufts of minor

setae on antennae distinctly developed, on segments V-X, VI-X, or VII-X

2'. Eye either constricted at middle of length or else divided into anterior and posterior oceh
luslike parts; tufts of minor setae either very inconspicuous on segments VII-X or else
absent

43

C. gilloglyi will be included and the key to related species improved by sub-
stituting the following couplets in the same key:

43(2'). Discal striole of pronotum much closer to median groove than to lateral margin; antennal

stylet very small C. heccarii Grouvelle

43'. Discal striole either closer to lateral margin than to median groove, or else equidistant

between them; antennal stylet large, obliquely truncate, chisel-like 44

44(43'). Parasutural (second) stria with complete series of setae ...................... 45

44'. Parasutural stria without setae 46

45(44). Anterior median pit of pronotum about one-half of width of pronotum opposite it; an-
terior and posterior parts of eye small, close together; male with deep median sulcus on

abdominal sterna I-III C. sulcigaster Bell

45'. Anterior median pit about one-fourth of width of pronotum; eye parts larger and further

apart; male with only a slight suggestion of median groove C. gilloglyi, n. sp.

46(44'). Anterior part of sternopleural groove pollinose; median groove of pronotum as wide as

90

Annals of Carnegie Museum

VOL. 69

anterior pit, sides of groove parallel; male with pollinose strip in midline of abdominal

sterna I-III C argus Bell and Bell

46'. Anterior part of sternopleural groove glabrous, obsolete; anterior median pit much wider
than median groove, latter tapered posteriorly; male abdomen without median pollinose
strip . . C. moldenkei Bell and Bell

The inclusion of C. gilloglyi in the beccarii group alters the definition of the
latter. The structure of the eye and the antennal stylet remain as group characters,
but the new species has antennal tufts and the male lacks tubercles on abdominal
sterna III and IV, so lack of tufts and presence of tubercles are removed from the
list.

The close similarity between C. gilloglyi and C argus suggests that these two
species might be sister species. This would imply that the minor setae were lost
independently in C. argus and in the ancestor of the remaining species. The paired
abdominal tubercles, well developed in the latter species, are very small in C.
argus and absent in C. gilloglyi^ suggestive of a secondary loss in the latter
species.

The discovery of C. gilloglyi in Panama gives added weight to the suspicion
that the type of C. argus, supposedly from the Philippines, is probably mislabeled.

Acknowledgments

We thank the following museum curators for sending loans of specimens: H. Schonmann (NMW),
L. Hendrich (MNHB), M. J. D. Brendell (BMNH), D. Kavanaugh (CAS), W. Schawaller (SMNS), J.
Liebherr (CUIC), E. Riley (TAMU), R. Anderson (CMNO), and J. Rawlins and R. Davidson (CMNH).
We thank also the following entomologists who sent specimens from their private collections: A. R.
Gillogly, G. Onore, A. Riedel, and A. Smetana. We especially thank Dr. G. A. Samuelson of the
Bishop Museum in Honolulu, Hawaii, for returning some paratypes of Rhyzodiastes convergeus for
comparison with the Admiralty Islands specimens. We are particularly grateful to those who allowed
us to retain specimens for our collection. We also want to thank Erika Geiger (UVM) for her accurate
typing of the final manuscript and Mary Morgan and Kimberly Layfield (UVM) for their work on the
final revisions.

Literature Cited

Arrow, G. J. 1901. Remarks upon the Genus Rhysodes, with descriptions of some new Oriental
species. Annals and Magazine of Natural History, 7:83-89.

. 1942. The beetle Family Rhysodidae, with some new species and a key to those at present

known. The Proceedings of the Royal Entomological Society of London (B), 11:171-183.

Bell, R. T. 1970. The Rhysodini of North America, Central America and the West Indies. Miscel-
laneous Publications of the Entomological Society of America, 6:289-324.

. 1973. A new species of Clinidium from Guatemala. Proceedings of the Entomological So-
ciety of Washington, 75:279-282.

. 1977. Ergebnisse der Bhutan-Expedition 1972 des Naturhistorischen Museums in Basel (Col:

Rhysodidae). Entomologica Basiliensia, 2:151-158.

. 1994. Beetles that cannot bite: Functional morphology of the head of adult rhysodines

(Coloptera: Carabidae or Rhysodidae). Canadian Entomologist, 126:667-672.

. 1998. Where do the Rhysodini (Coleoptera) belong? Pp. 261-272, in Museo Regionale di

Scienze Natural! Torino, Phylogeny and Classification of Caraboidea, XX International Congress
of Entomology, August 1996, Florence, Italy.

Bell, R. T, and J. R. Bell. 1979. Rhysodini of the world. Part II, Revisions of the smaller genera
(Coleoptera: Carabidae or Rhysodidae). Quaestiones Entomologicae, 15:377-446.

. 1982. Rhysodini of the world. Part III, Revisions of Ornoglyrnmius Ganglbauer (Coleoptera:

Carabidae or Rhysodidae) and substitutions for preoccupied names. Quaestiones Entomologicae,

18:124-259.

. 1985. Rhysodini of the world. Part IV, Revisions of Rhyzodiastes Fairmaire and Clinidium

Kirby, with new species in other genera (Coleoptera: Carabidae or Rhysodidae). Quaestiones
Entomologicae, 21:1-172.

2000

Bell and Bell — New Rhysodine Beetles

91

. 1987. Rhysodine beetles in the Geneva collection: A new species of Yamatosa and a major

range extension for Omoglymmius sakuraii Nakane (Coleoptera: Carabidae or Rhysodidae). Revue
Suisse de Zoologie, 94:683-686.

. 1993. Rhysodine beetles (Insecta: Coleoptera: Carabidae or Rhysodidae): New species, new

data, and revised keys to Omoglymmius (Subgenera Omoglymmius) and Pyxiglymmius. Annals of
Carnegie Museum, 62:165-185.

Grouvelle, a. 1903. Synopsis de Rhysodides et descriptions d’especes nouvelles. Revue
d’Entomologie, 22:85-148.

Hovorka, O. 1997. New Clinidium species from Ecuador (Coleoptera: Carabidae: Rhysodini). Acta
Societatis Zoologicae Bohemiae, 61:19-22.

Lewis, G. 1888. On the Family Rhysodidae. XL The Annals and Magazine of Natural History, 2:76-

85.

Nakane, T. 1978. New or little-known Coleoptera from Japan and its adjacent regions XXIX (in
English). Reports from the Faculty of Science, Kagoshima University (Earth Sciences and Biol-
ogy), 11:129-134.

, ^ U 'TiilfW t »'».jr.’iu»'T-^, K l^|Mim • | • -i - ^v. - •-n

‘ -M • ■• .V.'O"- .iU- r*.-TAV:...

Y* ^ ^ ^ ’- ^ f* ; ' ‘ •^*1' ■'* Nt

•‘ ■ .‘lip

« .. ■ -J. • w

A.„ '■> ' 'f^'

■*’’*■* ' , ( *■'■ r-' j ***'^

**■.’’ j v;’ <y ./5I' '■ ’ ’ ; ■. ■*'!*',

t’ ,^**

’ ’ . ■ ■-} ■■ i»%^r *.' /

it I ^

. .«*

»•< ’

t‘ V'

••<■' V (-‘‘

1’ . • .'■ . ‘ .v.'" ■ ilv ’v.v; *.*^i.tr

> • ^■' iH:^ *i^>^'’snrj^ fci ■•* *

' ;.. -tv U'\ pt.. ™ • ■ : •; ^

. V " ’ vr- ^.'

tAlh’.'Ilt V ■ ' ■*' '■ ./-;■' ‘f!?*

•v }.i;>* VJ' A. ViTt‘^-^ •''' i'* ■ A j> f ■Jlufi -.r'

>Af’\ ^ur.ni^;,i*! , .»•,,. , • f,\.,-.^j,, :fcrpii^>p^:r . .a

••>..« • • ■.• ! ,'-• .»! Nt: K# '■ ,.

* .-., ■ . ,- ^ {■••• w ^ . . ■ . >-•• <, ^•-..»fy^

; ' -■ .-, 6-

m.

«?i^.

>'4. ^

iikjikji 'J * V ■'*. »•

— ' } \ ^ '^ . ■ ■

t'i, . -■■. * ■ ’ ■' ’ •< ■4'. . . ' *iv , j' » . ■5'.,,

‘ I "-^V ■ I A '(%'*•.- -• I

■■ g. ?-» ■ -wfcy.H '

f . " * 2 • '// 'tif «*■•■'

M - T Vr's V* .(.‘'■V-if. -ifii'wi iftr.

S ^ if^.
H'l-’ • ■-

• .'?.v

* i i

’ fif ' •■., <,/,,■ fs», ' 4 .

■ ‘ • . r- -^ ' Tft.J^r^vUtA

■ ■ i'» .4’ ■ -fc

Tft..4r>.utA -

r* ■ - i '" *‘i» -'fp5^wvr>;.4. Vkv^^i i>|l 't.V

: ri

■.«5g ' iStiv.5»>'t‘ T • att-lVu*^ rM4ffe|u/fn’t5n|ii!a|!(^'.. • ' ’'.-i •■',»

'“ ^ ■.■x<- . ■ •'‘* :;>;•■■ K'=.

,;-.-ivliB;,ii,^tij|n^itei^. -'■, ■ ’•■■ ; V n^v-.-v ,-.• ( :.■(»);•'

J, •ifc'^,fwdt''vj.; '■■!-•. ^» jiipSi ■ '>i . r-%^'f\hvinu\ '-..•.-^r

ANNALS OF CARNEGIE MUSEUM

VoL. 69, Number 2, Pp. 93-134

23 May 2000

THE MICROTINE RODENTS FROM THE PIT LOCALITY IN PORCUPINE
CAVE, PARK COUNTY, COLORADO

Christopher J. Bell^

Anthony D. Barnosky^

Abstract

This report presents the results of an analysis of the microtine rodents from the Pit locality, one of
four localities within Porcupine Cave that have a relatively long stratigraphic sequence. At 2900 m
elevation. Porcupine Cave is the highest elevation site in North America to have produced a diverse
microtine rodent assemblage. At least 11 species are distributed through 14 stratigraphic levels: Phen-
acomys gryci, Phenacomys sp., Mimomys (Cromeromys) cf. M. virginianus, Ondatra sp., Allophaiomys
pUocaenicus, Terricola meadensis, Mictomys cf. M. meltoni, Microtus paroperarius, Microtus sp. (not
M. paroperarius), Lemmiscus curtatus, and Lemmiscus sp. All but one of these {Phenacomys sp. not
P. gryci) occur within a single stratigraphic level (level 4), making the assemblage of microtines
unique in species composition and among the most diverse known from any fossil deposit. The
diversity of the assemblage probably results from the location of the site within a topographically
diverse ecotonal region that allowed sampling of a wide range of nearby microhabitats by the carni-
vores and raptors that initially collected the bones. The unique species assemblage is likely related to
the high elevation of the site, which differentiates it physiographically from any other fossil deposit
that has produced microtine rodents in abundant numbers, and perhaps because the Rocky Mountain
backbone served as both a dispersal corridor and refugium as taxa adjusted biogeographic ranges in
response to glacial-interglacial transitions and other climatic changes.

The particular assemblage of species, when compared with previously dated occurrences of micro-
tine species, suggests that levels 4-8 of the excavation date to between approximately 750,000 and
850,000 YBR This age assessment is consistent with paleomagnetic data which indicate that level 8
is older than the Bruhnes-Matuyama boundary (780,000 YBP). The microtine biochronology suggests
that the upper levels of the Pit (levels 1-3), date somewhere between 250,000 and 780,000 YBP, but
it is not yet possible to be more precise.

The new data presented herein are important in revising previously reported interpretations of the
age of the Pit locality (e.g., the level 4/3 transition is now thought to be between 750,000 and 850,000
YBP rather than the previous interpretation of near 400,000 YBP). In addition, these data document
the occurrence of a unique microtine species assemblage which has implications for biochronological
schemes that rely on microtines, and provide a framework for interpreting the timing and effects of
climate change on the other 70+ species that occur in this very important middle Pleistocene site.

Key Words: Irvingtonian, Arvicolinae, biochronology, climate change

Introduction

Porcupine Cave is situated high in the Rocky Mountains in South Park, Park
County, Colorado. At 2900 m elevation, it is one of the highest fossil vertebrate
sites of Pleistocene age in North America, and has the most taxonomically diverse
fauna of any high=eIevation site on the continent. So far at least 80 species of

^Department of Integrative Biology and Museum of Vertebrate Zoology, University of California,
Berkeley, California 94720. Present address: Department of Geological Sciences, University of Texas
at Austin, Austin, Texas 78712. cjbeil@mail.utexas.edu.

^Mountain Research Center, Montana State University, Bozeman, Montana, 59717. Present address:
Museum of Paleontology and Department of Integrative Biology, University of California, Berkeley,
California 94720. tonyb@fossil.berkeley.edu.

Submitted 16 March 1998.

93

94

Annals of Carnegie Museum

VOL. 69

mammals, birds, reptiles, and amphibians have been identified. Ongoing research
in the cave continues to reveal a tremendous wealth of fossil vertebrate material,
and the site is rapidly earning its place as the most important middle Pleistocene
(Irvingtonian) locality in North America.

A history of the discovery and early investigations of Porcupine Cave was
presented by Bamosky and Rasmussen (1988). Continuing investigations at the
cave led to discovery of new localities and faunas, with a concomitant dramatic
increase in the number of specimens recovered. In spite of these advances, the
temporal span of the faunas, determination of the specific age of any given locality
within the cave, and the correlation between discrete faunas and localities within
the cave remain elusive goals. The most reliable age estimates were and are
generated through analysis of the biochronologic implications of the fauna, despite
efforts to apply paleomagnetic and electron-spin resonance dating techniques. Mi-
crotine rodents have been used for the last 50 years as a primary tool in construc-
tion of regional and continental scale biochronologies throughout the Holarctic,
and biochronologic age estimates for the various deposits in Porcupine Cave are
based on the stage of evolution and taxonomic composition of the microtines.
The temporal resolutions reportedly obtainable through careful study of microtines
in the Pliocene and Pleistocene range from on the order of hundreds of thousands
of years to tens of thousands of years, making them a powerful tool for relative
age determinations in a time period plagued by a lack of broadly applicable
radiometric dating techniques and the difficulty of correlating to the geomagnetic
polarity time scale (see below).

In this paper, we present a complete summary of the microtine rodent fauna
from the Pit locality within Porcupine Cave that was excavated by the Carnegie
Museum of Natural History from 1985 to 1990. Although work by the Denver
Museum of Natural History continues in the cave today, the Pit excavation is
now completed and that phase of the project is ended. Data from the Pit sequence
provide a baseline against which other deposits in the cave can be evaluated as
they are completely excavated and studied. The Pit locality represents one of only
four excavation sequences within Porcupine Cave from which faunas can be
placed in precise superpositional relationship. Our goals here extend beyond mere
identification of species and documentation of the change in taxonomic compo-
sition in microtines at the site through time. We specifically address some of the
problems encountered in attempts to place the Porcupine Cave faunas within the
existing microtine rodent biochronological framework, present estimates of the
temporal span of the Pit sequence, and discuss implications of these data for age
determinations elsewhere. The data presented here represent a more complete
summary of the fauna and update all previously published reports on microtine
rodents from the Pit (Bamosky and Rasmussen, 1988; Wood and Bamosky, 1994;
Barnosky et ak, 1996a, I996b\ Bell, 1997, 1998).

Abbreviations

CM, Carnegie Museum of Natural History, Pittsburgh, Pennsylvania; DMNH,
Denver Museum of Natural History, Denver, Colorado; KU, University of Kansas
Museum of Natural History, Lawrence, Kansas; SBCM, San Bernardino County
Museum, Redlands, California; UCMP, University of California Museum of Pa-
leontology, Berkeley, California; USNM, United States National Museum, Wash-
ington, D.C.

2000

Bell and Barnosky — Porcupine Cave Microtines

95

Table 1. — Institutional localities and locality numbers for vertebrate fossils recovered from different
excavation units in Porcupine Cave. * denotes multiple discrete excavations from a single room; NA
= locality number not assigned as of this writing; — = no collection housed at institution.

Locality

CM

DMNH

KU

UCMP

Badger Room

1928

942/1351

—

V93176

Come-Along Room

—

1353

—

—

Cramped Quarters

—

1346

—

—

Crystal Room

—

1345

—

V94014

Damp Room

1929

NA

—

V93178

Ferret Room

1930

1342

—

V93179

Fissure Dump Pile

—

1348

—

—

Fissure Fill A

—

1344

—

V98022

Fissure Fill B

NA

—

—

NA

Generator Dome Room

—

1347

—

—

Gypsum Room

1926

—

—

V93174

Gypsum Room SE corner

Site

—

1343

—

—

Gypsum Room Closet

—

—

NA

—

Memorial Day Room

—

1352

—

—

New Passage

1931

—

—

V93177

Pit*

1925

—

NA

V93173

Velvet Room*

1927

644

—

V93175

Velvet Room Mark’s Sink

—

1349

—

—

Velvet Room Will’s Hole

—

1350

—

—

PC General

1932

1354

—

V97002

The Localities in Porcupine Cave

At least 21 separate excavations in Porcupine Cave, conducted by parties rep-
resenting four institutions (Table 1), have yielded vertebrate fossils. Most exca-
vations were test pits or exploratory investigations, but as of 1998 four extensive
excavations have been conducted (or are ongoing) in stratified deposits at least
1-m thick. One of these excavations was in the Pit locality, and the other three
are located in the Velvet Room (Fig. 1). The first excavation within the Velvet
Room was conducted by CM crews and was situated on a slightly elevated portion
of the room along the northern wall. A second, more expansive excavation was
opened by the DMNH on the west wall next to the entry passage into the room.
This excavation is ongoing and continues to produce a wealth of material. A third
excavation, which began as an exploration pit seeking a passage to another room,
recently produced what appear to be some of the oldest vertebrate fossils from
the cave. Detailed stratigraphic excavation of this pit, named Mark’s Sink, was
begun by DMNH crews during the 1996 field season.

Background and Previous Research

Since the first publication about Porcupine Cave fossils (Barnosky and Ras-
mussen, 1988), papers or theses have been completed about the carnivores from
all localities within the cave (Anderson, 1994, 1996), the prairie dogs from the
Pit (Rouse, 1997), and climatic implications of portions of the Pit fauna (Wood
and Barnosky, 1994; Barnosky et al., 1996b).

In 1981, Don Rasmussen made the first systematic collection of vertebrate
fossils from the cave (in the Gypsum Room). In 1985, small collections were
made by Barnosky and Rasmussen in the Pit and Badger Room localities. Be-

96

Annals of Carnegie Museum

VOL. 69

NORTH

PORCUPINE CAVE

10 METERS

COLORADO

Fig. 1. — Location and map of Porcupine Cave. The map was produced by members of the Colorado
Grotto of the National Speleological Society from a survey that took place in 1987 and was made
available for this report by D. Rasmussen. Selected localities that have produced fossil vertebrates are
labeled with the name of the institution that conducted excavations and at which specimens are de-
posited. Material in this report comes from the Pit locality, in which the location of the approximately
2 m X 4 m excavation is plotted.

tween 1986 and 1990, CM crews under the direction of Bamosky conducted full
excavations in the Pit and the Velvet Room and made test excavations and/or
surface collections in the Badger Room, Crystal Room, Damp Room, Gypsum
Room, Ferret Room, Fissure Fill A, Fissure Fill B, and the New Passage Room.
UCMP field parties (again directed by Bamosky) continued work on the localities
originally worked by CM until 1993. In 1987, DMNH crews began a systematic
study that included detailed mapping of the cave and all fossil vertebrate localities,
additional test pit and surface sampling from known localities, the discovery and
documentation of several new localities, and an extensive new excavation in the
Velvet Room. The DMNH study still is underway. In 1994, a small team from
KU conducted a limited excavation adjacent to the CM Pit excavations. Speci-
mens from the various excavations are curated at CM, DMNH, UCMP, and KU.

Stratigraphy

The Pit locality (CM loc. 1925, UCMP loc. V93173) occurs in the northeastern
part of the cave approximately 30 m from the modem cave entrance (Fig. 1). A
grid system consisting of eight 0.9 1-m (3-ft) squares provided horizontal control
of the excavation (Bamosky and Rasmussen, 1988:fig. 2). Each square was ex-
cavated by natural stratigraphic levels or 10-cm intervals (whichever was thinner)
to an approximate depth of 235-252 cm where sediments terminated in a bedrock

2000

Bell and Barnosky — Porcupine Cave Microtines

97

floor. Additional details of excavation and screening procedures were provided
by Barnosky and Rasmussen (1988).

The stratigraphy in the Pit sequence consists primarily of alternating beds of
homogenous, unindurated silts and fine sands (“loess” of Barnosky and Ras=
mussen, 1988) interbedded with more consolidated beds consisting primarily of
hard clay nodules or concretions. At least the larger of the nodules appear to be
formed around a center “core” consisting of a small piece of carbonate or bone.
Some concretions effervesce readily and nearly completely disintegrate in hy-
drochloric acid, indicating carbonate cementation. The Pit stratigraphy is irregu-
larly punctuated with flowstones of varying thickness, including one that appears
to have been a stable surface for some length of time (designated as a “paleo-
floor”; level 8A). Two types of flowstone occurred in the excavation: apparently
pure layers of calcium carbonate (generally less than 5 cm in thickness), and
breccias tightly cemented by a calcium carbonate matrix. The breccias appear to
have been indurated as calcium carbonate-laden waters trickled through them.
Major sedimentological changes in the sequence were interpreted to be the result
of climate change (Barnosky and Rasmussen, 1988) and subsequent faunal anal-
yses supported this conclusion (Wood and Barnosky, 1994; Barnosky et al.,
1996Z?; see below). Fourteen discrete stratigraphic levels were numbered sequen-
tially from the top to the bottom of the section. Preliminary description of strati-
graphic levels 1-9 was provided by Barnosky and Rasmussen (1988), but only
levels 1-4 were entirely excavated at the time that report was written (lower levels
were originally described based on samples from a bucket auger core). Figure 2
presents a schematic summary of the stratigraphic section in the Pit. Descriptions
of levels listed here are in part taken from the discussion in Barnosky and Ras-
mussen (1988:274). Stratigraphic thickness is given as range of depth of level(s)
below a permanent datum established at the surface of the Pit prior to excavation
(datum = 0 cm). Color descriptions were obtained by matching samples to the
Geological Society of America rock color chart (1984).

Levels 1-3. — Moderate yellowish brown dust with a flourlike consistency. Dis-
tributed in a cone that is 81 -cm thick at the southeast comer of square 1 and
tapers to 5 -cm thick approximately 2 m to the northwest. Very soft. 0 cm below
datum at the southeast corner of square 1 to ~76 cm at the northwest corner of
square 2 (Barnosky and Rasmussen, 1988).

Levels 4-5. — Consolidated clay, dark yellowish brown in color. Upper 15 cm
composed of clay concretions (nodules) averaging 2-5 cm in diameter, grading
downward to concretions averaging 1 cm in diameter. Distributed in a cone that
is thickest in the vicinity of square 3 (63 cm) and tapers to the north, east, and
west. Moderately hard. “76 to —139 cm. At the base of level 5 in square 2 a
hard, indurated surface marked the end of the level and provided two paleomag-
netic samples. Depths at the base of level 5 ranged from — 1 12 cm in the southeast
comer of square 2 to “117 cm along the west profile (Barnosky and Rasmussen,
1988).

Level 6. — The top of level 6 was a broad, indurated rocky surface no more
than 5.08 cm thick. Below it was a moderately hard clay mixed with loose dust,
similar in appearance to level 5, but with more white gypsum(?) crystals inter-
mixed. The bottom of the level was marked by large chunks of white crystalline
material and very soft, loose gray dust, with sparse pea-sized nodules, approxi-
mately 20% white crystalline flecks. In screenwashing, about one-half of residue

98

cm

below

datum

Annals of Carnegie Museum

VOL. 69

Loose Dust

8

7

6

5

4

3

2

1

Clay Nodules

each square=0.91 m
Bedrock Indurated Breccia

Fig. 2. — Fence diagram showing the basic stratigraphy of the Pit excavation. The excavation grid at
the bottom is a plan view that enlarges the grid outline plotted on Figure 1. Squares 1, 2, 3, 5, 6, and
7 were excavated. The fence diagram shows schematic stratigraphy for the walls outlined in bold on
the grid; numbers on the walls identify the various stratigraphic levels mentioned in the text. The
north wall of the diagram shows the north wall of squares 7 and 6; the east wall is composed of
squares 6 and 2; the south wall of squares 2 and 3; and the west wall of squares 3 and 7. Stratigraphic
relationships became more complex in the center of the squares; see the explanation in the text and
the diagonal cross section through square 2 that was illustrated as figure 2 in Wood and Barnosky
(1994). Levels 1-3 were so thin and friable that they were not always differentiated in the areas of
the excavation that show in the fence diagram; specimens from levels 1-3 in these areas were labeled
“mixed level 1-3.” However, in square 1 and other portions of squares 2, 3, 5, 6, and 7, levels 1-3
were much thicker and were easily differentiated by depth measurements; these areas provided the
specimens that were stratigraphically assigned to level 1, 2, or 3. In general, layers identified as “loose
dust” are thought to represent warm, arid climates (interglacials), and those labeled “clay nodules”
and “indurated breccia” are thought to represent cool, moist climates (glacials).

2000

Bell and Barnosky — Porcupine Cave Microtines

99

was white crystalline fragments. Depths at base range from — 1 14 cm in southern
edge of square 2 to —130 cm in northeast comer of square 3.

Level 7. — Very soft, loose, gray dust with scattered pea-sized nodules; abundant
white crystalline flecks concentrated at base of level. Depths at base of level range
from — 120 cm in northeast comer of square 6 to — 145 cm at center of west wall
(between squares 3 and 7).

Level 8. — Loosely consolidated gray dust mixed with nodules; laterally abuts
or grades into level 8A.

Level 8A. — ^Indurated surface (paleo-floor), sitting on material similar to levels
1-3 but considerably more consolidated; two paleomagnetic samples were taken
from here. Base depths ranged from —158 cm in southwest corner of square 6 to
— 145 cm in southeast comer of square 3.

Level 9. — Extremely well-cemented level with fractures (crevices); bone brec-
cia in crevices in square 7; part of the bone breccia and paleo-floor is interpreted
as a fossil woodrat nest based on the presence of abundant fossil fecal pellets;
underlain in part by gypsum(?) powder. Depths at base ranged from —153 cm in
northwest corner of square 7 to —162 cm in northeast comer of square 6.

Level 10. — Large (1-5 cm diameter) nodules similar to those in levels 4-5;
reversely graded downwards to pea-sized nodules (0.5 cm diameter); a dusty
matrix increasingly fills voids downwards until matrix becomes more abundant
than nodules at base; in square 6, this level consists of a gypsum(?) powder
(brown/white) similar to the matrix in which nodules are contained in the other
squares; all overlying a hard flowstone. Depths at base of level 10 range from
— 168 cm in southeast comer of square 2 to — 183 cm in center of the west profile
(between square 3 and square 7).

Level 11. — Grayish- white dust; white gypsum crystalline flakes with more
brown dust near the top; becomes whiter downwards. Bottom of level is defined
by first occurrence of pinkish crystalline flakes. Only a slight color change breaks
the level. Depths at base of level 11 range from —202 cm in southeast comer of
square 2 to —212 cm in southwest comer of square 7 (—204 cm elsewhere).

Level 12. — Surface consists of pink crystalline flakes; this level is much more
fine grained than level 11, with a higher percentage of clay-sized material. Almost
100% white powder. About 10-12-cm thick in square 3; base of level marked by
mottled orangeish-brown material. Occasional large rock clasts (up to 14 cm in
diameter). Depths at base of level 12 range from —214 cm on south wall to —212
cm in southwest comer of square 7 (no level 12 in square 7).

Level 13. — Mottled orangeish-brown material with occasional large rock clasts
up to 14 cm in diameter; mostly largely devoid of bone. Depths at base range
from —234 cm in northwest comer of square 7 to —231 in southeast comer of
square 2.

Level 74.— Essentially the same lithology as level 13; an arbitrary level change,
with possible slight darkening in color. Largely devoid of bone. Depth at base of
excavation ranges from -235 cm in northwest comer of square 2 to -252 cm in
southwest comer of square 6.

Results

The Microtine Rodent Fauna

Microtine rodents constitute a major component of the specimens from the Pit
locality. In their preliminary report on the fauna from Porcupine Cave, Barnosky

100

Annals of Carnegie Museum

VOL. 69

Lingual

Fig. 3. — Lower right first molar of Microtus sp., illustrating the dental terminology used in the text.

and Rasmussen (1988) listed eight microtine rodent taxa, seven of which were
recovered from sediments within the Pit sequence. The updated list of 11 taxa
reported here reflects further investigation of the fauna from deeper stratigraphic
levels in the Pit as well as refined taxonomic identifications of some previously
reported material. Dental terminology is illustrated in Figure 3 and follows that
of Repenning (1992) unless otherwise noted.

Systematic Accounts

Family Muridae, Illiger, 1815
Subfamily Arvicolinae, Gray, 1821
Genus Phenacomys, Merriam, 1889
Phenacomys gryci. Repenning, 1987
(Fig. 4C)

Mate rial. —Specimens from the Pit (UCMP V93173, CM 1925): UCMP 155193 (left Mj, level 3);
UCMP 155194 (left M„ level 4); UCMP 155195 (right M;, level 10).

Identification. — Molars rooted and lacking cementum in the reentrant angles;
lingual alternating triangles and reentrants are asymmetrically elongated; usually
only five triangles present on M,, the fourth (second labial) usually preserving at
least some indication of a ''Mimomys kante” and the fifth (third lingual) broadly
confluent with a highly variable anterior cap (Repenning et al., 1987; Repenning
and Grady, 1988).

Discussion.-M^s from the type population of P. gryci in the Fish Creek fauna
of Alaska show a highly variable anteroconid complex (Repenning et al., 1987).
USNM 264298 and 264299 (illustrated in Repenning et al., 1987:fig. 6D, E) show
the closest resemblance to specimens identified from Porcupine Cave. The spec-
imens from Porcupine Cave conform well with the diagnosis of P. gryci, but

2000

Bell and Barnosky — Porcupine Cave Microtines

101

Fig. 4.— Lower first molars of microtine rodents from Porcupine Cave. A. Occlusal view of CM 66365,
Phenacomys sp., right M,. B. Labial view of CM 66365, Phenacomys sp., right C. Occlusal view
of UCMP 155195, Phenacomys gryci, right Mj. D. Occlusal view of UCMP 155566, Mimomys (Crom-

eromys) cf. M. virginianus, left M,. Scale bars = 1 mm.

show a very reduced (in one case absent) ''Mimomys kante” structure on the
fourth triangle. The type population from the Fish Creek fauna in Alaska is re-
ported to date to approximately 2.4 Ma (Repenning et ah, 1987). The species was
also reported from the Froman Ferry sequence in Idaho, where its first appearance
(at approximately 1.6 Ma) was used to define the beginning of the Irvingtonian
component of that fauna (Repenning et ah, 1995); it is also reported from Cathe-
dral Cave in Nevada, in an Irvingtonian fauna with no precise age control (Bell,
1995).

Phenacomys sp. (not P. gryci)

(Fig. 4A, B)

Specimens from the Pit (UCMP V93173, CM 1925): CM 66365 (right M^, level 1);
UCMP 155192 (right M„ level 2).

Identification. — Molars rooted and lacking cementum in the reentrant angles;
lingual alternating triangles and reentrants show pronounced asymmetrical elon-
gation; five fully closed triangles on Mj, with a sixth well developed and confluent
with a hooked anterior cap; no indication of a "Mimomys kante” structure.

102

Annals of Carnegie Museum

VOL. 69

Discussion. — These specimens differ from any known P. gryci in having a
greater number of triangles on There are no intermediate forms between P.
gryci and these Phenacomys specimens preserved in the Pit sequence and a direct
ancestor-descendant relationship between these specimens and the older P. gryci
cannot be demonstrated. Repenning and Grady (1988) relied upon labial dentine
tract morphology on Mj to diagnose three subgenera of Phenacomys. Our attempts
to apply their methodology reveal that there may be greater variation in the mor-
phology of dentine tracts on Mj than was previously recognized. This, combined
with the fact that on two occasions one of us (CJB) was unable to replicate
subgeneric identifications made by Repenning, inspires conservatism in our iden-
tification of this material until a more thorough study of all known Phenacomys
can be completed. Other Irvingtonian reports of Phenacomys include P. brachy-
odus from the Cheetah Room fauna in West Virginia (Repenning and Grady,
1988) and Phenacomys sp. from Cumberland Cave in Maryland (Guilday, 1971),
the Little Sioux fauna in Iowa (Guilday and Parmalee, 1972; referred to as “Coun-
ty Line” fauna by Repenning, 1987) and from Wilson Valley in Kansas (Hibbard,
1944; Guilday and Parmalee, 1972; this appears to be the record referred to the
“Cudahy fauna” by Hibbard, 1970 [see also Hibbard, 1976]). The genus is ques-
tionably reported from the SAM Cave fauna in New Mexico (Repenning, 1992:
33). Repenning (1987) listed Phenacomys sp. from the Java fauna in South Dakota
but R. Martin (1989) identified this material as a new species of Hibbardomys.

Genus Mimomys, Forsyth-Major, 1902
Subgenus Cromeromys, Zazhigin, 1980
Mimomys cf. M. virginianus. Repenning and Grady, 1988

(Fig. 4D)

Material. — Specimens from the Pit (UCMP V93173, CM 1925): UCMP 155188 (right Mj fragment,
level 5); UCMP 155189 (left dentary with Mi_2, level 6); UCMP 155191 (left M, fragment, level 8A);
UCMP 155566 (left Mj, level 8A); UCMP 156158, 156159 (one left and one right M', level 5);
UCMP 156160, 156161 (one left and one right M', level 7); UCMP 156162, 156163 (one left and
one right M', level 10); UCMP 156164 (left M^, level 5); UCMP 156165 (right M^, level 7); UCMP
156166 (right M^, level 8A); UCMP 155190 (left M^ level 7); UCMP 156167 (right M2, level 4);
UCMP 156168 (left M2, level 8A); UCMP 156169 (left M,, level 8 A).

Identification. — Molars rooted, with cementum in the reentrant angles; usually
five triangles present on M,, the fifth broadly confluent with a variable anterior
cap; at least some expression of a '"Mimomys kante” is present on Mj, and is
frequently well developed in a position slightly anterior to the fourth triangle.
Discussion. — The status of the genus Mimomys in North America is a subject of
some debate. In his long-standing effort to document intercontinental dispersal
patterns. Repenning (1980, 1987; Repenning and Grady, 1988; Repenning et al.,
1990) recognized the morphological similarities between North American and
Eurasian forms and emphasized an Old- World/New-World connection by retaining
the name Mimomys for the North American forms (although this is done in the
absence of direct evidence for an immigrant status in North America of most of
the taxa in question). An alternative view recognizes the unique morphological
features (notably enamel microstructure or schmelzmuster) and emphasizes the
endemic evolution seen within many of the North American lineages subsequent
to their first appearance on the continent (von Koenigswald, 1980; von Koenig-
swald and Martin, 1984a; L. Martin, 1989). In the latter scheme, the different

2000

Bell and Barnosky — Porcupine Cave Microtines

103

lineages traditionally recognized by Repenning as subgenera are elevated to full
generic rank (Cosomys, Ogmodontomys, Ophiomys, Cromeromys).

The first report of a M/mowys-like microtine from the Pleistocene of North
America was published in 1972 under the name Mimomys monahani (L. Martin,
1972). This species was later placed in the monotypic genus Loupomys based on
the primitive nature of its enamel microstructure (von Koenigswald and Martin,
1984<2); it is known only from the type locality in Nebraska. The presence of a
Mimomys-\\k& species in the Irvingtonian fauna of Hamilton Cave, West Virginia,
was noted by Repenning and Grady (1988), who placed this material in a new
species which they named Mimomys virginianus. A third record of a Pleistocene
Mimomys came from the Java fauna in South Dakota (R. Martin, 1989) and also
received a new name, Mimomys dakotaensis. The age of the Java fauna is some-
what problematic as it contains several taxa typical of the Blancan; R. Martin
(1989) considered it to be earliest Pleistocene and Repenning et al. (1990, 1995)
placed it in the Irvingtonian I microtine rodent division. Of the two Mimomys
M,s included in the Java fauna, one (designated as the holotype) is broken; the
second is not illustrated or discussed. Mimomys dakotaensis is distinguished from
M. virginianus by its size (larger than M. virginianus, although the measurement
of only one specimen from Java was published), by having significantly less
cementum in the reentrant angles and by the extremely high dentine tract at the
position of the Mimomys kante. The amount of cementum in specimens from
Porcupine Cave is more similar to that in M. virginianus than in M. dakotaensis.
Tentative referral to M. virginianus is made pending detailed examination of the
schmelzmuster of the Porcupine Cave material and comparisons with the complete
Hamilton Cave sample (which is not published in its entirety; C. Repenning,
personal communication). Other records of this species are from Cathedral Cave
in Nevada (Bell, 1995, as '"Mimomys {Cromeromys^ sp.”) and from several lo-
calities in the Yukon Territory (Repenning, personal communication).

Loupomys monahani, Mimomys virginianus, and M. dakotaensis share a re-
markably similar morphological appearance, most notably in the presence of ce-
mentum in the reentrant angles (distinguishing these species from all other North
American MimomyS'\ik& taxa). However, the schmelzmuster of Loupomys appears
to be quite primitive and is unique among North American microtines (von Ko-
enigswald and Martin, 1984<2). The schmelzmuster of M. dakotaensis was reported
to be similar to that of typical European species of Mimomys (R. Martin, 1989);
the schmelzmuster of M. virginianus and the Pit material has not been studied.

Mimomys virginianus was placed in the subgenus Cromeromys based on mor-
phological similarity with Old World forms assigned to the genus Cromeromys',
Repenning and Grady (1988:5) relegated Cromeromys to a subgenus to achieve
“a more balanced classification” with other North American forms that were
placed within various subgenera under the genus Mimomys. R. Martin (1989)
refrained from placing M. dakotaensis within any subgenus, but the presence of
cementum in the reentrant angles clearly argues for its alliance with M. virgini-
anus. R. Martin’s (1989) concern for subgeneric allocation of the North American
Mimomys having cementum in the reentrant angles stemmed from the fact that
Repenning and Grady (1988) did not distinguish Mimomys {Cromeromys) from
Mimomys (Pusillomimus), a subgeneric designation proposed by Rabeder (1981).

104

Annals of Carnegie Museum

VOL. 69

Genus Ondatra, Link, 1795
Ondatra sp.

Material. SpQcimQns from the Pit (UCMP V93173, CM 1925): UCMP 155567 (broken juvenile
right Ml, level 2); UCMP 155848 (right M', level 2); UCMP 155849 (left level 4); UCMP 155850
(right M2, level 1 or 2); CM 66522 (left M^, level 2); UCMP 155851 (left M^, level 4); CM 45409
(right M\ level 1); UCMP 155852, 155853 (one left and one right M^, level 4); UCMP 155854 (left
M^, level 5); UCMP 155855 (right M^ level 8A); UCMP 155856 (left M2 fragment, level 4).

Identification. — Molars rooted, with reduced cementum in reentrant angles, of-
ten forming a “network” pattern in lateral view; this feature, combined with large
size, distinguishes this genus from all other rooted microtines.

Discussion. — Muskrats are rare in the Pit fauna and unfortunately only one
(the most important tooth for identifying Ondatra species) specimen is available.
The Pit Ml is badly broken and is from a juvenile, so identification to species
cannot reliably be made. Muskrat specimens from Porcupine Cave were tenta-
tively referred to Ondatra cf. O. annectens by Bamosky and Rasmussen (1988);
this tentative identification may be correct, but definitive assignment to species
must await recovery of additional material. All known Pleistocene muskrats were
recently synonymizsed with the living species Ondatra zibethicus (R. Martin,
1993, 1996), yet formerly recognized species names were preserved and appended
to the specific name to form a trinomial nomenclature indicative of what Martin
calls “chronomorphs” (see also Krishtalka and Stucky, 1985).

Genus Allophaiomys, Kormos, 1933
Allophaiomys pliocaenicus, Kormos, 1933
(Fig. 5A, B)

Specimens from the Pit (UCMP V93173, CM 1925): UCMP 155000-155005 (three
right and three left Mj, level 4); UCMP 155006-155008 (two right and one left Mj, level 5); UCMP
155009, 155010 (two right M„ level 6); UCMP 155564 (right M„ level 6); UCMP 155011-155014
(two right and two left M,, level 7); UCMP 155015 (right dentary with I, and Mi_2, level 8A); UCMP
155016 (right dentary with I, and M,_3, level 8A); UCMP 155017 (left dentary with I, and Mi_25 level
8A); UCMP 155018, 155019 (one right and one left M„ level 8A); UCMP 155020, 155021 (two right
Ml, level 11).

Identification. — Rootless, ever-growing molars; Mi with a posterior loop and
three alternating closed triangles; primary wings well developed and broadly con-
fluent with a very simple anterior cap.

Discussion. — Allophaiomys plays an important role in the biochronology of
microtine rodents in Eurasia as well as in North America, where its complicated
history was recently reviewed by Repenning (1992). In North America, its first
appearance marks the lower boundary of the Irvingtonian I microtine rodent di-
vision east of the Rocky Mountains at approximately 1.9 Ma (Repenning, 1987;
Repenning et ah, 1990, 1995).

Allophaiomys pliocaenicus was originally described from Romania (Kormos,
1933) and is known from fossil faunas throughout much of the Holarctic (Re-
penning, 1992). Traditionally, A. pliocaenicus was viewed by most authors as a
single extinct species that had an expansive geographic distribution across the
Holarctic. More recently, evidence has come to light to suggest that its history,
especially in Europe, may be much more complex than previously thought and
that traditional treatments may actually be conflating members of a group of

2000

Bell and Barnosky — Porcupine Cave Microtines

105

closely related sibling species (Agusti, 1992; Agusti et al., 1993; Nadachowski
and Zagorodnyuk, 1996; see also R. Martin and Tesakov, 1988).

The oldest North American specimen associated with some form of external
age control was recovered from a core drilled in the Hansen Bluff area of southern
Colorado (Rogers et ak, 1992). The first magnetic polarity sample taken from
above the horizon in which the specimen was recovered (0.9 m above the spec-
imen) was normal and interpreted to be the lowest record of the Olduvai event
in the core; the first polarity sample from stratigraphically beneath the specimen
(0.3 m below) was reversed. The specimen therefore dates to either the very base
of the Olduvai Chron (at approximately 1.9 Ma) or just before it. The tooth in
question is an isolated M3 and cannot be positively identified as Allophaiomys.
Repenning (1992:32) identified the specimen and argued that the existence of any
other rootless microtine of that age and with that dental morphology (rootless and
having cement in the reentrant angles) in North America was “unbelievable.”

Other early North American records are from the Java fauna in South Dakota
(considered to be earliest Pleistocene; R. Martin, 1975, 1989) and the Wellsch
Valley fauna in Saskatchewan, Canada. The age of Wellsch Valley is not well
established; the sediments record the 780,000 YBP Brunhes/Matuyama boundary
and the fauna may be either latest Blancan or (more likely) earliest Irvingtonian
in age (Churcher, 1984; Barendregt et al., 1991; Repenning, 1992).

Two records were discovered since the completion of Repenning’s (1992)
review of North American Allophaiomys: Cathedral Cave in Nevada (Bell, 1995)
and the Little Dell fauna in northern Utah (Gillette et al., 1999). These records
are significant in that they document for the first time the presence of Allo-
phaiomys in faunas west of the Rocky Mountains, and lessen the distinction
between Irvingtonian faunas from the western and eastern United States faunal
regions proposed by Fejfar and Repenning (1992). The presence of this species
in the Porcupine Cave faunal assemblage was first recognized by Keesing
(1992).

Genus Terricola, Fatio, 1867
Terricola meadensis, (Hibbard, 1944)

(Fig. 5C, D)

Material. — Specimens from the Pit (UCMP V93173, CM 1925): UCMP 155196 (left dentary with
M,_2, surface); UCMP 155197-155199 (two left and one right M;, surface); UCMP 155200-155206
(four left and three right M„ level 1); CM 45402-45404 (three left M„ level 1); CM 45428, 45429
(two left Ml, level 1); CM 66307 (broken left Mi, level 1); CM 45405 (right M,, level 1); CM 45430
(broken right M„ level 1); CM 66199 (broken right Mi, level 1); UCMP 155207-155221 (five left
and ten right M,, level 2); UCMP 155222 (right dentary with M,_2, level 2); CM 45431-45434 (four
left Ml, level 2); CM 63601 (left M„ level 2); CM 66542 (left M„ level 2); CM 45435-45444 (ten
right Ml, level 2); CM 45446, 45447 (two right Mj, level 2); CM 66530 (right M„ level 2); CM
66560 (right M„ level 2); CM 66571 (right M„ level 2); UCMP 155223 (left M„ level 3); UCMP
155224 (right dentary with M„ level 3); UCMP 155225, 155226 (two right M„ level 3); UCMP
155227 (right dentary with Mi_2, level 3); UCMP 155228 (left Mj, level 3); UCMP 155229 (right
dentary with I, and Mi_3, level 3); UCMP 155230, 155231 (two2 right Mj, level 3); UCMP 155264
(left dentary with Ii and M,, level 3); CM 65558 (broken left Mj, level 3); CM 65563 (broken right
M„ level 3); CM 65587, 65588 (two right Mj, level 3); UCMP 155232-155243 (three left M[, two
broken left M,, four right Mj, and three broken right M,, level 4); UCMP 155565 (right M,, level 4);
UCMP 155244-155248 (two left Mj, one left M, fragment, and two right Mi, level 5); UCMP 155249-
155252 (one broken left M, and three broken right M,, level 6); UCMP 155253 (right M,, level 7);
UCMP 155847 (broken right Mj, level 8); UCMP 155254 (right M„ level 8A); UCMP 155255 (left
M„ level 1-3 undifferentiated); UCMP 155256-155259 (four left M„ level mixed).

106

Annals of Carnegie Museum

VOL. 69

Fig. 5. — Occlusal view of right lower first molars of microtine rodents from Porcupine Cave.
A. UCMP 155009, AUophaiomys pliocaenicus. B. UCMP 155012, Allophaiomys pliocaenicus.
C. UCMP 155206, Terricola mecidensis. D. UCMP 155227, Terricola meadensis.

Identification. — Rootless molars with cementum in the reentrant angles; M,
with a posterior loop and three closed triangles; primary wings (triangles 4 and
5) are confluent in a rhomb which is either closed off from the anterior cap or
connected to it via a narrow confluence; secondary wings well developed, with
labial reentrant angle 4 and lingual reentrant angle 5 well developed.

Discussion. — Clear recognition of the morphological distinction between the
dentition of pitymyines from that of Microtus dates back to Hinton (1923), but
there is no consensus among modem microtine scholars as to the taxonomic
treatment and systematic relationships of the group (Van Der Meulen, 1978; Re-
penning, 1983, 1992; R. Martin, 1987, 1993, 1995; Harris, 1988; Moore and
Janecek, 1990). We follow Repenning (1992) in recognizing Terricola as a valid
genus. Specimens of Terricola have been recovered from numerous localities
throughout the western two-thirds of North America, but are not known from east
of the Mississippi River (Repenning, 1983, 1992). The appearance of Terricola
meadensis in the fossil record of North America at approximately 850,000 YBP
is traditionally used to define the base of the Irvingtonian II microtine rodent
division (Repenning, 1987; Repenning et ah, 1990).

2000

Bell and Barnosky — Porcupine Cave Microtines

107

Fig. 6. — Occlusal view of lower first molars of microtine rodents from Porcupine Cave.
A. UCMP 155177, Mictomys cf. M. meltoni, right Mj. B. UCMP 155262, Lasiopodomys morphotype
of Microtus paroperarius, right Mj. C. UCMP 155562, Microtus paroperarius, right Mj. D. CM 66345,
M. paroperarius, left Mi. Scale bars = 1 mm.

Genus Mictomys, True, 1894
Mictomys cf. M. meltoni, (Paulson, 1961)

(Fig. 6A)

Madera/.— Specimens from the Pit (UCMP V93173, CM 1925): UCMP 155102 (right M„ surface);

CM 45414 (right dentary with Ii and Mi_2, level 1); CM 45415 (left M,, level 1, specimen missing);
CM 45416 (left Mi, level 1); UCMP 155103 (right dentary with li and M,.2, level 2); CM 45509,
45510 (two right Mi, level 2); UCMP 155104 (broken left Mi, level 2); UCMP 155105 (left M„ level
3?); UCMP 155106 (left Mi, level 3); UCMP 155107 (left Mi fragment, level 3); UCMP 155108 (left
Ml, level 3); UCMP 155109 (left M, fragment, level 3); UCMP 155110 (left level 3); UCMP
155111 (right Mi, level 3); UCMP 155112 (right dentary with Ij, Mi_2, and M3 fragment, level 3);
UCMP 155113 (right Mi fragment, level 3); UCMP 155841, 155842 (two left Mi, level 3); CM 65572
(right Ml, level 3); UCMP 155114-155120 (seven left M,, level 4); UCMP 155121 (broken left Mi,
level 4); UCMP 155122, 155123 (two left Mi, level 4); UCMP 155563 (left Mi, level 4); UCMP
155124-155127 (two left Mi fragments and two right Mj fragments, level 4); UCMP 155128-155134
(seven right M,, level 4); UCMP 155135-155137 (three broken right Mi, level 4); UCMP 155138-
155141 (four right Mi, level 4); UCMP 155142, 155143 (two right Mi fragments, level 4); UCMP
155144-155150 (two left Mi, three right Mi, and two right Mi fragments, level 5); UCMP 155151
(left dentary with Mi_2, level 5); UCMP 155844 (right Mi, level 5); UCMP 155845 (broken right Mi,
level 5); UCMP 155846 (broken left M,, level 5); UCMP 155152-155157 (three left Mi, one right
M,, one left Mi fragment, and one right Mi fragment, level 6); UCMP 155158-155161 (one right Mi,
three broken right Mi, level 7); UCMP 155162 (right dentary fragment with broken Mi, level 7);
UCMP 155163, 155164 (two right M, fragments, level 7); UCMP 155165 (left Mi, level 7); UCMP

108

Annals of Carnegie Museum

VOL. 69

155843 (left M,, level 7); UCMP 155166-155171 (five left and one right M,, level 8); UCMP 155172-
155177 (two left M,, two right Mj, and two left Mj fragments, level 8A); UCMP 155178-155182
(three left and two right Mj, level 10); UCMP 155183-155185 (one right M, and two right Mj
fragments, level 11); UCMP 155186 (left dentary with M,^2^ level 1-3 undifferentiated); UCMP
155187 (left Ml, level mixed); CM 45417 (left M', level 1); UCMP 155568 (right M', level 3); UCMP
155569-155594 (11 left and 15 right M', level 4); UCMP 155595-155609 (seven left and eight right
M', level 5); UCMP 155610-155616 (five left and two right M', level 6); UCMP 155617-155620
(two left and two right M‘, level 7); UCMP 155621-155625 (two left and three right M‘, level 8);
UCMP 155626-155633 (six left and two right M', level 8A); UCMP 155634-155637 (four left M',
level 10); UCMP 155638-155641 (one left and three right M', level 11); UCMP 155642 (right M',

level 12); UCMP 155644, 155645 (one left and one right M^, level 2); CM 65577 (left M^, level 3);

UCMP 156170 (right M^, level 3); UCMP 155646-155653 (five left and three right M^, level 4);
UCMP 155655-155665 (seven left and four right M^, level 4); UCMP 155666-155676 (seven left
and four right M^, level 5); UCMP 155677-155681 (one left and four right M^, level 6); UCMP
155682-155687 (six right M^, level 7); UCMP 155688-155690 (one left and two right M^, level 8);
UCMP 155691-155694 (two left and two right M^, level 8A); UCMP 155695-155697 (three left M^,
level 10); CM 45418, 45419 (one left and one right M-\ level 1); UCMP 155698 (left M^, level 3);

UCMP 155699-155715 (five left and 12 right level 4); CM 45511 (right M\ level 5); UCMP

155716-155720 (two left and three right M-', level 5); UCMP 155721-155727 (three left and four
right M-\ level 6); UCMP 155728-155733 (three left and three right M^, level 7); UCMP 155734-
155736 (one left and two right M-\ level 8); UCMP 155737 (left M^, level 8A); UCMP 155738 (right
M-\ level 9); UCMP 155739 (right M\ level 10); UCMP 155740, 155741 (two right M^, level 11);
UCMP 155742 (left M‘ or M^ fragment, level 8 A); UCMP 155743 (upper molar fragment, level 4);
UCMP 155744, 155745 (two left M2, level 3); UCMP 155746-155764 (11 left and eight right M2,
level 4); UCMP 155765-155770 (four left and two right M2, level 5); UCMP 155771-155775 (four
left and one right M2, level 6); UCMP 155776-155780 (one left and four right M2, level 7); UCMP
155781-155786 (two left and four right M2, level 8); UCMP 155787 (left M2, level 10); UCMP
155788-155790 (two left and one right M2, level 11); UCMP 155792 (right M3, level 2); UCMP
155793 (right M3, level 3); UCMP 155794-155802 (three left and six right M3, level 4); UCMP
155803-155805 (two left and one right M3, level 5); UCMP 155806-155809 (one left and three right
M3, level 6); UCMP 155810-155812 (one left and two right M3, level 7); UCMP 155813 (right M3,
level 8A); UCMP 155814, 155815 (one left and one right M3, level 10).

Identification. — Rootless molars with cementum in at least the lingual reentrant
angles of lower molars; as noted by von Koenigswald and Martin (1984Z?), the
axis of the lower tooth row is shifted to the labial edge; Mi with a posterior loop,
two closed but not alternating triangles, and a moderately variable anterior cap;
the first labial triangle of Mj (triangle 2 in other microtines) is absent due to
flattening of the first labial reentrant angle; enamel is positively differentiated,
with the anterior enamel band thicker than the posterior enamel band on each
triangle; posterior border of the first triangle is convex anteriorly or flat.

Discussion. — Mictomys kansasensis was named from the Kentuck assemblage
in Kansas (Hibbard, 1952) and was diagnosed primarily by the posteriorly ex-
tended margin of the lower incisor to a position about equal to the posterior edge
of M3. Mictomys meltoni was described from the Cudahy ash pit fauna in Kansas
(Paulson, 1961) and diagnosed by a combination of characters including positively
differentiated enamel, lack of a cementum-filled labial reentrant on M3, a posterior
extension of the lower incisor to a point just anterior to the anterior border of
M3, and a first triangle of Mj with an anteriorly convex (rather than concave)
posterior wall. The Mictomys material from Snowville, Utah (Repenning et ah,
1987; Repenning and Grady, 1988), consists of two specimens (left and right
partial dentaries), both cataloged under UCMP 124887. The left dentary retains
all teeth, the right is missing the M3. In all cheek teeth, enamel is positively
differentiated and the posterior border of the first triangle is concave anteriorly.
The schmelzmuster of these specimens has never been examined and their specific
affinity remains uncertain; they were previously regarded as representing the final

2000

Bell and Barnosky — Porcupine Cave Microtines

109

(and unnamed) intermediate step in the evolution of bog lemmings towards the
extant M. borealis (Repenning and Grady, 1988).

The majority of the Porcupine Cave Mj specimens have a convex or flat pos=
terior margin on the first triangle, similar to that described for M. meltoni, but the
expression of this character in M. kansasensis has not been explored. Examination
of resin casts of four MjS of this species from the Wathena fauna (Van Der
Meulen, 1978) in Kansas indicate that this morphology is at least sometimes
present in populations of M. kansasensis. In the few Porcupine Cave jaws of
Mictomys which preserve the necessary region, the posterior margin of the incisor
terminates at a point anterior to the M3, similar to the condition in the extant M.
borealis and in M. meltoni. Some variation in the expression of this character was
noted by Repenning and Grady (1988), and its reliability as a definitive character
to distinguish M. meltoni and M. kansasensis remains to be proven.

Specimens from Porcupine Cave have not as yet been subjected to an exami-
nation of schmelzmuster, which may be the only reliable way to consistently
distinguish M. meltoni and M. kansasensis (von Koenigswald and Martin, 1984Z?).
Triangles of Mj of Mictomys kansasensis are reported to have an extremely thin
posterior enamel margin that consists almost exclusively of radial enamel (the
exception to this being at the extreme lingual border of the triangle where some
lamellar enamel is present; von Koenigswald and Martin, 1984Z?). The only com-
parison of the schmelzmuster of Mj in these two species of Mictomys was re-
stricted to one specimen of M. kansasensis from each of three localities and one
specimen of M. meltoni from Cudahy ash pit (von Koenigswald and Martin,
1984^). This analysis does not establish a range of variation in enamel micro-
structure within a population and further analysis of multiple specimens from a
single stratigraphic horizon would be necessary to increase confidence in the use
of this character to discriminate between species. The importance of studying
multiple specimens from a single locality cannot be overstated. Few studies of
schmelzmuster in microtines have utilized more than one or two specimens from
a given locality, but a recent study of fossil remains of Microtus henseli from
Sardinia demonstrates that when sample size is increased, increased variation in
enamel microstructure pattern is evident (Mezzabotta et al., 1995).

Based on the morphology of the first triangle on Mj, the posterior margin of
the lower incisor, and the measurements presented by Wood and Barnosky (1994)
for these specimens, we tentatively refer the Porcupine Cave material to Mictomys
cf. M. meltoni. A more definitive identification must await a thorough reexami-
nation of all Irvingtonian Mictomys.

Genus Microtus, Schrank, 1798
Microtus paroperarius, Hibbard, 1944
(Fig. 6B-D)

Material. — Specimens from the Pit (UCMP V93173, CM 1925); CM 66345 (left Mj, level 1);
UCMP 155081-155084 (one left and three right Mj, level 1); UCMP 155085 (right dentary with M]_
2, level 2); CM 45482-45484 (two left and one right level 2); UCMP 155086-155089 (two left
and two right Mj, level 3); UCMP 155090 (left M„ level 3?); CM 45507 (left M^, level 3); CM 65559
(broken left M„ level 3); UCMP 155091-155095 (three left and two right Mi, level 4); UCMP 155262
(right M„ level 4); UCMP 156152 (right Mi, level 4); UCMP 155096 (right Mi, level 5); UCMP
155098, 155099 (one left and one right M„ level 7); UCMP 155562 (right Mi, level 7); UCMP
155100 (left dentary with Ii and M,_2 level mixed); UCMP 155101 (right M,, level mixed).

Identification. — Rootless molars with cementum in the reentrant angles; Mj

110

Annals of Carnegie Museum

VOL. 69

with a posterior loop followed by four closed alternating triangles; a fifth triangle
well developed, but confluent with the anterior cap; triangle 2 smaller than triangle
1 on Mj.

Discussion. — This species was originally diagnosed by Hibbard (1944) by the
presence of a posterior loop and four closed triangles on Mi, with a fifth triangle
confluent with the anterior cap, and with the cap showing evidence of lingual
reentrant angle 5 and labial reentrant angle 4. A subsequent analysis of new
material collected from the type locality resulted in an amended diagnosis (Paub
son, 1961) in which 20% of the population was said to have a closed fifth triangle.
Subsequent authors have followed Paulson’s amended diagnosis of the species
and many descriptions of fossil populations included a percentage of Mi with a
closed fifth triangle (although Paulson’s figure of 20% appears to be an artifact
of his sampling; Bell, 1997; Bell and Repenning, 1999). Until recently, this prac-
tice met with no difficulty because in all reported occurrences M. paroperarius
was the only Microtus species presumed to be present (although the presence of
a second species of Microtus in the Cudahy fauna in Kansas is now suspected;
Bell, 1997; Bell and Repenning, 1999). With the discovery of this species in the
Porcupine Cave assemblage (Barnosky and Rasmussen, 1988) a new challenge is
faced. The Porcupine Cave assemblage contains numerous specimens of Microtus
that cannot be assigned to M. paroperarius. For the purpose of this report, all
specimens identified as M. paroperarius conform with Hibbard’s (1944) primary
criterion in the original diagnosis: the fifth triangle is confluent with the anterior
cap. We recognize the possibility that some small percentage of the specimens
we identify as '"Microtus sp.” may actually be M. paroperarius, but we can find
no way to reliably separate them at this time. Five specimens of M. paroperarius
were previously reported from the Pit locality, with an average length of 2.61 nun
(Barnosky and Rasmussen, 1988). In our larger sample size from the Pit locality
{n = 22) the average length measurement is 2.64 mm. It is interesting that none
of the Porcupine Cave material approaches the large size reported for teeth of M.
paroperarius from the basal portions of the Hansen Bluff sequence in southern
Colorado (Rogers et aL, 1985; Repenning, 1992:57).

A single specimen from the Pit excavation (UCMP 155262) has a relatively
primitive morphotype resembling that of Lasiopodomys deceitensis. The name
Lasiopodomys is applied to three living species found in Russia and China (MuS“
ser and Carleton, 1993) and to numerous fossil populations throughout the Hob
arctic (Repenning, 1992). The species here called L. deceitensis was described
under the name Microtus deceitensis from the Cape Deceit fauna in Alaska (Guth^
rie and Matthews, 1971). This form was given the new name combination by
Repenning (1980; following a precedent set by Erbaeva [1976] for fossil forms
in Asia) and taken by him (Repenning, 1992) as “typical” of Lasiopodomys. In
Repenning ’s usage, Lasiopodomys is retained as a name for “intermediate forms
that have been assigned either to more primitive Allophaiomys or to the more
advanced Microtus"' (Repenning, 1992:47).

Most Irvingtonian records of Lasiopodomys deceitensis are from localities in
the eastern United States, including Hamilton Cave, West Virginia (Repenning
and Grady, 1988), Hanover Quarry, Pennsylvania (Guilday et ab, 1984), the
County Line fauna in Illinois (Miller et aL, 1994), and a population interpreted
to be intermediate in stage of evolution between Lasiopodomys and Microtus
paroperarius from Cumberland Cave, Maryland (Repenning, 1992). A very low
percentage (0.6%) of the Mi with four closed triangles from the type locality of

2000

Bell and Barnosky — Porcupine Cave Microtines

11

Fig. 7. — Occlusal view of lower first molars of Microtus sp. from Porcupine Cave. A. UCMP 155042,
left Ml. B. CM 45502, right C. UCMP 155045, right M,. Scale bars = 1 mm.

the Cudahy fauna also are Lasiopodomys morphotypes, but these are interpreted
as specimens of Microtus paroperarius with relatively primitive morphologies
(Bell, 1997; Bell and Repenning, 1999). There are now three localities from the
southwestern United States containing specimens potentially referable to Lasio-
podomys. A single specimen from the Anza-Borrego Desert in southern California
was described and illustrated by Repenning (1992:50, fig. 11 A). A single speci-
men from Cathedral Cave in Nevada and this record from the Pit locality in
Porcupine Cave constitute the only other records in western North America south
of Beringia. It is difficult to interpret the Anza-Bonego specimen since it is an
isolated find not associated with any other microtine rodent material. Both of the
cave faunas, however, have extensive microtine rodent faunas that include Micro-
tus paroperarius. This is significant since the only difference in the Mj of Lasio-
podomys and that of M. paroperarius is the development of secondary wings and
at least incipient constriction between the fifth triangle and the anterior cap (di-
mension B-B' of Repenning, 1992) in the latter. The specimens from Cathedral
and Porcupine caves are here interpreted as simple morphotypes of Microtus par-
operarius.

Microtus sp. (not M. paroperarius)

(Fig. 7A-C)

Specimens from the Pit (UCMP V93173, CM 1925): UCMP 155022 (left surface);
UCMP 155023 (left dentary with and Mi_2, west of square 1 towards ladder); UCMP 155024 (left
M/, south extension); UCMP 155025-155036 (seven left and five right Mj, level 1); CM 45485 (left
Ml, level 1); CM 45497 (left M„ level 1); CM 66319 (broken left M„ level 1); CM 66338 (left Mi,
level 1); CM 45486 (right Mi, level 1); CM 45498 (right Mi, level 1); CM 45502 (right Mi, level 1);
CM 65232 (right Mi, level 1); CM 66276 (right Mi, level 1); CM 66303 (right Mi, level 1); CM
66348 (right Mi, level 1); CM 66284 (broken left Mi, level 1); UCMP 155037 (broken right Mi, level
1); CM 66202 (broken right M„ level 1); UCMP 155038 (right Mi, level 1); UCMP 155039-155054
(six left and ten right Mi, level 2); UCMP 155055 (right dentary with Mi_2, level 2); UCMP 155056
(right dentary with Mi_2, level 2); UCMP 155261 (right dentary with Mj, level 2); UCMP 156153
(right dentary with I1-M2, level 2); UCMP 155057 (left Mi, level 2); CM 45487, 45488 (two left Mi,
level 2); CM 45503 (left Mi, level 2); CM 63606 (left Mi, level 2); CM 66534 (left Mi, level 2); CM
66593 (left M,, level 2); CM 45445 (right Mi, level 2); CM 45489-45494 (six right Mi, level 2); CM
45499 (right Mi, level 2); CM 66567 (right Mi, level 2); CM 66569 (right Mi, level 2); UCMP 155058
(left dentary with Ii and Mi, level 3); UCMP 155059-155061 (three right Mi, level 3); UCMP 155062,
155063 (two left Mi, level 3?); UCMP 155064 (left Mi, level 3); CM 45495, 45496 (one left and one
right Ml, level 3); UCMP 155561 (right Mi, level 3?); UCMP 155065-155067 (two left and one right
Ml, level 4); UCMP 156154-156156 (two left and one right Mi, level 4); UCMP 155068, 155069
(one left and one right Mi, level 5); UCMP 155070 (right Mi, level 6); UCMP 155071, 155072 (two

112

Annals of Carnegie Museum

VOL. 69

left Ml, level 1-3 undifferentiated); UCMP 155073, 155074 (two right dentaries with Mi^2> level 1-3
undifferentiated)); UCMP 155075-155079 (three left and two right Mj, mixed level); CM 45500,
45501 (two right M,, level mixed); UCMP 155080 (juvenile right M,, level 1).

Identification. — Rootless molars with cementum in the reentrant angles; Mj
with a posterior loop followed by at least five closed alternating triangles; triangle
2 distinctly smaller than triangle 1 on Mj; secondary wings extremely well de=
veloped.

Discussion. — There are at least 12 Microtus species extant in North America
that have Mj with five or six closed alternating triangles. These species include
M. breweri, M. californicus, M. chrotorrhinus, M. longicaudus, M. mexicanus, M.
miurus, M. montanus, M. oregoni, M. pennsylvanicus, M. richardsoni, M. town-
sendii, and M. xanthognathus (Hall, 1981). Quaternary paleontologists have been
struggling to discover means of reliably distinguishing these species in the fossil
record, but little progress has been made. The application of discriminant analysis
and morphometries to Microtus samples in New Mexico showed some promise
for distinguishing the five species included in the analysis (Smartt, 1972, 1977),
but no such study has yet been conducted that includes all North American taxa.

Some studies focused on certain morphological features such as size or the
complexity of as an aid in identification of Microtus species. The size criterion
was used in conjunction with cranial characters from partial skulls and dental
variation to identify the remains of M. xanthognathus in the eastern United States
(Guilday and Bender, 1960; Hallberg et aL, 1974; Guilday et aL, 1977, 1978),
but these studies did not consider M. richardsoni, another large species from the
northwestern United States. Similarly, the presence of with an additional pos-
terior dentine field is often regarded as positive evidence of the presence of M.
pennsylvanicus, yet this morphological feature is irregularly expressed in a num-
ber of different North American species, and is a regular feature in at least two
others (Bell, 1997; Bell and Repenning, 1999). A careful reading of the literature
on North American Quaternary Microtus demonstrates that modem geographic
distribution plays an important role in the taxonomic identification of fossil ma-
terial, if only in the choice of taxa to exclude from possible consideration. This
practice can result in faulty conclusions regarding the biogeographic and temporal
history of Microtus species in North America, and leads to potentially misleading
information that is then incorporated into biochronologies and other paleobiolog-
ical applications.

One specimen listed above (CM 45445) was previously reported as Pitymys
meadensis (= Terricola meadensis by our usage; Bamosky and Rasmussen,
1988). The specimens listed above also include specimens previously reported as
Microtus montanusiM. longicaudus (Bamosky and Rasmussen, 1988; Wood and
Bamosky, 1994; Bamosky et aL, 1996^). Preliminary identification of M. penn-
sylvanicus in the Porcupine Cave Velvet Room fauna was made on the basis of
the presence of M^ specimens showing a well-developed posterolingual dentine
field (Bamosky and Rasmussen, 1988; the species is included in the faunal list
in their table 2, p. 270, but is not discussed in the text). In light of the small
number of specimens with this morphology in the sample, and the infrequent
occurrence of this structure in at least five species of Microtus in North America
(Bell, 1997; Bell and Repenning, 1999), this identification is no longer is consid-
ered conclusive. It is possible that M. pennsylvanicus occurs in the Porcupine
Cave fauna, but this cannot be verified with the material at hand. Living specimens
were trapped in the vicinity of the cave by members of the DMNH Department

2000

Bell and Barnosky — Porcupine Cave Microtines

113

Fig. 8. — Occlusal view of lower first molars of microtine rodents from Porcupine Cave.
A. UCMP 155273, Lemmiscus curtatus, left B. UCMP 155310, L. curtatus, left Mj.

C. UCMP 155382, Lemmiscus sp., right M,. D. UCMP 155343, Lemmiscus sp., right M;. Scale
bars = 1 mm.

of Zoology in June 1988; M. montanus and M. longicaudus are also reported to
occur in the vicinity of the cave (Armstrong, 1972).

Genus Lemmiscus^ Thomas, 1912
Lemmiscus curtatus, (Cope, 1868)

(Fig. 8A, B)

Material. — Specimens from the Pit (UCMP V93173, CM 1925): UCMP 155265 (right dentary with
Ii and Mi_2, surface); CM 45457 (left dentary with Ij and Mj, level 1); CM 45458, 45459 (two right
dentaries with 1] and Mi, level 1); UCMP 155266-155290 (11 left Mi, two broken left Mi, and 12
right Ml, level 1); CM 45461 (left Mi, level 1); CM 66239 (left Mj, level 1); CM 66242 (left Mi,
level 1); CM 66277 (broken left Mi, level 1); CM 66279 (left Mj, level 1); CM 66283 (left Mi, level
1); CM 66304 (left M„ level 1); CM 66310 (left Mj, level 1); CM 66330 (left M„ level 1); CM
66343 (left Mi, level 1); CM 66346, 66347 (two left Mi, level 1); CM 45465 (right dentary with Mi_
2, level 1); CM 66190 (broken right Mi, level 1); CM 66201 (right Mi, level 1); CM 66220 (right Mi,
level 1); CM 66267 (broken right Mi, level 1); CM 66268 (right Mi, level 1); CM 66281 (broken
right Ml, level 1); CM 66282 (right level 1); CM 66301 (right Mi, level 1); CM 66308 (right Mi,
level 1); CM 66323, 66324 (two right Mi, level 1); CM 66328 (right Mi, level 1); CM 66334 (right
Ml, level 1); UCMP 155291-155295 (two left Mi, two right Mi, and one broken right Mi, level 2);
CM 66572 (left Mi, level 2); CM 66597 (left Mi, level 2); CM 66562 (right Mi, level 2); CM 66570
(right Ml, level 2); CM 66581 (right Mi, level 2); UCMP 155296, 155297 (one left and one right Mi,
level 3); CM 45468, 45469 (two left Mi, level 3); CM 65578 (left Mi, level 3); CM 45471, 45472

114

Annals of Carnegie Museum

VOL. 69

(two right Ml, level 3); UCMP 155298-155309 (seven left Mj, two right Mi, and three broken right
Ml, level 4); UCMP 155310-155312 (two left and one right Mi, level 5); UCMP 155313 (right Mi,
level 6); UCMP 155314 (broken left Mi, level 6); UCMP 155315 (broken left Mi, level 7); UCMP
155316, 155317 (two left Mi, level 8); UCMP 155318 (right Mi, level 9); UCMP 155319 (broken
right Ml, level 10); CM 65169 (left dentary with Ii and Mi_2, level 1-3 undifferentiated); UCMP
155320 (left dentary with Mi_3, level 1-3 undifferentiated); UCMP 155321 (left dentary with Mi_2,
level 1-3 undifferentiated); UCMP 155322 (right Mi, level 1-3 undifferentiated); UCMP 156157 (right
Mj, level 1-3 undifferentiated); UCMP 155323 (left dentary with li and Mi_2, south of extension area);
UCMP 155324-155328 (two left Mi, one broken left Mi, and two right Mi, Pit undifferentiated);
UCMP 155407, 155408 (one left and one right M^, surface); UCMP 155409-155423 (ten left and
five right M\ level 1); UCMP 155494 (left M^, level 1); UCMP 155424-155427 (one left and three
right M2, level 2); UCMP 155428 (left M^, level 3); UCMP 155429-155457 (12 left and 17 right M^,
level 4); UCMP 155458-155471 (seven left and seven right M^, level 5); UCMP 155472-155478
(three left and four right M^, level 6); UCMP 155479, 155480 (one left and one right M^, level 7);
UCMP 155481, 155482 (one left and one right M^, level 8); UCMP 155483-155486 (four right M^,
level 8A); UCMP 155487-155489 (one left and two right M^, level 10); UCMP 155490 (right M^,
level 11); UCMP 155491 (left M^, Pit undifferentiated); UCMP 155495, 155496 (one left and one
right M-\ surface); CM 45410-45413 (two left and two right M^ level 1); CM 66191 (left M\ level
1); CM 66260 (left M^, level 1); CM 66286 (left M^ level 1); CM 66265 (right M^, level 1); CM
66286 (right M^, level 1); UCMP 155497-155501 (four left and one right M^, level 1); UCMP 155502-
155504 (one left and two right M^, level 2); UCMP 155505 (right M\ level 3); UCMP 155506-
155525 (seven left and 13 right M\ level 4); UCMP 155526-155533 (four left and four right M-\
level 5); CM 45473 (left M^ level 5); UCMP 155534-155537 (one left and three right M^, level 6);
UCMP 155538, 155539 (two left M-\ level 7); UCMP 155540-155542 (three left M\ level 10); UCMP
155543 (left M^, Pit undifferentiated).

Identification. — Rootless molars with cementum in the reentrant angles; Mj
with a posterior loop and five or six closed alternating triangles (in most speci-
mens, triangle 6 is confluent with the anterior cap); incipient development of a
seventh triangle is very rare; triangle 2 of M, as large as or larger than triangle
1. of Lemmiscus is diagnostic in having an anterior loop, usually two alter-
nating triangles, and an elongated posterior loop. with anteroposteriorly en-
larged lingual reentrant angle.

Discussion. — Lemmiscus curtatus is reported from numerous Rancholabrean
localities in the western United States (Kurten and Anderson, 1980; Bamosky and
Rasmussen, 1988; Harris, 1993; Bell and Mead, 1998), but is rare from sites of
Irvingtonian age (the only other locality of that age being Cathedral Cave in
Nevada [Bell, 1995], and possibly the Kennewick Road Cut in Washington [Rens-
berger et al., 1984; Rensberger and Barnosky, 1993]). and specimens listed
in the material above probably include specimens from individuals showing the
primitive morphology of M, described below under ''Lemmiscus sp.”

Lemmiscus sp.

(Fig. 8C, D)

Ma/er/a/.— Specimens from the Pit (UCMP V93173, CM 1925): UCMP 155333-155338 (two left
and four right Mj, level 1); UCMP 155339 (right dentary with Ij and Mi_2, level 1); CM 45460 (left
M„ level 1); CM 45462 (left M„ level 1); CM 66214 (left M„ level 1); CM 45463, 45464 (two right
Ml, level 1); CM 65233 (right M,, level 1)); CM 65481 (right dentary with Mi_2, level 2); UCMP
155340-155343 (two left and two right Mi, level 2); CM 66547 (left Mi, level 2); CM 45466, 45467
(two right Ml, level 2); CM 66588 (right Mi, level 2); UCMP 155344-1553478 (one left and three
right Ml, level 3); UCMP 155348 (left Mi, level 3?); UCMP 155349 (left dentary with Mi_3, level 3);
UCMP 155350, 155351 (two right M„ level 3); CM 45506 (broken left M„ level 3); CM 65556 (left
Ml, level 3); CM 45470 (right Mi, level 3); CM 65562 (right Mi, level 3); CM 65579 (right Mi, level

3) ; UCMP 155352-155373 (nine left and 13 right Mi, level 4); UCMP 155374 (broken right M„ level

4) ; UCMP 155375-155378 (four right Mi, level 5); UCMP 155379, 155380 (one left and one right
M„ level 6); UCMP 155381 (right M„ level 7); UCMP 155382, 155383 (two right Mi, level 8);

2000

Bell and Barnosky — Porcupine Cave Microtines

15

UCMP 155384, 155385 (one left and one right M,, level 10); CM 65053 (right dentary with Ij and
Mi_3, level 1-3 undifferentiated).

Identification. — Rootless molars with cementum in reentrant angles; Mj with
posterior loop and four closed triangles, fifth triangle open and confluent with
anterior cap; only labial secondary wing (triangle 6) is developed; otherwise sim-
ilar to L. curtatus.

Discussion. — Specimens of Lemmiscus with a more primitive (four closed tri-
angles) morphology on M, were first described from the basal sections of the
Kennewick Roadcut in Washington; higher in the same section, specimens with
the typical L. curtatus morphology were present (Rensberger et ah, 1984; Rens-
berger and Barnosky, 1993). Lemmiscus specimens from SAM Cave in New Mex-
ico only show the primitive morphology and are reported to date to between
approximately 870,000 YBP and 875,000 YBP (Repenning, 1992). In both the
Kennewick locality and the Pit sequence in Porcupine Cave, there is evidence to
suggest that the Lemmiscus populations in these areas were undergoing progres-
sive morphological evolution in which the complexity of the Mj was increasing
(by the closing off of triangle 5). Under such circumstances, the loss of individuals
showing the primitive morphology would represent a pseudoextinction as opposed
to the loss of an actual species (Archibald, 1993). It is not certain whether this
primitive morphology merits recognition as a separate species. Until very recently
the primitive morphology of Lemmiscus was known only from Irvingtonian or
early Rancholabrean faunas in Porcupine Cave, Kennewick, Cathedral Cave, and
SAM Cave. However, several specimens were recently identified from the micro-
tine rodent fauna from Snake Creek Burial Cave (Bell and Mead, 1998), a late
Pleistocene locality in the central Great Basin radiocarbon-dated between 9460 ±
160 and 15,100 ± 700 YBP (Mead and Mead, 1989). This is the only record of
the primitive morphotype from sediments of late Rancholabrean age, and it was
not found in a survey of several hundred specimens of modern L. curtatus (A.
Barnosky and C. Bell, unpublished data). For the present, we simply refer to these
specimens as Lemmiscus sp., but recognize that they may represent a primitive
morphotype of Lemmiscus curtatus.

One of the specimens listed above (CM 45506) was previously reported as

Microtus paroperarius (Barnosky and Rasmussen, 1988).

Chronological Control

For much of the Irvingtonian mammal age, radiometric dating techniques for
vertebrate fossils in cave deposits are limited to uranium series dating or strati-
graphic association of fossiliferous deposits with a volcanic ash. Further compli-
cations in chronological ordering of faunas from this time stem from the difficulty
in correlating faunas with the geomagnetic polarity time scale. Most Irvingtonian
vertebrate faunas are from caves or thin stratigraphic sections, conditions under
which sequences of magnetic reversal events are not frequently detectable. In such
circumstances, paleomagnetic signatures are most useful to determine whether a
fauna is older than the Brunhes/Matuyama boundary (dated approximately
780,000 YBP); a reversed signature indicates that the fauna is older than 780,000
YBP. In short stratigraphic sequences where only normal polarity is recorded, it
is impossible to say on the basis of paleomagnetic data alone whether the fauna
is in the Brunhes Chron (and therefore younger than 780,000 YBP) or whether

116

Annals of Carnegie Museum

VOL. 69

Table 2. — Amino acid racemization results from samples taken from within Porcupine Cave. Relative
ages are expressed as a numeric sequence, with 1 being the oldest, and 3 the youngest. Absolute age
estimates were generated from a rate constant calculated using a mean annual temperature in the
cave of 51 °F. Absolute age estimates are here considered to be in error; see text for discussion.

AAR sample and locality

D/L ratio

Relative

age

Absolute age estimate

Badger Room; 2 teeth

0.241

1

82,334 YBP

Gypsum Room; horse tooth

0.170

2

52,666 YBP

Pit, level 2; broken horse tooth

0.250

1

86,165 YBP

Pit, level 6; 2 rodent incisors

0.249

1

85,738 YBP

Velvet Room (CM), level 1; horse tooth

0.1002

3

24,000 YBP

the fauna is encased in pre-780,000 YBP sediments deposited during a normal
polarity subchron within the Matuyama Chron (e.g,, Jaramillo or Cobb Mountain).

At Porcupine Cave, radiometric dating has not yet been possible, although the
recent discovery of a degraded ash bed within the DMNH Velvet Room exca-
vation may provide an opportunity for K-Ar dating. Adequate samples of the ash
have not yet been collected or submitted for analysis (D. Rasmussen, personal
communication, 1997). A horse molar (from levels 1-3 mixed) and flowstone
samples from the base of level 5 were submitted for uranium-series and electron
spin resonance dating, but results are pending. Other attempts to establish non-
biostratigraphic age control at Porcupine Cave included amino acid racemization
and paleomagnetic analyses.

Amino Acid Racemization. — Five samples of bone (taken from the Pit, Badger
Room, Gypsum Room, and CM Velvet Room excavations) were submitted to the
Amino Acid Dating Laboratory at Scripps Institution of Oceanography at the
University of California at San Diego for amino acid racemization in order to
determine relative ages of the localities (see Table 2). D/L enantiomeric ratio
results (from aspartic acid) indicate that the specimens from the Badger Room
and the Pit are the oldest of those analyzed. The Gypsum Room sample was next
oldest, and the Velvet Room sample was the youngest. The Velvet Room sample
was taken from the highest stratigraphic level within the CM excavation, and it
is not known whether the lower stratigraphic levels from the CM Velvet Room
excavation might overlap the younger units of the Pit. The relative ages as de-
termined by amino acid racemization are more or less consistent with relative
ages determined from biochronologic age assessments for the various localities,
but absolute age estimates derived from amino acid racemization (Table 2) are
not in accordance with approximate ages derived from biochronology. Absolute
age estimates were derived according to the calculations discussed by Bada (1985)
with rate constant {h) evaluations based on modem mean annual temperature
values in the cave measured at 10.5°C. There is an order of magnitude difference
between biochronologic age estimates based on microtines and the absolute ages
of localities calculated by amino acid racemization. This extreme difference is
probably the result of a number of factors. First, fossil bone does not represent a
closed system with respect to amino acid racemization. Free amino acids from
the surrounding environment (e.g., sediment) can be added to fossil bones after
burial; these free amino acids usually result in lower D/L ratios, and age esti-
mations that are consequently too young (Bada, 1985). Secondly, the age calcu=
lation is only as good as the rate constant evaluation, which in this case was
derived from an assumption of a stable temperature of 10.5°C within Porcupine

2000

Bell and Barnosky — Porcupine Cave Microtines

17

Cave over the last several hundred thousand years. Clearly, if the deposits rep-
resent glacial-interglacial transitions as discussed by Barnosky and Rasmussen
(1988), Wood and Barnosky (1994), and Barnosky et al. (1996Z?), stable temper-
atures through time would not be the case. Finally, although previous studies
indicate that after 80,000 to 100,000 years aspartic acid in fossil bones should be
completely racemized, fossil bones from Olduvai Gorge (dated by other means
at close to 600,000 YBP) still yielded aspartic acid ratios of approximately 0.075,
indicating that secondary aspartic acid was introduced in significant quantities
from sediments and groundwater (Bada, 1981, 1985). We believe that a similar
situation is likely to have occurred in Porcupine Cave and for the present, reject
the absolute age estimates derived from amino acid racemization.

Paleomagnetics. — Paleomagnetic samples were taken from the base of level 5
and from levels 8 A and 14 in the Pit sequence and were analyzed by V. A.
Schmidt at the University of Pittsburgh Paleomagnetic Laboratory shortly before
he died. Unfortunately, no viable paleomagnetic samples were obtained from
higher stratigraphic levels due to the friable nature of the sediment. Results of
the paleomagnetic analyses provided by Schmidt are difficult to interpret. Analysis
of the sample taken from the base of level 5 was equivocal. The data suggest that
levels 8 and 14 record mostly reversed polarities, but there are significant com-
ponents of intermediate and, to a lesser extent, normal components that make a
definitive reversed interpretation for these levels problematic. It is likely that at
least some of these sediments reliably record a reversed field and that at least the
lower levels of the Pit sequence (below level 7) can be considered to be older
than 780,000 YBP (Barnosky et al., 1996b).

Biochronology. — The taxonomic composition of the microtine rodent fauna
from the Pit sequence poses a challenging problem for biochronologic age esti-
mation. At least 1 1 microtine rodent species are present in the fauna as a whole;
as many as ten of these species occur within a single stratigraphic level (e.g.,
level 4; see Table 3) and seven or eight taxa are typical for most of the levels in
the Pit (these figures are based on consideration of non-M^ teeth for both Mi-
momys and Ondatra; see Table 3). Such diversity is rare within North American
microtine rodent faunas and the taxonomic associations in Porcupine Cave are
not found anywhere else on the continent. This particularly intriguing aspect of
the Pit fauna complicates attempts to place the sequence within the existing mi-
crotine rodent biochronology.

Previous age estimations for the upper stratigraphic levels in the Pit (Barnosky
and Rasmussen, 1988; Wood and Barnosky, 1994) were derived from preliminary
analysis of the microtine rodent fauna and were based upon the standard bioch-
ronological framework outlined by Repenning (1987) and Repenning et al. (1990).
In these schemes, the presence of Lemmiscus and specimens referred to Microtus
pennsylvanicus argued for an age no older than 400,000 YBP for stratigraphic
levels reported to containing these taxa.

In 1992, a significant change in the microtine biochronology was necessitated
with the recognition of Lemmiscus in the fauna from SAM Cave, New Mexico,
which dates to between 870,000 and 875,000 YBP (Repenning, 1992). In the
preliminary analysis of the SAM Cave fauna, the relatively primitive morphotypes
of Lemmiscus were interpreted to represent a gradual transition within that fauna
from Allophaiomys into Lemmiscus. It was this hypothesis that led to the sugges-
tion that the Allophaiomys morphotypes from Porcupine Cave might possibly be
three-triangle morphotypes of Lemmiscus. This suggestion was published as an

118

Annals of Carnegie Museum

VOL. 69

^ A

^ -S

C e

§ I

§ ^

^ S

a N

u s

I”

o

Q.

K •?

K

I

L

il

s ^

£|

^ i

^ s
C "G

I >

•S -S
S K
"2

S ^

'ii a
K e

5 6«

I II

S- N

1 2

K ^

P

2 a

I II

^ Co

0 — ;,
:§ ^

1^

60 “S3
'.S ^
C .2
^ 3

1 So

I -S

cn _'g

jj ^

3
e2

CO o

il § I I M I I

I I il « I S I

xs 11 I I I I 1 I I

Os CM O O CN I I CM

i § I I -

r- O — < i— I CM 1 CM

^

-H O O CM

I S X I I I

CM — < — i CO fO --x I O

S I I

r^—i'^CMCM'—

I X

00

.O^

cor^osi/ifO'^ro I—*

I SS I I S I

r- of- I m ^

so ^

msOvOCM iCMCM i—i 1^

l^r^ lo lo

CM VO »r) CM I CM CM I O

CNJ — I _H

SS I s I S I M I I

e

S

E 3

S|

s

g^i

5 o

a
s
•S
^ S
•3 5)

5 2

S Cj "§ Q

S gx ^

a s _

a s s ^ _

% 33f2;s?S;

SS

^ u

M &0

to

^ ^

S S

Q ^5 Q
U L C
C3 Q ^

s s a
^ ^ s
^ 3

a, a O

2000

Bell and Barnosky — Porcupine Cave Microtines

119

alternative hypothesis by Wood and Barnosky (1994) and Barnosky et al. (1996Z?),
but these authors preferred an interpretation in which Lemmiscus in the fauna was
restricted to a primitive four-triangle morphotype (as a possible new species of
Lemmiscus) and a more typical five-triangle morphotype similar to the modern L.
curtatus, with three-triangle specimens representing a true population of Allo-
phaiomys. After a challenge to the interpretation of an Allophaiomys-Lemmiscus
ancestor-descendaent relationship in SAM Cave, a more detailed analysis of the
fauna by C. A. Repenning failed to show a transition in morphology from a three-
triangled population to a four-triangled population and the hypothesis presented
by Repenning (1992) is now rejected on grounds of a more thorough analysis
(Bell, 1998). We follow Barnosky et al. {\996b) in considering the three-triangle
Ml in the Pit sequence to represent a valid population of Allophaiomys pliocaen-
icus. The earliest Lemmiscus known are from SAM Cave, New Mexico, and are
dated between 870,000 and 875,000 YBR These specimens all display a primitive
morphology in which the fifth triangle on Mj is confluent with the anterior cap.
Similar morphotypes are found in the basal (undated, but probably early Ran-
cholabrean) sections of the Kennewick Road cut locality (Rensberger et al., 1984;
Rensberger and Barnosky, 1993), and are known from a late Rancholabrean lo-
cality in eastern Nevada (Bell and Mead, 1998).

The most serious difficulty in determining an age for the different stratigraphic
levels within the Pit sequence is the apparent coexistence of taxa which before
have never been found in direct stratigraphic association, and are not known to
occur contemporaneously elsewhere. In the case of the lower stratigraphic levels,
there is independent evidence from paleomagnetic analyses that level 8 and below
are older than the Brunhes/Matuyama boundary at approximately 780,000 YBP
Above level 8, external age control is unavailable, and age determination must
be based on faunal composition.

There are three species from the Pit sequence {Phenacomys gryci, Mimomys
cf. M. virginianus, and Allophaiomys pliocaenicus) that argue for a greater antiq-
uity than previously considered possible for this deposit. At least one species
{Lemmiscus curtatus) would appear to indicate a younger age.

Phenacomys gryci was previously reported from three other localities. In the
Fish Creek fauna in Alaska (the type locality) it was determined to date to ap-
proximately 2.4 Ma (Repenning et al., 1987). In Froman Ferry, Idaho, specimens
of P. gryci were recovered from a much younger sequence spanning 1.6 to 1.5
Ma (Repenning et al., 1995). Phenacomys gryci first appears in the Pit sequence
in level 10, is absent until level 4, and makes its last appearance in level 3 (Table
3).

Mimomys (Cromeromys) virginianus is currently known only from the type
locality in the Cheetah Room fauna of Hamilton Cave, West Virginia, which is
reported to date to between 850,000 and 820,000 YBP (Repenning and Grady,
1988; Repenning 1992). A similar species described from the Java fauna in South
Dakota is considered by most authors to be early Irvingtonian in age (R. Martin,
1989; Repenning, 1992). Both of these previous North American records are from
pre-Brunhes age localities. Mimomys cf. M. virginianus is present in low abun-
dance in the Pit and is represented by Mj specimens only in levels 8, 6, and 5;
other teeth document its presence in levels 10, 7, and 4 as well (see Table 3).

The earliest well-dated occurrence of Allophaiomys pliocaenicus in North
America is from a drill core at the base of Hansen Bluff, Colorado, where a single
specimen was recovered from a horizon at or near the lower boundary of the

120

Annals of Carnegie Museum

VOL. 69

Olduvai magnetic polarity Chron at about 1.9 Ma (Repenning, 1992; Rogers et
aL, 1992). Specimens are also reported from early Irvingtonian faunas at Java,
South Dakota, and the Wellsch Valley fauna in Saskatchewan. The youngest re=
ported occurrence is from the Cheetah Room fauna in Hamilton Cave (840,000
YBP) or possibly from Cumberland Cave in Maryland (830,000 YBP); in any
event no records from North America are known to date younger than 825,000
YBP (Repenning, 1992). Allophaiomys occurs in levels 11 and 8-4 in the Pit
sequence.

Microtus paroperarius first appears in faunas reported to date between 830,000
and 840,000 YBP; Repenning and Grady (1988) suggested an evolutionary tran-
sition out of Lasiopodomys in the vicinity of Hamilton Cave, West Virginia, some-
time around 840,000 YBP. It is known from three other localities reportedly dating
to about 830,000 YBP (Cumberland Cave, Maryland; County Line, Illinois;, and
the base of the Hansen Bluff sequence in Colorado). Because of its presumed
endemic evolution in the eastern United States, M. paroperarius was never pro-
posed as a defining taxon at the base of the Irvingtonian II microtine rodent
division (traditionally drawn at 850,000 YBP based on the first appearance of
presumed immigrants), but it has played a role in characterizing faunas of this
age in the eastern and central United States (Repenning, 1987, 1992; Repenning
et ah, 1990). Its youngest reported occurrence is in Salamander Cave, South
Dakota (323,000 to 252,000 YBP; Mead et ah, 1996). Microtus paroperarius is
absent from the lower stratigraphic levels in the Pit, but makes its first appearance
in level 7, and persists through level 1.

Terricola meadensis, which is present in Porcupine Cave Pit from levels 1
through 8, currently is used as the primary defining taxon for the base of the
Irvingtonian II microtine rodent division. Its earliest occurrence is dated at ap-
proximately 820,000 YBP near the base of the Hansen Bluff sequence in Colo-
rado; an undated, but apparently early, occurrence was reported from the Anza-
Borrego Desert (Repenning, 1992). The youngest report of T. meadensis may be
in the Kennewick locality in Washington state, where it comes from deposits
estimated to be not much older than 328,000 YBP based on rates of formation
of pedogenic calcretes (Rensberger and Bamosky, 1993), or from Salamander
Cave, where it may be as young as 252,000 (Mead et ah, 1996). However, living
populations of the Jalapan pine vole in Mexico {Terricola quasiater, Pitymys
quasiater, or Microtus quasiater, depending on whose taxonomy is followed) have
a dental morphology indistinguishable from T. meadensis (Repenning, 1983).

The primitive morphotypes of Lemmiscus mentioned above first appear in level
10 of the Pit sequence and persist throughout the section. Specimens with a more
complex morphology similar to that seen in living L. curtatus also make their
first appearance in level 10 and persist to the top of the sequence. The specimens
from Porcupine Cave represent the earliest known record of Lemmiscus curtatus;
the oldest previously reported record is probably from the Kennewick fauna in
Washington (Rensberger et aL, 1984; Rensberger and Bamosky, 1993).

Attempts to place the Pit sequence into the existing biochronological framework
for microtines forces changes in the accepted temporal ranges of several taxa.
The most anomalous occurrence is that of Phenacomys gryci. This is in part
explained by the fact that so little is known about this species. The youngest
known occurrence of 1.5 Ma at Froman Ferry is still much older than the max-
imum temporal range of most other taxa in the Pit. It is noteworthy that the Pit
specimens are, without exception, among the most advanced morphotypes known

2000

Bell and Barnosky — Porcupine Cave Microtines

121

(Ma)

Taxa

0.

1 0.

2 0.

3 0.

4 0.

5 0.

6 0.

7 0.

8 0.

9 1.

0 1.

1 1.

5

P. gryci

M. virginmnm

A. pliocaenicus

M. paroperarius

Lemmiscus sp.

L. curtatus

T. meadensis

Microtus sp.

'W.

Fig. 9. — Temporal ranges of selected microtine rodent taxa in North America, based on previously
reported faunas and correlations. Data summarized from Repenning (1987, 1992), Repenning and
Grady (1988), Repenning et al. (1990, 1995), Mead et al. (1996), Bell (1998), and Bell and Mead
(1998).

for this species. Although they do fall within the morphological variation dis-
played in the type population (Repenning et al., 1987), the extremely complex
anterior cap morphologies seen in the Fish Creek sample are conspicuously absent
from the Pit sample. It is possible that as more becomes known of this species
and its evolution in North America, more sense may be made of the material
from the Pit. For now, we consider these specimens to represent the youngest
occurrence of the species, and count them among the most advanced morphotypes
described.

Setting Phenacomys gryci aside, plots of oldest and youngest reported occur-
rences elsewhere in North America of the other species in the fauna reveal only
a narrow temporal range during which potential sympatric associations could be
expected (Fig. 9). At some time near 850,000 YBP, both Allophaiomys and Mi~
momys (Cromeromys) disappear from the North American fauna. At approxi-
mately the same time (within 20,000 years or so to either side) Microtus paro-
perarius, the primitive morphotypes of Lemmiscus sp., and Terricola meadensis
all appear for the first time. If an age of 850,000 YBP is accepted for stratigraphic
levels in which sympatric associations of Allophaiomys, Mimomys, Microtus par-
operarius, the primitive morphotypes of Lemmiscus sp., and Terricola are found,
then the middle sections of the Pit sequence (levels 4, 5, and 7) would date to
that time.

It is unclear whether this age estimation is inconsistent with paleomagnetic data
because of the confusing paleomagnetic signature from level 5, which contained
some reversed component, but also contained a nearly equal proportion of normal

122

Annals of Carnegie Museum

VOL. 69

signatures. By level 8, a predominance of reversed signatures is consistent with
an age of older than 780,000 years.

The presence of both normal and reversed paleomagnetic signatures in level 5,
and to a lesser extent in level 8, may indicate that the middle section of the Pit
actually records the transition from the Matuyama Chron (reversed) to the Brunhes
Chron (normal). The magnetic signature of sediments deposited immediately pre-
ceding, during, and immediately after this event might be expected to be some-
what erratic (Opdyke and Channell, 1996). If levels 4 through 8 in the Pit do
capture the magnetic transition, then an age of around 750,000 to 780,000 YBP
would be applied to that part of the sequence. This would require younger oc-
currences of Allophaiomys and Mimomys than are elsewhere known (by approx-
imately 70,000-100,000 years). This explanation would nicely reconcile the
biochronological and paleomagnetic data.

An estimation of 750,000-780,000 years for levels 4 through 8 still requires a
very significant change in the accepted temporal range of Lemmiscus curtatus
(which is present in the Pit as low as level 10). Any interpretation of the age of
the Pit sequence will involve moving the first appearance datum of this species
back into the Irvingtonian because of its presence in the largely reversed sedi-
ments below level 8.

Temporal ranges of other species in the Pit fauna are not affected by this age
interpretation. Five-triangle Microtus were present in North America on the west
coast by as early as 1.4 Ma in the Anza-Borrego Desert (Zakrzewski, 1972;
Repenning, 1992), and probably earlier than that time, if Beringian immigrant
status for the southern California populations is to be believed. Without reliable
species identification of the Mictomys, Ondatra^ and Phenacomys specimens not
referable to P. gryci, evaluations of their temporal span cannot be made. It is
interesting to note that these more advanced morphotypes of Phenacomys do not
appear in the Pit sequence until very near the top of the section (level 2; see
Table 3). Ondatra is in such low abundance in the Pit that any statements re-
garding its status would be premature.

If the magnetic reversal in the Pit sequence represents the change from the base
of the Jaramillo into the middle Matuyama instead of the Brunhes/Matuyama
boundary, then an age of about one million years would apply to sediments in
level 8 of the Pit. This interpretation is rejected based on the biochronological
evidence. An age interpretation of approximately one million years would better
explain the presence of Mimomys and Allophaiomys in the Pit sequence, but would
force a much older age for the first appearance of Microtus paroperarius, Terri-
cola, and both species (or morphotypes) of Lemmiscus,

If the 252,000 YBP dates for Microtus paroperarius and Terricola meadensis
from Salamander Cave (Mead et ah, 1996) are correct, then the higher stratigraph-
ic levels of the Pit sequence (levels 1-3) probably are no younger than approxi-
mately 250,000 YBP. The potential temporal span of levels 1-3 is between
250,000 and 750,000 YBP. At present, a more precise age assessment of these
levels cannot be provided.

Discussion and Conclusion

The microtine rodent fauna from the Pit locality in Porcupine Cave is one of
the most taxonomically diverse in North America. It seems likely that the high
diversity and anomalous taxonomic associations are somehow related to the ex-

2000

Bell and Barnosky — Porcupine Cave Microtines

123

tremely high elevation of Porcupine Cave. At least three factors may have con-
tributed to the unique assemblage found in the Pit.

Taphonomic Pathways

It is possible that the taphonomic pathways responsible for the introduction of
bones and teeth into the site may have sampled taxa from a variety of microhab-
itats within the hunting ranges of the carnivores and raptors that apparently col-
lected most of the bones. The cave sits along a ridge overlooking a valley within
South Park, and a view from the top of the ridge today reveals an impressive
diversity of microhabitats and elevations within a relatively short distance from
the modern entrance. The taphonomic pathways that brought skeletal material
inside the Pit apparently included hunting of small mammals by predators, de-
position of raptor pellets or fecal material near the cave entrance, and collection
of and hoarding in their nests of the bone-laden pellets by woodrats (genus Neo-
toma; Barnosky and Rasmussen, 1988; Bamosky et al., 1996^). Several lines of
evidence substantiate this taphonomic interpretation. Abundant Neotoma skeletal
remains and fossilized woodrat midden matrix and fecal pellets in the Pit deposits
demonstrate the presence of these animals in and around the cave throughout the
period during which fossils were preserved. Predominance of the kind of small-
mammal bones (jaws, broken skulls, isolated teeth, disarticulated postcranial el-
ements, etc.) and bone-breakage patterns (e.g., teeth etched by digestive juices
and skulls missing their posterior half) commonly preserved in raptor pellets and
carnivore scat (Andrews, 1990) suggests that significant portions of the vertebrate
fossils from the cave were derived from these sources. In addition, observation
of modern Neotoma middens inside Porcupine Cave revealed that they contain
mummified Neotoma carcasses, raptor pellets, and carnivore scat with skeletal
representation similar to that found in the fossil deposits, and modem cow teeth
similar in size to the fossil ungulate teeth that occur in the Pit deposits. Obser-
vations by other authors also demonstrate the penchant of Neotoma for collecting
plant and animal remains (including raptor pellets and carnivore feces) from with-
in the region immediately surrounding their nests (Betancourt et al., 1990).

Detailed study of Neotoma cinerea accumulations indicate that fossil deposits
generated by woodrat activities such as occurred at Porcupine Cave reliably record
the taxonomic composition and probably relative abundance of small mammal
communities through time (E. Bamosky, 1992, 1994; Hadly, 1999). Therefore, it
seems reasonable to assume that at least part of the diversity in Porcupine Cave
is simply a reflection of relatively complete sampling of the taxa living within
the immediate vicinity of the site.

Introduction of Allocthonous Taxa

The contribution of fossil remains from relatively distant areas is more difficult
to assess. Raptorial birds and mammalian carnivores can introduce into cave de-
posits materials from distant locations, but most commonly their prey comes from
within a 5-km radius (Hadly, 1999). The abundant carnivore remains in Porcupine
Cave (Anderson, 1996) demonstrate that badgers, ferrets, other mustelids, wolves,
coyotes, and foxes were present and probably contributing to the deposition of
vertebrate materials near, and in some cases (e.g., the Badger Room) in, the cave.

Although transport of animal remains from within South Park seems likely,
there is little evidence to support the hypothesis that materials were transported

124

Annals of Carnegie Museum

VOL. 69

from long distances. A lower elevation (2300 m) fauna from Hansen Bluff, ap-
proximately 120-km south of Porcupine Cave, was described in detail by Rogers
et al. (1985). Of the six microtine rodents described from that locality, five are
present in the Pit sequence in Porcupine Cave. The single exception is a reported
occurrence of Clethrionomys sp. from locality PP2 near the base of the Hansen
Bluff sequence. Because of the significance of this record, both for interpretation
of the Porcupine Cave fauna and because of its importance for the definition of
the Irvingtonian II microtine rodent division (Repenning, 1987; Repenning et ah,
1990), we asked C. A. Repenning to reexamine the specimen and confirm the
identification. Closer inspection revealed that the specimen is a rooted microtine
molar, but is badly broken and lacks cementum in the reentrant angles. All known
specimens of living and fossil Clethrionomys have pronounced cementum in the
reentrant angles, and the material from Hansen Bluff is here transferred to Phen-
acomys (of unknown specific affinity).

The fact that all the microtines from Hansen Bluff are represented in Porcupine
Cave is interesting, but not necessarily surprising; Mictomys meltoni, Terricola
meadensis. Ondatra annectens (in Porcupine Cave as '"Ondatra sp.”), Microtus
paroperarius, and Microtus sp. are all known from lower elevations in the Great
Plains. Comparison of the faunal lists from Porcupine Cave (Bamosky and Ras-
mussen, 1988) and Hansen Bluff (Rogers et ah, 1985) reveals an overall similarity
in the taxonomdc composition of other small mammal species from the two lo-
calities, but two important exceptions occur. The heteromyid genera Perognathus
and Dipodomys are present in the Hansen Bluff sequence (at 2300-m elevation),
but are conspicuously absent from Porcupine Cave (all of the excavations within
the cave produced only one heteromyid specimen — an isolated molar of Dipo-
domys sp.). The historical distribution of heteromyids in Colorado includes re-
cords from at least as high as 2530 m for both Perognathus and Dipodomys
(Armstrong, 1972); a very early report (Coues and Yarrow, 1875) of Dipodomys
ordii taken from approximately 2804 m in the vicinity of Twin Lakes was con-
sidered doubtful by Armstrong (1972), If raptors or carnivores were bringing in
materials from lowland localities farther south or east, heteromyid remains would
be expected in the Pit deposit.

Dispersal Corridors

The position of the cave high in the Rocky Mountains may offer at least a
partial explanation of the anomalous assemblage of microtine rodents. It is pos-
sible that Porcupine Cave is situated near an ancient dispersal corridor along
which microtine species may have expanded their ranges from further to the north.
This hypothesis was proposed by Bamosky et al. (1996^) as an explanation for
what appeared to be an anomalously early occurrence of Microtus montanusIM.
longicaudus. Although these specimens are here identified simply as Microtus sp.,
they do appear rather abruptly in the Pit sequence at level 6 and persist through
the younger levels, increasing in relative abundance upsection (Fig. 10). Unfor-
tunately, the specific affinities of these specimens cannot be determined with the
material at hand, but it is possible that the dispersal of Microtus species with an
Mj with five closed triangles into the Rocky Mountain region was from north or
south along a mountain corridor. Fossil material of five-triangle Microtus is known
from earlier deposits in southern California (Zakrzewski, 1972; Repenning, 1992)
and also from the roughly contemporaneous Irvington gravels in north-central

2000

Bell and Barnosky — Porcupine Cave Microtines

125

A. pliocaenicus

P. gryci

Phenacomys sp. (not

Mimomys cf. M.

s

s

S

P.gO’ci) s'

virginianus s

2

2

2

2

2

3

3

3

3

3

4

^ 4 '

4

4

4

5

ii 5

5

; 5;

6

6

6'

'■

7

■■

7

7

' 7

8'

'wm

s'

’ s'

9

^ 9'

9

9

9

10

10

^ to'

10

10

. 11°

/ 11 "

ll'

ll'

(

) 50 100 t

1 50 100 (

) 50 100 {

) 50 100 0

Ondatra sp.

50

100

Relative Abundance of Taxon in Each Level

Fig. 10. — Relative abundance of microtine rodent taxa from the Pit excavation in Porcupine Cave.

California at a slightly lower latitude, and significantly lower elevation, than that
at Porcupine Cave (Savage, 1951). Similarly, the relatively late appearance in the
Pit sequence of Phenacomys not referable to P. gryci may indicate utilization of
a Rocky Mountain dispersal corridor (although the referral of the Hansen Bluff
Clethrionomys'" to Phenacomys could indicate its presence in Colorado at a
slightly earlier date).

More importantly, if the age interpretation given above is correct, the appear-
ance of T. meadensis in the Pit sequence ranks as one of the earliest occurrences
of the species in North America (Repenning, 1992), and certainly the Lemmiscus
curtatus record is the oldest known from the continent. Terricola meadensis has
long been regarded as an immigrant and used as a defining taxon for the beginning
of the Irvingtonian II microtine rodent division. The proposed immigration date
of this species at 850,000 YBP is based on its earliest occurrence at Hansen Bluff
at about 830,000 YBP (Repenning, 1987; Repenning et ah, 1990). The approxi-
mately contemporaneous first occurrence in two Colorado faunas, at Hansen Bluff
and in the Pit (level 8), provides support for a Rocky Mountain dispersal route
for this species. One problem with this interpretation is the occurrence of Terri-
cola at a relatively early age in southern California. A single specimen was re-
ported from reversely magnetized sediments of unknown age in the Anza-Borrego
Desert, (Repenning, 1992). One of us (CJB) recently identified an additional (and
potentially early) occurrence from the Elsinore Fault Zone in Riverside County,
California (SBCM locality 5.6.301; the fauna contains 15 Mj specimens of T.
meadensis).

The Pit sequence unquestionably records the earliest appearance of Mj showing
the morphology of the extant Lemmiscus curtatus. Prior to its discovery in Por-
cupine Cave, all records were of Rancholabrean age, the oldest probably from the
Kennewick locality in Washington state (Rensberger et al., 1984; Rensberger and

126

Annals of Carnegie Museum

VOL. 69

Barnosky, 1993). The place and time of origin and evolution of Lemmiscus cur-
tatus remain unknown, but there is an indication in both the Porcupine Cave Pit
sequence and in the Kennewick sequence that the earlier individuals with an Mj
with four closed triangles evolved into the more complex five^triangle forms.

Faunal Response to Climate Change

It is possible that the region around Porcupine Cave served as a high-elevation
refugium. for some microtine rodent taxa that appear to have been extirpated
elsewhere at earlier times. This hypothesis would apply specifically to Alio-
phaiomys and Mimomys, which elsewhere in North America disappear from fau-
nas between 820,000 and 840,000 YBP, and possibly to Phenacomys gryci, the
latest occurrence of which is 1 .56 Ma elsewhere. The refugium hypothesis would
gain support if it could be shown that the appearance (or persistence) of some
taxa in Porcupine Cave was correlated with climate change in the region.

Changes in the lithology of sediments within the Pit sequence were previously
interpreted to reflect climate change, with the nodule layers representing wetter
(and presumably glacial) periods and the fine “loess” deposits representing drier
(presumably interglacial) periods (Barnosky and Rasmussen, 1988; Wood and
Barnosky, 1994; Barnosky et al., 1996b). At least two such sedimentological
changes are recorded within the Pit sequence (Fig. 2). If this interpretation is
correct, the level 10/9 transition probably represents a change from glacial to
interglacial conditions, as does the transition between levels 4 and 3 (Fig. 2, 10).
Previous analyses of faunal change across sedimentary boundaries focused on the
level 4/3 transition since this was the only portion of the Pit collection that had
been adequately studied. In addition to the physical changes in the sediments
from level 3 to 4, changes in faunal composition across the stratigraphic levels
support the idea that the differences in lithology reflect climatic change. Three
faunal groups (squirrels, microtines, and reptiles) from levels 1 through 6 now
have been studied in sufficient detail to permit some tentative conclusions about
faunal response to climate change in the Porcupine Cave region.

In an analysis of the upper stratigraphic levels of the Pit (levels 1-5), Wood
and Barnosky (1994) presented dramatic faunal evidence in support of major
environmental change at the level 4/3 transition; levels 1-3 were composed of
nearly uniform, fine dust and levels 4 and 5 predominantly of reversely graded
clay nodules (with a change in nodule size defining the break between the levels).
Wood and Barnosky (1994) focused on changes in relative abundance in two
groups, sciurid rodents and microtine rodents (in the latter case drawing upon a
greater sample size than that reported by Barnosky and Rasmussen [1988] but
less than that reported here). Although in most cases all taxa analyzed were re-
covered from all stratigraphic levels, and changes in relative abundance were
detected within the homogenous sediment types (levels 1-3 and 4 and 5 respec-
tively), the most dramatic changes were seen at the level 4/3 transition. Among
the sciurids, a sharp decline in relative abundance of Marmota (from 81.7% to
17.4%) occurs concomitantly with increases in both Cynomys (from 1.4% to
16.3%) and Spermophilus (from 16.9% to 66.3%). Assuming the environmental
preferences of these genera have remained the same since the deposition of fossils
in the Pit, this change in the sciurid rodent fauna supports an interpretation of
decreased precipitation across the level 4/level 3 boundary because Marmota to-
day prefers wetter climates than do Cynomys and most species of Spermophilus

2000

Bell and Barnosky — Porcupine Cave Microtines

127

(Wood and Barnosky, 1994). The persistence of Marmota during effectively drier
climatic conditions might lend some support to the refugium hypothesis, since its
presence indicates that some mesic microhabitats remained within an overall drier
climatic regime.

Because reptile and amphibian fossils can in some instances provide informa-
tion on climate change (Holman, 1995), we examined the reptile and amphibian
fossils from the Pit sequence across the level 4/level 3 transition. Reptile and
amphibian fossils are relatively rare in all stratigraphic levels of the Pit, and most
of the remains are isolated vertebrae of small snakes, rendering estimates of num-
ber of individuals impossible. Nonetheless, an interesting change in the taxonomic
composition of the herpetofauna is clear. The oldest snakes appear in level 8A
(seven vertebrae) and are then sporadically encountered in other stratigraphic lev-
els (three vertebrae in level 7, one in level 6, five in level 4, three in level 3, six
in level 2, and three in level 1). All of the snakes below level 3 are identified as
natricines and probably represent fossils of Thamnophis, a widespread genus
throughout North America today whose members inhabit a wide variety of hab-
itats and elevations (Stebbins, 1985). Thamnophis is not uncommon in the im-
mediate vicinity of Porcupine Cave today. Natricines also are present in levels 1-
3 (nine vertebrae), but one colubrine fossil and two crotalid fossils (probably
representing Crotalus) are also present. Neither of these latter taxa are known in
the vicinity of the cave today (R. Finley, personal communication). Two other
fossils, from a sceloporine lizard and a pelobatid toad, cannot be placed in precise
stratigraphic levels, but their state of preservation and color more closely resemble
that of fossils from levels 1-3 than that of fossils from deeper levels. The change
in reptile diversity is probably correlated with increased temperatures around the
cave during the interglacial recorded in levels 1-3. In addition, the interglacial
recorded in the upper levels of the Pit was possibly warmer than the present
interglacial, since modem herpetofaunal diversity in the vicinity of the cave is
lower than that recorded in the upper levels of the Pit.

The changes in relative abundance and taxonomic diversity in the reptile and
squirrel components of the fauna support the conclusion that a change from a
glacial period to an interglacial period is recorded at the level 4/3 transition. This
transition records a change from relatively cooler, wetter (probably glacial) con-
ditions (level 4) to warmer, drier (probably interglacial) conditions (level 3) that
were probably warmer than current conditions. Previously, the putative glacial-
interglacial transition demonstrated by the level 4/3 boundary was thought to
correspond with oxygen-isotope stages near stage 12 (glacial) and stage 11 (in-
terglacial) (Barnosky and Rasmussen, 1988; Wood and Barnosky, 1994; Barnosky
et aL, \996a, 1996/?). However, the new biochronologic and paleomagnetic in-
terpretations presented here suggest the level 4/3 transition corresponds to a part
of the oxygen-isotope curve that dates between 750,000 and 850,000 YBP. The
istotope excursion from (glacial) stage 22 to (interglacial) stage 2 1 occurred about
875,000 YBP; and from (glacial) stage 20 to (interglacial) stage 19 about 800,000
YBP. At about 740,000 YBP, there was a significant warming event indicated by
the isotope excursion from (cold) stage 18 to (warm) stage 17 (Raymo et aL,
1997). It presently is impossible to say which of these warming events is recorded
by the Porcupine Cave Pit record, but it seems likely that one of them is.

Other clay-dust transitions (probably indicating changes from cool, mesic to
warm, dry conditions) in the Porcupine Cave Pit sequence (e.g., level 10 to 9;
see Fig. 2 ) are older than the level 4/3 transition, but probably still within the

128

Annals of Carnegie Museum

VOL. 69

780,000-900,000 range. This conclusion is based on the stratigraphic position of
these levels below sediments with a magnetically reversed signature, the species
assemblage of microtine rodents, and the rapid sedimentation rates in Holocene
deposits in which Neotoma nests played a major role. An example of such rapid
sedimentation rates was provided by Hadly (1996, 1999) who reported accumu-
lation rates in Lamar Cave, Yellowstone Park, which produced 2 m of fossiliferous
deposit similar to the Porcupine Cave deposit within 3000 years.

The distribution of microtine rodent taxa within the stratigraphic levels of the
Pit (Table 3) and the changes in their relative abundance (Fig. 10) are difficult to
explain in terms of climate change. Sample sizes are too low in levels 10, 9, and
8 for any meaningful statements to be made about faunal response to the climatic
transitions recorded in the sediments from those levels. The sample size for the
level 4/3 transition is adequate, but relative abundances for many of the taxa are
either stable {Lemmiscus curtatus, Lemmiscus sp.) or appear to reflect continuance
of trends begun prior to the major sedimentary change (Allophaiomys, Mictomys,
Terricola; see Fig. 10). One possible exception is Microtus sp. which begins a
sharp increase in relative abundance in level 3, but the major jump in relative
abundance is actually between levels 3 and 2, where sediments do not record a
major climatic change. It may be that the small mammals began responding to
the climate change prior to its reaching the threshold that caused a change in
sediment deposition, which would explain the earlier inception and continuation
of relative abundance trends across the sedimentary change at the level 4/3 bound-
ary.

These more complete data force slight modifications to the previous interpre-
tations of changes in relative abundance of microtine rodent taxa in the Pit (Bar-
nosky and Rasmussen, 1988; Wood and Barnosky, 1994; Bamosky et ah, 1996a,
1996^), but the substantive conclusions of these earlier reports are upheld. Pre-
vious discussions of changes in the microtine rodent fauna indicated that Alio-
phaiomys made its last appearance in the Pit in level 4, suggesting that the local
extirpation of this species (and quite possibly its extinction from North America,
as the Pit records represent the youngest known specimens of the species) cor-
responded with a major sedimentary change indicative of changing climate (Wood
and Bamosky, 1994; Bamosky et ah, 1996a, 1996^). This hypothesis is supported
by our further examination of the fauna. In addition, although no Mj specimens
of Mimomys cf. M. virginianus are present in level 4, there is an isolated M2
representing this genus (and presumably the same species as is represented by the
Ml material). The presence of this specimen in level 4 suggests the possibility
that Mimomys also was driven to extinction by the climate change recorded at
the level 4/level 3 transition.

The Possible Influence of Elevation

Although relatively little is known about Mimomys (Cromeromys) in North
America, the existing records (Cheetah Room in West Virginia, Porcupine Cave,
the Java fauna in South Dakota, and Cathedral Cave in Nevada) suggest that the
animal favored relatively high elevations or high latitudes. This raises the inter-
esting question about the potential influence of high elevation on biochronological
frameworks. It has long been recognized that both living and extinct faunas could
be partitioned into broad geographic provinces based on their occurrence at dif-
fering latitudes (Brown and Gibson, 1983). Faunal provinces of relatively limited

2000

Bell and Barnosky — Porcupine Cave Microtines

129

geographical extent are recognized in the late Pleistocene of North America (Gra-
ham, 1979; L. Martin and Hoffmann, 1987; Rogers et al., 1990; FAUNMAP
Working Group, 1996). Fejfar and Repenning (1992) proposed a series of pro-
vincial designations specifically applied to microtine rodents (microtine rodent
faunal regions). The proposed regions correspond to both latitudinal and longi-
tudinal differences in the distribution of fossil microtine rodents. The validity of
these regions is sometimes difficult to assess, but in principle the concept appears
sound. There is increasing evidence to support the idea that different biochron-
ologies must be applied to different geographical regions (intracontinental as well
as intercontinental). Differences in the taxonomic composition of microtine rodent
faunas on either side of the Rocky Mountains led Repenning et al. (1995) to argue
for the recognition of different temporal boundaries for the Irvingtonian I micro-
tine rodent division in the eastern versus western United States. The challenge in
implementing such regional biochronologies lies in the adequate establishment of
reliable, independent chronologies for the different proposed regions, and, espe-
cially in the case of the Irvingtonian, simply in identifying a sufficient number
of localities to permit regional faunal patterns to appear.

The unique species assemblage character of the Porcupine Cave microtine fauna
suggests the possibility that age relations of faunas from high elevations may not
be readily interpreted in terms of biochronologies established based on exami-
nations of low-elevation sites. Much of the microtine rodent biochronology in
North America west of the Mississippi River was developed based on faunas from
relatively low elevations. The difficulties in determining the age of various strati-
graphic levels within the Pit sequence may be a result of differences in Irving-
tonian community composition between high-elevation sites like Porcupine Cave
and contemporaneous low-elevation sites, just as is the case today (Armstrong,
1972). For the present, Porcupine Cave stands alone as a representative of a high-
elevation Irvingtonian microtine rodent fauna. Future discovery and investigation
of high-elevation sites in western North America may begin to provide the data
required to construct an independent biochronology. Any efforts to construct such
a biochronology must ultimately be anchored by reliable external age control.

Acknowledgments

We thank Frank and Connie McMurray for allowing excavations in Porcupine Cave, and Don
Rasmussen and the numerous volunteers who helped in excavation and fieldwork. We benefited from
conversations with Charles Repenning, Bob Martin, and Russ Graham. Helpful comments on earlier
versions of this report were provided by Charles Repenning, Harry Greene, Bill Berry, Geraldine
Swartz, and Jens Vindum. We are grateful to Pat Holroyd (UCMP) and Elizabeth Hill (CM) for access
to specimens in their care, to Healy Hamilton and Jeff Bada for their efforts with the Amino Acid
Racemization analysis, to the late Vic Schmidt for paleomagnetic analyses, and to the members of the
Colorado Grotto of the National Speleological Society and Don Rasmussen for preparing the map of
the cave and permitting us to publish it. Original artwork of the fossil specimens was prepared by
Kathryn G. Reynolds; Mary Poteet and Coco Kishi provided assistance with computer graphics.
Funding was provided by NSF Grant BSR-9 196082 (to Barnosky), and by the UCMP Samuel P.
Welles and Dorothy Hampton Welles Award and a travel grant from the Department of Integrative
Biology (to Bell). This paper was completed when ADB was in residence at the Mountain Research
Center, Montana State University.

Literature Cited

Agusti, J. 1992. The Allophaiomys complex in southern Europe. Geobios, 25:133-144.

Agusti, J., C. Castillo, and A. Galobart. 1993. Heterochronic evolution in the late Pliocene-early
Pleistocene arvicolids of the Mediterranean area. Quaternary International, 19:51-56.

130

Annals of Carnegie Museum

VOL. 69

Anderson, E. 1994. Evolution, prehistoric distribution, and systematics of Martes. Pp. 13-25, in
Martens, Sables and Fishers: Biology and Management (S. W. Buskirk, A. S. Harstead, M. G.
Raphael, and R. A. Powell, eds.). Cornell University Press, Ithaca, New York.

. 1996. A preliminary report on the Carnivora of Porcupine Cave, Park County, Colorado. Pp.

259-282, in Palaeoecology and Palaeoenvironments of Late Cenozoic Mammals: Tributes to the
Career of C. S. (Rufus) Churcher (K. M. Stewart and K. L. Seymour, eds.). University of Toronto
Press, Toronto, Ontario, Canada.

Andrews, P. 1990. Owls, Caves and Fossils. University of Chicago Press, Chicago, Illinois.

Archibald, J. D. 1993. The importance of phylogenetic analysis for the assessment of species turn-
over: A case history of Paleocene mammals in North America. Paleobiology, 19:1-27.

Armstrong, D. M. 1972. Distribution of mammals in Colorado. Monograph of the Museum of
Natural History, The University of Kansas, no. 3.

Bada, J. L. 1981. Racemization of amino acids in fossil bones and teeth from the Olduvai Gorge
region, Tanzania, East Africa. Earth and Planetary Science Letters, 55:292-298.

. 1985. Amino acid racemization dating of fossil bones. Annual Review of Earth and Planetary

Sciences, 13:241-268.

Barendregt, R. W., E E Thomas, E. Irving, J. Baker, A. M. Stalker, and C. S. Churcher. 1991.
Stratigraphy and paleomagnetism of the Jaw Face section, Wellsch Valley site, Saskatchewan.
Canadian Journal of Earth Sciences, 28:1353-1364.

Barnosky, a. D., and D. L. Rasmussen. 1988. Middle Pleistocene arvicoline rodents and environ-
mental change at 2900-meters elevation. Porcupine Cave, South Park, Colorado. Annals of Car-
negie Museum, 57:267-292.

Barnosky, A. D., E. A. Hadly, and C. J. Bell. 1996«. Climatic fluctuations and the evolution of
mammalian communities. American Quaternary Association, Program and Abstracts of the 14th
Biennial Meeting, Flagstaff, Arizona: 2-4.

Barnosky, A. D., T. I. Rouse, E. A. Hadly, D. L. Wood, E L. Keesing, and V. A. Schmidt, \996b.
Comparison of mammalian response to glacial-interglacial transitions in the middle and late Pleis-
tocene. Pp. 16-33, in Palaeoecology and Palaeoenvironments of Late Cenozoic Mammals: Trib-
utes to the Career of C. S. (Rufus) Churcher (K. M. Stewart and K. L. Seymour, eds.). University
of Toronto Press, Toronto, Ontario, Canada.

Barnosky, E. H. 1992. A positive role for taphonomy and time-averaging as illustrated by mammalian
fossils from Lamar Cave, Yellowstone National Park, WY. Geological Society of America Ab-
stracts with Programs, 24(6):2.

. 1994. Ecosystem dynamics through the past 2000 years as revealed by fossil mammals from

Lamar Cave in Yellowstone National Park, USA. Historical Biology, 8:71-90.

Bell, C. J. 1995. A middle Pleistocene (Irvingtonian) microtine rodent fauna from White Pine County,
Nevada, and its implications for microtine rodent biochronology. Journal of Vertebrate Paleon-
tology, 15 (Supplement to 3):18A.

. 1997. A revision of North American Irvingtonian (early and middle Pleistocene) microtine

rodent biochronology. Unpublished Ph.D. Dissert., University of California, Berkeley, California.

“. 1998. North American Quaternary Land Mammal Ages and the biochronology of North

American microtine rodents. Pp. 2-605-2-645, in. Dating and Earthquakes: Review of Quaternary
Geochronology and Its Application to Paleoseismology (J. M. Sowers, J. S. Noller, and W. R.
Lettis, eds.). Nuclear Regulatory Commission, NUREG/-CR, no. 5562.

Bell, C. J., and J. I. Mead. 1998. Late Pleistocene microtine rodents from Snake Creek Burial Cave,
White Pine County, Nevada. Great Basin Naturalist, 58:82-86.

Bell, C. J., and C. A. Repenning. 1999. Observations on dental variation in Microtus from the
Cudahy Ash Pit fauna, Meade County, Kansas and implications for Irvingtonian microtine rodent
biochronology. Journal of Vertebrate Paleontology, 19:757-766.

Betancourt, J. L., T. R. Van Devender, and R S. Martin (eds.). 1990. Packrat Middens: The Last
40,000 Years of Biotic Change. University of Arizona Press, Tucson, Arizona.

Brown, J. H., and A. C. Gibson. 1983. Biogeography. C. V. Mosby Company, Saint Louis, Missouri.

Churcher, C. S. 1984. Faunal correlations of Pleistocene deposits in western Canada. Pp. 145-158,
in Correlation of Quaternary Chronologies (W. C. Mahaney, ed.). Geo Books, Norwich, United
Kingdom.

CouES, E., AND H. C. Yarrow. 1875. Report upon the collections of mammals made in portions of
Nevada, Utah, California, Colorado, New Mexico, and Arizona, during the years 1871, 1872,
1873, and 1874. Pp. 35-129, in Report upon Geographical and Geological Explorations and
Surveys West of the One Hundredth Meridian, Volume 5. Government Printing Office, Washing-
ton, D.C.

Erbaeva, M. a. 1976. [The origin, evolution and intraspecific variation of Brandt’s vole from the

2000

Bell and Barnosky — Porcupine Cave Microtines

131

Anthropogene of western Transbaikal], Pp. 107-116, in Rodents Evolution and History of Their
Recent Fauna (1. M. Gromov, ed.). Academy of Sciences of the USSR Proceedings of the Zoo-
logical Institute, no. 66. (In Russian)

FAUNMAP Working Group. 1996. Spatial response of mammals to late Quaternary environmental
fluctuations. Science, 272:1601-1606.

Fejfar, O., and C. a. Repenning. 1992. Holarctic dispersal of the arvicolids (Rodentia, Cricetidae).
Pp. 205-212, in Mammalian Migration and Dispersal Events in the European Quaternary (W. von
Koenigswald and L. Werdelin, eds.). Courier Forschungsinstitut Senckenberg, no. 153.

Geological Society of America. 1984. Rock Color Chart. Geological Society of America, Boulder,
Colorado.

Gillette, D, D., C. J. Bell, and M. C. Hayden. 1999. Preliminary report on the Little Dell Dam
fauna. Salt Lake County, Utah (Middle Pleistocene, Irvingtonian land mammal age). Pp. 495-
500 in Vertebrate Paleontology in Utah (D. D. Gillette, ed.). Utah Geological Survey Miscella-
neous Publication 99-1.

Graham, R. W. 1979. Paleoclimates and late Pleistocene faunal provinces in North America. Pp. 49-
69, in Pre-Llano Cultures of the Americas: Paradoxes and Possibilities (R. L. Humphrey and D.
Stanford, eds.). The Anthropological Society of Washington, Washington, D.C.

Guilday, J. E. 1971. The Pleistocene history of the Appalachian mammal fauna. Pp. 233-262, in
The Distributional History of the Biota of the Southern Appalachians, Part III: Vertebrates (P. C.
Holt, R. A. Paterson, and J. P. Hubbard, eds.). Virginia Polytechnic Institute and State University
Research Division Monograph, no. 4.

Guilday, J. E., and M. S. Bender, i960. Late Pleistocene records of the yellow-cheeked vole,
Microtus xanthognathus (Leach). Annals of Carnegie Museum, 35:315-330.

Guilday, J. E,, and P. W. Parmalee. 1972. Quaternary periglacial records of voles of the genus
Phenacomys Merriam (Cricetidae: Rodentia). Quaternary Research, 2:170-175.

Guilday, J. E., P. W. Parmalee, and H. W. Hamilton. 1977. The Clark’s Cave bone deposit and the
late Pleistocene paleoecology of the central Appalachian mountains of Virginia. Bulletin of Car-
negie Museum of Natural History, no. 2.

Guilday, J. E., H. W. Hamilton, E. Anderson, and P W Parmalee. 1978. The Baker Bluff Cave
deposit, Tennessee, and the late Pleistocene faunal gradient. Bulletin of Carnegie Museum of
Natural History, no. 11.

Guilday, J. E., J. E P. Cotter, D. Cundall, E. B. Evenson, J. B. Gatewood, A. V. Morgan, A. D.
McCrady, D. M. Peteet, R. Stuckenrath, and K. Vanderwal. 1984. Paleoecology of an early
Pleistocene (Irvingtonian) cenote: Preliminary report on the Hanover Quarry No. 1 Fissure, Adams
County, Pennsylvania. Pp. 119-132, in Correlation of Quaternary Chronologies (W. C. Mahaney,
ed.). Geo Books, Norwich, United Kingdom.

Guthrie, R. D., and J. V Matthews, Jr. 1971. The Cape Deceit fauna — Early Pleistocene mam-
malian assemblage from the Alaskan Arctic. Quaternary Research, 1:474-510.

Hadly, E. a. 1996. Influence of Late-Holocene climate on northern Rocky Mountain mammals.
Quaternary Research, 46:298-310.

1999. Fidelity of terrestrial vertebrate fossils to a modern ecosystem. Palaeogeography, Pa-

laeoclimatology, Palaeoecology, 149:389-409.

Hall, E. R. 1981. The Mammals of North America. John Wiley and Sons, New York, New York.

Hallberg, G. R., H. a. Semken, and L. C. Davis. 1974. Quaternary records of Microtus xanthog-
nathus (Leach), the yellow-cheeked vole, from northwestern Arkansas and southwestern Iowa.
Journal of Mammalogy, 55:640-645.

Harris, A. H. 1988. Late Pleistocene and Holocene Microtus (Pitymys) (Rodentia: Cricetidae) in
New Mexico. Journal of Vertebrate Paleontology, 8:307-313.

— — — . 1993. Quaternary vertebrates of New Mexico. Pp. 179-197, in Vertebrate Paleontology in
New Mexico (S. G. Lucas and J. Zidek, eds.). New Mexico Museum of Natural History and
Science Bulletin, no. 2.

Hibbard, C. W. 1944. Stratigraphy and vertebrate paleontology of Pleistocene deposits of south-
western Kansas. Bulletin of the Geological Society of America, 55:707-754.

~ — — . 1952. Vertebrate fossils from late Cenozoic deposits of central Kansas. University of Kansas
Paleontological Contributions, Vertebrata, Article 2:1-14.

. 1970. Pleistocene mammalian local faunas from the Great Plains and central lowland prov-
inces of the United States. Pp. 395-433, in Pleistocene and Recent Environments of the Central
Great Plains (W. Dort, Jr., and J. K. Jones, Jr., eds.). Department of Geology, University of Kansas
Special Publication, no. 3.

- — . 1976. The localities of the Cudahy fauna, with a new ground squirrel (Rodentia, Sciuridae)

from the fauna of Kansas (late Kansan). Pp. 278-286, in Athlon: Essays on Palaeontology in

132

Annals of Carnegie Museum

VOL. 69

Honour of Loris Shano Russell (C. S. Churcher, ed.). Royal Ontario Museum, Life Sciences
Miscellaneous Publications, Toronto, Ontario, Canada.

Hinton, M. A. C. 1923. Diagnoses of species of Pitymys and Microtm occurring in the Upper
Freshwater Bed of West Runton, Norfolk. Annals and Magazine of Natural History, series. 9, 12:
541-542.

Holman, J. A. 1995. Pleistocene Amphibians and Reptiles in North America. Oxford University
Press, New York, New York.

Keesing, F. L. 1992. Mid-Irvingtonian occurrence of Allophaiomys pUocaenicus at Porcupine Cave,
Colorado. American Quaternary Association, 12th Biennia! Meeting, 24-26 August 1992, Pro-
gram and Abstracts: 46.

VON Koenigswald, W. 1980. Schmelzstruktur und Morphologic in den Molaren der Arvicolidae
(Rodentia). Abhandlungen der Senckenbergischen Naturforschenden Geselschaft, 539:1-129.

VON Koenigswald, W., and L. D. Martin. 1984fl. The status of the genus Mimomys (Arvicolidae,
Rodentia, Mamm.) in North America. Neues Jahrbuch ftir Geologic und Palaontologie Abhan-
dlimgen, 168:108-124.

. 1984/:?. Revision of the fossil and Recent Lemminae (Rodentia, Mammalia). Pp. 122-137,
in Papers in Vertebrate Paleontology Honoring Robert Warren Wilson (R. M. Mengel, ed.). Car-
negie Museum of Natural History Special Publication, no. 9.

Kormos, T. 1933. Neues Wiihlmause aus dem Oberpliocan von Piispokfurdo. Neues Jahrbuch fiir
Mineraiogie, Geologic, und Palaontologie Abhandlungen, Abteilung B, 69:323-346.

Krishtalka, L., and R. K. Stucky. 1985. Revision of the Wind River faunas. Early Eocene of central
Wyoming. Part 7. Revision of Diacodexis (Mammalia, Artiodactyla). Annals of Carnegie Muse-
um, 54:413-486.

Kurten, B., and E. Anderson. 1980. Pleistocene Mammals of North America. Columbia University
Press, New York, New York.

Martin, L. D. 1972. The microtine rodents of the Mullen Assemblage from the Pleistocene of north
central Nebraska. Bulletin of the Nebraska State Museum, 9:173-182.

. 1989. Plio-Pleistocene rodents in North America. Pp. 47-58, in Papers on Fossil Rodents in

Honor of Albert Elmer Wood (C. C. Black and M. R. Dawson, eds.). Natural History Museum
of Los Angeles County Science Series, no. 33.

Martin, L. D., and R. S. Hoffmann. 1987. Pleistocene faunal provinces and Holocene biomes of
the central Great Plains. Pp. 159—165, in Quaternary Environments of Kansas (W. C. Johnson,
ed.). Kansas Geological Survey Guidebook Series, no. 5.

Martin, R. A. 1975. Allophaiomys Kormos from the Pleistocene of North America. Pp. 97-100, in
Studies on Cenozoic Paleontology and Stratigraphy in Honor of Claude W. Hibbard (G. R. Smith,
and N. E. Friedland, eds.). University of Michigan Museum of Paleontology, Papers on Paleon-
tology, no. 12.

. 1987. Notes on the classification and evolution of some North American fossil Microtus

(Mammalia; Rodentia). Journal of Vertebrate Paleontology, 7:270-283.

-. 1989. Arvicolid rodents of the early Pleistocene Java local fauna from north-central South

Dakota. Journal of Vertebrate Paleontology, 9:438-450.

. 1993. Patterns of variation and speciation in Quaternary rodents. Pp. 226-280, in Morpho-
logical Change in Quaternary Mammals of North America (R. A. Martin and A. D. Barnosky,
eds.). Cambridge University Press, Cambridge, United Kingdom.

• . 1995. A new middle Pleistocene species of Microtus (Pedomys) from the southern United

States, with comments on the taxonomy and early evolution of Pedomys and Pitymys in North
America. Journal of Vertebrate Paleontology, 15: 17 1-186.

. 1996. Dental evolution and size change in the North American muskrat: Classification and

tempo of a presumed phyletic sequence. Pp. 431-457, in Palaeoecology and Palaeoenvironments
of Late Cenozoic Mammals; Tributes to the Career of C. S. (Rufus) Churcher (K. M. Stewart and
K. L. Seymour, eds.). University of Toronto Press, Toronto, Ontario, Canada.

Martin, R. A., and A. Tesakov. 1998. Introductory remarks: does Allophaiomys exist? Pp. 1-7 in
The early evolution of Microtus (R. A. Martin and A. Tesakov, eds.). Paludicola, 2.

Mead, E. M., and J. I. Mead. 1989. Snake Creek Burial Cave and a review of the Quaternary
mustelids of the Great Basin. Great Basin Naturalist, 49:143-154.

Mead, J. L, C. Manganaro, C. A. Repenning, and L. D. Agenbroad. 1996. Early Rancholabrean
mammals from Salamander Cave, Black Hills, South Dakota. Pp. 458-482, in Palaeoecology and
Palaeoenvironments of Late Cenozoic Mmammais: Tributes to the Career of C. S. (Rufus) Church-
er (K. M. Stewart and K. L. Seymour, eds.). University of Toronto Press, Toronto, Ontario, Canada.

Mezzabotta, C., F Masini, and D. Torre. 1995. Microtus {Tyrrhenicola) henseli, endemic fossil

2000

Bell and Barnosky — Porcupine Cave Microtines

133

vole from Pleistocene and Holocene localities of Sardinia and Corsica: Evolutionary patterns and
biochronological meaning. Bollettino della Societa Paleontologica Italiana, 34:81-104.

Miller, B. B., R. W. Graham, A. V. Morgan, N. G. Miller, W. D. McCoy, D. E Palmer, A. J.
Smith, and J. J. Pilny. 1994. A biota associated with Matuyama-age sediments in west-central
Illinois. Quaternary Research, 41:350-365.

Moore, D. W., and L. L. Janecek. 1990. Genic relationships among North American Microtus
(Mammalia: Rodentia). Annals of Carnegie Museum, 59:249-259.

Musser, G. G., and M. D. Carleton. 1993. Family Muridae. Pp. 501-755, in Mammal Species of
the World: A Taxonomic and Geographic Reference, Second Edition (D. E. Wilson and D. M.
Reeder, eds.). Smithsonian Institution Press, Washington, D.C.

Nadachowski, a., and I. Zagorodnyuk. 1996. Recent Allophaiomys-like species in the Palaearctic:
Pleistocene relicts or a return to an initial type. Acta Zoologica Cracoviensia, 39:387-394.

Opdyke, N. D., and j. E. T. Channell. 1996. Magnetic Stratigraphy. Academic Press, Inc., San
Diego, California.

Paulson, G. R. 1961. The mammals of the Cudahy fauna. Papers of the Michigan Academy of
Science, Arts, and Letters, 46:127-153.

Rabeder, G. 1981. Die Arvicoliden (Rodentia, Mammalia) aus dem Pliozan und dem alteren Pleis-
tozan von Niederosterreich. Beitrage zur Palaontologie von Osterreich, 8:1-373.

Raymo, M. E., D. W. Oppo, and W. Curry. 1997. The mid-Pleistocene climate transition: A deep
sea carbon isotope perspective. Paleoceanography, 12:546-559.

Rensberger, j. M., and A. D. Barnosky. 1993. Short-term fluctuations in small mammals of the
late Pleistocene from eastern Washington. Pp. 299-342, in Morphological Change in Quaternary
Mammals of North America (R. A. Martin and A. D. Barnosky, eds.). Cambridge University
Press, Cambridge, United Kingdom.

Rensberger, J. M., A. D. Barnosky, and P. Spencer. 1984. Geology and paleontology of a Pleis-
tocene-to-Holocene loess succession near Kennewick, Benton County, Washington. Eastern Wash-
ington University Reports in Archaeology and History, 100-39:1-105.

Repenning, C. A. 1980. Faunal exchanges between Siberia and North America. Canadian Journal of
Anthropology, 1:37-44.

■. 1983. Pitymys meadensis Hibbard from the Valley of Mexico and the classification of North

American species of Pitymys (Rodentia: Cricetidae). Journal of Vertebrate Paleontology, 2:471-
482.

— . 1987. Biochronology of the microtine rodents of the United States. Pp. 236-268, in Cenozoic

Mammals of North America: Geochronology and Biostratigraphy (M. O. Woodburne, ed.). Uni-
versity of California Press, Berkeley, California.

— “. 1992. Allophaiomys and the age of the Olyor Suite, Krestovka Sections, Yakutia. U.S. Geo-
logical Survey Bulletin, 2037:1-98.

Repenning, C. A., and F. Grady. 1988. The microtine rodents of the Cheetah Room fauna, Hamilton
Cave, West Virginia, and the spontaneous origin of Synaptomys. U.S. Geological Survey Bulletin,
1853:1-32.

Repenning, C. A., E. M. Brouwers, L. D. Carter, L. Marincovich, Jr., and T. A. Ager. 1987. The
Beringian ancestry of Phenacomys (Rodentia: Cricetidae) and the beginning of the modern Arctic
Ocean borderland biota. U.S. Geological Survey Bulletin, 1687:1-31.

Repenning, C. A., O. Fejfar, and W.-D. Heinrich. 1990. Arvicolid rodent biochronology of the
Northern Hemisphere. Pp. 385-417, in International Symposium: Evolution, Phylogeny and Bio-
stratigraphy of Arvicolids (Rodentia, Mammalia) (O. Fejfar and W.-D. Heinrich, eds.). Geological
Survey, Prague, Czechoslovakia.

Repenning, C. A., T. R. Weasma, and G. R. Scott. 1995. The early Pleistocene (latest Blancan-
earliest Irvingtonian) Froman Ferry fauna and history of the Glenns Ferry Formation, south-
western Idaho. U.S. Geological Survey Bulletin, 2105:1-86.

Rogers, K. L., C. A. Repenning, R. M. Forester, E. E. Larson, S. A. Hall, G. R. Smith, E.
Anderson, and T J. Brown. 1985. Middle Pleistocene (late Irvingtonian; Nebraskan) climatic
changes in south-central Colorado. National Geographic Research, 1:535-563.

Rogers, K. L., E. E. Larson, G. Smith, D. Katzman, G. R. Smith, T. Cerling, Y. Wang, R. G.
Baker, K. C. Lohmann, C. A. Repenning, P. Patterson, and G. Mackie. 1992. Pliocene and
Pleistocene geologic and climatic evolution in the San Luis Valley of south-central Colorado.
Palaeogeography, Palaeoclimatology, Palaeoecology, 94:55-86.

Rogers, R. A., L. D. Martin, and T. D. Nicklas. 1990. Ice-age geography and the distribution of
native North American languages. Journal of Biogeography, 17:131-143.

Rouse, T. I. 1997. Climate change and morphological evolution in the genus Cynomys. Unpublished
Ph.D. Dissert., University of California, Berkeley, California.

134

Annals of Carnegie Museum

VOL. 69

Savage, D. E. 1951. Late Cenozoic vertebrates of the San Francisco Bay region. University of
California Publications, Bulletin of the Department of Geological Sciences, 28:215-314.

Smartt, R. a. 1972. Late Pleistocene and Recent Microtus from southcentral and southwestern New
Mexico. Unpublished M.S. Thesis, Department of Biology, The University of Texas at El Paso,
El Paso, Texas.

. 1977. The ecology of late Pleistocene and Recent Microtus from south-central and south-
western New Mexico. The Southwestern Naturalist, 22:1-19.

Stebbins, R. C. 1985. A Field Guide to Western Reptiles and Amphibians. Houghton Mifflin Com-
pany, Boston, Massachusetts.

Van Der Meulen, A. J. 1978. Microtus and Pitymys (Arvicolidae) from Cumberland Cave, Maryland,
with a comparison of some New and Old World species. Annals of Carnegie Museum, 47:101-
145.

Wood, D. L., and A. D. Barnosky. 1994. Middle Pleistocene climate change in the Colorado Rocky
Mountains indicated by fossil mammals from Porcupine Cave. Quaternary Research, 41:366-375.

Zakrzewski, R. J. 1972. Fossil microtines from late Cenozoic deposits in the Anza-Borrego Desert,
California, with the description of a new subgenus of Symiptomys. Natural History Museum of
Los Angeles County Contributions in Science, 221:1-12.

ANNALS OF CARNEGIE MUSEUM

VoL. 69, Number 2, Pp. 135-144

23 May 2000

NEW LOWER MISSISSIPPIAN TRILOBITES FROM THE CHOUTEAU

GROUP OF MISSOURI

David K. BrezinskP

Research Associate, Section of Invertebrate Paleontology

Abstract

Reexamination of existing trilobite collections from the Kinderhookian (Lower Mississippian) Chou-
teau Group of central and northeastern Missouri indicates that two different suites of trilobites are
present in these two areas of the state. Moreover, the study of these collections has led to the erection
of a new genus and four new species. The new genus, Ameropiltonia, is based on a new species, A.
lauradanae. This genus and species is commonly confused with Breviphillipsia sampsoni (Vogdes).
ElUptophillipsia rotundas, n. sp., differs from the type species of this genus by possessing a rounded
frontal lobe to the glabella. The other new species, Perexigupyge chouteauensis and Richterella hes-
sleri, are present in the Compton Limestone of Marion and Ralls counties of northeastern Missouri.

Variations in trilobite species found in the Compton Limestone of central Missouri and the north-
eastern part of the state are interpreted to be environmentally related. It appears that the lime mudstone
and wackestone lithologies characteristic of the Compton Limestone of central Missouri were deposited
in a low-energy, subtidal shelf setting. The lime packstone-grainstone strata of northeastern Missouri
are interpreted to have formed as a tidal sand belt on the eastern margin of the Burlington shelf.

Key Words: Missouri, trilobites, Mississippian, Kinderhookian, Chouteau

Introduction

Trilobites are well known from the Lower Mississippian (Kinderhookian)
Chouteau Group of Missouri. Vogdes (1888, 1891), Branson and Andrews (1938),
Hessler (1963, 1965), and Brezinski (1986, 1988^?) are several of the more salient
references dealing with trilobites from the Kinderhookian strata of Missouri. Col-
lection of new material and reevaluation of preexisting collections have brought
to light a number of previously undescribed species and a new genus. Moreover,
reexamination of known collections indicates that broadly distributed biofacies
were present during the Kinderhookian. These regional trilobite biofacies parallel
distinct lithofacies within the Chouteau Group. The purpose of this paper is to
briefly describe the lithofacies and associated trilobite faunas present in the Chou-
teau Group of Missouri, and to describe a new genus and four new species of
trilobites.

The trilobites described in this report were recovered from three localities.
Locality 1 is located in central Missouri along the abandoned MKT Railroad right-
of-way, at the base of the river bluffs of the Missouri River, one mile north of
the town of Easley, Boone County. The remaining two localities are located in
northeastern Missouri. Locality 2 is along County Road 183, one mile south of
Warren, Marion County. Locality 3 is located along the spillway of the Clarence
Cannon Dam, Ralls County.

' Maryland Geological Survey, 2300 Saint Paul Street, Baltimore, Maryland 21218.
Submitted 4 April 1998.

135

136

Annals of Carnegie Museum

VOL. 69

El] ^ ^ ^

Emergent Shelf

Dolomite Limestones Shales

Fig. 1. — Generalized early Mississippian paleogeography of Missouri (modified from Lane and De
Keyser, 1980:fig. 8). Location of the three trilobite localities (labeled 1-3) used in this study with
their relative stratigraphic position illustrated on cross section. Cross-section A-B-C section line il-
lustrates relative stratigraphic distribution of formations within the Chouteau Group of Missouri, mod-
ified from Thompson (1979). No vertical scale implied.

Morphological terminology used in this paper follows that described by Whit-
tington (1997). Collections utilized in this study were made by J. L. Carter, A.
Kollar, and the author and are reposited in the invertebrate fossil collections of
Carnegie Museum of Natural History (CM).

Regional Facies

The Kinderhookian strata of Missouri is a complex mosaic of intertonguing
limestone, dolomite, and shale that were lumped together as the Chouteau Group
by Thompson (1979). The composite units include the Northview Shale, Sedalia
Dolomite, Compton Limestone, and Hannibal Shale (Fig. 1). The composite units
that make up the Chouteau Group, and especially the Compton Limestone, exhibit
a broad facies change from central to eastern Missouri (Thompson, 1979). In
central Missouri (Boone and Pettis counties), where the trilobite faunas have been
well described (Branson and Andrews, 1938; Hessler, 1963, 1965; Brezinski,
1986, 1988a), the main fossiliferous unit, the Compton Limestone, consists of

2000

Brezinski — New Mississippian trilobites

137

nodular- to medium-bedded, argillaceous, fossiliferous, lime mudstone to wack-
estone. King (1980) interpreted this lithology as having formed in an open shelf
environment of deposition, below storm wave base. Trilobites characteristic of
this facies include Breviphillipsia sampsoni (Vogdes), Comptonaspis swallowi
(Shumard), Griffithidella welleri (Branson and Andrews), Dixiphopyge arma-
tus(Vogdes), and Ameropiltonia lauradanae, n. gen. and n. sp. Uncommon faunal
components include Proetides colemani Hessler, Elliptophillipsia rotundas, n. sp.,
and Brachymetopus brezinskii Hahn and Hahn.

To the east, in Marion and Ralls counties (Fig. 1), the lime wackestone of the
Compton Limestone is replaced through lateral facies change. In this area an
unnamed Compton Limestone equivalent consists of medium-bedded, locally
cross-bedded, highly fossiliferous lime packstone to grainstone. The unnamed
Compton equivalent appears to have been deposited at a much higher energy level
than the nodular-bedded wackestone of the central part of the state. This is in-
dicated by the lack of carbonate mud and the presence of cross-bedded grain-
stones. This regional facies variation is interpreted to represent the change from
subtidal, low-energy, carbonate mud deposition of the inner shelf to tidally influ-
enced, sand deposition of the outer shelf. Consequently, the grainstone facies
within the Chouteau of northeastern Missouri are interpreted as representing a
shelf margin shoal deposits.

The trilobite fauna present in the packstone-grainstone facies of the Compton
Limestone of northeastern Missouri is quite different than that of the argillaceous,
lime mudstone- wackestone of central Missouri. In northeastern Missouri the most
common trilobites include Griffithidella newarkensis Hessler, Proetides insignis
(Winchell), Richterella snakedenensis Hessler, and Perexigupyge chouteauensis,
n. sp. A rarer component is Richterella hessleri, n. sp. This fauna is interpreted
to have inhabited high-energy environments of deposition rather than the quieter
subtidal environments which existed in central Missouri.

In western Illinois, the equivalent Chouteau Limestone undergoes an additional
facies change and rapidly thins eastward (Lineback, 1969). In what is now the
Mississippi River Valley, the grainstone facies is replaced by a thin (< 3 m) dark-
gray, bioturbated, lime wackestone. Brezinski (1998) interpreted this facies as
having formed in a deep-water, sediment-starved setting. The depth of the water
in this basinal environment may have exceeded 180 m (Lineback, 1969).

The trilobite fauna from the Chouteau Limestone of Illinois differs from either
of the Chouteau lithofacies observed in Missouri. The Illinois fauna consists of
Pudoproetus chappelensis (Hessler), Griffithidella doris (Hall), Phillibole planu-
caudus (Brezinski), and Thigriffides roundyi (Girty). Brezinski (1998) proposed
that the Illinois fauna was similar to that found in the Welden Limestone of
Oklahoma and the Chappel Limestone of Texas because these areas shared a
similar depositional setting. These interpreted deep-water species are generically
different from the fauna found within the presumed shelf deposits of Missouri.

Systematic Paleontology

Order Proetida Fortey and Owens, 1975
Family Phillipsiidae Oehlert, 1886
Genus Ameropiltonia, new genus

Type Species. — Ameropiltonia lauradanae, n. sp.

Other Species Assigned. — Only the type species is currently assigned.

138

Annals of Carnegie Museum

VOL. 69

Diagnosis. — Exoskeleton highly vaulted, covered with coarse tubercles. Ce-
phalon strongly arched in transverse and longitudinal profile. Genal angle round-
ed. Glabella parallel-sided to mildly forward-tapering, reaching to overhanging
anterior border. Pygidium semicircular, ribs sharply overhanging pleural furrows,
interpleural furrow lacking.

Remarks. — In his erection of the genus Breviphillipsia, Hessler (1963) desig-
nated Phillipsia sampsoni Vogdes as the genotype (i.e., type species). As a type
species, B. sampsoni has been somewhat of an enigma because Hessler (1963:pl.
61 and 62) included a number of specimens that differ considerably from the
holotype. It appears that all of the specimens, except the holotype, that Hessler
assigned to B. sampsoni differ in various characters from the holotype of Phillip-
sia sampsoni. The holotype (Hessler, 1963:pl. 61; fig. 15-17, 21) has a forwardly
tapering, broadly rounded glabella that is mildly constricted at y. The anterior
facial sutures diverge from y to (3, and the posterior segment has a relatively long
and straight e to ^ section. The anteriorly rounded glabella extends to the anterior
border furrow, but not to the anterior margin, and the pygidium has a well-defined
interpleural furrow which gives the pygidial ribs anterior and posterior bands.
Furthermore, the glabella is covered with small granules and the pygidium has
no prosopon. Most of the coarsely ornamented specimens illustrated by Hessler
(1963:pl. 61, fig. 22, 23, 27, 28; pi. 62, fig. 1) are not assignable to B. sampsoni.
The disparity of shape of the glabellae among the specimens suggests that the
coarsely ornamented specimens do not even belong to the genus Breviphillipsia.
The species B. sampsoni is herein restricted to those specimens from central
Missouri similar to the holotype in that they exhibit the tongue-shaped glabella
that extends to the anterior border furrow and a pygidium that exhibits both
pleural and interpleural furrows. Thus, most fossils referred to B. sampsoni (see
Levi-Setti, 1975; Brezinski, 1986:fig. 4) belong to both a different genus and
species. A new genus, Ameropiltonia, is established to include species from the
Compton Limestone of Missouri that exhibit the coarsely tuberculate prosopon.

Comparison. — Ameropiltonia is similar to Piltonia in shape of the glabella,
ornamentation, and outline and character of the pygidium. It is distinguished from
the latter genus by the lack of a preglabellar field, anterior facial sutures that trace
very close to the dorsal furrow rather than diverge anteriorly, and a shorter py-
gidium that has fewer rings and ribs. American representatives of Piltonia such
as P. tuberculata (Meek and Worthen, 1870) and P. eurybathrea (Hessler, 1963)
have much longer pygidia, a short preglabellar field, and anterior facial sutures
that diverge moderately from y to (3. Another closely related genus, Eocyphinium,
has a forwardly expanding glabella and longer pygidia than Ameropiltonia.

Range. — Late Kinderhookian.

Ameropiltonia lauradanae, new species
(Fig. 2A-E)

Breviphillipsia sampsoni Hessler, 1963:pl. 61, fig. 13, 19, 20 (not pi. 61, fig. 15-17, 21, 22, 27, 28;
pi. 62, fig. 1); Hahn and Hahn, 1972:120-121; Brezinski, 1986:fig. 4.1, 4.5, 4.8.

Diagnosis. — Highly vaulted species with coarse tuberculation. Glabella outline
quadrate, slightly anteriorly tapering to parallel-sided, broadly rounded anteriorly,
strongly arched in longitudinal and transverse profile, overhanging anterior mar-
gin. Dorsal furrow deep, narrow. Palpebral lobes small, anteriorly located. Pygid-

2000

Brezinski — New Mississippian trilobites

139

Fig. 2. — A-E. Ameropiltonia laurcidanae , n. gen. and n. sp. A-D. Holotype specimen from the Comp-
ton Limestone, Boone County, Missouri, Locality 1, CM 451 15, X 3.5. D. Paratype specimen from
Locality 1, CM 451 16, X 3.0. F— J. ElliptophilUpsia rotundus, n. sp. F, H-J. Holotype specimen from
the basal Compton Limestone, Boone County, Missouri, Locality 1, CM 45 11 8, X 4.0. G. Paratype
specimen, CM 45119, X 4.0.

ium highly vaulted, with robust, wide, rounded axis, and deeply incised pleural
furrows.

Holotype. — A complete exoskeleton from the Compton Limestone at locality
1, Boone County, Missouri, CM 45115, collected by the author.

Paratypes. — A partial exoskeleton and a pygidium from locality 1, CM 45116 and 451 17, collected
by the author.

Description. — Exoskeleton outline oval, vaulting high. Cephalon outline parabolic, genal angle

140

Annals of Carnegie Museum

VOL, 69

sharply rounded. Glabella with straight sides, nearly parallel to very slightly forward-tapered, broadly
rounded anteriorly; sides vertical, mildly arched crest in transverse profile; strongly longitudinally
arched. Frontal lobe with prosopon of coarse polygonal tubercles, meeting and/or overhanging anterior
border. SI narrow, deep, straight; S2-S4 marked by separation of the coarse prosopon. LI outline
triangular. Dorsal furrow deepest between eyes, shallowest anteriorly. Occipital furrow deep, sinuous,
overhung by LI. Occipital lobe wide at axial line. Palpebral lobes small, crescentic, ornamented by a
line of tubercles along outer edge. Anterior branch of facial sutures parallel and close to sides of
glabella, sharply rounded at (5. Posterior branch of facial sutures with a short straight section at e.
Eyes small, reniform, with vertical sides, and a row of tubercles at the base. Librigenae descending
steeply from eye to narrow, shallow, lateral border furrow. Lateral border sharply crested, underturned
on its outer edge, with a row of small tubercles on its inner edge. Posterior border furrow deep,
relatively wide. Posterior cephalic border wide, with a row of posteriorly directed tubercles at crest.

Thorax of nine equal tergites. Axial lobe robust, rings wide, semicircular in transverse profile
ornamented with two rows of small tubercles. Ring furrows deep, narrow, sinuous. Dorsal furrow,
deep, relatively straight. Pleural fields proximally flattened, shaiply curved at fulcrum, nearly vertical
at distal ends. Pleural tips sharply angular.

Pygidium outline semicircular, highly vaulted. Axis robust, wide, semicircular in transverse profile,
mildly arched in longitudinal profile, sharply rounded at posterior terminus, not reaching posterior of
pygidium, composed of ten to 1 1 rings. Rings wide, sharply crested, ornamented with a row of coarse
tubercles that are posteriorly directed, deep. Ring furrows deep, straight to slightly bowed forward
over axis. Pleural fields strongly arched in transverse profile, mildly arched in longitudinal profile,
composed of ten to 1 1 ribs that extend behind the axis. Posteriormost ribs defined by rows of tubercles.
Anterior ribs wide, sharply crested, and ornamented with a row of coarse tubercles that overhang very
deep pleural furrows that extend to and overhang margin.

Remarks, — Brezinski (1986:fig. 4) interpreted variations in coarseness of the
prosopon within Breviphillipsia sampsoni as the result of phenotypic plasticity of
individuals occurring in different depositional environments. The taxonomic re-
finement and reassignment in the current paper suggest otherwise. The shape of
the glabella and presence of interpleural furrows on the individuals illustrated in
figures 4.3 and 4.4 suggest that these specimens may be assignable to Breviphil-
lipsia. Figures 4.1, 4.5, and 4.8 lack interpleural furrows, and thus appear to be
assignable to Ameropiltonia. It is not clear now whether the coarsely ornamented
specimens illustrated by Hessler (1963:pl. 61, fig. 22, 27, 28; pi. 62, fig. 1) are
compacted specimens assignable to A. lauradanae or to a different species of
Ameropiltonia.

Distribution. — Chouteau Formation of Boone County, Missouri; Cuyahoga For-
mation of Ohio.

Etymology. — Named for Laura Dana Brezinski.

Genus Elliptophillipsia Hessler, 1963
Elliptophillipsia rotundas, new species
(Fig. 2F-J)

Diagnosis. — Glabella long, broadly rounded in the frontal lobe, slightly con-
stricted at the palpebral lobes. Anterior border narrow, sharply rounded at crest,
SI arcuate; LI suboval.

Holotype. — A nearly complete exoskeleton from the basal Compton Limestone
at locality 1, Boone County, Missouri (see Brezinski, 1986:location 1), CM 45118,
collected by the author.

Paratypes. — A partial exoskeleton and a cranidium, from the same locality, CM 45119, 45120,
collected by the author.

Description. — Exoskeleton outline elliptical, vaulting low. Cephalon outline parabolic. Glabella
long, narrow, slightly constricted between anterior end of palpebral lobes, broadly rounded in the
frontal lobe, exhibiting granular prosopon. SI relatively deep, arcuate; LI suboval. S2-S3 faint. Oc-

2000

Brezinski — New Mississippian trilobites

141

cipital furrow straight, shallow, narrow; occipital lobe wide at axial line. Dorsal furrow relatively deep,
slightly sinuous. Anterior border narrow, sharply crested; anterior border furrow deep in front of
glabella and much shallower in front of anterior fixigenae. Palpebral lobes large, about one-quarter of
total cranidial length, anteriorly located. Anterior section of the facial sutures straight, slightly ante-
riorly divergent; posterior section with a long straight section at e. Eyes large, about one-third of total
cranidial length, outline hemispherical; ocular platform naiTow, flat. Lateral border furrow broad, deep,
shallowing and narrowing posteriorly, extending well out onto genal spine. Border sharply crested,
narrow. Posterior border fun'ow shallow, narrow; posterior border wide, evenly convex. Genal spine
short, stout.

Thorax of nine subequal segments. Axis wide, evenly arched in transverse profile, rings wide,
furrows narrow. Dorsal furrow narrow, relatively deep. Pleural fields nan'ow, flat adjacent to dorsal
furrow, gently arched at fulcrum, descending to sharply rounded pleural tips.

Pygidium short, outline semicircular, vaulting low. Axis narrow, sharply posteriorly tapering, acutely
rounded at terminus, approximately 0.8 total pygidial length. Axis semicircular in transverse profile,
straight in longitudinal profile, composed of nine rings ornamented at crest with a row of small
granules. Ring furrows shallow, narrow, deepest over axis. Dorsal fuiTow nanow, relatively deep.
Pleural fields mildly arched in transverse profile, flat in longitudinal profile, composed of eight to nine
ribs. Anterior ribs composed of a broad anterior band and a short, nanow posterior band. Posterior
ribs not separated into bands. Border narrow, flat.

Remarks. — ■ElUptophillipsia rotundus, n. sp., differs from E. ellipticus (Meek

and Worthen) by having a constriction to the glabella between the palpebral lobes,
and a frontal lobe that is rounded. The holotype of ElUptophillipsia ellipticus (see
Hessler, 1963:pl. 62, fig. 16, 17, 21) has a forward-tapering glabella that is squared
off in front, and an anterior border furrow that is wider and deeper. No other
North American Carboniferous trilobite genus displays such an axial elongation
to the glabella.

The long, tongue-shaped glabella of Elliptophillisia rotundus is similar to that
seen in the European genus Linguaphillipsia. Linguaphillipsia differs from Ellip-
tophillipsia by having an elongate pygidium.

Distribution. — Late Kinderhookian of Boone County, Missouri, and Jersey
County, Illinois.

Etymology. — Rotunda, Latin, refers to the rounding of the anterior of the glabella.

Genus Perexigupyge Brezinski, 1988Z?

Perexigupyge chouteauensis, new species
(Fig. 3A-G)

Diagnosis. — Glabella tongue-shaped, widest between the palpebral lobes. Py-
gidium outline semicircular, evenly convex, low relief.

Holotype. — A fragmentary cranidium, from the grainstone facies of the Comp-
ton Limestone, at locality 2, Marion County, Missouri, CM 45121, collected by
the author.

Paratypes. — Four pygidia and an external and internal mold of a cranidium from localities 2 and
3 of Marion and Ralls counties, Missouri, CM 45122, 45123, collected by A. Kollar, J. L. Carter, and
the author.

Description. — Cranidium low in relief and vaulting. Glabella long, smooth, tongue-shaped, widest
between the palpebral lobes, tapering forward to about y, rounded anteriorly, extending to broad,
shallow, anterior border furrow. Anterior border rounded. SI short, shallow, arcuate; S2 short, per-
pendicular to shallow dorsal furrow; S3 slightly anteriorly directed adaxially. LI small, oval, incom-
pletely isolated from frontal lobe. L2-L3 small, faint. Palpebral lobes long, nanow. Anterior branch
of facial sutures with a long, straight section, diverging moderately from y to (3, shaiply rounded at
(3. Occipital furrow narrow, deep, straight. Occipital lobe wide, flat.

Pygidium outline semicircular, vaulting low. Axis short, narrow, mildly arched in transverse profile,
straight in longitudinal profile, strongly posteriorly tapering, sharply rounded at terminus, reaching
0.75 total pygidial length, composed of 1 1 to 12 smooth, wide (transverse) rings. Ring furrows shallow.

142

Annals of Carnegie Museum

VOL. 69

Fig. 3. — A-G. Perexigupyge chouteauensis, n. sp. A. Partially exfoliated holotype cranidium. Locality
2, Marion County, Missouri, CM 45121, X 4.0. B. Cast of external mold of partial paratype cranidium,
CM 45122a, X 4.0. C. Internal mold of paratype cranidium, CM 45122b, X 3.5. D. Nearly complete
paratype pygidium, CM 45123a, X 3.0. E-G. Dorsal, posterior, and lateral views of paratype pygidium,
CM 45123b, X 4.0. H— J. Richterella hessleri, n. sp., partial holotype cephalon from Locality 2, Marion
County, Missouri, CM 45 124, X 3.5.

narrow, straight, shallower laterally. Pleural fields mildly arched in transverse profile, straight in lon-
gitudinal profile, composed of eight to nine ribs. Ribs consist of broad, slightly elevated anterior
bands, and narrower, lower, posterior bands. Interpleural furrows shallow. Anterior bands extend onto
border nearly to margin. Border relatively wide and of equal width along entire margin, exhibits same
slope as pleural fields. Margin sharply iinderturned.

Remarks -Perexigupyge chouteauensis, n. sp., differs from P. hodgesi Brezin-
ski, 1988, by having a cranidium with much lower relief, and a glabella that is
tongue-shaped rather than nearly cylindrical. The pygidium of P. chouteauensis
is much lower in relief, and has more poorly defined pygidial rings and ribs.
Perexigupyge gerki Brezinski, 1988, is similar to P. chouteauensis with the for-
wardly tapering glabella, but differs from the latter species by having a more
triangular outline to the glabella, narrower, more forwardly located palpebral
lobes, and more arcuate trace to the anterior facial sutures which causes them to
appear more broadly divergent. The low relief and vaulting of P. chouteauensis
is similar to species of Spergenaspis as illustrated by Brezinski (1987). Indeed,
the shape of the glabella and smoothness of the pygidium are suggestive of a
species of Spergenaspis. However, P. chouteauensis lacks a preglabellar field

2000

Brezinski — New Mississippian trilobites

143

which characterizes Spergenaspis, and possesses the extension of the anterior band
of the pygidial ribs onto the border, a character present in Perexigupyge. It seems
plausible, inasmuch as both genera tend to be present in similar environmental
settings, that morphological convergence occurred between these two genera.
Conversely, it is possible that Spergenaspis descended from Perexigupyge.

Distribution. — Present in the late Kinderhookian of Marion and Ralls counties,
Missouri.

Etymology. — Named for the Chouteau Group from which the type material was collected.

Genus Richterella Hessler, 1965
Richterella hessleri, new species
(Fig. 3H-J)

Diagnosis. — Cranidium moderately vaulted, glabella smooth, with parallel sides
and a slight constriction between the palpebral lobes, anterior border very narrow,
rounded, border furrow absent. Palpebral lobes narrow, facial sutures with short
posterior section, and long, straight, slightly diverging anterior section.

Holotype. — An incomplete cephalon from the grainstone facies of the Compton
Limestone of locality 2, Warren, Marion County, Missouri, CM 45124, collected
by the author.

Description. — Cephalon semicircular to slightly parabolic in outline, moderately vaulted, relief low.
Cranidium longitudinal profile flat becoming more strongly arched anteriorly, evenly rounded in trans-
verse profile. Glabella with nearly parallel sides, a very slight constriction between the palpebral lobes.
SI shallow, arcuate; LI suboval. S2 straight, shallow; L2 subrectangular. S3 and L3 obsolete. Frontal
lobe bluntly rounded anteriorly, meeting a very narrow, rounded border. Anterior branch of facial
sutures long, straight, only slightly diverging from 7 to (3, sharply rounded at p. Posterior section of
facial sutures with a short, straight, parallel section, otherwise posteriorly diverging. Occipital furrow
shallow, narrow, straight except where overhung by LI; occipital lobe broad, flat. Dorsal furrow,
narrow, relatively deep between eyes, shallower forward. Palpebral lobes long, narrow. Eyes large,
with distinct, deep, ocular furrow that widens in front of the eye. Genal fields smooth, slightly inflated,
steeply descending into deep, narrow, lateral border furrow that becomes broader and shallower an-
teriorly. Lateral border narrow in dorsal view, relatively broad, evenly rounded in lateral view, with
faint terrace lines. Posterior border furrow, deep, narrow, directed to the posterior laterally; posterior
border wide, evenly rounded.

Remarks. — Richterella hessleri, n. sp., is diagnosed by the shape of the glabella
which is parallel" sided, and by the bluntly rounded frontal lobe and narrow an-
terior border. Richterella hessleri can be distinguished from R. snakedenensis
Hessler, 1963, by the more posterior location of the palpebral lobes, medially con-
stricted glabella, and wider anterior border in the latter species. The shape of the
glabella is similar to Perexigupyge hodgesi Brezinski, 1988. Richterella hessleri
can be distinguished from P. hodgesi by the presence of an anterior border furrow,
and by the more distinctively deeper dorsal, occipital, and glabellar furrows in
the latter species.

Distribution. — Known from the late Kinderhookian Chouteau Group of Marion
County, Missouri.

Etymology. — Named in honor of R. R. Hessler who erected the genus Richterella.

Acknowledgments

Collections from Marion County, Missouri, were made available by J. L. Carter and A. Kollar of
the Carnegie Museum of Natural History, Section of Invertebrate Paleontology. Reviewers’ comments
are greatly appreciated. Photographic facilities were provided by the Maryland Geological Survey.

144

Annals of Carnegie Museum

VOL. 69

Literature Cited

Branson, E. B., and D. Andrews. 1938. Subclass Trilobita. Pp. 113-122, in Stratigraphy and Pale-
ontology of the Lower Mississippian of Missouri, Part I (E. B. Branson), University of Missouri
Studies, 13.

Brezinski, D. K. 1986. Trilobite associations from the Chouteau Formation (Kinderhookian) of central
Missouri. Journal of Paleontology, 60:870-881.

. 1987. Spergenaspis: A new Carboniferous trilobite genus from North America. Annals of

Carnegie Museum, 56:245-251.

— . 1988fl. Revision and redescription of some Lower Mississippian trilobites from the Chouteau

Formation (Kinderhookian) of central Missouri. Journal of Paleontology, 62:103-110.

. 19881?. Trilobites of the Gilmore City Limestone (Mississippian) of Iowa. Journal of Pale-
ontology, 62:241-245.

. 1998. Lower Mississippian trilobites from starved basin deposits of the south-central United

States. Journal of Paleontology, 72:718-725.

Fortey, R. a., and R. M. Owens. 1975. Proetida: A new order of trilobites. Fossils and Strata, 4:
227-239.

Hahn, G., and R. Hahn. 1972. Trilobitae Carbonic! et Permici III. Pp. 333-531, in Fossilium Catal-
ogus 1, Animalia (F. Westphal, ed.). Ysel Press, Gravenhague, The Netherlands.

Hessler, R. R. 1963. Lower Mississippian trilobites of the Family Proetidae in the United States,
Part I. Journal of Paleontology, 37:543-563.

— 1965. Lower Mississippian trilobites of the Family Proetidae in the United States, Part II.

Journal of Paleontology, 39:248-264.

King, D. T. 1980. Genetic stratigraphy of the Mississippian System in central Missouri. Unpublished
Ph.D. Dissert., University of Missouri, Columbia, Missouri.

Levi-Setti, R. 1975. Trilobites. University of Chicago Press, Chicago, Illinois.

Lane, H. R., and T L. De Keyser. 1980. Paleogeography of the late early Mississippian (Tournaisian
3) in the central and southwestern United States. Pp. 149-162, mPaleozoic Paleogeography of
the West-central United States (T. D. Fouch and E. P. Magathan, eds.). West-central United States
Paleogeography Symposium 1, Rocky Mountain Section of the Society of Economic Paleontol-
ogists and Mineralogists.

Lineback, J. a. 1969. Illinois Basin — Sediment-starved during the Mississippian. American Associ-
ation of Petroleum Geologists Bulletin, 53:112-126.

Meek, F. B., and A. H. Worthen. 1870. Description of new species and genera of fossils from the
Palaeozoic rocks of the western states. Proceedings of the Academy of Natural Sciences of Phil-
adelphia, 2:22-56.

Oehlert, D. V. 1886. Etude sur quelques trilobites du groupe des Proetidae. Bulletin de la Societe
d’Etudes scientifiques d’Angers, 15:121-143.

Thompson, T. L. 1979. The Mississippian and Pennsylvanian (Carboniferous) systems in the United
States — Missouri. U.S. Geological Survey Professional Paper 1110.

VoGDES, A. W. 1888. Description of two new species of North American Carboniferous trilobites.
New York Academy of Science Annals, 4:69-105.

. 1891. On some new Sedalia trilobites. Saint Louis Academy of Science Transactions, 5:615-

618.

Whittington, H. B. 1997. Morphology of the exoskeleton. Pp. 1-85, in Treatise on Invertebrate
Paleontology, Trilobita, Part O, Revised (H. B. Whittington, B. D. E. Chatterton, S. E. Speyer, R.
A. Fortey, R. M. Owens, W. T. Chang, W. T. Dean, R A. Jell, J. R. Laurie, A. R. Palmer, L. N.
Repina, A. W. A. Rushton, J. H. Shergold, E. N. K. Clarkson, N. V. Wilmont, and S. R. A. Kelly,
eds.). University of Kansas Press, Lawrence, Kansas.

INSTRUCTIONS FOR AUTHORS

ANNALS OF CARNEGIE MUSEUM consist of
contributions to the earth sciences (including paleontol-
ogy), life sciences, and anthropology, in 30 by 46 picas
format (127 by 195 mm or 5 by 7% inches). Submit all
manuscripts to the Office of Scientific Publications. Au-
thors should give particular attention to scientific con-
tent, format, and general style for the ANNALS. Manu-
scripts that do not conform to the style of the AN-
NALS will be returned to the author immediately.
Every manuscript will be subjected to external peer re-
view. Authors should submit the names, addresses,
phone and fax numbers, and e-mail addresses of at least
four qualified potential reviewers for each manuscript.
Authors will be asked to subsidize, if funds are available,
any or all costs of publication (approximately $ 100/page
printed). A detailed set of instructions to authors may be
requested from the Office of Scientific Publications.

Manuscript Style. — Articles should include the fol-
lowing items in this order: title page, abstract, text (with
desired headings), acknowledgments, literature cited, ta-
bles, figure captions, and copies of illustrations. All
manuscripts must be typed double-spaced on standard
8Vi- by 1 1-inch white bond paper, with at least one-inch
margins all around, and submitted in quadruplicate — an
original for the editors, and three review copies. All
pages should be numbered, including tables, literature
cited and the list of figure captions. Only correspon-
dence quality or better dot matrix printouts will be ac-
cepted; draff mode is unacceptable. A monospaced font,
(e.g.. Courier) is required. A disk in WordPerfect x.x or
ASCII is helpful.

Title Page. The title should be brief, include the ani-
mal or plant group involved be placed on a separate
page, and appear two to three inches below the top mar-
gin. Include the author’s name(s) and the affiliations of
non-CMNH authors. In the case of multiple authorship,
indicate the address to which proofs should be sent.

Abstracts. Abstracts should be short, but substantive,
and included at the head of the first page of text.

Text. Do not right justify text or break (hyphenate) a
word at the end of a line. Footnotes and acknowledg-
ments as footnotes are unacceptable. All text is double-
spaced in a monospaced font (e.g.. Courier).

Tables and Figure Legends. The list of figure legends
and each table should be typed consecutively on indi-
vidual pages separate from the text. Tables must be dou-
ble-spaced throughout with no vertical lines.

Measurements. Metric units should be used, if possi-
ble.

Literature Cited. All references to literature in text
must appear in the Literature Cited section. The data
(author, date, and page) in both citations must agree. Do
not abbreviate the titles of periodicals or serials. The
following style, double-spaced, should be used in Lit-
erature Cited:

1) Two authors in a journal series:

Soltis, D. E., and P. S. Soltis. 1992. The distri-
bution of selling rates in homosporous ferns.
American Journal of Botany, 79:97-100.

2) Same authors repeated — use three-em dash:

. 1923. The fauna of the Ardyn Obo For-
mation. American Museum of Natural History
Novitates, 98:1-5.

3) Same authors plus a third author — repeat all authors:

Knutson, L. V, R. E. Orth, and W. L. Murphy.
1986. Catalog of Sciomyzidae (Diptera) of
America north of Mexico. Entomography, 4:
1-53.

4) Chapter in an edited volume:

Rausch, R. L. 1963. A review of the distribution
of Holarctic mammals. Pp. 29-43, in Pacific
Basin Biogeography (J. L. Gressitt, ed.).
Bishop Museum Press, Honolulu, Hawaii.

5) Unpublished dissertation:

Smith, J. P. 1976. Review of Eocene Mammals.
Unpublished Ph.D. Dissert., University of
California, Berkeley, California.

6) Book:

White, M. J. D. 1961. The Chromosomes. Meth-
euen and Co., Ltd., London, United Kingdom.

7) Journal articles with usual volume and issue number:

Anderson, W. I. 1969. Lower Mississippian con-
odonts from northern Iowa. Journal of Pale-
ontology, 43:916-928.

Figures. Including all illustrative materials (line art,
halftones, photographs), figures are to be numbered in
Arabic numerals. Four sets of figures are required, one
(original artwork) for reproduction, three for reviewers.
Photocopies of photographs for reviewers are usually not
acceptable but are adequate for line drawing review cop-
ies. Figures may not be larger than 17 by 12 inches.
Reducing figures is the responsibility of the author. All
figures must be reducible to a maximum of 127 by 195
mm (30 by 46 picas) without loss of clarity. Line copy
should be designed for reduction to % or Vi or actual size.
Typewritten figure copy will not be accepted. Pho-
tographic figures should be submitted at actual repro-
duction size, if possible.

Rectangular halftone figures should be abutted, with-
out intervening spaces. The printer will insert narrow
white spaces during the reproduction process. All figures
must have minimally one-inch borders all around. Each
figure should be given a protective cover and identified
on the back side.

Lettering and/or a magnification scale (linear metric
scale) for rectangular halftone figures should be placed
directly on the photo, not in a blank space between pho-
tos. The scale or lettering for closely cropped photos can
be placed in blank areas close to the figure.

Proof. — The author should answer all queried proof
marks and check the entire proof copy. Return corrected
page proof with the edited manuscript promptly to the
Office of Scientific Publications.

If an author chooses to make extensive alterations to
a paper in proof stage, the author will bear the cost.
Original manuscripts will not be returned unless request-
ed. Figures will be returned to the author only if re-
quested prior to publication.

■

ISSN 0097-4463

AS
3CP
Vlli
(\l M

ANNALS

0/ CARNEGIE MUSEUM

THE CARNEGIE MUSEUM OF NATURAL HISTORY

4400 FORBES AVENUE • PITTSBURGH, PENNSYLVANIA 15213

9 AUGUST 2000

VOLUME 69

NUMBER 3

CONTENTS

ARTICLES

Reassessment of the North American pelobatid anuran Eopelobates guthriei

Amy C. Henrici 145

Homology and phylogenetic implications of some enigmatic cranial features

in galliform and anseriform birds

Richard L. Zusi and Bradley C. Livezey 157

A new species of Carpocristes (Mammalia: Primatomorpha) from the Middle
Tiffanian of the Bison Basin, Wyoming, with notes on carpolestid phy-
togeny K. Christopher Beard 195

5^^\THS0A///^y^

Editors, ANNALS, BULLETIN and SPECIAL PUBLICATIONS:

D. S. Berman, Ph.D.

D. R. Watters, Ph.D.

J. R. Wible, Ph.D.

Orr Goehring, Managing Editor

Manuscripts, subscriptions, orders for individual numbers, and changes of address
should be sent to:

Office of Scientific Publications
Carnegie Museum of Natural History
4400 Forbes Avenue
Pittsburgh, PA 15213-4080
Phone (412) 622-3287
Fax (412) 622-8837
E-mail: scipubs@clpgh.org

ANNALS OF CARNEGIE MUSEUM is published quarterly by The Carnegie Museum of Natural
History, 4400 Forbes Avenue, Pittsburgh, Pennsylvania 15213-4080, by the authority of the Board of
Trustees of Carnegie Institute.

© 2000 Carnegie Institute.

@ This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).

ANNALS OF CARNEGIE MUSEUM
VoL. 69, Number 3, Pp. 145-156

9 August 2000

REASSESSMENT OF THE NORTH AMERICAN PELOBATID ANURAN

EOPELOBATES GUTHRIEI

Amy C, Henrici

Scientific Preparator, Section of Vertebrate Paleontology

Abstract

Eopelobates guthriei Estes, 1970, is based on a partial skull and associated right scapula from the
Early Eocene Wind River Formation (Lysitean), Fremont County, Wyoming. Reexamination of the
holotype and only known specimen reveals that it should no longer be regarded as Eopelobates because
it lacks characters considered to be diagnostic of that genus. Comparison to other pelobatids indicates
it is most similar to Scaphiopus and Spea in possession of an elongate postchoanal ramus of the vomer.
It compares more closely with Scaphiopus in its lack of hypossification of cranial bones and possession
of a long, low, arcuate ventral flange of the pterygoid. There is no evidence to suggest that it represents
a new genus, but because it is not known if the postcranial skeleton was specialized for burrowing,
as in Scaphiopus and some other pelobatids, it is only tentatively referred to Scaphiopus as cf. S.
guthriei (Estes, 1970). Two derived characters distinguish cf. Scaphiopus guthriei from other Sca-
phiopus: 1) frontoparietal narrowest just posterior of the supraorbital flange and 2) otic ramus of
squamosal long and thin. Assuming assignment to Scaphiopus is correct, then the temporal range for
Scaphiopus can be extended back from the Middle Oligocene to the Lower Eocene.

Key Words: Anura, Pelobatidae, Eopelobates, Scaphiopus, Lower Eocene

Introduction

In 1929 Parker named and described Eopelobates anthracinus as the generic
holotype for Eopelobates on the basis of a single specimen that was collected
from strata now considered to be uppermost Oligocene in age (von Koenigswald
et aL, 1992) from Rott, near Bonn, Germany. Since Parker’s (1929) description
of E. anthracinus, seven other species from North America, Europe, and Asia
have been included in this genus. However, the generic assignment of most of
these taxa has been either changed or questioned. Eopelobates bayeri, from the
Oligo^Miocene of the Czech Republic, is regarded as either very closely related
to E. anthracinus, differing only in its larger size (Spinar and Rocek, 1984; Rocek,
1995), or synonymous with E. anthracinus (SancMz, 1998). Eopelobates lepto-
colaptus and E. sosedkoi from the Upper Cretaceous of Mongolia and Uzbekistan,
respectively, have been reassigned to the gobiatid Gobiates (Spinar and Tatarinov,
1986; Rocek and Nessov, 1993). Rocek (1981), in a monographic study of Pe-
lobates fuscus, observed that both Pelobates and Eopelobates have a frontoparietal
derived from three ossifications, the typical paired elements with the addition of
a median element situated posterior to them. The three ossifications are easily
identified in tadpoles, but in adults the only indication of the presence of the
posteromedial element is that it prevents the median suture from reaching the
posterior end of the frontoparietal complex. On the basis of this character, Rocek
(1981) suggested that only two species, E. anthracinus and E. bayeri, should be
retained in the genus Eopelobates. Published information about the frontoparietal
of E. hinschei from the Middle Eocene of Geiseltal, Germany (Kuhn, 1941; Estes,

Submitted 28 June 1999

145

146

Annals of Carnegie Museum

VOL. 69

1970), led Rocek (1981) to suspect that the posteromedial element is absent in
this taxon, and thus it should not be regarded as Eopelobates. In addition, Sanchiz
(1998) maintained that although the species is valid, it lacks a diagnosis and is
in need of restudy. More recently, Wuttke (1988) has reidentified Propelodytes
wagneri, from the Middle Eocene of Messel, Germany, as Eopelobates wagneri.

Two species of Eopelobates have been described from North America: E. guth-
riei from the Lower Eocene Wind River Formation of Wyoming (Estes, 1970)
and E. grandis from the Lower Oligocene Chadron Formation of North Dakota
(Zweifel, 1956). Rocek (1981) concluded that both species are not Eopelobates
but rather may be more closely related to the North American spadefoots Sca-
phiopus and Spea. This conclusion was based on his determination from published
descriptions and illustrations that the frontoparietals of E. guthriei and E. grandis
also lack the posteromedial element, the quadratojugal is absent in E. guthriei and
possibly misidentified in E. grandis, and the columella is present in E. guthriei.
These are characters which all occur in Scaphiopus and Spea but not Eopelobates
(Rocek, 1981).

The purpose of this study is to provide a revision of E. guthriei that considers
the more recent published information about this genus and pelobatids in general.
Minor preparation of the holotype and only known specimen has also revealed a
feature important to its taxonomic assignment. A redescription and reassessment
of E. grandis is being prepared separately.

The correct identification of North American Eopelobates is important for un-
derstanding pelobatid evolutionary history and paleobiogeography. Eopelobates
has been thought to be a primitive pelobatid that is ancestral to the spadefoots
(Parker, 1929; Estes, 1970; Spinar, 1972; Savage, 1973), and Estes (1970) spec-
ulated that E. guthriei was close to the origin of spadefoots. The timing and place
of divergence of spadefoots has changed in accordance with new additions to the
fossil record, but, in general, the Eocene-earliest Oligocene was considered to be
an important time for spadefoot evolution and dispersal (Savage, 1973; Estes, in
Sage et al., 1982). More recently. Sage et al. (1982) have suggested that the
modern spadefoots (Pelobates, Scaphiopus, and Spea) diverged during the Cre-
taceous, and the spadefoot morphotype then subsequently underwent little change.
Their divergence time was based on immunological evidence.

Abbreviations

Anatomical. — a, angular; c, choana; co, columella; cp, crista parotica; fo, fe-
nestra ovalis; fp, frontoparietal; ipfm, impression of pars facialis of maxilla; jf,
jugular foramen; m, maxilla; md, mandible; n, nasal; op, otic plate of squamosal;
pa, parasphenoid; pcv, postchoanal ramus of vomer; pf, prootic foramen; pfm,
pars facialis of maxilla; pm, premaxilla; ppm, palatine process of maxilla; pt,
pterygoid; q, quadrate; s, sphenethmoid; sc, scapula; sq, squamosal; v, vomer; vf,
ventral flange of pterygoid.

Institutional. — CM, Carnegie Museum of Natural History, Pittsburgh, Penn-
sylvania; FMNH, Field Museum of Natural History, Chicago, Illinois; MCZ, Mu-
seum of Comparative Zoology, Harvard University, Cambridge, Massachusetts.

Description and Comparison op ""Eopelobates"" guthriei

The holotype and only known specimen of ""Eopelobates"" guthriei (MCZ
3493) consists of a moderate-sized, incomplete skull and associated incomplete

2000

Henrici — Reassessment of Eopelobates guthriei

147

Fig. 1. — Photographs of holotype of cf. Scaphiopus guthriei, MCZ 3493. A, dorsal view; B, ventral

view; C, right lateral view; and D, occipital view. Scale bar = 5 mm.

right scapula (Fig. 1, 2) collected from the Lower Eocene Lysite Member of the
Wind River Formation in the Wind River Basin, Fremont County, Wyoming. The
skull is missing bones of the snout, left temporal region, and most of the lower
jaws; the preserved bones are in articulation or are very closely associated. Dis-
tortion of the right temporal region has resulted in anterior rotation of the otic
plate of the squamosal away from the dorsal surface of the right exoccipital-
prootic complex. The ventral ramus of the squamosal is missing its base and is
pushed inward. Also, the medial ramus of the right pterygoid is not preserved in
articulation with the exoccipital-prootic complex.

Estes (1970) argued that the flattened, medially concave dorsal skull roof was
not an artifact of preservation but was natural and similar to the skull of other
Eopelobates. Evidence that the skull was dorsoventrally compressed does exist,
however. The left maxilla is not preserved in near vertical orientation but, rather,
slopes outward. In conjunction with this, the lateral process of the nasal, which
bears evidence of crushing by the numerous cracks running through it, is flattened
and oriented horizontally instead of a more vertical orientation. The anterior por-
tion of the left frontoparietal is more depressed than the right, and the dorsal edge
of the right squamosal is preserved at the same level as the right frontoparietal
rather than at a more ventral level. Additionally, the long axis of the occipital
condyles is oriented horizontally instead of having the typical, for pelobatids,
transverse orientation.

The undistorted skull roof of MCZ 3493 probably resembled that of Scaphiopus
skinneri from the Middle Oligocene of North Dakota (Estes, 1970) in being flat
and sloping slightly downward anteriorly. It should be noted that currently known
specimens of Eopelobates are all preserved as flattened skeletons, so the outline
of their dorsal skull roof cannot be accurately determined.

148

Annals of Carnegie Museum

VOL. 69

Fig. 2. — Diagrammatic drawings of holotype of cf. Scaphiopus guthriei, MCZ 3493. A, dorsal view;
B, ventral view; C, right lateral view; and D, occipital view. Scale bar = 5 mm.

Dermal ornamentation occurs on the frontoparietals, nasals, maxillae, and squa-
mosal. Although the dermal ornamentation is somewhat eroded and covered by
matrix in places, it can be discerned that it consists of a system of grooves and
ridges bearing tubercles, as well as being slightly reticulated in places. This or-
namentation pattern most closely resembles that occurring in Scaphiopus hoi-
brooki. Spinar and Rocek (1984) used as a diagnostic character of Eopelobates
the presence of a posteromedial element in the frontoparietal, resulting in a fron-
toparietal complex that is derived from the fusion of three ossifications. A fron-
toparietal complex derived from three ossifications also occurs in Pelobates (Ro-
cek, 1981). In contrast, the frontoparietals of MCZ 3493 are paired (Estes, 1970);
the median suture clearly extends the entire length between the two halves, in-
dicating that a posteromedial element is absent. The subrectangular frontoparietals
bear supraorbital flanges (otic processes in Estes, 1970) that reach their greatest
width at approximately one-third their length from the anterior end. The fronto-
parietal is waisted and narrowest just posterior to the supraorbital flange. At the
posterolateral comer of each frontoparietal the posterolaterally oriented posterior
tip caps a small boss on the dorsal surface of the exoccipital-prootic complex.
This boss was referred to as the paroccipital process by Estes (1970), which is
misleading because the paroccipital process is part of the opisthotic, a bone that
is absent in anurans. Anteriorly, the right frontoparietal seems to be complete, but
the left is not. The posterior edges of the nasals are irregular, which suggests that
some bone is missing, although Estes (1970:fig. 13B) illustrates them as being
complete. The exposed dorsal surface of the sphenethmoid bears faint impressions
of what probably were the posterior and medial edges of the nasals. These im-
pressions suggest that the complete nasals allowed considerably less dorsal ex-
posure of the sphenethmoid than that illustrated by Estes (1970:fig. 13B).

2000

Henrici — Reassessment of Eopelobates guthriei

149

Fig. 3. — Comparison of some pelobatid nasal bones. A, cf. Scaphiopus guthriei, holotype, MCZ 3493;
B, Scaphiopus holbrooki, CM 32300; C, Spea bombifrons, CM 48932; D, Eopelobates bayeri (from
Spinar, 1972). Scale bars = 2 mm. No scale for D.

Flattening of the left nasal has distorted its shape somewhat. Its short lateral
process is preserved in articulation with the maxilla. Bone making up the some-
what concave anterolateral edge of the nasal is rounded, indicating that the shape
of this edge is not the result of breakage. It is apparent that the anterolateral
margin of the nasal was not straight but was probably concave as in Scaphiopus,
Spea, and Pelobates (Fig. 3). A straight anterolateral margin coupled with a long
and slender lateral process is considered a diagnostic character for Eopelobates
(Spinar and Rocek, 1984). All that remains of the right nasal is its posteromedial
corner.

Only the posterior end of the right maxilla is preserved, and its posterior pro-
cess is broken off. The left maxilla is missing its anterior and posterior ends.
Small, bicuspid, pedicellate teeth are preserved along the pars dentalis of the left
maxilla, and a few tooth bases are preserved on the right. Bicuspid, pedicellate
teeth occur in Eopelobates (Wuttke, 1988), as well as in other pelobatids.

The right squamosal is nearly complete, missing only the base of its ventral
ramus, and the left squamosal is absent. The zygomatic ramus is deep proximally,
bearing a concavity along its ventral margin, and tapers distally where it articu-
lates with the zygomatic process of the maxilla. The otic ramus becomes deeper
distally, although it is not as deep as and is only slightly shorter than the zygo-
matic ramus. It bears a longer and deeper concavity on the ventral margin than
that occurring on the zygomatic ramus. The otic plate, which has rotated anteriorly
from its articulation with the crista parotica, extends slightly beyond the midpoint
between the frontoparietal and lateral edge of the skull and does not contact the
frontoparietal, as occurs in most megophryines. A dorsal process, which in all
Pelobates species except P. fuscus articulates with the superior lateral process of
the frontoparietal (Rocek, 1981), is absent. In P. fuscus a ligament bridges the
squamosal and frontoparietal (Rocek, 1981). The quadrate forms a wedge between
the pterygoid and ventral ramus of the squamosal. A quadratojugal is not pre-
served. As noted by Estes (1970), it was either absent or was lost during dis-

150

Annals of Carnegie Museum

VOL. 69

A

B

Fig. 4. — Comparison of vomer region of palate. A, cf. Scaphiopus guthriei, holotype, MCZ 3493; B,
Scaphiopus holbrooki, CM 18719. Note that dorsoventral compression of skull figured in A has caused
lateral displacement of part of the postchoanal ramus of the vomer and jaw elements. Scale bars = 5 mm.

placement of the posteroventral comer of the maxilla and ventral ends of the
squamosal and quadrate. Among pelobatids only Scaphiopus and Spea lack a
quadratojugaL

The sphenethmoid (ethmoid in Estes, 1970) is incompletely preserved, lacking
the right lateral process, most of the anterior process, and portions of the ventral
and anterior surfaces of the left lateral process. All that remains of the right vomer
is the proximal portion of the postchoanal ramus and bone forming the postero-
medial border of the internal nares. The somewhat more complete left vomer (Fig.
4A) has a raised area medial to the postchoanal ramus that Estes (1970) suggested
was the vomerine tooth plate. The lateral end of the postchoanal ramus ends in
a break, and lateral to this is a thin, fragmented rod of bone identified as a
probable palatine by Estes (1970). It should be noted that discrete palatines (neo-
palatine of Tmeb, 1993) do not occur in any known pelobatid (Caneatella, 1985).
Various authors have argued that in pelobatids each palatine fuses with either the
postchoanal ramus of the vomer (Rocek, 1981) or the maxilla (Kluge, 1966;
Zweifel, 1956). A study on the development of Spea bombifrons by Wiens (1989)
revealed that the elongate postchoanal ramus of the vomer arises from either the
vomer ossification or an independent ossification. Wiens (1989) also observed
that the palatine process of the pars facialis of the maxilla does not represent a

2000

Henrici — Reassessment of Eopelobates guthriei

151

palatine fused with the maxilla, but rather is derived from the maxillary ossifi-
cation. In MCZ 3493 the broken medial end of the “palatine” is preserved in
line with the broken lateral end of the postchoanal ramus, and the gap that sep-
arates them was filled with glue. Removal of matrix along the anterior edge of
the lateral portion of the “palatine” reveals that it is continuous with the pars
facialis of the maxilla, indicating that it is the palatine process of the maxilla
rather than a discrete palatine. An impression posterior and lateral to the palatine
process indicates that a wedge of the pars facialis is missing. The remaining
portion of the “palatine” consists of a thin bone that is sutured to the pars facialis
of the maxilla. This piece of bone is assumed to be the rest of an elongate post-
choanal ramus of the vomer for several reasons: it is preserved sutured to the
pars facialis of the maxilla; its broken end is close in size to the broken, lateral
end of the postchoanal ramus of the vomer; and it is preserved in alignment with
and separated only by a small gap from the rest of the postchoanal ramus. Pre-
sumably, this separation occurred when the skull was flattened. The only pelobatid
genera possessing an elongate postchoanal ramus of the vomer that articulates
with the palatine process of the maxilla are Scaphiopus (Fig. 4B) and Spea (Can-
natella, 1985). The anterior portion of both vomers is missing.

The three rami of the pterygoid are broad proximally and taper distally. As in
Scaphiopus, a long, low, arcuate ventral flange runs along the lateral edge of the
pterygoid, extending from the base of the anterior ramus to near the tip of the
posterior ramus (Fig. 5). The anterior ramus of the pterygoid articulates with the
medial edges of the zygomatic ramus of the squamosal and maxilla. The dorsal
margin of the distal end of the medial ramus is broken off. Matrix and the over-
lying right lateral ala of the parasphenoid obscure the ventral portion of the distal
end of the medial ramus.

The prootic foramen is widely emarginate as in other pelobatids, except Spea,
where it is completely surrounded by bone. The occipital condyles appear to be
widely separated, but this is probably the result of dorsoventral compression and
bone loss. The right occipital condyle is kidney shaped, whereas the left has a
more circular outline due to bone loss along the medial margin of the condyle.
The shape and position of the right occipital condyle indicates that the condyles
were probably narrowly separated, which is typical for pelobatids (Lynch, 1971).
A columella is present, as noted by Estes (1970).

A crushed, incomplete right scapula, mistakenly identified as a left by Estes
(1970), is exposed in medial aspect and rests against the anterior edge of the right
lateral ala of the parasphenoid. Most of the anterior half of the scapula is pre-
served, but missing are all of the pars glenoidalis except for its base, the distal-
most edge of the scapular blade, and the posterodistal corner of the scapula.
Enough of the scapula is preserved to determine that it is long. As in other
pelobatids, the scapula bears a large and bulbous pars acromialis, which indicates
that the clavicle articulated with its ventral edge rather than its anterior edge.
When the clavicle articulates with the anterior edge of the scapula, the pars ac-
romialis narrows considerably distally. Although only the base of the pars gle-
noidalis is preserved, it was obviously a distinct process that was separated from
the pars acromialis by a cleft. The anterior edge of the scapula is strongly concave
and lacks an anterior lamina (Estes, 1970), as in Scaphiopus and Spea. Presence
of an anterior lamina was demonstrated by Henrici (1994) to be a synapomorphy
of Eopelobates, Macropelobates, and Pelobates.

152

Annals of Carnegie Museum

VOL. 69

Fig. 5. — Comparison of some pelobatid pterygoid bones. A, cf. Scaphiopus guthriei, holotype, MCZ
3493; B, Scaphiopus holbrooki, CM 18719; C, Spea bomhifrons, CM 48932; D, Pelobates cultripes,
CM 55769; E, Leptobrachiwn hasselti, FMNH 131998. Anterior is to the right and medial to the
bottom of the page. Scale bar = 5 mm.

Discussion

Estes (1970:313) assigned MCZ 3493 to Eopelobates based on its possession
of “ a concave skull roof, approximately subequal orbital and temporal open-
ings, as well as the distinctive shape of the squamosal and ethmoid.” As noted
in the description, dorsoventral compression of the skull of MCZ 3493 most likely
caused the skull roof to appear concave. Also, as previously mentioned, all known
specimens of Eopelobates are preserved as flattened skeletons, and the configu-
ration of their dorsal skull roof cannot be accurately determined. The other char-
acters used by Estes (1970) are either not preserved or are not diagnostic at the
generic level. The sphenethmoid of MCZ 3493 is incomplete anteriorly, making
comparison to other pelobatid sphenethmoids impossible. Estes (1970) observed
that the orbital and temporal openings are of subequal size in Megophrys and
Eopelobates, whereas in pelobatines the orbit is enlarged and the temporal area
reduced. However, this character is difficult to interpret and varies intergenerically.
The shape of the squamosal is diagnostic at the species level for at least Pelobates
and Scaphiopus and thus is not a useful character for differentiating pelobatid
genera.

Four of the characters currently considered as diagnostic for Eopelobates (Spi-
nar and Rocek, 1984; Sanchiz, 1998) can be analyzed in MCZ 3493. Polarity of
these characters is based on the phylogenetic analysis of pelobatoids in Henrici
(1994) or outgroup comparison (using methodology of Wiley et al., 1991) in

2000

Henrici — Reassessment oe Eopelobates guthriei

153

which pelodytid and pipoid anurans comprise the outgroup. The characters are as
follows: 1) frontoparietal is derived from three ossifications (derived state); 2)
frontoparietal does not contact squamosal (primitive state); 3) nasal has a straight
anterolateral margin and a long, slender lateral process (derived state); and 4)
dermal sculpturing is pitted and lacks tubercles (state indeterminate). The first
character also occurs in Pelobates, the second occurs in Macropelobates, Sca-
phiopus, and Spea, and the latter two characters appear to be unique. The first of
these characters, incorporation of a posteromedial element into the frontoparietal
complex, warrants further discussion, because it has had some bearing on theories
of anuran phytogeny. Rocek (1981) homologized the posteromedial element with
the median extrascapular of Eusthenopteron, considered it to be a primitive char=
acter, and used it to form the basis of a phytogeny that placed Eopelobates and
Pelobates outside of Salientia (sensu Sanchiz, 1998). Milner (1988) pointed out
that the occurrence of an extra ossification wedged between the parietals and
postparietals, which he identified as the interparietal, is not unprecedented and
observed its occurrence in several temnospondyls. He (Milner, 1988:63) suggested
that the interparietal “ ... is a recurrently occurring derived condition” and fur-
ther stated that the posteromedial element in Eopelobates and Pelobates is either
a neomorph or a character reversal and, thus, is derived and most likely defines
a subclade within Pelobatidae. The phylogeny of Pelobatoidea proposed by Hen-
rici (1994) supports Milner’s theory that the occurrence of the posteromedial
element in the frontoparietal complex of Eopelobates and Pelobates is derived
and represents a synapomorphy of a subclade within Pelobatidae.

Of the four diagnostic characters, MCZ 3493 is similar to Eopelobates and
other pelobatines except Pelobates only in the primitive character of lack of con-
tact between frontoparietal and squamosal. It differs from Eopelobates in three
of these characters: 1) frontoparietal is paired (primitive state); 2) anterolateral
margin of nasal is not straight but probably was concave, and the lateral process
is short (Fig. 3; primitive state); and 3) dermal sculpturing consists of ridges and
grooves, which are arranged in a slightly reticulated pattern in places, and tuber-
cles (state indeterminate). Another distinction from Eopelobates, as well as Ma-
cropelobates and Pelobates, is the absence of an anterior lamina of the scapula.
Presence of an anterior lamina was determined by Henrici (1994) to be a syna-
pomorphy of Eopelobates, Macropelobates, and Pelobates. Based on these dif-
ferences, it is apparent MCZ 3493 should not be regarded as Eopelobates.

Comparison of MCZ 3493 to other pelobatids reveals that it most closely re-
sembles Scaphiopus and Spea in possession of an elongate postchoanal ramus of
the vomer that articulates with the palatine process of the maxilla (Fig. 4). Among
pelobatids this character is unique for Scaphiopus and Spea and has been regarded
as one of several synapomorphies uniting them (Cannatella, 1985; Henrici, 1994;
Maglia, 1998). It should be mentioned that besides pelobatids, the postchoanal
ramus of the vomer is known to articulate with the palatine process of the maxilla
in at least some species of Discoglossus: D. sardus (Piigener and Maglia, 1997)
and D. pictus (pers. obs.). However, in these species of Discoglossus the post-
choanal ramus is not elongate, in contrast to that of Scaphiopus, Spea, and MCZ
3493.

The skull of MCZ 3493 compares more closely with Scaphiopus than with
Spea. Spea exhibits cranial hypossification that is evidenced by the loss of dermal
ornamentation, reduced ossification of the frontoparietals and nasals which allows
dorsal exposure of the frontoparietal fontanelle and sphenethmoid, and reduction

154

Annals of Carnegie Museum

VOL. 69

of the otic plate and the zygomatic and otic rami of the squamosal with consequent
loss of contact between the squamosal and maxilla (Kluge, 1966; Wiens, 1989;
Maglia, 1998). This hypossification of the skull does not occur in MCZ 3493 or
other pelobatids and is thought to be paedomorphic in Spea (Wiens, 1989). Sea-
phiopus and MCZ 3493 possess a similarly shaped ventral flange of the pterygoid
that is long, low, and arcuate, and differs from the low, straight ventral flange in
Spea, and the prominent, short ventral flange in Pelobates and Eopelobates (Fig.
5). Leptobrachium and presumably other megophryines lack a ventral flange.
Maglia (1998), in a cladistic analysis of extant pelobatoids, considered the pres-
ence of a ventral flange to be derived and its absence to be primitive. It stands
to reason, then, that the different shapes of the ventral flange in pelobatids rep-
resent independently derived states.

Because there is no evidence suggesting that MCZ 3493 represents a new
genus, it is tentatively referred to Scaphiopus on the basis of the following char-
acters: 1) presence of an elongate postchoanal ramus of the vomer that articulates
with the palatine process of the maxilla, 2) lack of hypossification of cranial
bones, and 3) presence of a long, low, arcuate ventral flange of the pterygoid. It
is acknowledged, however, that the first character is based on somewhat frag-
mentary evidence. This, together with the lack of a postcranial skeleton for MCZ
3493, which in Scaphiopus and some other pelobatids is specialized for burrow-
ing, is the reason for only tentatively referring MCZ 3493 to Scaphiopus. Con-
fident generic assignment of MCZ 3493 must await discovery of more complete
specimens to determine if it has a postcranial skeleton similar to Scaphiopus.
Assuming assignment to Scaphiopus is correct, then two unique characters dis-
tinguish it from other species of this genus. These represent two of the three
characters used by Estes (1970) in his diagnosis of MCZ 3493 as a new species
of Eopelobates, but are described here using different terminology: 1) frontopa-
rietal is waisted and narrowest just posterior of the supraorbital flanges and 2)
otic ramus of squamosal is relatively thin and almost as long as the zygomatic
ramus. These characters are unique among pelobatids and, based on comparison
to pelodytids and pipoids, are judged to represent the derived state in cf. S. guth-
riei.

Several assumptions can be made about pelobatid evolution and paleobiogeog-
raphy if assignment of MCZ 3493 to Scaphiopus is correct. The temporal range
of Scaphiopus can now be extended back from the Middle Oligocene to the Early
Eocene. The currently known first occurrences of extant and extinct spadefoots
indicates that their dispersal across Laurasia must have occurred very early in the
Paleogene, if not earlier. In addition to the Early Eocene record of Scaphiopus,
the oldest-known record of Pelobates is P. decheni from the Late Eocene of
Belgium (Bohme et ah, 1982). Additionally, the oldest known record for the
extinct spadefoot, Macropelobates, is M. osborni from the Early Oligocene of
Mongolia (Noble, 1924). Unfortunately, the fossil record of spadefoots is too
sparse to allow speculation on their center of origin or dispersal routes. Finally,
the presence of extant spadefoots in both North America and Europe by the Late
Eocene adds support to the theory of Sage et al. (1982) that they are an ancient
and morphologically stable group that most likely diverged during the Cretaceous.

Acknowledgments

M. Sander and Z. Rocek are thanked for their generous hospitality during my visit to their respective
institutions. I am grateful to Charles J. Kubit who cheerfully and patiently printed the photographs in

2000

Henrici — Reassessment of Eopelobates guthriei

155

Figure 1. A. R. Milner and Z. Rocek provided helpful discussion about pelobatids. The following

institutions loaned specimens used in this study: Field Museum of Natural History, Harvard Museum

of Comparative Zoology, and the United States National Museum, Washington D. C. This project was

supported in part by funds from the M. Graham Netting Research Fund.

Literature Cited

Bohme, W., Z. Rocek, and Z. Spinar. 1982. On Pelobates decheni, 1861, and Zaphrissa eurypelis
Cope, 1866 (Amphibia: Salientia: Pelobatidae) from the Early Miocene of Rott near Bonn, West
Germany. Journal of Vertebrate Paleontology, 2:1-7.

Cannatella, D. C. 1985. A phylogeny of primitive anurans (archaeobatrachians). Unpublished Ph.D.
Dissert., University of Kansas, Lawrence, Kansas.

Estes, R. 1970. New fossil pelobatid frogs and a review of the genus Eopelobates. Bulletin Museum
Comparative Zoology, 139:293-339.

Henrici, A. C. 1994. Tephrodytes brassicarvalis, a new genus and species (Anura: Pelobatidae), from
the Arikareean Cabbage Patch Beds of Montana, USA, and pelodytid-pelobatid relationships.
Annals of Carnegie Museum, 63:155-183.

Kluge, A. J. 1966. A new pelobatine frog from the Lower Miocene of South Dakota with a discussion
of the evolution of the Scaphiopus-Spea complex. Contributions in Science, Los Angeles County
Museum, 113:1-26.

Koenigswald, W von, T. Martin, T. Mors, and H. U. Pfretzschner. 1992. Die oberoligozane
Wirbeltierfauna von Rott bei Hennef am Siebengebirge. Synonymien und Literatur 1828-1991.
Decheniana, 145:312-340.

Kuhn, O. 1941. Die Eozanen Anura aus dem Geiseltale nebst einer iibersicht iiber die fossilen Gat-
tungen. Nova Acta Leopoldina, 10:345-376.

Lynch, J. D. 1971. Evolutionary relationships, osteology, and zoogeography of leptodactyloid frogs.
Miscellaneous Publications, University of Kansas Museum of Natural History, 53:1-238.

Maglia, a. M. 1998. Phylogenetic relationships of extant pelobatoid frogs (Anura: Pelobatoidea):
evidence from adult morphology. Scientific Papers, University of Kansas Museum of Natural
History, 10:1-19.

Milner, A. R. 1988. The relationships and origins of living amphibians. Pp. 132-182, in Evolutionary
Biology and Classification of the Tetrapods, Volume 1, (M. J. Benton ed.). Systematics Associa-
tion Special Volume 35A, Clarendon Press, Oxford, United Kingdom.

Noble, G. K. 1924. A new spadefoot toad from the Oligocene of Mongolia with a summary of the
evolution of the Pelobatidae. American Museum Novitates, 132:1-23.

Parker, H. W. 1929. Two fossil frogs from the Lower Miocene of Europe. Annals and Magazine of
Natural History, 4:270-281.

PuGENER, L., AND Maglia, A. M. 1997. Osteology and skeletal development of Discoglossus sardus
(Anura: Discoglossidae). Journal of Morphology, 233:267-286.

Rocek, Z. 1981. Cranial anatomy of frogs of the family Pelobatidae Stannius, 1856, with outlines of
their phylogeny and systematics. Acta Universitatis Carolinae-Biologica, 1980(1-2): 1-164.

— . 1995. A new specimen of Eopelobates (Anura: Pelobatidae) from the Tertiary near Bonn

(Germany) and the problem of E. anthracinus-E. bayeri relations. Palaontologische Zeitschrift,
69:283-287.

Rocek, Z., and L. A. Nessov. 1993. Cretaceous anurans from central Asia. Paleontographica Abtei-
lung A, 226:1-54.

Sage, R. D., E. M. Prager, and D. B. Wake. 1982. A Cretaceous divergence time between pelobatid
frogs {Pelobates and Scaphiopus): Immunological studies of serum albumin. Journal of Zoology,
London, 198:481-494.

Sanchiz, B. 1998. Encyclopedia of Paleoherpetology, Part 4, Salientia, Verlag Dr. Friedrich Pfeil,
Miinchen, Germany.

Savage, J. M. 1973. The geographic distribution of frogs: patterns and predictions. Pp. 351-445, in
Evolutionary Biology of the Anurans (J. L. Vial ed.). University of Missouri Press, Missouri.

Spinar, Z. V. 1972. Tertiary Progs from Central Europe. W. Junk, The Hague, Netherlands.

Spinar, Z. V., and Z. Rocek. 1984. The discovery of the impression of the ventral side of Eopelobates
anthracinus Parker, 1929 holotype. Amphibia-Reptilia, 5:87-95.

Spinar, Z. V, and L. P. Tatarinov. 1986. A new genus and species of discoglossid frog from the
Upper Cretaceous of the Gobi Desert. 1986. Journal of Vertebrate Paleontology, 6:113-122.

Trueb, L. 1993. Patterns of cranial diversity among Lissamphibia. Pp. 255-343, in The Skull. Volume
2. Patterns of Structural and Systematic Diversity (J. Hanken and B. K. Hall eds.). The University
of Chicago Press, Chicago, Illinois.

156

Annals of Carnegie Museum

VOL. 69

Wiens, J. J. 1989. Ontogeny of the skeleton of Spea bombifrons (Anura: Pelobatidae). Journal of
Morphology, 202:29-51.

Wiley, E. O., D. Siegel-Causey, D. R. Brooks, and V. A. Funk. 1991. The compleat cladist. The
University of Kansas Museum of Natural History Special Publication, 19:1-158.

WuTTKE, M. 1988. Untersuchungen zur Morphologic, Palaobiologie und Biostratinomie der mitteleo-
zanen Anuran von Messel. Mit einem Beitrag zur Aktuopalaontologie von Anuren und zur Wei-
chteildiagenese der Wirbeltiere von Messel. Unpub. Ph.D. Dissert., Johannes-Gutenberg-Univer-
sitat, Mainz, 350 pp, Germany.

ZwEiEEL, R. G. 1956. Two pelobatine frogs from the Tertiary of North America and their relationships
to fossil and Recent forms. American Museum Novitates, 1762:1-45.

ANNALS OF CARNEGIE MUSEUM

VoL. 69, Number 3, Pp. 157-193

9 August 2000

HOMOLOGY AND PHYLOGENETIC IMPLICATIONS OP
SOME ENIGMATIC CRANIAL FEATURES IN
GALLIFORM AND ANSERIFORM BIRDS

Richard L. Zusi'

Research Associate, Section of Birds

Bradley C. Livezey

Associate Curator, Section of Birds

Abstract

Two landmarks of the temporal region of the skull in most birds are the zygomatic process (pro-
cessus zygomaticus) and the postorbital process (processus postorbitalis). The morphology and ho-
mology of these processes in gallinaceous birds (Galliformes) and waterfowl (Anseriformes), however,
are not clear. Anseriformes usually are said to lack a processus zygomaticus. By contrast, the processus
zygomaticus of many Galliformes often is described as connected to the tip of the processus postor-
bitalis, forming a temporal arch. Olson and Feduccia (1980a) cited these cranial differences as evidence
opposing a hypothesis of sister relationship between the two orders, an hypothesis having a substantial
history of advocacy (Seebohm, 1889; Shufeldt, 1901; Delacour, 1954; Johnsgard, 1965; Cracraft,
1981a, 1986; Schulin, 1987). Dzerzhinsky (1982, 1995) contradicted the proposal by Olson and Fed-
uccia (1980a), interpreting the two processes as completely fused in Anseriformes, forming a unique
“sphenotemporal process,” which he averred to have been derived evolutionarily from the condition
found in the Galliformes.

In the present study, we examined skulls and jaw muscles of juvenile and adult specimens of selected
taxa from both orders to test these opposing hypotheses, and found that: (a) the processus zygomaticus
is small or lacking in adult Galliformes, and absent in all Anseriformes; (b) the processus zygomaticus
is connected to the tip of the processus postorbitalis by an ossified aponeurosis of m. adductor man-
dibulae externus (aponeurosis zygomatica) in adults of most galliforms, whereas the aponeurosis zyg-
omatica of anseriforms has a linear origin along the os squamosum as far as the processus postorbitalis;
the aponeurosis zygomatica is ossified in Anhimidae and unossified in Anatidae; (c) a laterally exposed
fossa of the temporal region (fossa musculorum temporalium) is reduced in Galliformes and absent
in Anseriformes; (d) pars superficialis and pars zygomatica of m. adductor mandibulae externus are
shifted rostrad in Galliformes and Anseriformes, and (e) pars articularis of m. adductor mandibulae
externus is much enlarged in both orders. Based on these observations, we conclude that the parts of
musculus adductor mandibulae externus of Anseriformes have been misinterpreted in a number of
previous studies, perhaps reflecting confusion about associated processes and fossae. These findings
are interpreted with respect to the homology of the osteological features and their associated muscles.
The distribution of the included states supports the growing consensus for a sister relationship between
the Galliformes and Anseriformes.

Key Words: Anseriformes, cranium, Galliformes, homology, myology, osteology, processus postor-
bitalis, processus zygomaticus

Introduction

Historical Background

A series of classic, nineteenth-century works — e.g., Blanchard (1859), Eyton
(1867), Furbringer (1888), Seebohm (1888, 1889, 1890, 1895), Gadow and Se-

* Curator Emeritus, Division of Birds, National Museum of Natural History, Smithsonian Institution,
Washington, D. C. 20560.

Submitted 24 August 1999.

157

158

Annals of Carnegie Museum

VOL. 69

lenka (1891), Gadow (1892, 1893), and Beddard (1898) — mark the advent of
important contributions of comparative anatomy to an understanding of the evo-
lutionary relationships of birds. Despite diverse, intensive study of avian osteoh
ogy and myology since that time (e.g., Hofer, 1945, 1950, 1955; Starck and
Bamikol, 1954; Yudin, 1961, 1965; Burton, 1984), however, a number of features
of the avian skull have proven problematic for systematists, including the pro-
cessus basipterygoidei of the os parasphenoidale (Weber, 1993), the os lacrimale
(Cracraft, 1968), and the os pterygoideum (Hofer, 1945; Jollie, 1957; Weber,
1993). In recent decades (Hennig, 1966; Wiley, 1981), with the development of
explicit methodologies for the reconstruction of evolutionary relationships based
on anatomical characters, the elucidation of the homologies and transformations
of anatomical structures is recognized to be of paramount importance (Nelson,
1994; Doyle, 1996; Sanderson and Hufford, 1996).

Two features of the avian skull — the processus postorbitalis and the processus
zygomaticus (Fig. 1) — serve as landmarks for the structure and function of the
jaws (Biihler, 1981) but have proven problematic for systematists concerned with
the Anseriformes. In most birds, the processus postorbitalis provides the dorsal
attachment for the ligamentum postorbito-mandibulare, a structure important in
several aspects of cranial kinesis (Kripp, 1933; Zusi, 1962, 1967; Bock, 1964);
the processus zygomaticus supports the origin of a major aponeurosis of the ex-
ternal adductor muscle of the mandibula. Absence of the processus postorbitalis
is unusual and typically associated with a reduction or loss of the ligamentum
postorbito-mandibulare. Absence of the processus zygomaticus, however, does not
necessarily signify a change in the musculus adductor mandibulae extemus. In
the Anseriformes, the processus postorbitalis and adjacent fossae are uniquely
modified among extant birds (Fig. 2), and the processus zygomaticus is essentially
absent in Anseriformes (e.g., Gadow, 1892; Olson and Feduccia, 1980<3). The
processus postorbitalis has become an integral part of a reconstruction of the regio
temporalis (Fig. 1) on the lateral surface of the neurocranium, that part of the
skull enclosing the brain and sensory capsules (de Beer, 1937; ICVGAN, 1983;
Baumel and Witmer, 1993). In Anseriformes, the stout, rostrally oriented proces-
sus postorbitalis also serves as origin for a part (pars zygomatica) of musculus
adductor mandibulae extemus that is independent of the processus postorbitalis
in most avian taxa (Fig. 1). Thus the profound modification of these structures in
waterfowl is of considerable functional and phylogenetic interest.

Conformation of the regio temporalis and mandibular rami reflects the structure
of the primary, superficial adductor acting on the mandibula, m. adductor man-
dibulae extemus (abbreviated hereafter as AME). This complex, multipennate
muscle consists of two or more parts, each of which is more or less distinct and
associated with one or more aponeuroses that provide extensive surface for the
attachment of muscle fibers (Fig. 3). Despite several thorough anatomical surveys
of the avian cranium (Lakjer, 1926; Hofer, 1950; Starck and Barnikol, 1954;
Weber, 1993), delimitation of the parts of AME is somewhat arbitrary for many
taxa because the fibers of one part may blend imperceptibly with those of another
and some aponeuroses are shared by two or more parts. Disagreements about
terminology and homology of the parts of AME in the Anseriformes are consid-
erable (e.g., Lakjer, 1926; Starck and Barnikol, 1954; Dzerzhinsky, 1982; Weber,
1996).

These complexities notwithstanding, selected aspects of the AME and osteo-
logical features of the regio temporalis have figured prominently in studies of a

2000

ZUSI AND LiVEZEY CRANIUM OF GaLLIFORMES AND AnSERIFORMES

159

fos. mus. temp.

mand. arcus jug. pr. zyg.

lig. post. -mand. AME cor.

AME sup. lig. post. -zyg.

AME art. ext. AME art. int.

Fig. 1. — Diagrams of facies lateralis cranii of a hypothetical bird showing; (A) critical osteological
features highlighted in the clear rectangle; and (B, C) critical myological features. See Methods for
list of abbreviations used here and in the following figures.

diversity of avian orders (e.g., Mdller, 1932; Hofer, 1945, 1950; Fiedler, 1951;
Bamikol, 1952, 1953; Bas, 1954, 1955; Fisher and Goodman, 1955; Bams, 1956;
Simonetta, I960a-b, 1963, 1968; Zusi and Storer, 1969; Merz, 1963; Van der
Klaauw, 1963; Yudin, 1965; Bock and McEvey, 1969; Richards and Bock, 1973;
Burton, 1974fl-c, 1984; Morioka, 1974; Bhattacharyya, 1980, 1989; Cracraft,
1982; Johnson, 1984; Zusi and Bentz, 1984; Van Gennip, 1986; Elzanowski,
1987; Zusi, 1993; Dzerzhinsky, 1999), including members of the Galliformes
(Burggraaf, 1954; Burggraaf and Fuchs, 1954, 1955; Fuchs, 1954, 1955; Jollie,
1957; Fujioka, 1963; Dzerzhinsky and Belokurova, 1972; Dzerzhinsky, 1980;
Weber, 1996) and Anseriformes (Davids, 1952; Starck and Bamikol, 1954; Good-

160

Annals of Carnegie Museum

VOL. 69

impr. AME cor. impr. AME art.

Fig. 2. — Facies lateralis cranii of: (A) an adult specimen of Meleagris gallopavo (USNM 556372),
(Galliformes: Meleagrididae); (B) an adult specimen of Sarkidiornis melanotos (USNM 490276), (An-
seriformes: Anatidae); and (C) an adult specimen of Niimida meleagris (USNM 430657), Galliformes:
Numididae). Features unlabelled or controversial in the literature are indicated by question marks.
Scale bar = 1 cm.

man and Fisher, 1962; Zweers, 1974; Dzerzhinsky, 1982; Jager, 1990). In addition
to comparatively traditional studies, the complex has been examined with respect
to structural details and functional roles of the constituent parts (e.g., Gans and
Bock, 1965; Bock, 1964, 1968; Zweers, 1974; Dzerzhinsky, 1982).

The processus zygomaticus of Galliformes was described as well developed by
Gadow (1892), Shufeldt (1909), and Baumel and Witmer (1993), but as absent
or vestigial by Verheyen (1956). Starck and Barnikol (1954) found that the process
zygomaticus was small in juvenile Gall us, and that the aponeurosis of AME, pars
zygomatica, originating on the processus, ossified during maturation; they then
referred to the combined process and ossified aponeurosis as the zygomatic pro-
cess. Hofer (1950) considered the processus zygomaticus to be strong in Melea-
gris and Tetrao, taxa in which the aponeurosis zygomatica is ossified, but inter-
preted the processus to be lacking in Numida, in which the aponeurosis is not
ossified (Fig. 2). Traditionally, the processus zygomaticus was said to be lacking
in Anseriformes (e.g., Gadow, 1892; Olson and Feduccia, 1980a).

2000

ZUSI AND LiVEZEY CRANIUM OF GaLLIFORMES AND AnSERIFORMES

161

AME cor.

B

apon. zyg. apon. temp. apon. zyg. AME art. ext.

Fig. 3. — Diagrams of m. adductor mandibulae externus (AME) complex of a hypothetical bird: (A)
AME superficialis; (B) AME coronoidea; (C) AME zygomatica; and (D) AME articularis. Aponeuroses
shown in black.

Olson and Feduccia ( 1980^:4) argued against a close relationship between the
Galliformes and Anseriformes, stating that “ . . . the tip of the postorbital process
fuses with that of the zygomatic process in Galliformes, leaving a foramen, where-
as in the Anseriformes the zygomatic process is absent.” This anatomical inter-
pretation was part of a larger proposal in which the hypothetical “transitional
shorebirds” (purportedly exemplified by the fossil Presbyornis) were considered
ancestral to several modern orders (Livezey, 1991a), including waterfowl (Fed-
uccia, 1977, 1978, I9ma-b, 1994, 1995, 1996; Olson and Feduccia, 1980Z?). By
contrast, Dzerzhinsky (1995:327-328) concluded that “ in the Anseriformes,
the [ossified muscular aponeurosis from the zygomatic process] fuses to the post-
orbital process over its entire caudo ventral border to form a complete sphenotem-
poral process . . . Dzerzhinsky (1995) considered the “sphenotemporal pro-
cess” of the Anseriformes to be derived from the condition characteristic of the

162

Annals of Carnegie Museum

VOL. 69

Galliformes, bolstering his argument for a sister relationship between the two
orders. Still other workers simply omitted these disputed features from consid-
eration with respect to either order (Livezey, 1986; Cracraft, 1986, 1988; Ericson,
1996, 1997).

Objectives of Study

In this paper we interpret the homologies of osteological features unique to the
Anseriformes through comparative study of skeletal and spirit (fluid-preserved)
specimens of juvenile and adult examples of anseriform and galliform birds. Of
primary concern are the nature of the processus zygomaticus, homologues of the
aponeurosis zygomatica of AME, and the relationship of both with the processus
postorbitalis in Anseriformes. This examination is associated with several new
proposals regarding nomenclature for selected anatomical features. Finally, a com-
parison of the Anseriformes and Galliformes provides a perspective on the evo-
lutionary derivation of the condition of these features in waterfowl.

Materials And Methods
Specimens and Related Data

Criteria for Determination of Age, — Although none of the museum specimens
studied herein was of known age in relation to hatching, we use the terms chick,
juvenile, immature and adult as progressive stages of development based on size
and degree of fusion of suturae cranii. Chicks are birds within a few days of
hatching with fully evident suturae cranii. Juveniles are larger, even approaching
full size, and their neurocranial suturae are variously unfused, Immatures are
essentially full size with the suturae cranii fused except for those between the
processes frontales of the ossa nasales and the neurocranium. In adults all neu-
rocranial suturae are fused; only those between the processes nasales of the paired
premaxillae may remain distinct. Adults often display a more robust skull than
that of immatures.

Osteological Specimens. — Comparisons of adult skeletons of Galliformes and
Anseriformes were based on the entire skeleton collection of USNM, as well as
selected taxa from other museums (AMNH, BMNH, YPM). Taxa in which adult
specimens were compared with one or more specimens of chicks, juveniles, or
immatures are as follows: Galliformes: Megapodiidae-— freycinet,
Leipoa ocellata; Cracidae- — Ortalis vetula, Penelope jacuacu, P. superciliaris, P.
purpurascens, Crax rubra, C. alector, C. fasciolata, Aburria pipile; Meleagridi-
dae — Meleagris gallopavo; Tetraonidae — -Lag opus lagopus, L. mutus, Tetrao te-
trix, Bonasa bonasia, B. umbellatus, Centrocercus urophasianus; Phasianidae--”
Alectoris chukar, Francolinus adspersus, F. capensis, F. sephaena, F. pondicer-
ianus, F. francolinus, Arborophila crudigularis, A. brunneopectus, Bambusicola
thoracica, Ithaginia cruentatus, Lophura leucomelanos, Gallus domesticus, Cros-
soptilon crossoptilon, Catreus walUchii, Chrysolophus pictus, C. amherstiae,
Pavo cristatus, P. muticus; Numididae—Numida meleagris; Odontophoridae— -
CalUpepla squamata, Lophortyx calif ornica, Colinus virginianus. Anseriformes:
Anhimidae- — Chauna torquata; Ana.tida.e-—Dendrocygna bicolor, Anser caerules-
cens, A. canagicus, Branta canadensis, B. bernicla, Cygnus atratus, C. bewickii,
C. columbianus, Tachyeres pteneres, T. patachonichus, Tadorna radjah, T. ta-
dorna, Casarca ferruginea, Chloephaga hybrida, C. picta, Heteronetta atricap-
illus, Oxyura jamaicensus, Anas platyrhynchos, Aythya americana, Somateria

2000

ZUSI AND LIVEZEY CRANIUM OF GALLIFORMES AND ANSERIFORMES

163

mollisima, Histrionicus histrionicus, Melanitta perspicillata, M. fusca, Clangula
hyemalis, Bucephala clangula, B. islandica, Mergus merganser, M. serrator.

Spirit Specimens. — Spirit specimens (adults unless specified) dissected for com-
parison of jaw muscles were as follows: Galliformes: Megapodiidae — Megapo-
dius freycinet; Cracidae- — Ortalis vetula; Meleagrididae — Meleagris gallopavo;
Tetraonidae — Dendragopus canadensis’, Phasianidae — Alectoris graeca, Gallus
domesticus, Francolinus capensis’, Numididae — Numida meleagris’, Odontophor-
idae — Lophortyx gambelii. Anseriformes: Anhimidae — Chauna torquata’, Anser-
anatidae — Anseranas semipalmata (chick); Anatidae — Dendrocygna bicolor
(chick), D. autumnalis, Anser albifrons. Anas crecca, A. versicolor (chick), A.
acuta (chick), Mergus merganser.

Nomenclature and Classification of Galliformes and Anseriformes

For the Galliformes, we adopted the families recognized by Sibley and Monroe
(1990), except that we elevated the three major groups included by them in Phas-
ianidae (Tetraonidae, Meleagrididae, and Phasianidae sensu stricto) to family rank,
as accorded them by Peters (1934) and Wetmore (1951), and used by del Hoyo

et al. (1994).

For purposes of reference in comparative descriptions, tables, and figures, we
adopted the higher-order, phylogenetic classification of waterfowl proposed by
Livezey {1991 a-b). The essentials of this framework are as follows:

Order Anseriformes (Wagler, 1831). — -Waterfowl
Suborder Anhimae Wetmore & Miller, 1926

Family Anhimidae Stejneger, 1885. — Screamers
Genus Anhima Brisson, 1760.- — Homed screamer
Genus Chauna Illiger, 1811 .^-Crested screamers
Suborder Anseres Wagler, 1831.— Tme waterfowl
Superfamily Anseranatoidea (Sclater, 1880)

Family Anseranatidae (Sclater, 1880)

Genus Anseranas Lesson, 1828. — Magpie goose
Superfamily Anatoidea (Leach, 1820). — Typical waterfowl
tFamily Presbyomithidae Wetmore, 1926
Genus Presbyornis Wetmore, 1926
Family Anatidae Leach, 1820. — True ducks, geese and swans

Myological Technique

Dissection of jaw musculature was performed by RLZ on one specimen of each
taxon. The specimens, of varying age and provenance, had been fixed in formalin
and preserved in alcohol. Attention was focussed on m, adductor mandibulae
extemus, the muscle most often associated with the cranial features of concern in
this paper. Although this muscle is largely superficial and readily accessible, an
understanding of its complexity could be gained only through knowledge of its
internal stmcture of aponeuroses and associated muscle fibers. After illustrating
the superficial aspect of the muscle, all fibers were removed systematically, leav-
ing intact the complex of interdigitating, aponeurotic origins and insertions. The
identity of the major aponeuroses (coronoidea, superficialis, zygomatica, paraco-
ronoidea externa and interna, and articularis — Fig. 3) could then be determined
in most instances and the different taxa compared. Uncertainties were resolved
by inspection of the opposite muscle, usually with only partial dissection.

164

Annals of Carnegie Museum

VOL. 69

Anatomical Nomenclature

General Nomenclatural References.-— O^l^ologicdl and arthrological nomencla-
tore, respectively, followed Baumel and Witmer (1993) and Baumel and Raikow
(1993), much of which remained unchanged from the first code proposed by the
International Committee on Avian Anatomical Nomenclature (ICAAN); in the
latter, osteology was treated by Baumel (1979a), arthrology by Baumel (1979^),
and myology by Vanden Berge (1979). The two codes prepared by the ICAAN
were paralleled by standards for veterinary anatomists (Komarek, 1979; Komarek
et al. 1982), which in turn were intended to stabilize names used most frequently
by avian anatomists in recent decades (e.g., Bellairs and Jenkin, 1960; Berger,
1960, 1966). Osteological features labeled using ICAAN nomenclature were fig-
ured in substantial detail elsewhere (Butendieck, 1980; Butendieck and Wissdorf,
1982).

Myoiogical Nomenclature. — Weber (1996) recently compiled a synonymy of
terms used in major myoiogical studies of the cranium and mandibula. In this
paper, myoiogical nomenclature (listed below) for parts of AME mainly follows
Weber (1996), with any synonyms from Vanden Berge and Zweers (1993) given
in parentheses:

M. adductor mandibulae externus (AME)
pars coronoidea (rostralis; temporalis, or
rostralis temporalis)
caput temporale
caput mediate
pars superficialis (lateralis)
pars zygomatica (ventralis; medialis)
pars articularis (profunda; caudalis)
caput interna
caput externa

M. pseudotemporalis superficialis

M. adductor mandibulae posterior (adductor mandibulae
caudalis)

M. depressor mandibulae

The parts of AME will be abbreviated throughout the paper as AME corono-
idea, AME superficialis, AME zygomaticus, and AME articularis, and the heads
of the latter as AME articularis intemis and AME articularis extemis.

The name m. adductor mandibulae posterior was used traditionally until the
compilation by Vanden Berge (1979), in which the term “posterior” was changed
routinely to “caudal.” Under this nomenclatural convention, the name for this
muscle became M. adductor mandibulae caudalis. However, recognizing the pos-
sibility of confusion with AME articularis (also called AME caudalis; see above),
Vanden Berge and Zweers (1993) proposed a new name- — “M. adductor mandib-
ulae ossis quadrati” — -while retaining “adductor mandibulae caudalis” as an ac-
ceptable alternative. We retain the traditional name (m. adductor mandibulae pos-
terior) in the present study because it is used universally in the pertinent literature
on galliforms and anseriforms.

Each of the parts of m. adductor mandibulae externus (AME) is built around
one or more major aponeuroses (as well as some smaller, unnamed aponeuroses),
an architecture first emphasized for establishment of homologies by Bamikol
(1952) and that is evolutionarily conservative despite many adaptive modifications

2000

ZUSI AND LiVEZEY CRANIUM OF GaLLIFORMES AND AnSERIFORMES

165

among avian taxa exhibited in the avian jaw mechanism (Starck and Bamikol,
1954; Zusi, 1962; Dzerzhinsky and Podanova, 1974; Dullemeijer, 1951, 1952;
Weber, 1996). We use the following designations for the aponeuroses of the parts
of the AME, largely after Weber (1996):

AME coronoidea. — aponeurosis coronoidea and aponeurosis temporalis;

AME superficialis.— aponeurosis superficialis; Weber (1996) included AME
superficialis under AME zygomatica, but we provisionally recognize it
here pending a broader comparison of avian taxa;

AME zygomatica. — aponeurosis zygomatica;

AME articularis.-^aponeurosis paracoronoidea externa, aponeurosis paraco-
ronoidea interna, and aponeurosis articularis.

Arthrological Nomenclature. — With respect to nomenclature of ligaments and
joints, we mainly follow Baumel and Raikow (1993). However, in reference to
the complex of ligamenta included by those authors under the name “ligamentum
postorbitale,” we distinguish three separate ligamenta for purposes of this study,
thereby formalizing the substantial variations in attachments noted for this com-
plex (e.g., Lebedinsky, 1921; Zusi and Storer, 1969; Elzanowski, 1987; Jager,
1990): ligamentum postorbito-mandibulare (connecting processus postorbitalis
with the mandibula), ligamentum postorbito-jugale (connecting processus postor-
bitalis with the arcus jugalis), and ligamentum postorbito-zygomaticum (connect-
ing processus postorbitalis with the processus zygomaticus). The last of these
three names is synonymous with the “ligamentum zygomaticum” provisionally
recognized by Elzanowski (1987) and Weber (1996). Although this complex is
extremely variable among taxa and the included ligaments vary in discernability
within the fascia temporalis in which they are sometimes imbedded (Hofer, 1950;
Bamikol, 1953; Bas, 1954; Zusi, 1975; Elzanowski, 1987; Weber, 1996), we con-
cluded that separate, completely descriptive names for these important ligamenta
were critical for the clarity of comparative descriptions.

Osteological Nomenclature. — A term of long-standing in osteological nomen-
clature of most tetrapods is “fossa temporalis” or “temporal fossa,” traditionally
associated with the origin of AME coronoidea. However, among birds, this some-
times prominent feature of the regio temporalis marks the origins of more than
one muscle. For this study, it was critical to ascertain by dissection the relation-
ships of specific muscles associated with specific osteological features of the neu-
rocranium, and in this context a vague term encompassing a series of distinct,
nonhomologous states was not only useless but also misleading. Accordingly, the
“fossa temporalis” of birds has relevance for comparative study only as a broad,
topographic area — a variably differentiated site of origin for one or more unspec-
ified mandibular muscles. For this reason, we propose the explicitly descriptive
“fossa musculorum temporalium” (new term), for “fossa for muscles of the tem-
poral region,” as a replacement for the misleading, traditional name.

The fossa musculorum temporalium of a given species could comprise the
impressiones deriving from one or more of four muscles — AME coronoidea, m.
pseudotemporalis superficialis, AME articularis extemus, and even m. depressor
mandibulae; the term “impressio temporalis” was proposed by van Gennip (1986)
for the scar of a portion of m. depressor mandibulae in the Rock Dove (Columba
livia). In addition, included muscles may occupy different portions of the fossa
without osteological delimitation of the subdivisions. We recommend reference to
the impression of the pertinent muscle whenever it is known (e.g., impressio m.
AME coronoidea).

166

Annals of Carnegie Museum

VOL. 69

A more obscure term in avian osteology is “fossa subtemporalis” or “subtem-
poral fossa.” Typically occupied by AME articularis extemus, this concavity also
has been cited as the position of origo m. depressor mandibulae (Baumel and
Witmer, 1993). Some authors (e.g,, Vickers-Rich et ah, 1995) regarded the con-
fluence of impressiones AME coronoidea and articularis as the temporal fossa in
Megalapteryx (Dinornithiformes), worsening the ambiguity of the term with re-
spect to the homologies of the impressiones involved. Consequently, we abandon
the term fossa subtemporalis and refer to this feature using a directly descriptive
alternative— “impressio AME articularis.

Abbreviations Used To Label Figures

Abbreviations of anatomical terms used in the accompanying figures are listed
below in alphabetical order:

AME art.— musculus adductor mandibulae extemus, pars articularis
AME art. ext.— musculus adductor mandibulae extemus, pars articularis,
caput externa

AME art. int.— musculus adductor mandibulae extemus, pars articularis,
caput interna

AME cor- — musculus adductor mandibulae extemus, pars coronoidea

AME sup.— musculus adductor mandibulae extemus, pars superficialis

AME zyg.— musculus adductor mandibulae extemus, pars zygomatica

apon. art. — aponeurosis articularis

apon. par. ext. — aponeurosis paracoronoidea externa

apon. par. int.— aponeurosis paracoronoidea interna

apon. cor. — aponeurosis coronoidea

apon. sup.— aponeurosis superficialis

apon. temp. — aponeurosis temporalis

apon. zyg.— aponeurosis zygomatica

apon. zyg, oss.— aponeurosis zygomatica ossificans

arcus jug. — arcus jugalis

arcus suborb.— arcus suborbitalis

crist. AME art.— crista musculi adductoris mandibulae extemus, pars
articularis

crist. zyg.— crista zygomatica

fos. mus. temp.- — -fossa musculomm temporalium

impr. AME art.— impressio musculi adductoris mandibulae extemus, pars
articularis

impr. AME cor. — -impressio musculi adductoris mandibulae extemus, pars
coronoidea

lam. parasph.— lamina parasphenoidalis
lig. lac.^mand. — -ligamentum lacrimo-mandibulare
lig. post.-mand.— ligamentum postorbito-mandibulare
lig. post.-zyg. — ligamentum postorbito-zygomaticum
lig. suborb. — ligamentum suborbitale

m. add. mand. post.— musculus adductor mandibulae posterior
mand.- — mandibula

meat, acust. ext.— meatus acusticus extemus
orb. — orbita
os front. — os frontale

2000

ZUSI AND LIVEZEY CRANIUM OF GaLLIFORMES AND AnSERIFORMES

167

OS lacr.-— OS lacrimale

os lat.-sphen. — os laterosphenoidale

os par. — os parietale

os squam. — os squamosum

pr. otic. quad. — processus oticus quadrati

pr. postorb. — processus postorbitalis

pr. zyg, — processus zygomaticus

regio temp. — regio temporalis

rost. parasph.— rostrum parasphenoidale

sut. front.-squam. — sutura fronto-squamosa

sut. lat.” squam.— -sutura laterospheno-squamosa

tuber. — tuberculum

Concepts and Diagnosis of Homology

General Principles and Terms. — The concept of the homologue, first defined
by Owen (1843) as “ . . . the same organ in different animals under a variety of
form and function” {fide Panchen, 1994), originated and principally remains with-
in the context of comparative anatomy (Boyden, 1943, 1947; Patterson, 1982;
Van Valen, 1982; Young, 1993; Sluys, 1996). Despite the antiquity and generally
narrow context of the concept, the issue of homology and its practical application
remain the subject of substantial controversy (Rieppel, 1980, 1992, 1994; Roth,
1984, 1988, 1991; Sattler, 1984). Subsequently, however, the concept and criteria
for the diagnosis of homology have been recognized as equally vital and chal-
lenging for phylogenetic interpretation of characters as diverse as DNA sequences
(Patterson, 1988; Mindell, 1991; Hillis, 1994; Brower and Schawaroch, 1996),
proteins (Fitch, 1970), metric abstractions (Bookstein, 1994; Fink and Zelditch,
1995; Zelditch et al., 1995, 1998; Adams and Rosenberg, 1998; Rohlf, 1998;
Swiderski et al., 1998; Zelditch and Fink, 1998), and behavioral repertoires (Wen-
zel, 1992; Greene, 1994). In that Owen (1843) originally contrasted “ana-
logues”— -phylogenetically independent structures having common functions in
different taxa^ — from “homologues,” it is not surprising that an emphasis on func-
tion persists in the diagnosis of homology and the utility of characters for phy-
logenetic reconstruction (Bock, 1967, 1977, 1979, 1989; but see Cracraft, 1981Z?).
In its most essential form, the definition of homologues is structures having a
common evolutionary origin (Simpson, 1959; Bock, 1963^); an important impli-
cation of this definition is that homologues would share ontogenetic bases (Wag-
ner, 1989a-^, 1994; Goodwin, 1994; Hall, 1994, 1995).

These theoretical essentials of homology, however instructive, provide little in
the way of practical methodology for the recognition of homologues in a phylo-
genetic context. In this work, we essentially applied the three classical criteria of
Remane (1952, 1956): (1) similarity of position; (2) quality of resemblance (see
also Inglis, 1966); and (3) continuance of similarity through intermediate forms
(Wiley, 1981). Problems of homology do not increase necessarily with the com-
plexity of the characters in question; in fact, structural detail often provides the
distinctions essential to diagnosis of homology, rendering many anatomical sys-
tems more amenable to such determinations than simple features such as sequence
data (Wagele, 1995; McShea, 1996).

Specific Criteria Applied. — Use of published descriptions of jaw muscles is
complicated by different terminologies and different judgments about homologies.

168

Annals of Carnegie Museum

VOL, 69

We hypothesize homologies from similarity in external and internal structure of
muscles, and location of muscles in relation to each other and to associated bony
features. Within a phylogenetic context, inference of homology assumes an iter-
ative nature, in which a priori assessments of homology can be questioned on
the basis of the most parsimonious interpretation of the totality of evidence (Ste-
vens, 1984; Bryant, 1989; de Pinna, 1991; Haszprunar, 1992, 1998; Lipscomb,
1992; Coddington, 1994; McKitrick, 1994; Brooks, 1996; Hawkins et aL, 1997;
but see Lauder, 1994). The confirmatory advantages of this process hinge on the
validity of the delimitation of characters and composite states (Pogue and Mick-
evich, 1990; Barriel and Tassy, 1993). In this work, we emphasize the primary
assessment of homology by comparative study, relegating most phylogenetic im-
plications of this study to companion works (Livezey, 1997a, 1998).

Innervation is not included here because the muscles under discussion-^parts
of the AME — are supplied mainly by branches of nervus trigeminus mandibularis;
this complex varies sufficiently within species (Bamikol, 1953, 1954) to suggest
that data from single specimens could be misleading. Available information on
the associated systema cardiovasculare (Baumel, 1993) also provided no critical,
ancillary clues to homology of subdivisions of the AME (e.g., Richards, 1968).

Examination of crania of very young birds was critical for discernment of most
or all suturae cranii. Therefore direct study of prepared skeletons and fluid-pre-
served specimens of juveniles was performed for as many relevant taxa as pos-
sible, supplemented by reference to the literature on the ontogeny of cranial el-
ements and overlying musculature in a diversity of avian taxa (e.g., Edgeworth,
1907; Jollie, 1957; Hogg, 1978). Although all parts of the AME are derived from
a single primordium (McCleam and Noden, 1988), the study of juveniles provided
additional insight into the structure of AME in that parts of this complex were
clearly separable even in early developmental stages.

Changes in aspects of osteological or myological features during development
per se, however, were not used to infer directions of evolutionary change among
taxa, but instead as a means for delimitation of homologous anatomical structures
that are rendered less distinguishable in adults through variation in function and
selection pressures (Hanken and Hall, 1993). The relevance of such information
to the polarity of character states (i.e., the “ontogenetic criterion”) remains con-
troversial (Nelson, 1978; Alberch, 1985; de Queiroz, 1985; Kluge and Strauss,
1985; Kraus, 1987; Mabee, 1989, 1993; Wheeler, 1990; Williams et al., 1990; de
Pinna, 1994; Meier, 1997). In the present paper, references to phylogenetic po-
sition and polarity were based on previous works in which outgroup comparisons
were employed (Livezey, 1986, 1989, 1991, 1995a-£:, 1996a-c, 1991 a-c, 1998).

Pertinent Anatomical Issues

Processus Postorbitalis.^ThQ processus postorbitalis usually arises from the
caudolateral border of the orbita. With few exceptions, the processus postorbitalis
is largely derived from the os laterosphenoidale, with variable contribution from
the rostral portion of the os squamosum in most neognathous taxa. Typically the
processus is oriented roughly perpendicularly to the long axis of the mandibula
and it serves as origin for the ligamenta postorbito-mandibulare, postorbito-jugale,
and ligamentum suborbitale, which extend ventrally to the mandibula and arcus
jugalis and rostrally to the os lacrimale or os ectethmoidale, respectively. In some
taxa, the complex of ligamenta arising from the processus postorbitalis includes,

2000

ZUSI AND LiVEZEY CRANIUM OE GaLLIFORMES AND AnSERIFORMES

169

in addition to the comparatively conspicuous ligamentum postorbito-mandibulare,
a variably distinct component (ligamentum postorbito-zygomaticum) that attaches
on the processus zygomaticus (Starck and Bamikol, 1954). The ligamenta arising
from the processus postorbitalis may be slender or absent, and the processus
postorbitalis correspondingly reduced.

Therefore the processus postorbitalis may support the complex comprising the
ligamenta postorbitalia as well as the ligamentum suborbitale. These ligamenta
ossify in some taxa such that, in adult birds, the processus postorbitalis may
appear to be extended ventrally by dorsal ossification of the ligamentum postor-
bito-mandibulare, rostrally by partial or complete ossification of the ligamentum
suborbitale, or caudally by ossification of the ligamentum postorbito-zygomati-
cum.

Processus Zygomaticus. — This processus of the os squamosum is located on
the regio temporalis of the cranium between the processus postorbitalis and the
meatus acusticus externus, in many taxa immediately rostral or dorsal to the me-
atus. Oriented rostro ventrally, or sometimes extended laterally, it supports the
aponeurosis zygomatica of AME zygomatica; in some taxa this aponeurosis be-
comes ossified at its base, effectively extending the processus. Although the AME
zygomatica and aponeurosis zygomatica are usually present in birds, the processus
zygomaticus may be absent or reduced to an indistinct crista in some taxa (e.g.,
Sulidae, Phalacrocoracidae, Ardeidae, Phoeniculidae).

Fossa Musculorum Temporalium. — Usually this impressio is occupied largely
or wholly by the AME coronoidea in neognathous birds, but in paleognathous
birds the area supports m. pseudotemporalis superficialis wholly or in part (Hofer,
1945; Webb, 1957; Elzanowski, 1987; Weber, 1996). Additional complexity of
the muscles associated with this fossa were noted above.

Impressio AME Articularis. — This variably distinct depression, sometimes
termed “fossa subtemporalis,” lies caudal or ventral to the fossa musculorum
temporalium in some taxa, and is partly delimited by the processus zygomaticus
and the meatus acusticus externus. Occasionally it has been regarded as part of
fossa musculorum temporalium. This impressio is occupied by AME articularis
externus.

AME. — This complex muscle (Fig. 1, 3) arises variously from the fossa mus-
culorum temporalium, processus zygomaticus, processus oticus quadrati, and im-
pressio AME articularis in most birds. Pars coronoidea usually occupies part or
all of the fossa musculorum temporalium and facies lateralis of aponeurosis tem-
poralis and inserts on aponeurosis coronoidea of the mandibula. Pars zygomatica
originates mainly from the medial surface of aponeurosis zygomatica and has a
fleshy insertion on the lateral surface of the mandibula. Pars superficialis origi-
nates from the lateral surface of aponeurosis zygomatica and, in some taxa, from
the fascia temporalis, ligamentum postorbito-mandibulare, and the ligamentum
postorbito-zygomaticum. It inserts rostrally on aponeurosis superficialis and the
adjacent mandibular surface. Caput interna of pars articularis originates from apo-
neurosis articularis and the processus oticus of os quadratum and inserts on apo-
neuroses paracoronoidea interna and externa and on the adjacent portion of the
mandibula; some taxa have a caput externa of AME that originates from the
impressio AME articularis and on part of the lateral surface of aponeurosis zyg-
omatica, and inserts mainly on aponeurosis paracoronoidea externa and on the
mandibula.

In the absence of the processus zygomaticus, we consider the point of attach-

170

Annals of Carnegie Museum

VOL. 69

ment of the aponeurosis zygomatica, which lies between the fossa musculorum
temporalium and origo AME articularis, to be homologous among taxa. Although
aponeurosis zygomatica is the major aponeurosis of AME zygomatica, it also
receives fibers rostrolaterally from AME superficialis, dorsomedially from AME
coronoidea, and ventrolaterally from AME articularis. Association with these three
parts of the AME is characteristic of the aponeurosis zygomatica in various neog-
nathous birds, whether the aponeurosis occurs as an extension of processus zy™
gomaticus or originates from the cranium in the absence of the processus. Within
a few orders (e.g. Pelecaniformes, Ciconiiformes, Coraciiformes, Passeriformes),
the AME articularis extemus and its impressio on the cranium are enlarged in
some taxa, occupying a significant portion of the regio temporalis immediately
caudal to fossa musculorum temporalium (Eiedler, 1951; Bamikol, 1952; Richards
and Bock, 1973).

Results

Processus Zygomaticus

Galliformes. — We found the processus zygomaticus to be present in most
chicks and juveniles as an inconspicuous crista or tuberculum on the ventrolateral
facies of os squamosum, between the processus postorbitalis and meatus acusticus
extemus (Eig. 4). The crista is typically oriented obliquely on a parasagittal plane
along the ventral edge of os squamosum, and aponeurosis zygomatica arises from
it as a flat band passing rostroventrally.

Anseriformes. — We found no clear evidence of a processus zygomaticus on the
os squamosum in juveniles, immatures, or adults of many Anseriformes; in some
taxa, however, a tuberculum on the crista of origin of the aponeurosis zygomatica
at its caudal extremity may represent the processus (Eig. 5E). However, the con-
formation of the rostral portion of os squamosum resembles a processus in some
juvenile anatids. Os squamosum borders the entire length of the processus pos-
torbitalis at sutura laterospheno-squamosa in anhimids, and its basal (dorsal) one-
half in anatids. In most Anatidae, the processus postorbitalis is strongly angled
rostroventrad, and sutura laterospheno-squamosa conforms to this orientation of
the processus, as seen in immature specimens (Eig. 5). Eurthermore, sutura fronto-
squamosa of anatids is located closer to the dorsal limit of the processus postor-
bitalis than in galliform birds, and in some anatids the sutura lies only slightly
above the processus postorbitalis (Eig. 5). In the latter case, the dorsoventrally
compressed, anterolateral facies of os squamosum, in combination with its some-
times pointed rostral extremity, offers a spurious resemblance to a processus zy-
gomaticus in direct association with the processus postorbitalis.

Ossification of Aponeuroses

Galliformes. — Our survey of skeletal specimens of Galliformes of all age clas-
ses revealed that almost all chicks and juvenile specimens, and some immatures,
lacked ossification of aponeurosis zygomatica as indicated by its absence from
prepared skeletons. Occasional specimens of chicks and juveniles had tiny splints
of ossified aponeurosis attached to the zygomatic process by syndesmosis or syn-
ostosis. By contrast, ossified portions of the aponeurosis were an integral part of
most immature and all adult specimens of many galliform taxa (Eig. 4, 6).

During development, ossification of the aponeurosis begins basally and extends
rostrad to the level of the processus postorbitalis or beyond, but never to the

2000

ZUSI AND LiVEZEY CRANIUM OF GaLLIFORMES AND AnSERIFORMES

171

impr. AME cor. pr. zyg.

apon. zyg. oss.

sut. front. -squam. os squam. impr. AME cor.

Fig. 4. — Facies lateralis cranii of selected Galliformes: (A) Leipoa ocellata (USNM 346351), immature
(Megapodidae); (B) L. ocellata (USNM 345086), adult (Megapodidae); (C) Penelope jacquacu
(USNM 345564), immature (Cracidae); (D) P. purpurascens (USNM 613959), adult (Cracidae); (E)
Meleagris gallopavo (USNM 611021), chick (Meleagrididae); (F) M. gallopavo, juvenile (USNM
501018); (G) M. gallopavo (USNM 556388), immature (Meleagrididae); and (H) M. gallopavo (USNM
556372), adult (Meleagrididae). Scale bar = 1 cm.

172

Annals of Carnegie Museum

VOL. 69

os squam. sut. front. -squam.

A B

os squam. impr. AME art. os squam. impr. AME art.

C D

sut. lat.“Squam. os squam. pr. postorb. impr. AME art.

Fig. 5. — Facies lateralis cranii of selected Anseriformes: (A) Chauna torquata (BMNH 1954-5-3),
juvenile (Anhimidae); (B) Anatinae sp. (AMNH 8737), chick (Anatidae); (C) Anas platyrhynchos
(BMNH 1986-48-1), juvenile (Anatidae); (D) Clangula hyemalis (AMNH 6046), juvenile (Anatidae);
(E) Dendrocygna bicolor (USNM 501992), juvenile (Anatidae); and (F) D. javanica (USNM 343514),
adult (Anatidae). Scale bar = 1 cm.

rostral extremity of the aponeurosis. The juncture of aponeurosis zygomatica os-
sificans with processus zygomaticus is usually synostotic, but occasionally the
juncture is syndesmotic even in adults, indicating the limited extent of the pro-
cessus. Typically, the aponeurosis zygomatica ossificans is markedly flattened
lateromedially or in a ventrolateral-dorsomedial plane, bet instead it may be ir-
regular, with longitudinal plicae (e. g., Aepypodius, Chauna; Fig. 7). In the latter
case, the delimitation of the processus from a stout, ossified aponeurosis in adults

2000

ZUSI AND LiVEZEY CRANIUM OF GaLLIFORMES AND AnSERIFORMES

173

impr. AME cor. impr. AME art. impr. AME cor.

crist. AME art. impr. AME art.

Fig. 6. — Diagrams of the typical arrangement of the aponeurosis zygomatica of: (A) a generalized
neornithine bird; (B) a galliform; (C) an anhimid; and (D) a member of the Anseres. Membranous

portion is cross-hatched, ossified portion is shown in black.

is more difficult, but the longitudinal patterning of robust and complicated apo-
neuroses is continuous with their ossified portion (Zusi, personal observation).
Although no juvenile of Aepypodius was examined, we compared adults of Alec-
tura lathami with the illustration of a juvenile (Weber, 1996:fig. 3). The aponeu-
rosis zygomatica ossificans of adult A. lathami resembles that of Aepypodius, but
the processus zygomaticus of the juvenile is a robust tuberculum as in other
megopodes. Aponeurosis zygomatica, whether or not ossified, may have no con-
nection with the processus postorbitalis (Megapodiidae). However, in most taxa
this aponeurosis is anchored to the processus by a connection with ligamentum
postorbito-zygomaticum, which is ossified in adults of most galliforms (Phasian-
idae, Tetraonidae, Meleagrididae, Odontophoridae, and some Cracidae).

Anseriformes. — In the anhimid Chauna, the aponeurosis zygomatica has a lin-
ear attachment along lamina lateralis cranii from a point rostral to meatus acus-
ticus externus to, or nearly to, the processus postorbitalis, where the aponeurosis
passes medial or ventromedial to the terminus of the processus (Fig. 7). Rostral
to the processus postorbitalis, the aponeurosis is free from the cranium (Fig. 6).
In adults the aponeurosis is ossified to a point level with, or more often, rostral

174

Annals of Carnegie Museum

VOL. 69

os front. crista AME art.

Fig. 7. — Facies rostrolateralis cranii (A) and facies lateralis cranii (B-E) of adult specimens of: (A, B)
Chauna torquata (USNM 614547), (Anseriformes: Anhimidae); (C) Anseranas semipalmata (USNM
347638), (Anseriformes: Anseranatidae); (D) Sarkidiornis melanotos (USNM 490276), (Anseriformes:
Anatidae); and (E) Aepypodius arfakensis (YPM 7594), (Galliformes: Megapodiidae). Scale bar =
1 cm.

2000

ZUSI AND LiVEZEY CRANIUM OF GaLLIFORMES AND AnSERIFORMES

175

to the processus postorbitalis, but as in galliform birds, the rostrahmost portion
of the aponeurosis does not ossify.

Anatids show little or no ossification of aponeurosis zygomatica (Fig. 6). The
caudal-most portion of the aponeurosis arises on the cranium from the rostral
limit of impressio AME articularis or sometimes within the impressio, and con-
tinues rostrad along a linea or weak crista as far as the apex of the processus
postorbitalis; beyond the processus the aponeurosis extends rostroventrad toward
the mandibula, independent of the cranium. Ossification, if any, produces a low,
irregular crista (crista zygomatica) where the aponeurosis meets the cranium
(Fig. 6).

Processus Postorbitalis and Processus ''Sphenotemporalis'’

Galliformes. — The processus postorbitalis of Galliformes is well developed,
straight, and oriented approximately ventrad or somewhat rostrad, and continuous
with the margo (rima) caudalis of the orbita. The processus arises largely from
the os laterosphenoidale, and meets os squamosum only at its dorsal limit. The
ligamentum postorbitale-mandibulare is strong and unossified except at its ex-
treme dorsal limit in some Tetraonidae.

Anseriformes. — In the Anhimidae, the processus postorbitalis is not well de-
fined as a process in lateral view, but in rostral perspective it constitutes a ventrally
directed hook of os laterosphenoidale (Fig. 7). Caudal to the apex of the processus
postorbitalis, the lamina lateralis cranii is undercut ventrally, forming an over-
hanging crest that extends caudodorsally from the processus postorbitalis. Dorsal
to this crista AME articularis (new term), the regio temporalis forms a wedge-
shaped area defining an angle of 60-70° and delimited by the caudal contour of
the orbita and by crista AME articularis (Fig. 7). Dzerzhinsky (1982) termed this
wedge the “sphenotemporal process.” Aponeurosis zygomatica lies adjacent and
medial to crista AME articularis, arising along the facies medialis of the crista as
far as the apex of the processus postorbitalis, to which it passes ventromediad.

The skull of a large anhimid chick (Chauna) exhibits a sutura laterospheno-
squamosa that borders the processus postorbitalis for almost its full length. By
contrast, in chicks and immatures of the Anatidae, os squamosum borders only
the base of the processus postorbitalis (Fig. 5). A processus zygomaticus is not
visible in any anseriform, and the aponeurosis zygomatica takes its caudalmost
point of origin well caudal to the processus postorbitalis (Fig. 6). These facts
suggest that the “sphenotemporal” process was formed (in an evolutionary sense)
not by fusion of processes, but more likely by medial retreat of impressio AME
coronoidea and rostral extension of origo aponeurosis zygomatica to the processus
postorbitalis.

Anseres differ from anhimids in several respects. First, the processus postor-
bitalis of os laterosphenoidale extends well beyond os squamosum (Fig. 5). Sec-
ond, the processus postorbitalis usually extends rostrad to form the caudoventral
margin of the orbita, and the sutura laterospheno-squamosa often is angled cor-
respondingly. Third, the homologue of the crista AME articularis of anhimids is
not a well-defined crest in Anseres (except Anseranas), but rather a faint linea or
crest extending caudad from margo caudoventralis of the processus postorbitalis
roughly parallel to crista zygomatica and continuous with the margo dorsalis of
impressio AME articularis. Aponeurosis zygomatica attaches just medial or ven-

176

Annals of Carnegie Museum

VOL. 69

AME zyg.

AME cor.

AME

m. add. mand. post.

m. add. mand. post.

AME art. ext. AME art. int.

lig. lac. -mand.

lig. post. -mand.

AME sup. AME art. ext.

m. add. mand. post
AME art. ext. AME art. int
AME sup
AME zyg.

m. add. mand. post.

AME art. int.
AME cor.

m. add. mand. post.

m. add. mand. post.

Fig. 8. — Detailed illustrations of the AME complex (left lateral views) of: (A) Ortalis vetiila (USNM
344381), adult (Galliformes: Cracidae); (B) Alectoris graeca (USNM 540255), adult (Galliformes:
Phasianidae); (C) Chauna torquata (USNM 508682), adult (Anseriformes: Anhimidae); (D) Anseranas

2000

ZUSI AND LiVEZEY CRANIUM OF GaLLIFORMES AND ANSERIFORMES

177

tromedial to crista AME articularis (Fig. 6, 7). Fourth, the aponeurosis zygomatica
is unossified or forms only a roughened crista along its origin on os squamosum.

M. Adductor Mandibulae Externus

Our findings on the structure of AME in terms of muscle fibers and aponeuroses
agree essentially with those of Weber (1996) for the Megapodiidae, Dzerzhinsky
(1980) for the Cracidae, and Dzerzhinsky and Belokurova (1972) for the Tetraon-
idae. Also, our inferences concur in large part with those by Davids (1952) and
Zweers (1974) for the Anatidae, and with those by Dzerzhinsky (1982) for the
Anhimidae. However, our interpretations of homology of muscle parts differ from
the conclusions of those authors to varying degrees.

We applied the following anatomical generalizations derived from the literature
on avian jaw musculature to the interpretation of structure in Galliformes and
Anseriformes: (a) muscle fibers from all four parts of AME originate on aponeu-
rosis zygomatica in the Galliformes, Anseriformes, and other avian orders (Fig.
3); (b) by contrast, fibers to aponeurosis superficialis represent largely AME su-
perficialis, those to aponeurosis coronoidea represent mainly AME coronoidea,
and those to aponeuroses paracoronoidea externa and interna represent primarily
AME articularis; (c) some fibers of AME superficialis are inseparable from some
fibers of AME coronoidea and AME zygomatica; and, (d) similarly, fibers of
AME zygomatica blend with those of AME articularis. Based on the above con-
ventions, we conclude that AME articularis is much enlarged in galliform and
anseriform birds, and that AME zygomatica and AME superficialis are thereby
displaced rostrally (Fig. 8). Although our interpretations differ radically from
those of Lakjer (1926) and his followers (see below and Table 1), they agree
substantially with those of Dzerzhinsky (1982) and Weber (1996).

In anseriforms, AME superficialis and AME zygomatica are distinct and typical
in form except that both arise from aponeurosis zygomatica rostral to its attach-
ment on the processus postorbitalis, and from an additional, short aponeurosis
arising on that processus. Unique to adult specimens of Dendrocygna is an ossified
ligamentum suborbitale, a structure fused with the tip of the processus postorbitale
and os lacrimale, and forming an arcus suborbitalis (Shufeldt, 1914; Schipler,
1926; Livezey, 1995^). In adults, some fibers of AME superficialis arise from the
arcus rostral to the processus postorbitalis (Fig. 8).

The interpretation of homologies within galliformes is clouded by certain spe-
cializations. AME superficialis and AME zygomatica are variously developed or
merged within Galliformes, but, except for the Megapodiidae, their origins from
aponeurosis zygomatica are supported by the processus postorbitalis, as they are
in the anseriforms. These combined muscle parts blend also with AME articularis
externus in phasianid galliforms. Commensurate with the blending of muscle parts
are modifications of aponeuroses superficialis and paracoronoidea externa, which
form a continuous sheet that inserts along the facies lateralis of the mandibula

semipalmata (USNM, uncataloged), chick (Anseriformes: Anseranatidae); (E) Dendrocygna bicolor
(CM 2117), chick (Anseriformes: Anatidae); (F) D. autumnalis (CM 5247), adult (Anseriformes:
Anatidae); (G) Anas versicolor (USNM 345162), chick (Anseriformes: Anatidae); and (H) Anas acuta
(USNM 225218), chick (Anseriformes: Anatidae), with AME superficialis, zygomatica, and articularis
externa removed. Scale bar = 5 mm.

178

Annals of Carnegie Museum

VOL. 69

Table 1. — Anatomical nomenclatures applied to the m. adductor mandibularis externus complex
(AME) of selected non-anseriform and anseriform birds by Lakjer ( 1926), Starch and Barnikol (1954),

and the present study.

Non-anseriform birds

Anseriform birds

Lakjer (1926)^

III profundus^

I superficialis

III profundus^

II medialis

Present study’’

AME coronoidea

AME superficialis

AME articularis

AME zygomatica

Lakjer (1926)^

[absent]

I superficialis a

I superficialis b

I superficialis c

III profundus

II medialis

Present study^

AME articularis

externus

AME zygomatica

AME superficialis
AME articularis
internus

AME coronoidea

Starck and Barnikol

Present study’’

Starck and Barnikol

Present study'’

(1954)g

(1954)”

Aponeurosis 1

AME coronoidea

Aponeurosis 1

AME articularis

portion

portion (right)

externus (part)

Aponeurosis 1

AME zygomatica

portion (middle)

Aponeurosis 1

AME superficialis

portion (left)

et coronoidea

Aponeurosis 2

AME zygomatica et

Aponeurosis 2

AME articularis

portion

superficialis

portion

externus (part)

Aponeurosis 3

AME articularis

Aponeurosis 3

AME articularis

portion

internus et

portion

internus

externus

^ Cepphus grylle.

Cepphus grylle.

Melanitta nigra.

Anseranas semipalmata, Dendrocygna bicolor, D. autumnalis, Anser albifrons. Anas versicolor, and

A. acuta.

® Labelled “rostr” in some figures.

^Labelled “kaud” in some figures,
s Various non-passeriform taxa.

^ Anas platyrhynchos.

and merges with aponeurosis paracoronoidea interna in at least some members of
all galliform families.

Impressio AME Coronoidea

Galliformes. — The impressio AME coronoidea is variable in size but always
limited in extent among galliform birds, and it occupies the lamina lateralis cranii
between the processus postorbitalis and processus zygomaticus. In many taxa, but
most prominently in the Megapodiidae, the fossa is barely perceptible in lateral
view and the impressio is rotated mediad, occupying portions of ossae squamosum
and laterosphenoidale (Fig. 4, 7, 9). In those taxa characterized by an ossified
ligamentum postorbito-zygomaticum, the fossa is partially enclosed laterally
(Fig. 4).

Anseriformes. — Most waterfowl lack a laterally exposed impressio AME co-
ronoidea. Instead, the impressio occupies a comparatively medial position, and is
largely or completely overhung by crista AME articularis (Anhimidae, Anseran-
atidae) or crista zygomatica (most Anatidae). From a ventrolateral perspective,
impressio AME coronoidea is visible at the junctura of the os squamosum and

2000

ZUSI AND LiVEZEY CRANIUM OF GaLLIFORMES AND AnSERIFORMES

179

pr. postorb. fos. mus. temp.

A

B

pr. postorb. impr. AME cor. impr. AME cor. impr. AME art.

C D

Fig. 9. — Facies lateralis cranii (A, B) and facies ventrolateralis cranii (C, D) of: (A, B) Mergus
merganser (USNM 555255), adult, (Anseriformes: Anatidae), showing fossa musculares temporalis
and (B) origins of AME coronoidea (horizontal hatching) and AME articularis (vertical hatching),
inferred through dissection of USNM 505780; (C) Meleagris gallopavo (USNM 501018), juvenile
(Galliformes: Meleagrididae); and (D) Anas platyrhynchos (BMNH 1986-48-1), juvenile (Anserifor-
mes: Anatidae). Scale bar = 1 cm.

OS laterosphenoidale, between the tip of the processus postorbitalis and impressio
AME articularis (Fig. 9).

In most members of Anatidae, impressio AME coronoidea extends to the tip
of processus postorbitalis as a ventrolaterally directed planum. Processus postor-
bitalis is variably reduced in size and more ventrally directed within Mergini, and
exhibits a corresponding reduction in its involvement with impressio AME co-
ronoidea. Only in Lophodytes, Mergellus, and Mergus (Mergini) is impressio
AME coronoidea largely free from the processus postorbitalis and fully exposed
in lateral view, where it merges imperceptibly with impressio AME articularis
(Fig. 9). A less extreme but similar condition occurs in Biziura (Oxyurini). In
these birds, crista zygomatica is much reduced in prominence. However, it is
likely that AME coronoidea is obstructed largely or completely in lateral view by
AME articularis in all Anseriformes.

180

Annals of Carnegie Museum

VOL. 69

Impressio AME Articularis

Galliformes. — Usually, this impressio is small or absent in gallinaceous birds
(Fig. 9). The AME articularis originates extensively from the portions of aponeu-
rosis zygomatica that lie both caudal and somewhat rostral to the processus pos-
torbitalis, whether or not the aponeurosis is ossified (Fig. 8). Although AME
articularis is enlarged, it does not expand dorsally on the regio temporalis as
mentioned above for some other avian taxa.

Anseriformes. — The impressio AME articularis is small and located medially
in Anhimidae and Anseranatidae, but the AME articularis expands rostrally along
aponeurosis zygomatica as in Galliformes. Most Anatidae have a small- to me-
dium-sized impressio AME articularis (Fig. 5, 7). When small, the fossa occupies
the cranium between the caudal attachment of aponeurosis zygomatica and the
meatus acusticus extemus as in some Galliformes. Anseres resemble galliforms
and anhimids in their relatively large AME articularis, which attaches rostrad
along the caudal section of aponeurosis zygomatica to the processus postorbitalis.
In addition, crista AME articularis gives rise to a superficial aponeurosis that
extends rostrad, roughly parallel and dorsal to the linear origin of aponeurosis
zygomatica. The narrow area between the lines of origin of these aponeuroses
constitutes a rostral expansion of impressio AME articularis.

In some anatids (e.g., Mergus merganser), the impressio also expands dorsally
and crista zygomatica is much reduced; thus, the fossa includes impressio AME
coronoidea rostrally and impressio AME articularis caudally (Fig. 9). As there is
no linea separating these adjacent muscle scars in Mergus, the single fossa is best
referred to as fossa musculorum temporalium in this genus. Also, we could not
confirm the finding of Goodman and Fisher (1962) that AME coronoidea (their
AME medialis) occupies the entire fossa.

Discussion

Homology of Processus Zygomaticus

Among the Galliformes, the processus zygomaticus is represented by an indis-
tinct boss or crista. The processus is separated from the processus postorbitalis,
and in many taxa the aponeurosis zygomatica is supported by the processus pos-
torbitalis through ligamentum postorbito-zygomaticum (ossified or unossified).
Anseriformes are characterized by the absence of a distinct processus zygomati-
cus, but the homologous locus may be marked by a tuberculum. However, wa-
terfowl also are characterized by the extended origin of aponeurosis zygomatica
along a linea or crista extending from impressio AME articularis rostrad toward
or to the processus postorbitalis. Thus, in both the Galliformes and Anseriformes,
processus zygomaticus (or its homologous locus) lies caudal to the processus
postorbitalis.

This interpretation is contrary to the description by Dzerzhinsky (1982, 1995)
of a processus “sphenotemporalis” in Anseriformes and its evolutionary deriva-
tion by fusion of the processus zygomaticus (or ossified aponeurosis zygomatica)
with the processus postorbitalis. Our interpretation is influenced by the following
facts pertaining to Anseriformes: (a) much of processus postorbitalis is formed
exclusively from the os laterosphenoidale, a composition typical of many avian
orders; (b) there is no indication of a processus zygomaticus (much less a long
one) or an ossified aponeurosis zygomatica in skulls of juvenile, immature, or
adult Anseres; and, (c) an ossified aponeurosis zygomatica in anhimids fuses with

2000

ZUSI AND LIVEZEY CRANIUM OF GaLLIFORMES AND AnSERIFORMES

181

OS squamosum mainly caudal to processus postorbitalis, and in some specimens
passes medial or ventromedial to the apex of that processus. These distinctions,
although subtle, permit a more precise definition of homologous characters.

Evolution of Unique Anseriform Morphology

We found little evidence favoring any single ontogenetic or evolutionary mech-
anism that would best explain the probable transformation to anseriform mor-
phology from that of a common ancestor with galliforms. However, two such
hypotheses could be modelled after morphological states represented among the
galliform taxa we examined: (a) reduction and medial rotation of impressio AME
coronoidea and development of an overhanging crista (Megapodiidae);
(b) reconfiguration of the position or shape of the processus zygomaticus or os-
sified aponeurosis zygomatica (and/or processus postorbitalis) to constrict the im-
pressio AME coronoidea (Coturnix). A third hypothesis — ossification of the apo-
neurotic surface of AME coronoidea, thereby closing the fossa musculorum tern-
poralium — was not reflected in the morphology of any galliform examined.

Interpretation of Adductor Mandibulae Externus

Here we compare our interpretation of muscle homologies with those of authors
who have studied not only Galliformes and Anseriformes, but also a variety of
other orders. Terminology applied by each author to jaw muscles in taxa whose
structure is non-controversial provided the key to understanding their concepts of
muscle homologies in the more problematic Galliformes and Anseriformes.

In order to facilitate comparisons with the published works that disagree most
markedly with our interpretation of the myology of anseriforms, we present a
synonymy for the nomenclature applied to the AME by Lakjer (1926) and Starck
and Barnikol (1954), based on taxa for which their interpretations of the muscles
agree with those presented here (Table 1). We also show the interpretations of
both authors concerning the AME in the Anatidae, using both their terminology
and ours to highlight the differences in interpretation (Table 1). Lakjer (1926) did
not distinguish between AME articularis externus and internus. In essence, Lakjer
grouped our AME articularis externus, zygomaticus, and superficialis into AME
superficialis, and he considered AME coronoidea to be absent. Starck and Bar-
nikol (1954) synonymized our AME superficialis as part of AME zygomatica.
They combined our AME articularis externus (rostral portion), AME zygomaticus,
AME superficialis, and AME coronoidea into AME coronoidea. Their AME zy-
gomaticus is our AME articularis externus (caudal portion). These studies pos-
tulate an expansion and diversification of AME superficialis (Lakjer, 1926) or
AME coronoidea (Starck and Barnikol, 1954), whereas we hypothesize an ex-
pansion of AME articularis externus and a rostral displacement of AME zygo-
maticus and superficialis.

Dzerzhinsky (1982, 1995) identified muscle fibers that interconnect aponeuro-
ses zygomatica and paracoronoidea externa as AME superficialis rather than AME
articularis, and Weber (1996) regarded these same fibers as AME zygomatica
(which subsumes AME superficialis). Since these fibers are already incorporated
within a well-defined block of muscle tissue associated with aponeurosis para-
coronoidea externa in chicks of several Anseres (Fig. 8), we think it more likely
that the fibers in question represent AME articularis externus. Starck and Barnikol
(1954:12) included comparable fibers with AME articularis (their Ap. 3 portion)

182

Annals of Carnegie Museum

VOL. 69

Table 2. — Distribution of states of osteological characters associated with the AME complex in se-
lected taxa of Galliformes and Anseriformes, based on the present study.

Taxon

Processus

zygomaticus

Crista

zygomatica^

Fossa musculi
temporales’’

Impressio AME
articu laris'^

Galliformes

Megapodiidae

Small or absent

Absent

Partly medial and
lateral (latter

small)

Small or absent

Cracidae

Small or absent

Absent

Lateral (moderate)

Small or absent

Phasianidae

Small or absent

Absent

Lateral (moderate)

Small or absent

Anseriformes

Anhimidae

Absent

Present

Entirely medial

Small, medial

Anseranatidae

Absent

Present

Entirely medial

Small, lateral

Anatidae

Absent

Present

Typically medial

Small, lateral

® From processus postorbitalis.

Broadly equivalent to the “fossa temporalis” as traditionally defined (see text).
Broadly equivalent to the “fossa subtemporalis” of some authors (see text).

in Buteo buteo (Accipitridae). Under our interpretation it follows that, in Galli-
formes and Anseriformes, the origins of AME superficialis and AME zygomatica
are displaced to the portion of aponeurosis zygomatica rostral to the processus
postorbitalis, and their insertions typically are restricted to the mandibula rostral
to aponeurosis paracoronoidea externa.

Burton (1984) made comparisons between the AME of the Phoeniculidae (Cor-
aciiformes) and Anseriformes with special reference to portions originating on
the processus postorbitalis. He referred to these portions in both orders as the
“postorbital lobe” and suggested that the lobes were homologous and plesiomor-
phous in the two orders. His “postorbital lobe” in Anseriformes equates to our
AME superficialis and AME zygomatica. However, we found that AME zygo-
matica lies caudal to the “postorbital lobe” in Phoeniculus. Caput mediate of
AME coronoidea has not been discussed previously in this paper because it is not
present in the Galliformes and Anseriformes, but it occurs in Coraciiformes and
other avian orders (Richards and Bock, 1973; Burton, 1984). The “postorbital
lobe” of Phoeniculus may include elaborations of AME coronoidea medialis and
AME superficialis.

Interpretations of Characters and Phylogenetic Implications

Alternative Views of Characters, States, and Ordering. — Although a phyloge-
netic analysis incorporating the anatomical information described herein is beyond
the scope of this paper, the comparisons provide a framework for partitioning the
Galliformes and Anseriformes into several broad taxonomic groups (Tables 2-3).
Although most characters differ in the specific groupings suggested, most are
hierarchically consistent with each other; i.e., one character suggests a nested
subdivision of groups implied by another character (Table 3). Also, some char-
acters are redundant; e.g., complementarity of fossa musculorum temporalium
(Table 2) vs. origo AME coronoidea (Table 3).

Livezey (1997a) included one multistate, composite character (Appendix 1:
character 8) that was intended to summarize the anatomical changes described
herein (Table 3), one that emphasized the pattern of apparent changes in the
cranial skeleton (notably orientation of processus postorbitalis and prominence of

2000

ZUSI AND LIVEZEY CRANIUM OF GaLLIFORMES AND AnSERIFORMES

183

O u

5®

PJ 03

< o

■£ o

O

U 03

<1

.|l

O N

W.2

M U

•r Ci<

O S

e y

0 3 e

1 § a

O §“ ^

I 2 S

MOW)

§ o ^

iSI

.1-1 , »-i

o Q O S

n ^

a o

o

n ^

a o

q a

6 S
^ o
o

c« ■%

'd 3 'O 3

2^32^

■M (D ^ (U

O X>

6 B
^ o
o

•d P;d'0 3

2 ^ S 2 S

d gi .d -M D ..-I

g O Xi g O X)

o
P:i

■*^ W3 ■^.

d 3

TD

fi ^

u a
0)

s ^

o -a

U

2

a

o
■-S '2

o

OC

2

a

‘ o
- “2

o

a

11

^ -S

o ^

u

.S

M

bX)

d

C

2

c

o

Sm

0)

d

o

03

^ <

5 ^
.2 ^

.’S

xs a

d

0)

d

«

P 3

o o

d

d

d

s

s

o ^
b£)

0)

«

0)

!2

w

(Z1

'c«

c/)

^ o

'm

M

'S

«3

'S

I/)

O

C

O

C

;3 S

o

o

O

D

D

C3 O

•■s fi

c

2 d S S d

e

C

c

(U

u

<u

<u

Cfl

..d

g cp « d

c«

Xi

IZ)

<

£

<

<

<

’O

”0 "d
d p

^ 2

d d

3

M 2

o

X

O

'S

^1

p o

c«

(U
U

o

fi

a o

4_ e

X 3
1)

^■|

2-3

.— 1 d
2 '2
t«

2 w

d o

2 'C
d o

Cm

w O

t.i

rj^

6 o

HH- pc

184

Annals of Carnegie Museum

VOL. 69

processus zygomaticus). Nomenclatural differences between this earlier character
description and the present study aside, and in spite of the fact that the earlier
character description was based primarily on assessments of prepared skulls, the
four states delimited are consistent with the broad groupings substantiated by the
present study (Tables 2-3). Furthermore, it should be noted that ordering of the
previous coding scheme was not analytically influential, and the character had a
consistency index of 1.0 whether or not the four states were treated a priori as
ordered (Livezey, 1997(2:420). Also, in the final phylogenetic hypotheses, the
ordering implied by the sequence of states was preserved (Table 3), an ordering
that confirmed the general primitive-derived sequence implied by the summary
classification (Galliformes, Anhimidae, Anseranatidae, and Anatidae).

A majority of recent phylogenetic studies indicate that the basal polarities
(primitive states) of these anatomical characters for the Galliformes and Anseri-
formes (Tables 2-3) are sought most reliably among the paleognathous birds (Cra-
craft, 1988; Livezey, 1997a; Groth and Barrowclough, 1999). Although detailed
anatomical descriptions of these features among paleognaths are beyond the scope
of the present study, several published works (Bock, 1963^; Elzanowski, 1987;
Weber, 1996) provide a basis for some preliminary comparative comments. Pres-
ence of a strong processus zygomaticus in close proximity to the os quadratum
is diagnostic for paleognaths, suggesting that the reduction of the processus and
its disassociation from the os quadratum represent a synapomorphy of neognath-
ous birds (including Galliformes and Anseriformes). Similarly, the well-defined
AME articularis found in most neognaths is probably apomorphic relative to its
absence or minimal development in paleognaths. The distinctive pattern of de-
velopment of AME articularis described in this paper apparently is derived relative
to its usual form in other neognaths; the pattern includes enlargement of AME
articularis by expansion rostrally along aponeurosis zygomatica and along the
mandibula, and rostral displacement of AME zygomatica and AME superficialis.

Given these provisional inferences of polarity, the details of interpretation of
the anatomical states described (Table 3) have little or no impact on coding
schemes. All of the alternative interpretations of the plausible evolutionary events
that underlie these anatomical patterns considered here are consistent with the
four-state coding scheme used by Livezey (1997a), and most would permit any
of several alternative coding schemes (e.g., separate, binary characters treating
changes in three to five of the osteological or myological features involved).
Moreover, even the evolutionary interpretation proposed by Dzerzhinsky (1982) —
including the disputed hypothesis of the “sphenotemporal process” — would be
consistent with most or all of these alternative coding schemes, necessitating only
a revision of the accompanying descriptions of character states. This “transpar-
ency” of evolutionary interpretation in the coding of this character complex is
substantiated by the fact that an earlier draft of the description for this character
by Livezey (1997a) was based in large part on the interpretation by Dzerzhinsky
(1982), wherein the allocation of states among taxa was conserved.

Implications for Interordinal Relationships. — Data presented herein regarding
the skeleton and musculature associated with the AME complex (Table 3) and
provisional inferences of polarity based on the literature (Elzanowski, 1987; Dzer-
zhinsky, 1999) support (i.e., confirm without homoplasy) a sister relationship be-
tween the orders Galliformes and Anseriformes, as proposed in earlier anatomical
assessments by Dzerzhinsky (1982, 1995) and Livezey (1997a). This proposal
has been favored by a number of authorities for decades (e.g., Delacour, 1954;

2000

ZUSI AND LiVEZEY CRANIUM OF GaLLIFORMES AND AnSERIFORMES

185

Johnsgard, 1965), and is gaining support from recent morphological (Cracraft,
1988; Livezey, 1997a; Muller and Weber, 1998) and molecular analyses (Cracraft,
1981a; Cracraft and Mindell, 1989; Caspers et aL, 1997; Groth and Barrowclough,
1999). Further insights into this interordinal group, and the morphological trans-
formations which support this complex, may be gained through the continued
study of several fossil taxa recently inferred to be allied with the Anseriformes —
the Diatrymidae (Andors, 1992) and Dromornithidae (Murray and Megirian,
1998).

At least two cranial characters — absence of the ligamentum postorbito-zygo-
maticum, and medial displacement of impressio AME coronoidea — suggest that
megapodes possess a morphotype intermediate to the condition typical of most
Galliformes and that observed in modern Anseriformes. An “intermediate” con-
dition of this complex in the Megapodiidae relative to other Galliformes is con-
sistent with the majority view regarding it as the likely sister group of other
Galliformes (review by Crowe, 1988). These conclusions can be tested by broader
comparative studies within Aves (e.g., Livezey, 1997a), especially those empha-
sizing detailed comparisons of new character complexes and intensive sampling
of Galliformes and Anseriformes.

Acknowledgments

This research was supported by National Science Foundation (NSF) grant BSR-9396249 to Livezey,
NSF grant DEB-9815248 to Livezey and Zusi, and National Museum of Natural History grant
RI85337000 to Zusi. We thank the curators of the following museums for permitting study of their
collections or for lending specimens: U.S. National Museum of Natural History (USNM), American
Museum of Natural History (AMNH), The (British) Natural History Museum (BNMH), Carnegie
Museum of Natural History (CM), and Yale Peabody Museum (YPM). W. Boles donated a chick of
Anseranas from The Australian Museum (Sydney) to USNM in support of this work. Original illus-
trations in Figures 2, 4, 5, 7, and 9 were executed by Susan Escher; others were drawn by Zusi.
Personnel of the Office of Imaging, Printing, and Photographic Services of the USNM assisted with
the digitizing of plates for labelling and reproduction. Finally, we are grateful for the helpful comments
provided by three anonymous reviewers of this paper.

Literature Cited

Adams, D. C., and M. S. Rosenberg. 1998. Partial warps, phylogeny, and ontogeny: a comment on
Fink and Zelditch (1995). Systematic Biology, 47:168-173.

Alberch, P. 1985. Problems with the interpretation of developmental sequences. Systematic Zoology,
34:46-58.

Andors, A. V. 1992. Reappraisal of the Eocene groundbird Diatryma (Aves: Anserimorphae). Pp.
109-125, in Papers in Avian Paleontology Honoring Pierce Brodkorb (K. E. Campbell, Jr., ed.).
Contributions in Science No. 330. Natural History Museum of Los Angeles County, Los Angeles,
California.

Bams, R. A. 1956. On the relation between the attachment of jaw muscles and the surface of the
skull in Podiceps cristatus L., with some notes on the mechanical properties of this part of the
head. Parts I-IV. Verhandlingen der Koninklijke Nederlandse Akademie van Wetenschappen (Se-
ries C), 59:82-101, 248-262.

Barnikol, H. a. 1952. Korrelationen in der Ausgestaltung der Schadelform bei Vogeln. Gegenbaurs
Morphologisches Jahrbuch 92:373-414.

— . 1953. Vergleichend anatomische und taxonomisch phylogenetische Studien am Kopf der

Opisthocomiformes, Musophagidae, Galli, Columbae und Cuculi. Ein Beitrag zum Opisthocomus-
Problem. Zoologische Jahrbucher Systematische, 81:487-526.

. 1954. Zur Morphologic des Nervus trigeminus der Vogel unter besonderer Beriicksichtigung
der Acciptres, Cathartidae, Striges und Anseriformes. Zeitschrift fiir Wissenschaftliche Zoologie,
157:285-332.

Barriel, V., AND P. Tassy. 1993. Characters, observations and steps: comment on Lipscomb’s “Par-
simony, homology and the analysis of multistate characters.” Cladistics, 9:223-232.

186

Annals of Carnegie Museum

VOL. 69

Bas, C. 1954. On the relation between the masticatory muscles and the surface of the skull in Ardea
cinerea L. Part 1. Verhandlingen der Koninklijke Nederlandse Akademie van Wetenschappen
(Series C), 57:678-685.

. 1955. On the relation between the masticatory muscles and the surface of the skull in Ardea

cinerea L. Parts II-III. Verhandlingen der Koninklijke Nederlandse Akademie van Wetenschappen
(Series C), 58:101-108, 109-113.

Baumel, J. J. 1979. Osteologia. Pp. 53-121, in Nomina Anatomica Avium (J. J. Baumel, A. S. King,
A. M. Lucas, J. E. Breazile, and H. E. Evans, eds.). Academic Press, London, United Kingdom.

. 1979. Arthrologia. Pp. 123-174, in Nomina Anatomica Avium (J. J. Baumel, A. S. King,

A. M. Lucas, J. E. Breazile, and H. E. Evans, eds.). Academic Press, London, United Kingdom.

. 1993. Systema cardiovasculare. Pp. 407-475, in Handbook of Avian Anatomy: Nomina

Anatomica Avium, Second edition (J. J. Baumel, A. S. King, J. E. Breazile, H. E. Evans, and J. C.
Vanden Berge, eds.). Nuttall Ornithological Club, Cambridge, Massachusetts.

Baumel, J. J., and R. J. Raikow. 1993. Arthrologia. Pp. 133-188, in Handbook of Avian Anatomy:
Nomina Anatomica Avium, Second edition (J. J. Baumel, A. S. King, J. E. Breazile, H. E. Evans,
and J. C. Vanden Berge, eds.). Nuttall Ornithological Club, Cambridge, Massachusetts.

Baumel, J. J., and L. M. Witmer. 1993. Osteologia. Pp. 45-132, in Handbook of Avian Anatomy:
Nomina Anatomica Avium, Second edition (J. J. Baumel, A. S. King, J. E. Breazile, H. E. Evans,
and J. C. Vanden Berge, eds.). Nuttall Ornithological Club, Cambridge, Massachusetts.

Beddard, F. E. 1898. The Structure and Classification of Birds. Longmans, Green, and Company,
London, England.

Bellairs, a. d’A., and C. R. Jenkin. 1960. The skeleton of birds. Pp. 241-300, in Biology and
Comparative Physiology of Birds, Volume 1 (A. J. Marshall, ed.). Academic Press, New York,
New York.

Berger, A. J. 1960. The musculature. Pp. 301-349, in Biology and Comparative Physiology of Birds,
Volume 1 (A. J. Marshall, ed.). Academic Press, New York, New York.

. 1966. The musculature. Pp. 224-473, in Avian Myology (J. C. George and A. J. Berger,

eds.). Academic Press, New York, New York.

Bhattacharyya, B. N. 1980. The morphology of the jaw and tongue musculature of the common
pigeon, Columba livia, in relation to its feeding habit. Proceedings of the Zoological Society
(Calcutta), 31:95-126.

. 1989. Functional morphology of the jaw muscles of two species of imperial pigeons, Ducula

aenea nicobarica and Ducula badia insignis. Gegenbaurs Morphologisches Jahrbuch, 135:573-
618.

Blanchard, C. E. 1859. Recherches sur les caracteres osteologiques des oiseaux, appliquees a la
classification naturelle de ces animeaux. Annales des Sciences Naturelles (Serie 4), 11:1-141.

Bock, W. J. 1963^. Evolution and phylogeny in morphologically uniform groups. American Natu-
ralist, 92:265-285.

. 1963/?. The cranial evidence for ratite affinities. Pp. 39-54, in Proceedings XIII International

Ornithological Congress, Volume 1 (C. G. Sibley, ed.). American Ornithologists’ Union, Wash-
ington, D.C.

. 1964. Kinetics of the avian skull. Journal of Morphology, 114:1-42.

. 1967. The use of adaptive characters in avian classification. Pp. 61-74, in Proceedings XIV

International Ornithological Congress (D. W. Snow, ed.). Blackwell Scientific Publications, Ox-
ford, United Kingdom.

. 1968. Mechanics of one- and two-joint muscles. American Museum Novitates, 2319:1-45.

. 1977. Adaptation and the comparative method. Pp. 57-82, in Major Patterns in Vertebrate

Evolution (M. K. Hecht, P. Goody, and B. Hecht, eds.). Plenum Press, New York, New York.

. 1979. The synthetic explanation of macroevolutionary change — a reductionist approach. Pp.

20-69, in Models and Methodologies in Evolutionary Theory (J. H. Schwartz and H. B. Rollins,
eds.). Carnegie Museum of Natural History Bulletin Number 13. Carnegie Museum of Natural
History, Pittsburgh, Pennsylvania.

. 1989. The homology concept: its philosophical foundation and practical methodology. Zool-

ogische Beitrage (Neue Folge), 32:327-353.

Bock, W. J., and A. McEvey. 1969. Osteology of Pedionomus torquatus (Aves: Pedionomidae).
Proceedings of the Royal Society of Victoria (New Series), 82:187-232.

Bookstein, F. L, 1994. Can biometrical shape be a homologous character? Pp. 197-227, in The
Hierarchical Basis of Comparative Biology (B. K. Hall, ed.). Academic Press, New York, New
York.

Boyden, a. 1943. Homology and analogy: a century after the definition of “homologue” and “an-
alogue” of Richard Owen. Quarterly Review of Biology, 18:228-241.

2000

ZUSI AND LiVEZEY CRANIUM OF GaLLIFORMES AND AnSERIFORMES

187

. 1947. Homology and analogy: a critical review of the meanings and implications of these

concepts in biology. American Midland Naturalist, 37:548-669.

Brooks, D. R. 1996. Explanations of homoplasy at different levels of biological organization. Pp. 3-
36, in Homoplasy: The Recurrence of Similarity in Evolution (M. J. Sanderson and L. Hufford,

eds.). Academic Press, San Diego, California.

Brower, A. V. Z., and V. Schawaroch. 1996. Three steps of homology assessment. Cladistics, 12:
265-272.

Bryant, H. N. 1989. An evaluation of cladistic and character analyses as hypothetico-deductive
procedures, and the consequences for character weighting. Systematic Zoology, 38:214-227.

Buhler, P. 1981. Functional anatomy of the avian jaw apparatus. Pp. 439-468, in Form and Function
in Birds, Volume 2 (A. S. King and J. McLelland, eds.). Academic Press, London, United King-
dom.

Burggraaf, P. D. 1954. On the correlation between the skull structure and the muscles in the male
Phasianus colchicus L. II. The attachment of the musculus adductor mandibulae externus. Ver-
handlingen der Koninklijke Nederlandse Akademie van Wetenschappen (Series C), 57:292-303.

Burggraaf, P. D., and A. Fuchs. 1954. On the correlation between the skull structure and the muscles
in the male Phasianus colchicus L. I. General introduction. Verhandlingen der Koninklijke Ned-
erlandse Akademie van Wetenschappen (Series C), 57:286-303.

. 1955. On the correlation between the skull structure and the muscles in the male Phasianus

colchicus L. VII. General considerations. Verhandlingen der Koninklijke Nederlandse Akademie
van Wetenschappen (Series C), 58:98-100.

Burton, P. J. K. 191 Aa. Jaw and tongue features in Psittaciformes and other orders with special
reference to the anatomy of the Tooth-billed Pigeon (Didunculus strigirostris). Journal of Zoology
(London), 174:255-276.

. 1974^. Anatomy of head and neck in the Huia {Heteralocha acutirostris) with comparative

notes on other Callaeidae. Bulletin of the British Museum (Natural History), 27 (Zoology): 1-48.

1974c. Feeding and the Feeding Apparatus in Waders: A Study of Anatomy and Adaptations

in the Charadrii. British Museum (Natural History), London, United Kingdom.

1984. Anatomy and evolution of the feeding apparatus in the avian orders Coraciiformes

and Piciformes. Bulletin of the British Museum (Natural History), 47 (Zoology):33 1-443.

Butendieck, E. 1980. Die Benennung des Skeletts beim Truthuhn (Meleagris gallopavo) unter Be-
riicksischtigung der Nomina Anatomica Avium (1979). Ph.D. Dissert., Tierarztiliche Hochschule
Hannover, Hannover, Germany.

Butendieck, E., and H. Wissdorf. 1982. Beitrag zur Benennung der Knochen des Kopfes beim
Truthuhn {Meleagris gallopavo) unter Beriicksichtigung der Nomina Anatomica Avium (1979).
Zoologische Jahrbiicher (Abteilung fiir Anatomie), 107:153-184.

Caspers, G.-J., D. U. de Weerd, J. Wattel, and W W. de Jong. 1997. a-crystallin sequences support
a galliform/anseriform clade. Molecular Phylogeny and Evolution, 7:185-188.

Coddington, j. a. 1994. The roles of homology and convergence in studies of adaptation. Pp. 53-
78 in Phylogenetics and Ecology (R Eggleton and R. 1. Vane-Wright, eds.). Academic Press,
London, United Kingdom.

Cracraft, j. 1968. The lacrimal-ectethmoid bone complex in birds: a single character analysis.
American Midland Naturalist, 80:316-359.

—. 19Hla. Toward a phylogenetic classification of the Recent birds of the world (Class Aves).

Auk, 98:681-714.

-. 1981^. The use of functional and adaptive criteria in phylogenetic systematics. American

Zoologist, 21:21-36.

-. 1982. Phylogenetic relationships and monophyly of loons, grebes, and hesperornithiform

birds, with comments on the early history of birds. Systematic Zoology, 31:35-56.

1986. The origin and early diversification of birds. Paleobiology, 12:383-399.

. 1988. The major clades of birds. Pp. 339-361, in The Phylogeny and Classification of the

Tetrapods, Volume 1: Amphibians, Reptiles, Birds (M. J. Benton, ed.). Clarendon Press, Oxford,
United Kingdom.

Cracraft, J., and D. P. Mindell. 1989. The early history of modern birds: a comparison of molecular
and morphological evidence. Pp. 389-403, in The Hierarchy of Life (B. Fernholm, K. Bremer,
and J. Jornvall, eds.). Exerptica Medica, Amsterdam, The Netherlands.

Crowe, T. M. 1988. Molecules vs morphology in phylogenetics: a non-controversy. Transactions of
the Royal Society of South Africa, 46:317-334.

Davids, J. A. G. 1952. Etude sur les attaches au crane des muscles de la tete et du cou chez Anas
platyrhyncha platyrhyncha L. Parts I-III. Verhandlingen der Koninklijke Nederlandse Akademie
van Wetenschappen (Series C), 55:81-94, 525-533, 534-540.

188

Annals of Carnegie Museum

VOL. 69

de Beer, G. 1937. The Development of the Vertebrate Skull. Clarendon Press, Oxford, United King-
dom.

Delacour, J. 1954. Waterfowl of the World, Volume 1. Country Life, London, United Kingdom.

DePinna, M. C. C. 1991. Concepts and tests of homology in the cladistic paradigm. Cladistics, 7:
367-394.

. 1994. Ontogeny, rooting, and polarity. Pp. 157-172, in Models in Phylogeny Reconstruction

(R. W Scotland, D. J. Siebert, and D. M. Williams, eds.). Clarendon Press, Oxford, United King-
dom.

DE Queiroz, K. 1985. The ontogenetic method for determining character polarity and its relevance
to phylogenetic systematics. Systematic Zoology, 34:280-299.

Doyle, J. J. 1996. Homoplasy connections and disconnections: genes and species, molecules and
morphology. Pp. 37-66, in Homoplasy: The Recurrence of Similarity in Evolution (M. J. San-
derson and L. Hufford, eds.). Academic Press, San Diego, California.

Dullemeijer, P. 1951. The correlation between muscle system and skull structure in Phalacrocorax
carho sinensis (Shaw & Nodder): I-Ill. Proceedings of Koninklijke Nederlandse Akademie van
Wetenschappen (Series C), 54:247-259; 400-404; 533-536.

. 1952. The correlation between muscle system and skull structure in Phalacrocorax carbo

sinensis (Shaw & Nodder): IV. Proceedings of Koninklijke Nederlandse Akademie van Weten-
schappen (Series C), 55:95-102.

Dzerzhinsky, R. Ya. 1980. [Adaptive transformation of the maxillary apparatus in the evolution of
gallinaceous birds.] Pp. 148-158, in Morphological Aspects of Evolution (V E. Sokolov and N. S.
Lebedkina, eds.). Moscow State University, Moscow, USSR. [In Russian.]

. 1982. Adaptive features in the structure of maxillary system in some Anseriformes and

probable ways of evolution of the order. Zoologichesky Zhurnal, 56:1030-1041. [Russian with
English summary.]

. 1995. Evidence for the common ancestry of Galliformes and Anseriformes. Courier For-

schungsinstitut Senckenberg, 181:325-336.

. 1999. Implications of the cranial morphology of paleognaths for avian evolution. Pp. 267-

274, in Avian Paleontology at the Close of the 20th Century: Proceedings of the 4th International
Meeting of the Society of Avian Paleontology and Evolution, Washington, D. C., 4-7 June 1996
(S. L. Olson, ed.). Smithsonian Contributions to Paleobiology, 89:1-344.

Dzerzhinsky, R. Ya., and I. N. Belokurova. 1972. [On the comparative anatomy of the maxillary
musculature of birds. Jaw muscles of the Capercaillie {Tetrao urogallus).} Zoologichesky Zhurnal,
51:555-564. [Russian with English summary.]

Dzerzhinsky, R. Ya., and E. G. Potanova. 1974. [The system of tendinous formations as an object
of comparative myology of the maxillary apparatus of birds.] Zoologichesky Zhurnal, 53:1341—
1351. [Russian with English summary.]

Edgeworth, E H. 1907. The development of the head-muscles in Gallus domesticus, and the mor-
phology of head-muscles in the Sauropsida. Quarterly Journal of Microscopical Science (Series
2), 51:511-556.

Elzanowski, a. 1987. Cranial and eyelid muscles and ligaments of the tinamous (Aves: Tinamifor-
mes). Zoologische Jahrbiicher (Abteilung fiir Anatomic), 116:63-118.

Ericson, P. G. P. 1996. The skeletal evidence for a sister-group relationship of anseriform and
galliform birds — a critical evaluation. Journal of Avian Biology, 27:195—202.

. 1997. Systematic relationships of the palaeogene family Presbyornithidae (Aves: Anserifor-
mes). Zoological Journal of the Linnean Society, 121:429-483.

Eyton, T. C. 1867. Osteologia Avium; Or, a Sketch of the Osteology of Birds. R. Hobson, Salop,
England.

Feduccia, a. 1977. Hypothetical stages in the evolution of modern ducks and flamingos. Journal of
Theoretical Biology, 67:715-721.

. 1978. Presbyornis and the evolution of ducks and flamingos. American Scientist, 66:298-

304.

. 1980a. Evolution von Enten und Flamingos. Pp. 1243-1248, in Acta XVII Congressus

Internationalis Omithologici, Volume 2 (R. Nohring, ed.). Deutschen Omithologen-Gesellschaft,
Berlin, Germany.

. 1980/:?. The Age of Birds. Harvard University Press, Cambridge, Massachusetts.

. 1994. Tertiary bird history: notes and comments. Pp. 178-189, in Major Features of Verte-
brate Evolution (D. R. Prothero and R. M. Schoch, eds.). Paleontological Society, Knoxville, Ten-
nessee.

. 1995. Explosive evolution in Tertiary birds and mammals. Science, 267:637-638.

— . 1996. The Origin and Evolution of Birds. Yale University Press, New Haven, Connecticut.

2000

ZUSI AND LIVEZEY CRANIUM OF GaLLIFORMES AND AnSERIFORMES

189

Fiedler, W. 1951. Beitrage zur Morphologic der Kiefermuskulatur der Oscines. Zoologische Jahr-
biicher (Abteilung fiir Anatomic und Ontogenie Tiere), 71:235-288.

Fink, W. L., and M. L. Zelditch. 1995. Phylogenetic analysis of ontogenetic shape transformations:
a reassessment of the piranha genus Pygocentrus (Teleostei). Systematic Biology, 44:343-360.

Fisher, H. L, and D. C. Goodman. 1955. The Myology of the Whooping Crane, Grus americana.
University of Illinois Press, Urbana, Illinois.

Fitch, W. M. 1970. Distinguishing homologous from analogous proteins. Systematic Zoology, 19:
99-113.

Fuchs, A. 1954. On the correlation between the skull structure and the muscles in the male Phasianus
colchicus L. IIIA. The attachment of the musculus adductor mandibulae posterior and the mus-
culus adductor mandibulae internus. Verhandlingen der Koninklijke Nederlandse Akademie van
Wetenschappen (Series C), 57:454-470.

— 1955. On the correlation between the skull structure and the muscles in the male Phasianus

colchicus L. VI. Some remarks on a number of ligaments and other connective tissue connections.
Verhandlingen der Koninklijke Nederlandse Akademie van Wetenschappen (Series C), 58:114-
120.

Fujioka, T. 1963. On the origins and insertions of the muscles of the head and neck in fowl. Part I.
Muscles of the head. Japanese Journal of Veterinary Science, 25:207-226.

Furbringer, M. 1888. Untersuchungen zur Morphologic und Systematik der Vogel, zugleich ein
Beitrag zur Anatomic der Stiitz- und Bewegungsorgane, 2 Volumes. T. J. Van Holkema, Amster-
dam, Holland.

Gadow, H. 1892. On the classification of birds. Proceedings of the Zoological Society of London,
1892:229-256.

. 1893. Vogel. II. Systematischer Theil. In Klassen und Ordnungen des Their-Reichs (H. G.
Bronn, ed.). C. F. Winter, Leipzig, Germany.

Gadow, H., and E. Selenka. 1891. Vogel. I. Anatomischer Theil. In Klassen und Ordnungen des
Their-Reichs (H. G. Bronn, ed.). C. F. Winter, Leipzig, Germany.

Gans, C., and W. J. Bock. 1965. The functional significance of muscle architecture — a theoretical
analysis. Review of Anatomy, Embryology, and Cell Biology, 38:115-142.

Goodman, D. C., and H. I. Fisher. 1962. Functional Anatomy of the Feeding Apparatus in Waterfowl.
Southern Illinois University Press, Carbondale, Illinois.

Goodwin, B. 1994. Homology, development, and heredity. Pp. 229-247, in Homology: The Hierar-
chical Basis of Comparative Biology (B. K. Hall, ed.). Academic Press, San Diego, California.

Greene, H. W. 1994. Homology and behavioral repertoires. Pp. 369-391, in Homology: The Hier-
archical Basis of Comparative Biology (B. K. Hall, ed.). Academic Press, New York, New York.

Groth, j. G., and G. F. Barrowclough. 1999. Basal divergences in birds and the phylogenetic utility
of the nuclear RAG-1 gene. Molecular Phylogenetics and Evolution, 12:115-123.

Hall, B. K. 1994. Introduction. Pp. 1-19, in Homology: The Hierarchical Basis of Comparative
Biology (B. K. Hall, ed.). Academic Press, New York, New York.

. 1995. Homology and embryonic development. Pp. 1-37, in Evolutionary Biology, Volume

28 (M. K. Hecht, R. J. MacIntyre, and M. T Clegg, eds.). Plenum Press, New York, New York.

Hanken, j., and B. K. Hall. 1993. Mechanisms of skull diversity and evolution. Pp. 1-36, in The
Skull, Volume 3: Functional and Evolutionary Mechanisms (J. Hanken and B. K. Hall, eds.).
University of Chicago Press, Chicago, Illinois.

Haszprunar, G. 1992. The types of homology and their significance for evolutionary biology and
phylogenetics. Journal of Evolutionary Biology, 5:13-24.

. 1998. Parsimony analysis as a specific kind of homology estimation and the implication for

character weighting. Molecular Phylogenetics and Evolution, 9:333-339.

Hawkins, J. A., C. E. Hughes, and R. W. Scotland. 1997. Primary homology assessment, characters
and character states. Cladistics, 13:275-283.

Hennig, W. 1966. Phylogenetic Systematics. University of Illinois Press, Urbana, Illinois.

Hillis, D. M. 1994. Homology in molecular biology. Pp. 339-368, in Homology: The Hierarchical
Basis of Comparative Biology (B. K. Hall, ed.). Academic Press, New York, New York.

Hofer, H. 1945. Untersuchungen fiber den Bau des VogelschMels, besonders fiber den der Spechte
und SteiBhfihner. Zoologische Jahrbficher (Abteilung fiir Anatomie), 69:1-158.

= — — — 1950. Zur Morphologie der Kiefermuskulatur der Vogel. Zoologische Jahrbficher (Abteilung
fiir Anatomie), 70:427-556.

. 1955. Neuere Untersuchungen zur Kopfmorphologie der Vogel. Pp. 104-137, in Acta XI

Congressus Internationalis Ornithologici (A. Portmann and E. Sutter, eds.). Birkhauser Verlag,
Basel, Switzerland.

190

Annals of Carnegie Museum

VOL. 69

Hogg, D. A. 1978. The articulations of the neurocranium in the postnatal skeleton of the domestic
fowl (Gallus gallus domesticus). Journal of Anatomy (London), 127:53-63.

del Hoyo, J., a. Elliott, and J. Sargatal. 1994. Handbook of the Birds of the World, Volume 2.
Lynx Edicions, Barcelona, Spain.

ICVGAN (International Committee on Veterinary Gross Anatomical Nomenclature). 1983.
Nomina Anatomica Veterinaria, Third edition. Ithaca, New York.

Inglis, W. G. 1966. The observational basis of homology. Systematic Zoology, 15:219-228.

Jager, O. 1990. Die Kopfligamente der Anseriformes (Wolters, 1982). Ein Beitrag zum Grundplan
der Entenvogel. Acta Biologica Benrodis (Diisseldorf), 2:209-223.

JoHNSGARD, P. J. 1965. Handbook of Waterfowl Behavior. Constable, London, United Kingdom.

Johnson, R. 1984. The cranial and cervical osteology of the European Oystercatcher Haematopus
ostralegus L. Journal of Morphology, 182:227-244.

Jollie, M. T. 1957. The head skeleton of the chicken and remarks on the anatomy of this region in
other birds. Journal of Morphology, 100:389-436.

Kluge, A. G., and R. E. Strauss. 1985. Ontogeny and systematics. Annual Review of Ecology and
Systematics, 16:247-268.

Komarek, V. 1979. Anatomia Avium Domesticarum, Volume 1. Priroda, Bratislava, Czechoslovakia.

Komarek, V, L. Malinovsky, and L. Lemez. 1982. Anatomia Avium Domesticarum, Volumes 2
and 3. Priroda, Bratislava, Czechoslovakia.

Kraus, F. 1987. An empirical evaluation of the use of the ontogeny polarization criterion in phylo-
genetic inference. Systematic Zoology, 37:106-141.

Kripp, D. von. 1933. Der Oberschnabelmechanismus der Vogel. (Nach den Methoden der graphischen
Statik bearbeitet). Gegenbaurs Morphologisches Jahrbuch, 71:469-544.

Lakjer, T. 1926. Studien iiber die Trigeminus-versorgte Kaumuskulatur der Sauropsiden. C. A. Re-
itzel, Kopenhagen, Denmark.

Lauder, G. V. 1994. Homology, form, and function. Pp. 151-196, in Homology: The Hierarchical
Basis of Comparative Biology (B. K. Hall, ed.). Academic Press, New York, New York.

Lebedinsky, N. G. 1921. Zur Syndesmologie der Vogel. Anatomischer Anzeiger, 54:8-15.

Lipscomb, D. L. 1992. Parsimony, homology and the analysis of multistate characters. Cladistics, 8:
45-65.

Livezey, B. C. 1986. A phylogenetic analysis of Recent anseriform genera using morphological
characters. Auk, 103:737-754.

. 1989. Phylogenetic relationships of several subfossil Anseriformes of New Zealand. Uni-
versity of Kansas Museum of Natural History, Occasional Paper, 128:1-25.

. 1991. A phylogenetic analysis and classification of Recent dabbling ducks (Tribe Anatini)

based on comparative morphology. Auk, 108:471-508.

. 1995a. Phylogeny and evolutionary ecology of modern seaducks (Anatidae: Mergini). Con-
dor, 97:233-255.

. 1995/7. A phylogenetic analysis of the whistling and White-backed ducks (Anatidae: Den-

drocygninae) using morphological characters. Annals of Carnegie Museum, 64:65-97.

. 1995c. Phylogeny and comparative ecology of stiff-tailed ducks (Anatidae: Oxyurini). Wilson

Bulletin, 107:214-234.

. 1996c!. A phylogenetic analysis of modern pochards (Anatidae: Aythyini). Auk, 113:74-93.

. 1996/7. A phylogenetic reassessment of the tadornine-anatine divergence (Aves: Anserifor-
mes: Anatidae). Annals of Carnegie Museum, 65:27-88.

. 1996c. A phylogenetic analysis of the geese and swans (Anseriformes: Anserinae), including

selected fossil species. Systematic Biology, 45:415-450.

. 1997«. A phylogenetic analysis of basal Anseriformes, the fossil Preshyornis, and the inter-
ordinal relationships of waterfowl. Zoological Journal of the Linnean Society, 121:361-428.

. 1991b. An annotated phylogenetic classification of waterfowl (Aves: Anseriformes), includ-
ing selected fossil species. Annals of Carnegie Museum, 67:457-496.

. 1997c. A phylogenetic analysis of modern shelducks and sheldgeese (Anatidae, Tadomini).

Ibis, 139:51-66.

. 1998. Erratum. Zoological Journal of the Linnean Society, 124:397-398.

Mabee, P. M. 1989. An empirical rejection of the ontogenetic polarity criterion. Cladistics, 5:409-
416.

. 1993. Phylogenetic interpretation of ontogenetic change: sorting out the actual and artefactual

in an empirical case study of cetrarchid fishes. Zoological Journal of the Linnean Society, 107:
175-291.

McKitrick, M. C. 1994. On homology and the ontological relationship of parts. Systematic Biology,
43:1-10.

2000

ZUSI AND LiVEZEY CRANIUM OF GaLLIFORMES AND ANSERIFORMES

191

McClearn, D., and D. M. Noden. 1988. Ontogeny of architectural complexity in embryonic quail
visceral arch muscles. American Journal of Anatomy, 183:277-293.

McShea, D. W. 1996. Complexity and homoplasy. Pp. 207-225, in Homoplasy: The Recurrence of
Similarity in Evolution (M. J. Sanderson and L. Hufford, eds.). Academic Press, San Diego,
California.

Meier, R. 1997. A test and review of the empirical performance of the ontogenetic criterion. System-
atic Biology, 46:699-721.

Merz, R. L. 1963. Jaw musculature of the Mourning and White-winged doves. Publication of Uni-
versity of Kansas Museum of Natural History, 12:521-551.

Mindell, D. P. 1991. Similarity and congruence as criteria for molecular homology. Molecular Bi-
ology and Evolution, 8:897-900.

Moller, W. 1932. Biologisch-anatomische Studien am Schadel von Ara macao. Gegenbaurs Mor-

phologisches Jahrbuch, 70:305-342.

Morioka, H. 1974. Jaw musculature of the swifts (Aves, Apodidae). Bulletin of the National Museum
of Japan (Series A), 17:1-16.

Muller, W., and E. Weber. 1998. Re-discovery of a supposedly lost muscle in palaeognathous birds
and its phylogenetic implications. Mitteilungen aus dem Museum fiir Naturkunde in Berlin (Zool-
ogische Reihe), 74:11-18.

Murray, P. R., and D. Megirian. 1998. The skull of dromornithid birds: anatomical evidence for
their relationship to Anseriformes. Records of the South Australian Museum, 31:51-97.

Nelson, G. 1978. Ontogeny, phylogeny, paleontology, and the biogenetic law. Systematic Zoology,
27:324-345.

. 1994. Homology and systematics. Pp. 101-149, in Homology: The Hierarchical Basis of

Comparative Biology (B. K. Hall, ed.). Academic Press, New York, New York.

Olson, S. L., and A. Feduccia. 1980«. Presbyornis and the origin of the Anseriformes (Aves:
Charadriomorphae). Smithsonian Contributions to Zoology, 323:1-24.

. 1980^7. Relationships and evolution of flamingos (Aves: Phoenicopteridae). Smithsonian Con-
tributions to Zoology, 316:1-73.

Owen, R. 1843. Lectures on Comparative Anatomy and Physiology of the Invertebrate Animals,
Delivered at the Royal College of Surgeons in 1843. Longman, Brown, Green, and Longman,
London, England.

Panchen, a. L. 1994. Richard Owen and the concept of homology. Pp. 21-62, in Homology: The
Hierarchical Basis of Comparative Biology (B. K. Hall, ed.). Academic Press, New York, New York.

Patterson, C. 1982. Morphological characters and homology. Pp. 21-74, in Problems of Phylogenetic
Reconstruction (K. A. Joysey and A. E. Friday, eds.). Academic Press, New York, New York.

. 1988. Homology in classical and molecular biology. Molecular Biology and Evolution, 5:

603-625.

Peters, J. L. 1934. Checklist of Birds of the World, Volume 2. Harvard University Press, Cambridge,
Massachusetts.

Pogue, M. G., and M. F. Mickevich. 1990. Character definitions and character state delineations: the
bite noire of phylogenetic inference. Cladistics, 6:319-361.

Remane, A. 1952. Die Grundlagen des natiirlichen Systems, der vergleichendee Anatomie und der
Phylogenetik. Geest and Portig, Leipzig, Germany.

. 1956. Die Grundlagen des natiirlichen Systems, der vergleichenden Anatomie und der Phy-
logenetik, Second edition. Akademische Verlagsgesellschaft, Leipzig, Germany.

Richards, L. R, and W J. Bock. 1973. Functional anatomy and adaptive evolution of the feeding
apparatus in the Hawaiian honeycreeper genus Loxops (Drepanididae). Ornithological Monograph
15. American Ornithologists’ Union, Washington, D. C.

Richards, S. A. 1968. Anatomy of the veins of the head in the domestic fowl. Journal of Zoology
(London), 154:223-234.

Rieppel, O. 1980. Homology, a deductive concept? Zeitschrift fiir Zoologische Systematiks und Ev-
olutionforschung, 18:315-319.

— . 1992. Homology and logical fallacy. Journal of Evolutionary Biology, 5:701-715.

— . 1994. Homology, topology, and typology: the history of modern debates. Pp. 63-100, in

Homology: The Hierarchical Basis of Comparative Biology (B. K. Hall, ed.). Academic Press,
New York, New York.

Rohlf, F. j. 1998. On applications of geometric morphometries to studies of ontogeny and phylogeny.
Systematic Biology, 47:147-158.

Roth, V. L. 1984. On homology. Biological Journal of the Linnean Society, 22:13-29.

. 1988. The biological basis of homology. Pp. 1-26, in Ontogeny and Systematics (C. J.

Humphries, ed.). Columbia University Press, New York, New York.

192

Annals of Carnegie Museum

VOL. 69

. 1991. Homology and hierarchies; problems solved and unresolved. Journal of Evolutionary

Biology, 4:167-194.

Sanderson, M. J., and L. Hufford. 1996. Homoplasy and the evolutionary process: an afterword.
Pp. 327-330, in Homoplasy: The Recurrence of Similarity in Evolution (M. J. Sanderson and L.
Hufford, eds.). Academic Press, San Diego, California.

Sattler, R. 1984. Homology — a continuing challenge. Systematic Botany, 9:382-394.

SCHI0LER, E. L. 1926. Danmarks Fugle, Volume 2. Nordisk Forlag, Kobenhavn, Denmark.

Schulin, R. 1987. Hind limb myology and phylogenetic relationships of the Australian Magpie Goose
Anseranas semipalmata (Latham). Zoologische Jahrbiicher (Abteilung fiir Anatomic und Onto-
genie der Tiere), 116:217-243.

Seebohm, H. 1888. An attempt to diagnose the suborders of the great Gallino-Gralline group of birds,
by the aid of osteological characters alone. Ibis, 30:415-435.

. 1889. An attempt to diagnose the suborders of the ancient Ardeino-Anserine assemblage of

birds by the aid of osteological characters alone. Ibis, 31:92-104.

. 1890. The Classification of Birds: An Attempt to Classify the Subclasses, Orders, Suborders,

and Some of the Families of Existing Birds. R. H. Porter, London, England.

. 1895. The Classification of Birds: An Attempt to Classify the Subclasses, Orders, Suborders,

and Some of the Families of Existing Birds. Supplement. R. H. Porter, London, England.

Shufeldt, R. W. 1901. On the osteology and phylogenetic position of the screamers (Palamedea:
Chauna). American Naturalist, 35:455-461.

. 1909. Osteology of birds. New York State Museum Bulletin, 130:1-381.

. 1914. Contributions to the study of the “tree-ducks” of the genus Dendrocygna. Zoologische

Jahrbiicher (Systematik, Geographie und Biologie Tiere), 38:1-70.

Sibley, C. G., and B. L. Monroe, Jr. 1990. Distribution and Taxonomy of Birds of the World. Yale
University Press, New Haven, Connecticut.

Simpson, G. G. 1959. Anatomy and morphology: classification and evolution, 1859 and 1959. Pro-
ceedings of the American Philosophical Society, 103:286-306.

SiMONETTA, A. M. 1960«. On the mechanical implications of the avian skull and their bearing on the
evolution and classification of birds. Quarterly Review of Biology, 35:206-220.

. 1960/?. Observazioni sull meccanica del cranio degli uccelli e sulla sua influenza

sull’evoluzione e classificazione degli uccelli (nota preliminare). Monitore Zoologico Italiano, 68:
1-14.

. 1963. Cinesi e morfologia del cranio negli uccelli non passeriformi. Studio su varie tendenze

evolutive. Parte la. Archivio Zoologico Italiano (Torino), 48:53-135.

. 1968. Cinesi e morfologia del cranio negli uccelli non passeriformi. Studio su varie tendenze

evolutive. Parte 11. Striges, Caprimulgiformes ed Apodiformes. Archivio Zoologico Italiano (To-
rino), 52:1-36.

Sluys, R. 1996. The notion of homology in current comparative biology. Journal of Zoological
Systematics and Evolutionary Research, 34:145-152.

Starck, D., and a. Barnikol. 1954. Beitrage zur Morphologic der Trigeminus-muskulatur der Vogel
(besonders der Acciptres, Cathartidae, Striges und Anseres). Gegenbaurs Morphologisches Jahr-
buch, 94:1-64.

Stevens, P. F. 1984. Homology and phylogeny: morphology and systematics. Systematic Botany, 9:
395-409.

Swiderski, D. L., M. L. Zelditch, and W. L. Fink. 1998. Why morphometries is not special; coding
quantitative data for phylogenetic analysis. Systematic Biology, 47:508-519.

Vanden Berge, j. C. 1979. Myologia. Pp. 175-219, in Nomina Anatomica Avium (J. J. Baumel,
A. S. King, A. M. Lucas, J. E. Breazile, and H. E. Evans, eds.). Academic Press, London, United
Kingdom.

Vanden Berge, J. C., and G. A. Zweers. 1993. Myologia. Pp. 189-250, in Handbook of Avian
Anatomy: Nomina Anatomica Avium, Second edition (J. J. Baumel, A. S. King, J. E. Breazile, H. E.
Evans, and J. C. Vanden Berge, eds.). Nuttall Ornithological Club, Cambridge, Massachusetts.

Van der Klaauw, C. J. 1963. Projections, deepenings and undulations of the surface of the skull in
relation to the attachment of muscles. Verhandlingen der Koninklijke Nederlandse Akademie van
Wetenschappen (Series 2), 55:1-247.

Van Gennip, E. M. S. J. 1986. The osteology, arthrology and myology of the jaw apparatus of the
pigeon (Columha livia L.). Netherlands Journal of Zoology, 36:1-46.

Van Valen, L. 1982. Homology and causes. Journal of Morphology, 173:305-312.

Verheyen, R. 1956. Contribution a I’anatomie et a la systematique des Galliformes. Bulletin de
Institut Royal des Sciences Naturelles de Belgique, 32(42): 1-24.

Vickers-Rich, R, P. Trusler, M. J. Rowley, A. Cooper, G. K. Chambers, W. J. Bock, P. R. Millener,

2000

ZUSI AND LiVEZEY CRANIUM OF GALLIFORMES AND AnSERIFORMES

193

T. H. Worthy, and J. C. Yaldwyn. 1995. Morphology, myology, collagen and DNA of a mum-
mified upland moa, Megalapteryx didinus (Aves: Dinornithiformes) from New Zealand. Tuhinga:
Records of the Museum of New Zealand Te Papa Tongarewa, 4:1-26.

Wagele, J.-W. 1995. On the information content of characters in comparative moiphology and mo-
lecular systematics. Journal of Zoological Systematics and Evolutionary Research, 33:42-47.

Wagner, G. P. 1989fl. The origin of morphological characters and the biological basis of homology.
Evolution, 43:1157-1171.

. 1989^7. The biological homology concept. Annual Review of Ecology and Systematics, 20:

51-69.

. 1994. Homology and the mechanisms of development. Pp. 273—299, in Homology: The

Hierarchical Basis of Comparative Biology (B. K. Hall, ed.). Academic Press, San Diego, Cali-
fornia.

Webb, M. 1957. The ontogeny of the cranial bones, cranial peripheral and cranial parasympathetic
nerves, together with a study of the visceral muscles of Struthio. Acta Zoologica, 38:81-203.

Weber, E. 1993. Zur Evolution basicranialer Gelenke bei Vogeln, insbesondere bei Hiihner- und
Entenvbgeln (Galloanseres). Zeitschrift fiir Zoologische Systematiks und Evolutionforschung, 31:
300-317.

. 1996. Das Skelet-Muskel-System des Kieferapparates von Aepypodius arfakianus (Salvador!,

1877) (Aves, Megapodiidae). Courier Forschungsinstitut Senckenberg, 189:1-132.

Wenzel, J. W. 1992. Behavioral homology and phylogeny. Annual Review of Ecology and System-
atics, 23:361-381.

Wetmore, a. 1951. A revised classification for the birds of the world. Smithsonian Miscellaneous
Collections, 117:1-22.

Wheeler, Q. D. 1990. Ontogeny and character phylogeny. Cladistics, 6:225-268.

Wiley, E. O. 1981. Phylogenetics: The Theory and Practice of Phylogenetic Systematics. John Wiley
and Sons, New York, New York.

Williams, D. M., R. W. Scotland, and S. Blackmore. 1990. Is there a direct ontogenetic criterion
in systematics? Biological Journal of the Linnean Society, 39:99-108.

Young, B. A. 1993. On the necessity of an archetypal concept in morphology: with special reference
to the concepts of “structure” and “homology.” Biology and Philosophy, 8:225-248.

Yudin, K. A. 1961. On the mechanism of the jaw in Charadriiformes, Procellariiformes, and some
other birds. Trudy Zoologichesky Institut (Leningrad), 29:257-302.

— . 1965. [The phylogeny and classification of the Charadriiformes.] Fauna of the USSR: Birds —

Volume II (1), Part 1. Zoological Institute (New Series), No. 91. Academy of Science, Leningrad,
USSR.

Zelditch, M. L., and W. L. Fink. 1998. Partial warps, phylogeny and ontogeny: a reply to Adams
and Rosenberg. Systematic Biology, 47:345-348.

Zelditch, M. L., W. L. Fink, and D. L. Swiderski. 1995. Morphometries, homology, and phyloge-
netics: quantified characters as synapomorphies. Systematic Biology, 44:179-189.

Zelditch, M. L., W. L. Fink, D. L. Swiderski, and B. L. Lundrigan. 1998. On applications of
geometric morphometries to studies of ontogeny and phylogeny: a reply to Rohlf. Systematic
Biology, 47:159-167.

Zusi, R. L. 1962. Structural adaptations of the head and neck in the Black Skimmer, Rynchops nigra,
L. Publication of the Nuttall Ornithological Club, 3:1-101.

1967. The role of the depressor mandibulae muscle in kinesis of the avian skull. Proceedings

of the United States National Museum, 123:1-28.

-. 1975. An interpretation of skull structure in penguins. Pp. 59-84, in The Biology of Penguins

(B. Stonehouse, ed.). Macmillan Press, London, United Kingdom.

. 1993. Patterns of diversity in the avian skull. Pp. 391-437, in The Skull, Volume 2: Patterns

of Structural and Systematic Diversity (J. Hanken and B. K. Hall, eds.). University of Chicago
Press, Chicago, Illinois.

Zusi, R. L., AND G. D. Bentz. 1984. Myology of the Purple-throated Carib {Eulampis jugularis) and
other hummingbirds (Aves: Trochilidae). Smithsonian Contributions to Zoology, 385:1-70.

Zusi, R. L., and R. W. Storer. 1969. Osteology and myology of the head and neck of the pied-
billed grebes (Podilymbus). Miscellaneous Publications of the University of Michigan Museum
of Zoology, 139:1-49.

Zweers, G. a. 1974. Structure, movement, and myography of the feeding apparatus of the Mallard
(Anas platyrhynchos L.): a study in functional anatomy. Netherlands Journal of Science, 24:323-
467.

•f- I«' A

mkrkM '•■ ‘'~>-

.^■■■,’ -.•>.' A) ' jJiTmyJi j-*. r

>. ' ’ -Mi"3 ' ' 'fi 1«'(' I

'Oi>/ . ’ f* «

• .'• i ••! >.• .»

• •^■;?' tr', "•,4'S.V A^j ■

f>* .'^ruajwpiwyiwiiMBp '* -

f

.. .•> ’Vh’ >^''1 ♦ •*! ' %f;

"■ .-w- A r-<X4:WMA ^ ^

1 *V|;XC4*4^-v,ii»i."-J . !■< -»( '\i^.\'*\ '. ■-;jv - ^ , . ,1,

'■■■ ; ^ - - .' * ., ■ ' i - ■ Vi' .

' ■ 'tt> , . vvn IV

"■ ■ ^ ' ‘'■■- ' »■• H -' ■

* ' ^ j*‘- 'f ■■'

, ■ ■ ■: Tk.. . • ' ■• ; •

(, r /;*l ' ‘M/'K *''■

y»

y* ■■fl(ljUM|M» ■ ' .*

ft t ... * ^ ^ J ' ■•

,. sf.ir . W-’ •

■' ■ ■

>«il ^

■■■'' ■ V -'v.'iV’ ■ -» ?*■ -.T *

“fj-: ,

.' ..'. S .-»•' ■. ■ . „•’■

- ■ '-f:;* ■' . . ?

‘ . ' ♦ .

u .■ { j-: n • f

.*.^' .r€^'

. >4 . * '■f*

<‘ ■‘..''.A '

I! ^ ’ %r : - , * iw

'4 u^*r ;, V .' . , fr ‘ .■'■•

V v.

► AJ.'- »•,-*,•■*<

■• ?/ • .-/.•K ' •'X A'^U' ■'•

uh" .wiiypiWM ^fT‘r

i»-vr% i''^IWilSIIB^

V— . xf .

"^♦‘V ''•. »»M/^)ljlyr? ■> 5?IW^ »sj 4<;J J‘,.-n u»s.^S«r'' ^

' * > jw • I -u» 1 4 . ■!

^ " ■ v.H

•'r kAK\. .:■ b .V i et » hj^ ’ \| "‘I

I . 'V' } y. - f\K} Uf>\r

’*•■ ‘ *? * .’ 1 ’ ‘ */ :-r ■ y .' -I J^:> J r> V < s' » • ■ cv^Rna

’*'1WE3P<; *' ' - - • ■ * * ••% 2 V

■'***■-■ ' •' ,, v.^.^' Jn-- i^tlT . .'.y • . .'^ip

,., j'r' ■ .v.'i’i'-t;':." ‘ ’.^ '■

/ • . ■ \ •• , ^. :i.' ^f,■^H^r^v>vVa ■;> 'V» is ' vpX

.,1 I ' V; ^ .■ M(r‘ •''■ r ‘-'^ *'- ■ J VV?-^ OjrtI '* '

'-ft ^

v*- • r

' - •“

'i 7'* ■ ■■'^■'- .1' • r, ,

'4tB5>S ■■ '

' j —4: , •=»

r V M

ANNALS OF CARNEGIE MUSEUM

VoL. 69, Number 3, Pp. 195-208

9 August 2000

A NEW SPECIES OF CARPOCRISTES (MAMMALIA: PRIMATOMORPHA)
FROM THE MIDDLE TIFFANIAN OF THE BISON BASIN, WYOMING,
WITH NOTES ON CARPOLESTID PHYLOGENY

K. Christopher Beard

Associate Curator, Section of Vertebrate Paleontology
beardc @ carnegiemuseums.org

Abstract

A new species of Carpolestidae, Carpocristes rose!, is described from a middle Tiffanian locality
in the Bison Basin, Fremont County, Wyoming. A phylogenetic analysis of all known carpolestid
species identifies C. rosei as the most basal member of Carpocristes, a genus known from Tiffanian
localities in North America and the Bumbanian Wutu Formation, Shandong Province, People’s Re-
public of China. Available biostratigraphic and phylogenetic data suggest that Carpocristes originated
in North America.

As reconstructed here on the basis of dental characters, phylogenetic relationships among carpoles-
tids as a whole are highly compatible with the stratigraphic distributions of individual species. Nev-
ertheless, stratigraphic disjunctions between inferred sister taxa imply ghost lineages that lasted roughly
5.6 Ma (for early members of the chronolestine clade), 2.3 Ma (for early members of the lineage
culminating in Elphidotarsius wightoni), and 1.2 Ma (for early members of Carpocristes), respectively.

Key Words: Paleocene, Carpolestidae, Phylogeny, Paleobiogeography

Introduction

The Carpolestidae are a family of plesiadapoid primatomorphs characterized
by pronounced specializations of the posterior premolars (Rose, 1975). All car-
polestid taxa other than Chronolestes simul, for which the monotypic subfamily
Chronolestinae has been erected (Beard and Wang, 1995), possess a bladelike or
“plagiaulacoid” P4 that occluded with equally specialized, poly cuspate (Bik^
nevicius, 1986). With some notable exceptions (Fox, 1984), these highly derived
posterior premolars became progressively elaborate through time (Rose, 1975,
1977), and their morphology has formed the principal basis for reconstructing
phylogenetic relationships within the group (Rose, 1975; Krause, 1978; Fox,
1984; Beard and Wang, 1995; Bloch and Gingerich, 1998).

Carpolestids are fairly common components of North American Paleocene
mammal faunas, where they range in age from late Torrejonian (To3) to late
Clarkforkian (Cf3) (Archibald et al., 1987). Recently, the first Asian carpolestids
were described from the Wutu Formation in the Wutu Basin, Shandong Province,
People’s Republic of China (Beard and Wang, 1995). One of these Asian carpo-
lestids, Carpocristes oriens, is morphologically very similar to North American
Carpocristes hobackensis and Carpocristes cygneus, and all three of these species
are thought to form a clade within Carpolestidae. This particular clade is inter-
esting not only because of its widespread geographic distribution, but also because
it appears to document a relatively rare example of the successful invasion of
Asia by an endemic North American mammal (Beard and Wang, 1995; Beard,
1998).

Submitted 23 December 1999

195

196

Annals of Carnegie Museum

VOL. 69

The purpose of this paper is to describe a new species of Carpocristes from
the Bison Basin in south-central Wyoming. This new species, the oldest and most
primitive yet known for the genus, further substantiates the North American origin
of Carpocristes. It also suggests that carpolestids enjoyed their highest species
richness during the middle Tiffanian, when at least three species — Elphidotarsius
wightoni, Carpodaptes hazelae, and the new species of Carpocristes described
below — are known.

Paleocene mammals from outcrops of the Fort Union Formation in the Bison
Basin of south-central Wyoming were first reported by Gazin (1956). Additions
and emendations to the Paleocene mammalian faunas of the Bison Basin have
been made by McGrew and Patterson (1962), MacIntyre (1966), Van Valen (1966,
1978), Szalay (1973), Gingerich (1976, 1983), Sloan (1987), Gunnell (1989), and
Thewissen (1990). Carpolestids have not previously been reported from the Bison
Basin.

CM is the abbreviation used to designate specimens in the collections of the
Section of Vertebrate Paleontology, Carnegie Museum of Natural History, Pitts-
burgh, Pennsylvania. Measurements of tooth length (L) and width (W) follow
those of Rose (1975:fig. 1).

Systematic Paleontology

Class Mammalia Linnaeus, 1758
Mirorder Primatomorpha Beard, 1991
Superfamily Plesiadapoidea Trouessart, 1 897
Family Carpolestidae Simpson, 1935
Subfamily Carpolestinae Simpson, 1935
Carpocristes Beard and Wang, 1995
Carpocristes rosei, new species
(Fig. 1)

Holotype. — CM 40567, left dentary fragment preserving the crowns of P4-M3
and the complete or partial alveoli for several anterior teeth; only known specimen
(Fig. 1).

Type Locality. — Bison Basin Ridge locality, CM loc. 1035. Geographic and
stratigraphic data for this locality, on file in the Section of Vertebrate Paleontology
(CM), indicate that this is the same locality Gazin (1956) referred to as the Ledge
locality. The mammalian fauna from the Bison Basin Ledge locality correlates
with middle Tiffanian zone Ti3 (Archibald et ah, 1987).

Known Distribution. — Middle Tiffanian (Ti3) of the Bison Basin, Fremont
County, south-central Wyoming.

Diagnosis. — P4 differs from that of other species of Carpocristes in being rel-
atively taller and anteroposteriorly shorter, with larger ultimate apical cusp, only
minor posterolingual excavation, and weaker crest uniting main shearing blade
with talonid cusp. M, trigonid cusps less widely splayed than in other species of
Carpocristes. P4 further differs from that of Carpocristes hobackensis, Carpo-
daptes, and Carpolestes in being absolutely shorter anteroposteriorly. P4 further
differs from that of Carpolestes, Carpocristes hobackensis, and Carpocristes or-
iens in having fewer apical cusps. P4 further differs from that of Carpodaptes in
having ultimate apical cusp displaced posteroinferiorly.

Etymology. — For Kenneth D. Rose, whose monograph on the Carpolestidae (Rose, 1975) remains
a standard reference for the group.

2000

Beard — New Carpocristes from the Bison Basin

197

Fig. 1. — Carpocristes rosei, n. sp,, holotype, CM 40567. Left dentary fragment preserving P4=M3 and
whole or partial alveoli for anterior teeth in occlusal (A), buccal (B), and lingual (C) views. Note
prominent region of alveolar bone resorption beneath P4 in buccal view (B). Scale bar = 2 mm.

Description. — Anteriorly, the posterior part of the alveolus for Ii is preserved in CM 40567. This
relatively large alveolus is nearly horizontal in orientation, reflecting the procumbent disposition of Ij
in carpolestids. At least two small alveoli are present between the crown of P4 and the alveolus for
I,, although this part of the dentary is damaged in the holotype. The more posterior of these two
alveoli supported a diminutive, single-rooted P3 by analogy with other carpolestid taxa in which the
crown of this tooth is preserved (Rose, 1975; Beard and Wang, 1995). The more anterior alveolus, of
which only the posterior part is preserved in CM 40567, must have housed another single-rooted tooth
that was similar in size to P3. The homology of the latter tooth locus is uncertain given the fragmentary

198

Annals of Carnegie Museum

VOL. 69

Fig. 2. — Camera lucida tracings of the labial profiles of P4 among selected Carpolestinae (after Krause,
1978:fig. 7). Samples depicted are as follows: (A) Carpocristes cygneus, Roche Percee local fauna, n
= 19; (B) Carpocristes cygneus. Swan Hills site 1, = 4; (C) Carpocristes hohackensis, Dell Creek
Quarry, n = \\ (D) Carpocristes sp.. Police Point local fauna, n = 3; (E) Carpocristes rosei. Bison
Basin Ridge locality, « = 1; (F) Carpodaptes hazelae, Scarritt Quarry and Cedar Point Quarry, n = 2.

nature of the only known specimen of C. rosei, but it must have been either Cj or P2. That which is
preserved of the anterior part of the dentary in CM 40567 does not differ appreciably from comparable
parts of the dentary in Carpodaptes hazelae and Carpocristes cygneus.

The most noteworthy feature on the labial aspect of the dentary is a prominent area of missing
bone beneath the crown of P4, which may have been due to pathology.

The labial profile of P4 (L, 2.00 mm; W, 1.30 mm) closely resembles that of Carpodaptes hazelae,
from which it differs primarily in being smaller and in the placement of the ultimate apical cusp (Fig.
2). P4 in Carpocristes rosei is relatively taller and anteroposteriorly shorter than in other species of
Carpocristes. Five apical cusps are present, which is fewer than occur on P4 in Carpocristes oriens
(n = 7; Beard and Wang, 1995), and at the low end of the range for Carpocristes cygneus (n — 5-7,
mode = 6; Krause, 1978). In contrast to its more anterior position in Carpodaptes, in Carpocristes
rosei the ultimate apical cusp is located almost equidistant between the talonid cusp and the penulti-
mate apical cusp. This posteroinferior displacement of the ultimate apical cusp is a diagnostic apo-
morphy for Carpocristes (Beard and Wang, 1995:fig. 14). In C. rosei the ultimate apical cusp is large
and cuspidate, in contrast to the condition in other species of Carpocristes, in which this cusp is
smaller and less readily distinguished from the crest that unites the talonid cusp with the remainder
of the P4 blade. The latter crest itself is decidedly weaker in C. rosei than in other species of Car-
pocristes, which is a second reason that the ultimate apical cusp is so manifest in C. rosei. Lingually,
vertical ribs are well developed beneath all of the apical cusps of P4. The posterolingual part of the
crown is only weakly excavated. As a result, the anterior apical cusps are located almost directly

2000

Beard — New Carpocristes from the Bison Basin

199

2.0 1

1.8 -

•S

L6 “

L4 -

1.2 -

1.0 ■ — —I — — I ■ ■ » — 1

1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0

P4 length, in mm

Fig. 3. — Bivariate plot of P4 dimensions in the holotype of Carpocristes rosei (CM 40567) and a
sample of Carpocristes cygneus from the Roche Percee local fauna of Saskatchewan [C. cygneus data
from Krause (1978 Table 2)].

anterior to the ultimate apical cusp, rather than being tilted or displaced lingually with respect to the
ultimate apical cusp, as is the case in other species of Carpocristes.

The lower molars of C. rosei differ little from those of other species of Carpocristes, and need not
be described in detail here. In C. rosei the paraconid and metaconid cusps on Mi are less widely
splayed than is the case in other species of Carpocristes. Unlike C. hobackensis, there is no devel-
opment of a lingual cingulid on the trigonid of Mj in C. rosei. In general, the lower molar cusps in
all species of Carpocristes seem to be less inflated and bear better-developed crests than do their
homologues in Carpodaptes and Carpolestes. Measurements of the lower molars in CM 40567 are as
follows: Ml L, 1.20 mm; Mi W, 1.20 mm; M2 L, 1.10 mm; M2 W, 1.25 mm; M3 L, 1.80 mm; M3 W,
1.10 mm.

Discussion. — Carpocristes rosei is most easily confused with Carpocristes cyg-
neus, which is closely related to the new form. However, extensive comparisons
between the holotype of C. rosei and casts of specimens of C. cygneus from the
Roche Percee local fauna of Saskatchewan described by Krause (1978) revealed
consistent morphological differences, which are summarized in the diagnosis. P4
in the holotype of C. rosei also falls outside the range of metric variation observed
in the large sample of C. cygneus from Roche Percee (Fig. 3).

The most nearly complete and most reliably referred specimens of C. cygneus
are derived from sites belonging to late Tiffanian zone Ti4. While these late
Tiffanian specimens differ from the holotype of C. rosei in ways that support a
species-level distinction, fossils from various middle Tiffanian (Ti3) sites that
have previously been referred to C. cygneus appear to be more problematic. Some
or even all of these specimens may ultimately prove to belong to C. rosei rather

200

Annals of Carnegie Museum

VOL. 69

than to C. cygneus. For example, based on published descriptions and illustrations
(Krishtalka, 1973:fig. 16; Krause, 1978:fig. 7), the carpolestid from the Police
Point local fauna of southeastern Alberta is very similar to C. rosei, and these
specimens may well document a northern range extension for this species (Fig.
2). Likewise, it is possible that middle Tiffanian specimens from localities along
the Blindman River in central Alberta referred by Fox (1990) to C. cygneus
pertain to C rosei instead. However, the latter specimens have yet to be described
or illustrated, so that their species-level attribution remains uncertain. Finally,
specimens from the Williston Basin of western North Dakota referred by Holtz-
man (1978) to C. cygneus may instead belong to C. rosei.

Carpolestid Phylogeny and the Phylogenetic Position of
Carpocristes rosei

Introduction

The broad outlines of carpolestid phylogeny have been established for at least
a quarter of a century, when Rose (1975) published his monograph on the group.
Relationships among generic-level taxa of North American carpolestids have re-
mained stable throughout this interval, although the affinities of two Asian genera
have proven to be somewhat more controversial (Beard and Wang, 1995; Bloch
and Gingerich, 1998). Despite this virtual consensus at the generic level, certain
species-level relationships within the group have so far defied resolution. For
example, there is little or no consensus regarding which of the four described
species of Elphidotarsius is most closely related to the clade that includes Car-
podaptes, Carpocristes, and Carpolestes, despite virtual unanimity regarding the
paraphyly of Elphidotarsius (Rose, 1975; Krause, 1978; Fox, 1984). Similarly,
the affinities of both Carpolestes and Carpocristes with respect to the phyloge-
netically more basal species traditionally included in Carpodaptes have yet to be
resolved satisfactorily, although Carpodaptes jepseni is often cited as being close-
ly related to Carpolestes (Rose, 1975; Bloch and Gingerich, 1998). Various work-
ers have entertained the possibility that Carpolestes and some or all of the species
now included in Carpocristes are more closely related to each other than either
are to Carpodaptes (Dorr, 1952; Rose, 1975; Krause, 1978). However, this con-
clusion conflicts with the phylogeny published by Beard and Wang (1995: fig. 14),
in which a relatively basal dichotomy between Carpocristes and Carpolestes was
inferred, with both Carpodaptes hazelae and Carpodaptes jepseni being more
closely allied to Carpolestes than to Carpocristes. A major obstacle to resolution
of these lower-level relationships among carpolestids is our poor knowledge of
the anatomy of many of the relevant taxa, which are often documented by unique
or extremely fragmentary specimens.

Methods

In an attempt to clarify these details of carpolestid phylogeny, I extended the
taxon-character matrix published by Beard and Wang (1995: appendices 1, 2) to
include additional dental characters and all known carpolestid species (Appendices
1, 2). Multistate characters were treated as ordered in cases in which convincing
evidence for a morphocline exists. For example, the number of apical cusps on
P4, which varies from two to eight among carpolestids, was treated as an ordered
character on the assumption that the number of apical cusps on P4 evolved con-
secutively rather than randomly. Likewise, certain other characters were treated

2000

Beard — New Carpocristes from the Bison Basin

201

as irreversible when, as in the case of the presence or absence of a particular
tooth locus, evolutionary reversal seems highly implausible. All character state
transformations were weighted equally. Details regarding character status are pro-
vided in Appendix 1 . Trees were rooted by designating Pronothodectes matthewi,
Chronolestes simul, and a hypothetical ancestor as outgroups to carpolestine taxa.
Phylogenetic analysis of this enhanced dataset using PAUP 3.1.1 (Swofford, 1993)
yielded nine most parsimonious trees, a strict consensus of which is illustrated in
Figure 4.

Results

The topology of the strict consensus tree (Fig. 4) is fully consistent with that
published by Beard and Wang (1995:fig. 14), although additional taxa are included
here. As has long been assumed, Elphidotarsius emerges as a highly paraphyletic
assemblage of species that successively approximates more derived carpolestids
(Fig. 4, nodes 2-4). Both E. shotgunensis and E. russelli appear to be more closely
related to the Carpodaptes + Carpocristes + Carpolestes clade than is either E.
florencae or E. wightoni (Fig. 4, node 4). However, neither of the former species
of Elphidotarsius is represented by relatively complete material, and their derived
P4 structure forms the only unambiguous character support presently available for
the phylogenetic relationships reconstructed here. More nearly complete material
of either or both of these species will be necessary to evaluate their affinities
adequately. As originally proposed by Fox (1984), E. wightoni appears to be
related more closely to the Carpodaptes + Carpocristes + Carpolestes clade than
is E. florencae (Fig. 4, node 3).

Among more advanced carpolestids (Fig. 4, node 5), a fundamental dichotomy
appears to define two basic clades. One of these clades is equivalent to the genus
Carpocristes (Fig. 4, node 6), the most basal species of which is Carpocristes
rosei. Relationships among more advanced species of Carpocristes are the same
as those proposed by Beard and Wang (1995). That is, Carpocristes cygneus
appears to be the sister group of a clade consisting of Carpocristes hobackensis
and Carpocristes oriens (Fig. 4, nodes 7, 8). The sister group of Carpocristes
appears to be a clade that includes both Carpodaptes and Carpolestes.

The phylogenetic position of Carpodaptes hazelae, inferred here to be the sister
group of all other species of Carpodaptes and Carpolestes (Fig. 4, node 9), is the
weakest node on the consensus tree. This node occurs in only 35% of bootstrapped
trees, and is supported by a single character transformation (the loss of P2, which
occurred four times in parallel within Carpolestidae according to the character
walk optimized here). Regardless of the affinities of Carpodaptes hazelae, and
despite Beard and Wang’s (1995) transfer of C. cygneus and C. hobackensis (both
formerly included in Carpodaptes) to Carpocristes, Carpodaptes continues to
emerge as a paraphyletic group of species that successively approximates Car-
polestes. The type species of Carpodaptes, C. aulacodon, appears to be the sister
group of a clade that includes Carpodaptes jepseni and Carpolestes (Fig. 4, nodes
10, 11). Relationships among species of Carpolestes agree with those recently
proposed by Bloch and Gingerich (1998): C. dubius appears to be the sister group
of a clade that consists of C. nigridens and C. simpsoni (Fig. 4, nodes 12, 13).

Discussion

Perhaps the most notable result from the phylogenetic analysis of carpolestids
performed here is the cohesion of the Carpocristes clade with respect to other

202

Annals of Carnegie Museum

VOL. 69

Carpolestes simpsoni (Cf2)
Carpolestes nigridens (Cfl)
Carpolestes dubius (Ti5)
Carpodaptes jepseni (Ti4)
Carpodaptes aulacodon (Ti4)
Carpodaptes hazelae (Ti2)
Carpocristes oriens (Bu)
Carpocristes hobackensis (Ti5)
Carpocristes cygneus (Ti3)
Carpocristes rosei (Ti3)
Elphidotarsius russelli (Til)
Elphidotarsius shotgunensis (Til)
Elphidotarsius wightoni (Ti3)
Elphidotarsius florencae (To3)
Chronolestes simul (Bu)
Pronothodectes matthewi (To3)
Hypothetical Ancestor

Fig. 4. — Strict consensus of nine most parsimonious trees recovered from branch-and-bound search
in PAUP 3.1.1 (Swofford, 1993) of character-taxon matrix given in Appendix 2. For details regarding
character status, see Appendix 1. Tree length = 75; consistency index (excluding uninformative char-
acters) = 0.757. Numerical values above stems of clades indicate percent frequency with which those
clades were supported in 100 bootstrapped trees. Earliest known stratigraphic occurrence of each
species is given in parentheses. For details regarding stratigraphic occurrence and estimated antiquity
of each species, see Appendix 3.

Using the ACCTRAN character-state optimization algorithm, synapomorphies supporting each node
are as follows (see Appendix 1 for description of character states): Node 1 (Carpolestidae), Character
14 (0— >1), Character 25 (0-^1), Character 32 (0-^1), Character 37 (0-»l), Character 38 (0-^1), Char-
acter 41 (0-^1); Node 2 (Carpolestinae), Character 3 (0— >1), Character 4 (0— »1), Character 9 (O—^l),
Character 16 (0-^1), Character 18 (0-^1), Character 19 (0-^1), Character 20 (0-^1), Character 26
(0->l), Character 32 (1-^3), Character 39 (0-»l), Character 42 (0-^1); Node 3, Character 5 (0-^1),
Character 6 (0^1), Character 11 (0~>1), Character 12 (1^2), Character 40 (0-->l); Node 4, Character
5 (1-^2), Character 7 (l->2). Character 13 (l->2). Character 17 (0-a1), Character 33 (0-»l), Character
34 (0^1); Node 5, Character 22 (0-^1), Character 24 (0-4 1), Character 32 (3->4); Node 6 {Carpocris-
tes), Character 10 (0^1), Character 15 (0-4 1), Character 21 (0-^1), Character 35 (0— >1), Character
36 (0-4 1); Node 7, Character 32 (4-^5); Node 8, Character 7 (2-^3), Character 8 (0-^1), Character
10 (1— >2), Character 15 (1^2), Character 27 (0->l), Character 28 (0-4 1), Character 31 (0-^1); Node
9, Character 23 (0-^1); Node 10, Character 6 (1^2), Character 8 (0-4 1), Character 29 (0->l); Node
11, Character 32 (4-45); Node 12 {Carpolestes), Character 27 (0-4l), Character 28 (0^1), Character
30 (0-4 1), Character 32 (5-47); Node 13, Character 8 (1-42).

2000

Beard — New Carpocristes from the Bison Basin

203

advanced carpolestids. This supports the recognition of Carpocristes as a genus
distinct from Carpodaptes, all species of which appear to share more recent com-
mon ancestry with Carpolestes than with Carpocristes. Given the phylogenetic
relationships depicted in Figure 4, cladogenesis between Carpocristes and other
carpolestids must have occurred sometime prior to the late early Tiffanian (Ti2).
This inference is based on the earliest known occurrence of the Carpodaptes +
Carpolestes clade, which is provided by Carpodaptes hazelae at Scarritt Quarry
in the Crazy Mountains Basin of south-central Montana. All species of Carpocris-
tes other than C. oriens are restricted to North America. Both phylogenetic and
biostratigraphic data therefore imply that Carpocristes originated in North Amer-
ica prior to dispersing to Asia sometime during the late Paleocene (Beard and
Wang, 1995; Beard, 1998).

In general, there is marked agreement between the tree topology depicted in
Figure 4 and the stratigraphic ranges of carpolestid species. This can be quantified
using Huelsenbeck’s (1994) stratigraphic consistency index (SCI), which is the
proportion of internal nodes on a cladogram that are consistent with the strati-
graphic distributions of sister taxa divided by the total number of internal nodes.
Of the 13 internal nodes in Figure 4, ten are consistent with the stratigraphic
distributions of sister taxa, yielding an SCI of 0.769. Estimates of ghost lineage
durations for the stratigraphically inconsistent nodes are based on correlation of
Paleocene mammal-bearing strata from the Western Interior of North America
with the Geomagnetic Polarity Time Scale (Butler et ah, 1981, 1987; Berggren
et aL, 1995) and biostratigraphic correlations between North America and Asia
proposed by Beard and Dawson (1999) (Appendix 3). The longest ghost lineage
implied by the phytogeny depicted in Figure 4 and the stratigraphic distribution
of carpolestid species is that between Chronolestes and Carpolestinae. The du-
ration of this ghost lineage is equivalent to the difference between the earliest
known occurrence of a carpolestine (provided by Elphidotarsius florencae in zone
To3, estimated at 61.4 Ma) and the much younger occurrence of Chronolestes
(Bumbanian, estimated at 55.8 Ma), which is roughly 5.6 Ma. A shorter ghost
lineage of about 2.3 Ma duration separates the anachronistically young Elphido-
tarsius wightoni (zone Ti3, estimated at 58.3 Ma) from the earliest known oc-
currence of its sister taxon (zone Til, estimated at 60.6 Ma). In contrast, the
duration of the ghost lineage implied by the earliest known occurrences of Car-
pocristes (Ti3, estimated at 58.3 Ma) and the Carpodaptes + Carpolestes clade
(Ti2, estimated at 59.5 Ma) is only about 1.2 Ma.

Despite the stratigraphically dense succession of carpolestids known from Pa-
leocene basins in western North America, significant episodes of their evolution-
ary history remain undocumented by the fossil record. The documentation of
incompatibility between robust phylogenetic trees and the stratigraphic distribu-
tions of individual taxa is a useful endeavor because it highlights these elusive
episodes of evolutionary history, thereby providing a guide to fertile areas of
future research.

Acknowledgments

A previous version of this manuscript benefited from the comments of three anonymous reviewers.
I thank J. I. Bloch, D. W. Krause, K. D. Rose, and M. T. Silcox for providing comparative casts of
carpolestids, L. Krishtalka for clarifying details regarding the type locality of Carpocristes rosei, M.
Klingler for rendering the figures, and A. R. Tabrum for preparation, molding, and casting. Ongoing

204

Annals of Carnegie Museum

VOL. 69

research on Paleocene mammals from Wyoming is facilitated by the Wyoming State Office of the
Bureau of Land Management (paleontological reources use permit PA98-WY-042).

Literature Cited

Archibald, J. D., W A. Clemens, P. D. Gingerich, D. W. Krause, E. H. Lindsay, and K. D. Rose.
1987. First North American land mammal ages of the Cenozoic Era. Pp. 24-76, in Cenozoic
Mammals of North America: Geochronology and Biostratigraphy (M. O. Woodburne, ed.). Uni-
versity of California Press, Berkeley, California.

Beard, K. C. 1998. East of Eden: Asia as an important center of taxonomic origination in mammalian
evolution. Pp. 5-39, in Dawn of the Age of Mammals in Asia (K. C. Beard and M. R. Dawson,
eds.). Bulletin of Carnegie Museum of Natural History, no. 34.

Beard, K. C., and M. R. Dawson. 1999. Intercontinental dispersal of Holarctic land mammals near
the Paleocene/Eocene boundary: Paleogeographic, paleoclimatic and biostratigraphic implications.
Bulletin de la Societe Geologique de France, 170:697-706.

Beard, K. C., and J. Wang. 1995. The first Asian plesiadapoids (Mammalia: Primatomorpha). Annals
of Carnegie Museum, 64:1-33.

Berggren, W. a., D. V. Kent, C. C. Swisher III, and M.-P. Aubry. 1995. A revised Cenozoic
geochronology and chronostratigraphy. Pp. 129-212, in Geochronology, Time Scales and Global
Stratigraphic Correlation (W. A. Berggren, D. V. Kent, M.-P. Aubry, and J. Hardenbol, eds.).
Society for Sedimentary Geology (SEPM), Special Publication 54, Tulsa, Oklahoma.

Biknevicius, a. R. 1986. Dental function and diet in the Carpolestidae (Primates, Plesiadapiformes).
American Journal of Physical Anthropology, 71:157-171.

Bloch, J. I., and P. D. Gingerich. 1998. Carpolestes simpsoni, new species (Mammalia, Proprimates)
from the late Paleocene of the Clark’s Fork Basin, Wyoming. Contributions from the Museum of
Paleontology, University of Michigan, 30:131-162.

Butler, R. E, P. D. Gingerich, and E. H. Lindsay. 1981. Magnetic polarity stratigraphy and bio-
stratigraphy of Paleocene and lower Eocene continental deposits, Clark’s Fork Basin, Wyoming.
Journal of Geology, 89:299-316.

Butler, R. E, D. W. Krause, and P. D. Gingerich. 1987. Magnetic polarity stratigraphy and bio-
stratigraphy of middle-late Paleocene continental deposits of south-central Montana. Journal of
Geology, 95:647-657.

Dorr, J. A., Jr. 1952. Early Cenozoic stratigraphy and vertebrate paleontology of the Hoback Basin,
Wyoming. Bulletin of the Geological Society of America, 63:59-94.

Fox, R. C. 1984. A new species of the Paleocene primate Elphidotarsius Gidley: Its stratigraphic
position and evolutionary relationships. Canadian Journal of Earth Sciences, 21:1268-1277.

. 1990. The succession of Paleocene mammals in western Canada. Pp. 51-70, in Dawn of the

Age of Mammals in the Northern Part of the Rocky Mountain Interior, North America (T. M.
Bown and K. D. Rose, eds.). Geological Society of America Special Paper no. 243, Boulder,
Colorado.

Gazin, C. L. 1956. Paleocene mammalian faunas of the Bison Basin in south-central Wyoming.
Smithsonian Miscellaneous Collections, 13 1(6): 1-57.

Gingerich, P. D. 1976. Cranial anatomy and evolution of early Tertiary Plesiadapidae (Mammalia,
Primates). University of Michigan Papers on Paleontology, 15:1-141.

. 1983. New Adapisoricidae, Pentacodontidae, and Hyopsodontidae (Mammalia, Insectivora

and Condylarthra) from the late Paleocene of Wyoming and Colorado. Contributions from the
Museum of Paleontology, University of Michigan, 26:227-255.

Gunnell, G. F. 1989. Evolutionary history of Microsyopoidea (Mammalia, ?Primates) and the rela-
tionship between Plesiadapiformes and Primates. University of Michigan Papers on Paleontology,
27:1-157.

Holtzman, R. C. 1978. Late Paleocene mammals of the Tongue River Formation, western North
Dakota. North Dakota Geological Survey, Report of Investigation, 65:1-88.

Huelsenbeck, j. P. 1994. Comparing the stratigraphic record to estimates of phylogeny. Paleobiology,
20:470-483.

Krause, D. W. 1978. Paleocene primates from western Canada. Canadian Journal of Earth Sciences,
15:1250-1271.

Krishtalka, L. 1973. Late Paleocene mammals from the Cypress Hills, Alberta. Special Publications,
The Museum, Texas Tech University, 2:1-77.

MacIntyre, G. T. 1966. The Miacidae (Mammalia, Carnivora). Part 1. The systematics of Ictido-
pappus and Protictis. Bulletin of the American Museum of Natural History, 131:115-210.

2000

Beard — New Carpocristes from the Bison Basin

205

McGrew, R O., and B. Patterson. 1962. A picrodontid insectivore(?) from the Paleocene of Wy-
oming. Breviora, 175:1-9.

Rose, K. D. 1975. The Carpolestidae: Early Tertiary primates from North America. Bulletin of the

Museum of Comparative Zoology, 147:1-74.

. 1977. Evolution of carpolestid primates and chronology of the North American middle and

late Paleocene. Journal of Paleontology, 51:536-542.

Simpson, G. G. 1928. A new mammalian fauna from the Fort Union of southern Montana. American

Museum Novitates, 297:1-15.

Sloan, R. E. 1987. Paleocene and latest Cretaceous mammal ages, biozones, magnetozones, rates of
sedimentation, and evolution. Pp. 165-200, in The Cretaceous-Tertiary Boundary in the San Juan
and Raton Basins, New Mexico and Colorado (J. E. Fassett and J. K. Rigby, Jr., eds.). Geological
Society of America Special Paper no. 209, Boulder, Colorado.

SwoEFORD, D. L. 1993. PAUP: Phylogenetic Analysis Using Parsimony, version 3.1.1. Computer
program distributed by the Illinois Natural History Survey, Champaign, Illinois.

Szalay, E S. 1973. New Paleocene primates and a diagnosis of the new suborder Paromomyiformes.
Folia Primatologica, 19:73-87.

Thewissen, j. G. M. 1990. Evolution of Paleocene and Eocene Phenacodontidae (Mammalia, Con-
dylarthra). University of Michigan Papers on Paleontology, 29:1-107.

Van Valen, L. 1966. Deltatheridia, a new order of mammals. Bulletin of the American Museum of

Natural History, 132:1-126.

— 1978. The beginning of the Age of Mammals. Evolutionary Theory, 4:45-80.

Appendix 1

Character Descriptions, with Character Status in Italics

1. Laterocone or disto-apical cusp on I‘: absent (0); or present (1).

2. Mediocone on E: absent (0); or present (1).

3. Mesio-basal cusp on P: absent (0); or present (1).

4. Lingual crest on P^: absent (0); or present (1).

5. Number of cusps on lingual crest of P^: none (0); one (1); or two (2) {ordered).

6. Parastyle on P^: absent (0); present (1); or present, with neomorphic cusp anterior to it (2)
{ordered).

7. Number of buccal cusps posterior to paracone on P^: none (0); one (1); two (2); or three (3)
{ordered).

8. Parastylar lobe on P^: does not project anteriorly with respect to lingual part of tooth (0); moderate
anterior projection (1); or extreme anterior projection (2) {ordered).

9. Median crest on P^: absent (0); or present (1).

10. Number of median crests on P^: one (0); two (1); or three (2) {ordered).

11. Position of primary median crest on P^: lingual, closely appressed to lingual crest (0); or labial,
widely separated from lingual crest (1).

12. Parastyle on P^: absent (0); present, single (1); or present, dual (2) {ordered).

13. Number of buccal cusps posterior to paracone on P'^: none (0); one (1); or two (2) {ordered).

14. Median crest on P"^: absent (0); or present (1).

15. Number of median crests on P^: one (0); two (1); or three (2) {ordered).

16. Position of protocone on P'*: anterior (0); or central (1).

17. Number of lingual cusps on P'^: one (0); or three (1).

18. Crest running anterior to paracone on P'^: absent (0); or present (1).

19. Position of paraconule on P'^: anterior (0); or central (1).

20. Postparaconule crista on P^: incomplete (0); or complete (1).

21. Size of upper and lower molars: larger than in Carpocristes spp. (0); or as in Carpocristes spp.

(1).

22. I3: present (0); or absent (1) {irreversible).

23. P2: present (0); or absent (1) {irreversible).

24. P3: double-rooted (0); or single-rooted (1) {irreversible).

25. Size of P3: unreduced (0); or reduced or absent (1) {irreversible).

26. Plagiaulacoid P4: absent (0); or present (1).

27. Vertical rib beneath ultimate apical cusp on lingual side of P4: present (0); or absent (1).

28. Crest uniting penultimate apical cusp with talonid cusp on P4: weak (0); or strong (1).

29. Vertical ribs beneath anterior apical cusps on lingual side of P4: vertically oriented (0); or steeply
inclined from base of tooth anteriorly to apical cusps posteriorly.

206

Annals of Carnegie Museum

VOL. 69

30. Position of P4 talonid cusp; well below the level of Mj trigonid (0); or elevated to near the
level of M, trigonid (1).

31. Ultimate apical cusp on P4: cuspate (0); or indistinct or absent, being incorporated within crest
uniting main shearing blade of P4 with talonid cusp (1).

32. Number of apical cusps on P4: one (0); two (1); three (2); four (3); five (4); six (5); seven (6);
or eight (7) {ordered).

33. Position of ultimate apical cusp (= metaconid) on P4: slightly to moderately lingual to penul-
timate apical cusp (0); or directly posterior to it (1).

34. Position of apical cusps immediately preceding penultimate apical cusp (= protoconid) on P4:
slightly to moderately lingual to penultimate apical cusp (0); or directly anterior to it (1).

35. Position of ultimate apical cusp on P4: near penultimate apical cusp (0); or more posterior in
position, roughly equidistant between penultimate apical cusp and talonid cusp (1).

36. Posterolingual excavation on P4: absent (0); or present (1).

37. Anteroposterior elongation of P4: absent (0); or present (1).

38. Exodaenodonty on P4: absent (0); or present (1).

39. Paraconid of M,: not widely splayed relative to metaconid (0); or widely splayed (1).

40. Talonid notch on M,: weak to absent (0); or strong (1).

41. Protoconid of M;: same height as paraconid and metaconid (0); or taller than paraconid and
metaconid (1).

42. Talonid of M,: similar in anteroposterior length to that of M2 (0); anteroposteriorly abbreviated

(1).

Appendix 2

Taxon-Character Matrix Used in Parsimony Analysis

Hypothetical Ancestor

0000?

00?0?

?000?

00000

00000

0???0

?0???

00000

00

Pronothodectes matthewi

1100?

01?0?

?110?

00000

01000

0???0

?0???

00000

00

Chronolestes simul

0000?

00?0?

?0010

00000

01011

000?0

010?0

01100

10

Elphidotarsius florencae

???10

01?10

OHIO

10111

01001

10000

03000

OHIO

11

Elphidotarsius wightoni

? ? ?11

11010

12110

10111

011?1

10000

03000

01111

11

Elphidotarsius russelli

9 9 9 9

9 9 9 9 9

9 9 9 9 9

9 9 9 9 9

0????

10000

03110

Oil??

? ?

Elphidotarsius shotgunensis

9 9 9 9 9

9 9 9 9 9

9 9 9 9 9

9 9 9 9 9

000?1

10000

03110

01111

11

Carpodaptes hazelae

11112

12010

12210

11111

01111

10000

04110

01111

11

Carpodaptes aulacodon

9 9 9 9 9

9 9 9 9 9

9 9 9 9 9

9 9 9 9 9

0??11

10010

04110

01111

11

Carpodaptes jepseni

9 9 9 9 9

9 9 9 9 9

9 9 9 9 9

9 9 9 9 9

0????

10010

05110

01111

11

Carpolestes duhius

???12

22110

12210

11111

01111

11111

07110

01111

11

Carpolestes nigridens

10112

22210

12210

11111

01111

11111

07110

01111

11

Carpolestes simpsoni

10112

22210

12210

11111

011?1

11111

07110

01111

11

Carpocristes rosei

9 9 9 9 9

9 9 9 9 9

9 9 9 9 9

9 9 9 9 9

1 ? ? ? 1

10000

04111

Hill

11

Carpocristes cygneus

10112

12011

12211

11111

111?1

10000

05111

Hill

11

Carpocristes hobackensis

9 9 9 9 9

9 9 9 9 9

9 9 9 9 9

9 9 9 9 9

11111

11100

15111

Hill

11

Carpocristes oriens

???12

13112

12212

11111

110?1

11100

16111

Hill

11

Appendix 3

Explanatory Notes for the Age Estimates Used Here for Carpolestids
and Related Taxa, based on Biostratigraphy and
Paleomagnetic Stratigraphy

Pronothodectes matthewi: Gidley Quarry, Crazy Mountains Basin, MT (To3). Assigned to upper part
of Chron 27r, based on paleomagnetic data from Silberling Quarry, Crazy Mountains Basin and Rock
Bench Quarry, Bighorn Basin (Butler et al., 1987). Absolute age estimated at 61.4 Ma (Berggren et
ah, 1995).

Chronolestes simuh Wutu Formation, Shandong Province, People’s Republic of China (Bumbanian).
Estimated to correlate with North American zone Cf2 based on biostratigraphy (Beard and Dawson,

2000

Beard — New Carpocristes from the Bison Basin

207

1999). Assigned to lowest part of Chron 24r, based on paleomagnetic data from the Bighorn Basin
(Butler et ah, 1981). Absolute age estimated at 55.8 Ma (Berggren et ah, 1995).

Elphidotarsius florencae: Gidley Quarry, Crazy Mountains Basin, MT; Rock Bench Quarry, Bighorn
Basin, WY (To3). Assigned to upper part of Chron 27r, based on paleomagnetic data from Silberling
Quarry, Crazy Mountains Basin and Rock Bench Quarry, Bighorn Basin (Butler et ah, 1987). Absolute
age estimated at 61.4 Ma (Berggren et al., 1995).

Elphidotarsius wightoni: University of Alberta localities DW-1 and DW-2, near Red Deer, Alberta
(Ti3). Fossils possibly pertaining to this species were cited by Fox (1990:59) from Aaron’s Locality
(Til or Ti2), but the stratigraphic range for the species is not extended here, pending formal description
of the relevant specimens. Assigned to upper part of Chron 26r, based on paleomagnetic data from
the Crazy Mountains Basin and Bighorn Basin (Butler et al., 1981, 1987). Absolute age estimated at
58.3 Ma (Berggren et ah, 1995).

Elphidotarsius shotgunensis: Keefer Hill local fauna (= Shotgun local fauna). Wind River Basin, WY
(Til). Assigned to lower part of Chron 26r, based on paleomagnetic data from the Crazy Mountains
Basin and Bighorn Basin (Butler et al., 1981, 1987). Absolute age estimated at 60.6 Ma (Berggren et
al., 1995).

Elphidotarsius russelli: Cochrane 2 locality, Porcupine Hills Formation, Alberta (Til). Assigned to
lower part of Chron 26r, based on paleomagnetic data from the Crazy Mountains Basin and Bighorn
Basin (Butler et al., 1981, 1987). Absolute age estimated at 60.6 Ma (Berggren et al., 1995).

Carpocristes rosei: Ridge locality [= Ledge locality of Gazin (1956)], Bison Basin, WY (Ti3). As-
signed to upper part of Chron 26r, based on paleomagnetic data from the Crazy Mountains Basin and
Bighorn Basin (Butler et al., 1981, 1987). Absolute age estimated at 58.3 Ma (Berggren et al., 1995).

Carpocristes cygneus: Swan Hills site 1, Alberta; Canyon Ski Quarry, Alberta; Roche Percee local
fauna, Saskatchewan (Ti4). Several possible records from zone Ti3 (e.g., from University of Alberta
locality DW-1 near Red Deer, Alberta; Police Point local fauna, Alberta) have been listed by Fox
(1990). Additional specimens from the Judson and Brisbane localities in western North Dakota (Ti3)
may also pertain to C. cygneus. Although the best samples of this species are known from Ti4 sites,
the earliest known records of C. cygneus are provisionally considered to be Ti3. Assigned to upper
part of Chron 26r, based on paleomagnetic data from the Crazy Mountains Basin and Bighorn Basin
(Butler et al., 1981, 1987). Absolute age estimated at 58.3 Ma (Berggren et al., 1995).

Carpocristes hobackensis: Dell Creek Quarry, Hoback Basin, WY (Ti5). Assigned to middle part of
Chron 25r based on paleomagnetic data from the Bighorn Basin, WY (Butler et ah, 1981). Absolute
age estimated at 57.0 Ma (Berggren et al., 1995).

Carpocristes oriens: Wutu Formation, Shandong Province, People’s Republic of China (Bumbanian).
Estimated to correlate with North American zone Cf2 based on biostratigraphy (Beard and Dawson,
1999). Assigned to lowest part of Chron 24r, based on paleomagnetic data from the Bighorn Basin
(Butler et ah, 1981). Absolute age estimated at 55.8 Ma (Berggren et al., 1995).

Carpodaptes hazelae: Scarritt Quarry, Crazy Mountains Basin, MT (Ti2); Cedar Point Quarry, Bighorn
Basin, WY (Ti3); Hand Hills West locality, Alberta (Ti3); various localities in the Paskapoo Formation
along the Blindman River near Red Deer, Alberta (Ti3); Joffre Bridge Roadcut, lower level, Alberta
(Ti3). Earlier specimens that may pertain to this species are known from the Keefer Hill local fauna
of the Wind River Basin, WY (Til) and from the Cochrane 2 site in Alberta (Til). However, pending
fuller description of these specimens, the earliest reliable occurrence of C. hazelae is here regarded
as Ti2. Assigned to the middle part of Chron 26r, based on paleomagnetic data from Scarritt Quarry,
MT (Butler et al., 1987). Absolute age estimated at 59.5 Ma (Berggren et al., 1995).

Carpodaptes aulacodon: Mason Pocket locality, Animas Formation, San Juan Basin, CO (Ti4). As-
signed to lower part of Chron 25r, based on paleomagnetic data from the local section (Butler et al.,
1981). Absolute age estimated at 57.4 Ma (Berggren et al., 1995).

Carpodaptes jepseni: Divide Quarry, Bighorn Basin, WY (Ti4). Assigned to lower part of Chron 25r,

208

Annals of Carnegie Museum

VOL. 69

based on paleomagnetic data from the Bighorn Basin (Butler et al., 1981). Absolute age estimated at
57.4 Ma (Berggren et al., 1995).

Carpolestes dubius: Princeton Quarry and nearby sites, Bighorn Basin, WY (Ti5). Assigned to middle
part of Chron 25r based on paleomagnetic data from the Bighorn Basin, WY (Butler et al., 1981).
Absolute age estimated at 57.0 Ma (Berggren et al., 1995).

Carpolestes nigridens: Bear Creek locality, northern Bighorn Basin, MT (Cfl); Big Multi Quarry,
Washakie Basin, WY (Cfl). Assigned to early part of Chron 25n based on paleomagnetic data from
the Bighorn Basin, WY (Butler et al., 1981). Absolute age estimated at 56.3 Ma (Berggren et al.,
1995).

Carpolestes simpsoni: various localities in the northern Bighorn Basin, WY (Cf2-Cf3). Assigned to
lowest part of Chron 24r, based on paleomagnetic data from the Bighorn Basin (Butler et al., 1981).
Absolute age estimated at 55.8 Ma (Berggren et al., 1995).

INSTRUCTIONS FOR AUTHORS

ANNALS OF CARNEGIE MUSEUM consists of
contributions to the earth sciences (including paleontol-
ogy), life sciences, and anthropology, in 30 by 46 pica
format (127 by 195 mm or 5 by 7% inches). Submit all
manuscripts to the Office of Scientific Publications. Au-
thors should give particular attention to scientific con-
tent, format, and general style for the ANNALS. Manu-
scripts that do not conform to the style of the AN-
NALS will be returned to the author immediately.
Every manuscript will be subjected to external peer re-
view. Authors should submit the names, addresses,
phone and fax numbers, and e-mail addresses of at least
four qualified potential reviewers for each manuscript.
Authors will be asked to subsidize, if fimds are available,
any or all costs of publication (approximately $ 1 00/page
printed). A detailed set of instructions to authors may be
requested from the Office of Scientific Publications.

Manuscript Style. — Articles should include the fol-
lowing items in this order; title page, abstract, text (with
desired headings), acknowledgments, literature cited, ta-
bles, figure captions, and copies of illustrations. All
manuscripts must be t5q)ed double-spaced on standard
814- by 1 1-inch white bond paper, with at least one-inch
margins all around, and submitted in quadruplicate — an
original for the editors, and three review copies. All
pages should be numbered, including tables, literature
cited, and the list of figure captions. Only correspon-
dence quality or better dot matrix printouts will be ac-
cepted; draft mode is unacceptable. A monospaced font,
(e.g.. Courier) is required. A disk in WordPerfect x.x or
ASCII is helpful.

Title Page. The title should be brief, include the ani-
mal or plant group involved, be placed on a separate
page, and appear two to three inches below the top mar-
gin. Include the author’s name(s) and the affiliations of
non-CMNH authors. In the case of multiple authorship,
indicate the address to which proofs should be sent.

Abstracts. Abstracts should be short, but substantive,
and included at the head of the first page of text.

Text. Do not right justify text or break (hyphenate) a
word at the end of a line. Footnotes and acknowledg-
ments as footnotes are unacceptable. All text is double-
spaced in a monospaced font (e.g.. Courier).

Tables and Figure Legends. The list of figure legends
and each table should be typed consecutively on indi-
vidual pages separate from the text. Tables must be dou-
ble-spaced throughout with no vertical lines.

Measurements. Metric units should be used, if possi-
ble.

Literature Cited. All references to literature in text
must appear in the Literature Cited section. The data
(author, date, and page) in both citations must agree. Do
not abbreviate the titles of periodicals or serials. The
following style, double-spaced, should be used in Lit-
erature Cited:

1) Two authors in a journal series:

Soltis, D. E., and P. S. Soltis. 1992. The distri-
bution of selling rates in homosporous ferns.
American Journal of Botany, 79:97-100.

2) Same authors repeated — use three-em dash:

. 1923. The fauna of the Ardyn Obo For-
mation. American Museum of Natural History
Novitates, 98:1-5.

3) Same authors plus a third author — repeat all authors:

Knutson, L. V, R. E. Orth, and W. L. Murphy.
1986. Catalog of Sciomyzidae (Diptera) of
America north of Mexico. Entomography, 4;
1-53.

4) Chapter in an edited volume:

Rausch, R. L. 1963. A review of the distribution
of Holarctic mammals. Pp. 29-43, in Pacific
Basin Biogeography (J. L. Gressitt, ed.).
Bishop Museum Press, Honolulu, Hawaii.

5) Unpublished dissertation:

Smith, J. P. 1976. Review of Eocene Mammals.
Unpublished Ph.D. Dissert., University of
California, Berkeley, California.

6) Book:

White, M. J. D. 1961. The Chromosomes. Meth-
euen and Co., Ltd., London, United Kingdom.

7) Journal articles with usual volume and issue number:

Anderson, W. I. 1969. Lower Mississippian con-
odonts from northern Iowa. Journal of Pale-
ontology, 43:916-928.

Figures. Including all illustrative materials (line art,
halftones, photographs), figures are to be numbered in
Arabic numerals. Four sets of figures are required, one
(original artwork) for reproduction, three for reviewers.
Photocopies of photographs for reviewers are usually not
acceptable but are adequate for line drawing review cop-
ies. Figures may not be larger than 17 by 12 inches.
Reducing figures is the responsibility of the author. All
figures must be reducible to a maximum of 127 by 195
mm (30 by 46 picas) without loss of clarity. Line copy
should be designed for reduction to % or Vz or actual size.
Typewritten figure copy will not be accepted. Pho-
tographic figures should be submitted at actual repro-
duction size, if possible.

Rectangular halftone figures should be abutted, with-
out intervening spaces. The printer will insert narrow
white spaces during the reproduction process. All figures
must have minimally one-inch borders all around. Each
figure should be given a protective cover and identified
on the back side.

Lettering and/or a magnification scale (linear metric
scale) for rectangular halftone figures should be placed
directly on the photo, not in a blank space between pho-
tos. The scale or lettering for closely cropped photos can
be placed in blank areas close to the figure.

Proof. — The author should answer all queried proof
marks and check the entire proof copy. Return corrected
page proof with the edited manuscript promptly to the
Office of Scientific Publications.

If an author chooses to make extensive alterations to
a paper in proof stage, the author will bear the cost.
Original manuscripts will not be returned unless request-
ed. Figures will be returned to the author only if re-
quested prior to publication.

:

ANNALS

0/ CARNEGIE MUSEUM

THE CARNEGIE MUSEUM OF NATURAL HISTORY

4400 FORBES AVENUE • PITTSBURGH, PENNSYLVANIA 15213

VOLUME 69 29 NOVEMBER 2000 NUMBER 4

CONTENTS

ARTICLES

The upper Paleolithic bone industry of Klithi Rock Shelter, northwest Greece

Sandra L. Olsen 209

Review of Miocene (Hemingfordian to Clarendonian) mylagaulid rodents

(Mammalia) from Nebraska William W. Korth 227

Index to Volume 69

281

Editors, ANNALS, BULLETIN and SPECIAL PUBLICATIONS:

D. S. Berman, Ph.D.

D. R. Watters, Ph.D.

J. R. Wible, Ph.D.

W. O. Goehring, Managing Editor

Manuscripts, subscriptions, orders for individual numbers, and changes of address
should be sent to:

Office of Scientific Publications
Carnegie Museum of Natural History
4400 Forbes Avenue
Pittsburgh, PA 15213-4080
Phone (412) 622-3287
Fax (412) 622-8837
E-mail: scipubs@clpgh.org

ANNALS OF CARNEGIE MUSEUM is published quarterly by The Carnegie Museum of Natural
History, 4400 Forbes Avenue, Pittsburgh, Pennsylvania 15213-4080, by the authority of the Board of
Trustees of Carnegie Institute.

© 2000 Carnegie Institute.

0 This paper meets the requirements of ANSi/NISO Z39.48-1992 (Permanence of Paper).

ANNALS OF CARNEGIE MUSEUM

VoL. 69, Number 4, Pp. 209-226

29 November 2000

THE UPPER PALEOLITHIC BONE INDUSTRY OF KLITHI ROCK
SHELTER, NORTHWEST GREECE

Sandra L. Olsen

Associate Curator, Section of Anthropology

Abstract

Little has been published to date on the range of bone artifacts from the upper paleolithic of Greece,
so even relatively small collections can yield significant information that assists in understanding the
technology of the region in the Late Pleistocene. The Klithi rock shelter represents an important period
(21,300-12,600 B.R) just before the end of the Ice Age when paleolithic hunters moved into a pre-
viously uninhabitable area in order to hunt ibex and chamois. Their bone artifacts reveal that they
conducted domestic activities in the rock shelter, wore ornaments, and obtained some objects outside
the gorge through trade or travel. Manufacturing techniques were simple, but aspects of needle fab-
rication seem to indicate an indigenous process not reported in western Europe or the Near East.

Key Words: bone artifacts, Greek archaeology, upper paleolithic

Introduction

The upper paleolithic site of Klithi is located about 40 km north of the city of
loannina, the provincial capital of Epirus, in northwest Greece (Fig. 1). The site
is situated in a rock shelter in the Voidomatis gorge (Fig. 2) and contains deposits
from 21,300-12,600 B.R Excavated by a team of scientists headed by Geoffrey
Bailey in 1983-1988, Klithi has yielded considerable information about the re-
peated use of the rock shelter over a span of 9,000 years.

The Klithi shelter (Fig. 3) is 25 m wide at the entrance, extends back into the
rock 10 m, and has an overhanging ledge 40 m above its floor (Adam, 1989:224).
Excavations revealed deposits of accumulated natural scree and rockfall from the
shelter’s ceiling, hearths and ash deposits, a lithic industry based heavily on flakes
and backed bladelets, red ocher, faunal remains dominated by ibex and chamois,
marine and terrestrial mollusk shells, and a small assemblage of bone artifacts.

The early prehistory of Epirus is poorly known, partly because of the rugged
regional terrain and partly because of the limited investigation it has received
until recently. Based on its position, northwest Greece might have served as a
cultural conduit between the Near East and Europe during the paleolilthic, but
instead the glaciated mountains of the region probably acted as a hurdle that
impeded the flow of people and ideas during the Pleistocene.

The Environment

Epirus is a region of great topographic and geographic variation. In the north-
eastern part of Epirus the Pindus Mountains reach heights over 2,500 m and form
deep vertical river gorges. Albania, to the northwest, is also mountainous. On the
southwest Epirus is bounded by the Ionian Sea and on the south by the Gulf of
Arta and the low relief of Akamania.

During the last glacial maximum between 20,000-16,000 b.p., Epirus probably

Submitted 29 August 1997.

209

210

Annals of Carnegie Museum

VOL. 69

Fig. 1. — Map of Greece locating Klithi, Kastritsa, and Asprochaliko.

presented an inhospitable environment for human occupation (Bailey et aL, 1990).
Bailey et al. (1993) attribute the paucity of sites from this temporal range to low
biomass, snow cover in winter, and a shortage of sources of water.

According to pollen studies (Bottema, 1974; Willis, 1989), the Late Wiirm of
this area was cold and semi-arid. The vegetational regime was forest-steppe, con-
sisting mainly of Artemisia-dommdiXQd grasslands associated with groves of oak
and pine. Winters were probably 5-6° C colder than today and summers were
arid and 2-3° C cooler than present (Prentice et al., 1992).

Today the site of Klithi is located about midway up a steep, vertical- sided
gorge. At an elevation of 500 m, it is surrounded by Mount Gamila (elevation
2,500 m) and other peaks of the Pindus chain. Currently, the rock shelter is
precipitously perched 30 m above the Voidomatis River. The difficult access to
the shelter at present is somewhat deceptive, however, because the valley has
undergone significant changes in the last 10,000 years. At the time of occupation,
the valley was partially filled with a river terrace referred to as the Aristi Unit up
to just 12 m below the rock shelter (Bailey et al., 1990). The river was a few

Olsen — Klithi Rock Shelter Bone Industry

211

2000

Fig. 2. — Voidomatis river gorge near Klithi rock shelter.

Fig. 3. — Klithi rock shelter.

212

Annals of Carnegie Museum

VOL. 69

meters below that and the contours of the valley would have been broadly U-
shaped. The terrace’s geologic constituents made it a suitable source for the small
nodules of flint used by the occupants of Klithi for manufacturing stone tools
(Adam, 1989:225; Bailey et al., 1990:148). Much of the terrace was gouged out
in the Holocene to expose the greater depths of the gorge.

The broad river valley and rising mountains apparently provided an ideal en-
vironment for ibex {Capra ibex) and chamois (Rupicapra rupicapra), the domi-
nant herbivores in the faunal assemblage at Klithi. With their gentler topography
and lower elevations, other regions like the loannina Basin to the south and the
coast to the west were home to red deer {Cervus elaphus).

Paleolithic Sites in the Region

In the 1960s, Eric Higgs (Dakaris et al., 1964; Higgs and Vita-Finzi 1966;
Higgs et al., 1967) led several expeditions into northwest Greece searching for
paleolithic sites. His surveys revealed three of particular interest: Asprochaliko,
Kastritsa, and Klithi (Fig. 1). The oldest site in Epirus containing a dated strati-
graphic sequence is Asprochaliko, an open-air occupation near the coast contain-
ing (from bottom to top): Typical Mousterian, Micro-Mousterian, and upper pa-
leolithic deposits. The earliest upper paleolithic radiocarbon date is 26,100 ± 900
b.p. (uncalibrated) (Bailey et al., 1983:22). The lower levels of the upper paleo-
lithic may represent Gravettian, moving up into Epi-Gravettian comparable to the
Italian industry known from 20,000 b.p. on (Palma di Cesnola, 1976; Bailey et
al., 1983:30). More recent dates of 17,200 ± 400 b.p. and 13,700 ± 260 b.p. for
deposits on the talus slope appear unreliable because of mixing (Bailey et al.,
1983:22).

Kastritsa is an upper paleolithic site located in a rock crevice near Lake loan-
nina about 40 km south of Klithi. Kastritsa has a longer temporal span than Klithi,
ranging from around 20,200 ± 480 b.p. (Bailey et al., 1983:29) to 12,400 ±210
b.p. (Adam, 1989:252), but ends 2,000 years before the abandonment of Klithi.
Based on its selection of stone tools, Kastritsa shows some affinity with the Epi-
Gravettian industries of Italy after 20,000 b.p. (Bailey et al., 1983). Its location
in the lowlands near the lake impacted strongly on the economy of its inhabitants
and their raw materials for bone artifacts. Unlike Klithi, where the occupants
survived primarily on ibex and chamois, the Kastritsa people were mainly red
deer hunters. The bone artifact assemblage from Kastritsa is compared to and
contrasted with Klithi ’s assemblage in this report.

The Bone Artifacts from Klithi

The sample of bone artifacts (Table 1) analyzed from the 1983-1985 excava-
tions at Klithi is small (n = 38 objects) in comparison to the abundance of un-
modified faunal remains and the vast lithic assemblage. Despite this fact, the
collection yields considerable information about the upper paleolithic bone tech-
nology in this locality and can be productively compared to the collections from
neighboring regions, like the loannina Basin. The worked bones are mostly de-
rived from cultural layers that date to between 17,400-12,600 B.R (calibrated).
Adam and Kotjabopoulou (1997) examined additional bone tools from later ex-
cavations, but no significant new information was uncovered.

The osseous artifacts contribute to the interpretation of the site by providing
possible evidence for seasonal migration of hunters. The faunal material shows a

Table 1. — KUthi Bone Artifacts.

2000

Olsen — Klithi Rock Shelter Bone Industry

213

u 0) w « u ai u
a d fl d d a d
o o o o o o o
^ ^ ^ ^ ^

flj gj
-

dddadfiddCdMfiSddfiddddd
OOOOOOOOOOGoOOOOOOOOOnOOO

bObObOmmbOm^ mbObQbomm^bOmboB mbObOOibQmmbObObj^S bD m bJ)-S

dddddSdigdddaedddddajdCfidddddfiCajdddS

B S ^ B o

« « « .

s s s «

'd 'c ’d
(S cd d
o o o d

1 "d

'd 'g

"d

"d

1

i cd

i i

rt

! cti

1

3 S

1

s s

6

1 E

1 S

CO c«5

s s s s s
^ ^

^

^ -2 »S -S »2
^ ^ ^ ^ ^

VO X L' M m VO

QO

sn ^

VO

|in>r)vomivo|voi IvomIvovoI

M VO

^ i?) (M I IT) m

i I ^ I I S I «

m CO IT) VO m VO

^ ^ I I 2 I I ^

I S I I ^ I

o ^

Ov ^
U M

O X

P ^ ov m n

U y VO ^

^ ^ ^

Q 00

OV

o

rO

o

o

o

ov

cn

O

cn

tTi

CN

m

O

(T)

o

o

O

o

Ov

Ov

w

ov

r!

ov

o

X

r-

o

w

>

OV

o

O

Ov

P

U

Ov

<

ov

m

Ov

u

Q

Q

P

P

>

.2

'cfl

u ^

P

<

U

m

m

o

ov

Ov

Q

%

U

CM

ov

Q

o

CO

>

00

Q

<

CM

O

OV

Ov

Q

<

m

OV

<

if)

n

P

X

CN|

p

c-l o

^ CVI

m

e^i

a

(N

m

B/B

r>-

CN

a

O

o< ^ C

J m

(N

r-

CM

P

\D

CM

a

IT) Q
CM ^

a<

T-i ^

CN ^

ov

CM

N>

CM

P

Q

CM

CM

P

o

CM

m

CM

CM

a

mm o\ (X omr^-mm *rivovncNr-'^cNO'^M ^vo^mooMcnoo mvoo^'-i
OO > 0(M0-v0»-i^^— iC^OO^O <-i^r-NC'^0

TOVO^^CNlCNlMOOOVO^MOO'^OOr-'^OVOioOCNiC^CVIC^Or'CNlOlOvCNl^O^O
loin^inco ^co!>CN|Vir^invom<Nin»-^voin^'^ voN-voinininvocnvomvooc^^en

mm<uffl<ffluup<<pijmmpuuuuu<umuuppup<<J'uu®up

d a d

^ ^ ^

■q ’o o "o

^ X ^ ^

^ ^ ^ ^ X

r • ^ ^ ^

ujS'K'^'K'S 'o’o'o 'SuoJ’o

rtxxxxx Ph^iOmPhO^ pM-g -g -g S Ph -*-<

OT M M M M X '-d ^ ‘d M X X ^ 'd ’d ^ ^

o o

B

1

o

s

1

u

a ia ^

3

00

OO

00

w) o g

'o

X

cd

cd

£d

cd X X

X

X

X

41

X

X

< <

^ ^ ^
< < <

^ ^ ^ ^ ^
< < < < <

a) « 0 u « 0)

d ^ ^ ^ ^ ^

^ 0) « « (D « 0)

> « « W « « (U

^ z ^ ^ ^ ^ z

S o d d

^ ^ X X

° ° X X
0) u « w

^ ^ "o 'o
0) 0) -y -y
« CD O O

^ ^ z z

o o

X X

-a d
« «
X X
o o

o o

d d d s ^

C^ cd Cti cd y

"d 'P X

e d d d o

0) O U 4) >

Ph p, a o.<>

214

Annals of Carnegie Museum

VOL. 69

Strong preference for hunting ibex and chamois in the gorge. Curated artifacts
made from red deer canines and antler indicate that the Klithi people either hunted
in other areas or, less likely, received trade items from outside the immediate
vicinity of the rock shelter.

The excavations at Klithi were very meticulous. The site was excavated in
quarter-meter square units, recording artifact depth by natural layers and 5 cm
arbitrary levels. Soil was dry- and wet-sieved with fine mesh. Flotation samples
were taken in some quads and one of these samples yielded the largest awl (Fig.
6). All artifacts were recorded three-dimensionally from the datum. Because of
Greek antiquities laws, all bone artifacts had to be analyzed in Greece and there-
fore could not be removed for scanning electron microscopic analysis or further
study.

Although both unworked and worked bones at Klithi were often severely com-
minuted by roof fall, trampling, and natural pedoturbation, the surfaces of most
of the artifacts are relatively unmarred by taphonomic processes. The limestone
rock shelter provided a nearly ideal environment for bone preservation because
of its alkalinity. Three of the objects exhibit root-etching, one is pitted by roots
or some other natural agent, and the surface of one is partially exfoliated. The
fairly good state of preservation allows interpretations of manufacturing tech-
niques and possible use to be made in many cases.

Burning of bone artifacts was quite common, with ten (26%) of the pieces
showing evidence of direct exposure to heat. The best explanation for the burning
of the worked bone is that the objects were discarded or lost in or near hearths.
The rock shelter floor is strewn with accumulations of osseous material, some of
which is incorporated in the hearths.

In general, the bone artifacts are in deposits consisting of a mixture of faunal
material (food refuse) and lithic tools and debitage. These deposits do not clearly
indicate functional work areas, such as manufacturing stations, and specific lo-
cations of bone tools do not, unfortunately, assist in extrapolating the functions
of the objects or the materials upon which they were used.

Raw Materials. — Given the volume of the faunal assemblage, it is clear that
raw material for the manufacture of bone implements and ornaments was plentiful.
Although most of the artifacts were made on medium manunalian long bones that
do not retain identifiable features, the cortical thickness and overall dimensions
of the objects suggest that they were manufactured from the limb bones of ru-
minants within the size range of the predominant species, ibex and chamois. There
was just one example each of the use of a long bone of a small mammal (fox-
sized) and a large mammal (red deer- sized).

Antler, whether modified or not, is extremely rare in the rock shelter deposits.
This is not surprising, given that the rocky terrain and mountain slopes that sur-
round the site are better suited to ibex and chamois than to cervids. There is a
sharp contrast between the use of a major resource like antler at Kastritsa, in the
loannina basin and Klithi in the Voidomatis gorge. Only a single antler tine (field
number B3001) shows cultural modification at Klithi, whereas Kastritsa produced
numerous pieces of worked antler. The shortage of antler as a raw material at
Klithi appears to be the most important factor in the differences in tool types
between the two areas.

Despite the rarity of antler and postcranial elements of red deer, their canines
do occur in the site as pendants. The frequency of these isolated teeth at Klithi
indicates a cultural preference which must have involved curation of a product

2000

Olsen — Klithi Rock Shelter Bone Industry

215

obtained either during forays outside the immediate hunting territories or derived
through trade with neighboring people.

Manufacturing Techniques. — The range of techniques employed in making
bone artifacts at Klithi is somewhat limited. Fortuitous splinters of long bones,
probably obtained when breaking bone for marrow extraction, were modified into
artifacts by means of simple longitudinal scraping. The most efficient method of
scraping bones to shave off excess material is to use the stout edge of a burin
facet. Burins are in the Klithi lithic assemblage, although they are not extremely
common (Adam, 1989:240). Other tools, including the edges of unretouched
flakes and the dorsal ridges of flakes, can also be used, but are normally less
effective.

Because most tools were made by scraping splinters of broken bone, manufac-
turing debitage is not common. The only debitage that has been observed in this
collection consists of tabular offcuts produced when needles were finished (Fig.
4 A-C). After the needle shaft and point were shaped, the basal tab that was used
to hold the needle blank during manufacture was removed by making an annular
groove around the needle shaft where it intersected with the tab and then snapping
the tab off at that point (see description under Needles and Needle Debitage
heading). Evidence of the groove-and-snap technique was not documented among
any other bone artifacts, although it may have been used rarely to cut out long
awl blanks.

Sawing with a sharp stone blade was performed to make a series of short,
transverse notches on large mammal bones, the function of which is unknown.

Perforations were made by two methods: gouging and drilling. Gouging, a
technique observed only on perforated canine pendants, was done by repeatedly
incising back and forth in one spot with a pointed tool, such as a piercer, graver,
or tip of a burin, to form a short groove or trough. After a groove was made on
one side, the object was turned over and a similar trough was made on the op-
posite surface until the bone was very thin at the center of the trough. A small
hole could then be easily made by punching through the dividing wall with the
same stone tool. The opening was then enlarged and smoothed by reaming it out
with the piercer or by drilling.

Biconical drilling, sometimes using a very delicate drill, was employed in the
manufacture of the eyes of needles. Biconical drilling involves the rotation of a
stone drill until a pit is made through about half the thickness of the material.
The object is then turned over and a similar pit is drilled in the opposite side in
the same area until the two pits connect and a perforation is made through the
material. The opening can then be enlarged by reaming. The profile of the per-
foration, as the term biconical implies, is hourglass-shaped in outline. This is
because a stone drill is usually tapered, with the narrower part at the tip. Reaming
opens up the center and reduces the conical slope of the perforation, however.
The needles show that a flat surface was first prepared on the basal portion of
the shaft so that the drill would not slide off the narrow, rounded surface. It is
likely that a tiny pit was made with the tip of the drill by hand at first to “spot”
the drill so that it did not slip when rotated.

Light grinding with an abrader, probably made of local granular stone, was
performed on the basal end of one needle (field number C4618) to smooth its
ragged surface after the tab was removed by the groove-and-snap technique.

The variety of manufacturing techniques expressed in the bone artifacts from
Klithi is limited and most of the objects could have been made within a few

216

Annals of Carnegie Museum

VOL. 69

Scale = 30 min

Fig. 4. Needle off-cuts and needles (with cross-sections): A. needle off-cut (C5223), B. needle off-cut (B5208), C. unfinished needle with tab still attached
(B5008), D. needle shaft and tip (B4206), E. needle base (C4618), E needle (C6214), G. needle (A53), H. needle (C6031).

2000

Olsen — Klithi Rock Shelter Bone Industry

217

minutes. Only the needles demonstrate a considerable investment of time and
labor. Further details of the step-by-step manufacturing processes are provided in
the description of individual artifact types.

Awls. — The awls in this collection (Fig. 5) were made on fortuitous splinters
by sharpening one end by means of longitudinal scraping with a stone tool such
as a burin. This easy process leaves behind no by-products of manufacture, except
other fragments indistinguishable from unutilized products of marrow extraction
and very fine shavings, which are not likely to be recovered. Scraping traces
extend up the edges of the shaft on only five of the awls (field numbers A46,
B3208, C6712, C3023, and C5419). The others retain their unmodified fracture
edges and the original outline of the splinter except at the tip.

One awl (Fig. 6) was shaped extensively, however. It was made on a metatarsal
of a ruminant (ibex/chamois-sized) and retained a small portion of the proximal
articular surface at the base. The narrow shaft was scraped longitudinally over its
entire surface, giving it an oval cross section and straight, gradually tapering sides.

Artifacts made on splinters that are only modified slightly are generally asym-
metric and irregular in shape. In the case of awls, it is mainly the tip, i.e., the
functional part, which is limited by certain size and shape constraints and therefore
receives the most attention during manufacture. Previous research (Olsen, 1984)
has shown that the optimal awl tip for working hides or making normal- sized
coiled baskets has a round to oval cross section and measures less than 4 mm in
diameter (measured at a standard point 5 mm up from the end). For the seven
awls from Klithi from which measurements could be taken, the average tip width
is 2 mm and the average thickness is 1.8 mm (Table 2). None exceeds 2.5 mm
in diameter. This means that these artifacts are too delicate to have been used as
projectile points, but would have been quite suitable for piercing soft hides or for
weaving fine coiled baskets. Seven of the awls exhibit use polish at the tip and
partially up the shaft along the edges. Polish on bone tools has not proven to be
reliably characteristic for particular contact materials.

Awls from Kastritsa were also made on splinters, but because red deer bones
served as the common raw material, overall dimensions and cortical thicknesses
are often much greater for the Kastritsa awls than for the small splinter awls made
on ibex or chamois bones at Klithi.

Needles and Needle Debitage. — The needles from Klithi (Fig. 4) were made
in a way that is significantly different from those common in other European and
Near Eastern paleolithic to neolithic sites (Stordeur, 1977, Stordeur-Yedid, 1978).
The more common method consisted of making two semi-parallel grooves in a
long bone and then removing a carefully-shaped blank by means of the groove-
and-snap technique (also known as the groove- and- splinter technique).

Needles at Klithi, however, were made from fortuitous splinters of slender
outline obtained by shattering long bones (Fig. 7A). The method of manufacture
involved scraping the splinter with longitudinal strokes down all sides with a
burin or similar tool until it had the proper morphology (Fig. 7B). By retaining
a broad tab at the base of a needle blank, the worker could grip the piece firmly
while scraping down the narrow shaft with the edge of a burin facet. This was
done so the needle could be shaped from base to tip without the fingers getting
in the way. It also eliminated the difficulty of trying to hold a thin, delicate needle
while scraping it with considerable pressure. When the needle shaft had been
properly shaped so that its sides were straight and slightly tapered and its cross
section was round to oval, the tab was removed from the base by sawing an

218

Annals of Carnegie Museum

VOL. 69

Fig. 5. — Awls: A. C1014, B. B5007, C. C7007, D. B5608, E. B4810, F. awl made on ibex/chamois tibia, C3023.

2000

Olsen — Klithi Rock Shelter Bone Industry

219

Scale = 30 mm

Fig. 6. — Two views of large awl made on an ibex/chamois metatarsal (C6712).

annular groove around the shaft and snapping it in two (Fig. 7C). This produced
a tabular off-cut which retained the rough edges of the original splinter. Two of
these off-cuts (Fig. 4 A, B) were recovered, as well as the base of an unfinished
needle with the tab still attached (Fig. 4C). Each tab exhibits traces of the termini
of longitudinal scraping facets and annular incisions made by the groove-and-
snap technique that frees the tab from the needle. The presence of manufacturing
debitage and an unfinished piece indicates that needles were made on site at Klithi.

220

Annals of Carnegie Museum

VOL. 69

Table 2. — Awl Tip Dimensions.

Artifact #

Tip Width (mm)

Tip Thickness (mm)

A46

2.00

1.80

B3005

1.60

1.60

B5007

1.50

1.50

B5805

1.80

1.80

C1014

2.10

2.00

C6026

2.50

1.70

C6712

2.50

2.00

Mean

2.00

1.80

Evidence for the removal of a tab by the groove- and^ snap technique is visible on
the base of a needle from Kastritsa, as well.

As the manufacturing process continued, the base of the needle was abraded
to smooth away the rough manufacturing traces left when the tab was removed.
Direct evidence of abrading is visible on one needle base (Fig. 4E), but is not
detectable on others because use polish has obliterated any possible trace of it.

Before the eye was made, two opposing surfaces near the base were scraped
flat to ease spotting of the drill (Fig 7D). This is clearly demonstrated by the
scraping facets on the bases of two needles (A53 and C4618) (Fig. 4G).

One unfinished (Fig. 4G) and three finished needles (Fig. 4E, F, H), retaining
all or part of their eyes, show that these perforations were made by means of
biconical drilling. Very fine holes were made by first drilling on one side of the
base until the middle was reached and then turning the needle over and drilling
on the opposite side until the perforations met (Fig. 7E). The unfinished needle

Scale = 30 mm

ABC

Fig. 1 .■ — Needle manufacturing stages: A. an appropriately-shaped unmodified splinter is selected, B.
the needle preform is shaped by longitudinal scraping, C. the tabular off-cut is removed by the groove-
and-snap technique, D. the base is ground and the basal shaft is flattened to spot the drill where the
eye will be made, E. the eye is made by biconical drilling.

2000

Olsen — Klithi Rock Shelter Bone Industry

221

Table 3. — Needle Tip Dimensions.

Artifact #

Tip Width (mm)

Tip Thickness (mm)

B4206

1.20

1.20

C6031

1.50

1.50

C6214

1.60

1.20

Mean

1.40

1.30

(Fig. 4G) was apparently abandoned because it was drilled too close to the edge
from both sides and the two holes are misaligned.

One large needle (Fig. 4H) was recovered nearly intact. Its length is about 7
cm, basal width is 3.5 mm, and basal thickness is 1.2 mm. The tips of three
needles are preserved and could also be measured (Table 3). The average tip
width and thickness for needles is less than that of splinter awls.

Most of the needles show some use polish at the tip and along the shaft. One
needle tip (field number C6214) has a polished shaft and overlying scraping facets
at the end indicating that it was resharpened after use. Handling polish on the
base of the unfinished needle (Fig. 4G) suggests that this tool was recycled into
an awl when drilling of the eye failed.

Notched Bones. — An interesting, but perplexing group of artifacts from Klithi
consists of four long-bone shaft fragments, each of which has one or more rows
of short, transversely incised lines (Fig. 8D-G). In all of the examples, the surface
of the bone was first prepared by longitudinal scraping to smooth it and remove
rugosities. Transverse notches or grooves were then made by sawing with a sharp
stone flake or blade. All the fragmentary pieces have from eight to ten notches
preserved per row, but there is no way to determine how many notches a whole
artifact would have had. One of the fragments (Fig. 8D), made on an ibex/chamois
metapodial, has two parallel rows of incisions. Powdered red ocher was rubbed
into the notches on two of the specimens (Fig. 8D-E). Three of the notched bones
were charred brown to light gray (Fig. 8D-F).

Only one of these artifacts bears a polish on the notched surface (Fig. 8F).
Because the notches are not worn in three of the objects and are filled with
pigment in at least two, it is unlikely that they served as musical sounding rasps
(Queen, 1978; Olsen, 1979; Lund, 1981). They may, instead, have been parts of
ornaments (none are complete) or perhaps were tallies of some sort. The pigment
could have been used for decorative purposes or simply to make the incisions on
the clean white bone more visible.

Notched bones are found in the upper paleolithic of France beginning with the
Chattelperonian and occur in the Near East at such sites as Mugharet el Kebara
(Garrod, 1954:177), the terminal Aurignacian of ha-Yonim (Davis, 1974), the
upper paleolithic of Ksar Akil (Tixier, 1974), and the Kebaran of Jiita (Copeland
and Hours, 1977).

No notched bones were identified among the Kastritsa faunal material.

Red Deer Canine Pendants. — Four perforated red deer canines were found in
the Klithi excavations (Fig. 8A-C). These were made by first gouging a shallow
trough in the tooth root on both sides, followed by punching a small hole through
the thinnest part of the wall between the grooves, and reaming or drilling the
perforation from both sides to open it up to the desired diameter (2.7-4 mm).

Two of the teeth (Fig. 8B, C) show polish that indicates that a thong or cord
passed through the hole and rubbed on the top of the perforation’s rim as the

222

Annals of Carnegie Museum

VOL. 69

Fig. 8.— Red deer canine pendants (A-C) and notched bones (D-G): A. B2010, B. C1421, C. C6215, D. C6713, E. A62, E A39, G. 3208.

2000

Olsen — Klithi Rock Shelter Bone Industry

223

tooth hung downward. The teeth could have either been attached to a necklace
or sewn to clothing or bags.

Deer canines were significant objects across Europe and the Near East in pre-
historic times. Generally only occurring in male deer, they were apparently quite
prized by a number of cultures. Kastritsa produced at least ten canine pendants
and many more unmodified canines, but there red deer is the dominant animal in
the faunal assemblage. At El Wad, in Israel, a Natufian headdress included bone
beads carved to resemble deer canines — a kind of counterfeiting of a rare com-
modity (Garrod and Bate, 1937). Similar globular bone beads and genuine per-
forated canines have been found at ha-Yonim Cave (Bar Yosef and Tchemov,
1970).

At Klithi, it seems peculiar to find four of these objects because antler and deer
bones are rare in the faunal assemblage. What this may suggest is that some of
the occupants of Klithi ventured outside the gorge on hunting forays, as part of
their seasonal round, perhaps downstream on the more open flood plain of the
Voidomatis where red deer would find a more suitable habitat. Otherwise, the
canine pendants may represent trade objects from other cultures in neighboring
regions. These pendants would have been “curated” simply by wearing them,
perhaps for years before they finally broke or were dropped.

Worked Antler Tine. — Only one piece of worked antler appears in this collec-
tion. It consists of a tine fragment that has been scraped down to sharpen the tip.
The most logical use of this artifact would be as a pressure flaker, but unfortu-
nately the end that would have borne traces of use wear has been broken off.

Conclusions

It is useful to make comparisons between the Klithi bone artifacts and those
from the nearby site of Kastritsa. The bone technology of the two sites is basically
the same, with the modifications being made on splinters by scraping with a stone
tool. Awls, needles, and perforated canines occur at Kastritsa and Klithi. In ad-
dition, however, Kastritsa yielded large quantities of bipointed antler projectile
points and grooved ruminant incisors that could have been sewn on clothing or
worn as pendants. Because many of the Kastritsa awls are made on red deer long
bones and occasionally retain an articular condyle, they are significantly larger
than the splinter awls from Klithi made from smaller chamois or ibex bones. The
differences between the two bone assemblages reflect availability of specific raw
materials (i.e., red deer incisors, antlers, and bones) at Kastritsa, rather than any
real technological dissimilarities. No substitutes for antler points were made from
ibex or chamois long bones at Klithi, perhaps because bone is less resilient and
more susceptible to shattering on impact than antler (MacGregor and Curry, 1983;
Arndt and Newcomer, 1986). The cortical thickness of ibex and chamois bones
is also at the lower end of the range that is normally required to make a reliable,
reusable projectile point. A small bone bipoint was found at Kastritsa, but this
may have been a gorge or barb from a composite weapon used for fishing or
hunting.

When the Kastritsa and Klithi lithic assemblages were compared (Adam, 1989:
250-251), there were numerous technological and typological differences, as well
as similarities. Some of the variance can be attributed to the fact that most of the
lithic material from Klithi consisted of small, brittle pieces of flint from the Voi-
domatis river bed, while more variety in size and quality was available at Kas-

224

Annals of Carnegie Museum

VOL. 69

tritsa. This difference in lithic raw material parallels the differences observed in
the bone artifact assemblages, where antler was available at Kastritsa. Affinities
are strong in both bone and stone assemblages. However, several stone tool types
found at Kastritsa are missing in the Klithi assemblage, and there are some im-
portant details that differ in lithic production techniques (Adam, 1989; Bailey,
1997). These significant differences suggest that contemporaneous levels of the
two sites represent two related cultures, rather than the same culture.

The collection from Klithi, although small, has already yielded information not
only about bone artifact technology, but also regarding how adjustments are made
to suit the availability of raw material and how some of the stone tools, such as
burins, piercers, and drills, may have been employed. The presence of red deer
canines and a fragment of worked antler suggests that some artifacts or raw ma-
terial may have been brought in from hunting forays to other areas or through
trade. Manufacturing debitage from needles indicates that these fine objects were
made at Klithi.

Analysis of small collections of bone artifacts from upper paleolithic sites in
this poorly known region is important in establishing the extent of contact between
Greece and its neighbors at this time. Almost all manufacturing techniques for
bone industries were established in the Near East and western Europe during the
upper paleolithic, so learning how the industry was advancing in an intermediate
location like Epirus can be quite instructive. This relatively isolated region shows
only minimal use of the groove-and-snap technique compared to its neighbors,
but also documents a rather innovative method of needle manufacture.

The worked bones at Klithi are derived from 3 m of cultural layers that date
to between 21,300-12,600 B.P. (Gowlett et al., 1997). Tool types are consistent
through time, as far as can be discerned from this small sample. However, most
of the material in this collection comes from the upper five levels, dating to
between about 17,400 and 12,600 B.P. This makes the assemblage comparable to
the Epi-Gravettian in Italy. By the Gravettian, perforated red deer canines appear
at the Grotta del Broion in the Berici Hills near Mossano, Italy (Broglio, 1995:
57). Notched awls and other bones have been reported in the Sauveterriano
(9,500-8,200 B.P), or middle Mesolithic, which follows the Epi-Gravettian, at
Romagnano III, 10 km south of Trent, Italy (Broglio, 1995:104-108). In addition,
Romagno III produced a number of tools made on red deer antler and bone,
including axes and spatulas. During the French Magdalenian, c. 14,000-12,500
b.c., bone needles were made using the groove-and-splinter technique (Stordeur-
Yedid, 1978).

Needles manufactured in the fashion documented at Klithi are not reported in
either western Europe or the Near East. This technique appears to be an indige-
nous and isolated trait of the Klithi culture.

Splinter awls like those from Klithi occurred in western Europe and the Near
East from the Aurignacian on, but these are nearly universal and are not useful
cultural markers. Antler tine flakers are even more ancient and are also too ubiq-
uitous to shed light on specific cultural identity. Notched bones were found in the
Near East at upper paleolithic sites like Ksar Akil, Lebanon, (Tixier, 1974) and
later. Perforated canines and counterfeit imitations appear somewhat later than
Klithi in the Natufian at El Wad (Garrod and Bate, 1937) and ha-Yonim Cave
(Bar Yosef and Tchemov, 1970), in Israel.

This analysis has demonstrated that even small collections of bone artifacts can
contribute to the accumulated knowledge of prehistory. Bone artifacts are valuable

2000

Olsen — Klithi Rock Shelter Bone Industry

225

indicators of both shared and unique technological innovations. Close inspection
demonstrates that these artifacts reveal much about the methods and step-by-step
processes involved in their manufacture. This, in turn, assists in assigning uses to
implements made of stone or other materials. In some cases use wear can sup-
plement overall morphology to indicate probable function (Olsen, 1984). Identi-
fying the likely material on which bone artifacts were used contributes indirect
evidence for the use of perishable materials, like worked hides or baskets. Because
osseous artifacts may be curated far longer than unmodified bones, they can pro-
vide important information about the access a people may have had to species
not well-represented as food refuse at a seasonal camp site like Klithi. Such was
the case with the piece of antler and perforated canines of red deer. Finally, it is
useful for the faunal analyst to understand how the prehistoric people were ex-
ploiting raw materials obtained from the animals that they hunted. These data
supplement the evidence inferred from cutmarks on bones that indicate the taking
of hides, tendons, meat, horns, antlers, hoofs, and other materials.

Acknowledgments

I would like to express my gratitude to Geoffrey Bailey for giving me the opportunity to analyze
the bone artifacts from Klithi and to participate in the excavations of the site. Funding for the Klithi
excavations was provided by grants from the British Academy, the British School at Athens, the
National Geographic Society and the Society of Antiquaries.

Literature Cited

Adam, E. 1989. A Technological and Typological Analysis of Upper Palaeolithic Stone Industries of
Epirus, Northwestern Greece. B.A.R. International Series 512, Oxford, United Kingdom.

Adam, E., and E. Kotjabopoulou. 1997. The organic artefacts from Klithi. Pp. 245-258 in Klithi:
Palaeolithic Settlement and Quaternary Landscapes in Northwest Greece (G. Bailey, ed.). Mc=
Donald Institute for Archaeological Research, Cambridge, United Kingdom.

Arndt, S. L., and M. H. Newcomer. 1986. Breakage patterns on prehistoric bone points: An exper-
imental study. Pp. 165-173, in Studies in the Upper Palaeolithic of Britain and Northwest Europe
(D. A. Roe, ed.). B.A.R. International Series 296, Oxford, United Kingdom.

Bailey, G. N. 1997. Klithi: Palaeolithic Settlement and Quaternary Landscapes in Northwest Greece.
McDonald Institute for Archaeological Research, Cambridge, United Kingdom.

Bailey, G. N., P. L. Carter, C. S. Gamble, and H. P. Higgs. 1983. Asprochaliko and Kastritsa:
Further investigations of palaeolithic settlement and economy in Epirus (north-west Greece). Pro-
ceedings of the Prehistoric Society, 49:15-42.

Bailey, G. N,, J. Lewin, M. G. Macklin, and J. C. Woodward. 1990. The “older fill” of the Voi-
domatis valley, north-west Greece and its relationship to the palaeolithic archaeology and glacial
history of the region. Journal of Archaeological Science, 17:145-150.

Bailey, G. N., G. King, and D. Sturdy. 1993. Active tectonics and land-use strategies: A palaeolithic
example from northwest Greece. Antiquity, 67:292-312.

Bar-Yosef, O., and E. Tchernov. 1970. The Natufian bone industry of ha-Yonim Cave. Israel
Exploration Society, 20:141-150.

Bottema, S. 1974. Late Quaternary Vegetation History of Northwestern Greece. Groningen Univer-
sity Press, Groningen, The Netherlands.

Broglio, a. 1995. Guide Archeologiche: Preistoria e Protostoria in Italia, Vol. 4, Paleolitico, Me-
solitico e Neolitico dellTtalia Nord-orientale. A.B.A.C.O. Edizioni, Forli, Italy.

Copeland, L., and F. Hours. 1977. Engraved and plain bone tools from Jiita (Lebanon) and their
early Kebaran context. Proceedings of the Prehistoric Society, 43:295-301.

Dakaris, S. L, E. S. Higgs, and R. W. Hey. 1964. The climate, environment and industries of stone
age Greece: Part I. Proceedings of the Prehistoric Society, n.s., 30:199-243.

Davis, S. 1974. Incised bones from the Mousterian of Kebara Cave (Mount Carmel) and the Auri-
gnacian of ha-Yonim Cave (western Galilee), Israel. Paleorient, 2:181-182.

Garrod, D. a. E. 1954. Excavations at the Mugharet Kebara, Mt. Carmel, 1931: The Aurignacian
industries. Proceedings of the Prehistoric Society, n.s., 20, pt. 2:155-192.

226

Annals of Carnegie Museum

VOL. 69

Garrod, D. a. E., and D. M. A. Bate. 1937. The Stone Age of Mount Carmel, I. Clarendon Press,
Oxford, United Kingdom.

Gowlett, J., R. Hedges, and R. Housley. 1997. Klithi; the AMS Radiocarbon dating programme
for the site and its environs. Pp. 27-39, in Klithi: Palaeolithic Settlement and Quaternary Land-
scapes in Northwest Greece (G. Bailey, ed.). McDonald Institute for Archaeological Research,
Cambridge, United Kingdom.

Higgs, E. S., and C. Vita-Finzi. 1966. The climate, environment and industries of stone age Greece:
Part III. Proceedings of the Prehistoric Society, n.s., 33:1-29.

Higgs, E. S., C. Vita-Finzi, D. R. Harris, and A. E. Fagg. 1967. The climate, environment, and
industries of stone age Greece: Part III. Proceedings of the Prehistoric Society, n.s., 33:1-29.

Lund, C. 1981. The ethnomusicology of Scandinavia. World Archaeology, 13:246-265.

MacGregor, A. G., and J. D. Currey. 1983. Mechanical properties as conditioning factors in the
bone and antler industry of the third to the thirteenth century A.D. Journal of Archaeological
Science, 10:71-77.

Olsen, S. L. 1979. A study of bone artifacts from Grasshopper Pueblo, AZ P:14:l. The Kiva, 44:
341-373.

. 1984. Analytical Approaches to the Manufacture and Use of Bone Artifacts in Prehistory.

Unpublished Ph.D. Thesis, Institute of Archaeology, University of London, London, United King-
dom.

Palma di Cesnola, A. 1976 Le leptolithique archaique en Italie. Pp. 66-99 in Perigordien et Grav-
ettien en Europe. (B. Klima, ed.). Colloque XV, U.L, S.P.P. IX Congress, Nice.

Prentice, I. C., J. Guiot, and S. P. Harrison. 1992. Mediterranean vegetation, lake levels and pa-
laeoclimate at the last glacial maximum. Nature, 360:658-660.

Queen, L. 1978. Southwestern Indian musical instruments. The Smoke Signal, 35:1-23.

Stordeur, D. 1977. La fabrication des aiguilles a chas, observation et experimentation. Pp. 251-256,
in Methodologie Appliques a ITndustrie de POs Prehistorique (568) (H. Camps-Fabrer, ed.).
C. N. R. S., Paris, France.

Stordeur- Yedid, D. 1978. Aiguilles (les) a chas au Paleolithique. Gallia Prehistoric, Supplement 13.
C. N. R. S., Paris, France.

Tixier, j. 1974. Poingon decore du paleolithique superieur a Ksar ‘Aquil (Liban). Paleorient, 2:187-
192.

Willis, K. J. 1989. Late Quaternary vegetation history of Epirus, northwest Greece. Unpublished
Ph.D. Dissert., Cambridge University, Cambridge, United Kingdom.

ANNALS OF CARNEGIE MUSEUM

VoL. 69, Number 4, Pp. 227-280

29 November 2000

REVIEW OF MIOCENE (HEMINGFORDIAN TO CLARENDONIAN)
MYLAGAULID RODENTS (MAMMALIA)

FROM NEBRASKA

William W. Korth^

Research Associate, Section of Vertebrate Paleontology

Abstract

Mylagaulid rodents have been widely recognized in faunas from the Miocene of North America for
over 125 years, but a thorough review of the family at the species level has not been done in nearly
a century. A large sample of specimens is now available for study from throughout the Miocene
section in Nebraska. This sample permits a reexamination of the systematics and phylogeny of this
family. It also has allowed for the study of other unique attributes of this family such as sexual
dimorphism.

Three subfamilies of mylagaulids are recognized: Mylagaulinae, Mesogaulinae, and Promylagauli-
nae. The Promylagaulinae are not considered within the scope of this study. The Mesogaulinae are
limited to the nominal genus and restricted to the early Hemingfordian. Only two species of Meso-
gaulus are recognized, M. ballensis Riggs, and M. paniensis (Matthew); only the latter is known from
Nebraska. The Mesogaulinae appear both temporally and morphologically ancestral to the later, more
advanced mylagaulines.

The Mylagaulinae is comprised of six genera; Mylagaulus Cope, Ceratogaulus Matthew, Hesper-
ogaulus Korth, Umhogaulus n. gen., Pterogaulus n. gen., and Alphagaulus n. gen. Epigaulus Gidley
is considered a junior synonym of Ceratogaulus. All but Hesperogaulus are known from Nebraska.
Alphagaulus is the most primitive and contains species from the late Hemingfordian and early Bar-
stovian. Two species are present in the Nebraska record, A. vetus (Matthew) and a new species, A.
tedfordi. Two other previously described species are referred to this genus, A. pristinus (Douglass),
and A. douglassi (McKenna). The remaining four genera of mylagaulines from Nebraska represent
distinct lineages, all of which begin in the Barstovian and continue into the Hemphillian except
Umhogaulus, which ranges from the late Hemingfordian into the Barstovian only. Three new species
are recognized among these genera: Umhogaulus galushai, Ceratogaulus anecdotus, and Pterogaulus
harharellae . Mylagaulus monodon Cope is referred to Umhogaulus, and M. laevis Matthew and “M. ”
cambridgensis Korth are referred to Pterogaulus. Each of the recognized lineages increases in size
through time except Mylagaulus and Umhogaulus, which decrease in size. It is evident that the pres-
ence of nasal horns is not a sexually dimorphic character but defines Ceratogaulus as a unique genus
as other genera are defined by other forms of ornimentation of the nasal bones.

There is also a biogeographic limitation of these genera. Ceratogaulus, Umhogaulus, and Ptero-
gaulus are restricted to the northern Great Plains; Mylagaulus is found only in Florida, northwestern
Kansas, and possibly Nebraska. Hesperogaulus is present in Barstovian to Hemingfordian faunas from
the Great Basin only.

Key Words: Mylagaulidae, Miocene, Biostratigraphy, Phylogeny, Sexual dimorphism, Fossorial

Introduction

Cope (1878) first named Mylagaulus sesquipedalis from the Sappa Creek fauna
of Kansas. This species was represented by (AMNH 8329, later to be named
as the holotype) and an isolated incisor and P4 (AMNH 8330). A few years later.
Cope (1881fl) named an additional species, Mylagaulus monodon, from the Bar-
stovian Driftwood Creek fauna of Nebraska (see Voorhies [1990^] for age of
fauna). The latter species was represented by a partial mandible with P4 (AMNH

* Rochester Institute of Vertebrate Paleontology, 265 Carling Road, Rochester, New York 14610.
Submitted 10 September 1999.

227

228

Annals of Carnegie Museum

VOL. 69

8327). Later in the same year. Cope (1881^) included these unusual rodents in
their own family, the Mylagaulidae. Riggs (1899) named a second genus of my-
lagaulid, Mesogaulus, based on a new species, M. ballensis, from the Heming-
fordian North American Land Mammal Age of Montana.

The first cranial material of a mylagaulid to be described was from Colorado
in beds that are now known to be Barstovian in age. Matthew (1901) originally
identified this specimen (AMNH 9043) as Mylagaulus monodon but later named
it as the holotype of a new species, M. laevis Matthew (1902). The year after the
description of the first skull of Mylagaulus, Matthew presented a review of the
family. In this review, he described the skull of a mylagaulid from the same
horizon as that of the skull of M. laevis that had “a pair of large connate processes
on the nasals ...” (Matthew, 1902:291). Matthew named a new genus and spe-
cies for this skull, Ceratogaulus rhinocerus. He suggested the possibility that the
unique horn cores of this new skull were a male character of Mylagaulus. He
rejected this idea in the end because there is no evidence of any marked sexual
dimorphism in any other rodents, fossil or Recent (also see Matthew, 1924).

Since Matthew’s (1902) review of the Mylagaulidae, five additional genera of
mylagaulids have been named. Cranial material is known for four of them: Pro-
mylagaulus (McGrew, 1941), Trilaccogaulus (Nichols, 1976; Korth, 1992), Epi-
gaulus (Gidley, 1907; Hibbard and Phillis, 1945), and Hesperogaulus (Shotwell,
1958; Korth, 1999a). In addition, the skulls of several species of Mylagaulus and
Mesogaulus have been described (Matthew, 1924; McKenna, 1955; Shotwell,
1958; Fagan, 1960; Wilson, 1960; Wahlert, 1974; Galbreath, 1984; Munthe,
1988). Of the genera with known cranial material, only Epigaulus has the horn
cores on the nasal bones as in Ceratogaulus.

It has been the practice in the past to refer isolated cheek teeth and mandibles
of mylagaulids from Barstovian to Hemphillian times to a species of Mylagaulus
if no cranial material is associated. The identification of either horned genus,
Ceratogaulus or Epigaulus, is only done when cranial material with evidence of
horns is known. Previously, no dental morphology was used to separate the homed
species from the hornless species. In fact, the assignment of the hornless skulls
to the genus Mylagaulus is completely arbitrary because the holotype of the type
species of the genus is an isolated P"^. Hibbard and Phillis (1945) even suggested
that the Clarendonian species from Kansas, Epigaulus minor (with known pres-
ence of horns) and Mylagaulus sesquipedalis (type species of the genus), may be
synonyms because of their dental similarity and the lack of cranial material of
the latter. This suggests that the type species of Mylagaulus might have had horn
cores, eliminating the previously defined generic difference between Mylagaulus
and the homed genera.

In Matthew’s (1902) review, he also noted the fossorial adaptations of the
postcranial skeleton of mylagaulids. All subsequent descriptions of postcranial
material of mylagaulines have verified Matthew’s observations (Gidley, 1907;
Hibbard and Phillis, 1945; Fagan, 1960).

In the century since Matthew’s (1902) review of the Mylagaulidae, numerous
specimens have been collected from Miocene-aged horizons in Nebraska. The
bulk of these were first collected by Ted Galusha and Morris Skinner of the Frick
Laboratories in the 1930s through the 1970s. These collections are now housed
at the American Museum of Natural History. More recently, beginning in the
1970s, extensive collections of Miocene manunals have been made by M. R.
Voorhies and field parties from the University of Nebraska State Museum. These

2000

Korth — Miocene mylagaulid rodents from Nebraska

229

extensive collections contain numerous cranial, dental, and postcranial specimens
that are the basis for this study.

The definition of the Promylagaulinae used below follows that of Rensberger
(1980). Cranial features of promylagaulines are based on those described by
McGrew (1941), Wahlert (1974), Nichols (1976), Munthe (1988), and Korth

(19991?).

Methods

Cranial Measurements and Indices. — The placement and orientation of skull
measurements are presented in Figure 1. Horn placement index (NHI) is calcu-
lated by dividing the distance from the apex of the nasal horn to the anterior edge
of the nasal by the total length of the skull. Postorbital process size index (POI)
is calculated by dividing the length of the postorbital process on one side of the
skull by the minimum width of the postorbital constriction. Cranial width index
(W/L) is calculated by dividing the posterior width of the skull by the antero-
posterior length of the skull. The angle of the occipital bone (OA) is measured
with respect to the plane of the palate. Cranial indices for all mylagaulid species
for which they are known are presented in Table 1.

Dental Measurements and Terminology. — Measurements of the premolars of
mylagaulids used here represent the maximum width and length of the tooth. The
occlusal length was used only in specimens in very late stages of wear where the
occlusal measurements are the maximum. Abbreviations for dental dimensions
appearing in tables are as follows: a-p, anteroposterior length; tr, transverse width.

The terminology for the fossettes (-ids) of mylagaulid premolars is difficult
because different schemes have been proposed by virtually everyone who has
worked on this family (Riggs, 1899; McGrew, 1941; Shotwell, 1958; Rensberger,
1979; Munthe, 1988; Korth, 1994). The difficulty with using any system of nam-
ing the fossettes is that the homologies of the fossettes are difficult to determine
except in very primitive forms that have only a few fossettes that are traceable
from the unworn surface of the tooth. Here, the fossettes of the cheek teeth will
be referred to only by their position on the tooth (e.g. posterolingual, etc.). The
only fossette that appears to be significant in terms of the evolutionary changes
in mylagaulids is the anterocentral (or anterobuccal) fossette of This fossette
is forked anteriorly and the separation of the branches of the forked end is a
significant character that is used to define genera. In the following text this fossette
will be referred to as the parafossette. Since the number of fossettes (-ids) on the
premolars varies somewhat within a given sample of any one species, the mean
number of fossettes, as well as the modal number, are used for comparisons
between species and populations (Table 2). Any reference to cusps or styles on
the premolars is taken from the terminology of Wood and Wilson (1936).

Reference to early, middle, or late stages of wear on the premolars will be used
as defined by Korth (1999a); teeth that have been worn to a level within the first
20% of the presumed total height of the tooth are considered to be in early wear,
and those in the presumed last 20% of the total crown height are considered to
be in a late stage of wear. All others are considered to be in middle or moderate
wear (central 60% of the crown height). Clearly, these are approximations. How-
ever, the pattern of premolars in early and late wear is distinct (see section on
variation in Conclusions); lateral sides of the tooth tilt toward the center of the

230

Annals of Carnegie Museum

VOL. 69

2000

Korth — Miocene mylagaulid rodents from Nebraska

231

Table 1. — Cranial indices of mylagaulids.

W/L POI NHI OA

Trilaccogaulus lemhiensis
Galbreathia novellas
Mesogaulus paniensis
Alphagaulus vetus
Alphagaulus pristinus
Alphagaulus tedfordi
Umhogaulus galushai
Umbogaulus monodon
Ceratogaulus rhinocerus
C. cf. rhinocerus
Ceratogaulus minor
Ceratogaulus anecdotus
Ceratogaulus hatcheri
Pterogaulus laevis
Pterogaulus sp.
Pterogaulus barbarellae
Hesperogaulus gazini
Hesperogaulus wilsoni

0.46

0.27

0.52

0.24

0.82

0.32

0.77-0.83

0.26-0.36

0.31

1.22

0.25

0.87

0.35

0.87

0.35-0.36

0.95

0.19

0.85-0.88

0.12-0.15

0.93

0.12

1.00

0.12

0.95

0.11

0.90-1.04

0.39-0.41

0.83-0.90

0.34-0.46

0.85-1.03

0.40-0.57

0.97

0.31-0.35

0.80-0.87

0.22-0.25

0.19-0.21

110°

90°

90°

60°-75°

79°

90°

75°

75°

70°

0.16-0.22

67°-68'

0.29

62°

0.30

60°

0.38

63°

68°-75‘^

54°-60'

52°-60'

55°-5r

50°

occlusal surface in early wear, and strongly away from the center of the tooth in
late wear.

Abbreviations for Institutions. — AMNH, American Museum of Natural History,
New York, New York; CM, Carengie Museum of Natural History, Pittsburgh,
Pennsylvania; FAM, Frick Collections of the AMNH, New York, New York;
FMNH, Field Museum of Natural History, Chicago, Illinois; UCMP, University
of California Museum of Paleontology, Berkeley, California; UNSM, University
of Nebraska State Museum, Lincoln, Nebraska; UOMNH, University of Oregon
Museum of Natural History, Eugene, Oregon; USNM, United States National
Museum of Natural History, Smithsonian Institution, Washington, D. C.

B lOSTR ATIGRAPH Y

The biostratigraphy of the middle and later Tertiary of Nebraska used below
generally follows that presented by Tedford et al. (1987) for the Hemingfordian
through Clarendonian. However, there are two instances where this paper deviates
from the latter work. First, Voorhies (1990fl) argued convincingly that the Bar-
stovian-Clarendonian boundary, based on the Nebraska stratigraphic section,
should be at the base of the Burge Member of the Valentine Formation as was
initially intended in the original definition of the Clarendonian (Wood et aL, 1941).
This differs from Tedford et al. (1987), who included the fauna of the Burge
Member in the latest Barstovian (late-late Barstovian). In the following text, Voor-
hies’ interpretation will be used and the Burge fauna will be included as the early
Clarendonian (Fig. 2).

Fig. 1. — Cranial measurements of mylagaulids. A. Total length. B. Total posterior width. C. Width of
postorbital constriction. D. Distance of apex of nasal horn from anterior end of nasal. E. Length of
postorbital process. F. Angle of occipital. Cranial indices used in text: NL = D/A; POI = E/C; W/
L = B/A.

232

Annals of Carnegie Museum

VOL. 69

Table 2. — Number of fossettes (-ids) on premolars of my lagauUds.

X

Mode

Range

Mesogaulus paniensis

p4

5.0

5

5-5

P4

5.2

5

4-6

Mesogaulus ballensis

p4

—

—

—

P4

4

—

—

Alphagaulus vetus

p4

6.7

7

5-9

P4

5.9

5

5-9

Alphagaulus tedfordi

p4

6

—

—

P4

6

—

—

Alphagaulus pristinus

p4

6.0

6

6-6

P4

5.3

5

5-6

Alphagaulus douglassi

p4

P

7

—

—

Umbogaulus galushai

^4

p4

8.3

9

7-9

P4

8.4

8

7-11

Umbogaulus monodon

p4

7.9

8

6-10

P4

7.3

7

5-9

Ceratogaulus . rhinocerus

P4

8.0

7

7-10

P4

8.2

7

8-10

Ceratogaulus cf. rhinocerus

P4

6.7

6-7

6-8

P4

7.1

7

5-9

Ceratogaulus minor

p4

6

—

—

P4

7

—

—

Ceratogaulus anecdotus

p4

7.5

7

6-9

P4

7.4

7

6-10

Ceratogaulus hatcheri

p4

8

—

—

P4

9

—

—

Pterogaulus laevis

p4

6.5

6

5-9

P4

5.7

5

5-7

Pterogaulus sp.

P'*

7.2

8

5-10

P4

7.0

6

6-11

Pterogaulus barbarellae

P4

7.8

8

6-10

P4

6.5

6

5-8

Pterogaulus cambridgensis

p4

7.3

8

6-9

P4

7.0

7

6-8

Second, Tedford et al. (1987:fig. 3) figured the “Sand Canyon Beds” as bios-
tratigraphically equivalent to the Olcott Formation. However, the former (based
on the fauna from Observation Quarry in Dawes County, Nebraska) appears to
be older than the fauna from the type area of the Olcott Formation in Sioux
County, and is assumed here to represent an earlier deposit.

The Miocene biostratigraphy of Wyoming was not presented by Tedford et al.
(1987), so the position and correlation of the Split Rock fauna (late Hemingfor-
dian) is based on the interpretation presented in Munthe (1988:fig. 3).

Systematic Paleontology

Order Rodentia Bowdich, 1821
Family Mylagaulidae Cope, 1881
Subfamily Mesogaulinae, new subfamily

Definition. — Intermediate-sized mylagaulids; first molar lost with eruption of
last permanent premolar; retained, but cheek teeth higher crowned than in
promylagaulines; rudimentary roots present on premolars at advanced stage of

2000

Korth — Miocene mylagaulid rodents from Nebraska

233

NALMA

STRATIGRAPHY

Clarendonian

late

Ash Hollow
Formation

Merritt Dam Member

middle

Cap Rock Member

early

Valentine

Formation

Burge Member

Barstovian

iate

Crookston Bridge Member

middle

Cornell Dam Member

early

Olcott Formation

“Sand Canyon Beds”

Hemingfordian

late

Sheep Creek Formation

early

Running Water Formation

Marsland Formation

Fig. 2. — Miocene stratigraphic sequence for the Nebraska section used in this paper. Width of units
drawn equally, not intended to represent absolute duration (time) or thickness of section.

wear; retains outline of mesostyle and parastyle and P4 retains outline of me-
tastylid (lost in mylagaulines); upper incisor with smooth anterior enamel surface
(grooved in mylagaulines); skull not as broad as in mylagaulines, but broader than
in promylagaulines (W/L = 0.80); occipital not anteriorly tilted (OA = 90°); sin-
gle sagittal crest on skull (doubled in mylagaulines); postcranial skeleton with
f os serial adaptations.

Included Genus.- — --Mesogaulus Riggs, 1899.

Mesogaulus Riggs, 1899

Type Species.— Mesogaulus ballensis Riggs, 1899.

Included Species.— Mesogaulus paniensis (Matthew, 1902).

Range. — Hemiegfordian of Montana, Colorado and Nebraska.

Diagnosis. — -Only genus of the subfamily.

Discussion. — The last published classification of mylagaulids at the species

234

Annals of Carnegie Museum

VOL. 69

level cited six species of Mesogaulus ranging from the Hemingfordian to Barsto-
vian (Korth, 1994:109). However, only two of those species are considered here
as species of this genus. Mesogaulus pristinus and M. proximus from the early
Barstovian of Montana (Douglass, 1903) are shown to be synonyms and are re-
ferable to a new genus of primitive mylagauline (see below). Mesogaulus nov-
ellus, from the Hemingfordian of Wyoming and Nebraska (Matthew, 1924; Black
and Wood, 1956; Munthe, 1988), was referred to a new genus of specialized
promylagauline, Galbreathia (Korth, 1999^). Finally, M. praecursor from the
Hemingfordian of Nebraska (Cook and Gregory, 1941) has been shown previously
to be a junior synonym of M. paniensis (Wilson, 1960). All of the conclusions
presented in the following discussions regarding Mesogaulus are based on the
only two species of the genus recognized here, M. ballensis and M. paniensis.

In his first description of Mesogaulus, Riggs (1899) noted that the morphology
of the genus was intermediate between Mylagaulus and an ancestral form. Me-
sogaulus has developed cranial, dental, and postcranial features that are clearly
mylagauline-like, but are more primitive than any true mylagaulines. The skull,
while being posteriorly broadened and generally more robust than that of pro-
mylagaulines (Wilson, 1960; Galbreath, 1984; Korth, \999b), has not attained the
degree of development of these features found in other mylagaulines. Similarly,
the dentition is advanced over that of promylagaulines in hypsodonty and com-
plexity of the occlusal surface, however, is still retained in adult individuals,
a promylagauline feature. Features of the limb bones are similarly developed with
the hypertrophy of the forelimb seen in later Tertiary mylagaulines (Galbreath,
1984; Korth 1999Z?).

Mesogaulus is a likely ancestor to later mylagaulines. Its age, early Heming-
fordian, is also intermediate between promylagaulines and mylagaulines (see
Korth, 1994 for age ranges of promylagaulines). The only exception is the late
occurrence of the derived promylagauline Galbreathia (Korth, 1999^).

Mesogaulus paniensis (Matthew, 1902)

(Fig. 3A, B, 4)

Mylagaulus paniensis Matthew, 1902.

Mesogaulus paniensis (Matthew) Cook and Gregory, 1941.

Mesogaulus praecursor Cook and Gregory, 1941.

Type Specimen. — AMNH 9361, partial mandible with incisor and P4 (Matthew,
1902:fig. 4).

Referred Specimens. — Additional topotypic specimens listed in Galbreath (1953:95) and Wilson
(1960:51-53); referred specimens from Nebraska cited in Cook and Gregory (1941:551) as Mesogau-
lus praecursor, and FAM 6551 1, nearly complete skull with complete dentition, associated mandibles,
and some postcranial bones.

Horizon and Locality. — Topotypic specimens from Quarry A, Martin Canyon
beds, Logan County, Colorado; referred specimens from “four miles north of
Agate” (Cook and Gregory, 1941:551), Marsland Formation, Sioux County, Ne-
braska; FAM 65511 from Runningwater Formation, Cottonwood Creek, Dawes
County, Nebraska.

Fig. 3. — Dentitions of species of Mesogaulus. A, B. Mesogaulus paniensis, FAM 65511. A. P^-
M2. B. P4, M2-M3. C. Holotype of M. ballensis, FMNH P 25223, P4.

2000

Korth — Miocene mylagaulid rodents from Nebraska

235

236

Annals of Carnegie Museum

VOL. 69

msc

2000

Korth — Miocene mylagaulid rodents from Nebraska

237

Age, — Early Hemingfordian (early Miocene).

Emended Diagnosis.-— Vim & or more fossettes (-ids) on P4, more than in M.
ballensis.

Description. — Galbreath (1953, 1984) and Wilson (1960) have fully described the dentition of M.
paniensis along with some skull material and postcranial elements. A nearly complete skull with
associated limb elements, FAM 65511, is clearly referable to M. paniensis. The skull is generally low,
and posteriorly broadened as in mylagaulines, but has not attained the degree of posterior expansion
seen in the latter (W/L = 0.82). The postorbital processes are intermediate in size (POI = 0.32) which
appears to be primitive for the family. The angle of the occipital is 90°, again a primitive condition,
and similar to promylagaulines. However, the positions of all cranial foramina are identical to those
of later mylagaulines (WaMert, 1974:fig. 13).

Dentally, the upper incisors of M. paniensis also lack the broad, shallow groove along their medial
border that is present in all mylagaulines, and a is retained in all adult individuals of M. paniensis
(this tooth is lost in all mylagaulines). The humeri of FAM 65511 are proportioned as in later myla=
gauliees and are clearly broader than in promylagaulines (Table 3).

Discussion.— In virtually all features of the skull, dentition, and postcranial
skeleton, M. paniensis is intermediate between promylagaulines and later, more
derived mylagaulines. Cook and Gregory (1941) believed that additional speci-
mens of Mesogaulus from Nebraska represented a new species, M. praecursor,
but Wilson (1960) demonstrated that this latter species was a junior synonym of
M. paniensis. Although Matthew (1902) argued against synonymy, it is possible
that the type species of the genus, M. ballensis Riggs (1899), is synonymous with
the much better represented M. paniensis because the former is known only from
the holotype, FMNH P 25223, which contains heavily worn cheek teeth (Fig. 3C).
This argument cannot be settled until additional topotypic material of M. ballensis
is recovered and described.

Subfamily Mylagaulinae Cope, 1881^

Definition.— mylagaulids; absent; last premolar greatly enlarged (more
than twice the size of molars); at least first molars lost with eruption of permanent
premolars; no evidence of roots on premolars; upper incisor with broad, shallow
groove along medial border of anterior surface of enamel; posterior width of skull
nearly equal to anteroposterior length of skull (W/L = 0.77-1.04); occipital tilted
anteriordorsally (OA < 75°) except in most primitive species; sagittal crest dou-
bled (= parasagittal crests); primitively, small boss of rugose bone present dorsally
on anterior end of nasal bones; fossorial adaptations of postcranial skeleton greatly
enhanced (ratio of distal width to length of humerus greater than 0.43).

Included Genera.— Mylagaulus Cope, 1878; Ceratogaulus Matthew, 1902;
Hesperogaulus Korth, 1999a; Pterogaulus n. gen.; Umbogauius n. gen.; and Ai-
phagaulus n. gen.

Fig. 4. — Skull of Mesogaulus paniensis, FAM 65511. A. Dorsal view. B. Ventral view. C. Left lateral
view (zygoma removed). Abbreviations for cranial foramina; asc, alisphenoid canal; bu, buccinator;
dpi, dorsal palatine; euc, eustachian canal; fo, foramen ovale; hy, hyoid; ifo, infraorbital; in, incisive;
ipm, interpremaxillary; ju, jugular; ms, mastoid; msc, masticatory; op, optic; pom, posterior maxillary;
ppl, posterior palatine; spl, sphenopalatine; spn, sphenoidal fissure; sty, stylomastoid; t, temporalis.
Diagonal lines indicate broken areas. Bar scale = 1 cm.

238

Annals of Carnegie Museum

VOL. 69

Table 3. — Humeral index (distal width/length) of species of mylagaulids.

Galhreathia novellas

0.38

Mesogaulus paniensis

0.43

Alphagaulus tedfordi

0.43

Umbogaulus galushai

0.43

Ceratogaulus rhinocerus

0.42-0.45

Ceratogaulus minor

0.45

Ceratogaulus hatcheri

0.44

Pterogaulus laevis

0.43-0.44

Pterogaulus sp.

0.45

Alphagaulus, new genus

Type Species. — Mylagaulus vetus, Matthew, 1924.

Referred Species. — Alphagaulus pristinus (Douglass, 1903), A. douglassi (McKenna, 1955), and A.
tedfordi n. sp.

Range. — Late Hemingfordian of Wyoming and Nebraska; early Barstovian of
Nebraska and Montana.

Diagnosis. — Range in size from the smallest to intermediate- sized mylagau-
lines; skull not as wide posteriorly as other mylagaulines (W/L = 0.77-0.83); low
bosses of rugose bone at anterior end of nasals; postorbital process intermediate
in size (POI = 0.25-0.35); premolars simpler than other mylagaulines — fewer fos-
settes (minimum of five fossettids on P4); P^ parafossette remains forked until
very late wear stage.

Discussion. — In morphology of the skull and dentition, as well as temporal
occurrence (late Hemingfordian to early Barstovian), Alphagaulus is transitional
between Mesogaulus and more advanced mylagaulines. In overall size and the
morphology of the parafossette on P^, the species of Alphagaulus appear to be
divisible into two groups. Both A. vetus (Matthew, 1924) and A. pristinus (Doug-
lass, 1903) are more gracile, and the buccal branch of the parafossette on P'^
separates from the rest of the fossette in very late stages of wear. Alphagaulus
douglassi and the new species described below have much more robust skulls
than the former species and the lingual branch of the parafossette on P^ separates
first at an earlier stage of wear. These differences suggest possible ancestral re-
lationships to specific later, more derived mylagaulines.

In addition to the two species of this genus recognized from Nebraska, two
previously described species are referred here to Alphagaulus as well. Douglass
(1903) based the species Mylagaulus? pristinus on a single mandible with a newly
erupted P4 (CM 742). However, he sectioned the premolar at the level of its
maximum dimensions, providing a view of the premolar at an advanced stage of
wear for comparison with other species (Fig. 6E). Alphagaulus pristinus differs
from all other species of the genus in being smaller in size. In the same article,
Douglass (1903) described an additional species, Mylagaulus proximus, from the
same horizon as A. pristinus. It was also based on a single specimen, CM 843, a
mandible with dP4-M3 with an erupting P4. This specimen is nearly identical to
the holotype of A. pristinus in size and morphology (Douglass, 1903:189-190,
fig. 27). Therefore, M. proximus is believed to be a junior synonym of A. pristinus.

Alphagaulus pristinus most closely approaches A. vetus in size and morphology.
Now that a sufficient number of specimens of both of these species is known

2000

Korth — Miocene mylagaulid rodents from Nebraska

239

Table 4. — Measurements of premolars c/ Alphagaulus vetus/ram different areas. Topotypic material
is from Sheep Creek Formation of Nebraska. Split Rock data from Munthe (1988: table 2), no ranges
were given. Statistical abbreviations: n, number of specimens; x, mean; OR, size range; s, standard

deviation; cv, coefficient of variation.

n

X

OR

S

CV

Late Hemingfordian

(Split Rock, Wyoming)*

P** a-p

3

7.4

—

0.4

4.8

tr

3

5.7

—

0.3

5.1

P4 a-p

4

8.2

—

1.1

13.4

tr

4

4.3

—

0.3

6.7

Late Hemingfordian (Sheep Creek Formation, Nebraska)

P"^ a-p

4

7.8

7.5-8.0

0.2

2.6

tr

5

5.2

5.0-5.6

0.2

4.3

P4 a-p

12

8.1

7.3-9.3

0.7

8.2

tr

12

4.6

4.1-5.2

0.3

7.3

Early Barstovian (Observation Quarry, Nebraska)

P*^ a-p

15

8.0

7.3-9.0

0.5

6.6

tr

15

4.7

4.7-6. 1

0.4

7.8

P4 a-p

31

8.4

6.9-9.3

0.6

7.6

tr

31

4.4

3. 3-5.6

0.5

12.0

(Table 4; Sutton and Korth, 1995:table 3); it is evident that A. pristinus is indeed
smaller than representative samples of A. vetus.

A partial skull from the Barstovian of Montana, AMNH 21307, is assigned to
A. pristinus based on size and occurrence. It is smaller than skulls of A. vetus
(width of occipital slightly more than half that of skulls of A. vetus), but the angle
of the occipital (OA = 79°) and relative size of the postorbital processes
(POI = 0.31) are very similar to these measurements in skulls of the type species.
The sagittal crest is also doubled as in skulls of A. vetus.

Alphagaulus pristinus is restricted to the early Barstovian of Montana. Al-
though Black (1961) and Sutton and Korth (1995) referred specimens from this
area to other species {Mesogaulus paniensis and Mylagaulus vetus, respectively)
they are all clearly referable to A. pristinus.

Alphgaulus douglassi is represented by a single specimen, UCMP 44694, a
partial skull with cheek teeth (McKenna, 1955: fig. 1). This species is also refer-
able to Alphagaulus. The major difference between A. douglassi and the other
species is in the separation of the lingual fork of the parafossette on P"^ (buccal
fork separates in A. vetus and A. pristinus). This separation apparently occurs at
an earlier stage of wear than in A. vetus. McKenna (1955) noted that the skull of
A. douglassi had longer parietals than other known species of mylagaulids. This
clearly indicates that the angle of the occipital is nearly vertical, as in A. vetus,
although the posterior end of the skull is missing on the only known specimen
of A. douglassi. In more derived mylagaulines, the parietals are shortened by the
anterior tilting of the occiptal bone. The skull of A. douglassi is very robust and
has a single, broad sagittal crest, unlike that of A. vetus and all known mylagau-
lines that have a doubled sagittal crest.

Alphagaulus vetus (Matthew, 1924)

(Fig. 5, 6A-D; Table 4)

Mylagaulus vetus Matthew, 1924.

Mesogaulus vetus (Matthew) Cook and Gregory, 1941.

240

Annals of Carnegie Museum

VOL. 69

2000

Korth — Miocene mylagaulid rodents from Nebraska

241

Fig. 6. — Dentition of Alphagaulus vetus and A. pristinus. A-D. Alphagaulus vetus from early Barsto-
vian Observation Quarry, Nebraska. A. FAM 65547, left P^. B, C. FAM 65536. B. left M^. C. left
P4, M2-M3, D. FAM 65542, left P4. E. Alphagaulus pristinus, holotype, CM 742, little-worn occlusal
surface (above) and surface of cross section at midheight of tooth (below). Bar scale = 5 mm.

Fig. 5. — Skull of Alphagaulus vetus, FAM 65514. A. Dorsal view. B. Lateral view. C. Ventral view.
Bar scale = 1 cm.

242

Annals of Carnegie Museum

VOL. 69

Type Specimen. — AMNH 18905, right mandible with incisor and P4, M2 (Sutton
and Korth, 1995:fig. 3H).

Referred Specimens. — Topotypic specimens: AMNH 20504, 20507, 90734; FAM 65515, 65517-
65520, 65523, 65526, 65527. From Observation Quarry: FAM 65532, 65534-65536, 65538-65551,
65556, 65558, 65559, 65561. Also see Munthe (1988:24) for list of referred specimens from Split
Rock fauna.

Horizon and Locality. — Holotype and some referred specimens from the late
Hemingfordian Thompson Quarry, Sheep Creek Formation of Sioux County, Ne-
braska; other referred specimens from the late Hemingfordian Split Rock local
fauna of Wyoming and the earliest Barstovian Observation Quarry, Dawes Coun-
ty, Nebraska.

Age. — Late Hemingfordian to early Barstovian (late-early to early-middle Mio-
cene).

Emended Diagnosis. — Larger than A. pristinus; skull not as deep (dorsoven-
trally) or robust as in A. douglassi and A. tedfordi; in late stages of wear buccal
branch of parafossette of separates from the remainder of the fossette (lingual
branch separates in A. douglassi).

Description. — The proportions of the skull of A. vetus are intermediate between that of Mesogaulus
and later mylagaulines. In general, the skull is low and broad as in all mylagaulines (W/L = 0.77-
0.83; similar to that of Mesogaulus) but is not as posteriorly expanded as other mylagaulines (W/
L > 0.88). Similarly, the angle of the occipital (OA) ranges from 60° to 75°, much less than in
Mesogaulus (OA = 90°) but not as low as in advanced mylagaulines where the angle is as low as 50°
(Table 1). The postorbital process is intermediate in size (POI = 0.26-0.36), similar to that of Meso-
gaulus.

A small bump is present on the nasal bones at their anterior end. It is variable in size but is present
on specimens that preserve the nasals from the late Hemingfordian and early Barstovian as well. All
other features of the skull and mandible are similar to those described by Munthe (1988) for the
species, and by Wahlert (1974) for the family.

The number of fossettes on the upper premolar varies from five to nine. The majority of specimens
(80%) have six or seven (x = 6.7, mode = 7). The largest fossette is the parafossette, which remains
forked until very late stages of wear. Of the over 20 specimens examined, only two showed the
separation of the buccal branch of the parafossette. Both of these specimens were in very late stages
of wear. The remainder of the fossettes on follow a basic pattern: two on the anterior half of the
tooth (one buccal and one lingual to the parafossette); and three on the posterior half of the tooth,
roughly aligned with the three on the anterior half of the tooth. The seventh fossette is usually a small,
circular one along the buccal edge of the tooth near its center. Heavily worn specimens often eliminate
one or more of these fossettes, reducing their number. Specimens with more than seven fossettes
appear to be the result of one of the larger fossettes dividing into two or more smaller fossettes.

The lower premolar also consists of anteroposteriorly elongated fossettids, typical of mylagaulines.
The number of fossettids ranges from nine to five (x = 5.9, mode — 5). As with the upper premolars,
the number of fossettids generally is reduced with age. Those specimens with more than five or six
fossettids are usually younger individuals. The specimens that are at a moderate stage of wear usually
have only five or six. Again, as with the upper premolars, additional fossettids on P4 are produced by
the splitting of one or more of the larger fossettids into two. There is a general pattern in the location
of the five main fossettids: three along the lingual edge of the tooth, and two along the buccal side
of the tooth.

Discussion. — Alphagaulus vetus is the best represented species of the genus. It
is also one of the best represented species of mylagaulines. There is virtually no
difference in the size ranges and morphology of the samples of A. vetus from the
late Hemingfordian of Wyoming and Nebraska and the early Barstovian of Ne-
braska (Table 4). The size of the boss at the anterior end of the nasals varies
somewhat, but is always present on specimens that preserve the nasals. Alpha-
gaulus vetus is slightly larger than A. pristinus. In a very late stage of wear of
P^, the buccal branch of the parafossette separates from the rest of the fossette.

2000

Korth — Miocene mylagaulid rodents from Nebraska

243

unlike A. douglassi. This only occurs in very few specimens, all of which are of
senile individuals.

Alphagaulus vetus is morphologically intermediate between Mesogaulus and
later, more advanced mylagaulines in the construction of the skull. The skull of
A. vetus is broader and has a lower angle of the occipital than Mesogaulus, but
is not as broad as in other mylagaulines and has a steeper occipital angle. The
dentition is also intermediate, having a greater number of fossettes (-ids) on the
premolars than Mesogaulus but fewer than other mylagaulines.

Alphagaulus tedfordi, new species
(Fig. 7, 8, 9B, C)

Type and Only Specimen. — FAM 65711, nearly complete skeleton.

Dental Measurements of Holotype. — P^, a-p 7.50 mm, tr 5.70 mm; P4, a-p 6.74 mm, tr 5.40 mm.

Horizon and Locality.— Wot Springs drainage system (below level of Obser-
vation Quarry), NE %, SE %, section 4, T31N, R47W, Dawes County, Nebraska.

Age. — Latest Hemingfordian or earliest Barstovian.

Diagnosis. — Slightly larger than A. douglassi; skull robust with vertically ori-
ented occipital (OA = 90°); lingual fork of parafossette on separates first (as
in A. douglassi); retains outline of mesostyle, not becoming completely oval
as in other species.

Etymology. — Patronym for R. H. Tedford.

Description. — The postcranial skeleton is similar to that of other mylagaulines with robust limbs,
especially the forearms and manus. The ratio of distal width to length of the humerus is within the
range that of other mylagaulids (Table 3). The skull is also robust and retains a number of features
of Mesogaulus. The angle of the occipital is 90°, as in Mesogaulus and the promylagaulines (Table
1) and the postorbital process is primitively intermediate in size (POI = 0.25). However, there are a
number of features of the skull that make this specimen clearly a mylagauline. The skull is wider
posteriorly than in Mesogaulus and most other mylagaulines (W/L = 1.22) and there are bosses of
rugose bone on the anterior ends of the nasals. These bosses are larger than those of A. vetus, but
considerably smaller than those of Umbogaulus (described below). There appears to be no difference
between the arrangement of sutures or foramina on the skull of A. tedfordi and of other mylagaulines.
The auditory bullae have the characteristic elongated external meatus of all mylagaulids.

The upper premolar of the type specimen is in a moderate stage of wear and has six fossettes; four
elongated fossettes (two on the anterior half, two on the posterior half), and a minute buccocentral
fossette. The lingual branch of the parafossette has separated to form the minute seventh fossette. This
arrangement of the major fossettes is very similar to that of Mesogaulus and advanced promylagau-
lines. The separation of the lingual branch of the parafossette is characteristic of A. douglassi (Fig.
9A) and all species of Ceratogaulus and Mylagaulus. The occlusal outline of P^ retains the outline of
the mesostyle, a character of Mesogaulus and promylagaulines.

The lower premolar has six fossettids, reminiscent of the premolars of A. vetus, but is much more
robust than the latter. Both the upper and lower premolars are wider relative to length than in A. vetus.

Discussion. — Alphagaulus tedfordi retains a number of primitive Mesogaulus-
like characters of the skull and dentition. The angle of the occiptial is vertical
and the postorbital processes are intermediate in size, as in Mesogaulus. Similarly,
the outline of the mesostyle is preserved on another primitive character for
mylagaulines. However, the posterior widening of the skull of A. tedfordi is much
greater than that of Mesogaulus and promylagaulines, and the presence of bumps
on the nasal bones is also a mylagauline trait.

Alphagaulus tedfordi more closely resembles A. douglassi than the other spe-
cies of the genus. Both A. tedfordi and A. douglassi are more robust than A. vetus
and A. pristinus, and the lingual branch of the parafossette on separates at an

244

Annals of Carnegie Museum

VOL. 69

Fig. 7. — Photograph of skeleton of A. tedfordi, holotype, FAM 65711. Dorsal view above, ventral
view below. Scale is in cm.

2000 Korth — Miocene mylagaulid rodents from Nebraska

245

Fig. 8. — Stereo photograph (dorsal view) of skull of A. tedfordi, holotype, FAM 65711. Scale

246

Annals of Carnegie Museum

VOL. 69

Fig. 9. — Dentitions of the holotypes of Alphagaulus douglassi and A. tedfordi. A. Alphagaulus doug-
iassi, UCMP 44694, P^, B, C. Alphagaulus tedfordi, FAM 65711. B. Right F. C Right P4.

Bar scale = 5 mm.

earlier stage of wear. A. tedfordi can be distinguished from A. douglassi by its
slightly larger size and the retention of the outline of a mesostyle on P^.

The more robust skull and separation of the lingual rather than the buccal fork
of the parafossette on in A. tedfordi and A. douglassi suggest primitive positions
of these species with respect to later species of Ceratogaulus that share these
same characters. The small bosses on the nasals can also be considered primitive
to the nasal horns of Ceratogaulus. The nasal bosses on a new genus, Umbogaulus
(described below), are much larger than those of A. tedfordi and the premolars
have many more fossettes (-ids) than the latter, distinguishing it from Umbogau-
lus.

Umbogaulus, new genus
Type Species. — Umbogaulus galushai n. sp.

Referred Species. — Umbogaulus monodon (Cope, 188 to).

Range. — -Late (and possibly early) Barstovian of Nebraska.

Diagnosis. — Larger than species of Ceratogaulus; large bosses on anterior end

2000

Korth — Miocene mylagaulid rodents from Nebraska

247

of nasals; premolars circular to oval in occlusal outline with multiple minute
fossettes (-ids), more complex than Ceratogaulus; branches of parafossette of
remain attached until very late stages of wear; anterocentral fossettid of P4 usually
anteriorly forked or widened.

Etymology. — Greek, umbos, knob or boss; and gaulos, bucket.

Discussion. — Umbogaulus differs from Ceratogaulus in the presence of large
bosses at the anterior end of the nasals rather than paired conical “horns.” Den-
tally, Umbogaulus has a greater number of fossettes complexity on the premolars
than even the most advanced Hemphillian species of mylagaulines. The greatest
difference in the premolars of advanced species of Ceratogaulus and Umbogaulus
is that the premolars of Ceratogaulus are relatively longer compared to width
than are those of Umbogaulus, and all the fossettes (-ids) are narrow, nearly
straight, and anteroposteriorly oriented. In Umbogaulus the fossettes are often
rounder and the teeth are relatively much wider (P"^ nearly round in occlusal
outline).

Umbogaulus galushai, new species
(Fig. 10, 11 A; Table 5)

Type Specimen. — FAM 65576, left P4.

Referred Specimens.— ¥ KM 65571-65575, 65590-65592, isolated premolars; FAM 65577-65582,
edentulous mandibles; FAM 65566, 65567, partial skulls; FAM 65586, two humeri.

Horizon and Locality. — Observation Quarry, “Sand Canyon beds,” Dawes
County, Nebraska.

Age. — Earliest Barstovian (middle Miocene).

Diagnosis. — Largest species of the genus; greater number of fossettes (-ids) on
the premolars (P"^: x = 8.3, mode 9; P4: x = 8.4, mode 8).

Etymology. — Patronym for T. Galusha.

Description. — The two skull fragments of this species preserve only the anterior half of the skull.
Both show enlarged bony bosses at the end of the nasals, and are generally heavily built. The pre-
maxillaries are laterally splayed at their dorsal extent to accommodate the enlargement of the anterior
end of the nasals and development of the nasal bosses. The centers of these bosses are more anterior
than the centers of the horn cores in even the most primitive species of Ceratogaulus. Due to breakage,
little else can be determined about the skulls.

The premolars are generally larger than in other mylagaulids and are proportionally wider than in
species of Ceratogaulus (ratio of width to length of approximately equals 0.75). The upper pre-
molars (represented by only four specimens) are nearly circular in occlusal outline. The number of
fossettes ranges from 7 to 9 with 9 as the modal number. The anterior branches of the parafossette
do not separate on any of the available specimens, suggesting that any separation will occur in only
very late stages of wear. There are several small, circular to oval fossettes along the buccal margin
of the tooth posterior to the parafossette. The lingual fossettes are elongated.

The lower premolars are less circular in outline, being distinctly longer than wide but still relatively
wider than in other mylagaulines. The number of fossettes ranges from 7 to 1 1 (x = 8.4, mode = 8),
more than in most other mylagaulines except the most advanced Hemphillian species of Hesperogaulus
(Shotwell, 1958; Korth, 1999«). The fossettids are arranged anteroposteriorly in three rows — lingual,
central, and buccal. The buccal row usually consists of only two elongated fossettids and the central
row of three. The greatest variability occurs in the lingual row of fossettids. The number of fossettids
in this row can range from two to as many as five. The size and shape of the fossettids in the lingual
row are clearly dependent on the number. If the fossettids are small and circular, there can be many
of them, if elongated, there are fewer fossettids. On approximately 50% of the specimens of P4, the
anterocentral fossettid is anteriorly forked or widened. In some cases a minute fossettid is in the place
of one of the branches of the fork.

248

Annals of Carnegie Museum

VOL. 69

Fig. 10. — Premolars of Umbogaulus galushai. A. FAM 65568, right B. FAM 65574, left P4. C.
FAM 65575, left P4. D. FAM 65576, holotype, left P4. Bar scale = 5 mm.

Discussion. — Although the morphology of the nasal bosses and their position
on the skull of Umbogaulus galushai appear to be primitive relative to species of
Ceratogaulus, the size and complexity of the cheek teeth exceed that of any
species of the latter, so it is unlikely that U. galushai is the ancestor to any species

2000

Korth — Miocene mylagaulid rodents from Nebraska

249

of Ceratogaulus even though the temporal occurrence (earliest Barstovian) is ap-
propriate for an ancestor of Ceratogaulus.

Among all species of mylagaulids, Umbogaulus galushai most closely resem-
bles the holotype of ''Mylagaulus'' mondon (AMNH 8327). The latter is slightly
smaller than U. galushai with fewer fossettids on P4. “M.” monodon cannot be
referred to the same species as U. galushai, but is referable to the same genus.

Umbogaulus monodon (Cope, 1881a)

(Fig. IIB, 12; Table 5)

Mylagaulus monodon Cope, 1881«.

Type Specimen. — AMNH 8327, partial mandible with incisor and P4.

Referred Specimens.— AMNH 13866, 13868, 13869,17215, 18899, 21466, 22044, 22045, 81071,
81074, 81077, 81080, 81081, 81084, 81085, 81592, 95047, 95054, 95056, 95057; FAM 65713-65716,
65785, 65944, isolated premolars; FAM 65695, mandible with cheek teeth; FAM 65016, 65712, 65784,
partial skulls.

Horizon and Locality. — Holotype from Ogallala Group, Driftwood Creek (Haz-
ard Homestead quarry), Hitchcock County, Nebraska; all referred specimens from
various localities in the Olcott Formation, Sioux County, Nebraska.

Age. — Early and late Barstovian.

Diagnosis. — Slightly smaller than U. galushai; fewer fossettes on the premolars
(P^: X = 7.9, mode 8; P4: x = 7.3, mode 7).

Description. — The nearly complete skull of this species lacks only the nasal bones. It is evident
that it is a robust skull, similar to that of U. galushai and is posteriorly widened (W/L = 0.87). The
only evidence for the formation of large bosses on the anterior end of the nasals is the splaying of
the premaxillaries along the naso-premaxillary suture. This morphology is present in the skulls of U.
galushai that preserve the nasals. The postorbital processes are intermediate in size, probably reflecting
the primitive conditon in mylagaulids (POI = 0.35). The occipital is slightly tilted anteriorly (OA
= 75°) but not as much as in most other advanced mylagaulines. In all other features, the skull of U.
monodon is like the cranial material of all other mylagaulines.

The premolars, as in the type species of the genus, are larger and relatively wider than in other
mylagaulines. The parafossette of remains forked until late stages of wear, when either of the
branches may be separate. The number of fossettes varies from six in a very young individual to as
many as ten. Both the mean and modal number of fossettes are fewer than in U. galushai (x = 7.9,
mode 8).

The lower premolars are very similar to those of the type species, differing only in being slightly
smaller (Table 5) and having slightly fewer fossettids (x = 7.3, mode = 7). The anterocentral fossettid
of P4 is also forked or widened as in U. galushai. This occurs in approximately 65% of the specimens,
a higher percent than in U. galushai.

Discussion. — Mylagaulus monodon was the second species of the genus de-
scribed by Cope (1881a). It was from the Republican River beds, Hitchcock
County, Nebraska. This species was based on a single mandible with an incisor
and P4 (AMNH 8327). The locality from which this specimen was collected has
been determined to be late Barstovian in age (Fiorillo, 1988; Voorhies, 1990a).

Nearly all specimens of large mylagaulids from the Barstovian to the Hem-
phillian that have not been associated with cranial material, as well as some
hornless skulls, have been referred to Mylagaulus monodon (Wilson, 1937a,
1937^; McGrew, 1941; Gregory, 1942; Webb, 1969; Baskin, 1979; Korth, 1998).
In spite of this universal use of the species name, the holotype of M. monodon
does not appear to match any of this referred material. Even though more collec-
tions have been made at the type locality of this species (Fiorillo, 1988), no
topotypic material has been found to define better the species itself. The holotype

250

Annals of Carnegie Museum

VOL. 69

2000

Korth — Miocene mylagaulid rodents from Nebraska

251

Table 5. — Measurements of premolars o/ Umbogaulus. Statistical abbreviations as in Table 4.

n

X

OR

5

cv

Umbogaulus galushai

a-p

4

10.88

10.20-12.00

0.68

6.28

tr

4

7.21

6.40-7.65

0.48

6.73

P4 a-p

21

12.02

10.30-13.25

0.79

6.56

tr

22

6.53

5.50-7.55

0.58

8.81

Umbogaulus sp., cf. U. monodon

p4

a-p

28

10.24

8.50-12.30

0.97

9.40

tr

27

6.26

5.50-7.60

0.46

7.28

P4

a-p

32

11.05

4.50-13.15

1.00

9.06

tr

32

5.54

4.75-7.10

0.55

9.84

of M. monodon differs from the species of Mylagaulus, which are characterized
by having smaller, more simplified lower premolars (fewer fossettids). The type
premolar of M. monodon is much larger with more fossettids. It is clear that “M.”
monodon cannot be referred to Mylagaulus.

All of the referred specimens of Umbogaulus monodon listed above are from
the early Barstovian Olcott Formation. There is no morphological difference be-
tween this sample and the holotype of the species, but there is a distinct age
difference between the holotype (late Barstovian) and the referred material (early
Barstovian).

Among known species of mylagaulids, the lower premolars of Umbogaulus
galushai most closely resemble the type specimen of M. monodon, therefore the
latter is referred to Umbogaulus. The lower premolars of U. galushai are distinct
from those of U. monodon in being slightly larger and having more fossettids.

Mylagaulus Cope, 1878

Type Species. — Mylagaulus sesquipedalis Cope, 1878.

Included Species. — Mylagaulus kinseyi Webb, 1966, and M. elassos Baskin,
1980.

Range. — Possibly late Barstovian of Kansas and Nebraska; Clarendonian and

Hemphillian of Florida.

Emended Diagnosis. — with posterobuccal fossette C-shaped (open poster-
obuccally) and parafossette with lingual branch separated first in early wear (as
in Ceratogaulus); P^ oval in occlusal outline (no indication of stylar cusps); P4
simpler, with fewer fossettids (six) than other advanced mylagaulines.

Discussion.— Baskin (1980) demonstrated that the two species of Mylagaulus
from the Clarendonian and Hemphillian of Florida, M. kinseyi and M. elassos,
belonged to a lineage that reduced size through time. The only species of Myla-
gualus from the Great Plains with smaller size was the type species, M. sesqui-
pedalis. The separation of the lingual branch of the parafossette of P"^, generally
smaller size, reduced complexity (fewer fossettes [-ids] on the premolars), and

Fig. 11. — Skulls of Umbogaulus. In both A and B top is left lateral view, bottom is dorsal view. A.
Umbogaulus galushai, FAM 65566. B. Umbogaulus monodon, FAM 65016. Nasals missing on B. Bar
scale = 1 cm.

252

Annals of Carnegie Museum

VOL. 69

Fig. 12. — Premolars of Umbogaulus monodon. A. AMNH 8327, holotype, left P4. B. FAM 65695,

right P4, M2-M3. C. FAM 95057, right P^. D. AMNH 22044F, right P^ E. FAM 95052, left P^ Bar
scale = 5 mm.

2000

Korth — Miocene mylagaulid rodents from Nebraska

253

the C"Shaped posterobuccal fossette on are features that are shared by M.
sesquipedalis and the Florida species, making these three species a distinctive
genus.

The only significant difference between M. sesquipedalis and the Florida spe-
cies is its larger size. The temporal occurrence of M. sesquipedalis^ possibly latest
Barstovian (see discussion below), fits well into the sequence of species of My-
lagaulus from larger to smaller through time. Being the earliest species, M. ses-
quipedalis is the largest species, Mylagaulus elassos from the Clarendonian is
slightly smaller, and M. kinseyi from the late Hemphillian is the smallest species
(see Baskin, 1980:fig. 1).

Mylagaulus sesquipedalis Cope, 1878

(Fig. 13)

Type Specimen. — AMNH 8329, left P'^.

Possible Referred Specimen. — AMNH 8330, left P4.

Horizon and Locality. — Holotype from Ogallala Group, Sappa Creek, Rawlins
County, Kansas. Referred specimen from the “Republican River Beds,” Drift-
wood Creek, Hitchcock County, Nebraska.

Age. — ?Late Barstovian.

Emended Diagnosis. — Largest species of the genus.

Discussion. — Cope (1878) erected Mylagaulus sesquipedalis based on an iso-
lated upper premolar. The age of the fauna from Sappa Creek, from which the
holotype of M. sesquipedalis came, is uncertain but has been referred to by pre-
vious authors as Clarendonian (Hibbard and Phillis, 1945:552; Baskin, 1980).
However, it appears that the Sappa Creek fauna may be older than previously
thought. The AMNH specimens from the Sappa Creek fauna include those col-
lected by Cope’s field parties in the late 1870s, and a few specimens collected by
George Sternberg for the Frick Collections in the early 1930s. Besides the holo-
type of Mylagaulus sesquipedalis, the collected fauna consists of five mammalian
taxa: a horse, Protohippus medius; a rhino, Teleoceras fossiger; a peccary, Pros-
thennops servus; a mastodont, Tetralophodon campester; and a merycodont artio-
dactyl, Ramoceros kansanus. Protohippus medius, Prosthennops servus, and Te-
tralophodon compester are represented by the holotypes of each of these species.
The holotype of P. medius (AMNH 8360) is the skull of a senile individual where
the teeth are completely worn, eliminating nearly all of the occlusal pattern. This
species of Protohippus has not been recognized by any later workers (notably
Osborn, 1918; and Stirton, 1940). The specimen has facial fossae consistent with
those of Protohippus, and a size equivalent to that of P. perditus (R. L. Evander,
personal communication) which is limited to the late Barstovian of Nebraska
(Voorhies, 1990^). The merycodont similarly limits the age of the fauna because
Ramoceras does not extend any later than the late Barstovian in the Nebraska
section (Voorhies, 1990^). However, the peccary is very large and is similar to
Hemphillian species. There is clearly a mixing of faunas in the Sappa Creek fauna.
The mastodont is an advanced species which suggests a younger, possibly Hem-
phillian age. The rhino is known elsewhere to be late Clarendonian or early Hem-
phillian in age. Clearly, further collecting in the type area of Sappa Creek is
necessary to verify the age of the fauna.

Matthew (1902) reported on a second specimen of M. sesquipedalis, AMNH
8330, that included a lower premolar, an incisor, and a metacarpal. He stated that

254

Annals of Carnegie Museum

VOL. 69

Fig. 13. — Dentition of Mylagaulus sesquipedalis. A. holotype, AMNH 8329, right B. paratype,
AMNH 8330, left P4. Bar scale = 5 mm.

the referred specimen was from the same locality as the holotype of M. monodon.
Driftwood Creek, which is late Barstovian in age (Fiorillo, 1988; Voorhies,
1990a). At present, only the holotype and referred premolar are in the AMNH
collections. The isolated incisor and metacarpal of the referred specimen cannot
be located. While the reference of the referred P4 to M. sesquipedalis may be
questionable, it is morphologically consistent with the remainder of the species
referred to the genus, M. kinseyi and M. elassos. This specimen was previously
figured by Cope and Matthew (1915:pl. CXIXc, fig. 10), but was mistakenly
labelled as the holotype. The occurrence of the referred specimen from a late
Barstovian indicates that the species occurred at this time.

Both Hibbard and Phillis (1945) and Korth (1998) discussed the possibility of

2000

Korth — Miocene mylagaulid rodents from Nebraska

255

the synonymy of the Clarendonian species Epigaulus minor and M. sesquipedalis,
noting that no conclusion could be drawn simply because of the lack of compar-
ative material. The upper premolars of E. minor are slightly larger than the ho-
lotype of M. sesquipedaiis and they are also wider relative to length. The mor-
phology of the parafossette on is the same on both species, but the morphology
of the posterobuccal fossette on specimens of E. minor is not the rounded C-
shape that is diagnostic of M. sesquipedaiis. If, however, these species do prove
to be synonymous in the future, Myiagaulus would have priority, making the
presence of horns on the nasals a generic character for Myiagaulus.

Ceratogauius Matthew, 1902

Type Species.- — ^Ceratogauius rhinocerus Matthew, 1902,

Diagnosis.— EmgQ mylagaulines; horns present on nasal bones; postorbital pro-
cesses progressively reduced (POI ^ 0.20); slope of occipital in Hemphillian spe-
cies not as low as in other genera (OA = 60^-63°); occipital crest forms straight,
transverse line (not deflected anteriorly as in other genera); upper premolars rel-
atively wide (width to length ratio ^ 0.68); lingual branch of parafossette sepa-
rates first on P^; anterolingual fossette attaches to lingual branch of parafossette
on P^ with wear; anterocentral fossettid of P4 V-shaped in early wear.

Included Species.— Ceratogauius hatcheri (Gidley, 1907), C. minor (Hibbard
and Phillis, 1945), and C. anecdotus, n. sp.

Early Barstovian to Hemphillian (middle to late Miocene) of the
northern Great Plains.

Discussion.— Ceratogauius differs from all other mylagaulids in the possession
of horns (or hom cores) on the nasal bones. The progressive changes in the genus
through time are the more posterior placement of the nasal horns and reduction
of the postorbital processes (Fig, 14). The nasal horns not only are placed pro-
gressively more posteriorly on the skull, but the horns also have narrower bases
and are generally taller. In the late Barstovian C. rhinocerus, the horns are pointed
but have a very wide base, giving them a pyramid-like profile. In the Clarendonian
specimens of the former, the nasals are slightly more posterior (NHI = 0.16-0.22).
In the Hemphillian C. hatcheri, the horns have narrow bases, are circular in cross
section, and are nearly dorsal to the orbits (NHI > 0.28). Similary, the postorbital
processes are reduced through time: in C rhinocerus they are moderately large
(POI = 0.18-“0.26), probably near the primitive condition; in Ceratogauius sp.,
cf. Ceratogauius rhinocerus in the early Clarendonian, the processes are smaller
(POI = 0.12-0.15); and, ultimately, in Hemphillian species the postorbital pro-
cesses are nearly absent (POI = 0.11).

Gidley (1907) named Epigaulus for a species, E. hatcheri, from the Hemphil-
lian of Kansas, noting that it differed from the type of Ceratogauius rhinocerus
(only known species of the genus at that time) from the Barstovian in the more
posterior position of the nasal horns, more enlarged premolars, and presence of
cement around the prem.olars. All of these characters are now known to represent
gradational changes in the known species of Ceratogauius from throughout the
entire time of occurrence of the genus (Table 1). There is no evidence indicating
that Epigaulus exists as a genus separate from Ceratogauius, but more likely
represents the most derived species.

Ceratogauius is geographically restricted to the northern Great Plains from
Kansas to Saskatchewan. Species of Ceratogauius are frequently found in faunas

256

Annals of Carnegie Museum

VOL. 69

Fig. 14. — Outlines of skulls of Ceratogaulus showing progressive changes through time (position of
nasal horns and reduction of postorbital process). Dorsal view on left, left lateral view on right. A.
Barstovian C. rhinocerus, AMNH 9456 (holotype). B. Early Clarendonian C. sp., cf. C rhinocerus,
FAM 65385. C. Hemphillian C. hatcheri (holotype) USNM 5485. Bar scale = 1 cm.

2000

Korth — Miocene mylagaulid rodents from Nebraska

257

Fig. 15. — Premolars of Ceratogaulus rhinocerus. A, B. AMNH 9456 (holotype), Barstovian. A. Left
P^, M2--M7 B. Left P4. C, D. Ceratogaulus sp., cf. C. rhinocerus, early Clarendoeian (Burge). C. FAM
65385, left r. D. FAM 65408, left P4. Bar scale = 5 mm.

that also contain species of a hornless genus of mylagauline described below (see
discussion of Pterogaulus).

Ceratogaulus rhinocerus Matthew, 1902
(Fig. 14C, 15A, B; Table 6)

Type Specimen.— AMNH. 9456, nearly complete skull with associated lower
jaw (Matthew, 1902:fig. 1).

258

Annals of Carnegie Museum

VOL. 69

Table 6. — Measurements of premolars of Ceratogaulus. Statistical abbreviations as in Table 4.

n

X

OR

5

cv

Ceratogaulus rhinocerus

Late Barstovian (Crookston Bridge Member)

P^ a-p

3

9.82

9.78-9.88

—

—

tr

3

6.91

6.57-7.10

—

. —

P4 a-p

4

11.04

10.25-12.35

—

—

tr

4

6.45

6.10-6.61

—

—

Late Barstovian (Devil’s Gulch Member)

P"^ a-p

3

9.50

8.30-11.60

—

—

tr

3

6.70

5.70-8.40

—

—

P4 a-p

1

9.20

—

—

—

tr

2

5.75

5.70-5.80

—

—

Ceratogaulus sp., cf. C. rhinocerus

Early Clarendonian (Burge

Member)

P^ a-p

15

9.16

8.35-9.80

0.53

5.79

tr

16

6.21

5.40-7.00

0.56

8.99

P4 a-p

20

10.23

8.70-11.45

0.79

7.73

tr

21

5.43

4.90-6.15

0.38

7.00

Ceratogaulus anecdotus

Late Clarendonian (Merritt

Dam Member)

P^ a-p

3

10.16

9.10-12.20

—

—

tr

3

6.86

6.20-7.90

—

—

P4 a-p

3

12.66

10.00-14.70

—

—

tr

4

5.55

5.05-6.70

—

—

Referred Specimens.— AMNH 18899; UNSM 122005, 122007, 122010, 122009.

Horizon and Locality. — Holotype from Pawnee Creek beds, Logan County,

Colorado; AMNH 18899, from Olcott Formation, Sioux County, Nebraska;
UNSM specimens from Crooks ton Bridge Quarry, Crookston Bridge Member,
Valentine Formation, Cherry County, Nebraska.

Age. — Early to late Barstovian (middle Micoene).

Emended Diagnosis. — Homs on nasals positioned more anteriorly than in other
species (NHI = 0.19-0.21), and not as tall; postorbital processes small, triangular
flanges (POI = 0.19), larger than in other species.

Decription. — Matthew (1902) fully described and figured the skull of Ceratogaulus rhinocerus. The
upper premolar has from five to ten fossettes (x = 7.4; mode = 7). While the mean number of fossettes
on is slightly higher than other species of the genus, the modal number is identical for all species
of Ceratogaulus (Table 2). The parafossette is the longest of the fossettes and its lingual branch
separates from the rest of the fossette in early stages of wear. A single lingual fossette runs most of
the length of the tooth, occasionally uniting anteriorly with the lingual branch of the parafossette once
it has separated. The remainder of the fossettes on are smaller and directed anteroposteriorly.
Frequently, there is one minute, circular fossette present along the buccal edge of the tooth just
posterior to the center of the tooth. P'^ is not perfectly oval in occlusal outline. There is always a slight
concavity along the buccal edge of the tooth (created by a vestige of the mesostyle) with its deepest
curvature near the middle of the tooth.

The lower premolar has seven to ten fossettids (x = 7.7; mode = 7). There are three elongated
fossettids along the lingual side and three along the buccal side of the tooth with one or more small,
circular accessory fossettids randomly positioned. The elongated fossettids are all oriented obliquely
(posterolingual to anterobuccal). The most anterior fossettid is V-shaped in early stages of wear with
the apex of the V pointed posteriorly. After a moderate amount of wear, the lingual arm of the V
separates from the rest of the fossettid, and ultimately disappears with additional wear.

2000

Korth — Miocene mylagaulid rodents from Nebraska

259

Fig. 16. — Skull of Ceratogaulus sp., cf. C rhinocerus, FAM 65385, from Burge Quarry (early Clar-
endonian). A. Dorsal view. B. Left lateral view. C. Ventral view. D. Anterior view. Bar scale = 1 cm.

Discussion. — Specimens of C. rhinocerus show a consistent morphology of the
skull and dentition throughout the entire range of the species. The dentition is
especially conservative in terms of morphological change. There is little or no
change in the size or morphology of the premolars from the early Barstovian to
the early Clarendonian.

Ceratogaulus rhinocerus differs from the late Clarendonian and Hemphillian
species of Ceratogaulus mainly in the morphology of the premolars. The upper
premolars of Ceratogaulus are relatively wide compared to their length (length
to width ratio of ^0.71). of C rhinocerus is not entirely oval in outline,
maintaining a slight concavity on the buccal margin between the mesostyle and
parastyle. Later species of Ceratogaulus have much narrower pfemolars length
to width ratio ^0.65), cement appears around the premolars at an earlier stage
of wear, and the premolars are oval in outline, lacking the slight buccal concavity
present on P"^ of C. rhinocerus.

Ceratogaulus sp., cf. C. rhinocerus
(Fig. 14B, 15C, D, 16; Table 6)

Referred Specimens.— ¥ AM 65012, 65013, 65370, 65371, 65373, 65374, 65375, 65377, 65385,
65388, 65398, 65399, 65403, 65404, 65406, 65408, 65412, 65413, 65418, 65420, 65422, 65476,
65489, 65833; UNSM 122006.

Horizon and Locality.— All specimens from various quarries (Burge, Midway,
White Face, Lucht, June, Gordon Creek) in the Burge Member, Valentine For-

260

Annals of Carnegie Museum

VOL. 69

mation, northcentral Nebraska (see Skinner and Johnson, 1984, for location of
quarries).

Age. — Early Clarendonian.

Discussion. — Dentally, it is virtually impossible to distinguish the Burge spec-
imens of Ceratogaulus rhinocerus from those from lower in the Valentine For-
mation. However, the cranial material is somewhat different. The Burge speci-
mens, referred here to Ceratogaulus sp., cf. C. rhinocerus, have more posteriorly
placed horns and more reduced postorbital processes (Table 1). The skulls are
also slightly larger. The slightly larger skull size is not reflected in the dimensions
of the dentition (Table 6) or the complexity of the premolars (Table 2). The horns
on the specimens from the Burge have narrower bases and are more nearly circular
in cross section than those from the late Barstovian.

It would be impractical to name a new species for the Burge specimens at this
time because the majority of the specimens recovered are dentitions which cannot
be separated from the Barstovian specimens morphologically. The Burge skulls
are clearly more advanced in having more posteriorly placed horns and smaller
postorbital processes (Fig. 14, 16),

Ceratogaulus anecdotus, new species
(Fig. 17; Table 6)

Epigaulus minor Hibbard and Phillis; Korth, 1997 (in part).

Mylagaulus sp., cf. M. monodon Cope; Korth, 1997 (in part).

Type Specimen. — FAM 65800, fragmentary skull with both upper premolars
and some postcranial fragments.

Referred Specimens. — FAM 65466, partial skull; FAM 65456, 65464 isolated P'^s; FAM 65456,
associated upper and lower premolars; FAM 65430, 65468, 65481, 65497; UNSM 101813, mandible
with P4, M2-M3.

Horizon and Locality. — Holotype and some referred specimens from Pratt
Quarry (UNSM locality Bw-123), Merritt Dam Member, Ash Hollow Formation,
Brown County, Nebraska; other referred specimens from various localities in the
Merritt Dam Member; FAM 65466 from “Prospect 28-18,” Cap Rock Member,
Ash Hollow Formation, Brown County, Nebraska.

Age. — Middle to late Clarendonian (early late Miocene).

Diagnosis. — Smaller than C. hatcheri, larger than other species of the genus;
premolars more elongated than in Ceratogaulus rhinocerus, but less than in C.
hatcheri (width to length ratio of P^ = 0.64-0.66); fewer fossesttes (-ids) on pre-
molars than in C. hatcheri; P"^ oval in occlusal outline; cement surounds premolars
in an earlier stage of wear than in C. rhinocerus; postorbital process subequal to
that of C. hatcheri and smaller than in C. rhinocerus (POI 0.12); nasal horn
less posterior than C. hatcheri, more posterior than in C. rhinocerus (NHI = 0.30).

Etymology. — Greek, anekdotos, unpublished or secret.

Description. — The nearly complete skull of a juvenile individual of C. anecdotus, FAM 65466, is
badly broken but most of the morphology of the skull is recognizable. A nearly complete skull in the
UNSM collections (field number 2-26-5-15) is being studied elsewhere, but allows for a determination

Fig. 17. — Dentition of Ceratogaulus anecdotus. A, B. Holotype, FAM 65800. A. Right P"^. B. Left P^.
C. FAM 65484, right P4. D. UNSM 101813, right P4. E. FAM 65497, right P4. Bar scale = 5 mm.

262

Annals of Carnegie Museum

VOL. 69

of the position of the nasal horn (NHI = 0.30) and the angle of the occipital (OA = 60°). The holotype
also contains fragments of the skull including a complete nasal bone that preserves the horn. The horn
is roughly circular in cross section and is posteriorly positioned. Its position is slightly more anterior
than in C. hatched but more posterior than in any skull of Ceratogaulus rhinocerus (Table 1). The
postorbital processes are more reduced than in any other species of Ceratogaulus except C. hatched.
The premolars are larger than those of C. rhinocerus, but smaller than those of C. hatched. The
number of fossettes on ranges from 6 to 9, similar to that of C. rhinocerus (Table 2). is narrower
than in C. rhinocerus and C. minor (width to length ratio = 0.65). One specimen of P'^ (FAM 65465)
has little wear and appears a little wider, within the range of C. rhinocerus. The arrangement of the
fossettes of P"* is similar to those of C. minor and C. rhinocerus, differing only in having several
minute fossettes on the posterior half of the tooth. The two P^s of the holotype are only moderately
worn and already are completely surrounded by cement. This appears to occur much earlier in C.
anecdotus than in C. rhinocerus.

The lower premolars of C. anecdotus differ from those of C. rhinocerus in the same way as do the
upper premolars. The range of the number of fossettids is 6 to 10, again similar to the range in C.
rhinocerus. The only other differences between P4 of C anecdotus and those of C. rhinocerus are
larger size and earlier appearance of cement around the tooth.

Discussion. — Ceratogaulus anecdotus is more advanced than C rhinocerus
based on the morphology of the horn (more posterior, circular cross section),
proportions of the premolar, reduction of the postorbital process, larger size, and
earlier occurrence of cement around the cheek teeth. It differs from C. hatcheri
in having a more anteriorly placed horn, smaller size, and fewer fossettes (-ids)
on the premolars.

Korth (1997) originally referred the holotype of C. anecdotus to Epigaulus
minor. However, this specimen differs from C minor not only in the cranial
features cited above, but also in its larger size and narrower upper premolars.
Similarly, a specimen figured by Korth (1997: fig. IB) as Mylagaulus sp., cf. M.
monodon, UNSM 101813, is clearly referable to C. anecdotus with the charac-
teristic V-shaped anterior fossettid on P4.

Pterogaulus, new genus

Type Species. — Mylagaulus laevis Matthew, 1902.

Referred Species. — P. cambridgensis (Korth, 2000), P. harbarellae n. sp., and Pterogaulus sp.

Range. — Barstovian to Hemphillian of northern Great Plains.

Etymology. — Greek, pteron, wing; and gaulos, bucket.

Diagnosis. — Intermediate to large mylagaulines; horns or bosses lacking on
nasals; frontals with enlarged, wing-like postorbital processes (POI = 0.39-0.57);
angle of occipital in advanced species generally lower than in other genera
(OA < 60°) but not as low as in Hesperogaulusi occlusal pattern of premolars
with anteroposteriorly oriented linear fossettes(“ids); in advanced species, first and
second molars lost with eruption of permanent premolars; buccal branch of par-
afossette on P^ separates from remainder of fossette first; premolars more antero-
posteriorly elongated than in Ceratogaulus (P"^ width to length ratio < 0.68).

Fig. 18. — Premolars of Pterogaulus laevis and P. barbarellae. A, B. Pterogaulus laevis, AMNH 9043,
holotype. A, Right B. Right P4. C, D. Pterogaulus barbarellae. C. FAM 65491, holotype, left P’^.
D. FAM 65493, left P4, M2. Bar scale = 5 mm.

264

Annals of Carnegie Museum

VOL. 69

Pterogaulus laevis (Matthew, 1902)

(Fig. 18A, B, 19A; Table 7)

Mylagaulus monodon (in part) Matthew, 1901.

Mylagaulus laevis Matthew, 1902.

Type Specimen. — AMNH 9043, anterior half of skull with associated mandible
and pelvis (Matthew, 1901:figs. 3, 4, 5, 6).

Referred Specimens. — KU 9807, 9808, 9969; and specimens from Sioux County, Nebraska cited in
Matthew (1924:78).

Horizon and Locality. — Holotype and KU specimens from Pawnee Creek For-
mation of northeastern Colorado (Matthew, 1924), referred specimens from Olcott
Formation, Sioux County, Nebraska (Matthew, 1924; Skinner et ah, 1977).

Age. — Barstovian (middle Miocene).

Emended Diagnosis. — Smallest species of the genus; postorbital process small-
er than in other species (POI = 0.39-0.41); angle of occipital greater than in late
Barstovian and Clarendonian species (OA = 68°-75°); parafossette of P^ more
persistent than in other species (buccal branch separates very late in wear); fewer
number of fossettes (-ids) on premolars than in later species (P^: x = 6.5,
mode = 6; P4: x = 5.7, mode = 5).

Discussion. — Matthew (1901, 1902, 1924) fully described and figured the skull
and dentition of Pterogaulus laevis. The skeleton was fully described by Fagan
(1960).

This species is both temporally older and morphologically more primitive than
later species of this genus. Its early Barstovian occurrence predates that of all
other species. In its smaller size and morphology of the skull (angle of occipital,
size of postorbital processes) and dentition (fewer fossettes, separation of para-
fossette fork later in wear), it is clearly more primitive than the later species.

Pterogaulus sp.

(Table 7)

Discussion. — Several authors have previously identified specimens of this spe-
cies from the Barstovian of Nebraska and Saskatchewan as Mylagaulus cf. M.
laevis (Storer, 1975; Voorhies, 1990Z?; Evander, 1999). The cranial and dental
features of this species are clearly intermediate between those of the early Bar-
stovian Pterogaulus laevis and the late Clarendonian species described below.
This previously described material, and additional material more recently recov-
ered by UNSM clearly represents a new species. However, the description and
naming of this species will not be presented here; it is part of a larger faunal
study currently underway by workers at UNSM.

The stratigraphic range of Pterogaulus sp. is similar to that of Ceratogaulus
rhinocerus, but extends from the middle Barstovian to the middle Clarendonian,
slightly later than the last occurrence of C rhinocerus (early Clarendonian). There
appears to be no discernible difference in morphology or size between specimens
from the lowest level (Norden Bridge Quarry) and those from the highest level
(Cap Rock Member), unlike specimens of C. rhinocerus, which show advance-
ments in the skull in the early Clarendonian specimens.

While Pterogaulus sp. and C. rhinocerus are contemporaneous for most of their
known records, they rarely occur at the same quarry. For example, at the Norden
Bridge Quarry, West Valentine Quarry, and Stewart Quarry (see Voorhies, 1990i3,

2000

Korth — Miocene mylagaulid rodents from Nebraska

265

Fig. 19. — Skulls of Pterogaulus, dorsal view (top), ventral view (center), and left lateral view (bottom).
A. Pterogaulus laevis, AMNH 17576. B. Pterogaulus barbarellae holotype, FAM 65491. Scale = 1
cm.

and Voorhies and Timperley, 1997, for location of quarries), where Pterogaulus
sp. is abundant, there are no specimens of Ceratogaulus. In other quarries from
the same horizon in the Valentine Formation, such as the Cornell Dam Quarry,
only specimens of C rhinocerus have been recovered. This suggests that these
species were mutually exclusive in terms of geography, possibly competing for
resources with one another.

266

Annals of Carnegie Museum

VOL. 69

Table 7. — Measurements of premolars o/ Pterogaulus, Measurements from middle Barstovian sample
o/ Pterogaulus sp. from Voorhies ( 1990b: A84). Statistical abbreviations as in Table 4.

n

X

OR

5

cv

Pterogaulus laevis

P'* a-p

15

7.87

7.00-9.00

0.58

7.4

tr

15

4.97

4.35-5.45

0.30

6.0

P4 a-p

11

8.18

7.40-9.30

0.51

6.2

tr 11

Pterogaulus sp. (middle Barstovian)

4.12

3.60-5.00

0.34

8.3

P"^ a-p

8

9.0

8.3-10.0

—

—

tr

8

5.5

4.8-6.5

—

—

P4 a-p

9

9.4

8.7-10.2

—

—

tr 9

Pterogaulus sp. (Clarendonian)

5.0

4.5-5.5

P‘* a-p

9

8.9

7.6-10.2

0.8

8.9

tr

9

5.8

5.4-6.4

0.3

5.2

P4 a-p

2

9.8

8.5-11.2

—

—

tr

2

5.0

4.75-5.2

—

—

Pterogaulus barbarellae

P'^ a-p

19

10.48

8.80-12.50

1.20

11.48

tr

18

6.27

5.55-7.50

0.62

9.91

P4 a-p

18

11.40

9.45-13.30

1.19

10.41

tr

18

5.43

4.85-6.85

0.51

9.36

Pterogaulus barbarellae, new species
(Fig. 18C, D, 19B; Table 7)

Mylagaulus ? Gregory, 1942.

Mylagaulus cf. monodon Webb, 1969 (in part).

Mylagaulus cf. M. mondon Voohries, 1990a.

Mylagaulus sp., cf. M. mondon Korth, 1997 (in part).

Mylagaulus monodon Cope, Korth, 1998 (in part).

Type Specimen. — FAM 65491, skull with P^.

Referred Specimens.— PAM 65008, 65435, 65436, 65439-65443, 65446, 65448, 65454, 65465,
65475, 65491-65496, 65498, 65499, 65787, 65788, 65791, 65797.

Horizon and Locality. — Holotype from Xmas Quarry, Xmas and Kat Channels,
Ash Hollow Formation, Cherry County, Nebraska; referred specimens from var=
ious late Clarendonian localities in Merritt Dam Member, Ash Hollow Formation,
Nebraska.

Age. — Late Clarendonian (early-late Miocene).

Diagnosis. — Larger than P. vetus and Pterogaulus sp.; postorbital processes
larger than in any other species (POI = 0.40^0. 57); angle of occipital less than
in other species (OA = 52°); fossettes(-ids) on premolars more elongated and an-
teroposteriorly aligned than in other species but similar in number (P^: x = 7,8,
mode = 8; P4: x = 6.5, mode = 6); first two molars lost with eruption of perma-
nent premolar.

Etymology. — Patronym for Barbara Lamb.

Description. — The skull of Pterogaulus barbarellae is more advanced than that of Pterogaulus sp.
in having larger postorbital processes and having a lower angle of the occipital (Table 1). It is also
clearly larger in size (Table 7). All other features of the skull appear to be the same as in all other
species of the genus, with no indication of a horn or boss on the nasals.

2000

Korth — Miocene mylagaulid rodents from Nebraska

267

The premolars of P. barbarellae are larger than in other species in proportion to the rest of the
dentition. It appears that with the eruption of the permanent premolar, the first two molars are shed
instead of just M' or Mj. has a nearly equal number of fossettes to of Pterogaulus sp. with a
similar range (6 to 10) but a slightly greater mean number (7.8 rather than 7.2 for Pterogaulus sp.).
The buccal fork of the parafossette is separated from the rest of the fossette on nearly all specimens.
The only specimens that retain this connection are those of very young individuals with little or no
wear on the teeth.

On P4, the number of fossettids is not greater than in Pterogaulus sp. but they are relatively longer
and more directly anteroposteriorly oriented, forming an almost linear pattern on the occlusal surface.
The length of most of the main fossettids is approximately half the total length of the tooth.

Discussion. — Pterogaulus barbarellae is larger and more derived than either
of the older species of the genus. It clearly follows the trends in morphology of
the skull (larger postorbital processes, lower angle of occipital) and dentition
(larger premolars, more elongated fossettes) that are recognizable in earlier spe-
cies. Another species of this genus, P. cambridgensis, is present in the Hemphil-
lian of Nebraska (Korth, 2000). The latter also follows these same morphoclines
for the skull and dentition.

Ceratogaulus anecdotus is contemporaneous with P. barbarellae, similar to the
case of C. rhinocerus and Pterogaulus sp. in the Barstovian and early Claren-
donian. However, the occurrence of the late Clarendonian species does not appear
to be mutually exclusive as is the case with the Barstovian species. In at least
two localities (Beaman Creek, Pratt Quarry) specimens of both C. anecdotus and
P. barbarellae are present. However, the fossil quarries in the late Clarendonian
“Xmas and Kat Channels” (Skinner and Johnson, 1984) have produced only
specimens of Pterogaulus. There are no specimens of Ceratogaulus in the AMNH
Frick Collections from these quarries, although they were extensively collected.

Conclusions

Variation in Premolars within Species. — Black and Wood (1956) made a de-
tailed study of the progressive changes in the number and shape of fossettes on
the cheek teeth of the Hemingfordian ''Mesogaulus"' novellus. Shotwell (1958)
sectioned the premolars of a number of specimens of later Tertiary mylagaulids
to demonstrate ontogenetic changes in the species he discussed. Korth (1997) also
discussed ontogenetic changes in premolars from the late Clarendonian specimens
that he referred to Mesogaulus monodon (referred here to Pterogaulus barbarel-
lae). The pattern of change in the occlusal morphology of premolars of mylagau-
lines in an individual, and the range of variation in a population, appear to be
fairly consistent throughout the family.

The number of fossettes in a single premolar of any species varies throughout
the life of the individual. In unworn specimens of premolars that are just reaching
the level of occlusion, the number of fossettes is the fewest. At this time the major
fossettes are often joined at their ends, forming a number of star-like fossettes
with many branches. As the tooth reaches an intermediate stage of wear, the major
fossettes separate from one another, and the highest number of fossettes is at-
tained. In most senile individuals with heavily worn premolars, the number is
near the maximum, with only a few of the smallest fossettes, which are shallower,
being lost. At this time the proportions of the tooth are such that the tooth is
much narrower than at any other time, and the fossettes(“ids) are generally more
anteroposteriorly aligned and closely packed together.

Because the premolar tapers toward its base, once the tooth has worn past its
maximum dimensions it becomes surrounded by a layer of cement. In Mesogaulus

268

Annals of Carnegie Museum

VOL. 69

and primitive mylagaulines, this occurs in very late stages of wear. In the more
advanced species, where the premolar is proportionally much larger, cement is
present at the occlusal surface at an earlier stage of wear.

The parafossette of also shows some changes through the life of an individ-
ual. Originally, the parafossette maintains its two anterior branches. In more prim-
itive species, these remain attached to the remainder of the fossette until very late
stages of wear. When there is a preferred pattern of separation of the anterior
forks of the parafossette, as in Pterogaulus or Ceratogaulus, this separation occurs
at progressively earlier stages of wear (Fig. 20). Ultimately, in most specimens
that are at extremely late stages of wear, both forks are separated from the main
part of the parafossette and are minute, circular fossettes.

The greatest source of variation in the number of fossettes(-ids) on the pre-
molars of a single sample of a mylagaulid species is the division of one of the
major (longer) fossettes into two or more smaller fossettes. It is not uncommon
to have specimens that have divided even two of these major fossettes into more,
smaller fossettes. In looking at the range of variation in number of fossettes in a
single sample, it appears that the calculation of the mean and modal number of
fossettes is very useful. In some genera, such as Umbogaulus, there is a distinct
progressive change over time in the number of fossettes among its species. The
minimum number of fossettes also appears to be consistent within a single species.
If the maximum number of fossettes is dependent on the division of the “major”
fossettes, the minimum number will be little changed within a single sample.

Variation in Premolars among Species. — There is a general trend in mylagau-
lids, from the early, more primitive species to the latest, most specialized species:
the number of fossettes(-ids) on the premolars increases. However, within any
single genus, the number appears to increase by “steps” rather than in the form
of a gradual change. The best example of a gradual change in the number of
fossettes is in Umbogaulus where the later species, U. monodon, has fewer fos-
settes on the premolars than the earlier U. galushai. This difference is in the mean
number of fossettes rather than in the range, maximum, or even minimum number.
The lineage with the best evidence for a gradual increase in the number of fos-
settes is the one represented by species previously referred to Mylagaulus cf.
laevis or M. cf. monodon from the Great Basin (Shotwell, 1958:table 2), later
assigned to Hesperogaulus (Korth, 1999<2). The range and mean number of fos-
settes in this lineage clearly increases from the early Barstovian sample {H. gazini)
through the Hemphillian sample {H. wilsoni).

Within most genera, the change in the fossettes is morphological, or the change
in the number of fossettes occurs in steps. For example, within the species of
Pterogaulus, the range and mean number of fossettes on the premolars change
very little from the early Barstovian P. laevis to the middle Hemphillian P. cam-
bridgensis (Table 2), but the relative size of the fossettes does change. In the early
species, the fossettes are elongated, but are no longer than about one-fourth the
total length of the premolar. In P. barbarellae and P. cambridgensis, the fossettes
are at least half the length of the tooth.

Evolutionary Changes in Mylagaulidae. — Even in the earliest known skulls of
promylagaulines (McGrew, 1941; Nichols, 1976) many of the derived characters
of mylagaulids are present: posteriorly broadened cranium with well-developed
occipital crests; elongated external auditory meatus; fusion of cranial sutures early
in life; enlarged last premolar; and increased hypsodonty of the cheek teeth. How-
ever, the degree of development of these features becomes much greater in the

2000

Korth — Miocene mylagaulid rodents from Nebraska

269

HESPEROGAULUS

PTEROGAULUS and

ALPHAGAULUS VETUS

OCERAT0GAULUS,

MYLAGAULUS, and

PRIMITIVE CONDITION

MESOGAULUS

Fig. 20. — Schematic diagrams of right showing the development of the parafossette in different
mylagaulids.

270

Annals of Carnegie Museum

VOL. 69

mylagaulines, where the cheek teeth attain nearly complete hypsodonty, anterior
molars are lost with eruption of the enlarged permanent premolar, and the pos-
terior width of the skull is nearly equal to its total anteroposterior length. Meso~
gaulus is clearly intermediate between the promylagauline condition and that of
the mylagaulines. Even in the broadening of the bones of the forearm, promyla-
gaulines appear less derived (Table 3; Korth, 19991?).

As in many families of mammals, there appears to be a disinct increase in size
through time, with only a few exceptions. Promylagaulines are generally small
species, while the majority of mylagaulines are among the largest rodents of their
time. In most genera, such as He spero gaulus, Pterogaulus, and Ceratogaulus, the
later species are larger than the earlier species (Korth, 1999a; Tables 6, 7). The
exception is Mylagaulus, which appears to decrease in size through time (Baskin,
1980).

The morphology of the cheek teeth follows a similar pattern in most genera.
The mylagaulines have many more fossettes on the premolars than do promyla-
gaulines, along with increased hypsodonty and relative size of the premolar com-
pared to the molars. The number or relative size of the fossettes(-ids) on the
premolars changes differently within different genera (see above discussion). As
the relative size of the premolars increases, the number of molars lost with the
eruption of the permanent tooth also increases. In promylagaulines (with the ex-
ception of Galbreathia novellus [Korth, 1999Z?]), and all the molars are retained
in adult individuals. In Mesogaulus, P^ is retained but the first molar is lost. In
all mylagaulines, at least the first molars are lost with the eruption of the pre-
molars. In advanced (Hemphillian) species of Ptero gaulus and Hespero gaulus,
the first two molars are shed.

Several changes have occurred in cranial morphology among mylagaulines as
well. The skull of all mylagaulines has the same overall shape. The skull is
posteriorly broadened with well-developed occipital crests, the cranium is rela-
tively low, the rostrum is short, and the zygoma is robust and broad. Primitively,
there are small swellings near the anterior end of the nasals. In all mylagaulids,
the cranial sutures over most of the skull completely fuse early in life and are
not traceable on adult individuals. The greatest changes in cranial morphology
involve the modifications of the nasal bones, relative development of the post-
orbital processes, and the angle of the occipital. In Alphagaulus, the most prim-
itive of the mylagaulines, the angle of the occipital ranges from 90°“60°, the
postorbital processes are intermediate in size (POI = 0.25-0.36), and small bosses
are present on the nasals. Within each lineage (genus) the changes in these features
are unique to each genus.

In the short-lived lineage of Umbogaulus, the only modification is in the size
of the bosses on the nasals. In Umbogaulus the bosses become very large and
retain a spherical shape. The angle of the occipital and relative size of the post-
orbital processes remain as in Alphagaulus. In Pterogaulus the postorbital pro-
cesses become greatly enlarged (POI = 0.34-0.57), the nasals are dorsally smooth
(small swellings lost), and the range of the angle of the occipital is lower
(OA = 52°-75°). The skull of Ceratogaulus develops paired horns (or hom cores)
on the nasals, and the postorbital processes are greatly reduced (POI = 0.1 1-0.19).
Within Ceratogaulus, the nasal horns develop progressively more posteriorly in
a temporal series of species. The angle of the occipital is not as low as in Pter-
ogaulus (OA = 60°“70°), but the shape of the occipital crest is modified. The
primitive shape of the occipital crests for Mesogaulus and mylagaulines in dorsal

2000

Korth — Miocene mylagaulid rodents from Nebraska

271

view is that it is normal to the long axis of the skull at its center, and as it descends
laterally, there is a distinct anterior flexure or bend in the crest at a point approx-
imately one-half the distance from the center of the skull to its lateral border.
However, in Ceratogaulus, the occipital crest is a straight line in dorsal view. The
anterior bend has been lost.

The skull of Hesperogaulus has been modified by greatly reducing the angle
of the occipital (POI = 50°) and broadening and flattening the nasals anteriorly
(Korth, 1999a). The occipital crests on skulls of Barstovian Hesperogaulus are
similar to those of other genera with anterior bends. However, in the Hemphillian
species, the flexion is lost and the crest forms a straight line as in Ceratogaulus.
This feature is clearly attained separately in Hesperogaulus and Ceratogaulus. It
is achieved in the former as part of the decrease in the occipital angle, which
pushes the anterior margin of the occipital farther anterior.

Phylogeny of the MylagauUnae. — It is impossible at this time to determine
whether any of the known Arikareean promylagaulines is directly ancestral to the
later mylagaulines. It is evident that the latest of the promylagaulines (late Hem-
ingfordian-early Barstovian) are more derived than Mesogaulus and are clearly
not the ancestors of any mylagaulines (Korth, 1999^). Rensberger (1979) sug-
gested that none of the species of promylagaulines that he studied were ancestral
to later mylagaulines. However, it is likely that some unspecialized promylagau-
line in the late Arikareean or early Hemingfordian was morphologically transi-
tional between the promylagaulines and Mesogaulus.

In the proportions of the skull and skeleton and in dental adaptations (enlarge-
ment of premolars, loss of molars), Mesogaulus is distinctly transitional between
promylagaulines and mylagaulines. Not only are the cranial proportions inter-
mediate between the latter two subfamilies, but the dentition is also similar to
that of mylagaulines and advanced over promylagaulines in some characters (loss
of molars), but similar to promylagaulines and more primitive than mylagaulines
in others (retention of P^). There is nothing in the morphology of Mesogaulus
that would exclude it from the ancestry of all mylagaulines (Fig. 21).

Within the Mylagaulinae, Alphagaulus is the earliest (late Hemingfordian-early
Barstovian) and most primitive genus. This is evidenced by the proportions of
the skull, presence of a small boss on the nasals, and angle of the occipital (Table
1), as well as the relatively fewer number of fossettes(-ids) on the premolars
(Table 2). At least two of the later lineages (genera) of mylagaulines can be
derived from species of Alphagaulus. Both A. vetus and A. pristinus have rela-
tively gracile skulls and separate the buccal fork of the parafossette on P"^ with
wear. These characters can be traced to species of Pterogaulus and even possibly
species of Hesperogaulus (Korth, 1999a). Alphagaulus douglassi and A. tedfordi
have more robust skulls, a larger nasal boss, and separate the lingual fork of the
parafossette on P"^; characters found in Ceratogaulus and Mylagaulus. Besides
Alphagaulus, five genera appear in the Barstovian. The earliest is Umbogaulus,
which is restricted to the Barstovian. This genus is generally primitive cranially,
but develops a much larger boss on the nasals than do species of Alphagaulus.
Dentally, this genus is far advanced over the other genera of the Barstovian in
the much greater number of fossettes on the premolars. Umbogaulus is short-
lived, restricted temporally to the Barstovian, and clearly not ancestral to any
other. The only character that nfight relate Umbogaulus with any other genus is
the enlarged bosses on the nasals, which are more similar to the nasal horns of

272

Annals of Carnegie Museum

VOL. 69

PROMYLAGAULINAE

MESOGAULiNAE (Mesogauius)

Alphagaulus

Umbogaulus

Rerogauius

Hesperogaulus

Ceratogaulus

Mylagaulus

Fig. 21. — Cladogram of relationships of mylagaulids. Explanation of nodes; 1. Mylagaulidae: pro-
trogomorphous zygomasseteric structure; uniserial microstructure of incisor enamel; incisors broad-
ened; low, broad, robust skull with heavy occipital crest and shortened rostrum; postorbital process
present on frontals; cranial sutures completely fuse relatively early; auditory bulla with elongated
external meatus; postcranial skeleton fossorially adapted; last premolars largest cheek teeth; cheek
teeth lophate, occlusal pattern reduced to isolated enamel “lakes” (cusps reduced); cheek teeth pro-
gressively hypsodont. 2. Larger size; posterior width of skull subequal to anteroposterior length; post-
cranial skeleton fossorially adapted (humeral index > 0.40); first molars lost with eruption of perma-
nent premolars; premolars at least twice the size of molars. 3. Mylagaulinae: lost; broad, shallow

groove on upper incisor; greater number of fossettes(-ids) on premolars (minimum number and six
for five for P4); premolars completely hypsodont (roots lacking); small paired bosses on antero-
dorsal end of nasal bones; double sagittal crest. 4. Largest mylagaulids; premolars nearly oval in
occlusal outline (loss of outline of stylar cusps); occipital anteriorly tilted (OA < 90°). 5. Umbogaulus,
nasal bosses greatly enlarged; number of fossettes(-ids) on premolars increased (x = 8-9); upper pre-
molars wider relative to length than other mylagaulines. 6. Pterogaulus, postorbital processes greatly
enlarged (POI = 0.35-0.54); bosses on nasals lost; buccal fork of parafossette of P"* separates first;
premolars narrower relative to length than in other mylagaulids. 7. Hesperogaulus, angle of occipital
lower than other mylagaulids (OA = 50-57°); bosses on nasal bones low and anteriorly broadened
and squared-off; either of the anterior branches of forked parafossette of P'^ may separate first. 8.
lingual fork of parafossette on P'^ separates first; premolars wider relative to length than in Pterogaulus.
9. Ceratogaulus, horns (horn cores) on nasal bones; postorbital process progressively reduced

2000

Korth — Miocene mylagaulid rodents from Nebraska

273

Ceratogaulus than to the condition in any other genus. Again, the premolars of
Umbogaulus are too complex to have been ancestral to Ceratogaulus.

A second lineage of mylagauline is represented by Pterogaulus. This genus is
restricted to the northern Great Plains and is characterized by the loss of nasal
bosses and enlargement of the postorbital process. It ranges from the Barstovian
to the medial Hemphillian. Dentally, it can be distinguished by the separation of
the buccal fork of the parafossette on which occurs at a progressively earlier
wear stage in later species. This lineage shows through time an increase in skull
size and proportional size of the premolar, increased size of the postorbital pro-
cess, and increased length of the fossettes on the premolars.

Ceratogaulus is restricted temporally and geographically to the same time span
as Pterogaulus, and they are often found in the same faunal horizons. It, too,
shows a distinct increase in size over time along with the more posterior postition
of the nasal horns and reduction of the postorbital process (Table 1).

A fifth genus, Hesperogaulus, from the Great Basin, originates in the Barsto-
vian and exists until the Hemphillian, and exhibits an increase in size as well as
complexity of the premolars (Korth, 1999 a).

Biogeography of the Mylagaulinae. — Promylagaulines appear in the fossil re-
cord during the Arikareean and are limitied geographically to the northern Great
Plains and Rocky Mountains (McGrew, 1941; Nichols, 1976; Rensberger, 1979,
1980; Korth, 1992). In the early Hemingfordian, species of Mesogaulus appear
in approximately the same area (Riggs, 1899; Galbreath, 1953; Wilson, 1960).
By the late Hemingfordian, the first mylagaulines appear alongside advanced pro-
mylagaulines, again limited to the northern Great Plains and Rocky Mountains.
During the Barstovian the remainder of the genera of mylagaulines appear. Three
of these genera {Umbogaulus, Pterogaulus, and Ceratogaulus) are limited geo-
graphically to the northern Great Plains. However, another distinctive lineage,
Hesperogaulus, appears in the northern Great Basin (Shotwell, 1958; Korth,
1999a). The more primitive mylagaulines {Alphagaulus and Umbogaulus), along
with the specialized promylagaulines (Korth, 1999^), do not survive the end of
the age. The Clarendonian and Hemphillian occurrences of the four surviving
genera show a distinct geographic isolation that began in the Barstovian. Hes-
perogaulus continues into the Hemphillian in Nevada and Oregon (Shotwell,
1958; Korth, 1999a). Clarendonian and Hemphillian species of Mylagaulus are
known only from Florida (Webb, 1966; Baskin, 1980), even though this lineage
probably originated in the Great Plains alongside two other genera, Ceratogaulus
and Pterogaulus in the Barstovian or Clarendonian. The latter two genera are
completely restricted to the northern Great Plains throughout their fossil records.

The geographic distribution of the species of mylagaulids (Fig. 22) suggests a
number of conclusions. First, the origin and early radiation of the family was
within the northern Great Plains and Rocky Mountains. Then, with the rapid
diversification of mylagauline genera beginning in the Barstovian, there is a geo-
graphic dispersal of these genera with one lineage migrating to the Great Basin

(POI = 0.24-0.1 1); occipital crest more robust than in other mylagaulines and forms straight transverse
crest (no anterior bend on either side as in other mylagaulines); anterocentral fossettid on P4 V-shaped
in early wear (apex pointed posteriorly), lingual branch separates in early wear. 10. Mylagaulus:
posterobuccal fossette on P"^ C-shaped; reduction in size and complexity of premolars.

274

Annals of Carnegie Museum

VOL. 69

Arikareean Hemingfordian Barstovian Clarendonian Hemphiifian

1 ...1 -J- -J_—

Hesperogaulus

GREAT BASIN

Alphagaufus

Mesogaulus ^

— ^

ROCKY

MOUNTAINS

3’

Mesogaulus S

Alphagaulus

Umbogaulus

NORTHERN
GREAT PLAINS

Ceratogaulus

\

FLORIDA

Mylagaulus

Fig. 22. — Biostratigraphic and geographic occurrence of mylagaulids.

(Hesperogaulus) and another to Florida (Mylagaulus). Only two lineages remain
in the Great Plains until the end of the Hemphillian.

Among the specimens of the AMNH Frick collections are a number of myla-
gaulids from the Barstovian of New Mexico and the Hemphillian of Texas. These
specimens were not included in this study; however, the New Mexico material
may represent yet another morphologically and geographically distinct lineage.
This lineage cannot be directly related to the Texas material from the Hemphillian
at this time, but may prove to be related taxa.

Sexual Dimorphism. — In his initial description of Ceratogaulus rhinocerus,
Matthew (1902) suggested that the horns on the nasals of this species were pos-
sibly a sexually dimorphic character, and that the skull of Mylagaulus"^ laevis,
from the same horizon which lacked horns, was the skull of a female individual.
He rejected this idea (Matthew, 1902, 1924) mainly because there was no known
example in rodents. Recent or fossil, that had this scale of dimorphism. It is now
evident, with the discovery and description of numerous mylagaulid skulls, that
the horns on the skulls of species of Ceratogaulus are a distinctive generic char-
acter and not a sexual one due to the fact that other cranial characters (size of
postorbital processes) and dental characters (morphology of parafossette on
and anterocentral fossettid on P4) clearly separate the homed mylagaulids {Cer-
atogaulus) from the hornless mylagaulids (Pterogaulus) of the Great Plains.

However, in one lineage of mylagaulines there may be sexual dimorphism. In

2000

Korth — Miocene mylagaulid rodents from Nebraska

275

the species of mylagaulids from the Great Basin, cited by Shotwell (1958) that
were later referred to Hesperogaulus (Korth, 1999a), there appears to be a dif-
ference in the relative robustness of the skulls of individuals with no evidence of
a difference in dental dimensions or morphology. Skulls of H. gazini from the
Barstovian and H. wilsoni from the Hemphillian of Oregon (Korth, 1999a) are
divisible into larger and smaller sizes, suggesting a dimorphic condition. In no
other species is there even a suggestion of sexual dimorphism. However, this
disparity in size of skulls in Hesperogaulus may be due to the lack of an adequate
sample for comparison.

Problematical Species. — ^There are examples of specimens or species of my-
lagaulids that are not clearly assignable to known genera or lineages of mylagau-
lids. These include Mylagaulodon and a pair of unusual specimens from the Clar-
endonian of Oregon.

Mylagaulodon anguiatus was first named by Sinclair (1903) from the Heming-
fordian of Oregon, and was believed to represent a transitional form between
meniscomyine aplodontids and mylagaulids. Matthew (1924) argued that the pre-
molars of the holotype of M. anguiatus (UCMP 1652) were deciduous teeth and
that the specimen was a juvenile of Mylagaulus. Despite Matthew’s suggestion,
later authors continued to consider Mylagaulodon as the basal mylagaulid
(McGrew,1941).

Subsequent identifications of specimens of M. anguiatus from the Great Plains
(McGrew, 1941; Skwara, 1988) appear to be specimens of other mylagaulids
(Korth, 1992:91). Korth (1992) suggested that M. anguiatus represented an ad-
vanced meniscomyine aplodontid raliiei than a mylagaulid, later questionably in-
cluding M. anguiatus in the Meniscomyinae (Korth, 1994). Whether the type of
Mylagaulodon is an advanced meniscomyine or a juvenile mylagaulid cannot be
determined at this time. However, it does appear that it does not represent a
species transitional between the mylagaulids and an aplodontid ancestor as had
been previously suggested.

Shotwell (1958: fig 13) identified two isolated P^s from the Clarendonian of
Juntura Basin as Epigaulus minor (UOMNH F6165, F6166). Later, he (Shotwell
and Russell, 1963) suggested that these two premolars were from the same in-
dividual. These two specimens are unique among all described mylagaulids. The
fossettes are unusually branched and there is a deep reentrant valley along the
buccal side of the tooth that is reminiscent of the mesoflexus of the upper cheek
teeth of beavers. These specimens clearly belong to a mylagaulid, but do not fit
into any genus recognized at the present time. These specimens should be con-
sidered as belonging to an indeterminate mylagaulid.

Acknowledgments

Access to and/or loan of the specimens examined for this study were provided by the following
individuals: R. H. Tedford (AMNH), M. R. Voorhies (UNSM), D. Whistler (LACM), L. R. Martin
(KU), and A. D. Barnosky (UCMP). Assistance was also given in locating and processing the speci-
mens by R. G. Corner (UNSM) and R. L. Evander (AMNH). This project was funded in part by
grants from the American Museum of Natural History, Buffalo State College (Foundation Grant), and
private contributors to the Rochester Institute of Vertebrate Paleontology. The photographs for figures
7 and 8 were made by Lorraine Meeker (AMNH). Photographic equipment and access to a camera
lucida for many of the drawings were graciously provided by G. McIntosh of the Rochester Museum
and Science Center. Earlier versions of this manuscript were critically reviewed by J. H. Wahert, M.
R. Dawson, and an anonymous reviewer.

276

Annals of Carnegie Museum

VOL. 69

Literature Cited

Baskin, J. A. 1979. Small mammals of the Hemphillian age White Cone local fauna, northeastern
Arizona. Journal of Paleontology, 53:695-708.

. 1980. Evolutionary reversal in Mylagaulus (Mammalia, Rodentia) from the late Miocene of

Florida. American Midland Naturalist, 104:155-162.

Black, C. C. 1961. Rodents and lagomorphs from the Miocene Fort Logan and Deep River formations
of Montana. Postilla, 48:1-20.

Black, C. C., and A. E. Wood. 1956. Variation and tooth replacement in a Miocene mylagaulid
rodent. Journal of Paleontology, 30:672-684.

Cook, H. J., and J. T. Gregory. 1941. Mesogaulus praecursor, a new rodent from the Miocene of
Nebraska. Journal of Paleontology, 15:549-552.

Cope, E. D. 1878. Descriptions of new extinct Vertebrata from the Upper Tertiary and Dakota for-
mations. Bulletin of the U. S. Geological and Geographical Survey of the Territories, 1878:379-
396.

. 188M. Review of the Rodentia of the Miocene period of North America. Bulletin of the U.

S. Geological and Geographical Survey of the Territories, 6:361-386.

. 1881/?. The Rodentia of the American Miocene. American Naturalist, 15:586-587.

Cope, E. D., and W. D. Matthew. 1915. Hitherto unpublished plates of Tertiary Mammalia and
Permian Vertebrata. American Museum of Natural History, Monograph Series 2.

Douglass, E. 1903. New vertebrates from the Montana Tertiary. Annals of Carnegie Museum, 2:
145-199.

Evander, R. L. 1999. Rodents and lagomorphs (Mammalia) from the Railway Quarries local fauna
(Barstovian, Miocene) of Nebraska. Paludicola, 2:240-256.

Fagan, S. R. 1960. Osteology of Mylagaulus laevis, a fossorial rodent from the Upper Miocene of
Colorado. University of Kansas Paleontological Contributions, Vertebrata, 9:1-32.

Fiorillo, a. R. 1988. Taphonomy of Hazard Homestead Quarry (Gallala Group), Hitchcock County,
Nebraska. Contributions to Geology, University of Wyoming, 26:57-97.

Galbreath, E. C. 1953. A contribution to the Tertiary geology and paleontology of northeastern
Colorado. University of Kansas Paleontological Contributions, Vertebrata, 4:1-120.

— . 1984. On Mesogaulus paniensis (Rodentia) from Hemingfordian (middle Miocene) deposits

of northeastern Colorado. Pp. 85-89, in Papers in Vertebrate Paleontology Honoring Robert War-
ren Wilson (R. M. Mengel, ed.), Carnegie Museum of Natural History Special Publication no. 9.
Gidley, j. W. 1907. A new horned rodent from the Miocene of Kansas. Proceedings of the U. S.
National Museum, 32:627-636.

Gregory, J. T. 1942. Pliocene vertebrates from Big Spring Canyon, South Dakota. University of
California Publications, Geological Sciences, 26:307-446.

Hibbard, C. W., and L. F. Phillis. 1945. The occurrence of Eucastor and Epigaulus in the Lower
Pliocene of Trego County, Kansas. University of Kansas Science Bulletin, 30:549-555.

Korth, W. W. 1992. Fossil small mammals from the Harrison Formation (late Arikareean: earliest
Miocene), Cherry County, Nebraska. Annals of Carnegie Museum, 61:69-131.

. 1994. The Tertiary Record of Rodents in North America. Plenum Press, New York, New

York.

. 1997. Additional rodents (Mammalia) from the Clarendonian (Miocene) of northcentral Ne-
braska and a review of Clarendonian rodent biostratigraphy of that area. Paludicola, 1:97-111.

. 1998. Rodents and lagomorphs (Mammalia) from the late Clarendonian (Miocene) Ash

Hollow Formation, Brown County, Nebraska. Annals of Carnegie Museum, 67:299-328.

. \999a. Hesperogaulus, a new genus of mylagaulid rodent (Mammalia) from the Miocene

(Barstovian to Hemphillian) of the Great Basin. Journal of Paleontology, 73:938-944.

. 1999/?. A new genus of specialized promylagauline mylagaulid rodent (Mammalia) from the

early Miocene (Hemingfordian-Barstovian). Journal of Vertebrate Paleontology, 19:752-756.

— . 2000. A new species of mylagaulid rodent (Mammalia) from the late Miocene (Hemphillian)

of Frontier County, Nebraska. Paludicola, 2:265—268.

Matthew, W. D. 1901. Fossil mammals of the Tertiary of northeastern Colorado. Memoirs of the
American Museum of Natural History, 1:355-447.

. 1902. A horned rodent from the Colorado Miocene, with a revision of the Mylagauli, beavers,

and hares of the American Tertiary. Bulletin of the American Museum of Natural History, 16:
291-310.

. 1924. Third contribution to the Snake Creek fauna. Bulletin of the American Museum of

Natural History, 50:59-210.

2000

Korth — Miocene mylagaulid rodents from Nebraska

277

McGrew, P. O. 1941. The Aplodontoidea. Field Museum of Natural History, Geological Series, 9:
1-30.

McKenna, M. C. 1955. A new species of mylagaulid from the Chalk Cliffs local fauna, Montana.
Journal of the Washington Academy of Sciences, 45:107-110.

Munthe, J. 1988. Miocene mammals of the Split Rock area, Granite Mountains Basin, central Wy-
oming. University of California Publications in Geological Sciences, 126:1-136.

Nichols, R. 1976. Early Miocene mammals from the Lemhi Valley of Idaho. Tebiwa, 18:9-47.

Osborn, H. F. 1918. Equidae of the Oligocene, Miocene, and Pliocene of North America. Memoirs
of the American Museum of Natural History, 2:1-326.

Rensberger, J. M. 1979. Promylagaulus, progressive aplodontoid rodents of the early Miocene.
Contributions to Science, Natural History Museum of Los Angeles County, 312:1-16.

— . 1980. A primitive promylagauline rodent from the Sharps Formation, South Dakota. Journal

of Paleontology, 54:1267-1277.

Riggs, E. S. 1899. The Mylagaulidae: an extinct family of sciuromorph rodents. Field Columbian
Museum, Geological Series, 1:181-187.

Shotwell, j. A. 1958. Evolution and biogeography of the aplodoetid and mylagaulid rodents. Evo-
lution, 12:451-484.

Shotwell, J. A., and D. E. Russell, 1963. Mammalian fauna of the Upper Juntura Formation, the
Black Butte local fauna. Transactions of the American Philosophical Society, 53:42-69.

Sinclair, W. S. 1903. Mylagaulodon, a new rodent from the Upper John Day of Oregon. American
Journal of Science, 15:143-144.

Skinner, M. F., and F. W. Johnson. 1984. Tertiary stratigraphy and the Frick Collection of fossil
vertebrates from north-central Nebraska. Bulletin of the American Museum of Natural History,
178:1-368.

Skinner, M. E, S. M. Skinner, and R. J. Gooris. 1977. Stratigraphy and biostratigraphy of the late
Ceeozoic deposits in central Sioux County, western Nebraska. Bulletin of the American Museum
of Natural History, 158:263-370.

Skwara, T, 1988. Mammals of the Topham local fauna: early Miocene (Hemingfordian), Cypress
Hills Formation, Saskatchewan. Natural History Contributions, Saskatchewan Museum of Natural
History, 9:1-169.

Stirton, R. A. 1940. Phytogeny of North American Equidae. University of California Publications
in Geological Sciences, 25:165-198.

Storer, I. E. 1975. Tertiary mammals of Saskatchewan. Part III: the Miocene fauna. Life Science
Contributions to the Royal Ontario Museum, 103:1-134.

Sutton, J. E, and W. W. Korth. 1995. Rodents (Mammalia) from the Barstovian (Miocene) Anceney
local fauna, Montana. Annals of Carnegie Museum, 64:267-314.

Tedford, R. H., M. F. Skinner, R. W. Fields, J. M. Rensberger, D. P. Whistler, T. Galusha, B. E.
Taylor, J. R. Macdonald, and S. D. Webb. 1987. Faunal succession and biochronology of the
Arikareean through Hemphillian interval (late Oligocene through earliest Pliocene rocks) in North
America. Pp. 153-210, in Cenozoic Mammals of North America; Geochronology and Biostratig-
raphy (M. O. Woodbume, ed.). University of California Press, Berkeley, California.

VooRHiES, M. R. 1990a. Vertebrate biostratigraphy of the Ogallala Group in Nebraska. Pp. 115-151,
in Geologic Framework and Regional Hydrology, Upper Cenozoic Blackwater Draw and Ogallala
Formation, Great Plains (T. C. Gustavson, ed.). Bureau of Economic Geology, University of Texas,
Austin, Texas.

. 19901?. Vertebrate paleontology of the proposed Norden Reservoir area Brown, Cherry, and
Keya Paha counties, Nebraska. Technical Report 82-09, Division of Archeological Research,
University of Nebraska-Lincoin, Lincoln, Nebraska.

VooRHiES, M. R. and C. L. Timperley. 1997. A new Pronotolagus (Lagomorpha: Leporidae) and
other leporids from the Valentine Railway quarries (Barstovian, Nebraska), and the archaeolagine-
leporine transition. Journal of Vertebrate Paletoetology, 17:725-737.

Wahlert, j. H. 1974. The cranial foramina of protrogomorphous rodents; an anatomical and phy-
logenetic study. Bulletin of the Museum of Comparative Zoology, 146:363-410.

Webb, S. D. 1966. A relict species of the burrowing rodent, Mylagaulus, from the Pliocene of Florida.
Journal of Mammalogy, 47:401-412.

. 1969. The Burge and Minnechaduza Clarendonian mammalian faunas of north-central Ne-
braska. University of California Publications, Geological Sciences, 78:1-191.

Wilson, R. W. 1937a. New middle Pliocene rodent and lagomorph faunas from Oregon and Cali-
fornia. Carnegie Institute of Washington Publications, 487:1-19.

- — - — 19371?. Pliocene rodents of western North Am.erica. Carnegie Institute of Washington Pub-
lications, 487:23-73.

278

Annals of Carnegie Museum

VOL. 69

. 1960. Early Miocene rodents and insectivores from northeastern Colorado. University of

Kansas Paleontological Contributions, Vertebrata, 7:1-92.

Wood, A. E., and R. W. Wilson. 1936. A suggested nomenclature for the cusps of the cheek teeth
of rodents. Journal of Paleontology, 10:388-391.

Wood, H. E., II, R. W. Chaney, J. Clark, E. H. Colbert, G. L. Jepsen, J. B. Reeside, Jr., and C.

Stock. 1941. Nomenclature and correlation of the North American continental Tertiary. Bulletin
of the Geological Society of America, 52:1—48.

2000

Korth — Miocene mylagaulid rodents from Nebraska

279

Appendix 1

Cranial measurements of mylagaulid specimens examined in this study (see Figure 1 for orientation
of measurements). Abbreviations for cranial measurements: L, length; PW, maximum posterior width;
POC, minimum postorbital constriction; POP, length of postorbital process; OA, angle of the occipital
with the plane of the palate; NL, horizontal distance of the apex of the nasal horn from the anterior
end of the nasals (Ceratogaulus only). All measurements in mm. Asterisk indicates approximate
measurement.

Specimen

L

PW

POC

POP

OA

NL

Horizon

Mesogaulus paniensis

FAM 65511

48.05

39.5

11.05

4.05

90°

—

Marsland Fm., NE

Alphagaulus vetus

AMNH 18903

—

51.2

14.1

4.3

—

—

Sheep Creek Fm., NE

AMNH 20507

58.7

51.6

15.4

4.9

75°

—

Sheep Creek Fm., NE

FAM 65515

62.5

50.1

14.6

—

60°

—

Sheep Creek Fm., NE

FAM 65532

62*

51.5

15.8

5.8

70°

—

Observation Quarry, NE

FAM 65533

—

—

16.2

4.1

—

—

Observation Quarry, NE

Alphagaulus pristinus

AMNH 21307

—

22.65

11.20

3.50

79°

—

“Deep River Beds,” MT

Alphagaulus tedfordi

FAM 65711

64.7

78.8

14.60

3.60

90°

—

?Hemingfordian, NE

Umbogaulus monodon

FAM 65016

79.3

69.35

20.00

7.10

75°

—

Olcott Fm., NE

Ceratogaulus rhinocerus

AMNH 9456

66.7

70.6

17.50

4.55

70°

9.85

Pawnee Creek, CO

FAM 65815

—

—

23.75

4.20

—

—

Crookston Bridge Mb., NE

FAM 65014

—

—

—

—

—

19*

Devil’s Gulch Mb., NE

FAM 65015

—

57.4

—

—

—

—

Devil’s Gulch Mb., NE

FAM 65802

—

66.8

—

—

—

—

Devil’s Gulch Mb., NE

FAM 65012

70.7

60.3

16.60

2.05

—

12*

Burge Mb., NE

FAM 65013

—

67.8

—

—

68°

—

Burge Mb., NE

FAM 65385

79.35

70.0

17.40

2.15

67°

17.40

Burge Mb., NE

FAM 65833

—

70.4

17.00

2.50

—

—

Burge Mb., NE

FAM 65489

76.2

60.7

18.00

2.40

—

18.30

Burge Mb., NE

Ceratogaulus anecdotus

FAM 65466

63.0

63.0

17.25

2.12

—

17.5

Cap Rock Mb., NE

Pterogaulus laevis

AMNH 17576

58.0

56.0

12.45

5.40

—

—

Olcott Fm., NE

FAM 65471

63.6

52.4

12.00

5.85

—

—

Olcott Em., NE

Pterogaulus sp.

FAM 65008

67.35

56.2

19.30

6.60

57°

—

Burge Mb., NE

FAM 65009

—

54.5

15.80

5.95

54°

—

Burge Mb., NE

FAM 65010

• —

55.0

17.60

6.15

55°

—

Burge Mb., NE

FAM 65470

66.5

59.8

19.00

7.90

48°

—

Cap Rock Mb., NE

Pterogaulus barbarellae

FAM 65435

67.9

61.6

17.40

8.90

—

—

Merritt Dam Mb., NE

FAM 65436

—

—

16.70

9.50

—

—

Merritt Dam Mb., NE

FAM 65491

72.9

63.3

18.60

9.30

52°

—

Merritt Dam Mb., NE

FAM 65499

69.2

73.0

19.90

9.05

52°

—

Merritt Dam Mb., NE

280

Annals of Carnegie Museum

VOL. 69

Appendix 2

Classification of Mylagaulidae included in this paper

Family Mylagaulidae Cope, 1881
Subfamily Mesogaulinae, n. subfam.

Mesogaulus Riggs, 1899

Mesogaulus ballensis Riggs, 1899
Mesogaulus paniensis (Matthew, 1902)

Subfamily Mylagaulinae Cope, 1881
Alphagaulus, n. gen.

Alphagaulus vetus (Matthew, 1924)

Alphagaulus pristinus (Douglass, 1903)

Alphagaulus douglassi (McKenna, 1955)

Alphagaulus tedfordi, n. sp.

Umbogaulus, n. gen.

Umbogaulus galushai, n. sp.

Umbogaulus monodon (Cope, 1881)

Mylagaulus Cope, 1878

Mylagaulus sesquipedalis Cope, 1878
Mylagaulus kinseyi Webb, 1966
Mylagaulus elassos Baskin, 1980
Ceratogaulus Matthew, 1902

Ceratogaulus rhinocerus Matthew, 1902
Ceratogaulus hatcheri Gidley, 1907
Ceratogaulus minor (Hibbard and Phillis, 1945)

Ceratogaulus anecdotus, n. sp.

Pterogaulus, n. gen.

Pterogaulus laevis (Matthew, 1924)

Pterogaulus cambridgensis (Korth, 2000)

Pterogaulus sp.

Pterogaulus barbarellae, n. sp.

INDEX TO VOLUME 69

CONTENTS

ARTICLES

New record of Ctenodus (Osteichthyes: Dipnoi) from the Carboniferous of Montana .......

A. Kemp and R. Lund 1

Snake fauna associated with the “Earliest Recent” mammalian fauna in northeastern North

America J. Alan Holman 5

New Atokan productoid brachiopods from the Upper Carboniferous Ladrones Limestone of
southeastern Alaska, with a preliminary note on the phylogeny and classification of the Tribe
Retariini Stanislav S. Lazarev and John L. Carter 1 1

Eocene decapod crustaceans from Pulali Point, Washington

. . Carrie E. Schweitzer, Rodney M. Feldmann, Annette B. Tucker, and Ross E. Berglund 23

Rhysodine beetles (Insecta: Coleoptera: Carabidae): New species, new data. II

Ross T. Bell and Joyce R. Bell 69

The microtine rodents from the Pit locality in Porcupine Cave, Park County, Colorado .....

Christopher J. Bell and Anthony D. Bamosky 93

New Lower Mississippian trilobites from the Chouteau Group of Missouri

David K. Brezinski 135

Reassessment of the North American pelobatid anuran Eopelobates guthriei ..........

. Amy C. Henrici 145

Homology and phylogenetic implications of some enigmatic cranial features in galliform and

anseriform birds ....................... Richard L. Zusi and Bradley C. Livezey 157

A new species of Carpocristes (Mammalia; Primatomorpha) from the Middle Tiffanian of the

Bison Basin, Wyoming, with notes on carpolestid phylogeny K. Christopher Beard 195

The Upper Paleolithic bone industry of Klithi Rock Shelter, northwest Greece .......

. Sandra L. Olsen 209

Review of Miocene (Hemingfordian to Clarendonian) mylagaulid rodents (Mammalia) from

Nebraska William W. Korth 227

NEW TAXA

NEW SUBFAMILIES

tMesogaulinae, new subfamily 232

NEW GENERA AND SPECIES

fAlphagaulus, new genus . 238

"tAlphagaulus tedfordi, new species 243

'tAmeropiltonia, new genus .............................................. 137

281

282

Annals of Carnegie Museum

VOL. 69

'fAmeropiltonia lauradanae, new species .................................... 138

'\CarpiUus occidentalis, new species ....................................... 50

"fCarpocristes rosei, new species .......................................... 196

-\Caruthia, new genus ................................................. 12

^Caruthia borealis, new species .......................................... 13

fCeratogaulus anecdotus, new species ...................................... 260

Clinidium (CUnidium) gilloglyi, new species ................................. 88

Clinidium (sensu stricto) onorei, new species ................................ 85

'fElliptophillipsia rotundus, new species ..................................... 140

Kaveinga (Kaveinga) waai, new species .................................... 70

Omoglymmius (Omoglymmius) emdomani, new species ......................... 75

^Perexigupyge chouteauensis, new species ................................... 141

Plesioglymmius (Juxtaglymmius) negara, new species .......................... 73

'\Portumtes macrospinus, new species 39

'\Pterogaulus, new genus ............................................... 262

'\Pterogaulus barbarellae, new species ...................................... 266

-fPulaiius, new genus 41

tPulalius dunhamorum, new species 44

Rhyzodiastes (Temoana) mindoro, new species ............................... 79

Rhyzodiastes (Temoana) riedeli, new species ................................. 78

^Richterella hessleri, new species ......................................... 143

"fRugivestis girtyi, new species ........................................... 15

■fUmbogaulus, new genus ............................................... 246

fUmbogaulus galushai, new species 247

t Fossil taxa

AUTHOR INDEX

Barnosky, Anthony D. ................................................. 93

Beard, K. Christopher .................................................. 195

Bell, Christopher J. ................................................... . 93

Bell, Joyce R. ....................................................... 69

Bell, Ross T. ........................................................ 69

Berglund, Ross E. .................................................... 23

Brezinski, David K. ................................................... 135

Carter, John L. ....................................................... 11

Feldmann, Rodney M. ................................................. 23

Henrici, Amy C. ..................................................... . 145

JJolman, J. Alan ...................................................... 5

Kemp, Ann ......................................................... 1

Korth, William W. .................................................... 227

2000

Index to Volume 69

283

Lazarev, Stanislav S. .................................................. . 11

Livezey, Bradley C 157

Lund, Richard A. 1

Olsen, Sandra L. ..................................................... 209

Schweitzer, Carrie E 23

Tucker, Annette B. . 23

Zusi, Richard L. ..................................................... . 157

!•!

f i ^

A.-

pfTMm

•rjA'dir#

*5f

“7

IP.^

;. '. -. M’- -

ti _ •.

• -jS *, ',

\. '■

■ ‘j.-^-i'f. , ■*"■ . ■:^ r^^sDg2 ' j

■ .. ^ ^ • .fc ' . • '4iEaHF

M->V3a> i,^.* » -•• . ►’3jK^':^

*■^^•11^' ‘-‘■■M _. V. . trj t . ♦ *

■V* V ^

/ ,. . . .:y

'.0 §hffc i .nvunVMi;^

' • - ; - - , . ^J.

"/ ■• iJvV-rq

INSTRUCTIONS FOR AUTHORS

ANNALS OF CARNEGIE MUSEUM consists of
contributions to the earth sciences (including paleontol-
ogy), life sciences, and anthropology, in 30 by 46 pica
format (127 by 195 mm or 5 by 7% inches). Submit all
manuscripts to the Office of Scientific Publications. Au-
thors should give particular attention to scientific con-
tent, format, and general style for the ANNALS. Manu-
scripts that do not conform to the style of the AN-
NALS will be returned to the author immediately.
Every manuscript will be subjected to external peer re-
view. Authors should submit the names, addresses,
phone and fax numbers, and e-mail addresses of at least
four qualified potential reviewers for each manuscript.
Authors will be asked to subsidize, if funds are available,
any or all costs of publication (approximately $ 100/page
printed). A detailed set of instructions to authors may be
requested from the Office of Scientific Publications.

Manuscript Style. — Articles should include the fol-
lowing items in this order: title page, abstract, text (with
desired headings), acknowledgments, literature cited, ta-
bles, figure captions, and copies of illustrations. All
manuscripts must be typed double-spaced on standard
%V2- by 1 1-inch white bond paper, with at least one-inch
margins all around, and submitted in quadruplicate — ^an
original for the editors, and three review copies. All
pages should be numbered, including tables, literature
cited, and the list of figure captions. Only correspon-
dence quality or better dot matrix printouts will be ac-
cepted; draft mode is unacceptable. A monospaced font,
(e.g.. Courier) is required. A disk in WordPerfect x.x or
ASCII is helpful.

Title Page. The title should be brief, include the ani-
mal or plant group involved, be placed on a separate
page, and appear two to three inches below the top mar-
gin. Include the author’s name(s) and the affiliations of
non-CMNH authors. In the case of multiple authorship,
indicate the address to which proofs should be sent.

Abstracts. Abstracts should be short, but substantive,
and included at the head of the first page of text.

Text. Do not right justify text or break (hyphenate) a
word at the end of a line. Footnotes and acknowledg-
ments as footnotes are unacceptable. All text is double-
spaced in a monospaced font (e.g., Courier).

Tables and Figure Legends. The list of figure legends
and each table should be typed consecutively on indi-
vidual pages separate from the text. Tables must be dou-
ble-spaced throughout with no vertical lines.

Measurements. Metric units should be used, if possi-
ble.

Literature Cited. All references to literature in text
must appear in the Literature Cited section. The data
(author, date, and page) in both citations must agree. Do
not abbreviate the titles of periodicals or serials. The
following style, double-spaced, should be used in Lit-
erature Cited:

1) Two authors in a journal series:

Soltis, D. E., and P. S. Soltis. 1992. The distri-
bution of selling rates in homosporous ferns.
American Journal of Botany, 79:97-100.

2) Same authors repeated^ — ^use three-em dash:

. 1923. The fauna of the Ardyn Obo For-
mation. American Museum of Natural History
Novitates, 98:1-5.

3) Same authors plus a third author — repeat all authors:

Knutson, L. V, R. E. Orth, and W. L. Murphy.
1986. Catalog of Sciomyzidae (Diptera) of
America north of Mexico. Entomography, 4:
1-53.

4) Chapter in an edited volume:

Rausch, R. L. 1963. A review of the distribution
of Holarctic mammals. Pp. 29-43, in Pacific
Basin Biogeography (J. L. Gressitt, ed.).
Bishop Museum Press, Honolulu, Hawaii.

5) Unpublished dissertation:

Smith, J. P. 1976. Review of Eocene Mammals.
Unpublished Ph.D. Dissert., University of
California, Berkeley, California.

6) Book:

White, M. J. D. 1961. The Chromosomes. Meth-
euen and Co., Ltd., London, United Kingdom.

7) Journal articles with usual volume and issue number:

Anderson, W. 1 . 1969. Lower Mississippian con-
odonts from northern Iowa. Journal of Pale-
ontology, 43 : 9 1 6-92 8 .

Figures. Including all illustrative materials (line art,
halftones, photographs), figures are to be numbered in
Arabic numerals. Four sets of figures are required, one
(original artwork) for reproduction, three for reviewers.
Photocopies of photographs for reviewers are usually not
acceptable but are adequate for line drawing review cop-
ies. Figures may not be larger than 17 by 12 inches.
Reducing figures is the responsibility of the author. All
figures must be reducible to a maximum of 127 by 195
mm (30 by 46 picas) without loss of clarity. Line copy
should be designed for reduction to % or Vi or actual size.
Typewritten figure copy will not be accepted. Pho-
tographic figures should be submitted at actual repro-
duction size, if possible.

Rectangular halftone figures should be abutted, with-
out intervening spaces. The printer will insert narrow
white spaces during the reproduction process. All figures
must have minimally one-inch borders all around. Each
figure should be given a protective cover and identified
on the back side.

Lettering and/or a magnification scale (linear metric
scale) for rectangular halftone figures should be placed
directly on the photo, not in a blank space between pho-
tos. The scale or lettering for closely cropped photos can
be placed in blank areas close to the figure.

Proof. — The author should answer all queried proof
marks and check the entire proof copy. Return corrected
page proof with the edited manuscript promptly to the
Office of Scientific Publications.

If an author chooses to make extensive alterations to
a paper in proof stage, the author will bear the cost.
Original manuscripts will not be returned unless request-
ed. Figures will be returned to the author only if re-
quested prior to publication.

MECKMAAr

N JJ £

Bound-To-Please* ^ ^

may 01