, /3 // p

IG

number 2

november 1979

EDITORIAL STAFF

John E. Cooper, Editor
Alexa C. Williams, Managing Editor
John B. Funderburg, Editor-in-chief

Alvin L. Braswell, Curator of
Lower Vertebrates, N.C.

State Museum

John C. Clamp, Associate Curator
( Invertebrates ), N.C.

State Museum

Martha R. Cooper, Associate
Curator ( Crustaceans ), N.C.

State Museum

James W. Hardin, Department
of Botany, N.C. State
University

Board

David S. Lee, Chief Curator
of Birds and Mammals, N.C.

State Museum

William M. Palmer, Chief Curator
of Lower Vertebrates, N.C.

State Museum

Thomas L. Quay, Department
°f Zo°l°gy, N.C. State
University

Rowland M. Shelley, Chief
Curator of Invertebrates, N.C.

State Museum

Brimleyana, the Journal of the North Carolina State Museum of Natural His-
tory, will appear at irregular intervals in consecutively numbered issues. Con-
tents will emphasize zoology of the southeastern United States, especially North
Carolina and adjacent areas. Geographic coverage will be limited to Alabama,
Delaware, Florida, Georgia, Kentucky, Louisiana, Maryland, Mississippi, North
Carolina, South Carolina, Tennessee, Virginia, and West Virginia.

Subject matter will focus on taxonomy and systematics, ecology, zoo-
geography, evolution, and behavior. Subdiscipline areas will include general in-
vertebrate zoology, ichthyology, herpetology, ornithology, mammalogy, and
paleontology. Papers will stress the results of original empirical field studies, but
synthesizing reviews and papers of significant historical interest to southeastern
zoology will be included.

Suitability of manuscripts will be determined by the Editorial Board, and ap-
propriate specialists will review each paper adjudged suitable. Final ac-
ceptability will be decided by the Editor. Address manuscripts and all cor-
respondence (except that relating to subscriptions and exchange) to Editor,
Brimleyana , N. C. State Museum of Natural History, P. O. Box 27647, Raleigh,
NC 27611.

In citations please use the full name — Brimleyana.

North Carolina State Museum of Natural History

North Carolina Department of Agriculture
James A. Graham, Commissioner

CODN BRIMD 7
ISSN 0193-4406

Cretaceous Dinosaurs of North Carolina

Donald Baird and John R. Horner
Museum of Natural History, Princeton University,

Princeton, New Jersey 08544

ABSTRACT. — Isolated bones of Late Cretaceous dinosaurs from the
Black Creek Formation (Campanian) of Sampson and Bladen counties,
accumulated over the last eleven decades, are described and interpreted
in the light of more nearly complete material known from elsewhere. A
medium-sized carnivorous dinosaur is comparable to the tyrannosaurids
Dryptosaurus and Albertosaurus. One toe bone represents the ostrich-mimic
Ormthomimus. The enigmatic Hypsibema crassicauda , originally based on a
mixture of three suborders, is restricted to tail vertebrae and syn-
onymized with Parrosaurus; it is probably a sauropod of huge size but un-
certain family. Duckbill dinosaur bones are generically indeterminate
but evidently belong to the Hadrosaurinae. “Hadrosaurus tripos ” is not a
Cretaceous dinosaur but a Pliocene whale. Also misidentified as
dinosaurian are remains of the gigantic crocodile Deinosuchus
[Phobosuchus] rugosus, which probably preyed on the amphibious hadro-
saurs.

A revised list of the Phoebus Landing local fauna includes the
sharklike fishes Aster ac anthus , Scapanorhynchus , Squalicorax, Ischyrhiza , and
Brachyrhizodus; the bony fishes Paralbula and Pycnodus; the turtles Triohyx
and Taphrosphys ; the mosasaurid sea-lizards Tylosaurus and Platecarpus;
the crocodiles Deinosuchus and Leidyosuchus; and four dinosaurian genera.
Homonymy of Coelosaurus Leidy, 1865, with Coelosaurus [Owen] 1854
necessitates the transfer of C. antiquus Leidy to Ormthomimus .

INTRODUCTION

Our knowledge of the Cretaceous dinosaurs of North Carolina is, un-
fortunately, derived entirely from isolated and often fragmentary bones
that were brought to light during extremely sporadic episodes of
collecting activity. In the antebellum period an impressive amount of
pioneering work was done by Ebenezer Emmons, state geologist from
1851 to 1863. Although Cretaceous reptile bones (mosasaur, crocodile
and turtle) are illustrated in Emmons (1858), he seems not to have dis-
covered any dinosaur material — or if he did, it was lost to science with
the wartime destruction of the Survey collections (Stuckey 1965).

Thus the first discovery of dinosaurs within the state must be credited
to the versatile and energetic Washington Caruthers Kerr, an alumnus of
Chapel Hill who had done postgraduate work under Agassiz and others
at Harvard. Beginning his service as state geologist in 1864, in the final
desperate months of the war, and persevering through the troubled and

Brimleyana No. 2: 1-28. November 1979

1

2

Donald Baird and John R. Horner

impoverished era of reconstruction, Kerr re-established the Survey and
directed its activities until his death in 1885. The fossil vertebrates col-
lected by Kerr and his assistants were sent to Philadelphia for description
by the leading vertebrate paleontologist of the day, Edward Drinker
Cope, and were deposited for permanent safekeeping in the United States
National Museum. Cope published the Survey’s discoveries as they came
to hand, and subsequently (1875) compiled a synopsis of all the fossil ver-
tebrates that had been found in the state.

A hiatus in paleontological work then intervened, ending only in 1905
when a survey of the coastal plain deposits was undertaken as a
cooperative project of the United States Geological Survey and the North
Carolina Geological and Economic Survey. In the course of this work a
number of Cretaceous reptile bones were recovered at Phoebus Landing
on the Cape Fear River by Lloyd W. Stephenson and his associate,
Edward W. Berry (the elder). This material was to have been described
in Part II of the State Survey’s Report (Volume V), '‘The Cretaceous For-
mations of North Carolina,” but unfortunately the series was discon-
tinued after publication of Part I (Stephenson 1923). Aside from the iden-
tifications by Charles W. Gilmore published in Stephenson’s 1912 report
and the subsequent citation by Lull and Wright (1942) of specimens
believed to be hadrosaurian, Stephenson’s material remained un-
described in the National Museum.

After another lengthy hiatus, in the 1950s geologists from the Univer-
sity of North Carolina at Chapel Hill undertook systematic survey work
along the Cape Fear River. Their published reports (e.g. Brett and
Wheeler 1961; Heron and Wheeler 1964) added greatly to our un-
derstanding of Cretaceous sedimentology and stratigraphy. One
dinosaurian byproduct of this research was the hadrosaur humerus
(described below) recovered by Everett Brett and Walter H. Wheeler at
Milepost 49. Wheeler (1966) reported on a mandible of the mosasaurid
lizard Tylosaurus that he collected at Phoebus Landing in 1964.

A specific search for vertebrate fossils was begun in 1963 by Halsey W.
Miller, Jr., then on the faculty of High Point College. Through his ex-
cavations of 1964-1965 at Phoebus Landing the number of known
specimens was greatly increased, making it possible to prepare a
preliminary faunal list (Miller 1967, 1968). All specimens recovered were
presented to the Academy of Natural Sciences of Philadelphia and the
Sternberg Museum of Fort Hays Kansas State College.

The purpose of this paper is to gather together the piecemeal finds of
dinosaur bones from North Carolina and interpret them in the light of
more nearly complete material from elsewhere. As the foregoing historical
summary implies, the Tarheel state has Cretaceous vertebrates in store

Cretaceous Dinosaurs

3

for those who seek them, although hitherto the seekers have been few and
long between. We hope our contribution will stimulate further delving
into this fallow field.

ABBREVIATIONS

AMNH, American Museum of Natural History.

ANSP, Academy of Natural Sciences of Philadelphia.

PU, Princeton University Museum of Natural History.

SM, Sternberg Museum, Fort Hays Kansas State College.

UNC, University of North Carolina at Chapel Hill.

USNM, National Museum of Natural History, Smithsonian Institution.

LOCALITIES

Figure 1 shows the sites from which dinosaurian (or supposedly
dinosaurian) bones have been recovered.

(1) Marl pits of James King, Sampson County. According to Kerr
(1875: 198-199) King’s farm lay “some 10 miles from the depot [Faison’s
Depot], southwest .... This locality is on the waters of Six Runs Creek.”
The heterogeneous collection of bones upon which Cope based Hypsibema
crassicauda was found in the “blue marl,” now the Black Creek Formation.

(2) ,Marl pit of W. J. Thompson, Sampson County, “about ten miles
distant from the marl pit in which the Hypsibema was found” (Cope
1875: 40). This appears to be the locality mentioned in passing by Kerr
(1875: 198) : “ [Blue marl sample] No. 35 is from the farm of J. C. Pass, a
mile and a half west of Faison’s Depot .... Another well known outcrop
in the same neighborhood is at Dr. Thompson’s. ...”

Stephenson’s (1923) map shows an extensive area of Duplin Marl
(Miocene or Pliocene) extending southward from Faison. In the same
region Stuckey’s (1958) map indicates only an insignificant, mile-wide
patch of Duplin (Yorktown) lying west of Six Runs Creek. The dis-
crepancy between these maps is probably less a matter of fact than of
emphasis, the later cartographer choosing to ignore the patchy, surficial
remnants of a once-extensive Cenozoic cover. Under the circumstances it
is easy to see why Cope assumed the type specimen of Hadrosaurus tripos to
be a dinosaur vertebra from the Cretaceous “blue marl,” overlooking the
possibility (now confirmed) that it is a whale vertebra derived from the
Duplin Marl.

(3) Milepost (or Mile Board) 49 near Donahue Landing, south bank of
the Cape Fear River about 15 air miles (24 km) southeast of
Elizabethtown, Bladen County. A hadrosaurine humerus encrusted with
bryozoa and Exogyra spat was found at the base of the Peedee Formation,
evidently reworked from the underlying Black Creek Formation. Detailed

4

Donald Baird and John R. Horner

Fig. 1. Index map of Cretaceous dinosaur localities in North Carolina (for
descriptions see text). Formation boundaries are dotted. Formations in
descending order: Td, Duplin (Yorktown); Tch, Castle Hayne; Kpd, Peedee;
Kbc, Black Creek; Kt, Tuscaloosa. Latitude 35° North passes through Clinton.

Cretaceous Dinosaurs

5

analyses of this locality were provided by Brett and Wheeler (1961 : 67-69,
“Station 2”) and Heron and Wheeler (1964: 46, “Locality 12”).

(4) Phoebus Landing, south bank of the Cape Fear River below
Milepost 68, about 5 air miles (8 km) east-southeast of Elizabethtown,
Bladen County. The source of most of the dinosaur bones described in
this paper, Phoebus Landing was discussed by Stephenson (1912: 120;
1923: 10), Miller (1967), and Heron and Wheeler (1964: 42, “Locality
8”).

THE BLACK CREEK FORMATION

A succinct analysis of this formation was provided by Heron and
Wheeler (1964) in their guide to the Cretaceous strata along the Cape
Fear River. For the convenience of readers who lack ready access to the
guidebook we abstract pertinent passages here.

“The Black Creek Formation consists of poorly indurated, laminated
or thin-bedded clayey silts and silty clays, dark clays, and loose, coarse
light gray, well sorted sand, as well as thick lenses of this sand. These
sands, silts and clays contain substantial quantities of glauconite, lignite,
shell material, iron sulphide, and some amounts of phosphate, amber,
and marine microfossils. . . . Lignitized wood . . . occurs as logs and large
branches which may be partly silicified, or as mats of twigs and other
vegetable matter, as scattered fragments, and, rarely, as very thin layers
in some sands. Many sands are locally cemented and stained by ferric
iron.

“[Donald J. P.] Swift [unpubl. dissert., 1964] has distinguished four
types of stratification in the Black Creek: a fluvial stratification recognized
by nearly perfect separation of sediment into sand and clay, grouping of
strata into strata sets, sand beds with imperfectly developed third order
stratification of thin beds and laminae of sand separated by very thin
laminae of finer sand, and sands with an impoverished transition zone
suite of accessory minerals (glauconite and Foraminifera sparse or lack-

ing)-

“In fluviomarine bedding the sands are finer-grained, darker and more
clayey, the grouping of strata into strata sets is not as pronounced, and
strata sets are relatively continuous whereas the individual strata tend to
lens and bifurcate. The sediments carry a transition zone suite of ac-
cessory minerals, including Foraminifera, Ostrea fragments and up to 15%
of glauconite. Though marine, their stratification is controlled by river
flooding.

“A lagoonal stratification ranges from a lenticular thinly bedded type to
varieties where bedding is obscured by mottling produced by burrowing
invertebrates.

6

Donald Baird and John R. Horner

“Finally a littoral stratification is found only immediately beneath the
Peedee Formation. Strata are of pale sand with a median thickness of 2.5
feet, varying from 1 to 22 feet. . . . These beds are actually lenses when
seen in large enough outcrops.”

Given such varied circumstances of deposition, the ecologically heter-
ogeneous nature of the Black Creek vertebrate fauna is understandable.
Thus the thanatocoenosis at Phoebus Landing combines the remains of
fully terrestrial dinosaurs (carnosaurs and ornithomimids) with those of
amphibious, presumably paludal and lagoonal forms (hadrosaurs,
sauropods and giant crocodiles) and these in turn with remains of marine
mosasaurs and sharks.

Time relationships of the Black Creek Formation were interpreted by
Brett and Wheeler (1961: 102) as follows. “The stratigraphic situation
along the Cape Fear River indicates an onlap condition during Upper
Cretaceous time with an encroachment of the ocean onto a gently sloping
coastal plain. . . . Upstream and updip from Elizabethtown the Black
Creek Formation was, without doubt, deposited during Taylor time
represented by the Exogyra ponderosa zone (or older near the base);
downstream and downdip from Milepost 60 to at least Milepost 49, both
the Black Creek and Peedee Formations were also deposited during
Taylor time.” Thus the dinosaurs discussed in this report are Tayloran,
i.e. Campanian, late Cretaceous, in age.

SYSTEMATIC DESCRIPTIONS

Order SAURISCHIA
Suborder THEROPODA
Infraorder CARNOSAURIA
Family TYRANNOSAURIDAE
cf. Dryptosaurus Marsh, 1877, or Albertosaurus Osborn, 1905

Figs. 2A, 3A, 4B-B1

Two teeth from Phoebus Landing were assigned the tentative iden-
tification of Gorgosaurus ? by Miller (1967: 233): USNM 7199 (previously
cited as £ atomis ? [sic] by Stephenson 1912: 120) and the fragmentary
ANSP “15331” (actually 15332, two teeth, Fig. 2A). Aside from being
carnosaurian these teeth show no characteristics that would permit
generic identification. They are as similar to Dryptosaurus aquilunguis
(Cope) as to Gorgosaurus (i.e. Albertosaurus) ; and indeed, Stephenson’s in-
vocation of the Triassic tooth-genus gjatomus is not so far-fetched as it
sounds, for all carnosaurian cheek teeth are essentially alike.

Cretaceous Dinosaurs

7

A maxilla tentatively attributed to Gorgosaurus by Miller (1967) is here
reassigned to the crocodilian genus Deinosuchus and is discussed below in
the section on non-dinosaurian remains.

Three fragmentary femora are demonstrably carnosaurian. Two of
these, from the marl pits of James King in Sampson County, formed part
of the mixed lot of bones upon which Cope established Hypsibema
crassicauda (USNM 7189). The now-missing “left tibia'’ shaft section
(Cope 1875, PI. 6, Figs. 2-2a) appears instead to be part of a carnosaurian
left femur. In morphology it corresponds closely to the left femur of
Dryptosaurus aquilunguis but is slightly larger, about 10 cm rather than 9 cm
in maximum diameter. The badly-eroded fragment that Cope (1875, PI.
6, Figs. 1-la; our Figs. 4B-B1) interpreted as the distal extremity of a right
humerus proves on comparison to be a left femur. A more informative
specimen is ANSP 15330 (Fig. 3A), the distal one-third of a right femur
which Miller collected at Phoebus Landing but omitted from his papers.
Its distal condyles are much abraded and the anteromedial ridge that
bounds the origin of the femorotibialis muscle is broken away. So far as
the damaged state of the specimens permits comparison, ANSP 15330
and Cope’s “right humerus” are so nearly identical in size and
morphological detail that they might have come from the same individual
— although of course they were found many miles apart.

These femora are closely comparable to that of Dryptosaurus aquilunguis ,
the type specimen of which (ANSP 9995, Fig. 3B) came from the
“chocolate greensand bed” or New Egypt Formation of late Maestrich-
tian age near Barnsboro, Gloucester County, New Jersey. As the
photographs show, on the dorsal (anterior) surface the rugose area of
origin of the femorotibialis muscle is bounded proximally by a prominent
low welt that arcs across the face of the shaft and continues down its an-
teromedial edge as a crest to buttress the internal condyle. The break in
slope between shaft surface and muscle insertion is conspicuous in lateral

Fig. 2. Dinosaur teeth from Phoebus Landing, Bladen County. A, tyran-
nosaurid carnosaurian cf. Dryptosaurus or Albertosaurus , ANSP 15332. B,
hadrosaurine maxillary tooth in occlusal view; C, hadrosaurine dentary tooth in
lingual and profile views, ANSP 15333. Scale in mm.

Donald Baird and John R. Horner

view. The external condyle is prolonged proximally into a hemi-conical
swelling that plunges into the muscle scar.

On the ventral (posterior) surface the intercondylar fossa is relatively
broad and nearly flat-bottomed; on its inner side it extends proximally to
end in a popliteal pit (which is more prominently marked in the Phoebus
Landing femur than in USNM 7189 and the type of Dryptosaurus). Above
this pit the buttressing ridge from the inner condyle swings diagonally
across the shaft to merge with the more prominent ridge from the outer
condyle. The condyles themselves are too badly eroded to be informative.

cm.

Fig. 3. Tyrannosaurid femora (distal ends) in dorsal, medial, ventral and
lateral views. A, right femur from Phoebus Landing, ANSP 15330. B, left femur of
Dryptosaurus aquilunguis, ANSP 9995 (type), from Barnsboro, New Jersey. Scale in

Cretaceous Dinosaurs

9

Available femora of Albertosaurus libratus (Lambe) from the Oldman
Formation (AMNH 5458 and 5664) are either too crushed or too inac-
cessible (being embedded in panel-mounted skeletons) to be useful for
comparison. A right femur of Albertosaurus sarcophagus Osborn from the
lower Edmonton Formation (AMNH 5218) is closely comparable in both
proportions and morphological detail to -the Phoebus Landing femur and
that of Dryptosaurus. The right femur of Daspletosaurus torosus Russell
(AMNH 5438, paratype) from the Oldman Formation is essentially
similar to all the foregoing, differing mainly in that, on the ventral sur-
face, the ridge that runs proximally from the inner condyle crosses the
shaft at an angle of 32° to the femoral axis, as compared to 20° in the
Phoebus Landing femur and 23° in Dryptosaurus and Albertosaurus (angles
approximate). The lower shaft of the Daspletosaurus femur also appears
broader in proportion to its depth, although crushing makes this factor
uncertain.

The right femur of Alectrosaurus olsem Gilmore (AMNH 6554, syntype)
from the Iren Dabasu Formation of Mongolia differs conspicuously from
all the foregoing in morphology as well as in its slenderer proportions. On
the ventral surface it lacks strong ridges extending anteriorly from the
condyles, while on the dorsal surface the scar for the origin of the
femorotibialis muscle is much less sharply demarcated both proximally
and distally, with the result that no pronounced break in slope is evident
in lateral view.

In conclusion, so far as its preservation permits comparison
Dryptosaurus is not distinguishable from Albertosaurus on the basis of the
distal end of the femur. The femur from Phoebus Landing is extremely
similar to both, differing in that its diaphysis is slightly deeper relative to
width, so that the lumen is circular rather than oblate in section. It is less
similar to Daspletosaurus and quite different from Alectrosaurus. Having
made these comparisons, we do not feel that the material available per-
mits a generic identification for the North Carolina tyrannosaurid.

Infraorder COELUROSAURIA
Family ORNITHOMIMIDAE
cf. Ormthomimus Marsh, 1890

A single toe bone of an ostrich-mimic dinosaur from Phoebus Landing
(ANSP 15319) was described by Miller (1967: 232, PI. 3, Fig. 7) who
compared it to the first phalanx of the third digit of Ormthomimus altus
Lambe. This specimen is water-worn, with abraded proximal and distal
ends. The bone is symmetrical when viewed from either end. At its prox-
imal end the ventral surface is smooth and flat while the upper profile

10

Donald Baird and John R. Horner

forms a symmetrical arch. The proximal end bears a pair of shallow con-
cavities for the reception of the distal condyles of a preceding phalanx,
whereas if this were the first phalanx it would be singly concave to receive
the single condyle of the metatarsal. Thus the Phoebus Landing specimen
probably represents the second phalanx of the third digit. It closely
resembles phalanx III-2 of all the ornithomimids, including AMNH 2551
from Monmouth County, New Jersey.

The earliest name for an ornithomimid from the Atlantic coastal
deposits is Coelosaurus antiquus Leidy, 1865, based on a well-preserved tibia
(ANSP 9222) from the greensand of Burlington County, New Jersey.
However, Leidy ’s generic name proves to be a junior homonym of
Coelosaurus [Owen] 1854 (see Appendix). We accordingly transfer the
species antiquus to the oldest junior subjective synonym of Coelosaurus
Leidy, which is Marsh’s genus Ormthomimus. Fragmentary remains in-
distinguishable from Ormthomimus are found from the late Santonian or
early Campanian through the late Maestrichtian of New Jersey and
Delaware (Baird and Horner 1977). As the Phoebus Landing assemblage
is Campanian in age and belongs to the same faunal province as those of
New Jersey and Delaware, we concur with Miller’s provisional assign-
ment of the North Carolina specimen to Ormthomimus .

Suborder SAUROPODOMORPHA
Infraorder SAUROPODA
Family incerta
Hypsibema Cope, 1869

Neosaurus Gilmore in Gilmore and Stewart, 1945; non Nopcsa, 1923.
Parrosaurus Gilmore, 1945 (replacement name).

Hypsibema eras sic auda Cope

Hypsibema crassicauda Cope, 1869: 192; 1870: 122-G — 122-1 ( partim ), PI.
1, Figs. 15-1 5 A- 1 5 B ; 1871: 211-214 ( partim ); 1875: 36-40 {partim ), PI.
5, Figs. 2-2a-2b, PI. 6, Fig. 3 (only). Type species by monotypy.

Figs. 4A-A1, 5, 6

Type. — USNM 7189, a syntypic suite of mixed generic origin. We
designate as lectotype the caudal vertebra illustrated by Cope (our Figs.
4A-A1, 5A), collected by W. C. Kerr from James King’s marl pits in
Sampson County.

Hypodigm. — The lectotype and the following caudal vertebrae from
Phoebus Landing, Bladen County: USNM 6136, Berry and Stephenson
coll., 1907, cited erroneously as cotype by Lull and Wright (1942: 224)

Cretaceous Dinosaurs

11

Fig. 4. Hypsibema crassicauda, syntypes, USNM 7189. A-A^, lectotype caudal
vertebra in left lateral and dorsal views; B-B^, tyrannosaurid left femur (distal
end) in ventral and dorsal views; C-C^, hadrosaurian left tibia fragment; D-D^,
hadrosaurian right metatarsal II in anterior and posterior views. Scale in cm.

but see Miller (1967: 234); Fig. 5B. USNM 7093, 7094, Berry and
Stephenson coll., 1907. ANSP 15307A-C, three vertebrae, Miller coll.
(15307B figured by Miller 1967, PI. 4, Figs. 7-8 as Hadrosaurus ?); Figs. 6A-
A\ B-B1, C. ANSP 15338, Miller coll.; Figs. 6D-D1. USNM 10312, a
caudal neural spine, may pertain.

12

Donald Baird and John R. Horner

Discussion. — This taxon was born in confusion and has persisted in the
same state for more than a century. The bones from James King’s marl
pits (note the plural) that Kerr submitted to Cope were so obviously
water-tumbled and abraded that, even if they had been found in close
proximity, their organic association should have been considered un-
likely. Whether or not they were found together, the five bones that con-
stituted the original syntypic series can now be shown to represent three
suborders of dinosaurs. Cope, unfortunately, took their association for
granted. In deducing a single species from these disparate elements he in-
evitably created a monster incertae sedis for the confusion of subsequent
students.

Of the surviving syntypes illustrated in Fig. 4, A-A1 is the caudal ver-
tebra which inspired Cope’s specific name crassicauda (“thick-tailed”) and
which we designate as lectotype; the other bones are removed from
Hypsibema and described elsewhere in this paper. The fifth bone (now

Fig. 5. Hypsibema crassicauda , caudal vertebrae in posterior and ventral views. A,
USNM 7189, lectotype from James King’s marl pits; B, USNM 6136 from
Phoebus Landing. Scale in cm.

Fig. 6. Hypsibema crassicauda, referred caudal vertebrae from Phoebus Landing in
dorsal, lateral and anterior views. A-C, ANSP 15307A-C; D, ANSP 15338. Scale
in cm.

14

Donald Baird and John R. Horner

missing), which Cope took to be the shaft of the left tibia, appears instead
to be part of a carnosaurian left femur.

Material acquired since 1869 adds little to the morphological informa-
tion provided by the lectotype vertebra. The largest and presumably most
anterior vertebra (ANSP 15338) is slightly longer than its centrum height,
and the more posterior vertebrae become increasingly elongate, exactly as
illustrated by Gilmore in the tail of Parrosaurus. The ends of the centra are
hexagonal, wider than their height, and slightly amphicoelous (nearly
amphiplatyan) with chamfered rims. In well-preserved specimens the up-
per third of the end surface is rugose, typically showing a semi-sunburst
pattern that centers on the neural canal. In lateral view the centra are
marked by low, rounded, longitudinal ridges at mid-height; there are no
pleurocoeles. Ventrally the centrum is excavated by a broad, moderately
deep, boat-shaped sulcus which is bounded laterally by rounded ridges
that connect the anterior and posterior facets for the haemal chevrons.

The neural arch is centered at mid-length on the centrum and is com-
pletely co-ossified with it, so that only a difference in the orientation of
surface striae reveals the position of the neurocentral suture. The anterior
zygapophyses are short, terminating well behind the anterior margin of
the centrum; their articulating facets are inclined about 30° from the ver-
tical. The neural spine (only the base of which is preserved) arises from
the posterior half of the arch at a low angle (22° in the type); its base is
broadly ogival in cross-section. An experimental reconstruction in clay
shows us that the spine must have projected at least half a centrum length
behind its vertebra in order to articulate with the zygapophyses of the
next succeeding vertebra. Of course we cannot say whether the spine
maintained its initial angle or whether it became more bladelike distally.

The smallest vertebra is a topotypic specimen, now missing, illustrated
by Cope (1875, PI. 5, Figs. 2-2a-2b) as possibly belonging to a young in-
dividual of Hadrosaurus tripos. (In Cope’s plate, oddly, it is shown with the
lateral and end views upside down and the dorsal view captioned
“below.”) Its centrum height is given as 20.5 lines, i.e. 43.4 mm; height at
end, 38.1 mm; width of end, 45.1 mm; width of waist, 31.8 mm. In
morphology this centrum falls well within the range of variation of the
Hypsibema caudals from Phoebus Landing. Its small size and lack of co-
ossification with the neural arch indicate immaturity. Centrum dimen-
sions of some larger vertebrae are shown in Table 1.

The genus Parrosaurus. — A sequence of 13 caudal centra (USNM 16735)
recovered by well-diggers on the Chronister farm near Glen Allen,
Bollinger County, Missouri, was described by Gilmore as a new genus
and species, Neosaurus (later renamed Parrosaurus) missounensis. The source

Cretaceous Dinosaurs

15

TABLE 1. Hypsibema crassicauda and H. missounense, measurements of centra in
mm (slightly corrected for abrasion). Data on largest vertebra of H. missounense
(USNM 16735) from Gilmore and Stewert (1945).

Maximum length

Maximum width at posterior end

Height at center of posterior end

ANSP

USNM

ANSP

USNM

15338

7189

15308 A

16735

114

113

93

91

105

100

79

87

82

65

56

91

bed was the McNairy Sand Member of the Ripley Formation, considered
early Maestrichtian in age. In analyzing these vertebrae Gilmore ex-
plicitly noted their similarity to Hypsibema crassicauda ; but accepting
without question the hadrosaurian nature of the latter, he dismissed it
from comparison on the grounds that the vertebrae from Missouri could
not pertain to a member of the Hadrosauridae. Now that Hypsibema has
been freed of hadrosaurian encumbrances its generic identity with
Parrosaurus becomes obvious. Every morphological feature cited for the
Missouri vertebrae can be matched in those from North Carolina. Indeed,
the possibility of specific identity cannot be dismissed: but until better
material is available we prefer to maintain Hypsibema missounense

(Gilmore), n. comb., as a separate species.

\

Affinities of Hypsibema. — A genus known only by its tail bones is
necessarily somewhat difficult to characterize and classify. Regrettably,
the meager sampling of dinosaur bones from North Carolina affords no
additional element that can be attributed to Hypsibema crassicauda. Even
more regrettably, no attempt has apparently been made to recover the
rest of the skeleton of H. missounense which (we suspect) lies buried a few
meters underground on the Chronister farm. At present the caudal ver-
tebrae provide the only basis for comparison.

Hypsibema shares with the Hadrosauridae its slightly amphicoelous
caudal centra, the amidships position of its neural arch, and the shortness
of its anterior zygapophyses. Within that family (as John S. McIntosh has
pointed out to us) Hypsibema bears some similarity to a series of
diminutive caudal vertebrae assigned to Orthomerus transylvanicus by
Nopcsa (1928, PL 6, Fig. 4). But otherwise, as Gilmore justly noted in his
analysis of Parrosaurus , the caudals are distinctly un-hadrosaurian. Con-
spicuously lacking, indeed, is the feature that appears to have been a
basic adaptation of the hadrosaurs: a laterally-compressed tail height-
ened by neural and haemal spines to form an effective propulsive organ
for swimming. The Ceratopsia and Ankylosauria are excluded from com-
parison by the relative shortness of their caudal centra, while the other

16

Donald Baird and John R. Horner

varieties of Late Cretaceous ornithischians are all too small for considera-
tion.

As Gilmore concluded, the sauropod dinosaurs are the only group to
which this genus can be plausibly assigned. Judged ex pede Herculem by the
mass of its tail, Hypsibema must have been a dinosaur of brontosaurian
size, one worthy of a generic name meaning “high stride.” However, the
common sauropods of the Cretaceous — the Titanosauridae — are
characterized by procoelous caudal vertebrae with forward-set neural
arches and prezygapophyses that project beyond the end of the centrum:
this condition is exemplified by Alamosaurus from the Lower North Horn
formation (Maestrichtian) of Utah (Gilmore 1946). Hypsibema must
therefore belong to some other family, one in which the middle and
posterior caudals are amphicoelous to amphiplatyan. Gilmore, after
eliminating various possibilities, was unable to place the genus in any
family but concluded that it might be assigned tentatively to the
Camarasauridae on the basis of the “general make up” of the vertebrae.
Neither extensive comparison nor expert advice has enabled us to im-
prove on Gilmore’s treatment of the problem, so we must leave the family
assignment open until more diagnostic material is found.

Order ORNITHISCHIA
Suborder ORNITHOPODA
Family HADROSAURIDAE
Hadrosaurinae indet.

Figs. 2B-C, 4C-D, 7, 8

As might be expected, bones of duckbill dinosaurs form a large percent-
age of the total sample. Two mandibular fragments (ANSP 15306 and
USNM 7096) from Phoebus Landing were described by Miller (1967:
234). The first of these (Fig. 7) is part of a right lower jaw that includes
the basal portion of the coronoid process externally and the posterior end
of the grooved dentary (without teeth) internally. Also described by
Miller (as an extremely large postzygapophysis) is a right coronoid
process (ANSP 15329, Fig. 7) that evidently originated from a jaw com-
parable in size to ANSP 15306. Although the two specimens cannot be fit-
ted together, their identical coloring and size and the fact that they came
from the same locality suggest that they may pertain to the same indi-
vidual.

Two hadrosaurian teeth (ANSP 15333) from Phoebus Landing are
here described for the first time. A shed maxillary tooth (Fig. 2B) is of no
descriptive use below the family level. The dentary tooth (Fig. 2C) is un-

Cretaceous Dinosaurs

17

Fig. 7. Hadrosaurinae indet., unassociated right dentary fragment and coro-
noid process from Phoebus Landing, ANSP 15306 and 15329, in lingual and
labial views. Scale in cm.

worn and has a complete enameled crown and partial root or fang. The
crown is diamond-shaped, 18.0 mm high and 8.7 mm wide, with a labial-
lingual diameter of 6.6 mm measured at the base; its superior borders are
papillated along a slightly raised edge. The median carina is straight and
relatively low and not accompanied by secondary ridges. The angle be-
tween crown and root is nearly 140°, in agreement with the hadrosaurine
condition as described by Sternberg (1936); on the other hand, the ratio
between width of crown and labial-lingual diameter is characteristic of
the lambeosaurines. The same discrepancy was pointed out by Langston
(1960) in his description of the dentary teeth of Lophorothon atopus. Despite
this curious point of similarity, however, we feel that a positive assign-
ment of the North Carolina tooth to Lophorothon would be unwarranted.

18

Donald Baird and John R. Horner

A partial right scapula (ANSP 15322) from Phoebus Landing was
described by Miller (1967 : 233, PI. 4, Fig. 6). The fragment preserved ap-
pears to come from near the middle of the blade, above the acromial
ridge. Its edges are parallel and do not appear to be diverging dorsally.
According to Brett-Surman (1976) non-expanding scapular blades of this
sort are generally attributable to the Hadrosaurinae.

Two hadrosaurian limb bones, a right second metatarsal lacking its
distal end and a fragment of a small left tibia (“fibula” of Cope 1875, PI.
7, Figs. 1-la), are here removed from the syntypic series of Hypsibema
crassicauda (USNM 7189). The tibia fragment (Figs. 4C-C1) consists of a
badly eroded and water worn mid-shaft that is uncharacteristic below the
family level. Reconstructed, this tibia would have been about 600 mm
long, or about two-thirds the length of the tibia in the type skeleton of
Hadrosaurus foulkii Leidy (ANSP 10005). The metatarsal (Figs. 4D-D1) is
close in size and morphology to the corresponding element in Hadrosaurus
foulkii. A right third metatarsal (USNM 5963, Fig. 8B) from Phoebus
Landing was listed by Gilmore in Stephenson (1912: 120) as Trachodon
tripos ? (Cope). This bone is 340 mm long and has proportions similar to
those of Kritosaurus. But since generic differences have not been recognized
in the metatarsals of hadrosaurian dinosaurs, no generic identification is
possible at this time.

Fig. 8. Hydrosaurinae indet. A, distal half of right humerus from Milepost 49,
UNC 5735, in anterior and posterior views; B, right metarsal III from Phoebus
Landing, USNM 5963, in anterior and posterior views. Scale in cm.

Cretaceous Dinosaurs

19

Brett and Wheeler (1961 : 69) and Heron and Wheeler (1964: 46) men-
tioned a portion of a dinosaur leg bone collected at Milepost 49 in the
base of the Peedee Formation, having evidently been reworked from the
underlying Black Creek sediments. This bone (UNC 5735, Fig. 8A) is the
distal half of a very large right humerus, measuring fully 415 mm from the
condyles to the base of the deltopectoral ridge. The proportions of the
shaft are similar to those of Hadrosaurus foulkii and Kritosaurus incurvimanus
Parks (1920: 35, Fig. 9). Reconstructed, the humerus would have been
about 830 mm long, exceeded in length only by the lambeosaurine
humerus from Baja California described by Morris (1972). From the
proportions of other lambeosaurines Morris computed that the Mexican
hadrosaur would have been about 16.5 m (54 ft.) long. Although com-
putations of this sort may be possible with lambeosaurines, there is
evidence that such is not the case with hadrosaurines: an undescribed
hadrosaurine (AMNH 5465) from the lower Two Medicine Formation of
Montana has a humerus 765 mm long while its other limb elements are
the same length as those in a skeleton of Kritosaurus notabilis (AMNH
5350) that has a humerus only 490 mm long. Because of this variability in
humerus length within the subfamily Hadrosaurinae, we are not
prepared to estimate the North Carolina dinosaur’s size, beyond saying
that the humerus might have come from an individual exceeding 12 m (40
ft.) in length. The large metatarsal (USNM 5963) described above is
about the same size as that in the type skeleton of Kritosaurus incurvimanus ,
which has a total length of about 8.2 m (27 ft.).

In summary, all the hadrosaur material described above appears to
pertain to the subfamily Hadrosaurinae but none of the bones can be
identified on the generic level.

NON-DINOSAURIAN REMAINS

For the sake of completeness we append here a brief discussion of cer-
tain specimens from North Carolina that have been cited erroneously in
the literature as dinosaurian.

Class MAMMALIA
Order CETACEA
Suborder MYSTICETI
Family BALAENOPTERIDAE
“Hadrosaurus tripos ” Cope, nomen dubium
Hadrosaurus tripos Cope, 1869: 192; 1870: 122-I-J; 1875: 40-41, PI. 5, Figs.

1-la (only).

Material. — Type, USNM 7190, a large caudal vertebra. Cope’s initial
publication cited two caudals from the same locality: one “near the thir-
tieth caudal” with a centrum height of 4.5 in. (USNM 7190) and another,

20

Donald Baird and John R. Horner

“apparently terminal.” Cope’s subsequent and more detailed descrip-
tions ignored the second vertebra but added a third from another locality:
“A second [sic] and much smaller vertebra from the pit that furnished the
remains of Hypsibema crassicauda, belonged to a third individual, and
possibly to this species” (Cope 1870). The “terminal” vertebra is now
missing and the tentatively referred “second” vertebra, also missing, is at-
tributed by us to a juvenile Hypsibema crassicauda. Lull and Wright (1942:
145) cited as cotypes USNM 7190 and 7093, but since the latter specimen
is labeled as having been collected by Berry and Stephenson at Phoebus
Landing in 1907 it cannot be part of the type material; we take it to be a

Fig. 9. Balaenopterid whale, caudal vertebra in lateral, anterior, dorsal and
ventral views; lectotype specimen of “Hadrosaurus tripos, ” USNM 7190. From W.
J. Thompson’s marl pit in Sampson County near Faison, evidently from the
Duplin Marl. Scale in cm.

Cretaceous Dinosaurs

21

badly damaged vertebra of Hypsibema. Two additional caudals from
Phoebus Landing, USNM 7094 and 7095, were assigned to Hadrosaurus
tripos by Lull and Wright; the second is correctly identified (i.e. is ceta-
cean) but USNM 7094 is here transferred to Hypsibema crassicauda. Thus
USNM 7190, the vertebra upon which Cope’s species concept was prin-
cipally based, remains as the lectotype of Hadrosaurus tripos.

Locality and age. — W.J., Thompson’s marl pit in Sampson County, about
10 miles (16 km) from James King’s marl pit; collected by W. C. Kerr.
Cope’s original announcement of fossil reptiles from North Carolina
began with the curious statement that the remains were “of cretaceous
age, which were Intrusive in miocene beds.” As noted above, the actual
stratigraphic relationship is one of residual patches and erosional residue
of Tertiary marls overlying a Cretaceous terrane. Its zoological affinities
indicate that the type specimen of H. tripos (with other misidentified ceta-
cean bones from the “Cretaceous”) must have been derived from the
Duplin Marl. According to the Geological Society of America’s correla-
tion chart (Cooke et al. 1943) the Duplin is a lateral equivalent of the up-
per part of the Yorktown Formation. Its age is there indicated as latest
Miocene, but more recent investigation points to a Pliocene dating (Baum
and Wheeler 1977).

Discussion. — As Cope did, we have compared the type specimen of
Hadrosaurus tripos with the caudal vertebrae of the type species, H. foulku
Leidy\(type, ANSP 10005) — as well as those of numerous other
dinosaurian genera — but we cannot concur with his identification. In its
morphology, and particularly in the presence of zones of coalescence be-
tween the epiphyses and the body of the centrum, the vertebra is clearly
mammalian and evidently cetacean. This reinterpretation was confirmed
independently by Clayton E. Ray and Frank C. Whitmore, Jr. (pers.
comm., 1977), who identified the specimen as a caudal vertebra of a
balaenopterine whale and noted that its size is appropriate to Megaptera
expansa Cope, 1868. The latter species, however, was rejected by Kellogg
(1968: 116-118) on the grounds that Cope’s syntypic series is a mixture of
generically indeterminate vertebrae of two taxa from three localities in
Maryland and Virginia. Surely the nominal species Hadrosaurus tripos has
no greater claim to taxonomic validity than Megaptera expansa ; and while
it must be removed from the genus Hadrosaurus , we feel that there is no
genus of whale to which it can be justifiably transferred. We list it
therefore as “Hadrosaurus tripos ” Cope, nomen dubium.

The large vertebral centrum (SM 13025) from Phoebus Landing ten-
tatively assigned to Protamia by Miller (1968: 470, PI. 1, Figs. 1, 3) was
shown by Boreske (1974: 75, Fig. 26H) to be a cetacean caudal. Like the

22

Donald Baird and John R. Horner

type specimen of “Hadrosaurus tripos” it is evidently an erosional remnant
from the Duplin Marl.

Class REPTILIA
Order CROCODILIA
Suborder EUSUCHIA
Family CROCODYLIDAE
Deinosuchus Holland, 1909
Deinosuchus rugosus (Emmons 1858), n. comb.

The anterior portion of a right maxilla identified as Gorgosaurus ? by
Miller (1967: 232-233, PI. 3, Figs. 8-10; PL 4, Fig. 1; ANSP 15303) is
demonstrably crocodilian on the basis of its circular alveoli with dished
bottoms and the presence of a circular-sectioned, hollow tooth-root in the
most anterior alveolus. Brigaded with it under the same catalog number
are several unassociated and fragmentary bones (right jugal, left angular,
right squamosal?) that appear to represent the same species. The maxilla
is exceptionally deep and its dorsal surface slopes down abruptly to the
anterior sutural surface with which the premaxilla articulated. Huge size,
a deep snout, and a deep dorso-lateral saddle or notch between premax-
illa and maxilla are all characteristic of the crocodilian genus Deinosuchus
Holland, 1909, with which Phobosuchus Nopcsa, 1924, is objectively syn-
onymous (see Colbert and Bird 1954). Reassignment of the maxilla from
Phoebus Landing to Deinosuchus is supported by the evidence of an un-
described partial skeleton from Texas (Wann Langston, Jr., pers. comm.,
1978).

The first evidence of this giant crocodile in North Carolina consisted of
huge teeth found by Emmons (1858: 219-221, Figs. 38-39) in the
“miocene" marl at Elizabethtown, Bladen County. Emmons explicitly
recognized the possibility that his specimens might have been reworked
from older beds, and subsequent finds make it clear that their source
must have been the Cretaceous Black Creek Formation rather than the
Cenozoic marl. He assigned the teeth to Owen’s genus Polyptychodon as a
new species, P. rugosus. Leidy (1865: 17-18, 116; PI. 3, Figs. 22-23)
recognized the Cretaceous age and illustrated (as “undetermined”) two
additional teeth. Cope (1871) transferred Emmons’ species to
Thecachampsa and (1875, PI. 7, Fig. 3) illustrated a characteristic tooth
collected by W. C. Kerr. Hay (1902: 513) reassigned the species as
Crocodylus rugosus (Emmons). Teeth that are virtually indistinguishable
from Emmons' type specimens were collected at Phoebus Landing by
Miller (1967, PI. 2, Figs. 5-6; plus specimens not illustrated, ANSP
15308); and a quantity of undescribed material collected by Stephenson
and Berry is housed in the National Museum of Natural History.

Cretaceous Dinosaurs

23

As the morphology, distribution and taxonomy of Deinosuchus are
currently being studied by Langston, further analysis would be inap-
propriate here. The presence of this gigantic crocodilian in the Black
Creek fauna of North Carolina is significant to the study of dinosaurs in
view of Colbert and Bird’s (1954: 21) cogent suggestion that Deinosuchus
“may very well have hunted and devoured some of the dinosaurs with
which it was contemporaneous.” As Deinosuchus occurs in the hadrosaur-
bearing formations of Wyoming, Montana, Texas, North Carolina,
Delaware and New Jersey (our observation) — and doubtless elsewhere
— we feel justified in proposing that giant crocodiles rather than car-
nosaurian dinosaurs may have been the major predators upon the
amphibious hadrosaurs. In that case the predator/prey ratios computed
by Bakker (1972), who assumed that “tyrannosaurs were the only car-
nivores powerful enough to kill and dismember duck-bills, horned
dinosaurs and ankylosaurs,” may seriously misrepresent the ecological
realities.

Order CHELONIA
Genera indet.

A badly water-worn bone fragment (UNC 3370) from the basal Peedee
Formation at Milepost 49 on the Cape Fear River was illustrated by Brett
and Wheeler (1961, PI. 9, Fig. 5a-b) as a “dinosaur tarsal.” This bone is
the proximal end of the left femur of a large turtle. So far as comparisons
can be made it is similar to, but slightly larger than, the femur of
Taphrosphys sulcatus (Leidy) as illustrated by Gaffney (1975; PU 18707).
On the ventral surface of the inner trochanter it bears a transverse, canoe-
shaped depression lacking in Taphrosphys. A more precise identification of
the turtle represented is beyond our competence. The eroded and ablated
condition of the bone suggests that it has been reworked from the un-
derlying Black Creek Formation.

A small phalangeal bone (ANSP 15327) from Phoebus Landing that
Miller (1967: 232, PI. 3, Figs. 3-4) thought might belong to a small
coelurosaur appears instead to be a proximal pedal phalanx of a turtle,
probably Tnonyx.

THE PHOEBUS LANDING LOCAL FAUNA

Increased knowledge now makes it possible to update the provisional
faunal list prepared by Miller (1967, 1968). The tabulation offered here
should be considered an interim census, subject to emendation as ad-
ditional material is recovered and studied. For identification of fish
remains we are indebted to Gerard R. Case. A checklist of the
dinosaurian specimens is being published elsewhere (Horner 1979).

24

Donald Baird and John R. Horner

Class Chondrichthyes

Asteracanthus sp. [including Edaphodon ? cephalic spine of Miller 1968, PL 1,
Fig. 10] — hybodont shark.

Scapanorhynchus texanus (Roemer) [Carcharias? , Isurus of Miller 1967, PI. 1,
Figs. 1-5] — ghost shark.

Squalicorax pnstodontus (Agassiz) — galeoid shark.

Ischyrhiza mira Leidy [including Hypolophus? rostrum fragments of Miller
1968, PI. 1, Figs. 5-6] — sawfish.

Brachyrhizodus wichitaensis Romer [ Hypolophus ?, Protamia ? of Miller 1968,
PI. 1, Figs. 7-9, 13-14; but see Boreske 1974: 74] — rhinopterid ray.

Class Osteichthyes

Paralbula casei Estes [Egertoma of Miller 1967, PI. 1, Fig. 7] — ladyfish.

Pycnodus phaseolus Hay [ Egertoma ? of Miller 1967, PI. 1, Fig. 8; Gyrodus ?,
Anomoeodus , Bottosaums ? of Miller 1968, PI. 1, Figs, 4, 11-12, 15-16] —
deep-bodied coral-nibbler.

Class Reptilia

Order Chelonia

Trionyx spp. — soft-shelled turtles.

Taphrosphys sp. — side-necked turtle.

Order Squamata, Family Mosasauridae

Tylosaums sp. (Wheeler 1966) — shallow water marine lizard.
Platecarpus sp. — deeper water marine lizard.

Order Crocodilia

Deinosuchus rugosus (Emmons) — giant crocodile.

Leidyosuchus cf. L. formidabilis Erickson — ordinary sized crocodile.
Order Saurischia, Suborder Theropoda

cf. Dryptosaurus or Albertosaums sp. — medium sized carnivorous
dinosaur.

Ornithomimus (?) sp. — ostrich-mimic dinosaur.

Order Saurischia, Suborder Sauropodomorpha

Hypsibema crassicauda Cope — huge herbivorous dinosaur.

Order Ornithischia, Suborder Ornithopoda

Hadrosaurinae indet. — flat-headed duckbill dinosaurs.

FAUNAL AFFINITIES

In his 1967 analysis of the Phoebus Landing assemblage Miller was im-
pressed by its affinity with those of the Belly River Group (i.e. the
Oldman and Judith River Formations) of western Canada and the moun-
tain states. So far as the dinosaurs are concerned, the redeterminations

Cretaceous Dinosaurs

25

presented in this paper do not controvert that view as a generalization;
but we must point out that the material at hand does not permit the iden-
tification of Carolinian with Canadian dinosaurs on the generic level. In-
deed, the closest affinities of the North Carolina dinosaurs appear to lie
elsewhere.

As noted in the preceding pages, the sauropod Hypsibema is in-
distinguishable from Parrosaurus, a genus recorded from beds of early
Maestrichtian age in Missouri. The carnosaur from Phoebus Landing is
as similar to Dryptosaurus from the late Maestrichtian (and earlier) of New
Jersey as it is to Albertosaurus. The single toe bone of an ornithomimid
from North Carolina is comparable to ornithomimid bones from the early
Campanian to late Maestrichtian of New Jersey and Delaware, as well as
to Omithomimus from the Belly River beds. The hadrosaurian dentary
tooth from Phoebus Landing has no close counterpart except in
Lophorothon from the early Campanian of Alabama.

Considering the quality of the evidence available, we suggest that firm
conclusions about the affinities of the Black Creek dinosaur assemblage
are unwarranted at present. Caution seems particularly appropriate in
the light of our observation (Baird and Horner 1977) that the Upper
Cretaceous dinosaur faunas of eastern and western North America have
more in common with each other than had previously been recognized.
And for purposes of biostratigraphic correlation, certainly, the evidence of
the invertebrates must greatly outweigh that of the vertebrates.

APPENDIX

Homonymy of Coelosaurus [Owen] 1854 and Coelosaurus Leidy, 1865

The generic name Coelosaurus first appeared on page 15 of a catalogue of
the fossil reptiles and fishes in the museum of the Royal College of Sur-
geons of England, published in 1854. Although no author’s name appears
on the title page or elsewhere, the catalogue is known to be the work of
Richard Owen and is always cited as “Owen (1854).” It is a work of
original research and not a mere clerical compilation.

In that publication the name Coelosaurus is accompanied by a diagnosis
and etymology and is applied explicitly to a single bone, No. 58 in the
collection. No species name is associated with the generic name. Under
the International Code of Zoological Nomenclature (1961) we must ac-
cept the generic name Coelosaurus [Owen] 1854 as validly published: it
can be invalidated by neither its author’s anonymity [Art. 9(7), Recom.
51 A] nor its uninominal status [Art. ll(c)(i)]. It is thus a senior
homonym of Coelosaurus Leidy, 1865. As there is no appeal from the Law
of Homonymy, Leidy ’s name must be rejected.

26

Donald Baird and John R. Horner

So far as we can determine the name Coelosaurus [Owen] does not ap-
pear in the literature either before or after the 1854 publication cited. Its
author evidently abandoned it — not surprisingly, as it was based on a
mutilated vertebral centrum of indeterminate age that had been
redeposited in Pleistocene drift — for he made no allusion to it in his sub-
sequent studies on British fossil reptiles. Coelosaurus [Owen] is not listed
in the standard nomenclators of Neave, Sherborn, and Schulze-
Kukenthal, nor in the indexed bibliographies of vertebrate paleontology
compiled by Hay, Camp, and Gregory. We came across it entirely by ac-
cident. As a nomen oblitum it deserves oblivion in taxonomy although it
remains valid for purposes of homonymy.

ACKNOWLEDGMENTS. — For institutional courtesies and the loan of
specimens we are pleasantly indebted to Ms. L. Gay Vostreys of the
Academy of Natural Sciences of Philadelphia, Dr. Eugene S. Gaffney of
the American Museum of Natural History, Dr. Walter H. Wheeler of the
University of North Carolina at Chapel Hill, and Dr. Nicholas Hotton III
and Mr. Robert W. Purdy of the National Museum of Natural History,
Smithsonian Institution. Advice and criticism have been generously
provided by Messrs. Michael K. Brett-Surman and Gerard R. Case, and
Drs. John E. Cooper, Wann Langston, Jr., John S. McIntosh, Halsey W.
Miller, Jr., Clayton E. Ray, and Frank C. Whitmore, Jr. Our research
was supported by the William Berryman Scott Fund of Princeton
University.

LITERATURE CITED

Baird, Donald, and J. R. Horner. 1977. A fresh look at the dinosaurs of New Jer-
sey and Delaware. Bull. N. J. Acad. Sci. 22(2) : 50.

Bakker, Robert T. 1972. Anatomical and ecological evidence of endothermy in
dinosaurs. Nature 255(5359):81-85.

Baum, Gerald R., and W. H. Wheeler. 1977. Cetaceans from the St. Marys and
Yorktown Formations, Surry County, Virginia. J. Paleontol. 5/(3): 492-504.

Boreske, John R., Jr. 1974. A review of the North American fossil amiid fishes.
Bull. Mus. Comp. Zool. 745(l):l-87.

Brett, C. Everett, and W. H. Wheeler. 1961. A biostratigraphic evaluation of the
Snow Hill Member, Upper Cretaceous of North Carolina. Southeast. Geol
5(2):49-132, pi. 1-9.

Brett-Surman, Michael K. 1976. The appendicular anatomy of hadrosaurian
dinosaurs. Unpubl. MA thesis, Univ. California, Berkeley. 108 pp.

Colbert, Edwin H., and R. T. Bird. 1954. A gigantic crocodile from the Upper
Cretaceous beds of Texas. Am. Mus. Novit. 1688A-22.

Cooke, C. Wythe, J. Gardner and W. P. Woodring. 1943. Correlation of the
Cenozoic formations of the Atlantic and Gulf coastal plain and the Caribbean
region. Geol. Soc. Am. Bull. 54:1713-1723.

Cretaceous Dinosaurs

27

Cope, Edward D. 1869. [Remarks on Eschnchtius polyporus, Hypsibema crassicauda,
Hadrosaurus tripos , and Polydectes biturgidus.] Proc. Acad. Nat. Sci. Phila.
7569:192.

1870. Synopsis of the extinct Batrachia, Reptilia and Aves of North

America. Trans. Am. Philos. Soc. (N.S.) 74: i-viii, 1-252, pi. 1-14.

1871. Observations on the distribution of certain extinct Vertebrata in

North Carolina. Proc. Am. Philos. Soc. 72: 210-216.

1875. Synopsis of the Vertebrata whose remains have been preserved in

the formations of North Carolina. N. C. Geol. Surv. Rep. 7:29-52, pi. 5-8.

Emmons, Ebenezer. 1858. Agriculture of the eastern counties; together with
descriptions of the fossils of the marl beds. N. C. Geol. Surv. Rep., Raleigh, xvi
+ 314 pp.

Gaffney, Eugene S. 1975. A revision of the side-necked turtle Taphrosphys sulcatus
(Leidy) from the Cretaceous of New Jersey. Am. Mus. Novit. 2571:1-24.

Gilmore, Charles W. 1945. Parrosaurus, n. name, replacing Neosaurus Gilmore,
1945. J. Paleontol. 79:540.

1946. Reptilian fauna of the North Horn Formation of central Utah. U. S.

Geol. Surv. Prof. Pap. 210-C: 29-53.

, and D. R. Stewart. 1945. A new sauropod dinosaur from the Upper

Cretaceous of Missouri. J. Paleontol. 79:23-29.

Hay, Oliver P. 1902. Bibliography and catalogue of the fossil Vertebrata of North
America. U. S. Geol. Surv. Bull. 779:1-868.

Heron, S. Duncan, Jr., and W. H. Wheeler. 1964. The Cretaceous formations
along the Cape Fear River, North Carolina. Atl. Coastal Plain Geol. Assoc.,
Field Conference Guidebook 5:1-55.

Horner, John R. 1979. Upper Cretaceous dinosaurs from the Bearpaw Shale
(rparine) of south-central Montana with a checklist of Upper Cretaceous
dinosaur remains from marine sediments in North America. J. Paleontol.
55( 3):566-577.

International Code of Zoological Nomenclature. 1961. Int. Trust Zool. Nomen-
clature, London, xviii + 176 pp.

Kellogg, Remington. 1968. Fossil marine mammals from the Miocene Calvert
Formation of Maryland and Virginia, Part 5: Miocene Calvert mysticetes
described by Cope. U. S. Nat. Mus. Bull. 247: 103-132.

Kerr, Washington C. 1875. Physical geography, resume, economical geology.
N. C. Geol. Surv. Rep. 7:v + 120 pp.

Langston, Wann, Jr. 1960. The vertebrate fauna of the Selma Formation of
Alabama, Part VI. The dinosaurs. Fieldiana Geol. Mem. 5(6): 313-363.

Leidy, Joseph. 1865. Memoir on the extinct reptiles of the Cretaceous formations
of the United States. Smithson. Contrib. Knowl. 7-4(6) : 1 -1 35.

Lull, Richard S., and N. E. Wright. 1942. Hadrosaurian dinosaurs of North
America. Geol. Soc. Am. Spec. Pap. 49:1-242.

Miller, Halsey W. 1967. Cretaceous vertebrates from Phoebus Landing, North
Carolina. Proc. Acad. Nat. Sci. Phila. 779(5): 219-235.

1968. Additions to the Upper Cretaceous vertebrate fauna of Phoebus

Landing, North Carolina. J. Elisha Mitchell Sci. Soc. 54(4):467-471.

Morris, William J. 1972. A giant hadrosaurian dinosaur from Baja California. J.
Paleontol. 46(5) :777-779.

Nopcsa, Franz. 1928. Dinosaurierreste aus Siebenburgen, IV. Die Wirbelsaule
von Rhabdodon und Orthomerus. Palaeontol. Hungarica 7: 273-302.

28

Donald Baird and John R. Horner

[Owen, Richard]. 1854. Descriptive catalogue of the fossil organic remains of
Reptilia and Pisces contained in the museum of the Royal College of Surgeons
of England. Taylor & Francis, London, xix +184 pp.

Parks, William A. 1920. The osteology of the trachodont dinosaur Kritosaurus
incurvimanus . Univ. Toronto Studies, Geol. Ser. 77:1-76.

Stephenson, Lloyd W. 1912. The Cretaceous formations. N. C. Geol. Econ. Surv.
3:73-171 .

1923. The Cretaceous formations of North Carolina, Part 1, Invertebrate

fossils of the Upper Cretaceous formations. N. C. Geol. Econ. Surv. 5. xi + 604

pp.

Sternberg, Charles M. 1936. The systematic position of Trachodon. J. Paleontol.
7d(7):652-655.

Stuckey, Jasper L. 1958. Geologic map of North Carolina. N. C. Div. Mineral
Resour., Raleigh.

1 965. North Carolina: its geology and mineral resources. N. C. Dep. Con-

serv. Develop., Raleigh, xviii + 550 pp.

Wheeler, Walter H. 1966. A mosasaur mandible from the Black Creek Formation,
Cretaceous, of North Carolina. J. Elisha Mitchell Sci. Soc. 82 ( 2):92-93.

Accepted 29 May 1979

Revision of Appalachian Trechus
(Coleoptera: Carabidae)

Thomas C. Barr, Jr.

School of Biological Sciences,

University of Kentucky, Lexington, Kentucky 40506

ABSTRACT. — Appalachian Trechus are arranged in 2 subgenera, 4
species groups, and 28 species of which 6 species are polytypic. A new
key to species and subspecies includes all 40 taxa in the genus known
from Georgia, Kentucky, Maryland, North Carolina, South Carolina,
Tennessee, Virginia, and West Virginia. Thirteen new taxa are
described and illustrated: T. schwarzi scopulosus , new subspecies, North
Carolina; T. schwarzi saludae, new subspecies, North Carolina; T. can-
dy kei pisgahensis, new subspecies, North Carolina; T. haoe, new species,
North Carolina and Tennessee; T. tusquitee, new species, North
Carolina; T. valentinei, new species, North Carolina and Tennessee; T.
stupkai , new species, Tennessee; T. luculentus umcoi , new subspecies,
North Carolina and Tennessee; T. luculentus wayahensis , new subspecies,
North Carolina; T. nantahalae , new species, North Carolina; T. aduncus
toxawayi, new subspecies, North Carolina; T. aduncus coweensis, new sub-
species, North Carolina, and T. aduncus howellae, new subspecies, North
Carolina. New locality records are reported for 8 taxa. Trechus hydropicus
is polytypic, with subspecies T. h. hydropicus (Horn), T. h. beutenmuellen
Jeannel, T. h. avus Barr, and T. h. cams Barr.

INTRODUCTION

Species of Trechus are numerous and locally abundant in the southern
Appalachian mountains (Unaka and Blue Ridge provinces) of western
North Carolina, eastern Tennessee, and adjacent portions of Virginia,
Georgia, and South Carolina. One species, T. cumberlandus Barr, occurs in
the Cumberland plateau of Kentucky and Tennessee; T. tennesseensis Barr
exists in caves in the Appalachian valley of east Tennessee, and T.
hydropicus (Horn) ranges northward into easternmost Kentucky, western
Virginia and Maryland, and eastern West Virginia. Five Appalachian
species were recognized by Jeannel (1931) in his revision of North
American Trechini, and previously I described an additional 18 species
from the area (Barr 1962).

Since 1962 many fresh collections have been made. Study of this
material and comparison with my 1960 collections led to the present
paper. Thirteen new taxa are described, bringing the total for the region
to 40 taxa, which include 28 species, of which 6 are interpreted as
polytypic. In contrast, there are only 12 known native species of Trechus in

Brimleyana No. 2: 29-75. November 1979

29

30

Thomas C. Barr, Jr.

all the rest of North America. New distributional records are given for 8
taxa, yet no less than 12 taxa are still known from single collections at
single sites, suggesting that yet further field investigations and revisions
may be necessary before this large and interesting group of small moun-
tain beetles is properly understood from a purely taxonomic point of view.

There seems no special need to divide the Appalachian species of sub-
genus Trechus (males with first two protarsomeres dentate and setose
beneath) into two species groups, and I have lumped my “ carolinae group”
(Barr 1962:73) with the hydropicus group. The great majority of the species
belong to the endemic subgenus Microtrechus Jeannel, distinguished by
having only the first, instead of the first two protarsomeres dentate and
setose beneath in the male. This subgenus occurs in southwestern North
Carolina and adjacent parts of Tennessee, Georgia, and South Carolina.
Three distinctive species groups recognized by Barr (1962) have been
retained in the present revision, although I have shifted one species (77
verus Barr) from the uncifer group to the nebulosus group. In general, the
species of Trechus reflect the major biogeographic features of carabid dis-
tribution in the southern Appalachians as detailed by Barr (1969). Sub-
genus Trechus has most of its species north of Asheville and north of the
French Broad River valley, while the species of subgenus Microtrechus oc-
cur south of the French Broad, although T. (T.) schwarzi and T. (M.)
vandykei each have subspecies in the Black Mountains and on Pisgah
Ledge, respectively. Various species and subspecies are endemic to most
of the major mountain ranges of the area; lists of endemics by ranges are
presented below, and locations of the ranges are shown in figure 46.

The history of Trechus speciation in the southern Appalachians has
presumably involved vertical expansions and contractions of taxon ranges
in response to the colder, wetter climates of Pleistocene glacial maxima
alternating with warmer, drier regimes of interglacial periods (Barr 1962,
1969). For cold-limited species, suitable microhabitat areas waxed and
waned as did the continental glaciers, with the elevation of the maximum
permissible isotherm rising and falling like the level of the sea, from which
the cool, wet summits of the mountains emerged like islands or
achipelagoes. Because most Trechus species are cold-limited, their specia-
tion patterns are essentially insular, and vicar species and subspecies are
common.

It is thus not surprising that in the southern Appalachians we encoun-
ter arrays of closely similar taxa strung out along the major mountain
chains like beads on a string. For some of the altitudinally limited taxa,
the choice of assignment to the species or subspecies category is arbitrary:
should another glacial maximum ensue, these taxa which are today
allopatric could be brought together again (unlike their trechine cousins

Revision of Trechus

31

extrinsically isolated in inescapable cave systems). Applying Emerson’s
(1945) “practical” definition of a species, one could readily consider the
taxa here treated as polytypic 77 schwarzi, T. vandykei, 77 tennesseensis, and
77 aduncus as 11 or so distinct species because they are probably
genetically (extrinsically) isolated from each other at the present time.
However, intergradation between 77 hydropicus avus and 77 h. beutenmuellen
(Barr 1962:73) provides a morphological yardstick against which these
allopatric taxa can be compared. Polytypic 77 luculentus occurs at such low
elevations that existing gaps ought to pose no problems to occasional gene
flow between component subspecies. In most instances I have employed a
conservative treatment; if I have erred on the conservative side, I will at
least have indicated the close relationship of the taxa involved. For exam-
ple, all of the aduncus- group taxa ( = polytypic 77 aduncus in the present
paper) have, in my judgment, been derived from a single ancestral form
which became isolated relatively recently by the gradual restriction of
suitably cool, moist microhabitats to the summits of various mountain
ranges in the Great Balsams, Cowees, and Nantahalas. Perhaps 77
aduncus aduncus and 77 toxawayi may occasionally interbreed at some inter-
mediate geographic point, but 77 howellae and 77 coweensis are probably
absolute genetic isolates at the present time. My views on the probable
degree of isolation at present are indicated in the taxon accounts, so that
the reader may either accept my basically conservative interpretations or
superimpose equivocally feasible, more liberal interpretations of his own.

The addition of several new taxa has necessitated rewriting a key to
species and subspecies of Appalachian Trechus. All known species and
subspecies from the states of Georgia, Kentucky, Maryland, North
Carolina, South Carolina, Tennessee, Virginia, and West Virginia are in-
cluded. The present key makes greater use of external morphology than
my 1962 key, but determinations should always be checked by examina-
tion of an aedeagus where possible (cf. figs. 1-27 in Barr 1962, and figs.
18-43 in the present paper). The range of most species and subspecies is
quite limited, and the precise provenance of a specimen, if known, can be
used to limit still further the species under consideration.

The accounts of taxa other than those newly described take the form
of an annotated checklist; the minimal information given is full tax-
onomic citation and geographic range. Miscellaneous collecting informa-
tion and data on syntopic and sympatric relationships with other species
of the genus have been added when available.

32

Thomas C. Barr, Jr.

KEY TO SPECIES AND SUBSPECIES OF
SOUTHERN APPALACHIAN TRECHUS

1. Males with two segments of protarsus enlarged

{Trechus, sensu stricto) 2

Males with only the first protarsal segment enlarged

(subgenus Microtrechus ) 1 2

2(1). Pronotum about 0.7-0. 8 times as long as wide, total length

3. 5-5. 0 mm 6

Pronotum (Fig. 1) very transverse, less than 0.7 times as long as wide;

length 2. 5-3. 5 mm ( hydropicus ) 3

3(2). Elytra with 5 well-developed striae and trace of sixth stria, inner intervals
subconvex; aedeagal apex minutely knobbed (Fig. 18); western Virginia,

Maryland, eastern West Virginia hydropicus hydropicus (Horn)

Elytra with 4 or 5 complete striae, more shallowly impressed 4

4(3). Pronotum with hind angles blunt or sharp, elytra with 4 or 5 striae, at least
inner 3 striae moderately impressed; apex of aedeagus knobbed,

or thickened and inflected (Figs. 19, 21) 5

Pronotum with hind angles sharp, obtuse; elytra! striae very shallow,
inner 3 more pronounced than 4th, at best a trace only of 5th stria
present; apex of aedeagus unmodified (Fig. 20); Roan and Bald
mountains of Tennessee-North Carolina border, southeast to Black

and Great Craggy mountains, North Carolina

hydropicus beutenmuellen Jeannel

5(4). Elytra with 5 striae and trace of 6th (5th and 6th striae may both be very
shallow in westernmost part of range); apical recurrent groove long;
apex of aedeagus (Fig. 21) thickened and inflected; Grayson County,

Virginia, to Harlan County, Kentucky hydropicus canus Barr

Elytra with 4 striae and trace of 5th; apical recurrent groove short to
moderate; aedeagus as in Fig. 19; Blue Ridge in Avery and Ashe coun-
ties, North Carolina hydropicus avus Barr

6(2). Pronotum (Fig. 2) with sides subparallel in basal fourth, hind angles large
and nearly right; Roan Mountain, Tennessee, and adjacent North

Carolina roamcus Barr

Pronotum with sides convergent, hind angles small, various (Figs. 3, 4, 5)

7

7(6). Range: mountains of western North Carolina 8

Range: Cumberland plateau from Rockcastle County, Kentucky, to

Grundy County, Tennessee cumherlandus Barr

8(7). Range: Black and Great Craggy mountains and vicinity 9

Range: south and west of Asheville 11

9(8). Smaller, 3. 4-4. 2 mm, eye diameter equal to or greater than scape length;

pronotum sides feebly sinuate before hind angles; elytra with four
complete striae and trace of 5th stria; apical groove ending well in

advance of anterior apical puncture 10

Larger, 4.4-4. 8 mm, eye diameter less than scape length; pronotum sides
not sinuate, hind angles obtuse and sharp; elytra with five shallow,
complete striae and traces of 6th and 7th striae; apical groove short,
ending at level of anterior apical puncture carolinae Schaeffer

Revision of Trechus

33

10(9). Pronotum (Fig. 4) sides feebly sinuate before obtuse hind angles; aedeagus

apex slender, attenuate, tipped with more or less rounded knob

schwarzi scopulosus, new subspecies

Pronotum sides very briefly but distinctly sinuate before very small, right
hind angles; aedeagus apex not attenuate, tipped with knob bearing

sharp ventral cusp mitchellensis Barr

1 1 (8). Larger, 3. 8-4. 4 mm; pronotum (Fig. 3) a little less transverse, sides

shallowly sinuate before hind angles; aedeagus (Fig. 23) 1.31-1.39 mm
long; Mt. Pisgah area, Buncombe, Haywood, and Transylvania

counties, North Carolina schwarzi schwarzi Jeannel

Smaller, 3. 4-3,6 mm; pronotum (Fig. 5) more transverse, sides more deeply
sinuate before hind angles; aedeagus (Fig. 24) 1.08-1.16 mm long; gorge

of North Pacolet River, Polk County, North Carolina

schwarzi saludae, new subspecies

12(1 ). Aedeagus with apex rounded, slightly knobbed, or briefly reflexed; AND/
OR internal sac without large, sharp scales which obscure transfer

apparatus 13

Aedeagus (Figs. 32-35) with apex conspicuously reflexed and hooked;
internal sac armed with many large, sharp scales which obscure
transfer apparatus; smaller species, 2. 7-3. 9 mm .... {uncifer group) 15
13(12). Aedeagus (Figs. 25-31 ) with apex gradually attenuate, rounded or slightly
knobbed at tip; transfer apparatus not obscured by scales, consisting of
two lamellar copulatory sclerites, left small and triangular, right larger
and apically rounded; smaller species 2. 4-3. 3 mm long, aedeagus 0.43-

0.92 mm (vandykei group) 20

Aedeagus (Figs. 36-43) with apex broad, very briefly narrowed and slightly
reflexed at tip; transfer apparatus of heavily sclerotized copulatory
pieces, variable in structure but usually not as described above; medium

to large species, 3.3-5. 5 mm long, aedeagus 0.63-1.27 mm long

( nebulosus group) 27

14(12). Pronotum with sides not sinuate before hind angles, which are always

obtuse 15

Pronotum with sides at least feebly sinuate in basal tenth, hind angles
obtuse or less than obtuse; aedeagus as in Figs. 32-35 .... (aduncus) 17
15(14). Elytra with 3 inner longitudinal striae clearly impressed; aedeagus about

0.9-1 .5 mm long, apex greatly produced and slender 16

Elytra with at least 5 clearly impressed striae, disc rather depressed;
aedeagus about 0.7 mm long, apex briefly attenuate and produced;

Unicoi Mountains, North Carolina-Tennessee talequah Barr

16(15). Length 2. 7-3. 2 mm; aedeagus arcuate, apex abruptly narrowed and
produced, 0.88-1 .03 mm long; Great Smoky and Plott Balsam

mountains, North Carolina-Tennessee uncifer Barr

Length 3.2-3. 5 mm; aedeagus very long (1.47-1.49 mm) and straight, apex
greatly produced and not deflexed; Great Balsam Mountains, North
Carolina satanicus Barr

17(14). Elytra with two clearly impressed inner striae, 3rd and 4th striae (if

present) very shallow; Great Balsam Mountains and associated spurs,

North Carolina 18

Elytra with 5, rarely 4 or 6 striae; Cowee and Nantahala mountains,
North Carolina 19

34

Thomas C. Barr, Jr.

18(17). Aedeagus (Fig. 32) thicker, apex briefly produced; Great Balsam

mountain crest, also Pisgah Ledge aduncus aduncus Barr

Aedeagus (Fig. 33) more slender, apex produced and slightly swollen

before terminal hook; Toxaway Mountain

aduncus toxawayi , new subspecies

19(17). Aedeagus (Fig. 35) with median lobe strongly arcuate at middle, basal
bulb and anterior portion bent at right angles to apical portion;

apex with reflexed hook; Nantahala Mountains, North Carolina

aduncus howellae , new subspecies

Aedeagus (Fig. 34) with straight middle portion of median lobe, only
basal bulb bent at right angle; apex produced and knobbed; Cowee

Mountains, North Carolina aduncus coweensis , new subspecies

20(13). Length 2. 6-3. 2 mm, elytra narrower and with 2 to 4 clearly defined

striae; aedeagus smaller, 0.45-0.82 mm long 21

Length 3. 1-3.3 mm, elytra broadly oval, with at least 5 well-defined
striae; aedeagus 0.85-0.92 mm long; eastern Great Smoky and Plott

Balsam mountains, North Carolina-Tennessee subtilis Barr

21 (20). Larger, length 2. 8-3. 2, mean 3.0 mm; aedeagus 0.64-0.82 mm long
(Figs. 30, 31); elytra usually with 3 inner striae clearly defined .... 22
Smaller, length 2. 5-2. 9, mean 2.7 mm; aedeagus length 0.45-0.72 mm
(Figs. 25-29); elytra usually with either 2 or 4 inner striae clearly

defined 23

22(21). Aedeagus (Fig. 30) more slender, apex gradually attenuate, slightly
knobbed and feebly reflexed; widely distributed in southwestern

North Carolina, northeast Georgia, southeast Tennessee

barberi (Jeannel)

Aedeagus (Fig. 31) thicker, apex broader and reflexed, apical knob a
little larger; known only from western Great Smoky Mountains, North

Carolina-Tennessee tomtru Barr

23(21). Pronotum with hind angles obtuse, sides not sinuate before base (Figs. 9,

10) 24

Pronotum with hind angles right or slightly more than right, sides briefly

but distinctly sinuate before base (Figs. 6, 7, 8) 25

24(23). Hind angles of pronotum rounded (Fig. 10); aedeagus (Fig. 29) 0.58-0.72
mm long, apex gradually attenuate and terminating in small, reflexed

knob; Great Smoky Mountains, North Carolina-Tennessee

bowling i Barr

Hind angles not rounded (Fig. 9); aedeagus (Fig. 28) 0.49-0.58 mm long,
apex produced and bluntly truncate; Tusquitee, Snowbird, and Cheoah

mountains, North Carolina tusquitee , new species

25(23). Elytra with only the inner 2 striae clearly defined, but with traces of 3rd
and 4th striae; pronotum shallowly sinuate before hind angles;

parameres usually with 5 apical setae ( vandykei ) 26

Elytra with inner 4 striae more or less equally impressed; sides of
pronotum briefly but distinctly sinuate before small, more or less right
hind angles; parameres with 4 apical setae; Unicoi Mountains,

North Carolina-Tennessee haoe , new species

26(25). Pronotum sides more strongly rounded, hind angles nearly right (Fig. 6);

aedeagus (Fig. 25) a little larger, 0.46-0.51 mm, less arcuate; Black and
Great Craggy mountains north to Bald Mountains, North Carolina-

Revision of Trechus

35

Tennessee vandykei vandykei (Jeannel)

Pronotum sides more oblique, hind angles a little obtuse (Fig. 7);
aedeagus (Fig. 26) slightly smaller, 0.45-0.49 mm, more arcuate; Great

Balsams, Pisgah Ledge, and Cowee Mountains, North Carolina

vandykei pisgahensis, new subspecies

27(13). Eyes small and subconvex to flat, their short diameter 2/3 to 3/4 length

of scape 28

Eyes large and convex, their short diameter subequal to scape length,

or greater than scape length 35

28(27). Sides of pronotum distinctly sinuate before hind angles 29

Sides of pronotum not sinuate, or very briefly and almost imperceptibly

sinuate 31

29(28). Hind angles more or less right, apical recurrent groove long, joining or
directed toward 5th stria far in advance of anterior apical puncture . . .

( tennesseensis ) 30

Hind angles obtuse (Figs. 17, 45); apical recurrent groove joining 5th stria
a short distance in advance of anterior apical puncture; Nantahala

Mountains, North Carolina nantahalae , new species

30(29). Clypeus with pair of oblique striae internal to clypeofrontal grooves; caves

in Roane County, Tennessee tennesseensis tennesseensis Barr

Clypeus without oblique striae internal to clypeofrontal grooves; caves

and lowland forests in Great Smoky Mountains, Tennessee

tennesseensis tauncus Barr

31(28). Apical recurrent groove short, joining 5th stria at level of anterior apical
puncture; pronotum sides very feebly sinuate or not sinuate before hind

angles; hind angles small; elytra with 3 to 5 striae 32

Apical recurrent groove long, joining 5th stria well in advance of anterior
apical puncture; pronotum sides not sinuate, hind angles large, blunt,
obtuse (Fig. 44); elytra usually with 5 striae; spruce-fir forests in central
Great Smoky Mountains, North Carolina-Tennessee . novaculosus Barr
32(31). Elytra with 3 to 6 striae, but at most only 4 striae well impressed;

pronotum as in Figs. 11, 12; apex of aedeagus produced, knobbed, or

hooked; spruce-fir forest in Great Smoky Mountains 33

Elytra with at least 5 well impressed striae; pronotum as in Fig. 16; apex
of aedeagus broad, finely reflexed only at tip; caves in Tuckaleechee

Cove, Blount County, Tennessee tuckaleechee Barr

33(32). Pronotum (Fig. 1 1 ) sides scarcely or not sinuate; usually at least 5 or 6

striae visible, though only inner 2 to 4 are well impressed; aedeagus

0.92-1 .02 mm long 34

Pronotum (Fig. 12) sides finely sinuate immediately before the small,
right, hind angles; inner 2 elytral striae complete but lightly impressed,
3rd stria shallower, 4th evanescent; aedeagus (Fig. 38) 0.77 mm long
(unique holotype), apex produced and reflexed at tip; known only from

Ramsay Cascades area, Great Smoky Mountains, Tennessee

stupkai, new species

34(33). Elytra usually with only inner 2 striae well impressed, although striae 3
through 5 usually discernible as progressively obsolescent traces;
aedeagus (Fig. 37) slender and elongate, apex gradually attenuate and
hooked at tip, parameres slender, with 4 apical setae; eastern Great
Smoky Mountains verus Barr

36

Thomas C. Barr, Jr.

Elytra usually with 4 well impressed striae, stria (4 and) 5 obsolescent
trace; aedeagus (Fig. 36) much thicker, apex briefly produced, twisted
to left, deflexed, feebly knobbed; parameres shorter, broader,
non-styliform; with oblique apexes bearing 3-4 setae; central Great
Smoky Mountains, North Carolina-Tennessee . . valentinei , new species
35(27). Sides of pronotum distinctly sinuate before right or acute hind angles

(Figs. 13, 14, 15) ( luculentus ) 36

Sides of pronotum very feebly and briefly sinuate before obtuse hind

angles 38

36(35). Apical recurrent groove long, joining 5th stria well in advance of anterior
apical puncture; pronotum as in Figs. 14, 15; elytra with inner 3 striae

deeply impressed, inner 3 intervals convex 37

Apical recurrent groove short, joining 5th stria at level of anterior
apical puncture; pronotum as in Fig. 13; elytra with 4 inner striae
moderately impressed, 4 intervals subconvex; central Great Smoky

Mountains, North Carolina-Tennessee luculentus luculentus Barr

37(36). Aedeagus 0.95-1.04 mm long; outer striae feeble but discernible; Unicoi
Mountains, North Carolina-Tennessee . luculentus unicoi, new subspecies
Aedeagus 0.84-0.92 mm long; outer striae obsolete; Wayah Bald area in

Nantahala Mountains, North Carolina

luculentus wayahensis, new subspecies

38(35). Elytra with 3 to 5 moderately impressed striae 39

Elytra with all striae very feebly impressed, inner 2 or 3 striae barely
discernible; Plott Balsam Mountains, North Carolina . balsamensis Barr
39(38). Length 4. 5-5.0 mm; elytra with 4 or 5 striae; Plott Balsams and western

Great Balsam Mountains, North Carolina rosenbergi Barr

Length 3. 3-4.0 mm; elytra with 3 to 4 striae; central Great Smoky
Mountains, North Carolina-Tennessee nebulosus Barr

CLASSIFICATION OF APPALACHIAN TRECHUS

Trechus , sensu stricto
hydropicus group

hydropicus hydropicus (Horn) — western Maryland and Virginia, eastern
West Virginia (widespread taxon)
hydropicus avus Barr — Ashe and Avery counties, North Carolina
hydropicus beutenmuellen Jeannel — Black, Great Craggy, Bald, and Roan
mountains, North Carolina and Tennessee (widespread taxon)
hydropicus cams Barr — Grayson, Washington, and Lee counties, Virginia,
to Harlan and Letcher counties, Kentucky
schwarzi schwarzi Jeannel — Pisgah Ledge, Haywood-Buncombe-Tran-
sylvania counties, North Carolina

schwarzi scopulosus, new subspecies — Black and Great Craggy mountains,
Yancey-Buncombe-McDowell counties, North Carolina
schwarzi saludae , new subspecies — North Pacolet River gorge, Polk
County, North Carolina

Revision of Trechus

37

cumberlandus Barr — Cumberland plateau from Rockcastle County, Ken-
tucky, southwest to Grundy County, Tennessee (widespread taxon)
mitchellensis Barr — Black Mountains, Yancey, Buncombe, and
McDowell counties, North Carolina
carolinae Schaeffer — Black Mountains, Yancey County, North Carolina
roanicus Barr — Roan Mountain, Carter County, Tennessee, and Mitchell
County, North Carolina.

Figs. 1-5. Pronota of Trechus species, hydropicus group: 1. T. hydropicus hydropicus (Horn),
Mountain Lake, Virginia. 2. T. roanicus Barr, Roan Mountain, Tennessee. 3. T. schwarzi
schwarzi Jeannel, Mt. Pisgah, North Carolina. 4. T. schwarzi scopulosus, new subspecies,
Craggy Dome, North Carolina. 5. T. schwarzi saiudae, new subspecies, Melrose, North
Carolina.

Subgenus Microtrechus Jeannel
vandykei group

vandykei vandykei (Jeannel) — Black, Great Craggy, and Bald mountains,
North Carolina and Tennessee (widespread taxon)
vandykei pisgahensis , new subspecies — Great Balsam and Cowee moun-
tains, Buncombe, Haywood, Jackson, and Transylvania counties,
North Carolina

haoe, new species — Unicoi Mountains, Graham County, North Carolina
tusquitee , new species — Tusquitee/Valley River, Snowbird, and Cheoah
mountains, Cherokee, Clay, Graham, Macon, and Swain counties,
North Carolina

38

Thomas C. Barr, Jr.

bowlingi'Barr — Great Smoky Mountains, North Carolina and Tennessee
(abundant and widespread in the Smokies)
barben (Jeannel) — Unaka mountain and Blue Ridge provinces south and
west of Asheville in Georgia, North Carolina, South Carolina, and
Tennessee (widespread taxon)

tomtru Barr — Thunderhead Mountain, Blount County, Great Smoky
Mountains, Tennessee

subtilis Barr — eastern Great Smoky and Plott Balsam mountains,
Haywood and Jackson counties, North Carolina.

Figs. 6-10. Pronota of Trechus species, vandykei group: 6. T. vandykei vandykei (Jeannel),
Balsam Gap, North Carolina. 7. T. vandykei pis gahensis, new subspecies, Mt. Pisgah, North
Carolina. 8. T. haoe, new species, Haoe Lead, North Carolina. 9. T. tusquitee , new species,
Tusquitee Bald, North Carolina. 10. T. bowlingi Barr, Mt. Sterling, North Carolina.

uncifer group

uncxfer Barr — central Great Smoky Mountains east to Plott Balsam
Mountains, Sevier and Cocke counties, Tennessee, and Haywood and
Jackson counties, North Carolina

satamcus Barr — Great Balsam Mountains, Haywood County, North
Carolina

Revision of Trechus

39

aduncus aduncus Barr — Great Balsam Mountains, Haywood, Buncombe,
Jackson, and Transylvania counties, North Carolina
aduncus toxawayi , new subspecies — Toxaway Mountain, Jackson County,
North Carolina

aduncus coweensis, new subspecies — Cowee Mountains, Macon County,
North Carolina

Figs. 11-17. Pronota of Trechus species, nebulosus group: 11. T. valentinei, new species, Mt.
Kephart, North Carolina-Tennessee. 12. T. stupkai, new species, Ramsay Prong, Tennessee.
13. T. luculentus luculentus Barr, Clingmans Dome, North Carolina-Tennessee. 14. T.
luculentus unicoi, new subspecies, Haw Knob, Tennessee. 15. T. luculentus wayahensis, new sub-
species, Dirty John Creek, North Carolina. 16. T. tuckaleechee Barr, Tuckaleechee Caverns,
Tennessee. 17. T. nantahalae, new species, Burningtown Bald, North Carolina.

40

Thomas C. Barr, Jr.

aduncus howellae, new subspecies — Nantahala Mountains, Macon
County, North Carolina

talequah Barr — Unicoi Mountains, Graham County, North Carolina,
and Monroe County, Tennessee

nebulosus group

nebulosus Barr — central and eastern Great Smoky Mountains, North
Carolina and Tennessee

balsamensis Barr — Plott Balsam Mountains, Haywood and Jackson coun-
ties, North Carolina

valentinei , new species — central Great Smoky Mountains, North Carolina
and Tennessee

verus Barr — eastern Great Smoky Mountains, North Carolina and Ten-
nessee

stupkai , new species — Ramsay Prong, Sevier County, Tennessee, in
Great Smoky Mountains

novaculosus Barr — central Great Smoky Mountains, North Carolina and
Tennessee

luculentus luculentus Barr — Great Smoky Mountains, North Carolina and
Tennessee (widespread in the Smokies)
luculentus umcoi, new subspecies — Unicoi Mountains, Graham County,
North Carolina, and Monroe County, Tennessee
luculentus wayahensis, new subspecies — Nantahala and Tusquitee/Valley
River mountains, Clay and Macon counties, North Carolina
rosenbergi Barr — Plott Balsam and western Great Balsam mountains,
Haywood and Jackson counties, North Carolina
tuckaleechee Barr — Tuckaleechee Caverns, Great Smoky Mountains,
Blount County, Tennessee

tennesseensis tennesseensis Barr — Berry Cave, Roane County, Tennessee
tennesseensis tauncus Barr — Bull Cave and vicinity, Great Smoky Mon-
tains, Blount County, Tennessee

nantahalae , new species — Burningtown Bald, Nantahala Mountains,
Macon County, North Carolina

LOCAL MOUNTAIN FAUNAS OF TRECHUS SPECIES

These lists of guilds of associated species reflect changes in the names of
some of the species, as well as additions to the known fauna since publica-
tion of my 1962 paper on Appalachian Trechus. The Cowee and Nan-
tahala mountains must be added to the list of mountain ranges with en-
demic species or subspecies. Toxaway Mountain, North Carolina, har-
bors one endemic subspecies, T. aduncus toxawayi , along with the common

Revision of Trechus

41

T. barberi. An asterisk (*) indicates strict endemism; a dagger (f)

dicates the taxon is endemic to two adjacent ranges.

A) Roan Mountain, North Carolina-Tennessee

hydropicus beutenmuellen
*roanicus

B) Black and Great Craggy mountains, North Carolina

hydropicus beutenmuellen *carolinae

*schwarzi scopulosus vandykei vandykei

*mitchellensis

C) Great Balsam Mountains and Pisgah Ledge, North Carolina

schwarzi schwarzi *aduncus aduncus

\ vandykei pisgahensis *satanicus

barberi \rosenbergi

D) Cowee Mountains, North Carolina
f vandykei pisgahensis

barberi

*aduncus coweensis

E) Plott Balsam Mountains, North Carolina

barberi *balsamensis

\subtihs \rosenbergi

\uncifer

F) Great Smoky Mountains, North Carolina and Tennessee
*bowlingi

barberi
*tonitru

^ verus

*stupkai

*novaculosus

\subtilis

\uncifer

*nebulosus

*valentinei

*luculentus luculentus
*tuckaleechee (in cave)
*tennesseensis tauricus (in cave
entrance)

G) Nantahala and adjacent mountains, North Carolina

*tusquitee * luculentus wayahensis

barberi *nantahalae

*aduncus howellae

H) Unicoi Mountains, North Carolina and Tennessee
*haoe

barberi
*talequah
* luculentus umcoi

in-

42

Thomas C. Barr, Jr.

SYSTEMATIC ACCOUNTS
Trechus Clairville
Subgenus Trechus , sensu stricto
hydropicus group

Trechus (Trechus) hydropicus hydropicus Horn, new status

Figs. 1, 18

Trechus hydropicus Horn 1883:273. Type locality, “Virginia”; type depos-
ited in Academy of Natural Sciences, Philadelphia. Jeannel 1931:435.
Barr 1962: 71.

Few specimens of typical T. hydropicus were available prior to 1960,
most of them labeled “Va.”, “W.Va.”, and “Md.” I suggested (Barr
1962:72) that T. hydropicus and T. beutenmuellen Jeannel might eventually
prove to belong to a widespread polytypic species, but did not then have
sufficient fresh material to justify the recombination. The type of T.
hydropicus bears the label, “Va., Ulke”. I propose restriction of the type
locality to Bald Knob, Mountain Lake, Giles County, Virginia; T.
hydropicus is moderately abundant at that locality, and it is within the area
known to have been visited by Ulke.

In comparison with other subspecies, T. h. hydropicus has an aedeagus of
medium length (0.93-0.99 m), the apex of the median lobe straight and

0 0.5mm

» -*

Figs. 18-21. Aedeagi of Trechus species, left lateral view: 18. T. hydropicus hydropicus (Horn),
Mountain Lake, Virginia. 19. T. hydropicus avus Barr, Grandfather Mountain, North
Carolina. 20. T. hydropicus beutenmuellen Jeannel, Mt. Mitchell, North Carolina. 21. T.
hydropicus canus Barr, White Top Mountain, Virginia.

Revision of Trechus

43

minutely knobbed in left lateral view. The pronotum is two-thirds as long
as wide; the elytral disc is slightly flattened in the center, and its inner
longitudinal striae are more deeply impressed.

New locality records for T. h. hydropicus are as follows; unlike the very
general locations given on museum labels of 19th century collections,
these localities suggest- a distribution encompassing the Blue Ridge and
western mountains of Virginia, eastern West Virginia, possibly north into
western Maryland (“Md.” labels must apply to specimens from the
higher elevations in Maryland). VIRGINIA: Giles County. — Bald Knob
at Mountain Lake, elevation 4350 ft. (1325 m) (T. C. Barr — restricted
type locality); Cascades of Little Stony Creek, 2 mi. (3.2 km) nw of above
locality, elevation 3000 ft. (900 m) (S. B. Peck). Patrick County. — Pin-
nacles of Dan, along Blue Ridge Parkway near Vesta (R. Hoffman and L.
Knight). Rockbridge County. — Apple Orchard Mountain (R. Hoffman).
Tazewell County. — Burkes Garden (R. Hoffman). WEST VIRGINIA;
Pocahontas County. — Ravine n of Route 39, 4 mi. (6.4 km) nw Mill Point,
elevation 3500 ft. (1100 m) (T. C. Barr); Falls of Hills Creek, 4 mi. (6.4
km) nw Lobelia, elevation 3300 ft. (1000 m) (S. B. Peck).

Trechus ( Trechus ) hydropicus avus Barr, new combination

Fig. 19

Trechus beutenmullen avus Barr 1962: 72. Type locality, Grandfather Moun-
tain, Avery County, North Carolina; type deposited in United States
National Museum of Natural History (USNM).

This subspecies is known only from Grandfather Mountain and Three
Top Mountain (Ashe County), North Carolina. It differs from T.
hydropicus hydropicus in the slightly longer aedeagus (0.99-1.02), the apex of
which is more conspicuously knobbed and reflexed. The pronotum is only
0.6 as long as wide, and the elytral striae are more feebly impressed.
Trechus h. avus intergrades with T. h. beutenmuellen on Beech Mountain,
Avery County, North Carolina (see Barr 1962:73).

Trechus ( Trechus ) hydropicus beutenmuellen Jeannel, new combination

Fig. 20

Trechus Beutenmullen Jeannel 1931: 436. Type locality, Mt. Mitchell, Yan-
cey County, North Carolina; type deposited in Museum National
d’Histoire Naturelle, Paris (MNP) (not seen).

Trechus hydropicus: Schaeffer 1901 : 212. Casey 1918:410. Jeannel 1927:191.
Trechus beutenmullen beutenmullen'. Barr 1962:72.

This subspecies is abundant at higher elevations from the Black and
Great Craggy mountains (Buncombe, Yancey, McDowell counties) of

44

Thomas C. Barr, Jr.

North Carolina northward to the Bald Mountains along the Tennessee-
North Carolina border. The northernmost limit appears to be Roan
Mountain, Carter County, Tennessee — Mitchell County, North
Carolina. The aedeagus is the smallest among the four subspecies of 77
hydropicus (0.84-0.90 mm long), its apex slender, attenuate, slightly inflec-
ted but not knobbed. The pronotum is only 0.6 as long as wide, the elytral
disc is convex, and the elytral striae are feebly impressed; the elytral
apexes are more tapered than in other subspecies of 77 hydropicus.

Trechus (Trechus) hydropicus cams Barr, new combination

Fig. 21

Trechus beutenmulleri cams Barr 1962:73. Type locality, Whitetop Moun-
tain, Grayson County, Virginia (USNM).

Originally described from Whitetop Mountain in extreme southwest
Virginia, T. h. cams has more recently been collected in Bowling Cave,
0.75 mile (1.2 km) southeast of Ben Hur, Lee County, Virginia, elevation
1700 feet (500 m) (J. R. Holsinger); from the summit of Big Black Moun-
tain, Harlan County, Kentucky, elevation 4000 feet (1200 m) (T. C. Barr
and S. B. Peck); and from the north slope of Pine Mountain, Kingdom
Come State Park, Letcher County, Kentucky, elevation 2000 feet (600 m)
(T. C. Barr). The range of the subspecies thus extends across the Ap-
palachian valley floor from the Unakas to the Allegheny front; it probably
intergrades with T. h. hydropicus farther north, but intermediate popula-
tions have not yet been discovered.

The aedeagus of T. h. cams is long (0.96-1.02 mm), the apex
prominently inflected in left lateral view but not noticeably knobbed at
the tip. Externally T. h. cams is virtually indistinguishable from T. h.
hydropicus , with the pronotum fully two-thirds as long as wide, and the in-
ner elytral striae rather deeply impressed.

Trechus (Trechus) schwarzi schwarzi Jeannel, new status

Figs. 3, 23

Trechus Schwarzi Jeannel 1931:437. Type locality, “Roan High Knob”,
but corrected to Retreat, Haywood County, North Carolina; type
deposited in USNM. Barr 1962:74 (in part).

As previously noted (Barr 1962: 75), the type locality of T. schwarzi is
Retreat, near the present site of Sunburst and Lake Logan, in the Great
Balsam Mountains, Haywood County, North Carolina. Additional
material which I recently collected at Mt. Pisgah, Haywood-Buncombe
counties, and Buck Spring, Transylvania County, enabled me to make a

Revision of Trechus

45

Figs. 22-24. Aedeagi of Trechus species, left lateral view: 22. T. schwarzi scopulosus, new sub-
species, Craggy Dome, North Carolina. 23. T. schwarzi schwarzi Jeannel, Mt. Pisgah, North
Carolina. 24. T. schwarzi saludae, new subspecies, Melrose, North Carolina.

careful comparison of topotypic 77 schwarzi with similar populations in
the Great Craggy and Black mountains north of the French Broad River
valley. The latter proved to be taxonomically distinct and are described
below. The nominate subspecies is thus restricted to populations of
Pisgah Ledge, which is the eastern arm of the Great Balsams; 77 schwarzi
has not been taken at other localities in the Great Balsams.

My earlier description of 77 schwarzi (Barr 1962) was based on a
paratype from Retreat and is essentially correct, but the aedeagal sketch
which accompanied it (Fig. 7) applies to 77 s. scopulosus (described below).
A female Trechus specimen from Tusquitee Bald, Clay-Macon counties,
North Carolina, which I thought “may belong to schwarzi ” (Barr
1962:75), is assigned to T. ( Microtrechus) luculentus wayahensis (described
below), a species and subspecies now known to be abundant in the
Wayah Bald area, scarcely 10 miles (16 km) from Tusquitee Bald.

Nominate schwarzi is 3. 8-4. 4, mean 4.0 mm long. The hind angles are
rather prominent and usually acute because of a relatively deep sinuosity
in the lateral margins of the pronotum. Longitudinal striae of the elytra
are rather shallowly impressed, the third through the sixth progressively
obsolescent, the seventh absent. The aedeagus (1.31-1.39, mean 1.35 mm
long) is weakly arcuate, with large basal bulb, a more or less straight mid-
dle portion, and reflexed, attenuate, and knobbed apex.

46

Thomas C. Barr, Jr.

Trechus (Trechus) schwarzi scopulosus, new subspecies

Figs. 4, 22

Trechus schwarzi : Barr 1962:74, in part. NOT Jeannel 1931:437.

Etymology. — Latin scopulosus , “craggy; ” a reference to the type locality.

Description. — Differs from T. s. schwarzi in smaller size, less prominent
pronotal hind angles, deeper longitudinal striae of elytra; aedeagus
slightly smaller, more arcuate, apical knob more prominent. Length 3.4-
3.9, mean 3.6 mm. Sides of pronotum less deeply incised before hind
angles, which are small, sharp, and usually more or less right. Inner two
or three elytral striae distinctly and moderately impressed, outer striae
progressively obsolescent. Aedeagus 1.25-1.29, mean 1.26 mm long, simi-
lar to that of nominate schwarzi but more arcuate, consequently shorter;
parameres slightly shorter; apical knob much more prominent.

Type series. — Holotype male (American Museum of Natural History —
AMNH) and 18 paratypes, below summit of Craggy Dome, elevation
5600 feet (1700 m), Buncombe County, North Carolina, 22 July 1960, T.
C. Barr and M. C. Bowling.

Measurements (in mm). — Holotype male: total length 3.9, head 0.81
long X 0.84 wide, pronotum 0.71 long X 1.05 wide, elytra 2.33 long X
1.67 wide, antenna 1.92 long.

Distribution. — Known only from the Black and Great Craggy moun-
tains and the adjacent Blue Ridge, in Buncombe, McDowell, and Yancey
counties, North Carolina. Additional specimens, not made paratypes,
were collected on Mt. Mitchell, elevation 6600 feet (2000 m), in the Black
Mountains; and on the Pinnacle, elevation 5200 feet (1600 m), where the
Blacks join the Blue Ridge, near the Buncombe-McDowell-Yancey
county corner.

Trechus (Trechus) schwarzi saludae , new subspecies

Figs. 5, 24

Etymology. — Derived from Saluda, a town near the type locality.

Description. — Resembles T. schwarzi scopulosus in smaller size, less
prominent hind angles of pronotum, deeper longitudinal striae of elytra;
aedeagus smaller than in T. s. schwarzi or T. s. scopulosus , weakly arcuate
and with small apical knob as in nominate schwarzi. Length 3. 4-3. 7, mean
3.5 mm. Pronotum and elytra about as in T. s. scopulosus. Aedeagus 1.08-
1.16 mm long, weakly arcuate, apical knob small, apex rather sharply
deflexed.

Revision of Trechus

47

Type series. — Holotype male (AMNH) and 29 paratypes, one mile (1.6
km) east of Melrose, from small, wet ravines on the south side of the gorge
of the North Pacolet River, elevation 1300 feet (350 m), Polk County,
North Carolina, 13 July 1969, T. C. Barr.

Measurements (in mm). — Holotype male: total length 3.7, head 0.86
long X 0.74 wide, pronotum 0.74 long X 1.02 wide, elytra 2.14 long X
1.60 wide, antenna 1.66 long.

Distribution. — Known only from the type locality.

Discussion. — It is remarkable that this distinctive subspecies of T.
schwarzi should occur at such a low elevation, since both of the other sub-
species are known only at altitudes above 4500 feet (1400 m). The type
locality lies in second-growth forest between U. S. 176 and the Southern
Railway tracks a few miles southeast of Saluda. All specimens in the
series were collected from beneath mosses and liverworts on stones in or
at the edge of small, spring-fed brooks, where the cool temperature and
high humidity apparently create a suitable microenvironment. Trechus s.
saludae occurs farther from the eastern slope of the mountains and at a
lower elevation than any other North Carolina Trechus species yet known.
Melrose is approximately 35 miles (55 km) southeast of Mt. Pisgah, the
easternmost locality at which T. s. schwarzi has been collected.

The three taxa here placed in polytypic T. schwarzi are probably all ex-
trinsically isolated at the present time. However, survival of the T.
schwarzi stock at a comparatively low elevation near Saluda suggests that
other such insular populations may be scattered across the lowland areas
separating the three subspecies, perhaps restricted to cool, wet ravines in
the vicinity of springs. A gradual cooling of the regional climate could
unite these populations again.

Trechus ( Trechus ) cumberlandus Barr

Barr 1962:76, Fig. 9. Type locality, ‘Lish Steele Caves, Wayne County,

Kentucky; type deposited in USNM.

Trechus cumberlandus , closely similar to T. schwarzi but without a distinct
apical knob at the tip of the aedeagus, is a vicar species to T. schwarzi , oc-
curring in the Cumberland plateau of eastern Kentucky and Tennessee.
Elsewhere (Barr 1962, 1969) I interpreted the species as a Wisconsin
relic. The range of T. cumberlandus extends along the western margin of the
Cumberland plateau from Rockcastle County, Kentucky, to Grundy
County, Tennessee. It has been collected in eleven localities, as follows:
KENTUCKY: Rockcastle County. — Pine Hill Cave, at Pine Hill; Sinks of

48

Thomas C. Barr, Jr.

Roundstone Cave, 0.8 mi. (1.3 km) ene Pine Hill. Pulaski County. —
Hydens Cave, 1.3 mi. (2.1 km) ne Blue John. Wayne County. — ‘Lish
Steele Caves, 3 mi. (4.8 km) e Monticello (type locality); Johnson Fork
Cave, 0.4 mi. (0.6 km) e Burfield; Upper Blowing Cave, at Sunnybrook.
TENNESSEE: Overton County. — Falling Springs Cave, 1.25 mi. (2 km)
wsw Hanging Limb. Putnam County. — Sinkhole in Calfkiller Valley, 0.7
mi. (1.1 km) se Bee Rock (near Monterey). Cumberland County. — Jewett
Cave, 0.5 mi. (0.8 km) ese summit Hinch Mountain. Van Buren County. —
Gorge below Fall Creek Falls (in State Park). Grundy County. — At spring
in Savage Gulf, near Beersheba Springs. The last two localities are
epigean. The Kentucky localities are at elevations of 800 to 980 feet (240
to 300 m) and those in Tennessee at 1100 to 2200 feet (335 to 670 m). The
series are relatively small except in the case of Hydens Cave, where about
80 specimens were collected far back in the cave on wet, rotting leaves
beside a deep pool.

Trechus (Trechus) mitchellensis Barr

Barr 1962:75, Fig. 8. Type locality, Celo Mountain, Yancey County,
North Carolina; type deposited in USNM.

Trechus mitchellensis is known from three localities in the Black Moun-
tains (Celo Mountain, Mt. Mitchell, and Big Tom, all in Yancey
County), from the Pinnacle, where the Blacks intersect the Blue Ridge
(McDowell County), and from Balsam Gap, where the Blacks join the
Great Craggy Mountains (Buncombe County). All localities are at eleva-
tions between 4800 and 6500 feet (1460 to 2000 m) northeast of Asheville,
North Carolina. This species is moderately large (3.6-4. 2 mm), closely
similar to T. schwarzi scopulosus , with which it is both sympatric and syn-
topic. From T. s. scopulosus it is distinguished by the sharp ventral cusp on
the apical knob of the aedeagus and by the pronotum sides, which are
very briefly but distinctly sinuate before the small, right, hind angles. The
usual microhabitat is under moss carpets in spruce-fir forest.

Trechus ( Trechus ) carolinae Schaeffer

Schaeffer 1901 :212. Jeannel 1931 :439. Barr 1962:74, Fig. 6. Type locality,
Mt. Mitchell, Yancey County, North Carolina; type deposited in
AMNH.

Trechus carolinae is a very large (4. 5-5.0 mm) species for the Appala-
chian area. It is relatively rare, inhabiting deep spruce and fir needle duff
near the summits of the Black Mountains. On Mt. Mitchell I have taken a
small number of specimens by prying back the duff layer at the bases of

Revision of Trechus

49

low, wet cliffs near the summit. It coexists with four other Trechus species
(T. schwarzi scopulosus, T. mitchellensis, T. hydropicus beutenmuellen, T. van-
dykei vandykei ), but is easily distinguished by the large size. The apparent
rarity is probably real but could possibly reflect the comparatively inac-
cessible microhabitat in which it occurs.

Trechus (Trechus) roamcus Barr

Fig- 2

Barr 1962:73. Type locality, Roan Mountain, Carter County, Tennessee;
type deposited in USNM.

Trechus roanicus is known only from the type locality, where it occurs on
Roan High Bluff, the highest point on Roan Mountain (elevation 6313
feet) (1924 m). It is closely similar to T. hydropicus , but is larger (3. 8-4. 4
mm) and coexists with the smaller and much more abundant T. h.
beutenmuellen. The species occurs under moss carpets on north-facing
cliffs, in contrast to T. h. beutenmuellen , which is found not only in the same
microhabitat but also under wood chips, under moss on logs, and under
stones near seeps.

Subgenus Microtrechus Jeannel
vandykei group

Trechus ( Microtrechus ) vandykei vandykei (Jeannel), new status

Figs. 6, 25

Microtrechus Vandykei Jeannel 1927: 587, Figs. 1280-1285; 1931: 443. Type
locality, Black Mountains, North Carolina; type deposited in MNP
(not seen).

Trechus ( Microtrechus ) vandykei: Barr 1962:77 (in part).

Although geographic variation among the smaller beetles of the vandykei
group was previously noted (Barr 1962: 78), I separated out only the dis-
tinctive Smoky Mountains endemic, T. bowlingi, and incorrectly lumped
the remainder into T. vandykei. Further study of older collections and
much fresh material suggests that populations of the Great Balsams,
Pisgah Ledge, and Cowee Mountains differ (at least) subspecifically from
T. vandykei populations north of the French Broad valley, and that the
populations of (a) the Tusquitee/Valley River, Snowbird, and Cheoah
mountains and (b) the Unicoi Mountains, respectively, are best treated as
distinct species.

From T. bowlingi and T. tusquitee, which have very obtuse hind angles
on the pronotum, T. vandykei (s. lat.) is distinguished by the briefly but

50

Thomas C. Barr, Jr.

distinctly sinuate pronotum sides and the form of the aedeagus. From 77
haoe it is distinguished by having only two instead of four distinct elytral
striae; the aedeagus is less arcuate, and the parameres usually bear five
rather than four apical setae. Nominate vandykei differs from 77 v.
pisgahensis , which occurs south of the French Broad River valley, in the
nearly right hind angles of the pronotum and the slightly longer, slightly
less arcuate aedeagus.

Both subspecies of 77 vandykei are relatively common, but less so than
the somewhat larger 77 hydropicus beutenmuelleri or 77 barben. They occur in
leafmold or under moss carpets. Nominate vandykei is known from several
localities in the Black Mountains, the Great Craggy Mountains, at
Balsam Gap between these two ranges, and from the Pinnacle at the point
where the Blacks join the Blue Ridge (Yancey-Buncombe-McDowell
counties, North Carolina). It also occurs on Camp Creek Bald and Unaka
Mountain, in Greene and Unicoi counties, Tennessee, respectively. In the
Black Mountains it is sympatric with four other species of Trechus: 77
schwarzi scopulosus, 77 mitchellensis, 77 carolinae , and 77 hydropicus
beutenmuelleri.

Trechus ( Micro trechus) vandykei pisgahensis , new subspecies

Figs. 7, 26

Trechus (Microtrechus) vandykei: Barr 1962:77 (in part).

Etymology.— From Mt. Pisgah, the type locality.

Description. — Length 2. 5-2. 8, mean 2.7 mm. Closely similar to T. v.
vandykei , differing in more oblique sides and slightly obtuse hind angles of
pronotum; aedeagus more arcuate and slightly smaller, 0.45-0.49 mm
long.

Type series. — Holotype male (AMNH) and 29 paratypes, Mt. Pisgah,
elevation 5000 feet (1524 m), Haywood-Buncombe counties, North
Carolina, 21 May 1961, T. C. Barr; 15 additional paratypes, same
locality and collector, 26 July 1960, 12 September 1964, and 17 August
1978.

Measurements (in mm). — Holotype male: total length 2.79, head 0.52
long X 0.55 wide, pronotum 0.51 long X 0.71 wide, elytra 1.54 long X
1.12 wide, antenna 1.22.

Distribution. — Occurs in the Great Balsam Mountains, Pisgah Ledge,
and the Cowee Mountains, separated from the range of T. v. vandykei by
the valley of the French Broad River. Additional localities other than Mt.
Pisgah are Shining Rock Road and Graveyard Fields, Haywood County;

Revision of Trechus

51

near Devils Courthouse, Transylvania County; and summit of Whiteside
Mountain, Jackson County, North Carolina.

Discussion. — This subspecies commonly is found beneath carpets of
moss on boulders. All collections have been made at relatively high
altitudes, between 4700 and 5400 feet (1400 and 1600 m). Because of ap-
parent altitudinal restriction there is, in my judgment, little or no gene
flow between T. v. vandykei and T. v. pisgahensis at the present time. Syn-
topic Trechus species include T. schwarzi schwarzi, 77 aduncus aduncus , and
T. barberi.

Trechus ( Microtrechus) haoe , new species
Figs. 8, 27

Trechus (. Microtrechus ) vandykei: Barr 1962:77 (in part).

Etymology.— From Haoe Lead, the type locality.

Diagnosis. — A small species of the vandykei group, with non-sinuate
pronotum sides, small, obtuse hind angles, four elytral striae, and a short,
evenly arcuate aedeagus the apex of which is neither produced, knobbed,
nor truncate.

Figs. 25-28. Aedeagi of Trechus species, left lateral view: 25. T. vandykei vandykei (Jeannel),
Balsam Gap, North Carolina. 26. T. vandykei pisgahensis, new subspecies, Mt. Pisgah, North
Carolina. 27. T. haoe , new species, Haoe Lead, North Carolina. 28. T. tusquitee , new species,
Tusquitee Bald, North Carolina.

52

Thomas C. Barr, Jr.

Description. — Length 2. 6-2. 9, mean 2.7 mm. Form small, robust, sub-
convex; color piceous, shining, all appendages pale, contrasting. Head
feebly transverse; labrum evenly emarginate; eye diameter 0.12-0.14 mm,
a little less than or subequal to scape length. Pronotum about 0.7 as long
as wide, sides arcuate apical two-thirds, then convergent and not sinuate
before small, obtuse, hind angles; greatest width in apical fourth. Elytra
1.4 times longer than wide, oval, subconvex, four longitudinal striae pres-
ent, outer striae obsolescent, apical groove wide and rather short, ter-
minating a short distance anterior to anterior apical puncture. Aedeagus
0.45-0.48, mean 0.46 mm long, strongly arcuate and rather slender, more
so than in T. vandykei pisgahensis which it most closely resembles;
parameres with four apical setae.

Type series. — Holotype male (AMNH) and 21 paratypes, Haoe Lead,
elevation 4800 feet (1460 m) above Joyce Kilmer Memorial Forest,
Graham County, North Carolina, 25 July 1960, T. C. Barr and M. C.
Bowling.

Measurements (in mm). — Holotype male: total length 2.70, head 0.49
long X 0.58 wide, pronotum 0.55 long X 0.80 wide, elytra 1.66 long X
1.19 wide, antenna 1.20 long.

Distribution. — Known only from the type locality in the Unicoi Moun-
tains, along the Tennessee-North Carolina border.

Discussion. — Morphologically T. haoe is closest to T. vandykei pisgahensis,
but the geographic ranges of T. tusqmtee and T. bowlingi (two other closely
similar species) intervene, suggesting complete genetic isolation. The en-
tire type series was collected from beneath a moss carpet on a large
boulder. Associated with T. haoe , but less abundant, were T. talequah and
T. barben. Trechus luculentus umcoi, the only other species of the genus thus
far known from the Unicoi Mountains, has been collected nearby but not
at the same spot.

Trechus ( Microtrechus) tusquitee , new species

Figs. 9, 28

Trechus (Microtrechus) vandykei: Barr 1962:77 (in part).

Etymology. — From Tusquitee Bald, the type locality.

Diagnosis. — A small species of the vandykei group with obtuse, not
rounded, hind angles of the pronotum, four elytral striae, the aedeagal
apex produced and truncated.

Description. — Length 2. 6-3. 2, mean 2.8 mm. Form small, robust, sub-
convex; color piceous, shining, appendages all pale, contrasting. Head

Revision of Trechus

53

feebly transverse; labrum evenly emarginate; eye diameter 0.14-0. 15 mm,
a little less than or subequal to scape length. Pronotum less than 0.7 as
long as wide, sides arcuate apical two-thirds, then convergent and barely
sinuate or not sinuate before obtuse hind angles; greatest width in apical
fourth. Elytra 1.4 times longer than wide, oval, subconvex, four
longitudinal striae and trace of fifth present, apical groove wide and
rather short, terminating at or a little anterior to level of anterior apical
puncture. Aedeagus 0.49-0.58, mean 0.53 mm long, basal bulb large and
bent at right angle to median lobe, apex narrowed, produced, abruptly
truncate; parameres long, rather slender, each with four apical setae.

Type series. — Holotype male (AMNH) and 19 paratypes, Tusquitee
Bald, Clay-Macon counties, North Carolina, 4 August 1960, T. C. Barr
and M. C. Bowling.

Measurements (in mm). — Holotype male: total length 2.95, head 0.57
long X 0.60 wide, pronotum 0.54 long X 0.83 wide, elytra 1.71 long X
1.23 wide, antenna 1.28 long.

Distribution. — The species is known at present from three small moun-
tain ranges in southwestern North Carolina, the Tusquitee/Valley River,
Snowbird, and Cheoah mountains. Two additional localities from which
I have collected T. tusquitee are Joanna (^Teyahalee) Bald, Graham-
Cherokee counties, and Cheoah Bald, Graham-Swain counties. It may
also occur in the adjacent Nantahala Mountains, although none of the
small, vandykei-\ike species have yet been found there.

Discussion. — The altitudinal range of T. tusquitee is about 3500 to 5000
feet (1070 to 1525 m); it was collected near Old Road Gap on the slopes
of Tusquitee Bald, on the north-facing cliff at the summit of Tusquitee
Bald, at a spring on the same mountain at an intermediate elevation, at
about 4500 feet (1370 m) near the summit of Joanna Bald, and at 5060
feet (1540 m) on the summit of Cheoah Bald.

The broadly transverse pronotum with obtuse hind angles and wide
marginal gutter (Fig. 9) suggests affinity with 77 bowhngi, which inhabits
the Great Smoky Mountains northeast of the range of T. tusquitee , but the
smaller and apically truncate aedeagus indicates probable status as a dis-
tinct species. Taken with T. tusquitee on Tusquitee Bald were two other
species of Trechus, T. barben and T. luculentus wayahensis.

Trechus ( Microtrechus) bowlingi Barr
Figs. 10, 29

Barr 1962:78. Type locality, Mt. Kephart, Great Smoky Mountains,

Sevier County, Tennessee; type deposited in USNM.

54

Thomas C. Barr, Jr.

Figs. 29-31. Aedeagi of Trechus species, left lateral view: 29. T. bowlingi Barr, Mt. Kephart,
North Carolina-Tennessee. 30. T. barben (Jeannel), Sassafras Knob, North Carolina-South
Carolina. 31. T. tomtru Barr, Thunderhead (Mountain), Tennessee.

Trechus bowlingi is limited to the Great Smoky Mountains, where it is
a vicar species replacing the similar T. vandykei, T. tusqmtee , and T.
haoe. Like the latter species, it occurs in a variety of habitats, such as
moss carpets, rotting leaf litter, and wet soil beneath stones. Its
altitudinal range is about 3000 to 6500 feet (900 to 2000 m). Along the
crest of the central Smokies, between Newfound Gap and Clingmans
Dome, it coexists with two other small species, T. barben and T. uncifer ,
and four larger species, T. nebulosus, T. valentinei, T. luculentus , and T.
novaculosus. Characters useful in distinguishing T. bowlingi, T. barben ,
and T. uncifer are given elsewhere in the present paper in the discussion
of T. uncifer. Among the four small species of the vandykei group, T.
bowlingi is easily differentiated by the large aedeagus and the blunt, ob-
tuse hind angles of the pronotum.

Trechus (Microtrechus) barben (Jeannel)

Fig. 30

Microtrechus Barben Jeannel 1931:444, Figs. 55-57. Type locality, Retreat,
Haywood County, North Carolina; type deposited in USNM.

Trechus ( Microtrechus ) barben : Barr 1962:78.

Trechus barben occupies a range which includes the Great Balsams and
Pisgah Ledge, North Carolina, and extends southwestward to extreme
eastern Tennessee, the mountains of northeast Georgia, and the South

Revision of Trechus

55

Carolina border. Both altitudinally and geographically it is the most
widely distributed of the Unaka assemblage of Trechus species, ranging
from the highest summits in the Great Balsams, Plott Balsams, and Great
Smokies down to 2400 feet (730 m) in the gorge of the Whitewater River.
Additional collections made since 1962 serve firmly to establish its ubiq-
uity in the main portion of its range, and I list only the southern
peripheral records below.

GEORGIA: Towns County. — Summit of Tray Mountain, elevation
4400 feet (1340 m) near White-Habersham county corner. Rabun County.
— Coleman River gorge, 0.4 mile (0.6 km) above mouth of the river in
ravines on southwest slope of Straw Mountain. Gilmer County. — Quarry
near summit of Betty Mountain, elevation 3300 feet (1000 m) on crest of
Cohutta Mountain. Murray County. — Northeast slopes of Grassy Moun-
tain along Mill Creek, elevation 3100 feet (945 m). NORTH
CAROLINA — SOUTH CAROLINA: Sassafras Mountain, elevation
3500 feet (1070 m), Transylvania-Pickens county line. NORTH CAR-
OLINA-— TENNESSEE; 0.5 mile (0.8 km) south of Stratton Meadows
on Johns Knob, elevation 4600 feet (1400 m), Graham-Monroe county
line.

The larger species of the vandykei group include T. barben, T. tomtru, and
T. subtilis. The relationship of the latter two species to T. barben is un-
clear, but probably they represent isolates whose range has been invaded
by the ecologically more successful and more abundant T. barben. The
status of T. tomtru needs clarification: it is apparently well within the
geographic range of T. barben , but the two species have not been taken
syntopically, and T. tomtru is still known from a single site. On the other
hand, T. subtilis coexists with T. barben , although it is much less
abundant.

Trechus (Microtrechus) tomtru Barr
Fig. 31

Barr 1962:79. Type locality, Thunderhead, Great Smoky Mountains,

Blount County, Tennessee; type deposited in USNM.

Trechus tomtru is known only from the summit of Thunderhead, eleva-
tion 5500 feet (1675 m) in the western Great Smoky Mountains, on the
border between Blount County, Tennessee, and Swain County, North
Carolina. The entire type series of 19 specimens was collected from
beneath moss in a scrubby heath thicket just below the top of the moun-
tain on the north (Tennessee) side. The species is closely similar to T.
barben , but the apex of the pronotum is a little wider than the base, the
pronotum sides are slightly sinuate before the right to slightly obtuse hind

56

Thomas C. Barr, Jr.

angles, and the apical groove is longer and oblique to the suture. The
aedeagus, 0.76-0.81 mm long, is a little larger than that of 77 barben; its
apex is slender, slightly reflexed, and finely knobbed, the copulatory
pieces are broader, and the parameres are conspicuously thicker and
shorter.

Trechus ( Microtrechus) subtilis Barr

Barr 1962:80, Fig. 15. Type locality, Mt. Sterling, Haywood County,
North Carolina; type deposited in USNM.

This rare species is similar to 77 barberi , with which it is sympatric and
syntopic. Only nine specimens are known, six of them from the slopes of
Mt. Sterling in the eastern Great Smoky Mountains, and three from
Jones Knob (^Junaluska Balsam) in the Plott Balsam Mountains; both
localities are in Haywood County, North Carolina. In comparison with
T. barberi, 77 subtilis is a little larger and more robust, with propor-
tionately smaller eyes, slightly transverse head, more arcuate pronotum
sides which are shallowly but distinctly sinuate before the right to slightly
obtuse hind angles, proportionately wider elytra with five distinct
longitudinal striae, and an oblique apical groove. The aedeagus, 0.85-
0.92 mm long, is larger, more slender, and less arcuate than that of 77
barberi , but shows the same general pattern.

uncifer group

Trechus ( Microtrechus ) uncifer Barr

Barr 1962:80, Fig. 16. Type locality, Clingmans Dome, Sevier County,
Tennessee; type deposited in USNM.

This species is known only from the spruce-fir forests of the central
Great Smoky Mountains and from Water Rock Knob, in the Plott
Balsams, between altitudes of 5500 and 6500 feet (1675 and 2000 m).
Males are immediately distinguished by the dense spines of the internal
sac and the long, slender, produced, hooked aedeagal apex. In the
Smokies, the only other species of Trechus in the same size range and same
localities (Clingmans Dome, Mt. Buckley, Collins Gap, Mt. Collins,
Sugarland Mountain) are T. barberi and T. bowlingi. Trechus bowlingi dif-
fers from both T. uncifer and T. barberi in slightly smaller size (2. 6-2. 9 mm,
mean 2.7 mm), evenly convex elytra not flattened on the central part of
the disc, and certain pronotal characters: pronotum very transverse,
nearly a half wider than long, marginal gutter broad and extending past
hind angles onto sides of base, hind angles very obtuse and slightly round-

Revision of Trechus

57

ed, sides scarcely sinuate before angles. Externally 77 uncifer can usually
be distinguished from 77 barben by the following combination of charac-
ters: Color of body and appendages more contrasting; head, pronotum,
and elytra dark piceous to black, appendages rather pale, testaceous,
outer segments of antenna not darker than basal segments; femora
slightly infuscated, and tibiae and tarsi paler. Pronotum sides convergent
without trace of sinuation; hind angles obtuse. Eye small, its short
diameter less than length of scape.

I have collected 77 uncifer only under rocks in damp places, usually near
seeps or small springs. It is apparently never abundant, with only one to
four specimens occurring in the same spot where large numbers of 77
luculentus or 77 valentinei are found.

Trechus ( Microtrechus) satanicus Barr

Barr 1962, Fig. 17. Type locality, Graveyard Fields, Haywood County,

North Carolina; type deposited in USNM.

Superficially similar to T. uncifer , this species is known only from the
west end of Graveyard Fields near Devils Courthouse, in the Great
Balsam Mountains. It has not been taken farther west in the Balsams, nor
on Pisgah Ledge to the east, despite careful search. Trechus satanicus is
notable for the comparatively great length of the aedeagus, 1.47-1.49 mm
long, despite the small body size (length 3. 2-3. 5 mm). Within the uncifer
group, T. uncifer and T. satanicus form a subgroup and may be vicar
species. The remaining species, T. aduncus and T. talequah , form another
subgroup. Although T. satanicus lies within the range of nominate aduncus ,
the two species have not been collected syntopically.

Trechus ( Microtrechus ) aduncus aduncus Barr, new status

Fig. 32

Trechus ( Microtrechus ) aduncus Barr 1962:82. Type locality, Mt. Pisgah,

Haywood County, North Carolina; type deposited in USNM.

Trechus aduncus is a medium-small (3. 1-3.9 mm) species with eye
diameter equal to or slightly less than the scape length, the pronotum
0.65-0.70 times as long as wide, pronotum apex and base widths sub-
equal, the sides not or very feebly sinuate just before the small, right or
slightly obtuse hind angles, the elytra with three inner striae clearly im-
pressed and one to three additional, feebly impressed outer striae. The
aedeagus is moderately arcuate to strongly arcuate, the apex more or less
produced and terminating in a hook. The two copulatory pieces, which
are normally obscured by the spiny armature of the internal sac, are sim-
ple, subequal, and have triangular apexes.

58

Thomas C. Barr, Jr.

Figs. 32-35. Aedeagi of Trechus species, left lateral view: 32. T. aduncus aduncus Barr, Bear-
pen Gap, North Carolina. 33. T. aduncus toxawayi, new subspecies, Toxaway Mountain,
North Carolina. 34. T. aduncus coweensis, new subspecies, Yellow Mountain, North Carolina.
35. T. aduncus howellae , new subspecies, Big Butt, North Carolina.

Nominate aduncus , which has only one elytral stria beyond the inner
three and an aedeagus as in Figure 32, occurs along the crest of the Great
Balsam Mountains, from Richland Balsam eastward to Mt. Pisgah, in
Haywood, Transylvania, Buncombe, and Jackson counties, North
Carolina. I have collected it at Mt. Pisgah, Buck Spring, Devils
Courthouse, Bearpen Gap, Spot Knob, and Richland Balsam, at eleva-
tions ranging between 4600 and 6000 feet (1400 and 1800 m). Its usual
habitat is wet moss in contact with a rock surface, and it is sympatric and
syntopic at various localities with other Trechus species: T. barberi, T. van-
dykei pisgahensis, T. schwarzi schwarzi , and T. rosenbergi. The type (and only
known) locality for T. satamcus lies near the middle of the range of T. a.
aduncus , but the two species have not yet been collected at the same spot.

Four subspecies of T. aduncus are recognized in the present paper. All
are probably extrinsically isolated populations which have diverged re-
cently, and they are treated as a polytypic species to emphasize their close
relationship. Intergrades between the subspecies have not been found,
and (except possibly for nominate aduncus and toxawayi ) are unlikely to be
found because of the existence of low gaps between the ranges of the
subspecies.

Revision of Trechus

59

Trechus (Microtrechus) aduncus toxawayi , new subspecies

Fig. 33

Etymology. — From Toxaway Mountain, the type locality.

Description. — Length 3.2-3. 3 mm. Elytra with three inner striae clearly
impressed, at best a trace of fourth stria; elytra rather slender, elongate-
oval. Aedeagus slender, 0.91-0.97 mm long, less arcuate than in the other
subspecies of T. aduncus , apex slender and produced more than in
nominate aduncus , terminating in reflexed hook.

Type series. — Holotype male (AMNH), two male and one female
paratypes, 0.25 mile (0.4 km) west of the summit of Toxaway Mountain,
elevation 4600 feet (1400 m), Jackson County, North Carolina, 24 June
1970, T. C. Barr, Jr., and T. C. Barr, III.

Measurements (in mm). — Holotype male: total length 3.25, head 0.62
long X 0.68 wide, pronotum 0.66 long X 0.89 wide, elytra 1.89 long X
1.31 wide, antenna 1.57 long.

Distribution. — Known only from the type locality, at the end of a divide
(Tanasee Ridge) which extends 17 miles (27 km) south from the crest of
the Great Balsams. The ridge possibly permits continuity between T. a.
toxawayi and nominate aduncus , but no intermediate populations are
known.

Discussion. — The type series was collected from beneath a mat of
decaying leaves at the base of a low, sloping rock cliff, thus from very
typical T. aduncus microhabitat. Trechus barben was the only other species
of the genus taken on Toxaway Mountain.

Trechus ( Microtrechus ) aduncus coweensis new subspecies

Fig. 34

Etymology. — From Cowee Mountains, where the type locality is
situated.

Description. — Length 3. 4-3. 8, mean 3.7 mm. Elytra with at least five
visible striae, inner three striae more deeply impressed. Aedeagus 0.84
mm long in paratype, median lobe thicker than in T. a. aduncus or T. a.
toxawayi , apex slender, produced, slightly recurved, terminal hook knob-
like.

Type series. — Holotype male (AMNH), three male and one female
paratypes, summit of Yellow Mountain, elevation 5000 feet (1525 m), 4
miles (6.4 km) southwest of Glenville, Macon-Jackson counties, North
Carolina, 28 June 1969, T. C. Barr.

60

Thomas C. Barr, Jr.

Measurements (in mm). — Holotype male: total length 3.84, head 0.84
long X 0.74 wide, pronotum 0.68 long X 1.02 wide, elytra 2.17 long X
1.55 wide, antenna 1.61 long.

Distribution. — Known only from the type locality in the Cowee Moun-
tains, North Carolina.

Discussion. — The five specimens of T. a. coweensis were taken from
beneath wet leaves alongside the road to the fire tower, just below the
summit of Yellow Mountain at an approximate elevation of 5000 feet
(1525 m). A single T. barben was collected at the same spot. This sub-
species is geographically and morphologically a little closer to T. a.
howellae, from the Nantahala Mountains farther west, than to either of the
other two known subspecies of T. aduncus. The Cowees are separated from
the Great Balsams by the valley of the Tuckasegee River and from the
Nantahalas by the Little Tennessee River, thus there is every indication
of complete extrinsic isolation of this taxon.

Trechus ( Microtrechus) aduncus howellae , new subspecies

Fig. 35

Etymology. — Patronymic honoring Dr. Thelma Howell, director-
emeritus of the Highlands Biological Station.

Description. — Length 3. 3-3. 9, mean 3.7 mm. Elytra with at least five
visible striae, inner three striae more deeply impressed. Aedeagus 0.87-
0.88 mm long, very strongly and evenly arcuate in median lobe, which is
thick, as in T. a. coweensis ; apex more evenly attenuate and terminal hook
less reflexed than in T. a. aduncus , but not knob-like as in T. a. coweensis.

Type series. — Holotype male (AMNH) and seven paratypes, east face of
Big Butt, elevation 4800 feet (1460 m), Coweeta Hydrologic Laboratory,
Macon County, North Carolina, 21 July 1970, T. C. Barr, Sr., T. C. Barr,
Jr., and T. C. Barr, III. Two paratypes, same locality, 26 July 1969, T. C.
Barr, Jr.

Measurements (in mm). — Holotype male: total length 3.58, head 0.84
long X 0.68 wide, pronotum 0.68 long X 0.96 wide, elytra 2.05 long X
1.43 wide, antenna 1.61 long.

Distribution. — Known only from the type locality, in a research area
maintained by the United States Forest Service, in the Nantahala
Mountains.

Discussion. — The type series was collected from wet moss at the base of
a vertical cliff face beside the Appalachian Trail on the east face of Big

Revision of Trechus

61

Butt, a prominent peak in the Nantahala Mountains. The only other
Trechus species associated with T. a. howellae was T. barben , a pair of which
was also taken at the Big Butt site.

Trechus ( Microtrechus) talequah Barr

Barr 1962: 82, Fig.^20. Type locality, Haw Knob, Monroe County, Ten-
nessee; type deposited in USNM.

Trechus talequah is a Unicoi Mountains vicar species to T. aduncus.
Morphologically and geographically it is much farther removed from the
subspecies of T . aduncus than they are from each other. It is smaller and
flatter than T. aduncus , has deeply impressed elytral striae and a shorter
apical groove. The aedeagus is shorter and straighter, but its form
nevertheless suggests a close affinity with T. aduncus. The species coexists
at Haw Knob (type locality) with T. luculentus unicoi and on Haoe Lead,
Graham County, North Carolina, with T. haoe. Females and undissected
males are easily separated from T. haoe by the depressed form and deep
elytral striae.

nebulosus group

Trechus ( Microtrechus ) nebulosus Barr

Barr 1962: 86, Fig. 21. Type locality, Mt. Kephart, Sevier County, Ten-
nessee, and Swain County, North Carolina; type deposited in USNM.

Trechus nebulosus is a species of medium size (3. 3-4.0 mm) with large
eyes, the sides of the pronotum not sinuate, and the hind angles obtuse
and slightly reflexed. It is abundant beneath wet, fluffy moss which car-
pets north-facing cliffs along the crest of the Great Smoky Mountains
from Clingmans Dome to Mt. Guyot, including the Mt. LeConte spur. In
this region it coexists with three smaller species of Trechus (T. barben, T.
bowlingi, T. uncifer) and with four other species of the nebulosus group ( T . 1.
luculentus, T. valentinei, T. novaculosus , and T. verus). The only one of these
species in the same size range and also having large eyes is T. luculentus ,
but in that species the sides of the pronotum are distinctly sinuate before
the hind angles. Males of T. nebulosus are always easily determined if the
aedeagus protrudes or is removed, because the tip of the right copulatory
piece is large and twisted into the shape of a bird’s head.

The lowest elevation at which I have collected T. nebulosus is about 4600
feet (1400 m) from mossy boulders along Ramsay Prong, above the
Cascades, in Sevier County, Tennessee, on the north side of the Smokies.
At higher elevations it is the most common of the larger species in the
moss carpet assemblage.

62

Thomas C. Barr, Jr.

Trechus ( Microtrechus) balsamensis Barr

Barr 1962:87, Fig. 24. Type locality, Water Rock Knob, Haywood-

Jackson counties, North Carolina; type deposited in USNM.

This species is known only from the type locality, at an elevation of
6200 feet (1900 m) in the Plott Balsam Mountains. It occurs beneath
moss carpets and in spruce-fir needle duff at the base of low cliffs on the
north face of Water Rock Knob, in spruce-fir forest. It is syntopic with 77
rosenbergi and 77 barben at the type locality. Both 77 uncifer and 77 subtilis
have been taken in the Plott Balsams, but not at the Water Rock Knob
locality.

Trechus balsamensis resembles T. nebulosus in habitus and habitat select-
ed. It is characterized by large eyes, pronotum sides feebly sinuate only in
basal 0.05, hind angles small and approximately right, very feebly im-
pressed elytral striae, only the inner two or three of which are present;
aedeagus 0.93-1.05 mm long, of the same size and shape as that of T.
nebulosus , but apex much broader in dorsal view, the right piece scoop-
shaped, much larger than the left and without an apical twist.

Trechus ( Microtrechus ) valentinei, new species

Figs. 11, 36

Trechus ( Microtrechus ) verus: Barr 1962 (in part).

Etymology. — Patronymic honoring Dr. J. Manson Valentine for his
early studies of North American trechines.

Diagnosis. — Distinguished from other species of the nebulosus group by
small eyes, broad pronotum base with acute hind angles, pale coloration;
aedeagal apex attenuate and produced, finely knobbed, twisted to left;
parameres broad and non-styliform.

Description. — Length 3. 6-3. 9, mean 3.7 mm. Piceous, more or less pale,
shining, form moderately robust and subconvex. Head slightly longer
than wide; labrum shallowly emarginate; eyes small, their diameter 0.7
scape length; antenna about 0.45 body length. Pronotum three-fourths as
long as wide, conspicuously transverse; apex 0.9 times as wide as base
and only 0.7 maximum width, which occurs in apical third; margins con-
vergent behind, not sinuous; hind angles small, sharp, acute, laterally
produced, base slightly lobed; basal foveae large and deep. Elytra 0.4
longer than wide, disc subconvex, usually with five longitudinal striae,
4th and 5th more shallowly impressed; apical groove short, joining 5th
stria a short distance anterior to anterior apical puncture; scutellar stria
very short, obsolescent; anterior discal at level of fourth umbilicate.

Revision of Trechus

63

Aedeagus 0.92-1.02, mean 0.97 mm long; thick, strongly arcuate, basal
keel strong, apex attenuate, produced, curved to left, finely knobbed;
copulatory pieces exserting to right, right piece much larger, heavily
sclerotized, enfolding smaller left piece; internal sac armed with medial
band of dense, slender spines; parameres broad, non-styliform (Fig.
36A), in left lateral view, with seta-bearing surfaces peculiarly elongate,
obliquely truncate; usually only three setae on apex of each paramere,
rarely four.

'•j

Type series. — Holotype male (AMNH) and 12 paratypes, Appalachian
Trail near summit of Mt. Kephart, Sevier County, Tennessee, Great
Smoky Mountains National Park, altitude approximately 6000 feet (1800
m), 1 July 1960, T. C. Barr, M. C. Bowling, Joyce and R. T. Bell.

Measurements (in mm). — Holotype male: total length 3.80, head 0.78
long X 0.71 wide, pronotum 0.78 long X 1.02 wide, elytra 2.23 long X
1.61 wide, antenna 1.74 long.

Distribution. — Known only from the central Great Smoky Mountains of
Tennessee and North Carolina, from about 3500 to 6000 feet (1070 to
1830 m) in elevation.

Discussion. — In my earlier paper (Barr 1962) T. valentinei was confused
with T. verus, another pale species with small eyes. The older records cited
for 77 verus at Clingmans Dome, Sugarland Mountain, and Mt. Kephart
(Barr 1962:82) are incorrect and apply to T. valentinei. True T. verus is
known only from Mt. Sterling, Cataloochee Balsam, and Old Black, in
the eastern end of the Smokies.

Trechus valentinei is an inhabitant of the high spruce-fir forests of the
Great Smokies, where it occurs under rocks beside streams and springs
and in shallow ravines, often in company with the more abundant T. 1.
luculentus and occasionally with a rare specimen of T. uncifer or T.
novaculosus. In addition to the type locality I have taken this species on Mt.
LeConte, near the summit; Clingmans Dome, near the summit; Mt.
Collins, near the summit; Sugarland Mountain, near its junction with the
main crest of the Smokies. It can be readily distinguished from sympatric
species of Trechus by the small eyes and paler color (T. verus , also pale with
small eyes, is apparently allopatric), the broad pronotum base and acute
hind angles, and the slender, produced, sinistrally twisted aedeagal apex
(often visible without dissection). The non-styliform parameres are ex-
ceptional, duplicated in no other trechine species known to me (although
rather different non-styliform parameres occur in Italian and Yugoslavian
species of Orotrechus Muller).

64

Thomas C. Barr, Jr.

Figs. 36-39. Aedeagi of Trechus species, left lateral view (except 36A): 36. T. valentinei, new
species, Mt. LeConte, Tennessee. 36A. T. valentinei , ventral view, Ramsay Cascades, Ten-
nessee. 37. T. verus Barr, Mt. Sterling, North Carolina. 38. T. stupkai, new species, Ramsay
Prong, Tennessee. 39. T. luculentus luculentus Barr, Clingmans Dome, North Carolina-
Tennessee.

Revision of Trechus

65

Trechus ( Microtrechus) verus Barr
Fig. 37

Barr 1962:81. Type locality, Mt. Sterling, Haywood County, North

Carolina; type deposited in USNM.

When describing T. verus I placed it in the uncifer group, where it is
readily distinguished by its large size and the absence of heavy, scaly ar-
mature on the internal sac. Both characters, so highly diagnostic in the
uncifer group, are not at all out of the ordinary in the nebulosus group,
where it more properly belongs alongside the closely similar species T.
valentinei , with which I had earlier confused it. True T. verus is confined to
the eastern end of the Smokies in Haywood County, North Carolina, and
Cocke County, Tennessee, where it has been taken on Mt. Sterling, Old
Black, and Cataloochee Balsam. Like T. valentinei , which it resembles, T.
verus has small eyes and pale coloration and commonly occurs beneath
rocks in wet places. The aedeagus of T. verus is more slender than that of
T. valentinei , the apex is more conspicuously reflexed and not twisted to
the left, and the parameres are styliform.

Trechus ( Microtrechus ) stupkai , new species
Figs. 12, 38

Etymology. — Patronymic honoring Mr. Arthur Stupka, former Park
Naturalist and Research Biologist, National Park Service.

Diagnosis. — Eye diameter less than scape length; pronotum with apex
and base widths subequal, sides briefly sinuate before small, right, hind
angles; aedeagal apex briefly produced straight back, truncate at tip.

Description. — Length of unique holotype 3.3 mm. Form moderately
robust and subconvex; dark piceous, appendages all pale. Head one
fourth longer than wide; labrum evenly emarginate; eye diameter less
than length of scape. Pronotum three fourths as long as wide, apex and
base subequal, about 0.7 greatest width, which occurs in apical third;
sides rounded apical two thirds, then convergent, shallowly but distinctly
sinuate before small, about right, hind angles. Elytra 1.4 times longer
than wide, oval, subconvex, inner two striae moderately impressed, third
feebly impressed, fourth stria present only as evanescent trace; apical
groove wide and short, ending slightly in advance of anterior apical
puncture. Aedeagus of holotype 0.78 mm, basal bulb large and bent at
right angles to median lobe, apex rather abruptly produced straight back
and truncate with terminal button; copulatory pieces elongate-lobulate,
left three fourths as long as right, internal sac weakly armed with small,

66

Thomas C. Barr, Jr.

indistinct scales; parameres slender and elongate, apexes with four setae
each and not obliquely truncate as in T. valentinei.

Type series. — Holotype male (AMNH), a unique, in mossy rocks at
edge of Ramsay Prong, about 150 yards (135 m) above Ramsay
Cascades, elevation 4600 feet (1400 m), Sevier County, Tennessee, in
Great Smoky Mountains National Park, 25 May 1969, T. C. Barr.

Measurements (in mm). — Holotype male: total length 3.29, head 0.74
long X 0.60 wide, pronotum 0.62 long X 0.83 wide, elytra 1.85 long X
1.28 wide, antenna 1.39 long, aedeagus 0.78 long.

Distribution. — Known only from the type locality.

Discussion. — Efforts to obtain additional specimens of this distinctive
species have been unsuccessful. Trechus nebulosus, T. valentinei , and T.
bowlingi were collected in the same microenvironment along Ramsay
Prong, but even these otherwise common species are relatively rare at the
site.

Trechus (Microtrechus) novaculosus Barr
Fig. 44

Barr 1962:89, Fig. 27. Type locality, Clingmans Dome, Sevier County,

Tennessee; type deposited in USNM.

This large species, confined to the spruce-fir forests of the central Great
Smoky Mountains, is the largest species of Trechus in the Smokies. The
usual size range in the vicinity of Clingmans Dome is 4.4-4. 8, mean 4.6
mm, but J. Manson Valentine took three very large specimens, fully 5.5
mm long, on Mt. FeConte in May 1934. Seventeen specimens from Mt.
Kephart and along Anakeesta Ridge, which connects Mt. FeConte with
the main crest of the Smokies, are intermediate in length, 5. 1-5.5, mean
5.2 mm; these beetles, which I collected in August 1975, occurred several
inches below the surface of wet scree piles along the trails.

The aedeagus of Clingmans Dome area specimens, 1.22-1.24 mm long,
is smaller than the 1.23-1.32 mm length range seen in the Mt. FeConte-
Anakeesta Ridge material, but the basic pattern is identical: rather
strongly arcuate in lateral view, the median lobe sharply bent near the
middle and produced and attenuate at the apex; in dorsal view the apex is
simple and broadly spatulate. The copulatory sclerites are simple,
spatulate, quite elongate and narrow, almost razor-like in shape.

From other species of the nebulosus group in the Smokies, T. novaculosus
is readily distinguished by large size, small eyes, convergent (i.e., non-
sinuate) pronotum sides, and large, obtuse hind angles. I found it only

Revision of Trechus

67

under and among rocks in wet places near seeps and springs at altitudes
above 5500 feet (1675 m). In such microhabitats it is often associated with
77 valentinei, which was always more abundant, and rarely with 77 uncifer.
The available specimens suggest a possible geographic cline in size, with
total length increasing from Clingmans Dome along the crest to Mt.
Kephart, then along the Boulevard across Anakeesta Ridge to Mt.
LeConte. In the past three summers (1976, 1977, 1978) I have attempted
to find intermediate populations (e.g., near Newfound Gap) but have
been unable to do so. If the species is actually restricted to the very high
elevations where it has thus far been taken, then one is faced with the
alternative to a cline, that there are two genetically isolated populations
less than 10 miles (16 km) apart. The only specimens known thus far from
Mt. LeConte are those of Valentine, now in my private collection together
with the series from Mt. Kephart and Anakeesta Ridge. In addition to
Clingmans Dome the species has been taken on Mt. Buckley, Mt. Collins,
and along the trail between Clingmans Dome and Andrews Bald; all
specimens from these four localities fall into the 4.4-4. 8 mm size range.

Trechus (Microtrechus) luculentus luculentus Barr, new status

Figs. 13, 39

Trechus ( Microtrechus ) luculentus Barr 1962:88. Type locality, Clingmans

Dome, Swain County, North Carolina; type deposited in USNM.

Trechus luculentus s. lat. is characterized by large eyes, their diameter
equal to or greater than the length of the scape, and distinctly sinuate
sides of the pronotum. The apex of the aedeagus is not produced; the dor-
sal (right?) copulatory sclerite is large and scoop-shaped, and the ventral
(left?) sclerite is a small, elongate-triangular spicule. Since collection of
the type series in 1960, I have obtained specimens of T. luculentus from
many other localities in the Great Smoky, Unicoi, and Nantahala moun-
tains. Populations in the Unicois and Nantahalas differ subspecifically
from Smoky Mountain T. luculentus. The nominate race is distinguished
from the other subspecies by the combination of smaller hind angles of
the pronotum, usually five longitudinal elytral striae, and the small
aedeagus (length 0.63-0.73 mm).

In the central Smokies, T. 1. luculentus has proven to be widely dis-
tributed altitudinally, occurring under rocks or debris beside seeps and
springs or near water in little ravines. Near the summit of Clingmans
Dome it is found under rocks in little wet ravines and under broken drain
tiles near the septic disposal field. Along Roaring Fork and in Greenbrier
Cove it was taken from beneath leaves and sticks beside small tributary
streams at elevations of 3000 feet (915 m) and 2900 feet (885 m), respec-

68

Thomas C. Barr, Jr.

tively. A single specimen was found on Fish Camp Prong under moss car-
peting a low ledge at an elevation of 2200 feet (670 m). On West Prong of
Little Pigeon River (4000 feet) (1200 m) it occurred under rocks at the
edge of the stream, with 77 valentinei and Bembidion carolinense Casey. At
the south base of the Smokies (Swain County, North Carolina) 77
luculentus was extracted from leaves and wet gravels at 2250 feet (685 m),
beside a small tributary stream crossing the Deep Creek Trail. A single
female was collected on Kephart Prong under similar circumstances, at
an elevation of 3350 feet (1020 m). These collecting data have been given
in some detail to demonstrate that 77 luculentus is by no means restricted
to higher elevations, but should be able to bridge the low gaps between
mountain ranges. A limited amount of gene flow between nominate
luculentus and the other two subspecies is thus a reasonable proposition,
although intergrading populations have not yet been discovered.

Wherever it coexists syntopically with 77 valentinei, T. luculentus is
always two to three times more abundant. It occasionally is taken
together with 77 barben and 77 bowhngi , both conspicuously smaller
species, and rarely with 77 uncifer and 77 novaculosus at higher elevations.
In the central Smokies the only other large species of Trechus with large
eyes is 77 nebulosus , an inhabitant of wet, fluffy moss carpets (consequent-
ly almost never syntopic with 77 luculentus ); in 77 nebulosus the sides of the
pronotum are simply convergent to the base, and males are readily deter-
mined by the distinctive shape of the apex of the dorsal copulatory piece
which often protrudes from the apical orifice of the aedeagus.

Trechus ( Microtrechus) luculentus umcoi, new subspecies

Figs. 14, 40, 41

Trechus ( Microtrechus ) luculentus: Barr 1962:88 (in part).

Etymology. — From Unicoi Mountains, where the type locality is
situated.

Description. — Length 4. 3-4. 5, mean 4.4 mm. Differs from other sub-
species of T. luculentus in more prominently sinuous pronotum sides,
which are subparallel in basal 0.12-0.15; more prominent, acute, hind
angles of pronotum; more extensive longitudinal striation of elytra, with
all striae usually present, although 5th through 8th progressively
shallower; and larger aedeagus, 0.95-1.04, mean 1.00 mm long.

Type series. — Holotype male (AMNH) and 22 paratypes, Stratton
Meadows, elevation 4900 feet (1494 m), Monroe County, Tennessee, and
Graham County, North Carolina, 30 August 1964, T. C. Barr.

Revision of Trechus

69

Measurements (in mm). — Holotype male: total length 4.48, head 1.00
long X 0.83 wide, pronotum 0.89 long X 1.19 wide, elytra 2.54 long X
1.89 wide, antenna 1.68 long.

Distribution. — In addition to the type series, I have seen specimens from
the following localities. TENNESSEE: Monroe County. — Sugar Cove
Creek, elevation 3000 feet (900 m) (T. C. Barr, 1964); Laurel Top, near
Big Junction, elevation 5200 feet (1585 m) (T. C. Barr, 1960); Haw
Knob, elevation 5400 feet (1645 m) (T. C. Barr, 1960). NORTH
CAROLINA-: Graham County. — Joyce Kilmer Memorial Forest, elevation
3000 feet (900 m) (J. M. Valentine, 1950-51); Bemis Lumber Company
Camp (J. M. Valentine, 1951); Johns Knob, near Stratton. Meadows,
elevation 4800 feet (1465 m) (T. C. Barr, 1971). All localities are in the
Unicoi Mountains, and specimens are at present in my private collection.

Discussion. — Previously (Barr 1962: 88) I noted the occurrence of this
subspecies in the Unicoi Mountains but did not name it because insuf-
ficient specimens were available. I have now examined 55 specimens from
the seven stated localities. The subspecies has the same microhabitat
preferences of nominate luculentus and is prevalent beneath rocks and moss
in wet ravines and near seeps and springs.

Figs. 40-43. Aedeagi of Trechus species, left lateral view (except 41): 40. T. luculentus unicoi,
new subspecies, Stratton Meadows, North Carolina-Tennessee. 41. T. luculentus unicoi, apex,
dorsal view. 42. T. luculentus wayahensis, new subspecies, Dirty John Creek, North Carolina.
43. T. nantahalae, new species, Burningtown Bald, North Carolina.

70

Thomas C. Barr, Jr.

Trechus (Microtrechus) luculentus wayahensis , new subspecies

Figs. 15, 42

Etymology. — From Wayah Bald, a prominent topographic feature near
the type locality.

Description. — Length 3. 9-4.4, mean 4.1 mm. Differs from T. 1. luculentus
in having only four elytral striae and larger aedeagus, 0.84-0.95 mm long;
differs from T. 1. umcoi in having less prominent hind angles of pronotum,
only four elytral striae, and smaller aedeagus.

Type series. — Holotype male (AMNH) and 30 paratypes, ravine at
head of Dirty John Creek, southwest slope of Winespring Bald (near
Wayah Bald), elevation 4900 feet (1500 m), Macon County, North
Carolina, 3 July 1969, T. C. Barr.

Measurements (in mm). — Holotype male: total length 4.42, head 1.15
long X 0.87 wide, pronotum 0.84 long X 1.15 wide, elytra 2.48 long X
1.98 wide, antenna 2.11 long.

Distribution. — In addition to the type locality, this subspecies is known
from two other sites in Macon County, North Carolina: Berties Falls,
near Wayah Bald, elevation 4000 feet (1200 m) ; and Tusquitee Bald, near
the Macon-Clay county line, elevation 4000 feet (1200 m). Tusquitee
Bald, part of the Tusquitee/Valley River mountains, is 7 miles (11.3 km)
west of Wayah Bald across the Nantahala River valley.

Discussion. — The occurrence of this subspecies in the Nantahala Moun-
tains was first brought to my attention through the kindness of Dr. J.
Manson Valentine, who gave me three females he had collected at Bertie
Falls and Dirty John Creek in 1950. A single large female Trechus
specimen which I collected on Tusquitee Bald in 1960 was earlier referred
tentatively to T. schwarzi (Barr 1962:75), but this specimen properly
belongs with T. luculentus wayahensis. Two other species of Trechus, T.
barben and T. tusquitee , have been collected syntopically with T. wayahensis ,
but T. aduncus howellae and T. nantahalae, which also occur in the Nan-
tahalas, have not. Trechus nantahalae is readily distinguished from T.
luculentus by the smaller eyes, pale color, and obtuse hind angles of the
pronotum; the other three species are conspicuously smaller.

Trechus ( Microtrechus ) rosenbergi Barr

Barr 1962: 89, Fig. 26. Type locality, Water Rock Knob, Haywood and

Jackson counties, North Carolina; type deposited in USNM.

Trechus rosenbergi is a large (4. 5-5.0 mm) species, comparable in size
only to T. carolinae (Black Mountains) and T. novaculosus (central

Revision of Trechus

71

Smokies). It is known from only two localities: Water Rock Knob, in the
Plott Balsam Mountains, and Richland Balsam, Haywood County,
North Carolina, in the Great Balsam Mountains. At both collecting sites
it was taken deep in the mat of spruce-fir needles piled up against wet,
vertical rock faces. It is easily distinguished from the other species of
Trechus with which it is sympatric (77 aduncus, T. balsamensis, T. barben, T.
subtilis, T. vandykei) by size alone.

Trechus ( Microtrechus) tuckaleechee Barr

Fig. 16

Barr 1962:86, Fig. 22. Type locality, Tuckaleechee Caverns, Blount
County, Tennessee; type deposited in USNM.

This species is known only from Tuckaleechee Caverns near Town-
send, Tennessee, at the north side of the Great Smoky Mountains. It is
distinguished from other members of the nebulosus group by small eyes,
sinuate sides of the pronotum, hind angles sharp and nearly right, elytra
with five clearly impressed striae and short apical groove, and the apex of
the aedeagus broad, finely reflexed at the very tip. Morphologically it ap-
pears to be closest to T. tennesseensis and T. nantahalae.

Trechus ( Microtrechus ) tennesseensis tennesseensis Barr

Barr 1962:87, Fig. 23. Type locality, Berry Cave, Roane County, Ten-
nessee; type deposited in USNM.

This taxon is at present known from a single locality in the floor of
the Appalachian valley approximately 50 miles (80 km) west of the
Great Smoky Mountains. Berry Cave is located a quarter of a mile (0.4
km) west of the Tennessee River (mile 578.4) near Wright Bend, 0.15
mile (0.24 km) west of Tennessee Route 72, at an elevation of 840 feet
(255 m). Trechus t. tennesseensis resembles T. tuckaleechee , but the hind
angles of the pronotum are sharp and acute, and the aedeagus is much
larger. Neither T. tuckaleechee nor T. tennesseensis exhibit any regressive
modifications that would restrict them to cave habitats. Presumably
they are both relics of species that were more widely distributed during
glacial maxima.

Trechus ( Microtrechus ) tennesseensis tauricus Barr

Barr 1962: 87. Type locality, Bull Cave, Blount County, Tennessee; type
deposited in USNM.

This subspecies, known only from the sinkhole at the entrance to Bull

72

Thomas C. Barr, Jr.

Cave, Blount County, Tennessee, at the edge of Great Smoky Mountains
National Park (below the Rich Mountain Road, elevation 1840 feet
[560m]), differs from nominate T. tennesseensis in the absence of internal
clypeal grooves and outer elytral striae. The two taxa treated as polytypic
T. tennesseensis are probably quite isolated, at least extrinsically, at the
present time, but the morphological differences between them are so
slight that I consider them relatively recent (post-Wisconsin?) relics of a
previously more widely distributed species.

Trechus ( Microtrechus) nantahalae , new species
Figs. 17, 43, 45

Etymology. — From Cherokee nantahala , geographic name for the moun-
tain range in which the type locality occurs.

Diagnosis. — Resembles T. tennesseensis in small eyes, pale coloration,
sinuate sides of pronotum, and slender aedeagal apex; differs in obtuse
hind angles of pronotum, shorter apical recurrent groove, smaller
aedeagus, and knobbed right copulatory piece.

Description. — Length 4. 2-4. 4, mean 4.2 mm. Reddish piceous, shining;
microsculpure obsolete on central elytral disc; form robust and subcon-
vex. Head a little wider than long; labrum shallowly and singly
emarginate; frontal grooves broad and deep; eyes small and subconvex,
their diameter (about 0.15 mm) less than scape length; antenna about
0.45 body length. Pronotum 0.3 wider than long; apex and base subequal
in width and about 0.7 maximum width; widest in apical third; sides ar-
cuate, sinuous in basal sixth, hind angles small, sharp, slightly elevated,
and obtuse; basal foveae broad and deep, separated from marginal gutter
by low ridge; base oblique behind hind angles, thus slightly lobed. Elytra
0.75 as wide as long; inner three longitudinal striae moderately deep, in-
ner intervals subconvex, 4th and 5th striae shallow, outer striae obsolete;
apical recurrent groove attaining 5th stria a short distance in advance of
anterior apical puncture; scutellar stria very short, obsolescent; anterior
discal puncture slightly behind level of 4th umbilicate puncture.
Aedeagus of paratype 0.94 mm long, similar to that of T. tennesseensis but a
little shorter, apex slender in dorsal view; dorsal (right) copulatory piece
with apex twisted into knob. Female unknown.

Type series. — Holotype male (AMNH) and 4 male paratypes, 0.4 mile
(0.6 km) northwest of Burningtown Gap on the southwest slope of Burn-
ingtown Bald, at a seep along the Appalachian Trail, elevation 4300 feet
(1300 m), Macon County, North Carolina, 5 July 1969, T. C. Barr.

Revision of Trechus

73

Figs. 44-45. Trechus species, habitus sketch: 44. T. novaculosus Barr, large form (5.5 mm),
Mt. LeConte, Tennessee. 45. T. nantahalae, new species (4.3 mm), Burningtown Bald, North
Carolina.

Measurements (in mm). — Holotype male: total length 4.22, head 0.93
long X 0.81 wide, pronotum 0.84 long X 1.09 wide, elytra 2.45 long X
1.86 wide, antenna 1.92 long.

Distribution. — Known only from the type locality in the northern Nan-
tahala Mountains, southwestern North Carolina.

Discussion. — Trechus nantahalae is readily distinguished from the other
species of Trechus with which it is sympatric in the Nantahalas (T. barben,
T. aduncus howellae, T. luculentus wayahensis, T. tusquitee) by the small eyes,
pale coloration, and peculiarly obtuse hind angles of the pronotum which

74

Thomas C. Barr, Jr

_T C/3

C ^3

O

§

U _r»

>- c

<*>

■§G

-t-^ l4_|

c o

C/3

<U

Q- Z

\

0»-

Y V''v

/

, OL

UJ v<f

/

i £

X

/ *

) <

►—

/ =5,

» “-Z

"N_ /

~W

/ V

/ ^ 1

1 ^

Qh

52

•

4

j lZ ^

r* U>- -

» /<*
xo

\ ^

V

\ LLlf

V. “v

o

£ u

<u

Jg g*
« >-
C .Q

5-

13

Revision of Trechus

75

result in a lobed base. The five specimens of the type series, all males,
were collected from beneath stones in wet mud. No other Trechus species
were found in the immediate vicinity, nor has T. nantahalae yet been found
elsewhere in the Nantahala Mountains. In fact, the limited collections
thus far made of Trechus in the Nantahala and Tusquitee/Valley River
mountains suggest that the somewhat lower elevations along the crest and
the absence of spruce-fir forest provide a restricted number of suitable
microhabitats in contrast to the Great Smokies, Great Balsams, and Plott
Balsams. Of the four endemic taxa known at the present time only T.
luculentus wayahensis and T. tusquitee have been taken together (on Tus-
quitee Bald), in contrast to many syntopic species pairs known from the
other ranges mentioned.

ACKNOWLEDGMENTS. — This study was supported in part by
grants from the National Science Foundation (G-12968 and GB-5521 to
me, and through a grant-in-aid from the Highlands Biological Station,
GB-2496), the Theodore Roosevelt Memorial Fund, and the American
Philosophical Society (Penrose Fund No. 6180).

LITERATURE CITED

Barr, Thomas C., Jr. 1962. The genus Trechus (Coleoptera: Carabidae: Trechini)
in the southern Appalachians. Coleopt. Bull. 76:65-92.

1969. Evolution of the Carabidae (Coleoptera) in the southern Ap-
palachians. pp 67-92 in Holt, Perry C. (ed.). The distributional history of the
biota of the southern Appalachians, Part I: Invertebrates. Res. Div. Monogr.
1, Va. Polytech. Inst., Blacksburg. 295 pp.

Casey, Thomas L. 1918. Observations on the American Pogoninae, including
Trechus. Memoirs on the Coleoptera, vol. 8: 394-412. Privately printed, Lan-
caster, Pennsylvania.

Emerson, Alfred E. 1945. Taxonomic categories and population genetics. En-
tomol. News 56: 14-19.

Horn, George H. 1883. Miscellaneous notes and short studies of North American
Coleoptera. Trans. Am. Entomol. Soc. 10: 269-312.

Jeannel, Rene. 1927. Monographic des Trechinae. Morphologie comparee et dis-
tribution geographique d’un groupe de Coleopteres (2e livraison). L’Abeille
33: 1-592. "

1931. Revision des Trechinae de l’Amerique du Nord. Arch Zool. exp.

gen., 77:403-499.

Schaeffer, Charles. 1901. Synopsis of the species of Trechus , with the description of
a new species. Bull. Am. Mus. Nat. Hist. 14: 209-212.

Accepted 22 August 1979

1

Cretaceous Selachians from the Peedee Formation
(Late Maestrichtian) of Duplin County, North Carolina1

Gerard R. Case

129 Carlton Avenue, Jersey City, New Jersey 07306

ABSTRACT. — A shark-tooth fauna of Late Cretaceous (Navarroan,

Late Maestrichtian) age from the uppermost member of the Peedee For-
mation was recovered from three sites in southern Duplin County,
North Carolina. Species represented include Hybodus sp. 1; Squalicorax
kaupi and S. pnstodontus; the goblin shark Scapanorhynchus texanus;
Odontaspis sp.; Cretolamna appendiculata lata; C. biaunculata (the first New
World record of this species) and C. serrata; Plicatolamna cf. P. arcuata\ the
sawfish Ischyrhiza mira; and the ray Rhomb odus cf. R. binkhorsti. This
Peedee assemblage adds significantly to the still-limited roster of Late
Maestrichtian shark faunas of the world.

INTRODUCTION

The Peedee Formation is an Upper Cretaceous deposit of neritic
marine origin (Heron and Wheeler 1964) which spans the Tayloran and
Navarroan provincial stages (Brouwers and Hazel 1978), equivalent to
the Late Campanian and the entire Maestrichtian stages of Europe. At
the present time the deposit’s various phases of regression and deposition
are not named as members. This paper deals with the selachian fauna of
an “uppermost” member of the Peedee that is Navarroan (Late
Maestrichtian) in age. As shown below, the shark species of this member
correspond to those known Late Maestrichtian faunas found elsewhere in
the United States and other parts of the world.

The selachian fauna described here is distinct from that of the Black
Creek Formation (Miller 1967, 1968; Baird and Horner 1979), which un-
derlies the Peedee in North Carolina. A transition of several of the species
from the Black Creek up into the Peedee is normal, but other species ap-
pear in the Peedee that were not present in the Black Creek.

The marine vertebrate fauna of the Maestrichtian in North America is
poorly known. I have under study additional faunas of that age from
widely separated localities (in Maryland and Texas) that should, when
described, make our knowledge of the Late Maestrichtian much clearer.
One of these faunas, like that of the Peedee described here, contains only
megascopic species; the other contains a microfauna as well. Using this
undescribed material for comparison I endeavor here to place on record

Additions to the fossil vertebrate fauna of North Carolina, Part I.

Brimleyana No. 2: 77-89. November 1979

77

78

Gerard R. Case

for the first time a Late Maestrichtian/Navarroan shark fauna (incom-
plete as it is) from the state of North Carolina.

LOCALITIES

The sites from which the specimens in this report were recovered are all

in southern Duplin County. Site 1, by far the most productive, is a borrow

pit from which the county maintenance department extracts limestone for

use as road fill. It is located on County Road 1148 approximately 3.2 km

►

due east of its junction with the north-south U.S. Highway 117, about 2.4
km south of Rosehill. Its position is approximately 8.8 km diagonally
northwest of the Angola Bay Game Preserve and the Northwest Branch of
the Cape Fear River, at 34°52,30’' north latitude and 77°90,0” west
longitude. Fossils were first discovered at this locality by Mr. John C.
Golden, Jr., of North Charleston, South Carolina.

At Site 1 the Peedee Formation is exposed at a depth of approximately
4.5 to 6.1 m and consists of limestone with little traces of glauconite. The
lack of invertebrate steinkerns and foramnifera and the surficial wear on
the shark teeth strongly suggest that the fossils were reworked after
original deposition elsewhere and redeposited at their present location.
Unconformably underlying the Peedee is the Black Creek Formation,
which has yielded no fauna at Site 1 simply because the borrow pit has
not yet been dug deep enough. The overburden is a mixture of Castle
Hayne limestone (Claibornian, Upper Middle Eocene), Trent Marl
(Aquitanian, Early Miocene) and Duplin Marl (= Yorktown, Sahelian,
Late Miocene or more likely Early Pliocene) (Berry 1947; Richards 1950;
Cooke et al. 1943; Baum and Wheeler 1977). The shark faunas of the
Castle Hayne, Trent, and Duplin Marls at this locality will be described
in subsequent parts of this series.

Site 2 is the spoil heap of an irrigation pond alongside the Rosehill
Processing Plant (chicken Tenderers), approximately 2.4 km north of
Rosehill on U. S. Highway 117. Several species of the Peedee fauna are
quite common at this locality, although their condition is worn and
abraded, indicating reworking and subsequent redeposition. The over-
lying material at this site is also the Castle Hayne Limestone.

Site 3, the Superior Stone Quarry, is situated 3.2 km west of U. S.
Highway 117 at the Carroll intersection, approximately midway between
the towns of Warsaw and Magnolia. The predominant material recovered
from this now-defunct quarry is Castle Hayne Limestone with Late
Eocene shark teeth and echinoids. The underlying formation is the
Peedee, from which we obtained the same species as at Site 1. The ver-
tebrate remains also have been worn and abraded by wave action.

Cretaceous Selachians

79

SYSTEMATICS

Full synonymies for the species cited will be found in Cappetta and
Case (1975b) and Case (1978). The illustrated voucher specimens are
housed in the Museum of Natural History, Princeton University (PU);
additional specimens are in the author’s collection.

Class CHONDRICHTHYES
Subclass ELASMOBRANCHII
Clade HYBODONTIFORMES
Order SELACHII
Hybodus Agassiz, 1837
Hybodus sp. 1

Hybodus sp. 1 - Cappetta and Case, 1975:5, Fig. 2.

Material : — PU 22418, one lower lateral tooth lacking roots (Figs. 1-2).
Site 1.

Description : — An isolated central cusp showing little wear on the blade;
the root is missing as it is in most specimens. Plications on either side of
the blade, rising from the apron above the root separation to approx-
imately 1/5 the total height of the blade; plications more pronounced on
the labial face. The tooth faces inward and is slightly concave in lingual
aspect. It resembles a similarly preserved specimen from the Navesink
Formation (Middle Maestrichtian) of New Jersey (Cappetta and Case
1975b, Fig. 2), differing only in having a slight lateral apron.

Discussion : — Cappetta and Case described Hybodus sp. 1 on the basis of
incomplete cusps, with or without lateral cusplet traces. More recently
Case (1978) described as Hybodus montanensis a tooth from the Judith River
Formation (Campanian) of Montana, which shows strong, quite deep
plications occupying approximately 1/3 of the tooth cusps. This species
from the Western Interior is quite different from Hybodus sp. 1. The latter
will be named by Case on the basis of an almost complete specimen show-
ing a partial root and two lateral cusplets on one side, found in central
New Jersey in 1978.

Suborder ANACORACOIDEI
Family ANACORACIDAE
Squalicorax Whitley, 1939
Squalicorax kaupi (Agassiz)

Corax kaupi Agassiz, 1843:225, PI. 26a, Figs. 25-34; PI. 26, Figs. 4-8.

Material'. — PU 22419A, an upper antero-lateral tooth (Figs. 3-4), and
PU 22419B, a lower lateral tooth (Figs. 5-6); numerous additional teeth,

80

Gerard R. Case

anterior, lateral and posterior, from both upper and lower jaws. Sites 1
and 2.

Description: — Small-sized teeth, approximately 1 to 1.5 cm wide. The
teeth of Squalicorax kaupi are about 1/3 the size of those of Squalicorax
pristodontus . They show an indentation on the side of the blade directed
towards the commissure of the mouth, whereas the teeth of S. pristodontus
have a full, arcuate shape and no indentation of the blade. I assume that
the upper and lower teeth are quite similar, showing no heterodonty, so it
is difficult to be certain whether a given tooth comes from the upper or
lower jaw.

Squalicorax pristodontus (Agassiz)

Corax pristodontus Agassiz, 1843:224, PI. 26, Figs. 9-13.

Material : — PU 22420, a lower lateral tooth (Figs. 7-8). Site 1.

Description: — Teeth of fairly large size, averaging between 2 and 3 cm
wide. Large crescent-shaped blade with no indentation, as distinct from
S. kaupi. Roots, as with S. kaupi , are quite flattened, with or without an
apical foramen on the lingual face, and with or without fenestration or
multiple foramina on the upper portion of the root processes on the labial
face. Upper and lower teeth are similar, as with S. kaupi.

Discussion: — The teeth of S. pristodontus are quite distinctive. Although
some specimens have been recovered from formations as early as the Late
Campanian, the species is quite uncommon in pre-Maestrichtian out-
crops, and is by far more abundant in the Middle and Late Maestrichtian
stages of the Upper Cretaceous. Its geographic range is much like that of
S. kaupi. Squalicorax pristodontus is most abundant in the Middle Maestrich-
tian (Navesink Marl) of New Jersey, the Late Maestrichtian (Kemp Clay
Member, formerly the “Littig,” of the Navarro Group) in northeast and
central Texas, and the Late Maestrichtian (“ Couche trois” des phosphates) of
Morrocco, North Africa.

Clade EUSELACHIFORMES
Order EUSELACHII
Family MITSUKURINIDAE
Scapanorhynchus Woodward, 1889
Scapanorhynchus texanus (Roemer)

Lamna texana Roemer, 1852:29, PI. 1, Fig. 7.

Material: — PU 22421A, an upper lateral tooth (Figs. 9-10), and PU
22421 B, a lower anterior tooth (Figs. 11-12); about two dozen additional

Cretaceous Selachians

81

teeth, anterior, lateral and posterior, from both upper and lower jaws.
Sites 1 and 3.

Description. — Teeth of large size, averaging between 2 and 5 cm in
height from root tip to blade tip; upper and lower teeth are similar. The
anterior teeth of both jaws are elongate and sigmoidal, with or without
minute, sigmoidal lateral cusplets,with long striae on the lingual face of
the blade. The antero-lateral and lateral teeth are more flattened on their
labial and lingual faces and bear one or two lateral cusplets which differ
from those of the anterior teeth in being flattened and spade-like. Long
striae are seldom observed on the lingual face of the anterolateral, lateral,
or posterior teeth, but the beginning of a striation pattern can be observed
at the blade apron near the root on the labial face of well-preserved
specimens.

Discussion : — Scapanorhynchus texanus is a very common and ubiquitous
species in the Late Cretaceous of North America. It is most abundant in
the Late Campanian to Middle Maestrichtian of New Jersey and Texas,
and has recently been discovered in Late Campanian outcrops northeast
of Taroudannt, Morocco (Pierre Zennaro, pers. comm.).

Until recently the description of teeth from the various mouth positions
in Scapanorhynchus was quite haphazard, with numerous assignments to
various erroneous genera, causing quite a problem with synonymic
nomenclature in the literature. Capetta and Case (1975) compared the
fossil teeth of Scapanorhynchus with those in the jaws of the modern “goblin
shark,” Mitsukunna owstom. The teeth are basically the same, position for
position, except that many teeth of Mitsukunna are devoid of lateral ac-
cessory cusplets. Consequently Mitsukunna has been relegated to syn-
onomy and the living species should be cited as Scapanorhynchus owstom
(Jordan) (Bass et al. 1975:18). Now that the problem of differences be-
tween teeth from different positions has been cleared up, future authors
should be able to assign occurrences correctly to Scapanorhynchus and avoid
further synonymic confusion.

Scapanorhynchus texanus also occurs in the Black Creek Formation (Cam-
panian) at Phoebus Landing on the Cape Fear River, Bladen County,
North Carolina. The specimens, however, were mistakenly assigned to
the Tertiary genera Carchanas and Isurus by Miller (1967:223, Figs. 1-5).

Family ODONTASPIDAE
Odontaspis Agassiz, 1838
Odontaspis sp.

Material : — PU 22422, a lower antero-lateral tooth (Figs. 13-14); about
two dozen anterior and lateral teeth from both upper and lower jaws. Site 1.

82

Gerard R. Case

Figs. 1-26. Selachian teeth from the Peedee Formation of Site 1, Duplin County, North
Carolina, in labial and lingual views. Scale bar equals approximately 5 mm. 1-2, Hybodus sp.
1 (PU 22418); 3-4, Squalicorax kaupi, upper antero-lateral tooth (PU 224 19A); 5-6,

Squalicorax kaupi , lower lateral tooth (PU 2241 9B); 7-8, Squalicorax pnstodontus , lower lateral
tooth (PU 22420); 9-10, Scapanorhynchus texanus, upper lateral tooth (PU 22421A); 11-12,
Scapanorhynchus texanus , lower anterior tooth (PU 22421 B); 13-14, Odontaspis sp., lower
antero-lateral tooth (PU 22422); 15-16, Cretolamna appendiculata var. lata, upper lateral tooth
(PU 22423A); 17-18, Cretolamna appendiculata var. lata, lower anterior tooth (PU 22423B);
19-20, Cretolamna biaunculata, upper lateral tooth (PU 22424); 21-22, Cretolamna serrata,
lower antero-lateral tooth (PU 22425B); 23-24, Cretolamna serrata, upper lateral tooth (PU
22425A); 25-26, Plicatolamna cf. P. arcuata (PU 22426).

Cretaceous Selachians

83

Description : — Teeth comparable in size and superficial appearance to
those of the modern Odontaspis (Carcharias) taurus (Rafinesque). Erect cen-
tral blade with or without striae or plications upon either the lingual or
labial face; one or two lateral cusplets extending upward and outward
from the base of the blade near the root apron; roots robust in teeth of the
anterior and symphysial files, flattened in the lateral and posterior files.
Tooth cuspidate or sigmoidal depending upon its position in the jaw.

Discussion'. — Teeth of the Odontaspidae are quite difficult to sort and
separate, as the species are sometimes similar in characteristics. The
teeth from the Peedee Formation show no features of cusp design or root
structure that would serve to distinguish them from previously-known
species of Odontaspis , such as the Upper Cretaceous species 0. samhammen,
0. hardingi , and 0. holmdelensis described by Cappetta and Case from New
Jersey. Until more and better material becomes available, the North
Carolina teeth cannot be identified to species.

Family CRETOXYRHINIDAE
Cretolamna Glyckman, 1958
Cretolamna appendiculata lata (Agassiz)

Otodus latus Agassiz, 1843:271, PI. 32, Fig. 26.

Material'. — PU 22423A, an upper lateral tooth (Figs. 15-16), and PU
22423B, a lower anterior tooth (Figs. 17-18); two dozen anterior and
lateral teeth from both upper and lower jaws. Site 1.

Description : — Teeth as described in Cappetta and Case (1975b: 21).
The lower jaw teeth are more pointed and basically symmetrical, with
pinched roots, while the roots of the upper jaw teeth are more flattened
and their cusps are more squat. In the lateral to posterior files the blades
are more slanted towards the commissure. There are two constricted ac-
cessory cusps, triangular in shape (not sigmoidal as in the Odont-
aspidae), on the anterior lower teeth, while the accessory cusps on the up-
per teeth are much wider and flatter. Roots in the lingual aspect are with
or without an apical foramen, but all teeth lack a medial groove in that
area. On certain specimens, mostly lateral to posterior teeth, there are
vestiges of an additional and smaller accessory cusplet.

Discussion: — Cretolamna appendiculata lata is a most common species in
the Cretaceous of the Atlantic Coastal Plain, while it is rare in the Gulf
Coastal Plain (Texas in particular), and completely unknown in the
Western Interior where Plicatolamna takes its place as the dominant form.
It is common in the Maestrichtian of North Africa, especially in Morocco,
and is present but uncommon in European Cretaceous deposits. In the

84

Gerard R. Case

Cretaceous of New Jersey and Delaware C. a. lata ranks fourth in order of
abundance, after Scapanorhynchus texanus, Squalicorax (kaupi and
pristodontus ), and Plicatolamna arcuata.

Cretolamna biauriculata (Zittel in Wanner)

Otodus biauriculata (Zittel in MS) Wanner, 1902:148, PL XIX, Fig. 28.

Material : — PU 22424, an upper lateral tooth (Figs. 19-20); five other
teeth, anterior and lateral. Site 1.

Description : — The teeth are basically similar to those of C. a. lata but
differ conspicuously in bearing two divergent lateral cusps on either side
of the central cusp, the outer ones being the smaller. The roots are flat-
tened, with or without apical foramina, and lack a medial groove or
furrow such as is found in the Odontaspidae.

Discussion : — This is the first reported occurrence of C. biauriculata in the
New World. Its teeth are quite common in the Maestrichtian of Morocco,
a bit scarcer in Algeria and Tunisia, and quite rare in Israel, where Raab
(1963:27, PI. 1, Figs. 1-18) claims to recognize C. b. ?maroccana Aram-
bourg. It is extremely rare in European deposits.

Cretolamna serrata (Agassiz)

Otodus serratus Agassiz, 1838, PI. XXXII, Figs. 27-28

Material : — PU 22425A, an upper lateral tooth (Figs. 23-24), and PU
22425B, a lower antero-lateral tooth (Figs. 21-22); about two dozen other
teeth, anterior and lateral, from both upper and lower jaws. Site 1.

Description : — As described by Cappetta and Case (1975b), the typical
tooth has two lateral cusps on one side of the central cusp and one lateral
cusp on the other side. Otherwise the teeth are much like those of
Cretolamna appendiculata lata.

Discussion : — Teeth of C. serrata are found only in Late Maestrichtian
deposits. In North America the species ranges through the Atlantic and
Gulf Coastal Plains, but has not turned up in the Western Interior nor
west of the Rocky Mountains. It is quite common in North Africa
(Morocco).

Plicatolamna Herman, 1974
Plicatolamna cf. P. arcuata (Woodward)

Lamna arcuata Woodward, 1894:198, PI. VI, Fig. 10

Material : — PU 22426, a lower anterior tooth (Figs. 25-26). Site 1.

Description : — Tooth of average size, about 1 cm high and 8 mm max-

Cretaceous Selachians

85

imum width (although the right lateral cusp and a small portion of the
root are missing). It bears a sigmoidal lateral cusplet alongside a slender,
slightly sigmoidal central cusp.

Discussion : — The teeth of P. arcuata are quite common in beds of Late
Campanian to Middle Maesjxichtian age but quite uncommon in the
Late Maestrichtian. This specimen is not definitely identifiable as P.
arcuata , and its temporal occurrence is very late for that species.

Order BATOIDEA
Suborder GANOPRISTINIDEA
Family SCLERORHYNCHIDAE
Ischyrhiza Leidy, 1856
Ischyrhiza mira Leidy
Ischyrhiza mira Leidy, 1856:221.

Material : — PU 22427A-B, two rostral teeth (Figs. 27-30); three other
rostral teeth (no oral teeth). Site 1.

Description : — Rostral teeth of rather large size, ranging in length from
3.5 to 5.5 cm. No barb present on the lateral edge of the enameloid cusp
(as in Onchopristis) and no plications or striations on the enameloid faces.
The root base bears strongly rugose pleats which probably strengthened
the attachment to the rostral cartilage. Otherwise the rostral teeth are-
characteristically sclerorhynchoid.

Discussion : — Rostral teeth of the sawfish Ischyrhiza are common in Up-
per Cretaceous deposits of North America, especially in the range from
Late Campanian to Middle Maestrichtian. This genus is unknown out-
side the continental limits of the United States, for, as noted in previous
work (Case 1967), Sclerorhynchus replaces Ischyrhiza in the Upper
Cretaceous deposits of Europe and North Africa. Originally the rostral
teeth of Ischyrhiza were thought to be the oral teeth of a teleost fish (Leidy
1856; Fowler 1911) but were later identified as belonging to a primitive
sawfish that is known as a ganopristine because its denticles have a com-
bination of enameloid and osseous structure.

Suborder DASYATOIDEI
Family MYLIOBATIDAE
Rhombodus Dames, 1881
Rhombodus cf. R. binkhorsti Dames
Rhombodus binkhorsti Dames, 1881:1, Fig. 1.

Material : — PU 22428, a pavement tooth (Figs. 31-35); and a dozen ad-
ditional specimens. Sites 1, 2 and 3.

86

Gerard R. Case

Figs. 27-35. Batoid teeth from the Peedee Formation of Site 1, Duplin County, North
Carolina. Scale bar equals approximately 5 mm. 27-30, Ischyrhiza mu a , two rostral teeth in
dorsal and ventral views (PU 22427A-B); 31-35, Rhombodus cf. R. binkhorsti, pavement tooth
in occlusal, anterior, lateral, basal, and posterior views (PU 22428).

Cretaceous Selachians

87

Description : — Tooth-plates rhombic in outline, with short crowns con-
taining enameloid material and with large, bifurcated roots that comprise
approximately 2/3 of the total height. The crown has no rugose striation
on its occlusal surface (such as occurs in Myledaphus ), and its outer edge is
crimped or milled, no doubt to facilitate interlocking between adjacent
units in the tooth-pavement. Tooth-plates are of medium size, averaging
less than 1 cm in diameter.

Discussion : — Because of their worn state the teeth from the Peedee For-
mation cannot be definitely assigned to the species Rhombodus binkhorsti.
Despite its superficial similarity to Myledaphus bipartitus Cope, Rhombodus
appears to be more closely related to the myliobatid Hypolophus than to
the dasyatids, the skate family.

The isolated crushing teeth of Rhombodus are quite common in the Late
Maestrichtian and are found in all marine deposits of that age, especially
in Morocco, and in Texas, where they are second in number to the teeth
of Cretolamna s errata.

FAUNAL COMPARISONS

The Late Maestrichtian stage of the Upper Cretaceous system is a most
interesting period, and its vertebrate fauna is not very well known. Some
of the shark species persisted into the Paleocene while others became ex-
tinct.

Table 1, based on data from nine states, shows the stratigraphic dis-
tribution of the shark species known from the Peedee Formation of North
Carolina. This chart, of course, is not complete, but will require modifica-
tion as further discoveries are made. On the basis of current information
the shark species of the Peedee are characteristic of Late Maestrichtian
time although some are seen to occur as early as the Cenomanian.

Considering the Late Maestrichtian shark fauna as a whole, the follow-
ing forms are known to be of earlier origin: Heterodontus, Lonchidion,
Pseudocorax, Paranomotodon, Squalus, Ginglymo stoma, Brachaelurus , Mesiteia,
Rhinobatos, Sclerorhynchus, Ischyrhiza avomcola, Ischyrhiza mira, Ptychotrygon,
Cretorectolobus , and Protoplatyrhina. Appearing for the first time in the Late
Maestrichtian are Notorhynchus, Paleogaleus, Scyliorhinus, Dasyatis, Raja , and
Rhombodus , along with Cretolamna biauriculata and C. serrata. The only forms
that survived into the Paleocene are Odontaspis, Cretolamna appendiculata,
Rhombodus, Squalus, Ginglymo stoma, and Rhinobatos.

ACKNOWLEDGMENTS. — I would like to thank Mr. & Mrs. John C.
Golden, Jr., of North Charleston, South Carolina for their assistance, es-
pecially their generosity in donating several of the specimens in this
study. Thanks go out to Mr. E. Dewey Ross of Mesic, North Carolina, for

88

Gerard R. Case

Table 1. Stratigraphic distribution of species in the United States Upper Cretaceous Formations*

Stages

Species**

Cen

Tur

Con

San

Camp

Early

Mas

Mid

Mas

Late

Mas

Hybodus sp. # 1 .

X

X

X

Squalicorax kaupi

X

X

X

X

X

X

X

Squalicorax prisiodontus

X

X

X

X

Scapanorhynchus texanus

X

X

X

X

Odontaspis sp.

X

X

X

X

X

X

X

X

Cretolamna appendiculata

X

X

X

X

X

X

X

X

Cretolamna biauriculata

X

Cretolamna serrata

X

Plicatolamna sp.

X

X

X

X

Ischyrhiza mira

X

X

X

X

Rhombodus sp.

X

*Data from Upper Cretaceous faunal assemblages from the following states: New Jersey; Maryland;
North Carolina; Alabama; Arkansas; Texas; Montana; Wyoming and California.

**Known species recovered from Rosehill, North Carolina.
Cen = Cenomanian San = Santonian

Tur = Turonian Camp = Campanian

Con = Coniacian Mas = Maestrichtian

his assistance in the field. My appreciation goes to Dr. Donald Baird,
Department of Geological and Geophysical Sciences, Princeton Univer-
sity, for his kindness in reading the original manuscript and offering
suggestions for the improvement of the paper. The excellent photography
of the figures is the work of Mr. Richard E. Grant of Dallas, Texas. Ap-
preciation is extended to Mr. Richard Hamilton, Rockaway, New York,
for his kind assistance on the art preparation for the figures and tables.

This work was supported in part by a grant from the Griffis Fund of the
American Littoral Society: GF-ALS-1812. Additional assistance came
from the Scott Fund of Princeton University.

LITERATURE CITED

Arambourg, Camille. 1952. Les vertebres fossiles des gisements de phosphates
(Maroc-Algerie-Tunisie). Notes Mem. Div. Mines Geol. Maroc. #92. 372 pp.
Baird, Donald, andj. R. Horner. 1979. Cretaceous dinosaurs of North Carolina.
Brimleyana 2:1-28.

Cretaceous Selachians

89

Bass, A. J., J. D. D’Aubrey and N. Kistnasamy. 1975. Sharks of the east coast of
southern Africa: Pt. IV: The families Odontaspidae, Scapanorhynchidae,
Isuridae, Cetorhinidae, Orectolobidae and Rhiniodontidae. Oceanogr. Res.
Inst., (Durban) Invest. Rep. #39. 102 pp.

Baum, Gerald R., and W. H. Wheeler. 1977. Cetaceans from the St. Marys and
Yorktown Formations, Surry County, Virginia. J. Paleontol. 5/(3) :492-504.
Berry, E. Willard. 1947. Marls and limestones of Eastern North Carolina. Bull.

#54 N. C. Dep. Conserv. Dev., Raleigh. 16 pp.

Brouwers, Elisabeth M., and J. E. Hazel. 1978. Ostracoda and correlation of the
Severn Fm. (Navarroan: Maestrichtian) of Maryland. Soc. Econ. Paleontol.
Mineral., Paleontol. Monogr. 1. 52 pp.

Cappetta, Henri, and G. R. Case. 1975a. Selaciens nouveaux de Cretace du
Texas. Geobios (Lyon) 5(4):303-307.

and 1975b. Contribution a 1 ’etude des selaciens de Groupe Monmouth

(Campanien-Maestrichtien) du New Jersey. Palaeontogr. Abt. Palaeozool-
Stratigr. (1 5 1 ) : 1 -46.

Case, Gerard R. 1967. Fossil shark and fish remains of North America. Priv.
publ., New York. 20 pp.

1973. Fossil sharks: a pictorial review. Priv. publ., New York.

64 pp.

1978. New selachian fauna from the Judith River Formation (Campanian)

of Montana. Palaeontographica, A (160) : 1 76-205.

Cooke, C. Wythe, J. Gardner and W. P. Woodring. 1943. Correlation of the
Cenozoic Formations of the Atlantic and Gulf coastal plain and the Caribbean
region. Geol. Soc. Am. Bull. 54:1713-1723.

Emmons, Ebenezer. 1858. Agriculture of the eastern counties with descriptions of
the fossils of the marl beds. N. C. Geol. Surv. Rep., Raleigh. 314 pp.
Fowler, Henry W. 1911. A description of the fossil fish remains of the Cretaceous,
Eocene and Miocene Formations of New Jersey. Geol. Surv. N. J. Bull. #4. 192

pp.

Heron, S. Duncan, Jr., and W. H. Wheeler. 1964. The Cretaceous formations
along the Cape Fear River, North Carolina. Atl. Coastal Plain Geol. Assoc.,
5th Annu. Field Conf. Guidebook. 55 pp.

Miller, Halsey W. 1966. Cretaceous vertebrate fauna from the Phoebus Landing,
North Carolina. J. Elisha Mitchell Sci. Soc. 82( 2):1.

1967. Cretaceous vertebrates from Phoebus Landing, North Carolina. Proc.

Acad. Nat. Sci. Phila. 119{ 5):219-235.

1968. Additions to the Upper Cretaceous vertebrate fauna of Phoebus

Landing, North Carolina. J. Elisha Mitchell Sci. Soc. 84( 4):467-471.

Raab, M. 1963. Fossil fish and reptiles from Late Campanian phosphatic deposits
of the Negev region of Israel. Isr. J. Earth Sci. 12( l):26-40.

Richards, Horace G. 1950. Geology of the Coastal Plain of North Carolina.

Trans. Am. Philos. Soc., N.S. 40( 1 ) : 1 -83.

Stephenson, Lloyd W., P. B. King and R. W. Imlay. 1942. Correlation of the out-
cropping Cretaceous Formations of the Atlantic and Gulf Coastal Plain and
Trans-Pecos Texas. Geol. Soc. Am. Bull. 53: 435-448.

Accepted 29 May 1979

Freshwater Triclads (Turbellaria) of North America.

XII. Another New Cave Planarian from North Carolina,

Phagocata carolinensis n. sp.

Roman Kenk

m

Department of Invertebrate oology ,

National Museum of Natural History, Smithsonian Institution

Washington, D. C. 20560

ABSTRACT. — Phagocata carolinesis n. sp. from One Bat Cave, Burke
County, North Carolina, is a slender, unpigmented species with two
small eyes and is characterized by its peculiar penial anatomy and by
the development of an enormous vagina with special histological dif-
ferentiation. It belongs to a group of closely related species of Phagocata
distributed in the Appalachian region.

Through the courtesy of Dr. Cato O. Holler, Jr. of Old Fort, North
Carolina, I received some specimens of a new species of Phagocata , collec-
ted in a cave in Burke County, North Carolina. Dr. Holler is with the
North Carolina Cave Survey and an investigator in the Biological Survey
of North Carolina Caves.

Phagocata carolinensis , new species

Type-material (deposited in the U. S. National Museum of Natural
History [USNM], Smithsonian Institution, Washington, D. C.). — Holo-
type, set of sagittal sections on 3 slides (USNM 58431); paratypes, sagit-
tal and horizontal sections of 2 specimens on 12 slides (USNM 58432-
58433).

External features (Figs. 1A, IB). — The species is unpigmented (white),
externally indistinguishable from Phagocata angusta Kenk (1977) or P.
holleri Kenk (1979). In gliding locomotion the body is very slender,
reaching up to 13 mm in length and about 1 mm in width. The truncate
head has a somewhat wavy frontal margin and bears on each side a
lateral projection, slightly more distinct than that of P. holleri. Posterior to
these projections the head narrows, forming a kind of neck. Behind the
neck the lateral body margins diverge again gradually to reach the max-
imum width, then run parallel for the greater part of the body length, to
converge again and to meet at the rather rounded posterior end.

The two small eyes are situated close together at a considerable dis-
tance behind the frontal margin of the head. The intestine reaches

Brimleyana No. 2: 91-96. November 1979

91

92

Roman Kenk

Fig. 1. Phagocata carolinensis. A, photograph of living animal, X7; B, outline drawing of liv-
ing animal, with indication of eyes, pharynx, and copulatory apparatus.

anteriorly to the level of the eyes. The pharnyx is rather long, about 1/5
body length, and its root is inserted at the beginning of the fourth fifth of
the body. The copulatory apparatus is visible in the living specimen as a
transparent area occupying the anterior half of the postpharyngeal
region.

Anatomy. — The two eyes are rather small, the diameters of their pig-
ment cups measuring 26-37 /im.

In the reproductive system, the two ovaries are situated behind the
second or third lateral branches of the anterior intestinal trunk. Each
ovary is equipped with a rather large, lobed parovarium, usually located
laterally to the ovary. There are, however, cell accumulations also an-
terior to the ovaries that histologically resemble parovaria without being
connected with either the ovaries or the true parovaria.

The numerous testes are essentially ventral, although at full maturity
individual testicular follicles may extend almost to the dorsal body wall.
They are arranged on either side of the body in a broad longitudinal zone
beginning some distance behind the ovaries and terminating at the level
of the pharyngeal root. In the region of the testes, the thin vas deferens or
sperm duct on each side runs along the medial side of the ventral nerve
cord, somewhat above the longitudinal layer of the ventral integumental

New Cave Planarian

93

Fig. 2. Phagocata carolinensis, semidiagrammatic view of copulatory apparatus in sagittal
sectiop. ac, common atrium; am , male atrium; b, copulatory bursa; bd , bursal duct; bp, penis
bulb; gp, gonopore; la, lamellae; m, mouth; ode, common oviduct; ph, pharyngeal pouch; pp,
penis papilla; va, vagina; vd, vas deferens.

muscles. In the region of the pharynx, the vasa deferentia expand to form
the tortuous false seminal vesicles or spermiductal vesicles, filled with
sperm, which proceed posteriorly toward the copulatory complex.

The copulatory apparatus (Fig. 2) is located immediately behind the
pharyngeal pouch {ph). The genital aperture or gonopore {gp) is situated
about midway between the mouth {m) and the posterior end of the body.
It leads through a short narrow duct into the common atrium {ac) which
opens anteriorly into the male atrium {am) and posteriorly into the vagina
{va) of the bursal duct. Both the male and common atria are lined with a
nucleate and ciliated cuboidal epithelium.

The penis consists of a rather voluminous bulb {bp), the musculature of
which is rather feeble and developed mainly near its periphery, and a
short, bluntly conical papilla {pp) extending into the male atrium. The
configuration of the penial lumen is quite unique. Three to five transverse
lamellae {la) originate from the dorsal and ventral walls, alternatingly
projecting downward and upward. Each lamella is attached to the lateral
walls of the penis, as can be seen in horizontal sections through the organ.

94

Roman Kenk

Fig. 3. Photomicrographs of copulatory organs in sagittal section. A, Phagocata carolinensis,
X60; B, Phagocata holleri, X127.

This arrangement would permit the passage of sperm from the bulb to the
tip of the penis papilla only in a serpentine fashion. It is possible,
however, that the configuration of the penis lumen as it appears in the
slides obtains only when the penis is retracted. When the papilla is ex-
tended during copulation the passage may well straighten out. Both the
lamellae and the greater part of the outer surface of the papilla are
covered by a thin infranucleate epithelium with two underlying muscle
layers, a circular and a longitudinal one. There is no differentiation of the
penial lumen into a seminal vesicle and ejaculatory duct. At most, a small
anterior portion of the lumen may differ from the remaining part by hav-
ing a ciliated and nucleate epithelium. The vasa deferentia (vd) enter the
penis bulb ventrolaterally, turn dorsomedially, and unite at their opening
into the anterior end of the penial lumen.

The copulatory bursa (b) adjoins the wall of the pharyngeal pouch
(ph). Its outlet, the bursal stalk or duct (bd) proceeds posteriorly for a
short distance as a narrow duct situated somewhat to the left of the
midline and lined with a nucleate, ciliated epithelium. It then expands
into a large, irregularly lobed sac, the vagina (va), the greater part of

New Cave Planarian

95

which is also located on the left side. The epithelium lining the vagina
consists of large, apparently apocrine cells with basal nuclei (Fig. 3A),
quite different from those of the anterior part of the bursal duct. It ap-
pears that one of the functions of the vagina is the resorption of super-
fluous sperm after copulation, a process that may take place at many
locations in the turbellarians (cf. particularly Cernosvitov 1931). Sperm
was present in the vagina in two of the three specimens examined. The
vagina connects with the posterior end of the common atrium (ac). At the
transition, the nature of the epithelial lining changes from the vaginal
glandular epithelium to the nonglandular, ciliated epithelium of the
atrium.

The two oviducts or ovovitelloducts unite in the space between the bur-
sal stalk and the atria and form a common oviduct {ode) that opens from
the dorsal side near the junction of the male and common atria.

Distribution and ecology. — Four specimens were collected on 7 April 1979
by Dr. Cato O. Holler, Jr., Christopher Holler, and Keith Barnes in a
small seep toward the rear end of One Bat Cave, located in the Linville
Gorge Wilderness Area, Burke County, North Carolina. They were sent
to me alive. Three of the specimens proved to be sexually mature.

Taxonomic position. — Phagocata carolinensis belongs to a group of closely
related species of the genus distributed in the Appalachian region. Exter-
nally it cannot be distinguished from two other species, P. angusta and P.
holleri. The three have in common the lack of body pigment, a slender
shape, a truncate head end without prominent auricular projections, a
pair of very small eyes, and anatomically prepharyngeal ventral testes.
The principal differences are in the configuration of the copulatory com-
plex. The penial lumen in P. carolinensis shows the peculiar transverse
lamellae (at least in the retracted penis), while in P. holleri it contains
finger-shaped villi (compare Figures 3A and 3B) and in P. angusta no
special internal differentiations. Phagocata carolinensis has an extraor-
dinarily large lobate vagina with a modified histology; P. angusta , a villous
vagina of moderate size; and P. holleri , no distinguishable vaginal
formation. In P. holleri the copulatory apparatus is situated considerably
posterior to the pharyngeal pouch; in P. carolinensis the apparatus adjoins
the pouch closely; while P. angusta has an intermediate position. Details in
the histological differentiation of the epithelia of the copulatory complex
also differ in the three species.

Etymology. — The species is named carolinensis after its habitat, North
Carolina.

96

Roman Kenk

ACKNOWLEDGMENTS. — Thanks are due to Dr. Cato O. Holler, Jr.
for his generous collaboration in collecting and shipping the flatworms.
Dr. John C. Harshbarger of the Smithsonian Institution was helpful in
the preparation of the photomicrographs, and Dr. Marian H. Pettibone
kindly reviewed the manuscript.

LITERATURE CITED

Cernosvitov, L. 1931. Studien iiber Spermaresorption. III. Die Samenresorption
bei den Tricladen. Zool. Jahrb. Abt. Anat. Ontog. Tiere 54: 295-332, pi. 5-7.
Kenk, R. 1977. Freshwater triclads (Turbellaria) of North America. X. Three
new species of Phagocata from the eastern United States. Proc. Biol. Soc. Wash.
59:645-652.

1979. Freshwater triclads (Turbellaria) of North America. XI. Phagocata

hollen , new species, from a cave in North Carolina. Proc. Biol. Soc. Wash.
92( 2):389-393.

Accepted 30 August 1979

Freshwater Fishes of Croatan National Forest,

North Carolina, with Comments on the Zoogeography

of Coastal Plain Fishes

Fred C. Rohde

Department of £ 'oology , University of North Carolina,

CEapel Hill, North Carolina 27514

- George H. Burgess
Florida State Museum, University of Florida,

Gainesville, Florida 32611
and

G. William Link, Jr.

Institute of Marine Sciences, University of North Carolina,

Morehead City, North Carolina 28557

ABSTRACT. — In a survey of freshwater fishes in and near Croatan
National Forest, eastern North Carolina, 94 collections at 51 localities
yielded 5670 specimens representing 38 species. Sixteen additional
species have been reported from the Forest, but at least four are con-
sidered questionable. Twenty species of estuarine fishes were collected
during the study, including the first North Carolina record of the
Fourspine stickleback, Apeltes quadracus.

Examination of the distribution of fishes in the nine major river
systems draining the lower Coastal Plain of North Carolina reveals the
absence of a number of species in the smaller drainages (e.g. Shallotte,
New, White Oak and Newport rivers). Two possible reasons are lack of
suitable ecological conditions and the irregular distribution patterns ob-
served at the periphery of a species’ range.

INTRODUCTION

The fishes of the lower Coastal Plain of North Carolina have not been
as intensively sampled as those that occur in upland areas of the state,
primarily because of apparent lower species diversity and high frequency
of underwater obstructions. We initiated the survey on which this paper
is based in 1973 to determine what freshwater fishes occur in the Croatan
National Forest in eastern North Carolina. Observations on brackish
water and marine fishes were also included.

Several prior fish surveys made in the general vicinity of the Croatan
National Forest were either very localized (Bayless 1966; Turner and
Johnson 1973), or covered a wide area of the state (Bayless and Smith
1962; Keup and Bayless 1964; Davis and McCoy 1965). Seehorn (1976)

Brimleyana No. 2: 97-118. November 1979

97

98

Fred C. Rohde, George H. Burgess, G. William Link, Jr.

Fig. 1. Map of Croatan National Forest showing sampling stations. Heavy line depicts
forest boundary. Circles denote stations in this study; squares represent collections made by
Duke University Marine Laboratory (DUML) personnel or earlier workers at University of
North Carolina - Institute of Marine Sciences (UNC); triangles indicate collections made by
North Carolina Wildlife Resources Commission (NCWRC).

attempted to list all fishes known to occur or which might occur in
southeastern national forests, but his Croatan list was based almost solely
on a literature search (R. E. Jenkins, pers. comm.). We attempted to
sample a large number of freshwater localities in the Croatan National
Forest and adjacent streams in order to assess accurately the abundance
and distribution of its fishes.

STUDY AREA

Croatan National Forest comprises 382,716 hectares and occupies
parts of Carteret, Craven, and Jones counties (Fig. 1). It is drained by the
Neuse, White Oak, and Newport rivers; a fourth river within the forest,
the Trent, is tributary to the Neuse. Five natural lakes totaling 10,617

Croatan National Forest Fishes

99

hectares are present there. Most of the land not forested is in agricultural
use. There is little domestic or industrial pollution and, because the land
is generally level, there is little evidence of soil erosion.

Water was tannin stained and acidic, as is typical of the Coastal Plain,
and pH values of 4.6 - 9.0 have been recorded in the Neuse River within
the forest, 5.1 - 8.2 in the White Oak, 4.9 - 8.1 in the Newport, and 4.5 -
4.7 in the lakes (Bayless and Smith 1962; Davis and McCoy 1965; Bay-
less 1966). Water temperatures in our study ranged from 3.0 C in January
to 31.0 C in June, and the waters were usually clear.

Stream-side vegetation was typical for the Coastal Plain. Twenty
aquatic plant species were observed in the area: filamentous algae;
Sphagnum sp.; Taxodium distichum ; Typha sp.; Sparganium amencanum;
Potamogeton pectinatus ; JVajas sp.; Ruppia maritima; Sagittana sp.; Scirpus sp.;
Pontedena cor data; Juncus sp.; Saururus cernuus; Ceratophyllum demersum;
Polygonum punctatum; Nymphaea odorata ; Proserpinaca sp.; Myriophyllum
brasiliense; Hydrocotyle sp.; and Utnculana purpurea.

MATERIALS AND METHODS

From 3 September 1973 to 16 September 1978, 94 collections were
made at 51 localities in and near the forest (Fig. 1). All specimens were
collected either with a 3.0 m X 1.2 m, 3.1 mm mesh flat seine; a 6.1 m X
1.2 m, 6.4 mm mesh flat seine (used several times); or a 30.5 m gill net of
vahying mesh sizes (used only once). No attempt was made to standardize
the collection methods used because of variations in physiography at the
different sites. Fish were immediately preserved in 10% formalin and
returned to the laboratory for identification. Representative specimens
were deposited in the Institute of Marine Sciences - University of North
Carolina (UNC) collection. Water and air temperatures and pertinent
physical characteristics of each site were recorded. Although two sites
were sampled intensively as part of several life history studies, most
localities were sampled only once.

Data from collections made at 31 localities (Fig. 1) by the North
Carolina Wildlife Resources Commission (NCWRC) (Bayless and Smith
1962; Davis and McCoy 1965; Bayless 1966) were summarized and are
included. These collections were made with 5% emulsified rotenone. Data
from seine and gill-net collections made at 17 localities (Fig. 1) by earlier
workers at UNC-IMS and Duke University Marine Laboratory (DUML)
are also included. Additional records were obtained from the fish collec-
tions at the North Carolina State Museum of Natural History (NCSM)
and Duke University (DU). All localities at which collections were made
are listed in Table 1.

100 Fred C. Rohde, George H. Burgess, G. William Link, Jr.

Table 1. Sampling localities and dates (day/month/year). Numbers
refer to our stations, letters to those of others (see text).

I. Neuse River drainage

A. Craven County

I. Long Creek on SR 1700, 7.8 km ne North Harlowe,
13/10/73.

2. Bachelor Creek on SR 1224, 14.7 km w New Bern, 31/12/74.

3. Southwest Prong of Slocum Creek on SR 1746, 2.7 km wnw
Havelock, 4/1/75.

4. Hancock Creek on dirt road s off NC 101, 5.1 km se
Havelock, 1/6/75.

5. Cahooque Creek on SR 1716, 7.7 km ne Havelock, 1/6/75.

6. South Fork of Cahooque Creek on SR 1716, 7.7 km ne
Havelock, 1/6/75.

7. Tucker Creek on dirt road w of US 70, 7.5 km ene Havelock,
21/6/75.

8. East Prong of Brice Creek on SR 1100, 3.2 km sw Croatan,
21/6/75, 31/1/76, 27/2/76, 12/3/76, 2/4/76, 30/1/77,
23/2/77, 23/4/77, 14/5/77, 16/9/78.

9. Tributary to Brice Creek on Forest Road 121-2, 7.2 km w
Croatan, 21/6/75.

10. Drainage ditches on SR 1100 at jet. Forest Road 121-2, 4.8
km sw Croatan, 21/6/75, 9/4/77.

II. Tributary to Brice Creek on Forest Road 170, 8.8 km nw
Croatan, 31/1/76.

12. Tributary to Brice Creek on Forest Road 170, 8.8 km nnw
Croatan, 31/1/76.

13. Neuse River at Cherry Point at end of SR 1716, 11.0 km ne
Havelock, 18/5/76

14. Neuse River at Flanner Beach at end of SR 1107, 2.4 km ne
Croatan, 18/5/76, 16/9/78.

15. Reedy Branch on SR 1004, 11.0 km ene Pollocksville,
31/1/76.

16. Reedy Branch on SR 1340, 11.7 km se Pollocksville,
31/1/76.

17. Tucker Creek on dirt road e of US 70, 5.0 km se Croatan,
17/7/77.

A. Little Lake, 7.4 km wsw Havelock, 14/4/52, 29/4/52.

B. Ellis Lake, 8.5 km sw Havelock, 24/4/52, 29/8/57.

C. Tributary to Hancock Creek on NC 101, 3.7 km e Havelock,
14/4/52.

Croatan National Forest Fishes

101

D. Drainage ditch tributary to outlet Ellis Lake, 7.8 km sw
Flavelock, 14/4/52.

E. Neuse River at North Harlowe, 29/8/57.

F. Drainage ditch along US 70, 3.2 km se Havelock, 29/8/57,
14/4/62.

G. See #14, 19/4/56.

H. See #8, 2/5/71, 7/10/72.

E Long Creek on SR 1700, 7.8 km ne North Harlowe,
13/10/73.

J. Cahooque Creek at junction Hancock Creek, 6.7 km ne
Havelock, 20/5/76, 24/8/76.

B. Jones County

18. Mill Creek on SR 1108, 3.2 km s Pollocksville, 22/8/74,
7/12/74.

19. Island Creek on SR 1004, 8.2 km ene Pollocksville, 22/8/74.

20. Tributary to Trent River on SR 1337, 1.4 km nw
Pollocksville, 7/12/74.

21. Mill Run on NC 58, 5.4 km w Pollocksville, 7/12/74,
12/3/76, 28/8/76.

22. Mill Creek on SR 1004, 0.8 km se Pollocksville, 28/8/76.

23. Tributary to Trent River on SR 1340, 9.9 km ne
Pollocksville, 14/6/77, 19/11/77.

II. White Oak River drainage

A. Carteret County

24. Pettiford Creek on Forest Road 128, 8.0 km ne Cape Car-
teret, 1/1/75.

25. Unnamed creek on SR 1101, 0.5 km e Stella, 1/1/75.

26. Hadnot Creek on NC 58, 8.8 km nnw Cape Carteret,
8/6/76.

27. Hadnot Creek on SR 1104, 8.8 km n Cape Carteret, 8/6/76.

28. Drainage ditch on Forest Road 176, 12.8 km n Cape Car-
teret, 8/6/76.

29. Hunters Creek on NC 58, 13.6 km nnw Cape Carteret,
8/6/76.

K. White Oak River at Stella, 3/3/56.

L. White Oak River on NC 24 at Swansboro, 17/8/71.

M. See #24, 31/7/72.

B. Craven County

30. Great Lake, sw corner at end Forest Road 126, 17.3 km se
Maysville, 16/8/76.

31. Drainage ditch on SR 1100, 10.5 km e Maysville, 30/1/77.

102

Fred C. Rohde, George H. Burgess, G. William Link, Jr.

C. Jones County

32. Black Swamp Creek on NC 58, 4.8 km se Maysville,
7/12/74.

33. Drainage ditch along SR 1105, 9.8 km e Maysville,
31/12/74, 4/1/75, 16/8/75, 30/1/77, 23/2/77, 12/3/77,
19/3/77, 27/3/77, 3/4/77, 9/4/77, 16/4/77, 23/4/77,
30/4/77, 14/5/77, 14/6/77, 17/7/77, 14/8/77, 18/9/77,
16/10/77, 19/11/77.

34. Tributary to Black Swamp Creek on SR 1105, 6.2 km ese
Maysville, 4/1/75.

35. Drainage ditch on SR 1105, 200 m w Black Swamp Creek,
9.9 km e Maysville, 16/8/75.

36. Black Swamp Creek on SR 1105, 10.0 km e Maysville,
16/8/75, 9/4/77.

37. Holston Creek on NC 58, 10.1 km se Maysville, 8/6/76.

38. White Oak River on US 17, 2.2 km s Maysville, 28/8/76,
14/5/77.

39. White Oak River on SR 1118, 3.2 km nw Maysville,
28/8/76.

40. White Oak River, 0.5 km sw Haywood Boat Landing, 10.7
km se Maysville, 14/5/77.

41. White Oak River at Haywood Boat Landing, 10.2 km se
Maysville, 14/5/77.

N. See #33, 18/7/59.

D. Onslow County

42. Starkys Creek on SR 1434, 13.4 km nw Swansboro, 7/12/74.

43. Webb Creek on SR 1435, 9.0 km nw Swansboro, 7/12/74.

III. Newport River drainage
A. Carteret County

44. Mill Pond and below spillway in Black Creek on SR 1154,
3.3 km e Newport, 3/9/73, 1/6/75.

45. Northwest Prong of Newport River on SR 1 124, 3.2 km wnw
Newport, 22/8/74, 30/12/74, 4/1/75, 10/5/75, 16/9/78.

46. Drainage ditch on SR 1125, 7.4 km s Havelock, 1/1/75,
1/4/75, 5/4/77.

47. Unnamed tributary on US 70, 1.6 km s Newport, 1/3/75.

48. Shoe Branch on SR 1124, 1.3 km wnw Newport, 1/3/75.

49. Little Deep Creek on SR 1139, 2.9 km ne Newport, 1/3/75.

50. Deep Creek on SR 1133, 3.7 km n Newport, 1/3/75.

51. Pond, 0.5 km n of SR 1124, 1.3 km nw Newport, 1/3/75.

O. See #44, 17/8/56.

Croatan National Forest Fishes

103

P. See #45, 31/7/72, 7/6/73.

Q. Southwest Prong of Newport River on SR 1124, 6.9 km w
Newport, 7/6/73.

RESULTS

Freshwater Fishes

The following is a list of freshwater fishes collected in (or near) or
reported from the forest, with comments on dubious records and records
of interest. Table 2 provides a list of species by drainage and collector.

Lampetra aepyptera , Least brook lamprey

2 specimens, station 21 . Menhinick et al. (1974) listed four localities for
this species in North Carolina, one of which was station 21. Although
three trips were made specifically to collect this lamprey, only two am-
mocoetes were captured. We also visited two of the remaining three
localities and took lampreys (8 ammocoetes) at only the Pitt County site.
The habitat at these sites was marginal for lampreys. Bailey et al. (1977)
listed this species as of Special Concern in North Carolina. Because of
lack of suitable habitat at the few known localities we feel that this species
should be considered Threatened in the state.

Petromyzon mannus , Sea lamprey

One ammocoete (155 mm TL) was collected at station 21 on 21 Oc-
tober 1969 by J. R. Bailey, Duke University (DU uncat.). Two am-
mocoetes of L. aepyptera were taken concurrently.

Lepisosteus osseus, Longnose gar

4 specimens, stations 27, 38, 44. K, J, NCWRC.

Anna calva, Bowfin

NCWRC. Local fishermen reported this species from small ponds near
Newport. Turner and Johnson (1973) collected it at their uppermost sta-
tion in the Newport River.

Anguilla rostrata , American eel

237 specimens, stations 3-8, 16, 19, 21-27, 29, 36-39, 41, 42, 44, 45, 47-
50. B, C, D, P, NCWRC.

Dorosoma cepedianum , Gizzard shad
1 specimen, station 38. J, NCWRC.

Umbra pygmaea, Eastern mudminnow

188 specimens, stations 5-7, 8, 10, 28, 31, 33, 35, 36, 44, 46. N,
NCWRC.

104

Fred C. Rohde, George H. Burgess, G. William Link, Jr.

.Q

rtf

C

o

(->

rt5

u

o

Z

C/3

<u

S_

o

Uh

rtj

c

o

-t— >

rti

Z

c

rtj

rtj

o

u

U

t-<

rtS

oj

c

"O

c

k5

C

C/3

(-1

<u

_Q

E

3

C

Iv.

rtj

CJ

_o

"O

3

rt3

► ^ C/3

/ < <u

N — - C/3

__ CJ

cj "H

C

CJ

u

rt!

a,

CJ

JU

o

cj

c/3 C

<u

C/3

z £

cJ3 CJ
i_ _Q

£ e

rt3 ZS

Z C

-C

C/3

cj .3

U-I

rt5

C-4-.

U

o

O

C/3

hJ

u

V

JO

>,

E

T3

3

3

-t— >

C

C4)

T3

rt5
+— *

G

O

03

■M

C/3

V

bC

_4J

’o

_C

<U

O-,

Ut

C/3

CO

JJ

-Q

rt

h

S-I

CJ

>

5

C

o

Oh

£

<u

z

u,

<D

>

Z

rt3

O

CJ

4-»

'Jg

(->

CJ

_>

£

CJ

1/3

3

aj

z

, H-I

p Q

U

Z

£

U

z

C/3 Z'

'3 ^

~ C/3

, hJ

D Q

U

z

£

u

z

.2 -a

£ 3

H -tJ

~ C/3

. Z

D Q

U

Ph

£

u

z

3

+->

C/3

CO Tj-

O O

OO LO

xt-

vO vO

C> >p— I

CM Tf

LO LO

co oo
co co

1/3

C/3 ^

CJ O
• ^ * ,"H

.ti u

13 z
o o

-3 a
<-*— « C-»— '
o o

6 6

z z

C/I

aj

• r-H

U

<V

a,

(73

CO

GO rO cO C- cO

LO

co

cO CO

VO

o"

LO

LO

CO

LO

O CO VO o o

co^

CO

CO

OO

LO OO CO

3”

vO CO

LO

CO

LO

o

CO •'fr-

/-vO V~N

LC^ T-H_ U-3^

cO cO t'-~

1—1 TH

2

<*>

^3

35

<3

•s,

c o

a

£

<3

V

■as

V

2

s

<3

V3

3

S3

a -2
a ^

■Cc,

Vi

<3

<3

V

25

g a

'Ll <T

35 -2

5 t

^ s

•S

2 a
-c> ><

.5 °

<-3

25

•<«»

2

25

35

<-3

V. .25

•Sp §

S3

cO

cO

00 CO

r- co

CO

LO T— I
LO

vo CO

CO ' — 1

O CO
LO

CO LO

cO

co

CO ^

VO CN
cO I—1

cO CO
CO

vO CO

O LO
T— 1 CO

25

Vi

2

2

<2>

■

•gs

Co

K

3i C-3

Vl

3S

v*J

35

Vi

35

<o

V>

25

•s

35

3

2 -2
35 35

■2 JU

'Ll

35

-C3

Z'

**0

S3

2

■

* \

S3

*

^0

35

'O

<o

25

S3

33

•J3

33

»v ^

^ l=lo <J3 . va/3 ?3 sv: a g ^ a A U

§D | f § R 2 ^-1 I -§ ^ “S I ^ g

35

<5

■s

<o ~c;

Croatan National Forest Fishes

CM

CM

(M

CM

vO

cm oj OJ r-

LO LO Cs|

vo vo co cm lo m-

co

(M

VO

CO T-H CVI <0

O' 00 rH h-

CM T— I

u~> M" CM
T-C OO

M" ' — 1 M" CM CM cO

vO 00 r- CO O' CM

vO t''~ lO

fO

<N vO

O cO Cv cO O

00

LO vo CM

OO M O

vO vO

o-

oo

f-

LO t"~ vO vO

CO M-
CM

CM

vO

cO
M" T-H

o"

C" LO
O- -r-H

CM cO

r cm co

cO

CO

lO

Cv

vO

OO

CO cO cO

t vO r"~
CM 00

CO

Cv

•<-1 CM M" t— i
cv h-~ cm

' OO

CM

CM CO

CO

cO CM
rn OO

CM Cv

CO

o? ^ oT 'O'

M- CM CM

M" CM cO

CM

CO CVI

M" cO

cO

M" lo CM

LO OO Cv vO O'

— ' o

CM

cO lO cO
CM f-

O

o

LO

r- r- o m- co

vo LO cO

LO LO O VO LO CM M-
r-» M" CM

O.

<0

a

**so>

2

so

<o

So

a

_ r
•si ^

.a

a

<o

a

S

a

-c>

a

<0

a

■S

Vi

a

a

to

a

a

'£

Sri

a

a

a

a

a

Vi

K

LO

o

o

a

«o

a

a

a

a

s-2 ^

a1

a

-a

. ^

O ^

2 2
O Vj

a

* *>4

R

: <si

~a

~a

a

V.

a "X ~a «o

«-> .tvo 8 .2 -3

g s Is

a a s ^3
-a © -a a

■ .a , • -c; ■«_ a . vj

a

a

^ 3

a -a
•S I

‘o a
a

-a •tj

S a

Vi

a

a

a

a

«o

a

-a

2 g

2 2

**si to

R <3

ko “s»^i

CJ kj

<-o
O
* «**»

o

<o

a

-R

<o

R

<o

R O

c§ ^

<-0

a

So
* c««i

a

R

<o

a

O

-R

-R

2

* c»o

<s

R

r

<0

- S
fcuD

S§

s

2

a

<o

LO .

a <o

.§ t-
2

Lo

ta

sj kj ^4 ^

Co

a £

•*<a V-

a g
~a

Vi

a

I

5 i |

• «s» SO

S |

a
-a
a
a
a

2

a

«0
. <o

l*> a
a a
a

2
2

^ a

^ 3 5

ki

105

Cv

o

cO

vO Cv
OJ VO

o-

o-

CO

cO

O VO
co T— I
CM
CM

00

LO

co

LO LO
CO OO
vO
CM

c «

c

QJ

L0 C

<u C

•y 'u
c y
cu cu

c/3 cn

-*-* -t— i

o o
H H

106

Fred C. Rohde, George H. Burgess, G. William Link, Jr.

Esox americanus americanus, Redfin pickerel

100 specimens, stations 4, 6-10, 17, 18, 21, 22, 26, 27, 33, 35-37, 45-47,
49, 50. C, F, M, P, NCWRC.

E. mger, Chain pickerel

21 specimens, stations 1, 6, 8, 23, 27, 37. B, I, J, NCWRC.

Carassius auratus , Goldfish

B. Local fishermen reported it as present in a small pond near
Newport.

Cypnnus carpio , Carp

NCWRC. One specimen taken in the Newport River.

Hybognathus regius, Silvery minnow

36 specimens, stations 18, 23. NCWRC. The status of this form as a full
species and its relationships with other Hybognathus species requires
further investigation.

Notemigonus crysoleucas, Golden shiner

69 specimens, stations 6, 7, 8, 18, 25, 27, 32, 38, 40, 44, 45, 50. P,
NCWRC.

JVotropis altipinnis, Highfin shiner

Reported by Seehorn (1976) from the forest, but R. E. Jenkins (pers.
comm.) says it was included by mistake. Hubbs and Raney (1948) re-
corded it from a tributary on the north shore of the Trent River.

N. amoenus , Comely shiner

NCWRC. Taken only in Trent River and tributaries. Representative
specimens are in the NCSM fish collection.

N. bifrenatus , Bridle shiner

NCWRC. Three Tucker Creek specimens represent the southernmost
record for the species (Jenkins and Zorach 1970). We made collections at
two separate sites without success. Station 17 is close to the NCWRC site,
but we encountered high salinities (8 ppt) which had apparently forced
most freshwater fishes upstream. Bailey et al. (1977) said that N. bifrenatus
is probably extirpated at this locality.

N. chalybaeus , Ironcolor shiner

176 specimens, stations 18, 22, 27, 32, 38, 39, 45. C, P, NCWRC.

N. cummingsae, Dusky shiner

47 specimens, stations 16, 18, 19, 32. NCWRC.

Croatan National Forest Fishes

107

JV. hudsomus , Spottail shiner

3 specimens, station 23. NCWRC. Found only in Trent River and
tributaries.

JV. procne , Swallowtail shiner

NCWRC. Taken only in Trent River and tributaries. Representative
specimens are in NCSM collection.

Enmyzon oblongus oblongus , Creek chubsucker

70 specimens, stations 3, 6, 7, 19, 21, 23, 29, 38, 45, 46. J, NCWRC.

E. sucetta , Lake chubsucker

138 specimens, stations 8, 10, 33. H, NCWRC. We encountered some
difficulty in identifying certain Croatan specimens. Hanley (1976) stated
that E. oblongus oblongus and E. sucetta hybridize in the drainage ditches of
the forest probably as a result of habitat alteration.

Moxostoma spp., Redhorses

NCWRC tentatively recorded M. collapsum ( = anisurum) from a Trent
River tributary and M. papillosum from the White Oak River. R. E.
Jenkins (pers. comm.) believes that either M. amsurum or M. macro-
lepidotum may be present in Forest waters, as both have been collected
from the lower Neuse in the vicinity of New Bern. We did not collect
either species.

Ictalurus catus , White catfish
K, NCWRC.

/. natahs, Yellow bullhead

92 specimens, stations 7, 8, 10, 30, 32, 33, 36, 38, 45-47. B, H, P,
NCWRC.

I. nebulosus, Brown bullhead
J, NCWRC.

I. punctatus, Channel catfish

H, NCWRC. Possibly introduced to Neuse drainage (Jenkins et al.
1972).

Noturus gynnus, Tadpole madtom

15 specimens, stations 29, 38, 45, 49, 50. NCWRC.

JV. insigms , Margined madtom

23 specimens, stations 21, 40, 45. NCWRC.

Chologaster cornuta , Swampfish

803 specimens, stations 8, 10, 24, 33-35, 37, 45-47. N, NCWRC. Most

108

Fred C. Rohde, George H. Burgess, G. William Link, Jr.

specimens were collected at station 33 as part of an investigation into the
life history of this species.

Aphredoderus sayanus sayanus , Pirateperch

476 specimens, stations 3-8, 20-25, 27, 29, 32, 33, 35-39, 41, 42, 44-47,
49, 50. P, NCWRC.

Fundulus diaphanus diaphanus , Banded killifish

6 specimens, station 23. Taken only in a tributary of the Trent River.

F. lineolatus , Lined topminnow

237 specimens, stations 8, 10. H, NCWRC. Found only in drainage
ditches along Catfish Lake Road (SR 1100), where it is abundant. Wiley
(1977) recently revised the Fundulus notti species group. He considered F.
lineolatus a full species and resurrected the name F. escambiae for the
Florida panhandle populations. However, Ralph Yerger (pers. comm.)
has found what he believes to be intergradation of the two forms in the
median portions of several panhandle systems. Additional study of this
complex problem is certainly required.

Gambusia affims holbrooki , Mosquitofish

1129 specimens, stations 1, 5, 6, 8, 10, 11, 17, 18, 21, 23, 25-27, 30, 38,
44, 45, 47, 49, 50. A, B, C, E, H, NCWRC.

Acantharchus pomotis , Mud sunfish

41 specimens, stations 4, 8, 33, 34, 45-47, 50. NCWRC.

Centrarchus macropterus, Flier

136 specimens, stations 4, 7-9, 15, 18, 32, 35, 36, 45, 47. M, Q,
NCWRC.

Elassoma zonatum, Banded pygmy sunfish

NCWRC. One specimen collected at a site on the Trent River.

Enneacanthus chaetodon , Blackbanded sunfish

Sweeney (1972) reported the Blackbanded sunfish from a locality near
the forest, but indicated that the record is probably erroneous because
numerous efforts by collectors in this region have failed to collect the
species. Seehorn (1976) and Jenkins et al. (1975) perpetuated this possi-
ble error. The closest known records, according to Menhinick (ms), are in
Duplin, Lenoir, and Pender counties.

E. glonosus , Bluespotted sunfish

735 specimens, stations 1, 3-10, 15, 17, 21, 23-25, 27-31, 33, 35-39, 41,
44-47, 50. H, M, P, NCWRC.

Croatan National Forest Fishes

109

E. obesus , Banded sunfish

126 specimens, stations 8, 10, 31, 33, 35, 36, 45, 46. A, C, D, P,
NCWRC.

Lepomis auntus , Redbreast sunfish

22 specimens, stations 18, 21, 38, 44, 45, 49, 50. P, NCWRC.

L. gibbosus, Pumpkinseed

70 specimens, stations 5-8, 17, 18, 20, 21, 23, 25, 29, 38, 41, 45, 48. J, P,
NCWRC.

L. gulosus, Warmouth

89 specimens, stations 3, 7, 8, 10, 30, 37, 38, 44, 45. B, C, D, Q,
NCWRC.

L. macrochirus, Bluegill

172 specimens, stations 8, 18, 21, 25, 26, 38, 43-45, 50. B, C, D, Q,
NCWRC. Possibly introduced to the Neuse system (Jenkins et al. 1972).

L. marginatus, Dollar sunfish

Although reported by Seehorn (1976) as occurring in the forest, there
are no records to substantiate this. Closest verified localities are in north-
ern Craven County.

L. microlophus , Redear sunfish

NCWRC. Stocked by the NCWRC in the White Oak River.

Micropterus salmoides salmoides , Largemouth bass

24 specimens, stations 1, 23, 25, 32, 37, 38, 44, 45, 50. I, J, O, P,
NCWRC.

Pomoxis mgromaculatus , Black crappie

8 specimens, stations 8, 44. NCWRC. Possibly introduced into New-
port drainage (see Discussion).

Etheostoma fusiforme fusiforme , Swamp darter

38 specimens, stations 3, 6, 7, 21, 38, 44. B, NCWRC.

E. olmstedi olmstedi , Tessellated darter

176 specimens, stations 3, 16, 18, 21, 22, 38, 39, 45, 48. P, NCWRC.

E. sernferum , Sawcheek darter

152 specimens, stations 6-8, 22, 33, 38, 39, 44-47, 50. P, NCWRC.
Perea flavescens , Yellow perch

4 specimens, stations 8, 36. A, NCWRC. Possibly introduced to Neuse
system (Jenkins et al. 1972).

110

Fred C. Rohde, George H. Burgess, G. William Link, Jr.

Percina crassa roanoka , Piedmont darter

4 specimens, station 21. This species, rarely encountered on the lower
Coastal Plain, was collected over gravel in swifter portions of the creek.
Station 21 represents the easternmost known locality in North Carolina.

Estuarine Fishes

Twenty estuarine fishes were taken within the forest (Table 3). Most
were sporadic visitors during periods of increased salinity. One item of
note is the capture of a Fourspine stickleback, Apeltes quadracus, at station
23 on 14 June 1977, the first published record of a stickleback in North
Carolina.

Table 3. Species, numbers (N) and station numbers of estuarine fishes collected
in (and near) Croatan National Forest, North Carolina, in this study.

Species

N

Station

Alosa aestivalis

21

23

Brevoortia tyrannus

2

14

Anchoa mitchilli

3

17

Fundulus heteroclitus

1

14

Lucama parva

17

17

Membras martinica

31

13, 14

Memdia beryllina

48

13, 14, 17,23

Memdia memdia

103

13,14

Apeltes quadracus

1

23

Syngnathus fuscus

1

1

Morone amencana

3

14

Eucinostomus argenteus

12

17,26

Lagodon rhomboides

31

1, 13, 14, 17,26

Bairdiella chrysura

13

17

Leiostomus xanthurus

72

13, 14, 17

M ugil cephalus

5

14, 17

Mugil curema

4

14

Dormitator maculatus

2

23, 26

Gobiosoma bosci

5

17

Tnnectes maculatus

11

25, 27, 38,41,

44, 45, 48

Croatan National Forest Fishes

111

DISCUSSION

Fifty-one species of freshwater fishes representing 16 families have been
identified as occurring in or near the Croatan National Forest (Table 2,
plus Petromyzon mannus). Four additional species ( Notropis altipinnis, En-
neacanthus chaetodon, Lepomis marginatus, L. punctatus ) possibly occur there.
Sunfishes (Centrarchidae), well suited to the slow-moving, vegetated
waters of the Croatan National Forest, are the most speciose family with
12 species. Minnows (Cyprinidae) are also well represented with ten
species, but none is common.

Although major sections of the White Oak and Newport rivers are con-
tained in the forest, their faunas do not contribute as much to the overall
diversity as does the fauna of the Neuse River, which drains only a very
small part of the overall area. Forty-six species were collected in the
Neuse drainage, including 12 not taken in the Newport or White Oak
rivers. The Trent River, a tributary of the Neuse, forms the northern
boundary of the forest and accounts for much of the diversity. Eight of the
12 species confined to the Neuse drainage were taken only in the Trent or
its tributaries. The Newport and White Oak rivers had 33 and 34 species
respectively, of which only 3 were unique to one or the other system.

Our examination of fishes in the three Croatan National Forest
drainages prompted a review of the distributions of lowland (lower
Coastal Plain) fishes throughout North Carolina. Most distributional
data were obtained from regional faunal summaries (Jenkins et al. 1972;
Jenkins et al. 1975; Menhinick, ms) and systematic revisions (e.g.
Collette 1962; Snelson 1968; Yerger and Relyea 1968; Sweeney 1972),
but some unpublished records are based on collections of ours. Nine ma-
jor rivers drain into coastal North Carolina estuaries, including one (the
Cape Fear) that is composed of two systems on the Coastal Plain (Fig 2).
Some (e.g. the Shallotte, New, White Oak, and Newport rivers) are small
drainages confined to the lower reaches of the Coastal Plain, whereas
others (e.g. the Cape Fear, Neuse, Tar, and Roanoke rivers) are more ex-
tensive and originate in the Piedmont or the Appalachian Mountains.
The following discussion concerns fishes of those rivers lying totally or in
part below the Fall Line in North Carolina.

Nineteen species reach the northern termini of their distribution ranges
on the Coastal Plain of North Carolina or in nearby southeastern Virginia
(Table 4, plus Micropterus salmoides salmoides and Etheostoma serriferum). The
greatest faunal break occurs between the Cape Fear and the more
northern drainages, with eight terminations in the combined Cape Fear-
Northeast Cape Fear systems (see Jenkins et al. 1972 for additional dis-
cussion of this break). Five species terminate in the Chowan River in
North Carolina and Virginia (Jenkins et al. 1975). Only four fishes

1 12 Fred C. Rohde, George H. Burgess, G. William Link, Jr.

Fig. 2. Coastal drainages of eastern North Carolina lying below the Fall Line. Surry Scarp
drawn from White (1966).

( Lampetra aepyptera, Notropis bifrenatus, Etheostoma vitreum, Percina peltata
nevisense ) have southern termini in the Coastal Plain drainages. Noturus
funosus is confined to the Neuse and Tar drainages (Taylor 1969), but
may have been present in the Roanoke prior to habitat destruction by
dams and pollution (C. R. Gilbert, pers. comm.). Memdia extensa and
Etheostoma perlongum are endemic to Lake Waccamaw, and Fundulus wacca-
mensis is confined to lakes Waccamaw and Phelps (see zoogeographic dis-
cussions in Hubbs and Raney 1946; Jenkins et al. 1972; and Bailey et al.
1977). Phelps Lake F. waccamensis have not been critically examined, and
introduction from Lake Waccamaw is considered a possibility.

Twenty-four native species are uniformly distributed throughout
coastal North Carolina drainages: Lepisosteus osseus, Dorosoma cepedianum,
Anguilla rostrata, Umbra pygmaea, Esox americanus americanus, E. mger,
Notemigonus crysoleucas, Notropis chalybaeus, Enmygon oblongus oblongus, 1c-
talurus natahs, Noturus gynnus, N. insignis, Aphredoderus sayanus sayanus, Gam-
busia affinis holbrooki, Acantharchus pomotis, Centrarchus macropterus, En~

Croatan National Forest Fishes

113

neacanthus glonosus, Lepomis auritus, L. gibbosus, L. gulosus, L. macrochirus,
Micropterus salmoides salmoides , Etheostoma olmstedi olmstedi, and E. serriferum.
Of these, only M. s. salmoides and E. serriferum terminate in the Chowan
(Jenkins et al. 1975); all others extend farther northward. No southern
terminations are involved.

Using available distributional information, 41 species are known to be
absent from one or more Coastal Plain drainages in North Carolina.
Based on habitat preferences and distributional patterns of the species in-
volved, we believe that 13 of these will eventually appear in collections
from those drainages where they are now unknown (Table 4). Three
others (Elassoma evergladei, Notropis petersoni, N. cummingsae) have northern
termini in North Carolina and are not expected in more northerly
drainages. The remaining 25 species have what appears to be real dis-
tributional gaps. These may be further separated into two groups: a
lowland contingent composed of 5 species of sunfishes, and a larger group
of 20 species with predominantly upland affinities (Table 4). All are miss-
ing from the New, White Oak, and Newport rivers, and all but Lepomis
marginatus and L. punctatus are also absent from the Shallotte or Northeast
Cape Fear, or both.

The absence of the 5 lowland species of sunfishes ( Lepomis marginatus, L.
punctatus, Elassoma zonatum, Enneacanthus chaetodon, Pomoxis nigromaculatus)
from the New, White Oak, and Newport rivers is enigmatic. Since the
first three species reach terminations in nearby northern drainages
(Table 4), it is possible that their absence in these intermediate rivers
merely reflects the irregular distribution pattern often observed at the
periphery of a species’ range. Enneacanthus chaetodon is often absent from
habitats for which it is seemingly well-suited and its distribution is spotty
throughout most of its range (Jenkins et al. 1975; Burgess et al. 1977).
Enneacanthus chaetodon is likely to be discovered in one or more of these
drainages with further collecting. We have two records of Pomoxis
nigromaculatus from the Newport drainage, but since both are from a single
pond (Mill Pond) owned by a fishing club, we suspect introduction.
Although P. nigromaculatus is a common lowland form in much of North
Carolina (Table 4), its apparent absence in these three coastal drainages
may prove real. <

The remaining 20 species’ distributions are perhaps more easily ex-
plained. Most species involved are forms that prefer swifter flowing, up-
land streams, habitats which are scarce or absent in the New, White Oak,
and Newport drainages. The 30 meter high stand in sea level (repre-
sented today by the Wicomico Shoreline or Surry Scarp, Fig. 2) that oc-
curred in the Pleistocene may have influenced the distribution of these
fishes. The area occupied by the present-day Shallotte, New, White Oak,

114

Fred C. Rohde, George H. Burgess, G. William Link, Jr.

£

O

•

4— >

03

c

q-

X

13

O

X

13

~a

<u

+->

u

<u

a,

x

u

<u W
<u
c/2

13

bD

c

03

s_

03
£

£
o3 (*-1

'o °

m

-£ £

-C

.2 £

S-,

£ O
. £ ■*-'

JS II
CL,

o3

m -6
03

O

O

O

o3

£

o

n

u

13

— _Q
O o3
a C
03 5

U -2

c/3

_c <u
£
CT

a

O

2

u

JZ V

4— > n

w T3

C/3 ^0

*-> o

g £

£ u
D T3
■ — ' 13

D

~o

13

£
13

£

JS 3

qj O
in "O

M X

£

_o

4— >

£

_Q

u,

4— >

C/)

5

13

JD

o3

H

C/3

13

'a

o

bn

13

-I— '

o3

l)

SJU3UIUI03

"O

1)

U

£

2

j>S

3

•

C/3

C/3

O

a

3

o

13

£

03

fa

3{jein3q|y

-UEMOg^

OOIJUIEJ-JEJ^

9Sn3^[

iJocLwa^

TO 3;iqM

M9^

< -

^^xxxxxxxxx

XXXXXX«XX«w

^xxxxxx«xxx«

uyuuuuuuuuu^

nniGnnnnnrSrSn^

XXXXXXXX^Xw

XwXXXXXXwww

^XwwwwXX^XXX

JE3J ^d^Q

jsB3g;jo^[ ^^XXXXXXXX^XXX

JH3d 3d^0 ^xxxxxxxxxxxxx^

3U°nBqs wwXXXXXXXWhwhX

1/3

X

o

H

£

«

hH

es

H

1/3

Q

C/3

P

o
£

X v.

H 55

X §

50 a

o

a

C-3

a

a

C-5

S ^3

a

a

•'-j

a

i)

a

5

K

^ g

a ^

a

a a
~

<-3

3 §

■S ^

2 §

2 .1
a
a

a

a

a

a

8

<o

e |

o S ^ 53 ^ ^ eg

<0

<0

O

o' ^

k* O-5

; s ^ Lo

<0 0-5

^3 S
a a

^ -2

O

C-)

a

C-5

-a>

a

*2

a

a

a

v** ^

<-3

a

a ^

~a

■*<»

t-3

a u.

a

a

§ 2

a

a

a

y

a a

a

a

so

2

<->

a

a

a

O

<-5

a

"t^s

a

a

§

8
a

ra ;c 2

^ *3 ^

<o

1 2 ^

a «

,a .<o> .<-3

-Cl -Cl

S S £

-O v®

«N ^ ^

DISJUNCT DISTRIBUTIONS

Croatan National Forest Fishes

115

"O

<u

u

3

"O

O

_Q

O

s->

Oh

X X

^ X X

^ X X X

^ X X X X

X

a

o

8

8

o

K.

o c

?>.

Dh -Q

2: I

o a*

c

o

(U

c

H-J

(3

Uh

<u

>

o

-Q

<

rSrSAi^rSM'Ni^iii'S^AnrS

^ x”w XXXXXXXXX XX

CN

XS/SZ^S^^S^S/S/S/S^S/S/SZ^'v*

<N

rS m n rS rS r% rS rS rN ry r^i ^N *N *N

X X X X X ^ X

X X X X X ^^^^XX XX XXXXXXXXX

SZ1 S/ SZ< kZ

rS rS n rN

116

Fred C. Rohde, George H. Burgess, G. William Link, Jr.

and Newport rivers and most of the Northeast Cape Fear River was inun-
dated by encroaching seas; colonization by stream fishes began as the sea
regressed and rivers developed. Most species with preferences for slow-
moving, lowland waters were probably able to enter these newly develop-
ing systems, but others with upland affinities undoubtedly found the
ecological conditions unsuitable and never became established.

The distribution of Fundulus diaphanus diaphanus , a secondary freshwater
species (Myers 1938, 1951) that tolerates brackish waters, is not ex-
plainable by lack of upland habitat. Flowever, F. d. diaphanus prefers
gravel or sand bottoms and avoids areas of heavy siltation (Shapiro 1947),
and most lowland North Carolina rivers are heavily silted. In addition,
the species is at the southern periphery of its range in North Carolina,
and records are few south of the Neuse River. Ictalurus punctatus is similarly
distributed, but the extensive introductions of this species throughout its
range prevents assigning any significance to the pattern.

The diversity of fishes in the Croatan National Forest is quite high
compared to studies made on other Coastal Plain regions. Jenkins et al.
(1975) found 25 species in the Dismal Swamp and 43 species in the
Chowan system on the Coastal Plain. Both areas are considerably larger
than Croatan National Forest.

At present there is little evidence of degradation of the Forest waters,
and with proper management practices these waters should continue to
support a diverse ichthyofauna.

ACKNOWLEDGMENTS.— We would like to thank R. G. Arndt, J. E.
Cooper, C. R. Gilbert, D. S. Lee, S. W. Ross, and F. J. Schwartz for their
helpful comments for improvement of the manuscript. Many people
assisted with the collecting; we especially thank D. E. Fast and S. W.
Ross. J. G. Lundberg and students assisted with several collections. A. F.
Chestnut and F. J. Schwartz of the Institute of Marine Sciences - Univer-
sity of North Carolina provided equipment and vehicles. J. B. Sullivan of
Duke University Marine Laboratory and J. R. Bailey of Duke University
provided information on collections in their care. W. M. Palmer allowed
access to the fish collections at the North Carolina State Museum of
Natural History. E. F. Menhinick graciously permitted us to review his
unpublished distribution maps of North Carolina fishes.

Most of this work was carried out while the authors were at the In-
stitute of Marine Sciences - University of North Carolina.

Croatan National Forest Fishes

117

LITERATURE CITED

Bailey, Joseph R. and Committee. 1977. Freshwater fishes, pp. 265-298 in J. E.
Cooper, S. S. Robinson, and J. B. Funderburg (eds.). Endangered and
Threatened Plants and Animals of North Carolina. N. C. State Mus. Nat.
Hist., Raleigh, xvi + 444 pp.

Bayless, Jack D. 1966. Coastal lakes I. N. C. Wildl. Resour. Com., Raleigh. 11

pp.

, and W. B. Smith. 1962. Survey and classification of the Neuse River and

tributaries, North Carolina (Appendix separate). N. C. Wildl. Resour. Com.,
Raleigh. 33 pp.

Burgess, George H., C. R.. Gilbert, V. Guillory and D. C. Taphorn. 1977. Dis-
tributional notes on some north Florida freshwater fishes. Fla. Sci. 4(9(1) : 33-4 1 .
Collette, Bruce B. 1962. The swamp darters of the subgenus Hololepis (Pisces, Per-
cidae). Tulane Stud. Zool. 0:115-211.

Davis, James R., and E. G. McCoy. 1965. Survey and classification of the New -
White Oak - Newport rivers and tributaries, North Carolina (Appendix
separate). N. C. Wildl. Resour. Com., Raleigh. 19 pp.

Hanley, Robert W. 1976. Population phenetics of chubsuckers in North Carolina
( Enmyzon : Catostomidae). Masters thesis, Duke Univ., Durham. 182 pp.
Hubbs, Carl L., and E. C. Raney. 1946. Endemic fish fauna of Lake Waccamaw.
Misc. Publ. Mus. Zool. Univ. Mich. 65: 1-30.

, and 1948. Subspecies of Notropis altipinnis, a cyprinid fish of the eastern

United States. Occas. Pap. Mus. Zool. Univ. Mich. 5(9(5:1-20.

Jenkins, Robert E., E. A. Lachner and F. J. Schwartz. 1972. Fishes of the central
Appalachian drainages: their distribution and dispersal, pp. 43-117 in P. C.
Holt (ed.). The distributional history of the biota of the southern Ap-
palachians, Part III: Vertebrates. Res. Div. Monogr. 4, Va. Polytech. Inst.
State Univ., Blacksburg. 306 pp.

, L. A. Revelle and T. Zorach. 1975. Records of the blackbanded sunfish,

Enneacanthus chaetodon , and comments on the southeastern Virginia freshwater
ichthyofauna. Va. J. Sci. 26(3): 128-1 34.

, and T. Zorach. 1970. Zoogeography and characters of the American

cyprinid fish Notropis bifrenatus. Chesapeake Sci. / 7(3) : 174- 1 82.

Keup, Lowell, and J. Bayless. 1964. Fish distribution at varying salinities in
Neuse River basin, North Carolina. Chesapeake Sci. 5(3) :1 19-123.
Menhinick, Edward F., T. M. Burton and J. R. Bailey. 1974. An annotated
checklist of the freshwater fishes of North Carolina. J. Elisha Mitchell Sci. Soc.
96(l):24-50.

Myers, George S. 1938. Fresh-water fishes and West Indian zoogeography.
Smithson. Inst. Ann. Rep. 1937:339-364.

1951. Fresh-water fishes and East Indian zoogeography. Stanford Ichthyol.

Bull. 4(1):1 1-21.

Seehorn, Monte E. 1976. Fishes of southeastern national forests. Proc. 29th Annu.

Conf. Southeast. Assoc. Game Fish Comm.: 10-27.

Shapiro, Sidney. 1947. Geographic variation in Fundulus diaphanus , a cyprinodon-
tid fish. Ph.D. dissert., Univ. Michigan, Ann Arbor. 137 pp.

Snelson, Franklin F. 1968. Systematics of the cyprinid fish, Notropis amoenus , with
comments on the subgenus Notropis. Copeia 1 968(4) :776-802.

Sweeney, Edward F. 1972. The systematics and distribution of the centrarchid
fish tribe Enneacanthini. Ph.D. dissert., Boston Univ., Boston. 205 pp.

118

Fred C. Rohde, George H. Burgess, G. William Link, Jr.

Taylor, William R. 1969. A revision of the catfish genus Noturus Rafinesque with
an analysis of higher groups in the Ictaluridae. U. S. Natl. Mus. Bull. 282. 315

pp-

Turner, William R., and G. N. Johnson. 1973. Distribution and relative abun-
dance of fishes in Newport River, North Carolina. NOAA Tech. Rep. NMFS -
666. .23 pp.

White, William A. 1966. Drainage asymmetry and the Carolina capes. Geol. Soc.
Am. Bull. 77:223-240.

Wiley, E. O. 1977. The phylogeny and systematics of the Fundulus notti species
group (Teleostei: Cyprinodontidae). Occas. Pap. Mus. Nat. Hist. Univ. Kans.
66. 30 pp.

Yerger, Ralph W., and K. Relyea. 1968. The flat-headed bullheads (Pisces: Ic-
taluridae) of the southeastern United States and a new species of Ictalurus from
the Gulf Coast. Copeia 1968(2):36 1-384.

Accepted 31 May 1979

Genetic Variation in Cave-dwelling and Deep-sea Organisms,
with Emphasis on Crangonyx antennatus (Crustacea:

Amphipoda) in Virginia.

Gary W. Dickson1 and John C. Patton
Department of £ oology ,

University of Georgia, Athens, Georgia 30602

John R. Holsinger
Department of Biological Sciences,

Old Dominion University, Norfolk, Virginia 23508

AND

John C. Avise
Department of £ oology ,

University of Georgia, Athens, Georgia 30602

ABSTRACT. — Genetic variation was analyzed through electrophoretic
techniques in six populations of the troglobitic (i.e. obligatory cave-
dwelling) amphipod Crangonyx antennatus from Lee County, Virginia.
From the results of this investigation and those tabulated from previous
studies on a number of cave-dwelling species, genetic variability does
not appear to be substantially reduced in populations inhabiting sub-
terranean environments. The origin of normal levels of genetic
variability in cave-dwelling species may differ from those organisms in-
habiting another relatively stable environment, the deep-sea. The high
levels of genetic variability recorded in many deep-sea invertebrates are
thought to be due in part to the presence of large populations of these
species. In contrast, the small population sizes observed in cave-dwelling
organisms may allow species to expand their niches with an associated
increase in genetic variability.

INTRODUCTION

Similarities in the physical environments of caves and the deep-sea
have led to the suggestion that similar selection pressures may be
operating on their respective faunas (Poulson 1971). Caves and the deep-
sea share several important characteristics including absence of light,
reduced food supply, and relative stability of chemical and physical
properties. Similar biological adaptations have been observed in
organisms inhabiting abyssal (Menzies et al. 1973) and subterranean

Present address: Savannah River Ecology Laboratory, Drawer E, Aiken, S.C.

29801.

Brimleyana No. 2: 119-130. November 1979

119

120

Gary W. Dickson, et al.

(Poulson 1964) areas. The deep-sea and cave environments probably lie
at the extreme end of a continuum of environmental stabilities.

Several theories relating the level of genetic variability maintained in a
natural population to the degree of environmental heterogeneity have
been proposed (Levins 1968; Grassle 1972; Selander and Kaufman 1973;
Valentine 1976). The basis for these hypotheses is that increased genetic
variation would allow individuals to be better suited to spatial and tem-
poral heterogeneity of the environment. A population in a heterogeneous
physical environment or a population which perceives its environment as
coarse-grained, would be expected to maintain greater genetic variability
than a population living under a more homogeneous regime (Nevo 1976).
These hypotheses have received some support in experimental laboratory
populations of Drosophila maintained under constant versus variable con-
ditions (Powell 1971; McDonald and Ayala 1974). Genetic studies of
deep-sea and cave faunas should permit tests with natural populations of
the proposed correlation between genetic variability and environmental
heterogeneity.

Electrophoretic techniques may be employed to survey products of
structural genes chosen without prior bias with respect to level of
variability. Such studies have been conducted on various deep-sea inver-
tebrates including mollusks, echinoderms and crustaceans. A recent
review of these studies compiled by Siebenaller (1978) indicated that
abyssal fauna contain levels of genetic variability similar to those of
species living in other aquatic habitats.

In one of the first genetic studies of a cave organism, Avise and Selan-
der (1972) observed exceptionally low genetic variability in certain cave
populations of the fish Astyanax mexicanus. Results identified an additional
complicating factor influencing genetic variability but unrelated to selec-
tion pressures per se : drift and founder effect in the frequently small
cavernicole populations. Since that time several other studies of genetic
variation in cave species have been published. Because of the possibility of
drift lowering genetic variation in many cave populations, tests of the
proposed relationship between environmental and genetic heterogeneity
using cave organisms are likely to be one-sided. That is, the observation of
normal or high levels of genetic variability would tend to refute the
proposed correlation, while the observation of low variability could often
be attributed to either selection pressures or stochastic events.

Our investigation was conducted (1) to examine genetic variability in
another cave organism, the troglobitic (i.e. obligate cave-dwelling)
amphipod crustacean Crangonyx antennatus Packard, and (2) to summarize
the available literature concerning genetic variation in other cave species.

Genetic Variation

121

METHODS AND MATERIALS

In this study, populations of the troglobitic amphipod crustacean, C.
antennatus , were examined electrophoretically. This species is one of the
most common and widespread aquatic troglobites in the eastern United
States. It occurs in the Appalachian region of Virginia, Tennessee,
Georgia and Alabama (Holsinger 1969, 1972). Crangonyx antennatus oc-
cupies two distinctly different habitats within its range: gravel-bottom
streams and mud-bottom pools. Significant differences in morphology
(Dickson 1977a), behavior (Dickson 1977b) and population structure
(Dickson and Holsinger, in press) were observed between populations liv-
ing in these habitats.

Populations of C. antennatus were sampled from six caves in Lee County,
Virginia (Fig. 1). Although the populations are relatively large in caves of
this area, small collections were taken from each cave to minimize pop-
ulation disruption (Table 1). Three of the populations sampled inhabited
gravel-bottom streams (Spangler, Cope and Gallohan No. 2 caves) and
three inhabited mud-bottom pools (Roadside No. 1, Molly Wagle and
Sweet Potato caves). Mud-bottom pool habitats generally contained
greater quantities of available food, fewer predators and smaller numbers
of potential competitors than stream habitats. Additional ecological data
on these habitats were given elsewhere (Dickson and Kirk 1976; Dickson
1977a); geological data are found in Holsinger (1975).

After collection, amphipods were transported live to the laboratory and
stored at — 70°C until analyzed. Amphipods were electrophoresed on
horizontal starch gels using conditions and staining procedures similar to
those of Selander et al. (1971) and Ayala et al. (1972). Of 21 enzyme
systems screened on six gel types, eight allowed consistent scoring, with
two of those exhibiting polymorphic allozyme patterns interpreted as
segregating electromorphs. As previously noted by Gooch and Hetrick (in
press), amphipods are not well suited for electrophoretic study, probably
because of the release of inhibitory enzymes from the hepatopancreas
during homogenization.

The monomorphic systems included glucose-6-phosphate
dehydrogenase (G-6-P), indophenol oxidase (IPO), aldehyde oxidase
(AO), malate dehydrogenase (MDH), and general proteins I-II (GP-1,
GP-2). The polymorphic systems scored were phosphoglucose isomerase
(PGI) and phosphoglucomutase (PGM). The PGI locus exhibited three-
banded heterozygote patterns, suggesting that this enzyme is a dimer.
Two-banded heterozygote phenotypes were observed at the PGM locus,
as expected for a monomeric enzyme. These enzyme structures are con-
sistent with those recorded in other crustaceans, including lobsters
(Tracey et al. 1975) and crayfish (K. Stueck, in prep.). To determine

122

Gary W. Dickson, et al.

Fig. 1. Allele frequency distributions for phosphoglucose isomerase and phospho-
glucomutase in the study area. Insert shows location of the six caves sampled for Crangonyx
antennatus in relation to the major surface drainage in the Powell Valley, Lee County,
Virginia. 1. Roadside No. 1 Cave (RD), 2. Molly Wagle Cave (MW), 3. Sweet Potato Cave
(SP), 4. Cope Cave (C), 5. Spangler Cave (S), and 6. Gallohan No. 2 Cave (G).

whether electromorphs produced at the same locus in different popula-
tions were identical or different, amphipods from all six populations were
run on the same gels.

RESULTS

All population samples were monomorphic for the same electromorph
at six loci (G-6-P, IPO, AO, MDH, GP-1 and GP-2). Electromorph fre-
quencies for the two polymorphic loci are given in Table 1. A total of five
allelic products were distinguished for both PGI and PGM, although in
only one case (PGM, Molly Wagle Cave) were all five electromorphs ob-
served in a single cave. For eight loci in C. antennatus , polymorphism
(proportion of loci polymorphic) equals 0.25. The mean proportion of loci

Genetic Variation

123

Table 1.

Allele frequencies at polymorphic loci in Crangonyx antennatus popula-
tions. (N) = number of individuals examined. Abbreviations as in Fig.

1.

Locus

allele

RD

MW

SP

S

C

G

PGM

a

0.536

0.043

b

0.923

0.155

0.203

0.250

0.391

0.281

c

0.077

0.155

0.250

0.750

0.196

0.188

d

' —

0.107

0.391

—

0.369

0.250

e

—

0.036

0.141

—

—

0.156

(N)

(26)

(42)

(32)

(36)

(23)

(16)

PGI

a

0.200

—

—

—

—

0.800

b

—

0.400

0.500

0.500

0.200

—

c

—

0.100

—

0.500

0.800

0.200

d

0.800

0.400

0.500

—

—

—

e

—

0.100

—

—

—

—

(N)

( 5)

( 5)

( 5)

( 5)

( 5)

( 5)

heterozygous per individual (heterozygosity) can be estimated in two
ways: by direct count, and by expected heterozygote proportions in each
population assuming Flardy-Weinberg equilibrium. Heterozygosity
values per cave range from 0.021 (counted) and 0.058 (expected) to 0.276
(counted) and 0.124 (expected) (Table 2), and overall mean
heterozygosities (H), weighted by sample size, in C. antennatus equal 0.1 18
and 0.136, respectively. The greater spread in values for the counted
heterozygosities primarily results from assay in some populations of only
small numbers of specimens at the apparently monomorphic loci. At the
PGM locus there was a consistent tendency for pool populations to ex-
hibit significant heterozygote deficiencies relative to Hardy-Weinberg

Table 2. Heterozygosity (H) values in populations of Crangonyx antennatus. Ab-
breviations'as in Fig. 1.

Population

Frequency of heterozygous loci in
an average individual

Counted

Expected ±1 Standard Error

RD

0.021

0.058 ± 0.042

MW

0.130

0.161 ± 0.106

SP

0.102

0.149 ± 0.099

s

0.183

0.109 ± 0.073

c

0.276

0.124 ± 0.087

G

0.216

0.107 ± 0.073

124

Gary W. Dickson, et al.

o 45

a a

oj c

bO 45

' CO

3 2

0 CU

1 X->

o

c

<u

ca

-M

• ^

4—*

X

03

_o

X

bn

II

O

a

H

X

co

2

45

>

X

o3

45

45

G

bn

G

G

C

03

4— *

X

X

45

03

co

X

C

03

3

X

CO

• ■— 1

G

>

>—

in

X

C

G

o3

be

i-H

a

O

O

a

45

X

c

c

3

4->

c

15

II

5

’C

2

cd

>

45

-*—1

45

c

43

bn >
ju 3

X qj

03 >

4J '-2
15 03

43

T3

_>-

03

13

<u

<u

E< C1

O -£

x £2

x J2

o be

u o

a

4J H
J3

03 43

oS

_co C

>

X O
CO

<u

X

03

H

43

CJ

a

d

o

C/5

•- x

w wT

9 O

^ 6

&? j>4

3

Oh

c/3

O

be

'E N

x- o

u

45

45

E

o*G
6 o
2 J

c

03

be

Ih

o

CM

r--

C'

tH

45

X

G

<u

GO

-a

G

o3

45

CO

’>

<

r-

r-

ON

JD

X

Oh

1 °

1> I

* ^~1
a M
a 45
43 G5

4*1 OO

4l> N —

±3 ^
x: v_^

r--

a-

ON

c

o3

3

a

X

c

03

JU

X

Si

45

CM

r-

ON

45

X

C

J3

45

CO

X

G

03

45

C/3

">

<

CM

r-

0

03

1978

X

NO

C"

X

t"-

O

X

t-H

, *1

T— '

4—»

*

45

>4

03

-4— >

<U

03

X

-a

.

X

0

45

Xh

<V

G

bn

V

be

G

c

in

c

a

443

03

03

O

X

c5

hJ

r-

r-

ON

NO

e-

ON

X 03

03 i *

45

>4

JV

X

CJ

o

U

45

• 4— H

c

o

X

Si

o

X

CO

X

>4

c

X

03

0

X

d

0

i-H

t-H

03

0

t'-

r-

On

0

0

0

r— H

X

LD

G

7^

r-l h-

o3

[-1 ON

C/5

C/5

45

Si

X

c

c

O C/5
a C/5
c/5 45

c

X

o

Si

Oh

G

ON

rd

CM

_43

X
as id

Si X

4J _o

X ca

C/3 ‘C
C o3
O >

u

45

>4

o

o

io

X

LO

CM

X

U-)

CM

O

CM

O

o

ao

X

03 ”0

Ci ._

45 XI

X ca
c r. i 'a
C 03

° >

u

x

On’

45

Si

45

_c

’<u

a

CO

oo

CO

45

X

C/5

c

CA

T — 1 * —■

a

a

u y

45

*C 'u

X

X X

<V V

O

XX

>4

a

0

X

c5

X

CM

5

1

'a

0

03

° >
U ^

13 45

X o
x Cu

45

N

CM

CU

2

LO

CM

X

CM

LO

E

CO

oO

CO

X

X

X

X

0

O

0

0

a

a

a

a

430

CM

X

X

X

00

•

X

LO

rO

X

X

M"

X

X

Ph

Qh

X

X

<

1

<

I

<

X

T " <

2

2:

2:

<

0

00

0

0

0

O

CO

0

On

4— 1

co

LO

JJ

X

a,

_o

be

o

45

a

X

_o

bn

o

Si

H

43

<o X

a 03

a

03

a X

E 3

X 03

a a
kj y.

a H

45

a |
a

Si 1
U. — v

-Si 03
-a Si

a <u
<-> 3b
45 X
, K O

x E
a o
~a r
kC ex

45
+->

X

o

bn

o

Si

H

x I
a, 1

1/3

<-> 03
a a
be iJ
43

-c; Q,
«u O
a 45

S o
U

45

•*—>

X

_o

"bn

o

Si

<-5 E— 1

a ,

45

>

X

_2

kk bn

f s

a ^

I

45

X

_o

bn

o

Si

H

~a

03

a Si

be 45

53 o

-a x
-Cc, o
a 45

I o

U

So ^

'*■<> •' ^

-3 5

■‘k h

45

K £

45 UU

43 O

-a 45
’a' o
^^2

o3

^5

^ O

5 Q-

45 X

-*-U .

J Q

oo ■—

_45

X

X

JO

bn

5 2

15

a I

<-5 '

a

JJ

45

45

45

lc

7^23

a

X

Qa

X

45

X

X

O

X

r3

_o

JO

0

_o

bn

X

X

bn

0

bn

0

5_

O

0

a

a

H

a

H

bn
„ 0

H

H

X

X

3

-♦-*

CO

4-5

c

45

C/3

45

Si

(X

LO

CM

NO

CM

430

^H

OO

r-

r-

X

d

On

OO

V

u

4—1

CM

X

X

X

X

CM

X

03

X

G

Oh

O

O

O

4f— (

0

O

O

^_i

, ,

O

O

2

2

O

O

0

O

2

C3

2

O <2>

P 2

O

O

03

-4-J

03

OO

O

X

Si

45

K 1
45

bo as

X

■S, O
a X

t-

-43 -

C.

3 ON

o

a

ii

-§1 Q
o” O

U

x O

"a

kj

s

a

G

X

"c

X

u

o3

iu

<

~a

a

a

a

' c~U>

So

-C)

Su

to

U

H

CX oj

Xx
^ o

a .9*

2^
x 4! x
~r c
o c

■ ts

o3

X

o

co

Co

a

a

L-j

a

Si

a

S

5

a

O

03

X

O

_x

X

X

g

<

45

■*—>

X

_o

bn

o

3 H

'*-u

a
a
a
a

03

X

o

X

a^X

^ X
53 2

E o-

CO

c

_o

■4—*

03

4—*

’cj

X

45

45

X

X

45

4— »

G

45

co

G

O

45

45

X

X

G

O

45

J5

o

X

03

co

443

CO

45

’G

45

X

co

45

X

45

U

c

45

X

45

45

>

03

>-

Si

O

4— 1

03

bn

X

o

X

x ■£

45 •-

3 ^

x ~a

g w

o

45

CJ

45

Si

03

45
O
co
co
>S C3

co

45

45

X
45
45
3
X
o

Si

X -o

45 u

CJ

CCJ

Si

o3

X

u

Si

45

X

+1

’45

45

Si

45

CO

45

_G

03

>

C/3

CO

45

co

co

O

X

X .

45 cj

41 .

Si I

0 X

X"G

05 x

a O
co bn
c o
o H

03 x
3
X
03

x

3 a
X 45
O X

H X '

Genetic Variation

125

expectations. Heterozygosity values displayed no correlation with dis-
tance from the Powell River.

Sample sizes and numbers of loci scored were small for reasons noted
earlier, so the results must be interpreted with extreme caution. Estimates
of genetic variability fall near the middle of the range of values previously
reported in crustaceans from other aquatic habitats (H ranges from 0.038
in the lobster, Homarus amencanus , to 0.211 in the krill, Euphausia
distinguenda; see reviews in Nevo 1978 and Powell 1975). Heterozygosity
values in C. antennatus are considerably above mean values reported for
eight other cave species (H = 0.056, Table 3).

A striking aspect of the data is the high degree of allele frequency
heterogeneity among C. antennatus populations, particularly at the PGM
locus where sample sizes were fairly large (Table 1, Figure 1).
Heterogeneity among localities can be expressed by the standardized
genetic variance (FST), which provides a common scale for comparing the
relative levels of interlocality allele frequency heterogeneity at different
loci, or in different sets of populations (Wright 1978). For five elec-
tromorphs at the PGM locus, weighted FST = 0.300 among populations
of C. antennatus inhabiting the approximately 60 km2 area of the study.
This variance in allele frequency is greater than mean values among snail,
Helix aspersa , populations in different California cities or house mice, Mus
musculus, populations on different farms in Texas (Selander and Kaufman
1975). It is comparable to observed values between bluegill sunfish,
Lepomis macrochirus, populations inhabiting distinct reservoirs within large
southeastern drainages (Avise and Felley, in press). The microgeographic
heterogeneity in C. antennatus is consistent with the contention that caves
may be highly partitioned habitats, analogous to islands or archipelagos
(Culver 1971).

Genetic similarity values were calculated between pairs of cave popula-
tions using Nei’s (1972) identity statistic (T). This statistic can assume
values from 0 (no electromorphs shared) to 1 (identical electromorph fre-
quencies). For the populations of C. antennatus studied, I values fell be-
tween 0.858 and 0.968, within the range typical of conspecific populations
in other invertebrates. The identity matrix was employed to generate a
biochemical dendrogram (Fig. 2) according to the unweighted pair-group
method with arithmetic means (Sneath and Sokal 1973). There is no clear
tendency for stream and pool populations to cluster distinct from one
another. Based on the relatively distinct pattern of allele frequencies in
the population from Roadside No. 1 Cave (Figs. 1, 2; Table 1), it appears
that the Powell River has acted as a partial barrier to gene flow.

126

Gary W. Dickson, et al.

I

0.85 0.90 0.95 1.00

i 1 1 1

Roadside No.l

— POOL

Gailohan No. 2

— STREAM

Sweet Potato

— POOL

Molly Wagle

— POOL

Cope

STREAM

Spangler

STREAM

(185 0.90 095 LOO

Fig. 2. Dendrogram of C. antennatus populations based on genetic similarity values derived
from eight loci. Scale is in units of genetic identity, I. Cophenetic correlation equals 0.79

DISCUSSION

In addition to measurements of physical and chemical parameters,
studies of troglobitic organisms have yielded biological evidence which at-
tests to the relative environmental stability of cave systems. In general,
troglobitic species exhibit K-selected population characteristics including
late maturity, low reproductive rates, large size at hatching and increased
longevity (see reviews in Vandel 1965; Dickson and Holsinger, in press).
Population characteristics of this type have been associated with
relatively stable (i.e. predictable) habitats (MacArthur and Wilson
1967). Of these characteristics, perhaps the most remarkable is the ex-
tremely long life span reported for various troglobites (Poulson 1964;
Cooper 1975; Dickson and Holsinger, in press). Increased longevity has
also been described in deep-sea organisms (Turekian et al. 1975;
Engemann 1978).

From the results of the present investigation and previous studies on a
number of different cave species (Table 3), genetic variability does not ap-
pear to be substantially reduced in populations inhabiting subterranean

Genetic Variation

127

habitats. Values of genic heterozygosity in many cavernicolous popula-
tions are comparable to those of species from epigean environments (Nevo
1978). However, a cautionary note is required. Studies on troglobitic
organisms are still few in number, and generally have included only a
small number of loci. Results are not yet sufficient to eliminate the
possibility of a mild quantitative decrease in genic variation in cavern-
icolous populations. It does appear very likely that the genetic response of
troglobitic organisms to their environment (at least as evidenced by level
of genetic variability) is not qualitatively different from that of other
species. Analogous conclusions have been reached from studies of
organisms inhabiting abyssal areas (Siebenaller 1978; Costa and Bisol
1978).

The genetic variability observed in populations inhabiting stable en-
vironments (i.e. caves and deep-sea) may have different origins.
Siebenaller (1978) proposed that the relatively high variability observed
in many deep-sea invertebrates could be due to the presence of large pop-
ulation sizes and to interactions predicted by the time, population size
and divergence hypothesis of Soule (1976). In cave organisms, species and
population numbers are normally smaller than in most faunal counter-
parts living in other habitats, including abyssal areas (Poulson 1964).
The normal levels of genetic variability in these subterranean species
could be due to a process described as ecological release (Avise and Selan-
der 1972). In the generally depauperate cave communities, troglobites
could expand their niches and exploit resources which would normally be
shared by two or more species. Relatively high levels of genetic variability
might be maintained under these conditions.

If genetic variation is not neutral with respect to fitness, normal levels
of genetic variability in deep-sea and cave organisms could also be main-
tained by selection due to spatial or temporal heterogeneity. Although
caves and the deep-sea are considered more stable than most other
aquatic and terrestrial areas, physical habitat variability and seasonal
fluctuations in certain parameters are present (Poulson and White 1969;
Menzies et al. 1973). It is possible that the threshold of environmental
heterogeneity below which significant genetic variability is no longer
favored is still lower than the level of environmental variation in the
proverbially stable environments of caves and the deep-sea.

ACKNOWLEDGMENTS. — We are grateful to Mrs. Eleanor Berry,
owner of Molly Wagle Cave, and to other cave owners in Lee County,
Virginia, for their cooperation in allowing us access to their property. In
addition, we thank Karen L. Stueck, Raymond R. White, Donald G.
Buth and an anonymous reviewer for providing helpful comments and

128

Gary W. Dickson, et al.

criticisms on various drafts of this manuscript. Portions of this study were
supported by Contract DE-AC09-76SR00819 between the University of
Georgia and the U.S. Department of Energy.

LITERATURE CITED

Avise, John C., and J. Felley. In press. Population structure of freshwater fishes. I.
Genetic variation of bluegill ( Lepomis macrochirus) populations in man-made
reservoirs. Evolution.

, and R. K. Selander. 1972. Evolutionary genetics of the cave-dwelling

fishes of the genus Astyanax. Evolution 26: 1-19.

Ayala, Francisco J., J. R. Powell, M. L. Gracey, C.A. Mourao and S. Perez-Salas.
1972. Enzyme variability in the Drosophila willistoni group. IV. Genic variation
in natural populations of Drosophila willistoni. Genetics 70:113-139.

Carmody, George R., G. Murphy and S. B. Peck. 1972. Preliminary studies on
electrophoretic variation in cavernicolous Ptomaphagus beetles (Coleoptera,
Leiodidae, Catopinae). Ann. Speleol. 27:399-404.

Cockley, D. E., J. L. Gooch and D. P. Wetson. 1977. Genic diversity in cave-
dwelling crickets ( Ceuthophilus gracilipes ). Evolution 27:313-318.

Cooper, John E. 1975. Ecological and behavioral studies in Shelta Cave,
Alabama, with emphasis on decapod crustaceans. Ph.D. dissert., Univ. Ky.,
Lexington, xvi + 364 pp.

Costa, Rodolfo, and P. M. Bisol. 1978. Genetic variability in deep-sea organisms.
Biol. Bull. 755:125-133.

Culver, David C. 1971. Caves as archipelagoes. Bull. Natl. Speleol. Soc. 33: 97-

100.

Dickson, Gary W. 1977a. Variation among populations of the troglobitic
amphipod crustacean Crangonyx antennatus Packard living in different habitats.
I. Morphology. Int. J. Speleol. 9:43-58.

1977b. Behavioral adaptation of the troglobitic amphipod crustacean

Crangonyx antennatus to stream habitats. Hydrobiologia 56:17-20.

, and P. W. Kirk, Jr. 1976. Distribution of heterotrophic microorganisms

in relation to detritivores in Virginia caves. (With supplementary bibliography
on cave mycology and microbiology), pp. 205-226 in Parker, Bruce C., and M.
K. Roane (eds.). The distributional history of the biota of the southern Ap-
palachians. Part IV: Algae and fungi. Univ. Press Va., Charlottesville. 416 pp.

, and J. R. Holsinger. In press. Variation among populations of the

troglobitic amphipod crustacean Crangonyx antennatus Packard (Crangonyct-
idae) living in different habitats, III: Population dynamics and stability. Int. J.
Speleol.

Engemann, Joseph G. 1978. Indirect evidence shows deep-sea benthos may reach
extreme ages as individuals. Am. Zool. 18: 666.

Gibert, Janine. 1971. Analyse electrophoretique des proteines de Niphargus
(Crustace Amphipode hypoge) en fonction de facteurs physiologiques,
ecologiques et systematiques. These Doctorat de speciality, Faculte de Lyon,
France. 130 pp.

Giuseffi, Steven, T. C. Kane and W. F. Duggleby. 1978. Genetic variability in the
Kentucky cave beetle Neaphaenops tellkampfu (Coleoptera: Carabidae). Evolu-
tion 22:679-681.

Genetic Variation

129

Gooch, James L., and S. W. Hetrick. In press. The relation of genetic structure to
environmental structure: Gammarus minus in a karst area. Evolution.

Grassle, J. F. 1972. Species diversity, genetic variability and environmental uncer-
tainty. Proc. Eur. Mar. Biol. Symp. 5:19-26.

Hetrick, Steven W. 1975. Genetic studies of cave and spring populations of the
freshwater amphipod crustacean Gammarus minus Say. M. S. thesis. Old
Dominion Univ., Norfolk. 45 pp.

Holsinger, John R. 1969. Biogeography of the freshwater amphipod crustaceans
(Gammaridae) of the central and southern Appalachians, pp. 16-50 in Holt,
Perry C., (ed.). The distributional history of the biota of the southern Ap-
palachians, Part I: Invertebrates. Res. Div. Monogr. 1, Va. Polytech. Inst.,
Blacksburg. 295 pp.

1972. The freshwater amphipod crustaceans (Gammaridae) of North

America. Biota of Freshwater Ecosystems. Ident. Man. 5, U. S. Environ.
Protect. Agency, Washington. 89 pp.

1975. Descriptions of Virginia caves. Va. Div. Mineral Resour. Bull.

85. 450 pp.

Johnston, G. H., and G. R. Carmody. In press. Allozyme variation within and be-
tween populations of the cave spider Meta menardi (Latreille). Can. J. Zool.

Laing, C., G. R. Carmody and S. B. Peck. 1976a. Population genetics and
evolutionary biology of the cave beetle Ptomaphagus hirtus. Evolution 59:484-498.

1976b. How common are sibling species in cave-inhabiting inver-
tebrates? Am. Nat. 770:184-189.

Levins, Richard. 1968. Evolution in changing environments. Princeton Univ.
Press, Princeton, New Jersey. 120 pp.

MacArthur, Robert H., and E. O. Wilson. 1967. The theory of island
biogeography. Princeton Univ. Press, Princeton, New Jersey. 203 pp.

McDonald, John F., and F. J. Ayala. 1974. Genetic response to environmental
heterogeneity. Nature 250: 572-574.

Menzies, Robert J., R. Y. George and G. T. Rowe. 1973. Abyssal environment
and ecology of the world oceans. John Wiley and Sons, New York. 488 pp.

Merkle, Donald A., and S. I. Guttman. 1977. Geographic variation in the cave
salamander Eurycea lucifuga. Herpetologica 55:313-321.

Nei, Masatoshi. 1972. Genetic distance between populations. Am. Nat. 106: 283-
292.

Nevo, Eviatar. 1976. Adaptive strategies of genetic systems in constant and vary-
ing environments, pp. 141-158 in Karlin, S., and E. Nevo (eds.). Population
genetics and ecology. Academic Press, New York.

1978. Genetic variation in natural populations: Patterns and theory.

Theor. Popul. Biol. 75:121-177.

Poulson, Thomas L. 1964. Animals in aquatic environments: Animals in caves,
pp. 749-771 in Dill, D. B. (ed.). Handb. Physiol. Sect. 4, Adapt. Environ., Am.
Physiol. Soc., Washington.

1971. Biology of cave and deep-sea organisms: A comparison. Bull.

Nat. Speleol. Soc. 55:51-61.

, and W. B. White. 1969. The cave environment. Science 705:971-981.

Powell, leffrey R. 1971. Genetic polymorphisms in varied environments. Science
174: 1035-1036.

1975. Protein variation in natural populations of animals. Evol. Biol.

5:79-1 19.

130

Gary W. Dickson, et al.

Sbordoni, Valerio, E. de Matthaeis and M. C. Sbordoni. 1976. Phosphogluco-
mutase polymorphism and natural selection in populations of the cave cricket
Dolichopoda geniculata. Z. Zool. Syst. Evolutionsforsch. 7 4:292-299 .

Selander, Robert K., and D. W. Kaufman. 1973. Genic variability and strategies
of adaptation in animals. Proc. Natl. Acad. Sci. U.S.A. 7(9:1875-1877.

, and 1975. Genetic structure of populations of the brown

snail (Helix aspersa). I. Microgeographic variation. Evolution 29:385-401.

, M. H. Smith, S. Y. Yand, W. E. Johnson and J. B. Gentry. 1971.

Biochemical polymorphisms and systematics in the genus Peromyscus. I. Varia-
tion in the old-field mouse. Stud. Genet. VI. Texas Univ. Publ. 7103-49-90.
Siebenaller, Joseph F. 1978. Genetic variation in deep-sea invertebrate popula-
tions: the bathyal gastropod Bathybembix bairdii. Mar. Biol. 47:265-275.
Sneath, Peter H. A., and R. R. Sokal. 1973. Numerical Taxonomy. W. H.

Freeman and Co., San Francisco, California.

Soule, Michael. 1976. Allozyme variation: its determinants in space and time. pp.
60-77 in Ayala, Francisco J. (ed.). Molecular Evolution. Sinauer Assoc., Inc.,
Sunderland, Massachusetts.

Steiner, William W. M., E. A. Lisowski and D. Osterbur. 1977. Biochemical dif-
ferences in sympatric color morphs of an aquatic isopod ( Asellus brevicauda).
Comp. Biochem. Physiol. B Comp. Biochem. 56:371-374.

Tracey, M. L., K. Nelson, D. Hedgecock, R. A. Schleser and M. L. Pressick.
1975. Biochemical genetics of lobsters: genetic variation and the structure of
American lobster ( Homarus amencanus ) populations. J. Fish. Res. Board Can.
52: 2091-2101.

Turekian, K. K., J. K. Cochran, D. P. Kharkar, R. M. Cerrato, J. R. Vaisnys, H.
L. Sanders, J. F. Grassle and J. A. Allen. 1975. Slow growth rate of a deep-sea
clam determined by 228Ra chronology. Proc. Natl. Acad. Sci. U.S.A. 72:2829-
2832.

Valentine, James W. 1976. Genetic strategies of adaptation, pp. 78-94 in Ayala,
Francisco J. (ed.). Molecular Evolution. Sinauer Assoc., Inc., Sunderland,
Massachusetts.

Vandel, A. 1965. Biospeleology, the biology of cavernicolous animals. Pergamon
Press, New York. 524 pp.

Wright, Sewall. 1978. Evolution and the genetics of populations. Vol. 4.
Variability within and among natural populations. Univ. Chi. Press, Chicago.
580 pp.

Accepted 29 August 1979

A Photographic Technique to Study Tadpole Populations

C. Kenneth Dodd, Jr.

Office of Endangered Species , U.S. Fish and Wildlife Service,

Washington, D. C. 20240

ABSTRACT. — Although the use of photography to estimate the num-
bers of vertebrates in wild populations is a well established wildlife man-
agement technique, it has not been previously used in studies of tad-
poles in shallow breeding ponds. A pond used by the American toad,
Bufo americanus , was monitored through the breeding season of 1977 in
Fairfax County, Virginia. Tadpole clusters were photographed and the
prints used to evaluate this technique for working with such populations.
An example of how the technique may be used is provided. While espe-
cially valuable in certain types of studies (for instance, those involving
tadpoles in shallow ponds without much vegetation and containing one
or a few species), photographic analysis is probably of use in many field
studies.

INTRODUCTION

Use of photography to estimate the numbers of vertebrates in wild pop-
ulations is a well established wildlife management technique (Watson
1969; also see references in Caughley 1977 and Giles 1969) although it
has generally been employed in connection with aerial censuses.
However, photography has not been previously used to estimate the num-
bers of larval amphibians in breeding ponds. Since many species of
amphibians breed in shallow ponds where the larvae are readily visible,
such populations should be amenable to analysis by photographic
methods. I attempted to determine if photography could be used to study
one of these populations of tadpoles.

METHODS AND MATERIALS

The study area encompassed a shallow (initially 70 mm maximum
depth) temporary pond, 42.8 m2 in surface area, located near the con-
fluence of Indian Run and Back Lick Run, Fairfax County, Virginia.
Characteristic vegetation in and surrounding the pond included sedges
(Cyperaceae), Juncus sp., Bidens sp., Aster sp., Lespedeza sp., willow ( Salix
sp.) and river birch ( Betula nigra)', individual plants often formed clumps
and were scattered along the bottom and around the edges of the pond.
The pond was surrounded by a rather well-defined margin consisting of a
thick growth of vegetation on one side and a mud bank on the other. This

Brimleyana No. 2: 131-136. November 1979

131

132

C. Kenneth Dodd, Jr.

margin apparently represents the limit to which the pond fills during
periods of maximal rainfall. Tadpoles remained in the pond and were not
observed in any of the small pockets of water located in the margin.

Five study quadrats were selected and photographed with Kodachrome
64 slide film using a Nikon F camera with a 55 mm micro lens. A one
meter ruler was placed at the top of each area photographed to insure
uniformity of quadrat size; each quadrat measured 25 x 40 cm (see Fig. 1
for location of quadrats). The camera was hand held directly over the
area to be photographed, in such a way as to assure that the film plane
was parallel to the surface of the water and the scale of reference. If this is
not done the resulting shift in parallax could cause a scaling problem,
thus affecting area, and therefore density, estimates. Care was taken to
insure that movements of the observer and placement of the meter ruler
did not disturb the tadpoles and cause them to change their position.

Each slide was enlarged to a 17 x 25.5 cm print and the 25 x 40 cm
quadrat was marked off (Fig. 2). A clear plastic sheet was overlain on the
photograph and tadpoles marked onto the sheet as they were counted.
Thus, both the exact position and number of tadpoles could be scored for
each quadrat.

Photographic Technique

133

Fig. 2. Quadrat B showing the meter rule and the 25 x 40 cm study plot.

RESULTS AND DISCUSSION

Results of the photographic overlays are provided in Fig. 3. Tadpoles
were easily counted and marked; seldom was there a question as to
whether a black mark on the photograph should be counted. However,
tadpole shadows are a potential problem. Densities of tadpoles were
generally D-2 (20-50% of area covered by tadpoles, tadpoles within 1-5
cm of one another) or D-3 (51-89% of area covered, tadpoles mostly in
contact) in the sense of Beiswenger (1975); if densities greater than this
are encountered, photographic analysis might prove difficult.

Although the five study plots were apparently selected at random, I
made no attempt at complete objectivity, such as using a table of random
numbers to preselect quadrats. Therefore, an estimate of population size
based on total numbers of tadpoles from these sites would not provide a
statistically accurate picture of tadpole numbers. Methods for the estima-
tion of population sizes via direct counts within randomly selected
quadrats are available. Caughley (1977) provided several methods which
might be useful, notably the use of stratified random sampling as applied
to aerial surveys.

134

C. Kenneth Dodd, Jr.

Fig. 3. Positions of tadpoles in each of the study plots as determined from photographs.
The number of tadpoles in each quadrat is: A) N = 545, B) N = 599, C) N = 268, D) N = 213,
E) N=480.

As an example, had the quadrats used in this study been randomly
preselected, the stratified random sampling method would have yielded a
value of Yh = 180,188. The standard error rounded off to 34,000 would
give a population estimate (with 95% confidence) between 116,200 and
244,200 tadpoles. The literature on toads reports a wide range of eggs
deposited per female (2000 to 20,600). Therefore, using the high figure, 6
to 12 females produced these tadpoles; using the low figure, 58 to 122
females produced them. Other data gathered preliminary to this study
showed that 12 females actually bred at the pond. Assuming 100% suc-
cessful hatch of eggs, this number was therefore within the admittedly
large estimate predicted by tadpole numbers. On the other hand, if eggs

Photographic Technique

135

136

C. Kenneth Dodd, Jr.

per female were estimated, then each female would have deposited be-
tween 9680 and 20,350 eggs, again within the range of literature reports.
Obviously, knowledge of other aspects of a species’ biology would severely
limit basing predictions about anuran population breeding characteristics
solely on tadpole counts. However, depending on species, photographic
analysis of a tadpole population could be substituted for, or used to sup-
plement, counts in the field. Thus, photography could provide a valuable
technique for the estimation of certain amphibian population parameters.

There are obvious problems with the method. Deep ponds or those that
are very murky or clogged with vegetation would not be suitable for such
analysis, nor would ponds that harbor many species of tadpoles, since
tadpoles are difficult to identify without examining certain key charac-
teristics (Altig 1970). However, many types of tadpole populations would
seem to be particularly suited to photographic analysis, notably tadpoles
of those species which use shallow and/or temporary water sources, es-
pecially in xeric habitats. In addition, photographic analysis could be
valuable in monitoring the breeding status of species considered en-
dangered or threatened, such as the Houston toad, Bufo houstonensis. The
animals could be left undisturbed, thus avoiding the harassment which
may be prohibited by law without a special permit. Sequential
photography might also prove valuable in studies of the orientation
behavior of tadpoles over a period of time.

ACKNOWLEDGMENTS. — I would like to thank John Fay, Bruce
MacBryde, and LaVerne Smith for identifying the plants. Ronn Altig,
Paul V. Cupp, James J. Hebrard and an anonymous reviewer provided
valuable criticisms and comments on the manuscript.

LITERATURE CITED

Altig, Ronald. 1970. A key to the tadpoles of the continental United States and
Canada. Herpetologica 26:180-207.

Beiswenger, Ronald E. 1975. Structure and function in aggregations of tadpoles of
the American toad, Bufo americanus. Herpetologica 31: 222-233.

Caughley, Graeme. 1977. Analysis of vertebrate populations. John Wiley & Sons,
London. 234 pp.

Giles, R. H., Jr. (ed.). 1969. Wildlife management techniques. Wild. Society,
Washington, D. C. 623 pp.

Watson, R. M. 1969. Aerial photographic methods in censuses of animals. East
African Agric. For. J. 34: 32-37.

Accepted 8 July 1979

Rediscovery of the Sharphead Darter, Etheostoma acuticeps , in
North Carolina (Pisces: Percidae)

Richard T. Bryant, James P. Beets and Michael G. Ryon
Graduate Program in Ecology, University of Tennessee,

Knoxville, Tennessee 37916

ABSTRACT. — Etheostoma (Nothonotus) acuticeps , the Sharphead darter,
was known prior to October 1975 on the basis of only 37 specimens from
the South Fork Holston River, Tennessee, and the North Toe River,
North Carolina. The populations at all of these localities are either ex-
tremely tenuous or extirpated by impoundments and strip mining
runoff. In 1975 a TVA field crew discovered a healthy population in the
lower Nolichucky River below Davy Crockett Reservoir. A search of the
upper tributaries of the Nolichucky revealed a small population in the
lower 6 km of the Cane River, North Carolina. The Sharphead darter
inhabits the swiftest portion of riffles in water 15-40 cm deep. The sub-
strate is typically cobble and small boulders 8-20 cm in diameter, well
covered with riverweed, Podostemum ceratophyllum. Because of the small
size of the Cane River population and various threats to the water
quality of the river, the Sharphead darter warrants endangered status in
North Carolina.

INTRODUCTION

The Sharphead darter, Etheostoma acuticeps , is a poorly known member
of the subgenus Nothonotus. In fact, until October 1975, only 37 specimens
of this Fish had been collected and the species was considered extinct. It
was not recognized as distinct until 1947, when two specimens were
collected in the South Fork Holston River, Tennessee. A last attempt to
obtain the darter prior to the impoundment of South Holston Reservoir
resulted in only four specimens, indicating a meager population. The
species was described by Bailey (1959) on the basis of these six
specimens, and at that time the only known locality was under 58 m of
water.

In the 1960s L.W. Knapp and T. Zorach independently uncovered four
misidentified collections of E. acuticeps at the Academy of Natural Sciences
of Philadelphia totaling 28 specimens (Zorach 1972). These had been
labeled by Fowler (1936a,b) as Nothonotus rufilineatus. The collections were
made in June 1930 by J.G. Carlson and T. Kerr in the South Fork
Holston River, Tennessee, and the North Toe River, North Carolina. Ap-
parently the collections were not very exhaustive, as they included only a
few species. Etheostoma acuticeps was caught in respectable numbers,

Brimleyana No. 2: 137-140. November 1979

137

138

Richard J. Bryant, et al.

however, indicating much healthier populations than at the type locality,
but these populations are also undoubtedly now extirpated. The South
Fork Holston localities are either impounded or are tailwaters supporting
cold-adapted fishes (Jenkins and Burkhead 1975). The North Toe River
has been subjected to extensive siltation from mica strip mining in North
Carolina (Jenkins and Burkhead 1972, Saylor and Etnier 1976). The
mica and feldspar dust smothers the eggs and eliminates the interstitial
microhabitat of the darters. In recent years siltation of the North Toe has
been aggravated by highway construction that parallels the river nearly to
its headwaters.

Exhaustive attempts by Jenkins and Burkhead (1975) to obtain E.
acuticeps in the Holston drainage in 1972 resulted in only three specimens.
These were taken in the South Fork Holston River just above the im-
pounded portion of South Holston Reservoir, where there apparently is
enough habitat to support a small and extremely tenuous population. In
the fall of 1975, Charles Saylor and a TV A field crew looking for the Snail
darter, Percina tanasi , discovered the Sharphead darter in the Nolichucky
River between Davy Crockett and Douglas Reservoirs (Saylor and Etnier
1976). Davy Crockett Reservoir had acted as a settling basin and pro-
tected the lower Nolichucky population from the siltation that eliminated
the North Toe population. The discovery of this very healthy population
allowed better understanding of the exact microhabitat of the Sharphead
darter and fueled speculation by D. Etnier (pers. comm.) that the species
may exist in larger tributaries of the Nolichucky that are relatively free of
siltation. Bailey (1977) considered the Sharphead darter extirpated in
North Carolina, however, as it had not been collected in the state in 47
years. v

Sharphead Darter Rediscovery

139

RESULTS AND DISCUSSION

We collected Etheostoma acuticeps on 14 July 1977 in the Cane River just
below the US hy. 19W bridge, 1 km se of Sioux, Yancey County, North
Carolina. It has since been collected at several sites between the 19W
bridge and the confluence with the Toe River by E. F. Menhinick and his
ichthyology class from the University of North Carolina at Charlotte, and
by the authors. Attempts by both groups to find the fish above the 19W
bridge have been fruitless, leading us to believe that this is its upstream
limit.

The river above the bridge is slightly higher in gradient than below,
with larger rocks and less riverweed, Podostemum ceratophyllum. The only
member of the subgenus Nothonotus found there was the Greenfin darter,
Etheostoma chlorobranchium. The habitat below the bridge is more like that
in the lower Nolichucky. The Sharphead darter is found in the main body
of riffles, in fairly swift water 15-40 cm deep. The substrate is cobble and
small boulders, 8-20 cm in diameter, usually well covered with riverweed.
The lower Nolichucky has slightly smaller substrate and lower current
velocity. The lower Cane River drops about 2.3 m/km and the lower
Nolichucky only 1 m/km. The riffles are typically shaded by large trees
and bordered by beds of waterwillow, Justicia amencana. Etheostoma
acuticeps is not a numerous species at any of these sites. The largest collec--
tion made by Menhinick and class at the 19W bridge included only 14
specimens in a sample containing 80 other darters.

During November 1977 a flood in the Cane River watershed washed
away a road, a house and a service station at the 19W bridge. In order to
replace the road bed, bulldozers were used in the vicinity of the bridge, an
action which inadvertently eliminated Sharphead darter habitat. The
shade trees, waterwillow beds, and thick riverweed growth are gone, and
the area is broad, bare gravel. Just how much of this destruction is due to
the bulldozing is difficult to assess, but riffles above and below the bridge
seem to be unchanged and normal habitat. Intensive sampling at t
bridge on 23 September 1978 revealed mostly Gilt darters, Percina evides
few E. chlorobranchium and only one E. acuticeps. In addition, Big Creek,
which enters the Cane just above the bridge, has been channelized for
several kilometers upstream. This may have detrimental effects on the
lower Cane. Other problems upstream of the E. acuticeps population in-
clude two gravel washing operations, one of which, according to a local
game warden, was the source of a chemical spill that resulted in a fish kill
5 km long.

Etheostoma acuticeps is known in North Carolina from only the lower 6
km of the Cane River. It apparently is not numerous and its continued ex-
istence is not assured. Channelization, gravel washing, and possible

140

Richard J. Bryant, et al.

chemical spills may have detrimental effects on this species, as well as
others inhabiting the lower Cane River. The Sharphead darter should be
considered an endangered species in North Carolina.

ACKNOWLEDGMENTS.— The authors would like to thank Dr. E. F.
Menhinick, University of North Carolina at Charlotte, for the use of his
excellent field notes, and Dr. J. F. McCormick and the Ecology Program
at the University of Tennessee for support and the use of a vehicle.

LITERATURE CITED

Bailey, Joseph R. 1977. Etheostoma acuticeps Bailey. Sharphead darter, pp. 284-285
in Cooper, J. E., S. S. Robinson andj. B. Funderburg (eds.). Endangered and
Threatened Plants and Animals of North Carolina. N. C. State Mus. Natural
Hist., Raleigh, xvi + 444 pp.

Bailey, Reeve M. 1959. Etheostoma acuticeps , a new darter from the Tennessee River
system, with remarks on the subgenus Nothonotus. Occas. Pap. Mus. Zool.
Univ. Mich. 603:1-10.

Fowler, Henry W. 1963a. Notes on some Tennessee fishes. Fish Culturist 75:111.

1 936b. Freshwater fishes obtained in North Carolina in 1 930 and 1 934. Fish

Culturist 75:192-194.

Jenkins, Robert E., and N. M. Burkhead. 1975. Recent capture and analysis of
the sharphead darter, Etheostoma acuticeps , an endangered percid fish of the up-
per Tennessee River drainage. Copeia 1 975 (4) :731-740.

Saylor, Charles F., and D. A. Etnier. 1976. Discovery of the Sharphead darter,
Etheostoma acuticeps Bailey, in the lower Nolichucky River, Tennessee. ASB
Bull. 23{ 2):93-94. Abstract.

Zorach, Timothy. 1972. Systematics of the percid fishes Etheostoma camurum and E.
chlorobranchium , new species, with a discussion of the subgenus Nothonotus.
Copeia 1 972(3) :427-447.

Accepted 22 January 1979

Nesting Biology of Andrena ( Larandrena ) miserabihs Cresson and
Description of the Prepupa (Hymenoptera: Andrenidae)

Beth B. Norden and Aubrey G. Scarbrough

Department of Biological Sciences, Towson State University,

Baltimore, Maryland 21204

ABSTRACT. — Andrena ( Larandrena ) miserabihs Cresson was found nesting
in the sandy soil of the amphitheater in Highlands Hammock State
Park, Highlands County, Florida. Observations were made on 31
December 1975, 1-3 January 1977, and 4-7 March 1977. Circular tumuli
marked nest entrances and an average density of 11 nests per 15 cm2
area was found. Nest tunnels were ca. 5 mm in diameter and extended
from 28.5 cm to 47.0 cm deep. A single cell ca. 5.7 mm in diameter was
constructed, lined, and provisioned. Pollen balls averaging 2.9 mm in
diameter were dark yellow and consisted of Acer rubrum pollen. Prepupa
were cream-colored, C-shaped, and 5.2 mm to 7.3 mm long. Sphecodes sp.
and Solenopsis sp. were found in association with the nesting bees, and
evidence of nest destruction by the armadillo, Dasypus novemcinctus, was
noted.

Although the Holarctic genus Andrena is a common, extensive tax-
onomic group, little is known about the biology of many species (Rozen
1973, Davis and LaBerge 1975). A large colony of Andrena ( Larandrena )
miserabihs Cresson in Florida provided an opportunity to observe nest
biology on 31 December 1975, 1-3 January 1977, and 4-7 March 1977.
Adults and larvae also were collected during these visits for later
laboratory examination. All specimens retained are in the Entomology
Collection, Towson State University.

Study Site. — Andrena miserabihs was nesting in the soil of the
amphitheater in Highlands Hammock State Park, Highlands County,
Florida. The amphitheater consists of an oblong clearing ca. 9.0 m X 15.0
m in the mature hardwood hammock. It contains 12 rows of wooden
benches (Fig. 1) and is surrounded by hydric hammock dominated by
broadleaved evergreen trees including Liqmdamber styraciflua, Persa
palustns, Godoma lasianthus, Ulmus flondana, Nyssa sylvatica, Sabal palmetto ,
and Citrus aurantium. Epiphytes of the genera Epidendrum, Tillandsia , and
Polypodium are common on the trunks and limbs. All nests were located
within the clearing, 2 to 3 m inward from the peripheral vegetation, and
this was the only nesting site found within the park.

Nests of A. miserabihs were scattered throughout the clearing, beneath

Brimleyana No. 2: 141-146. November 1979

141

142

Beth B. Norden and Aubrey G. Scarbrough

Fig. 1. Study site, March 1977.

benches as well as in open walkways. As pointed out by Osgood (1972),
sites with well drained soil, good surface flow, and sparse plant cover are
most often selected for nest construction by solitary bees. The sandy soil
of the amphitheater provided these characteristics, while the hammock
floor was densely matted with plant debris. The loose leaf litter covering
the amphitheater floor presented no obstacle for bees to crawl beneath.
Davis and LaBerge (1975) suggested that loose debris protects nests
against beating rain and parasites, and may aid Andrena in nest recogni-
tion.

The amphitheater soil is classified as Sunniland-Bradenton which con-
sists of fine sand down to marl hardpan. It is moderately wet and has a
pH of 5. 0-6. 5. The sand is dark gray and contains fine carbon particles
from fires of previous years. The soil is permeated by a complex network
of roots at a depth of 2 to 5 cm below the surface. These roots, originating
from hammock trees, hindered excavation.

Nest Construction. — The density of nests was determined by counting the
number of nest openings within a 15 cm wire frame, placed at 1 m inter-
vals along a line transect diagonally crossing the amphitheater. An
average of 11 nests per 15 cm area (N = 30) was found, and the nests

Andrena Nesting Biology

143

were slightly more numerous toward the rear of the amphitheater. This
distribution is perhaps influenced by morning sunlight which reaches the
rear area earliest. Early warmth after cool January nights may allow
more foraging time.

Conspicuous, circular tumuli were usually present at nest entrances.
They ranged in diameter from 2.0 cm to 4.3 cm (x = 3.2 cm, N = 30).
Tumuli height ranged from 0.9 cm to 2.0 cm (x = 1.6 cm). Most nests
were clustered in groups of 2 to 5, with many tumuli overlapping. Solitary
nests were seldom found.

Nests were excavated with the aid of plaster of Paris poured into the
entrances. Tunnels descended vertically except to circumvent roots. They
were circular in cross-section, ca. 5.0 mm in diameter, and their walls
consisted of unlined, slightly compacted, moist sand. The main shaft ex-
tended from 28.5 cm to 47.0 cm deep (x = 37.8 cm, N = 46). At the bot-
tom of the vertical tunnel a lateral tunnel ca. 1.5 cm long terminated in a
single cell. Although additional cells may have existed, none was located.
Thorp and Stage (1968) reported difficulty in finding laterals of Andrena
placida which also nests in sand.

Most cells were oriented horizontally to the main tunnel. They were
slightly larger in diameter than the descending shaft (x = 5.7 mm, N =
18) and were lined with a thin, transparent waterproofing material. Cells
were shiny, smooth, and jet-black from soil carbon wetted as the cell was
lined. Microscopic examination showed that individual sand grains were
cemented together by the hardened lining material.

Provisioning. — Pollen balls were spherical, dark yellow, and uniformly
moist. Completed balls ranged in diameter from 2.5 mm to 3.2 mm (x =
2.9 mm, N = 25). As previously noted for other Andrena (Michener and
Rettenmeyer 1956, Linsley and MacSwain 1959, Davis and LaBerge
1975), incomplete balls were smaller, less smooth, and drier than com-
pleted balls.

Slides were prepared of provisions taken from three different nests ca. 3
m apart. Microscopic examination of several hundred pollen grains
proved them to be of a single type, probably Acer rubrum, typically a wind
pollinated species. We never observed bees upon the flowers as it was not
possible to follow them outside the amphitheater and no A. rubrum were in
the immediate area. The bees never flew through the hammock, but
always left or entered the area from above the surrounding trees. As noted
by Graenicher (1930), bees are seldom found in a shaded hammock
interior.

Adult bees were actively foraging when the study site was first visited
on 31 December 1975. A steady progression of individuals left the

144

Beth B. Norden and Aubrey G. Scarbrough

Fig. 2. Prepupa of Andrena miserabilis. a. Larva, lateral view. b. Head, lateral view. c.
Head, frontal view. d. Spiracle, frontal view. e. Spiracle, longitudinal section, f. Right
mandible, ventral view. g. Right mandible, inner view.

amphitheater and returned to the nesting site, usually carrying pollen.
Flight activity of several hundred bees became concentrated in a region
directly above their nests to a height of ca 1.5 meters. Zig-zag flight pat-
terns exhibited within this region by exiting and returning bees suggested
that they were conducting orientation maneuvers. Once above this region
they followed a straight line of flight.

Observations on 1-3 January 1977 were made during cooler tem-
peratures (7-9°C) under overcast skies and during intermittent showers.
Under these conditions activity was noticeably reduced, with little or no
zig-zag flight activity above the nests. Returning bees flew to the vicinity
of their nests, landed, and crawled (seldom directly) to their burrows.
When departing, bees remained on the sand outside their nests for several
minutes, then crawled 3 or 4 cm before taking flight. Males were found
crawling along the ground beside females, and in two instances males oc-
cupied burrows with females during a shower.

Andrena Nesting Biology

145

Description of Prepupa (Fig. 2). — Cream colored, C-shaped, 5.2 mm to
7.3 mm long (N = 15), with prominent transverse, dorsolateral tubercles.
Intersegmental furrows distinct. Head lacking setae, weakly sclerotized;
cleavage lines indistinct. Vertex rounded with two small paramedian con-
vexities. Antennae represented by low convexities arising from large,
rounded prominences. Parietal bands not apparent. Posterior thickening
of head capsules inconspicuous; hypostomal and pleurostomal thicken-
ings more pronounced. Posterior tentorial pits distinct. Epistomal suture
weak; labroclypeal suture distinct. Labrum bearing two small convexities
with clustered sensillae; tubercles absent. Mandibles robust; moderately
sclerotized, most heavily at cusp; apex attenuate and sharply pointed; up-
per apical margin with dense row of teeth; lower margin with row of
smaller denticles; ventral surface with row of denticles near lower margin,
denticles and spicules scattered distally; adductor apodeme twice as long
as abductor apodeme. Maxillae with indistinct separation of cardines and
stipites. Maxillary palpi very pronounced, twice as large as labial palpi.
Labium with prementum and postmentum undefined. Salivary opening a
U-shaped slit. Body naked; spiracular atria extending slightly above body
surface.

Predators and Parasites. — On 2 January 1977, three adult Sphecodes sp.
were noted in the nesting area, two of which were observed leaving
Andrena nest burrows. Members of this genus are frequently nest parasites
of bees (Bohart 1970). Also present at Andrena nest entrances were ants of
the genus Solenopsis, and an adult female Andrena was collected on 3
January 1977 with an ant firmly grasping both right wings (Fig. 3). The
bee crawled from a nest hole, attempted flight, then crawled along the
ground dragging the ant.

Fig. 3. Female Andrena miserabilis with Solenopsis attached to wings.

146

Beth B. Norden and Aubrey G. Scarbrough

On the mornings of 5 and 6 March 1977, tracks and feeding excava-
tions of the Nine-banded armadillo, Dasypus novemcinctus , were found in
the area containing the highest density of Andrena burrows. The armadillo
excavations extended to the depth of larval bee cells and fragments of
opened cells were present. It is doubtful that bee larvae were being sought
by armadillos; but if encountered larvae would likely be consumed, and
nests were destroyed by the digging.

ACKNOWLEDGMENTS. — We gratefully acknowledge the contribu-
tions of the following persons: Drs. W. E. LaBerge, Illinois Natural
History Survey, Urbana; S. W. T. Batra, U.S.D.A., Beltsville, Md.; and
E. S. Deevey, Florida State Museum, Gainesville, for bee, ant, and pollen
identifications respectively; and Dr. R. Shoemaker, M. Brenner, T.
Bladen, R. Franz, and A. Norden for technical assistance.

LITERATURE CITED

Bohart, G. 1970. The evolution of parasitism among bees. 41st Faculty Honor
Lecture, Utah State Univ., Logan. 33 pp.

Davis, L., Jr., and W. LaBerge. 1975. The nest biology of the bee Andrena (Ptilan-
drena) erigeniae Robertson (Hymenoptera: Andrenidae). 111. Nat. Hist. Surv.
Biol. Notes (95) : 1 - 1 6.

Graenicher, S. 1930. Bee-fauna and vegetation of the Miami region of Florida.
Ann. Entomol. Soc. Am. 23:153-174.

Linsley, E., and J. MacSwain. 1959. Ethology of some Ranunculus insects with
emphasis on competition for pollen. Univ. Calif. Publ. Entomol. 76:1-45.
Michener, C., and C. Rettenmeyer. 1956. The ethology of Andrena erythronn with
comparative data on other species (Hymenoptera: Andrenidae). Univ. Kans.
Sci. Bull. 37(16) :645-684.

Osgood, E., Jr. 1972. Soil characteristics of nesting sites of solitary bees associated
with the low-bush blueberry in Maine. Maine Life Sci. Agric. Exp. Stn. Tech.
Bull. 59:1-8.

Rozen, J-.Jr- 1973. Biology notes on the bee Andrena accepta Viereck (Hymenop-
tera, Andrenidae). J. N.Y. Entomol. Soc. 81 { 1 ) : 54-6 1 .

Thorp, R., and G. Stage. 1968. Ecology of Andrena placida with descriptions of the
larva and pupa. Ann. Entomol. Soc. Am. 67(6) : 1 580- 1 586.

Accepted 5 May 1979

SUBSCRIPTIONS AND EXCHANGES

The editors anticipate at least two issues of approximately 150 pages each annually. Rates
for subscriptions for all issues appearing within the calendar year:

Individual — United States $ 7.50

Individual — Foreign $10.00

Institution $12.50

Single issue purchase $ 4.50

All subscriptions must be paid in advance.

Issues will be available on an exchange basis to organizations and institutions publishing
general natural history and ecology journals or papers in a fairly regular schedule. Publica-
tions received on exchange will be placed in the State Museum’s H. H. Brimley Memorial
Library.

Address all subscriptions and requests for information on purchase and exchange to
Managing Editor, Bnmleyana, N. C. State Museum of Natural History, P. O. Box 27647,
Raleigh, NC 27611.

DATE OF PUBLICATION

Bnmleyana No. 1 was mailed on 29 March 1979.

COVER DESIGN

The cover of Bnmleyana was designed by Ray Hodges and Tim Dove.

ERRATA

A few typographical errors were missed in proofing Bnmleyana No. 1, but only one sub-
stantive error has been brought to our attention.

In the essay, “The Brothers Brimley: North Carolina Naturalists,” by John E. Cooper
(p. 8, para. 2, line 8) the statement was made that illustrations by Roger Tory Peterson ap-
peared in the 1919 edition of Birds of North Carolina. Peterson’s illustrations actually first ap-
peared in the 1942 edition. We are grateful to Eloise H. Potter, editor of The Chat , for bring-
ing this error to our attention.

PRODUCTION

Bnmleyana is typeset in Baskerville 10 pt. at photoset, inc., Raleigh, and printed by Con-
temporary Lithographers Incorporated, Raleigh.

REVIEWERS OF MANUSCRIPTS

We are indebted to many scientists who served as referees for manuscripts. Their names
and institutional affiliations will be provided in every third issue.

INDEX AND TABLE OF CONTENTS

An index of scientific names and a table of contents for 1979 will appear in Bnmleyana
No. 3, which will appear in the spring of 1980.

Brimleyana No. 2: 147. November 1979

147

1

I

*

INFORMATION FOR CONTRIBUTORS

Submit original and two copies of manuscripts to Editor, Brimleyana, North Carolina State
Museum of Natural History, P. O. Box 27647, Raleigh, NC 27611. In the case of multiple
authorship, indicate correspondent. Manuscripts submitted for publication in this journal
should not also be submitted elsewhere.

Preparation of manuscript. Adhere generally to the Council of Biology Editors Style
Manual, Fourth Edition. Use medium-weight bond paper, 8V2 X 11”, and leave at least an
inch margin on all sides. Double space all typewritten material.

The first page will be separate and contain the title and the author’s name and address.
Where appropriate, the title will indicate at least two higher categories to which taxa
belong. Example: Studies of the genus Hobbseus Fitzpatrick and Payne (Decapoda: Cam-
baridae).

A brief informative abstract on a separate sheet follows the title page, preceding the text. In-
dicative abstracts are not acceptable. Footnotes will be used only where absolutely
necessary, numbered consecutively throughout the paper.

Individuality of writing style and text organization are encouraged, but for longer papers
the INTRODUCTION, MATERIALS AND METHODS, RESULTS, DISCUSSION
and LITERATURE CITED format is preferable, with those headings centered and
capitalized. Headings plus sub-headings must be kept to a total of three levels.

Scientific names in taxonomic papers will include the author in first usage. Descriptions of
new taxa must be in accordance with the requirements of established international codes.
Etymology is desirable.

Last item in the text will be acknowledgments, with the body of the section preceded
thusly: ACKNOWLEDGMENTS. — Authors should verify that persons mentioned in
acknowledgments acquiesce in the wording.

Appendixes: place after acknowledgments and before literature cited.

Form for literature cited: Author’s last name, first name, middle initial. Year. Title. Jour-
nal (see BIOSIS list of Serials with Title Abbreviations) volume (number) :pages. Provide
total number of pages for books, dissertations, and theses. For second authors use initials
followed by last name. Examples:

Woodall, W. Robert, Jr., and J. B. Wallace. 1972. The benthic fauna in four small southern
Appalachian streams. Am. Midi. Nat. 5S(2):393-407.

Crocker, Denton W. and D. W. Barr. 1968. Handbook of the Crayfishes of Ontario. Univ.
Ontario Press, Toronto. 158 pp.

Authors, not the editor, are responsible for verifying references.

Form for citing references in text: parenthetical (Woodall and Wallace 1972:401), page
numbers optional, following a colon; for more than two authors use et al. (not italicized).

All tables go on separate sheets at the end of the manuscript. Do not use vertical lines in
tables. Indicate lightly in pencil in the margin of the original manuscript where tables and il-
lustrations would best fit.

Preparation of illustrations. Illustrations, including maps, graphs, charts, drawings,
and photographs, should be numbered consecutively as figures. They should not be larger
than 21.5 X 28 cm (8 Vi X 11”). Plates must be prepared and presented as they are to ap-
pear, not as groups or large sheets of items for arrangement by the editors. Do not mount in-
dividual photographs. The author’s name, title of the manuscript, figure number, and the
notation “Top,” should be penciled lightly on the back of every illustration. Lettering on
original drawings and maps should be of adequate size to permit proper reduction where
needed. Do not type on illustrations. Legends should be typed, double-spaced, on separate
sheets. Avoid indicating scale as “X life size.” Consult CBE Style Manual, pp. 39-45, for more
complete guidelines.

Page charges, reprints and proofs. A per page charge of $20 is expected from authors
who have funds available from institutions, grants, or other sources. Those without such
funds should so indicate in their correspondence with the Editor. This will not affect accept-
ance for normal publication. Contributors who pay full page costs will be furnished 100 free
reprints. Reprint order forms will be sent with galley proofs and are to be returned to the
Managing Editor. On papers with more than one author, it will be the responsibility of the
correspondent to assure that other authors have an opportunity to obtain reprints. Proofs
are to be corrected, signed and returned to the Managing Editor within 48 hours.
Changes in proofs other than type corrections will be charged to the author.

CONTENTS

Cretaceous Dinosaurs of North Carolina. Donald Baird

and John R. Horner 1

Revision of Appalachian Trechus (Coleoptera: Carabidae).

Thomas C. Barr, Jr 29

Cretaceous Selachians from the Peedee Formation

(Late Maestrichtian) of Duplin County, North Carolina.

Gerard R. Case 77

Freshwater Triclads (Turbellaria) of North America. XII.

Another New Cave Planarian from North Carolina,

Phagocata carolinensis n. sp. Roman Kenk 91

Freshwater Fishes of Croatan National Forest, North Carolina, with
Comments on the Zoogeography of Coastal Plain Fishes.

Fred C. Rohde, George H. Burgess and G. William Link, Jr 97

Genetic Variation in Cave-dwelling and Deep-sea Organisms,
with Emphasis on Crangonyx antennatus (Crustacea: Amphipoda)
in Virginia. Gary W. Dickson, John C. Patton,

John R. Holsinger and John C. Avise 119

A Photographic Technique to Study Tadpole Populations.

C. Kenneth Dodd, Jr 131

Rediscovery of the Sharphead Darter, Etheostoma acuticeps,
in North Carolina (Pisces: Percidae). Richard T. Bryant,

James P. Beets and Michael G. Ryon 137

Nesting Biology of Andrena ( Larandrena ) miserabilis Cresson

and Description of the Prepupa (Hymenoptera: Andrenidae).

Beth B. Norden and Aubrey G. Scarbrough 141

Errata and Miscellany 147