<71H

71

,/sNEr"

■- < v ■ -V.- > .: A. •:•

■

■Mh ■'■ '■

'

|M|£

A.

^ i

« | '■ - S

..

•••

‘A

Z/jZri

r

AA

p

! V.-: w 1

'.o' ’ - >

m

§1

J?S|:

.

A:;;v

-■ ■

mm

S»sSm i

AAv

'

.

> vv\e

5eV^°

r>/iA

Sti!

g

If

lit:

i

IS

II

■

t§s§

k

»iSk

.1 f

4;

ii\

; A | >

k

IlllilSl

<

JUL 3 1 1990

Ml

FIELD MUSEUM LIBRARY

14

I

number 16

july 1990

2 4 ■<>'> / » ?'»• . l"v -V > A A '* « <£ *• ‘ ^ A.

EDITORIAL STAFF

Frank J. Radovsky, Editor

Eloise F. Potter, Managing Editor

Sheree Worrell, Production Manager

RESEARCH CURATORIAL STAFF

North Carolina State Museum of Natural Sciences

Alvin L. Braswell William M. Palmer

Curator, Lower Vertebrates Curator of Lower Vertebrates

Mary K. Clark

Curator of Mammals

Frank J. Radovsky

Director, Research and Collections

John A. Gerwin

Curator, Birds

Rowland M. Shelley

Curator of Invertebrates

David S. Lee

Curator of Birds

Vincent P. Schneider

Research Associate, Paleontology

Brimleyana, the Journal of the North Carolina State Museum of Natural

Sciences, appears at irregular intervals in consecutively numbered issues. Contents

emphasize zoology of the southeastern United States, especially North Carolina

and adjacent areas. Geographic coverage is limited to Alabama, Delaware,

Florida, Georgia, Kentucky, Louisiana, Maryland, Mississippi, North Carolina,

South Carolina, Tennessee, Virginia, and West Virginia.

Subject matter focuses on systematics, evolution, zoogeography, ecology,

behavior, and paleozoology. Papers stress the results of original empirical field

studies, but synthesizing reviews and papers of significant historical interest to

southeastern zoology are included. Brief communications are accepted.

Appropriate specialists review each manuscript, and final acceptability is

determined by the Editor. Address manuscripts and all correspondence (except

that relating to subscriptions and exchanges) to Editor, Brimleyana, North

Carolina State Museum of Natural Sciences, P.O. Box 27647, Raleigh, NC

27611.

Address correspondence pertaining to subscriptions, back issues, and exchanges

to Brimleyana Secretary, N.C. State Museum of Natural Sciences, P.O. Box

27647, Raleigh, NC 27611.

In citations please use the full name — Brimleyana.

North Carolina State Museum of Natural Sciences

North Carolina Department of Agriculture

James A. Graham, Commissioner

CODN BRIMD 7

ISSN 0193-4406

First Record of the Rock Vole,

Microtus chrotorrhinus (Miller)

(Rodentia: Cricetidae), in Virginia

John F. Pagels

Department of Biology

Virginia Commonwealth University

Richmond, Virginia 23284

ABSTRACT. — A rock vole, Microtus chrotorrhinus, collected in

Bath Co., Va., represents the first record of this species in the state.

The collection site was at 1,036 m on Allegheny Mountain in George

Washington National Forest. The site was characterized by talus and

supported a northern hardwood forest.

A subadult male rock vole, Microtus chrotorrhinus (Miller), collected

in Bath Co., Va., on 28 July 1987, represents the first record to be

published of this species in the Commonwealth. The collection site, at

1,036 m elevation on the eastern slope of Allegheny Mountain (38° 14'

N, 79° 49' W) in the George Washington National Forest, is 16 km N of

Mountain Grove. The rock vole was captured on a west-facing slope

about 25 m above the channel of Lightner Run, a tributary of Little

Back Creek. The nearest of several West Virginia localities where M.

chrotorrhinus has been collected is about 32 km from this site (Kirkland

1977a, 1977b).

Forty Sherman live traps were set among talus and boulders in a

small area approximately 60 m by 60 m, 26-28 July and 12-15 August,

1987, for a total sampling effort of 240 trap nights. Traps were baited

with a mixture of rolled oats and peanut butter scented with oil of anise.

The rock vole (Virginia Commonwealth University Mammal Collection

No. 5006) was captured in a trap set within the talus with the opening of

the trap directed under a small rock overhang. Other species captured in

the same area were red-backed vole, Clethrionomys gapperi (Vigors) (13

specimens); short-tailed shrew, Blarina brevicauda (Say) (7); cloudland

deer mouse, Peromyscus maniculatus nubiterrae Rhoads (3); and masked

shrew, Sorex cinereus Kerr (1). Eight pitfall traps set along the stream

captured S. cinereus (4); smoky shrews, S. fumeus Miller (3); and rock

shrew, S. dispar Batchelder (1). These small mammals are some of the

same species reported as habitat associates of M. chrotorrhinus in West

Virginia (Kirkland 1977a, 1977b). In New York, Kirkland and Knipe

(1979) also captured the water shrew, Sorex palustris Hooper, with M.

chrotorrhinus. The only known Virginia locality for S. palustris is a site

along Little Back Creek only 1.6 km from Lightner Run (Pagels and

Tate 1976, Pagels 1987).

Brimleyana 16:1-3, July 1990

1

2

John F. Pagels

The capture site was characterized by talus and supported a northern

hardwood forest. Sugar maple, Acer saccharum, composed nearly 50%

of the trees 10 cm in diameter or greater within 100 m of the collection

site. Remaining trees were principally yellow birch, Betula lutea (16%);

black birch, B. nigra (10%); and basswood, Tilia americana (10%).

Scattered throughout the sampling area were small numbers of American

beech, Fagus grandifolia; northern red oak, Quercus rubra’, and hickory,

Carya sp. No conifers were observed. Understory consisted primarily of

mountain maple, Acer spiratum; witch-hazel, Hamamelis virginiana ;

and saplings of A. saccharum. The most prominent forbs were white

snakeroot, Ageratum altissima, and Dutchman’s pipe, Aristolochia

macrophylla. Rocks and talus were generally moss-covered, and stumps

and fallen trees in various stages of decomposition were prevalent.

Severe drought conditions prevailed in much of Virginia, including most

montane areas, during summer 1987, and the rocky channel of Lightner

Run was nearly dry on the collection date. Long-term climatic data

from nearby Marlinton, W.Va., are given by Pagels and Tate (1976).

Based on studies of M. chrotorrhinus throughout its range (see

Kirkland and Jannett 1982), the following observations can be noted.

(1) Boulders, talus, or rocks, as the name rock vole so aptly indicates,

are important features of M. chrotorrhinus habitat. (2) The rock vole

exploits subterranean portions of its rocky environment; subsurface

runways are observed and captures in subsurface sets are common. (3)

Water, whether surface or subsurface, is an important component of

rock vole habitat. (4) Microtus chrotorrhinus is most often associated

with a suite of small mammals having northern affinities, and in the

southern Appalachian Mountains appropriate habitat for all such species

is present only at relatively high elevations. (5) Although vegetation

associations are variable with respect to the rock vole’s habitat preferences

in various parts of its range, tree species that predominate at a given

collecting site in the southern Appalachian Mountains are kinds with

northern affinities, for example, red spruce (Picea rubens ), sugar maple,

and yellow birch.

Existing habitat in the southern Appalachian Mountains that appears

suitable for the rock vole is highly fragmented as a result of both

natural forces and human activities, especially burning and timbering

(Handley 1980). Opportunity for repopulation of sites from which M.

chrotorrhinus has been extirpated seems negligible because of the lack

of avenues of suitable habitat. If M. chrotorrhinus is found at other

locations in Virginia, these undoubtedly will be protected rocky and/or

talus sites that have remained relatively moist throughout historical

time. Certain factors provide a moderately optimistic outlook for the

Rock Vole in Virginia

3

rock vole in Virginia: (1) the healthy status of M. chrotorrhinus in

nearby West Virginia and (2) the presence in Virginia of apparently

suitable habitat at some high elevation sites, despite the discontinuity of

those sites.

ACKNOWLEDGMENTS. — Collection of the rock vole was made

ancillary to a study on the distribution of the northern flying squirrel

sponsored by the Nongame Wildlife and Endangered Species Program

of the Virginia Department of Game and Inland Fisheries. I am grateful

to R. W. Duncan, M. Fies, and K. Terwilliger for their support in that

effort. I thank R. Glasgow of George Washington National Forest for

his support and encouragement. Station manager B. Bocchicchio, who

provided access to the site, and Sara S. Bell, C.M. Kershner, and people

at the security station of the Virginia Power Bath County Pumped

Storage Station are gratefully acknowledged for many acts of kindness.

I am especially grateful to Ms. Bell for much assistance both before and

after the collection was made. I thank Dr. C. O. Handley, Jr., for

verifying the identification of the vole and Dr. M. F. Johnson for

identification of plant species. Dr. G. L. Kirkland, Jr., graciously

provided many helpful comments on an earlier draft of the manuscript.

LITERATURE CITED

Handley, C. O., Jr. 1980. Rock Vole. Microtus chrotorrhinus carolinensis

Komarek. Pages 574-577 in Threatened and Endangered Plants and Animals

of Virginia, D. W. Linzey, editor. Ctr. Environ. Stud., Va. Polytech. Inst,

and State Univ.

Kirkland, G. L., Jr. 1977a. The rock vole, Microtus chrotorrhinus (Miller)

(Mammalia: Rodentia) in West Virginia. Ann. Carnegie Mus. Nat. Hist.

46:45-53.

Kirkland, G. L., Jr. 1977b. Responses of small mammals to the clearcutting of

northern Appalachian forests. J. Mammal. 58:600-609.

Kirkland, G. L., Jr., and F. J. Jannett, Jr. 1982. Microtus chrotorrhinus.

Mammalian Species. 180:1-5.

Kirkland, G. L., Jr., and C. M. Knipe, 1979. The rock vole {Microtus

chrotorrhinus ) as a Transition Zone species. Can. Field-Nat. 93:319-321.

Pagels, J. F. 1987. The pygmy shrew, rock shrew and water shrew: Virginia’s

rarest shrews (Mammalia: Soricidae). Va. J. Sci. 38:364-368.

Pagels, J. F., and C. M. Tate. 1976. Shrews (Insectivora: Soricidae) of the

Paddy Knob-Little Back Creek area of western Virginia. Va. J. Sci. 27:202-

203.

Accepted 11 October 1988

Digitized by the Internet Archive

in 2017 with funding from

BHL-SIL-FEDLINK

https://archive.org/details/brimleyana161990

Occurrence of a Northern Cicada, Okanagana rimosa

(Homoptera: Cicadidae), in the Southern Appalachians

E. E. Brown

Box 343, Davidson, North Carolina 28036

AND

J. D. Brown

United States Forest Service

504 Justis Drive, Greeneville, Tennessee 37743

ABSTRACT. — The cicada Okanagana rimosa (Say) is reported from

five mountain counties in North Carolina and one in Tennessee. This

represents a range extension southward of ca. 300 miles (485 km) from

Rappahannock Co. in northern Virginia.

Virtually all of the relatively large cicadas of the eastern United

States are members of the genus Tibicen. Although smaller than most of

its congeners, Tibicen canicularis (Harris) is one of the characteristic

forms in the Northeast, and it perhaps extends farther north than any

other member of the genus. Okanagana is a large genus (several dozen

species) of mostly smaller western and southwestern forms. However,

according to Davis (1930), Okanagana rimosa (Say) and O. canadensis

(Provancher) are noteworthy in being distributed across much of

southern Canada and spilling down into the border states of the United

States. The distribution of O. canadensis appears to be slightly the more

northerly of the two. The most southerly eastern records for O. rimosa

have been in Pennsylvania, northeastern Ohio, and, as a helpful reviewer

reminds us, a specimen collected by Allard (1938) at about 3,500 feet

(1,070 m) on Mary’s Rock, Rappahannock Co., in northern Virginia.

For a number of years, one of us (E.E.B.) has had an interest in

cicadas. He was aware of the statement by Davis (1922) that, from an

elevation of 3,900 feet (1,190 m) on Bald Knob, Bath Co., in western

Virginia, a friend had brought to him a headless specimen apparently of

T. canicularis (Harris). Considering the numerous northern organisms

that range southward in the mountains, he surmised that the ranges of

the small northern forms O. rimosa and T. canicularis might extend

farther south in the Appalachians than had been reported. Consequently,

he searched along the Blue Ridge Parkway, then being unaware that his

hearing was not adequate to have detected the call of O. rimosa. When

J.D.B. relocated to Marion, N.C. (Grandfather District, Pisgah National

Forest), E.E.B. asked him to be alert for cicadas, especially small ones.

Brimleyana 16:5-7, July 1990

5

6

E. E. Brown and J. D. Brown

On 14 July 1983, J.D.B. collected a small, newly transformed

cicada in a patch of white pines near old N.C. Hwy. 105 (McDowell

Co.), which runs along the ridge just west of Linville Gorge and Table

Rock. On later examination, this specimen appeared to be a female of

O. rimosa.

On 22 July we went back up old N.C. Elwy. 105 (SR 1238). The call

of the cicada was a fine, long buzz that J.D.B. could hear but that the

older ears of E.E.B. failed to pick up then, though he later heard it

easily with Bionic Ear® equipment. Beginning about 7 miles (11.3 km)

north of the Canal Bridge on Lake James, J.D.B. could hear calls at

numerous stops along the road (among mixed pines and hardwoods)

over a stretch of about 5 miles (8 km) at an altitude of 2,800-3,300 feet

(855-1,005 m). We found two small, conspicuously banded nymphal

skins, both on young maples (Acer pennsylvanicum) among larger white

pines.

On 9 August J.D.B. hand-caught a male specimen about 1 foot (0.3

m) above the ground on a young white pine, just off Forest Service

Route 496, 3.2 miles (5.1 km) S of N.C. Elwy. 181), on the E side of

Table Rock-Sitting Bear (Elawksbill) Ridge at about 2,800 feet (855 m)

(Burke Co.). In the same area the next day, using a .22 caliber shotshell,

we collected another male. This specimen was about 18 feet (5.5 m)

above the ground, on the underside of, and far out on, a white pine

limb. It was nearly invisible in shadow and against a dark blotch of

bark.

The cicadas identified as O. rimosa agree with: (a) Heath’s (1978)

characterization of Okanagana, with narrow head, widely separated and

exposed tymbals, non-retractable uncus, etc.; (b) Say’s original de-

scription of O. rimosa as quoted by Davis (1919), including orange-

rufous markings on pronotum, mesonotum, edges of abdominal terga,

and bases of wings; (c) Alexander’s (1961) provisional key to eastern

species of cicadas; (d) Moore’s (1966) notations regarding the species in

Michigan; and (e) Davis’s (1926) notes touching the northeastern region

of the country.

Since 1983, in addition to the continued occurrence of the species

at the localities noted above, J.D.B. has heard specimens at several

other sites.

1984: First noted 12 June, Burke Co. Later heard on the ridge S of

Roseboro (Block #135) in Avery Co.

1985: First heard 11 June. Heard 25-28 June in the pines on the

Singecat Ridge area, N of Sunnyvale, in McDowell Co.

Northern Cicada in Southern Appalachians

7

1986: Heard 5 June on Joe White Mountain, just E of Mortimer,

Caldwell Co. Heard during July on Flat Top Mountain, near the

Tennessee line in Yancey County.

1987: On 1 May, heard on a knob S of Meadow Creek Mountain, in

Cocke Co., Tenn. During June it was heard at several other points

in Cocke Co.

These appear to be the first records of O. rimosa for the Carolina

region. They thus extend its known range some 300 miles (485 km)

southward from Rappahannock Co., Va., to the central mountain

counties of North Carolina.

Evidently, we have in the mountains not just an isolated brood or

two of this cicada, but a moderately widespread population, especially

at intermediate elevations where pines are present. Presumably it will

show up in some other mountain counties. Although present in

considerable numbers, it definitely is not a conspicuous form and must

be sought after.

Specimens will be deposited in the collections of the North Carolina

Department of Agriculture.

LITERATURE CITED

Alexander, R. D. 1961. Key to species of cicadas occurring in the United

States east of the Mississippi River. 3rd revision. 10 pp., mimeo. Distributed

by author.

Allard, H. A. 1938. Notes on some cicadas in Virginia and West Virginia. J.

N.Y. Entomol. Soc. 46:449-452.

Davis, W. T. 1919. Cicadas of the genera Okanagana, Tibicinoides and

Okanagodes, with descriptions of several new species. J. N.Y. Entomol.

Soc. 27:179-225.

Davis, W. T. 1922. An annotated list of the cicadas of Virginia with

description of a new species. J. N.Y. Entomol. Soc. 30:36-53.

Davis, W. T. 1926. The cicadas or harvest flies of New Jersey. N.J. Dept.

Agric., Circ. No. 97.

Heath, M. S. 1978. Genera of American cicadas north of Mexico. Ph.D.

dissertation, Univ. Florida.

Moore, T. E. 1966. The cicadas of Michigan (Homoptera: Cicadidae). Pap.

Mich. Acad. Sci., Arts & Lett. 51:75-96.

Accepted 24 October 1988

8

ATLAS OF NORTH AMERICAN FRESHWATER FISHES

by

D. S. Lee, C. R. Gilbert, C. H. Hocutt, R. E. Jenkins,

D. E. McAllister, J. R. Stauffer, Jr., and many collaborators

This very useful book provides accounts for all 777 species of fish

known to occur in fresh waters in the United States and Canada. Each

account gives a distribution map and illustration of the species, along

with information on systematics, distribution, habitat, abundance, size,

and general biology.

“. . . represents the most important contribution to freshwater fishes

of this continent since Jordan and Evermann’s ‘Fishes of North and

Middle America’ over 80 years ago.” — Southeastern Fishes Council

Proceedings.

1980 825 pages Index Softbound ISBN 0-917134-03-6

Price: $25, postpaid, North Carolina residents add 5% sales tax. Please make

checks payable in U.S. currency to NCDA Museum Extension Fund.

Send to FISH ATLAS, N.C. State Museum of Natural Sciences,

P.O. Box 27647, Raleigh, NC 27611.

ATLAS OF NORTH AMERICAN FRESHWATER FISHES

1983 SUPPLEMENT

by

D. S. Lee, S. P. Platania, and G. H. Burgess

The 1983 supplement to the 1980 Atlas of North American

Freshwater Fishes treats the freshwater ichthyofauna of the Greater

Antilles. In addition to this bound supplement, there are 19 accounts,

mostly species not described in 1980, in looseleaf form to be added to

the 1980 volume. Illustrated by Renaldo Kuhler.

1983 67 pages Index Softbound ISBN 0917134-06-0

Price: $5, postpaid. North Carolina residents add 5% sales tax. Please make

checks payable in U.S. currency to NCDA Museum Extension Fund.

Send to FISH ATLAS, N.C. State Museum of Natural Sciences,

P.O. Box 27647, Raleigh, NC 27611.

Distribution and Ecology of the Blackside Dace,

Phoxinus cumber landensis (Osteichthyes: Cyprinidae)

Christopher J. O’Bara

Tennessee Cooperative Fishery Research Unit

Department of Biology

Tennessee Technological University

Cookeville, Tennessee 38505

ABSTRACT. — A recent status survey of the blackside dace, Phoxinus

cumber landensis, resulted in nine new distributional records and a

better understanding of the species ecology. This threatened species is

now known from 30 streams; it inhabits approximately 27.0 km of

small headwater streams within the Upper Cumberland River basin.

Impacts from coal-mining activities have resulted in the loss of the

species from seven previously reported localities. Removal of riparian

vegetation and increased siltation appear to be the primary degrading

factors.

The blackside dace, Phoxinus cumberlandensis, is a rare cyprinid

restricted to small streams of the Upper Cumberland River basin of

Tennessee and Kentucky. The species was first discovered in 1975 and

was described by Starnes and Starnes (1978a). A status survey in 1978

recorded the species from 12 sites (Starnes and Starnes 1978b).

Additional localities were reported by Starnes (1981), Warren (1981),

and Burr and Warren (1986). The biology of the species was described

by Starnes and Starnes (1981). The species currently is listed as a

threatened species by the U.S. Fish and Wildlife Service.

Phoxinus cumberlandensis has been reported only from the Upper

Cumberland River basin. This has been confirmed by extensive surveys

in adjacent basins: Kentucky River (Branson and Batch 1981, 1983,

1984), Tradewater and Green rivers (Warren and Cicerello 1982), Big

South Fork River (O’Bara and Estes 1984), Little South Fork River

(Branson and Schuster 1982), and Powell River (Tennessee Valley

Authority 1975). In addition, Harker et al. (1979, 1980) reported on

surveys of Tygarts Creek and the Kentucky, Little Sandy, Licking, and

Upper Cumberland river basins.

The main purpose of the present study was to determine the current

status of P. cumberlandensis. Potential threats and habitat requirements

were also investigated.

STUDY AREA

The Upper Cumberland River basin has been traditionally defined

as that section upstream of Cumberland Falls (McGrain 1966). Although

Brimleyana 16:9-15, July 1990

9

10

Christopher J. O’Bara

no definitive geological records exist, Cumberland Falls is believed to

have been originally located near Burnside, Ky. The present falls are

situated approximately 72 km upstream of this presumed origin.

The entire Upper Cumberland River basin is within the Appalachian

Plateau Physiographic Province. Three physiographic sections lie within

the basin. The Cumberland Mountains section consists of two parallel

ridges with altitudes ranging from 600 to 1,295 meters. Headwater

streams in this region are generally of steep gradient. The Kanawha

section is a dissected plateau characterized by narrow valleys and

moderate-gradient streams. The Cumberland Plateau section is a broad

plateau of moderate relief. The Pottsville escarpment along the western

edge of the plateau contains steep-gradient streams (McGrain 1966).

METHODS AND MATERIALS

Sampling was conducted from October 1984 to August 1985.

Collections were made by using seines and backpack electrofishing

equipment. The habitat was characterized qualitatively at each site. The

analyzed characteristics included substrate, embeddedness, riparian

vegatation, stream morphometry, watershed land use, and canopy cover.

The approximate length of suitable habitat based on the previously

mentioned characteristics was measured upstream and downstream

from a known population location. Healthy populations were defined as

those that contained two or more year classes and numerically composed

more than 25% of the fish community.

RESULTS AND DISCUSSION

Of the 193 sites sampled, 30 contained P. cumberlandensis (Table

1). It appears that P. cumberlandensis has been extirpated from nine

streams (Table 2). Nine new populations were discovered (Table 1), but

only the one in Bucks Branch was considered healthy. An estimated

27.0 km of stream were inhabited by P. cumberlandensis , but only 13.0

km supported healthy populations.

Five population clusters were found (Table 1). Downstream of

Cumberland Falls, five streams were inhabited by P. cumberlandensis.

The other four population clusters were found in the Straight Creek

system (six populations), in the Jellico Creek system (five populations),

in the Clear Fork system (five populations), and in four small streams

that drain directly into the Cumberland River. These population clusters

are extremely important to the continued existence of the species. If

conditions improve in adjacent streams, natural recolonization is likely

to occur as a result of the fairly large numbers of P. cumberlandensis in

these population clusters. Reinvasion was documented in the Straight

Table 1. Location, number of individuals collected, and approximate habitat

length for all observed Phoxinus cumber landensis populations.

12

Christopher J. O’Bara

Creek system with the expansion of P. cumber landensis into four

streams that were devoid of the species when sampled by Starnes and

Starnes (1978b).

Five populations (Poor Fork, Brownies Creek, Davis Creek, Little

Clear Creek, Trammel Fork) are isolated and appear to be in jeopardy

because of low numbers or inadequate adjacent habitat. These

populations could be lost as a result of a single catastrophic event, and

natural recolonization would be extremely unlikely because of their

isolation.

Fish species typically found in association with P. cumberlandensis

included creek chub ( Semotilus atromaculatus), white sucker ( Catosto -

mus commersoni), blacknose dace ( Rhinichthys atratulus), stripetail

darter ( Etheostoma kennicotti), and central stoneroller ( Campostoma

anomalum). Some or all of these five species were generally found in

each stream containing P. cumberlandensis.

Gradient is believed to influence the distribution of P. cumberland-

ensis significantly. In streams of the Cumberland Mountains, the

gradient is usually high, which results in large populations of R.

atratulus. Low-gradient streams (Kanawha section) did not appear to

provide suitable hatitat for P. cumberlandensis. A 60:40 riffle/ pool ratio

appears to be preferred by P. cumberlandensis . I conclude that 42 of the

193 systems sampled did not historically support populations of P.

cumberlandensis owing to inadequate habitat (too high or low gradient,

streams too large, etc.), whereas the remaining 151 systems have adequate

habitat to sustain the species.

Habitat degradation resulting from human activities is the cause of

the apparent decline of P. cumberlandensis. Major degrading activities

are coal mining and associated disturbances (site preparation, road

maintenance). During the survey, mining occurred in only one of the 30

watersheds in which the blackside dace was found. That watershed,

Sims Fork, was impacted not directly by run-off from the coal mine,

but primarily by poor road construction and bridges not adequate to

enable the passage of large coal trucks, thus resulting in trucks passing

directly through the stream. Coal mining had occurred in five of the

other watersheds, but had been stopped because of the current economic

state of the coal industry. The absence of blackside dace from 101 of the

151 systems with theoretically adequate habitat for the species could

have resulted from coal-mining activities. The remaining 20 systems

with appropriate gradient but not inhabited by P. cumber lanensis have

been degraded by agriculture, road construction, impoundments, or

poor forestry practices. Extirpated populations were impacted by coal

mining (6), road construction (1), agriculture (1), and drought (1).

Blackside Dace 13

Table 2. Streams previously reported to contain Phoxinus cumber landensis

and from which the species apparently has been extirpated.

Coal mining, either by direct runoff or by secondary disturbances

such as poor or inadequate roads and bridges and mine preparation and

maintenance, appears to affect two important physical components of

the blackside dace habitat, riparian vegetation and substrate. First,

riparian vegetation, consisting of hemlock, rhododendron, ironwood,

river birch, and sycamore, with canopy cover exceeding 70 percent of

the stream is important. The significance of a natural, undisturbed

riparian vegetational zone appears to be two-fold: in preventing elevated

water temperature due to solar heating, and in providing submerged

root systems that appear to be preferred cover. Healthy populations had

access to extensive undercut, rooted banks. The riparian vegetational

zone is often removed or significantly reduced during mine preparation

and in road and bridge construction. Second, it is extremely important

that the substrate consist of a cobble-gravel mix in riffles, a bedrock-

boulder-silt combination in pools, and silt-free areas just downstream of

the riffles. Coal-mining activities increased siltation, thus degrading this

preferred substrate.

CONCLUSIONS

The continued existence of the rare and threatened P. cumberland-

ensis appears to depend on a number of factors. Viable population

clusters and improved habitat in adjacent streams are necessary for

natural recolonization to occur. Further physical isolation of populations

would only further jeopardize this species. It is to be hoped that

protection under stringent federal and state regulations will improve the

possibility for the continued existence of P. cumber landensis.

14

Christopher J. O’Bara

ACKNOWLEDGMENTS.— I thank D. L. Swann (Tennessee

Technological University) and R. G. Biggins (U.S. Fish and Wildlife

Service) for their assistance in field collections. J. B. Layzer, F. J.

Bulow, A. G. Bailey, D. A. Etnier, and R. R. Cicerello made helpful

comments on the manuscript. The study was funded by the U.S. Fish

and Wildlife Service-Endangered Species Office. The Tennessee Cooper-

ative Fishery Research Unit is funded jointly by the U.S. Fish and

Wildlife Service, Tennessee Technological University, and the Tennessee

Wildlife Resources Agency.

LITERATURE CITED

Branson, B. A., and D. L. Batch. 1981. Fishes of the Dix River, Kentucky.

Tech. Rep. No. 2. Ky. Nat. Preserves Comm., Frankfort.

Branson, B. A., and D. L. Batch. 1983. Fishes of the South Fork of the

Kentucky River, with notes and records from other parts of the drainage.

Southeast. Fishes Council Proceed. 4:1-15.

Branson, B. A., and D. L. Batch. 1984. Fishes of the Middle Fork of the

Kentucky River, Kentucky. Southeast. Fishes Council Proceed. 4:4-9.

Branson, B. A., and G. A. Schuster. 1982. The fishes of the wild river section

of the Little South Fork of the Cumberland River, Kentucky. Trans. Ky.

Acad. Sci. 43:60-70.

Burr, B. M., and M. L. Warren, Jr. 1986. A Distributional Atlas of Kentucky

Fishes. Tech. Rep. No. 4. Ky. Nat. Preserves Comm., Frankfort.

Harker, D. F., Jr., S. M. Call, M. L. Warren, Jr., K. E. Camburn, and P.

Wigley. 1979. Aquatic biota and water quality survey of the Appalachian

Province, eastern Kentucky. Tech. Rep. Ky. Nat. Preserves Comm.,

Frankfort.

Harker, D. F., Jr., M. L. Warren, Jr., K. E. Camburn, S. M. Call, G. J. Fallo,

and P. Wigley. 1980. Aquatic Biota and water quality survey of the Upper

Cumberland River basin. Tech. Rep. Ky. Nat. Preserves Comm., Frankfort.

McGrain, P. 1966. Geology of the Cumberland Falls State Park area. Univ.

Kentucky, Ky. Geol. Surv., Series X. Spec. Publ., Lexington.

O’Bara, C. J., and R. D. Estes. 1984. Recent collections of fishes from the Big

South Fork of the Cumberland River system, Tennessee and Kentucky.

Southeast. Fishes Council Proceed. 4:12-16.

Starnes, W. C. 1981. Listing package for the blackside dace, Phoxinus

cumberlandensis. U.S. Fish Wildl. Serv., Asheville, N.C.

Starnes, W. C., and L. B. Starnes. 1978a. A new cyprinid of the genus

Phoxinus endemic to the Upper Cumberland River drainage. Copeia

1978:508-516.

Starnes, W. C., and L. B. Starnes. 1978b. Status report on a new and

threatened species of Phoxinus from the Upper Cumberland River drainage.

Southeast. Fishes Council Proceed. 2:1-3.

Blackside Dace

15

Starnes, W. C., and L. B. Starnes. 1981. Biology of the blackside dace,

Phoxinus cumberlandensis. Am. Midi. Nat. 106:360-370.

Tennessee Valley Authority. 1975. Powell River basin: fish, bottom fauna, and

aquatic habitat. Tennessee Valley Authority, Norris, Tenn.

Warren, M. L., Jr. 1981. New distributional records of eastern Kentucky

fishes. Brimleyana 6:129-140.

Warren, M. L., Jr., and R. R. Cicerello. 1982. New records, distribution, and

status of ten rare fishes in the Tradewater and lower Green Rivers,

Kentucky. Southeast. Fishes Council Proceed. 3:1-7.

Accepted 1 February 1989

16

ENDANGERED, THREATENED, AND

RARE FAUNA OF NORTH CAROLINA

PART II.

A RE-EVALUATION OF THE MARINE AND

ESTUARINE FISHES

by

Steve W. Ross, Fred C. Rohde, and David G. Lindquist

This is the second in a series of reports by committees appointed in

1985 by the North Carolina State Museum of Natural Sciences to re-

evaluate the faunal lists presented in Endangered and Threatened Plants

and Animals of North Carolina (John E. Cooper, Sarah S. Robinson,

and John B. Funderburg, editors. N.C. State Mus. Nat. Hist., Raleigh,

1977), which is now out of print. The report on marine and estuarine

fishes by Ross, Rohde, and Lindquist treats one Endangered species, six

Vulnerable species, and four anadromous fishes that, while not formally

listed, are of some concern. Five species listed as being of Special

Concern in 1977 no longer warrant formal status. The publication

includes six original drawings by Renaldo Kuhler.

1988 20 pages Softbound ISBN 0-917134-17-6

Price: S3 postpaid. North Carolina residents add 5% sales tax. Please make checks

payable in U.S. currency to NCDA Museum Extension Fund.

Send order to: ETR MARINE FISHES, N.C. State Museum of Natural Sciences,

P.O. Box 27647, Raleigh, NC 2761 1.

Oviposition, Larval Development, and Metamorphosis in

the Wood Frog, Rana sylvatica (Anura: Ranidae), in Georgia

Carlos D. Camp, Charles E. Condee,

and D. Glenn Lovell

Department of Biology, Piedmont College

Demorest, Georgia 30535

ABSTRACT. — Oviposition and development in the wood frog, Rana

sylvatica , were investigated from February through July of 1987 in the

upper Piedmont of Georgia. Egg masses were laid in February and

March in groups ranging from 1 to 22 masses. The number of eggs

averaged 553 per mass. Mean egg diameter was 2.8 mm. Larvae, which

averaged 8.7 mm in total length, hatched 18 to 25 days after oviposition.

Larvae had reached a mean total length of 51.6 mm when they

metamorphosed 1 15 to 130 days after hatching. Newly metamorphosed

froglets had a mean snout-vent length of 18 mm.

The wood frog, Rana sylvatica LeConte, ranges from the tundra in

Alaska and Canada to upland areas of Georgia (Martof 1970) and

Alabama (Mount 1975). Various aspects of the life history of this frog

have been investigated in a number of localities (Martof and Humphries

1959, Herreid and Kinney 1966, 1967, Meeks and Nagel 1977, Howard

1980, Berven 1982a, b, Seale 1982, Seigel 1983). Davis and Folkerts

(1986) recently studied the life history of the wood frog in Alabama but

did not report specific data on larval development time or size at

metamorphosis. The only previous study describing larval development

in the southern portion of the wood frog’s range was by Meeks and

Nagel (1977) in northeastern Tennessee. We present here an account of

development in the wood frog in northeastern Georgia.

MATERIALS AND METHODS

This study was conducted in the Piedmont of Habersham Co., Ga.,

from February through July, 1987. Study areas were small, rain-filled,

temporary pools located at Nancytown Lake (NTL; 3.2 km S of Mt.

Airy, elevation = 280 m) and Roger’s Creek (RC1 and RC2; 6.4 km S of

Batesville, 450 m). Eggs were measured, and egg masses were counted

from an additional site on the Soque River (SR; 8.0 km W of Demorest,

350 m). NTL, RC2, and SR were woodland pools, whereas RC1 was

located in a pasture approximately 200 m from RC2. NTL and RC1

had abundant macrophytic vegetation; RC2 was heavily shaded and had

little vegetation; SR was a deep wheel rut in an old logging road and

contained no vegetation.

Brimleyana 16:17-21, July 1990

17

18

C. D. Camp, C. W. Condee, and D. G. Lovell

Dates of oviposition were recorded, and egg masses were counted

in the study sites and in additional breeding sites that were discovered

subsequently. Sample egg masses were removed and preserved in 10%

formalin. Eggs were counted in eight separate collected masses, and

samples (at least 10 eggs per mass) were taken from six of these masses

for determination of egg size. The diameters of both eggs and outer jelly

envelopes were measured to the nearest 0. 1 mm using a vernier caliper.

Egg masses at NTL, RC1, and RC2 were monitored daily until

hatching was complete. Dates when hatching began and ended were

recorded. Midday readings of water and air temperatures were taken at

NTL and RC1. Maximum depth was determined daily at all three sites.

Ten larvae were preserved at hatching in 10% formalin, measured for

total length to the nearest 0.1 mm, and staged (after Gosner 1960) with

the aid of a dissecting microscope.

A series of tadpoles was collected at RC2 as larvae neared meta-

morphosis (stages 39-41 of Gosner 1960), preserved in 10% formalin,

and measured for total length to the nearest 0.1 mm.

A partial drift fence (after Gibbons and Semlitsch 1982) was

constructed along one side of RC2 in order to collect newly meta-

morphosed individuals. Tadpoles were observed two or three times per

week until some of them began developing hind legs, at which time daily

monitoring of both the edge of the pool and the drift fence began.

Monitoring continued until metamorphosis and dispersal from the pool

were complete. Body lengths (snout-vent) of newly metamorphosed

froglets were measured to the nearest 0.5 mm in the field using a ruler,

and the animals were then released on the opposite side of the fence.

RESULTS

Eggs were first discovered at NTL on 12 February 1987, and newly

laid clutches appeared on 13 February. Adult male wood frogs were

collected at the site on 12 and 23 February. A total of nine egg masses

was located at NTL. These were deposited as individual masses scattered

throughout an area covering approximately 35 m2. Eggs were deposited

in RC1 (22 masses) and RC2 (18 masses) from mid-February to 2

March, with most oviposition events occuring from 20 to 24 February.

Most egg masses in these ponds were deposited as communal aggregates.

Limited breeding activities were observed in three small adjacent ponds,

where 1, 2, and 10 individual egg masses were recovered.

Midday water temperatures during periods of oviposition were 3-

18° C (mean = 9.1) at NTL and 6-15° C (x = 8.9) at RCI. Air temperatures

during this period were 1-23° C (x = 8.7) at NTL and 5-20° C (x = 9.5) at

RCI. Maximum depth averaged 9 cm at NTL and 33 cm at RCI during

this time.

Wood Frog in Georgia

19

Clutch sizes varied from 295 to 706 eggs per mass (x = 553 ±

139.05 SD; N = 8). No more egg masses were taken because of the small

number of egg masses deposited in each breeding pond (range = 1-22).

Eggs averaged 2.8 mm in diameter (range = 2.2-3. 3 mm; range in

mean diameter per clutch = 2.7-2. 9 mm). Outer jelly envelopes averaged

11.2 mm in diameter (eggs = 7.3-14.1 mm; mean per clutch = 10.4-12.3

mm).

Hatching began at NTL on 3 March and continued until 13 March.

Hatching began 24 February at RC1 and 26 February at RC2. Hatching

was completed 20 March at RC1 and 1 1 March at RC2. Hatching times

ranged from 18 to 25 days after the date of egg deposition. Water

temperatures during the hatching period averaged 12.3° C (range = 3-22)

at NTL and 1 1.7° C (range = 6-17) at RC1.

Hatching occurred at stage 20 (gill circulation) at an average total

length of 8.7 mm (range = 8. 3-9.0). Larvae reached an average maximum

total length of 51.6 mm (range = 49.0-55.6) before metamorphosing at

an average body length of 17.8 mm (range = 15.0-21.0). The first

metamorphosed frog was collected from RC2 on 21 June and the last

on 19 July. From the start of hatching to the start of metamorphosis

was 1 15 days, and from the end of hatching to the end of metamorphosis

was 130 days. NTL and RC1 were completely dry by the end of April

(22 and 25 April, respectively). Complete mortality of larval populations

in both sites was assumed, and this was supported by the observation of

large numbers of dead tadpoles at each site.

DISCUSSION

Rana sylvatica typically lays eggs in communal aggregates that are

hypothesized to represent a thermal adaptation to development in cold

climates (Wells 1977, Howard 1980, Seale 1982, Waldman 1982, Waldman

and Ryan 1983). In four breeding ponds containing nine or more egg

masses in our study, only two (RC1 and RC2) showed signs of communal

oviposition. The lack of aggregation of NTL and SR egg masses may

have been an aberration created by the small number of clutches in

those breeding pools.

Wood frog eggs from Georgia are only slightly smaller than those

reported from Alabama (2.9 mm; Davis and Folkerts 1986). The

tendency of wood frogs to deposit progressively larger eggs from the

north to the south in their geographic range has been noted by several

investigators (Herreid and Kenney 1967, Meeks and Nagel 1977, Davis

and Folkerts 1986).

Larvae in this study hatched at a smaller size than those in

Alabama (10.7 mm; Davis and Folkerts 1986). The size difference is

attributable to larvae in Georgia hatching at an earlier stage than those

20

C. D. Camp, C. W. Condee, and D. G. Lovell

in Alabama (stage 21). Meeks and Nagel (1977) reported a similar

hatching size to that in this study but did not report the larval stage at

hatching. Herreid and Kinney (1967) reported that Alaskan wood frogs

hatch at larval stage 20. Tadpole development time in our study was

longer than that reported for wood frogs elsewhere (Hinckley 1882,

Beilis 1957, Herreid and Kinney 1967, Meeks and Nagel 1977, Berven

1982b, Davis and Folkerts 1986). Berven (1982b) suggested that temper-

ature was a major factor in variation in wood frog larval development

periods.

Maximum total length reached by tadpoles in Georgia was similar

to that reported from most parts of the species range (Beilis 1957,

Herreid and Kinney 1967, Meeks and Nagel 1977). Size at metamorphosis

was similar in Georgia wood frogs to that reported elsewhere (Hildebrand

1949, Beilis 1961, Meeks and Nagel 1977, Berven 1982b).

Catastrophic mortality as a result of premature desiccation of NTL

and RC1 emphasizes the risky nature of breeding in temporary pools.

Desiccation has been identified as a major selective force for amphibians

breeding in temporary ponds (Semlitsch 1987).

ACKNOWLEDGMENTS.— We thank Carlton Bowers for access to

wood frog breeding sites on his property and Trip Lamb, Ray Semlitsch,

and J. Whitfield Gibbons for their comments on the manuscript.

LITERATURE CITED

Beilis, E. D. 1957. An Ecological Study of the Wood Frog, Rana sylvatica

LeConte. Ph.D. dissertation, Univ. Minnesota, Minneapolis.

Beilis, E. D. 1961. Growth of the wood frog, Rana sylvatica. Copeia

1961:74-77.

Berven, K. A. 1982a. The genetic basis of altitudinal variation in the wood

frog Rana sylvatica. I. An experimental analysis of life history traits.

Evolution 36:962-983.

Berven, K. A. 1982b. The genetic basis of altitudinal variation in the wood frog

Rana sylvatica. II. An experimental analysis of larval development. Oecologia

52:360-369.

Davis, M. S., and G. W. Folkerts. 1986. Life history of the wood frog, Rana

sylvatica LeConte (Amphibia: Ranidae), in Alabama. Brimleyana 12:29-50.

Gibbons, J. W., and R. D. Semlitsch. 1982. Terrestrial drift fences with pitfall

traps: an effective technique for quantitative sampling of animal populations.

Brimleyana 7: 1-16.

Gosner, K. L. 1960. A simplified table for staging anuran embryos and larvae

with notes on identification. Herpetologica 16:183-190.

Wood Frog in Georgia 21

Herreid, C. F., and S. Kinney. 1966. Survival of Alaskan wood frog larvae.

Ecology 47: 1 039- 1 04 1 .

Herreid, C. F., and S. Kinney. 1967. Temperature and development of the

wood frog, Rana sylvatica , in Alaska. Ecology 48:579-590.

Hildebrand, H. 1949. Notes on Rana sylvatica in the Labrador Peninsula.

Copeia 1949:168-172.

Hinckley, M. H. 1882. Notes on the development of Rana sylvatica LeConte.

Proc. Boston Nat. Hist. Soc. 22:85-95.

Howard, R. D. 1980. Mating behavior and mating success in wood frogs,

Rana sylvatica. Anim. Behav. 28:705-716.

Martof, B. S. 1970. Rana sylvatica. Cat. Am. Amphib. Rep. 86.1-86.4.

Martof, B. S., and R. L. Humphries. 1959. Geographic variation in the wood

frog, Rana sylvatica. Am. Midi. Nat. 61:350-389.

Meeks, D. E., and J. W. Nagel. 1977. Reproduction and development of the

wood frog, Rana sylvatica , in eastern Tennessee. Herpetologica 29:188-191.

Mount, R. H. 1975. The Reptiles and Amphibians of Alabama. Ala. Agric.

Exp. Sta., Auburn.

Seale, D. B. 1982. Physical factors influencing oviposition by the wood frog,

Rana sylvatica, in Pennsylvania. Copeia 1982:627-635.

Seigel, R. A. 1983. Natural survival of eggs and tadpoles of the wood frog,

Rana sylvatica. Copeia 1983:1096-1098.

Semlitsch, R. D. 1987. Relationship of pond drying to the reproductive success

of the salamander Ambystoma talpoideum. Copeia 1987:61-69.

Waldman, B. 1982. Adaptive significance of communal oviposition in wood

frogs (Rana sylvatica). Behav. Ecol. Sociobiol. 10:169-174.

Waldman, B., and M. J. Ryan. 1983. Thermal advantages of communal egg

mass deposition in wood frogs {Rana sylvatica ). J. Herpetol. 17:70-72.

Wells, K. D. 1977. The social behavior of anuran amphibians. Anim. Behav.

25:666-693.

Accepted 28 March 1989

22

FISHERMAN’S GUIDE

FISHES OF THE SOUTHEASTERN UNITED STATES

by

Charles S. Manooch, III

Illustrated by Duane Raver, Jr.

Remarkable for its breadth of coverage, this book details the

habits, range, and appearance of more than 250 species of fish, 150 of

which are illustrated in color. Each account also includes tips on

catching the fish and preparing it for the table.

Manooch is experienced in the field of Fisheries Management and

Technology, and Raver is nationally known for his paintings of wildlife.

“An excellent general reference book for scientific or non-scientific

audiences. ... It contains information not easily found in any other

source.” — Carter R. Gilbert, Curator of Fishes, Florida State Museum.

1984 376 pages Index Bibliography ISBN 0-917134-07-9

Price: $29.95, plus $1.25 for shipping. North Carolina residents add 5% sales

tax. Please make checks payable in U.S. currency to NCDA Museum

Extension Fund.

Send to FISHERMAN’S GUIDE, N.C. State Museum of Natural Sciences,

P.O. Box 27647, Raleigh, NC 27611.

Seasonal Diet of the Margined Madtom,

Noturus insignis (Osteichthyes: Ictaluridae),

in a North Carolina Piedmont Stream

Robert P. Creed, Jr.,1 and Seth R. Reice

Department of Biology, Coker Hall 010- A

University of North Carolina

Chapel Hill, North Carolina 27514

ABSTRACT. — The diet of four size classes of the margined madtom,

Noturus insignis, collected during 1 year from New Hope Creek,

Orange Co., N.C., was examined. Margined madtoms consumed a

wide variety of prey, but most (93.8%) were Diptera, Ephemeroptera,

Trichoptera, and Plecoptera. On average, fewer individual prey were

consumed during the winter than in any of the other three seasons.

Many prey species were most abundant in the madtom’s diet when

those prey appeared to be most abundant in the stream. The diversity

of the madtom’s diet increased as the fish increased in size. Small

madtoms consumed primarily chironomid larvae ( >70% of the diet).

Although chironomids were still a major prey item numerically for

large madtoms, 65-70% of their diet consisted of other taxa. We

hypothesize that the diet of the madtom diversifies with increasing size

because large madtoms are able to capture large prey successfully and

are able to forage in areas with high current velocities.

The margined madtom, Noturus insignis (Richardson), is a common

benthic fish in streams of eastern North America from New York to

Georgia (Rohde 1980). Despite the margined madtom’s widespread

distribution, characterization of its diet is based on only three individuals

collected by Flemer and Woolcott (1966) between 13 June and 22

December, 1958. Flemer and Woolcott found a dipteran larva, two

stoneflies, and unidentified insect and fish remains in the stomachs of

the margined madtoms they sampled. The present study had two major

goals. First, we wanted to provide more detailed information about the

diet of N. insignis in the southern part of the species range. Second, we

wanted to determine if the diet of margined madtoms changed with

either the season or the size of the fish.

METHODS

Margined madtoms wdre sampled from a series of three riffles and

three pools (about 150 m of stream) in New Hope Creek, a fourth-order

Present address: W. K. Kellogg Biological Station, Department of Zoology, Michigan

State University, Hickory Corners, MI 49060.

Brimleyana 16:23-32, July 1990

23

24

Robert P. Creed, Jr., and Seth R. Reice

stream that flows through the Duke Forest, in Orange County, N.C.,

approximately 1 1 km NE of Chapel Hill. Madtoms were collected using

an electroshocker in combination with dipnets, because strong currents

and the rocky bottom prevented the effective use of a seine. Shocked

fish were captured as they drifted into two large dipnets (0.5 m X 0.25

m) placed on the bottom about 0.5 m downstream from the anode ring.

Fish were collected from September 1982 through August 1983 with

collection dates approximately 1 month apart (range 27-41 days, x = 33

days). We restricted collection to a small section of the stream so that

any observed variation in diet would not be a result of variation in prey

availability at different sites.

Because madtoms are reported to be nocturnal feeders (Mayden

and Burr 1981, Burr and Mayden 1982, Moyle and Cech 1982), collections

were made between 1 and 2 hours after sunrise. Fish were preserved

whole in 95% ethanol within an hour of capture. In order to assess

whether madtoms consumed any prey during the day, fish were collected

twice on one date (27 July 1983), once in the morning and again

between 1700 and 1800 hours.

Prior to gut content analysis, fish were measured (total length in

mm). Then the foregut and intestine were removed. Visual estimations

of gut fullness were made for both the foregut and the intestine

following the methods of Yoshiyama (1980) and Creed (1985). Evaluation

of gut fullness was not initiated until March 1983. Five levels of fullness

were used: 0, lA, Vi, 1, 2, of which 0 signifies empty or with traces of

food, 1 is full, and 2 is distended. Contents of both the foregut and the

intestine were then removed and examined under a dissecting microscope.

Prey items were counted and identified to the lowest taxon possible

(often genus, occasionally species). These data were pooled by season,

because it was not always possible to sample adequate numbers of

madtoms in particular months. This was especially true of collections

made in January and February, when water levels were high and

effective sampling was difficult. The seasons referred to in the results

(fall 1982 through summer 1983) encompass the following collection

dates: fall - 28. IX, 28. X, 7. XII; winter - 6.1, 9. II, 1 6. Ill; spring - 14.IV,

ll.V, 21. VI; summer - 27. VII, 25. VIII.

RESULTS AND DISCUSSION

A total of 53 margined madtoms were sampled for gut content

analysis. Numbers of madtoms collected in each of the first three

sampling seasons were about equal (fall N = 10, winter N = 1 1, spring N

= 12), but about twice as many madtoms were collected during the

summer (N = 20).

Seasonal Diet of Margined Madtom

25

Table 1. Distribution of 53 margined madtoms collected from New Hope

Creek, Orange Co., N.C., by size class and season.

Madtoms, which ranged from 19 to 113 mm, were assigned to four

size classes (Table 1). Fish 31 mm or longer were collected throughout

the year. In fall and winter samples, however, most fish were in the

40-50 mm and 95-105 mm ranges, which closely match the ranges of

Clugston and Cooper (1960, figure 2) for age 0+ and 1+ fish during fall

and winter months. During the spring and summer we collected a

number of madtoms from an intermediate size class (61-90 mm).

Clugston and Cooper collected madtoms in that same size range during

the summer growing season. The data suggest that madtoms in the

intermediate size class in our study are probably age 0+ fish growing to

age 1 size. We collected only one young-of-the-year madtom, less than

31 mm long (19 mm), in July.

Captured madtoms almost always contained prey. The only

exceptions were two small fish (40 and 38 mm) collected in January. In

general, fish collected during the winter contained fewer prey than those

sampled during the rest of the year (Table 2). Ninety percent of the fish

collected from March through August (N = 31) had a foregut fullness of

l/i or greater. Most fullness values (84%) for the intestine were l/i or less.

These data suggest that the madtoms had fed throughout the night and

that processed food was starting to move into the intestine. Four of the

five madtoms sampled on the morning of 27 July had a foregut fullness

of 1 and the fifth had a foregut fullness of lA\ all had an intestinal

fullness of l/i or less. Three of the five individuals sampled that

afternoon had no prey in the foregut, and two contained 8 and 42 prey

items, respectively; all had an intestinal fullness of x/i or greater. Some

prey from the afternoon foregut samples were not broken up, which

suggests that they had been consumed recently. In general, however,

intestinal fullness increased from morning to afternoon. Therefore,

although the general trend was for nocturnal feeding, some madtoms

appeared to take prey during the day.

"O

x>

o

JD

o

x>

2

o

•*— »

"O cd

3 r \

6 u.

TD Z

x> r

c o

'So U

c5 <o

C 60

c c

cd

^ o

T3

_CD

'cx

£

c a

o ^

c/5 r-

Cd «3

x> m

in ^

o

£

cd

o

H

x>

6

2 Ci

m

60

•ss

CX W

c/5

x> _

•*-< ' '

c ~

§ w

"3 2

Pm w

c

o

X

cd

+->

<U

S~)

Dm

vO Tf (N

cd

VO VO Tf vO <N O VO

— < — vo

vO — •

CTv

o o o

o o o o o o

<N Tf

O O O rn O Tf

rf O O

r-~ O O rn o rn

O O —

Os VO Tf O <N m

x>

cd

-a

5

x.

o

R

-Q

a

§■

cj

3

V <u

O cd

-r -a

& •-£

2 <u

§ ^

6<) •

05 ~0

— 3 —

„ -R C

x> £> ^ P

Dm

x>

cd

T3

2 *

cx 3

o

V.

3

X) X

R

O

X

60

- x.

X>

&

§

£

x>

-t-»

cx

o

• ^■H

c

x>

cd

H

b

<x

§■

-R

§■

v.

»■»

60

>,

u

x>

♦->

CX 3

• 2 £

C o

X) s*

cd R

H £

a

•

r

. ~ ^Oh

T3 -R g •

C §. O 3

o S S3 C

X X P

vn

o o

CM

rt — ;

vn

cd

Xh

<U

-t-t

CX

o

X)

vn

u-5 (N

<N w

X I

C/5

cd

t-4

X>

-t— •

cd ^

X, O

x> o

■e <u

Dh rr^

O Dm

O

^ O

Du C

3 §

o 2

H S

SO —

o o o

■4*

Cj

‘M*

V.

<x

-R

cd

)m

<D

•*->

CX

o

<u

_, 05

2 3 -R w

<u R (J

cx »r •

O 2

ii Sj 3 C

o ^ Q D

VO

Ov

cd

Im „ .

<U y

w Cy

CX "O

o x

"O ^

X u

cx >.

x> Du

J

cd

Im

<u

-*->

CX

1 ^

(U

2 ^

c

o

C/5

cd

<D

i/5

■o

<D

2,

s

cd

C/5

X

C/5

o

£

cd

-t->

O

H

u

4J

2

3

C/5

O

<N

60

•S f?

CX w

C/5

<u

C

• ^H

cd

Ph

C

O

X

cd

■«_>

Os

so

Os

in — m

ir>

Tf

o o

Tf

Tf o

vO

<N

<U

cd

c:

_cd

<u

cd

TD

2

o

cd £

*M o

a> vm

Dh X

5 u

d>

cd

CX

D

a

«

cd

T3

2

o

c

o

Uh

X

u

IT)

— < vr> Tt vn

VT) OS ~

— H O —

m — ^t

000 —

ri M O

Tt — H

(N - Tt Os

^t

OO — O

(N O - O

»no 000 r-omo o

O Tf

(U

cd

TD

3

cx

H

a

-R

cj

O

*M

R

(U

cd

■d

•

3

cx

H

3 x

c

(U

cd

3

£

• r—4

C/5

(U

cd

£

_cd

5

3

4J

cd

”0? ^

cd cd

CX O

D _

cx 5

w 3

3 3

5 5

Co Co

3

05

o

«fc

4J

cd

<u

cd

T3

■d ’’3

’3 ’Sh

5

o

■3

O

3

S>»

t<X

s ^

<u

cd

3

"3

'Dh

2

pp

3

£

M M ^ C

B

vr>

O Tt

Os <N

*T5

<N OO

<N —

<N w

m

or ^ Tt

un —

— O vO (^5000 OO — »T5 000

00 <N

CN — '

05

3

o

~R

O

3

5

cd

Im

(U

■*-*

CX

s

cd

+->

o

H

x

c/l

cd

<D

H->

CX

5

6

c

cd

(U

s

m so rn

o

— o

0 — 0

— — o

cd u

Xh cd

- .2

CX :D

o c

VX u

S 60

cd

05

^3

O

3

* '■m*

R

2

4>

X

CX

<D

cx X

w

3

r 2

x> <x

^5 R

2 S

& §

a; i5

NO

ro

— <N — (N — — 1

ON

ro

— in

o —

o o

O — '

o o

NO

<N

o o

ON

ro

o o

CM O O

O <N

ro

O — '

— < tN 1

o o o

<N

o

2 «

O "O

K, ’S

o 42

c/3 <«

• !S o

CS <

< ^

cd

c

1-1

cd

o

cd

Uh

TO

So

sc

cd

cd "5

O O

o

cd

C/5

3

(-1

o

Cd

TD cd

o 3

a, o

rv O d

O o ^

CX >, cd

= u u

a

X

t-H

o

aJ

-o '~

© *

° (2

C/3

cd

l-H

o

TD

cd

u

cd

t-H

o

■o

cd

o

cd

T3

O U

”0 ^3

42 3 g1 cd

UD JO

cd

X3

o

CX

X

a,

O

cd

3

‘C

cd

6

£

so

O

Cd

fcH

o

c/3

c

cd

T3

O

cd

C/3

SO

-*-> cd

c/3 t-

o u

C/3

W)

00

O

42

C/3

•

tu

CX

o

c *-*

o

£ 3

„ so c

o D

H

cd

C/5

<u

fc

oo

Tf

ON

ON Tf

r-~ ro

NO ^

NO

ro ro

Tf «T3

NO ^

in

On ro

m cn|

<n w

r~

t-~ no

no ro

ro s — '

42

C/3

<L)

cx

So

o

c

3 §

o

H

ro m ro oo <N —

o — — r-

— ro tj-

«N <N — Tf d1

— in cn

^ on — r- m

<N — Tf

(N

Tf

O <N NO O

r-

m

On

(N

<o) — — '

Tf

<n

m

o o m on Tt

-H OO ^

o on tj- o Tt no o r--

<N 1 nO —

<N CO

Tf

NO

<N

»n — i o

o o

m -h

ON

—I CO *-H

— <N w

O NO

-^ © m

oo

m no

r- ^

O O CO o o o o

— ' O O O

O NO o —

<N

<N ni

in o r-

o o

CO O -H

NO

NO <N

CN w

42

C/3

5

3

C

o

<X

Q

C

X

Co

3 cS-'3'3

43 ^ P

cd

t-H

<D

*->

cd

t- o

D Vh

*r* D

0 e

C. a>

D 42

a cx

g w

42

CX o

W c

1 2

H S

cd

t-H

D

*->

cx

o

D

cd

TD

42

O

So

C/3

CX

O

X

-c

X

s?

§•

5

3

X

D

cd

3

o

So

C/3

* CX

o

t-H

~a

D

cd

~o

42 -a 4;

3 >"o

Uh 3h

H

-c

^ D

^ :

§•* o.

"Q 3 °

D

■3 > .’2

I

>< Dh

'O • — < Si ^

3 3 £ cx

135 D jg* ^ O «

•2

«-H

X

o

D

cd

"O

D «

g -3

3 S

<P cd

3

IN.

T3

C

D

o a

cx c

o -S

•22

Oh

^3

O

^>xfc

-s:

t

O

Q

cd

l-H

D

■»->

CX

o

42

o

3

D

42

C/3

cd

t-H

D

4->

cd

tn o

« 42

0,3

0 Jr

42 H

o

'C o

H «=

s ^

o 3

H S

aUnid. = unidentifiable beyond taxon listed.

28

Robert P. Creed, Jr., and Seth R. Reice

Margined madtoms consumed a wide variety of prey (Table 2). A

vast majority of the prey were benthic, but some terrestrial prey were

consumed. We assume that the terrestrial prey were encountered on the

stream bottom because madtoms were always observed on or in the

bottom. Of the benthic prey, immature aquatic insects in four orders

dominated in the diet (93.8%); these orders were Diptera, Ephemeroptera,

Trichoptera, and Plecoptera.

Dipterans, especially chironomid larvae, were the predominant

prey in the diet except in the spring (Table 2). Chironomids were found

in all but 2 of the 51 madtoms that contained food in their guts. The

mean number of chironomids consumed per fish in the fall and summer

was about the same. The number of chironomids eaten was lowest in

the winter, which appears to be a result of reduced feeding activity by

members of the 31-60 mm size class of madtoms, perhaps as a result of

the strong currents present on those dates. A majority of the prey (67%)

were consumed by large fish ( >90 mm) during the winter. As discussed

below, large madtoms consumed fewer chironomids than did small

madtoms. Reduction in the number of chironomids consumed in the

spring appears to have a different explanation. We observed dense

aggregations of large instars of Baetis (Ephemeroptera), prior to their

emergence in the spring, on the upper surface of many rocks. Encounters

between madtoms and Baetis probably increased at this time of year.

Increased feeding on Baetis probably led to reduced feeding on

chironomids in the spring. Indeed, there was an inverse relation between

the numbers of Baetis and of chironomids in the guts of the 12 madtoms

collected in the spring (r = -0.47, Pearson product -moment correlation).

Simulium spp. were also observed to emerge in the spring, the season

when large instars of this genus are probably encountered most frequently

by the margined madtom. Simulium pupae were found in the madtom

gut only in the spring.

Ephemeropteran nymphs other than Baetis were also important in

the madtom’s diet (Table 2). Heptageniid nymphs were found in the gut

throughout the year, with a peak in the summer. About half as many

ephemeropterans as dipterans were consumed over the entire year.

However, in terms of prey biomass, it is likely that Ephemeroptera often

contributed at least as much as Diptera, if not more, to the madtom’s

diet. In the spring Ephemeroptera outnumbered Diptera and probably

contributed more biomass to the diet as well.

There were few apparent trends in the consumption of trichopteran

larvae (Table 2). However, consumption of Hydropsychidae and

Hydroptila was greatest in the summer. Most Chimarra were consumed

during the summer and fall. The total number of Trichoptera consumed

29

Seasonal Diet of Margined Madtom

was extremely low during the winter. Overall, low numbers of Plecoptera

nymphs were consumed. Winter-emerging stoneflies (e.g. Taeniopteryx,

Strophopteryx , Amphinemoura, and Allocapnia ) were consumed when

large instars were most abundant in New Hope Creek.

Our results indicate that the diet of N. insignis is similar to that of

other species of Noturus (Mayden et al. 1980, Mayden and Burr 1981,

Burr and Mayden 1982, Miller 1984). Unlike Flemer and Woolcott

(1966), we did not find any fish remains in the madtoms sampled.

Seasonal variation in the composition of the diet was observed. We

believe it is attributable, in part, to seasonal variation in the abundance

of different prey taxa. For example, Baetis, Simulium, and winter-

breeding stoneflies were most abundant in the madtom’s diet when large

individuals of these taxa were most common in the stream. Miller

(1984) did not observe seasonal variation in the diet of Noturus munitus.

An interesting trend is apparent when the diet of the madtom is

analyzed by size class (Table 3). As madtoms increased in size, the

proportion of chironomid larvae consumed decreased relative to other

prey. That trend in chironomid consumption, though evident during

spring, summer, and fall (Table 4), is most pronounced in summer and

fall. In winter the proportions of chironomid larvae consumed by the

two intermediate size classes were almost identical; only the largest fish

had a fairly diverse diet. Chironomids increased in importance for

intermediate-sized madtoms (61-90 mm) during the winter, when their

movements were probably restricted by swift currents. In the spring, on

the other hand, the proportions of chironomids consumed by all size

classes were reduced. That appears to be a result of the increased

consumption of Baetis and Simuliidae by all sizes of madtoms, possibly

as a consequence of increased encounter rates with those taxa. In

general, though, as madtoms increased in size the importance of

chironomids in the diet decreased, while the importance of other

invertebrates, primarily Ephemeroptera and Simuliidae, increased. The

importance of chironomids as prey for the young of other stream fish

has been noted by Allen (1941), Scrimgeour (1986), and Weatherley

(1987). Mayden and Burr (1981), Burr and Mayden (1982), and Miller

(1984) also noted an increase in the diversity of the diet of other

Noturus species with increasing size of the fish.

We have considered three explanations for the more diverse diet of

large madtoms: (1) small madtoms may be restricted in habitat use by

piscivorous predators, (2) movement of small madtoms may be restricted

to areas of reduced current velocity, and (3) the mouths of small

madtoms are just too small to handle large prey. Because we collected a

majority of the madtoms in riffles, where largemouth bass, the dominant

30 Robert P. Creed, Jr., and Seth R. Reice

Table 3. Contribution (percent) of major prey taxa to the diet of four size

classes of margined madtoms.a

piscivore of New Hope Creek, are absent, the first explanation seems

inadequate. In addition, because madtoms are active primarily at night

and bass feed primarily during the day, predation risk would probably

be low for all size classes of the madtom. We did not directly measure

the swimming ability of madtoms under different current regimes.

However, two pieces of indirect evidence lend support to the idea that

the movement of small madtoms is influenced by current velocity. First,

both the total and mean number of prey consumed were lowest in the

winter, a period of higher than average discharge and of frequent

flooding (Reice 1981; Creed, personal observation). Most of the prey

consumed during the winter were eaten by the large madtoms, the only

size class also to have a fairly diverse diet. Second, Simulium and

Baetis, which composed about 38% of the diet of madtoms >90 mm

Seasonal Diet of Margined Madtom 31

Table 4. Contribution (percent) of chironomids to the diet of four size classes

of margined madtoms for each of the four seasons. a

long, were usually found on the upper surface of rocks. This was

especially true for the filter-feeding Simulium. These taxa made up only

9% of the diet of madtoms <60 mm long. Many of the prey were

obviously too large for small madtoms to handle, e.g. late instars of

Acroneuria, Stenonema , Isonychia, and Megaloptera, as well as crayfish.

Consequently, we favor the idea that diet diversifies with increasing size

because the larger madtoms (1) are able to capture larger prey successfully

and (2) are able to forage in areas of higher current flow. Our data

suggest that the diet of margined madtoms is strongly influenced by

size-specific prey capture abilities and current velocity. These and other

factors influencing the predatory behavior of many stream fishes,

especially nonvisual predators like madtoms, deserve further study.

ACKNOWLEDGMENTS.— We thank Jerry Diamond and Rob

Edwards for helping with the collection of the madtoms. Bill Cooper,

Dennis Mullen, and Mark Oemke reviewed earlier drafts of the

manuscript. Comments by two anonymous reviewers improved the

manuscript. This research was funded by NSF Grant DEB-8206910 to

S. R. Reice. Contribution No. 649 from the W. K. Kellogg Biological

Station, Hickory Comers, Mich.

LITERATURE CITED

Allen, K. R. 1941. Studies on the biology of the early stages of the salmon

(Salmo salar). 2. Feeding habits. J. Anim. Ecol. 10:47-76.

Burr, B. M., and R. L. Mayden. 1982. Life history of the brindled madtom

Noturus miurus in Mill Creek, Illinois (Pisces: Ictaluridae). Am. Midi. Nat.

107:25-41.

32

Robert P. Creed, Jr., and Seth R. Reice

Clugston, J. P., and E. L. Cooper. 1960. Growth of the common eastern

madtom, Noturus insignis in central Pennsylvania. Copeia 1960:9-16.

Creed, R. P., Jr. 1985. Feeding, diet, and repeat spawning of blueback herring,

Alosa aestivalis , from the Chowan River, North Carolina. Fish. Bull., U.S.

83:711-716.

Flemer, D. A., and W. S. Woolcott. 1966. Food habits and distribution of the

fishes of Tuckahoe Creek, Virginia, with special emphasis on the bluegill,

Lepomis m. macrochirus Rafinesque. Chesapeake Sci. 7:75-89.

Mayden, R. L., B. M. Burr, and S. L. Dewey. 1980. Aspects of the life history

of the Ozark madtom, Noturus albater, in southeastern Missouri (Pisces:

Ictaluridae). Am. Midi. Nat. 104:335-340.

Mayden, R. L., and B. M. Burr. 1981. Life history of the slender madtom,

Noturus exilis, in southern Illinois (Pisces: Ictaluridae). Occas. Pap. Mus.

Nat. Hist. Univ. Kansas 93:1-64.

Miller, G. L. 1984. Trophic ecology of the frecklebelly madtom Noturus

munitus in the Tombigbee River, Mississippi. Am. Midi. Nat. 1 1 1:8-15.

Moyle, P. B., and J. J. Cech, Jr. 1982. Fishes: An Introduction to Ichthyology.

Prentice-Hall Inc., Englewood Cliffs, N.J.

Reice, S. R. 1981. Interspecific associations in a woodland stream. Can. J.

Fish. Aquat. Sci. 38:1271-1280.

Rohde, F. C. 1980. Noturus insignis (Richardson), Margined madtom. Page

461 in Atlas of North American Freshwater Fishes (D. S. Lee, C. R.

Gilbert, C. H. Hocutt, R. E. Jenkins, D. E. McAllister and J. R. Stauffer,

Jr., editors). N.C. State Mus. Nat. Hist., Raleigh.

Scrimgeour, G. J. 1986. Prey selection by torrentfish, Cheimarrichthys fosteri

Haast, in the Ashley River, North Canterbury, New Zealand. New Zealand

J. Mar. Freshwater Res. 20:29-35.

Weatherley, N. S. 1987. The diet and growth of 0-group dace, Leuciscus

leuciscus (L.), and roach, Rutilus rutilus (L.), in a lowland river. J. Fish.

Biol. 30:237-247.

Yoshiyama, R. M. 1980. Food habits of three species of rocky intertidal

sculpins (Cottidae) in central California. Copeia 1980:515-525.

Accepted 10 April 1989

Population Dynamics of Adult Unionicola formosa

(Acari: Hydracarina), a Parasite of Anodonta imbecillis

(Mollusca: Bivalvia), in West Virginia

James E. Joy and Jeffrey W. Hively

Department of Biological Sciences

Marshall University

Huntington, West Virginia 25701

ABSTRACT. — Population dynamics of a parasitic aquatic mite,

Unionicola formosa , were studied at the McClintic Wildlife Station

(West Virginia) in two ponds that supported different densities of the

host mussel, Anodonta imbecillis. Pond 27 with 26.0 host indivi-

duals/m2 was categorized as a high-density pond, whereas Pond 14

with 8.6 host individuals /m2 was considered a moderate-density pond.

Collections were made monthly from May through November 1986.

All hosts in both ponds were infested by female mites, but only 57 of

90 hosts from Pond 14 and 60 of 79 from Pond 27 were infested by

males. Intensity of infestation, as mean adult mites/ host, was lowest in

May (5.4 for Pond 14; 19.7 for Pond 27) and highest in August (12.9

and 31.3) for those ponds. Although the number of mites per host was

positively correlated with host shell length for mussels in Pond 27,

there was little or no correlation in Pond 14. Mite sex ratios were

heavily female-biased at 10.7:1 in Pond 14, and 18.5:1 in Pond 27.

May (1983) demonstrated, by an illustration based upon math-

matical modeling, that a relationship existed between the number of

animal parasites per host and the host population density. In theory, at

relatively high parasite burdens and correspondingly high levels of host

mortality, a host population could be regulated by parasitic infestation

(Anderson and May 1978). Lanciani (1975) showed that increased

numbers of an ectoparasitic water mite, Hydryphantes tenuabilis

Marshall, reduced the rate of population increase of its aquatic insect

host, Hydrometra myrae Bueno, in a laboratory setting.

It is often difficult to assess the extent to which a parasite regulates

host population growth in nature. Therefore, the primary goal of this

study was to evaluate intensity of infestations of a parasitic aquatic mite

in two freshwater mussel populations at different densities. This study

was carried out during the 7-month seasonal period when mussels are

most active, and a secondary objective was to examine changes in

seasonal intensity levels of mites in the two host populations. We also

attempted to correlate intensity of infestation with host size, a rela-

tionship investigated in several previous mite/ mussel studies. This paper

Brimleyana 16:33-42, July 1990

33

34

James E. Joy and Jeffrey W. Hively

constitutes the first report of the parasitic mite Unionicola formosa

(Dana and Whelpley, 1836) from West Virginia.

MATERIALS AND METHODS

The subject of this study was a parasitic aquatic mite, Unionicola

formosa, and its freshwater mussel host, Anodonta imbecillis (Say,

1829). Work was carried out in two ponds at the McClintic Wildlife

Station, Mason Co., W.Va. The station, outlined on USGS Topographic

Map, Cheshire Quadrangle, Ohio— W.Va., is a 2,800-acre (1,135-ha)

wildlife sanctuary dotted by 35 ponds and managed by the W.Va.

Department of Natural Resources. Ponds 14 and 27 were chosen as

study sites because they harbored thriving mussel populations at

densities of 8.6 (moderate) and 26.0 (high) A. imbecillis individuals per

m2, respectively (Harmon 1987). Pond 14 had a surface area of

approximately 1.4 ha. It was a shallow pond (maximum depth of 2.4 m)

with a considerable amount of rooted aquatic vegetation (coon-tail,

Ceratophyllum demersum) arising from a silt/ clay substrate. Pond 27

had a surface area of approximately 0.75 ha. It also was shallow

(maximum depth of 1.6 m) with a silt/ clay substrate. With the exception

of a few small shoreline patches of cattail, this smaller pond was

virtually devoid of rooted aquatic vegetation.

A host sample was collected, by hand, from each pond monthly

from May through November, 1986. Collections were not random

because a randomized procedure resulted in a sample containing dispro-

portionately large numbers of mussels in the shell-length range of 65-79

mm. Because one goal was to estimate intensity of infestation relative to

host length, some additional effort was made to collect individuals with

a shell length <65 mm or >79 mm.

Each mussel was processed at the site where it was collected:

cleaned, measured for its shell length with vernier calipers to the nearest

0.1 mm, and opened by severing the adductor muscles with a # 60

autopsy scalpel blade. The entire open mussel was then placed in a

separate, labeled (pond designation, date, shell length) jar containing a

fixative of 10% buffered formalin acetate. This procedure precluded loss

of mites and exchange of mites between hosts. Hosts thus collected and

preserved were transported to the laboratory. In the laboratory mites

were collected from the bottom of the jars and from host soft tissues

with jeweler’s forceps and the use of a Zeiss stereomicroscope as

needed. Only adult mites were counted. Females were easily separated

from males on the basis of two or more of the following criteria: larger

body size, presence of eggs, shape of palps, and differences in anal plate

morphology (Vidrine 1986). The data from two collections made in the

same month were combined.

35

Unionicola formosa in West Virginia Mussels

Counts of female and of male mites were transformed (log j q [Y]

for female mites; logjQ [Y+l] for male mites). These data were then

backtransformed to show mean intensity levels (as mean number of

mites per infested host) with 95% confidence limits (Fig. 1A-C). To

detect seasonal differences in means, log-transformed data were used in

calculating F-values (ANOVA) on an AT&T PC 6300 computer with

Microstat® general-purpose statistics package developed by ECOSOFT,

Inc. A Texas Instruments statistial calculator was used for Mests (Table

1) and regression analyses (Fig. 2 and 3).

RESULTS

A total of 169 Anodonta imbecillis individuals— 90 from Pond 14

and 79 from Pond 27 — were examined for Unionicola formosa during

the 7-month study period. All host mussels were infested by female

mites (Fig. 1A), but male mites were recovered from only 57 of 90

(63.3%) and 60 of 79 (75.9%) hosts in Ponds 14 and 27, respectively

(Fig. 1B-C). The sex ratio of U. formosa was 10.7:1 (794 females: 74

males) in A. imbecillis from Pond 14 and 18.5:1 (1,737 females: 94

males) for the host sample drawn from Pond 27.

Mean intensity levels of female U. formosa in A. imbecillis

individuals from the high-host-density Pond 27 were significantly higher

than mean intensities for hosts in the moderate-density Pond 14 for

every month sampled from May through September (Table 1; Fig. 1A).

There was no statistical difference between the means for October, and

no comparison could be made for November when no mussels were

taken from Pond 27 (Table 1). Conversely, mean intensity levels of male

U. formosa in A. imbecillis were essentially the same for both ponds in

every month in which comparisons could be made (Table 1; Fig. 1B-C).

Although mean intensities of female mites increased seasonally

from May through September in Pond 27 (Fig. 1A), those increases

were not significant as determined by ANOVA on log-transformed data

{F - 2.166, 73 df; P - 0.0671). Seasonal variations in log-transformed

means for female mites in Pond 14 mussels (Fig. 1A) were, however,

significantly different ( F - 2.504, 83 df; P = 0.0282). An ANOVA on

log-transformed data revealed no significant differences in mean numbers

of male mites by season in either pond {F = 0.059, 54 df; P- 0.7681 and

F - 1.861, 50 df; P - 0.1062 for male mites in mussels from Ponds 27

and 14, respectively).

Adult mites were positively correlated with host shell length in the

high-host-density Pond 27 for every month sampled (Fig. 2). Cor-

relations between adult mites and host length in the moderate-host-

density Pond 14 were, however, largely nonexistent (Fig. 3).

36

James E. Joy and Jeffrey W. Hively

Fig. 1A. Back-transformed

mean number of female mites,

i.e. antilog [log Y]. Horizontal

lines and closed circles indicate

mean numbers of U. formosa

females in host mussels from

Ponds 14 and 27, respectively.

Vertical lines are 95% con-

fidence limits around the

means. Numbers above vertical

lines equal host sample size.

Because prevalence was 100%,

number of infested mussels is

the same as sample size.

Fig. IB. Back-transformed

mean number of male mites,

i.e. antilog [log Y+l]. Horizon-

tal lines indicate mean numbers

of U. formosa males in host

mussels from Pond 14. Vertical

lines are 95% confidence limits

around the means. Fractions

above vertical lines denote prev-

alence, with denominator the

host sample size and numerator

the number of hosts infested.

Fig. 1C. Back-transformed

mean number of male mites,

i.e. antilog [log Y+l]. Closed

circles indicate mean numbers

of U. formosa males in host

mussels from Pond 27. Vertical

lines are 95% confidence limits

around the means. Fractions

above vertical lines denote prev-

alence as in Fig. IB.

40-»

- 30

• />

o

\

«/»

20-

£

O

10

5

0-1

3n

Fig. 1 A

16

13

17

25

15

+

12

I

10

+

17

+

13

♦

15

8

+

«/)

O

o

• mm

S

*0

1

0

4

8

Fig. 1 B

'\7 J- '0,

no

4

15 - I I n 4.

L [ T T Is la 4

2-

0-1

M J J A S O N

37

Unionicola formosa in West Virginia Mussels

Table 1. Mean numbers of adult Unionicola formosa in Anodonta imbecillis in two ponds

compared by month.3

Female mites/host Male mites/host

aInitial /-tests based on raw data followed by log-transformed data in brackets.

^Significantly different at a - 0.05.

Cnd - not determined (mean - 1.0 and variance - 0 for male mites in mussels

from both ponds).

DISCUSSION

The impetus for this study came primarily from two sources: (1)

Vidrine’s (1980) suggestion that concentrated populations of mussels

harbored more U. formosa individuals than mussels in areas of lower

density; and (2) Dimock’s (1985) statement that “no association between

the population biology of the host and that of symbiotic mites has yet

been established.” Vidrine’s empirical observation should be evaluated

with consideration of his extensive work on aquatic mite/freshwater

mussel relationships. The present investigation strengthens, quantita-

tively, his generalization that mussels in a high-density population (as in

Pond 27) harbor significantly greater numbers of female mites than host

mussels in a moderate-density situation (as in Pond 14) (Table 1; Fig.

1A). Differences were so striking that mean numbers of female mites

Number Adult Mites per Host

Fig. 2. Scatter diagrams showing monthly relationship between number of

adult mites per host and host length. Each dot represents a single host from

Pond 27.

Number Adult Mites per Host

40 n

10-

MAY

r= .029

y = 5.03+ (0.008) x

FIG. 3 _

JUN

r= .447

y = -0.217+ (0.153) X

• •

40-i

10-1

JUL

r= .442

y= 1.93 + (0.120) X

AUG

r = -.172

y = 20.95 ■+■ (-0.118) x

40 n

10

SEP

r= .242

y= -0.053 +■ (0.113) X

• •

Host Shell Length in mm

Fig. 3. Scatter diagrams showing monthly relationship between number of

adult mites per host and host length. Each dot represents a single host from

Pond 14.

40

James E. Joy and Jeffrey W. Hively

recovered from Pond 27 hosts in May, July, and September exceeded

the maximum numbers of female mites collected from mussels in Pond

14. Still, these findings must be tempered somewhat because they

describe an association between host populations and the populations of

their acarine parasites in only two ponds.

Although sex ratios of Unionicola species vary considerably (Humes

and Jamnback 1950, Mitchell 1965, Gordon et al. 1979, Hevers 1980,

Dimock 1983, 1985), a female-biased situation is almost universal. This

condition, referred to as “harem-defense polygyny” by Dimock (1985),

was also seen in the present study with female: male sex ratios of 10.7:1

and 18.5:1 in Ponds 14 and 27, respectively.

Humes and Jamnback (1950) reported an inverse relationship

between prevalence of Najadicola ingens (Koenike, 1895) and size of

Elliptio complanata (Lightfoot, 1786) and Anodonta cataracta Say,

1817, whereas Mitchell (1965) found no correlation between host size

and any parameter of the population biology of Unionicola fossulata

(Koenike, 1895). Conversely, Gordon et al. (1979) and Dimock (1985)

cited positive correlations between host size and the presence of U.

formosa, with which our findings in Pond 27 concur (Fig. 2). Previous

workers tend to group all sample data on a single scatter plot to show

correlations between host shell length and number of mites present.

There is a possibility, however, that this approach obscures seasonal

correlations. For example, if a disproportionate number of small mussels

are examined in the spring, with predominantly larger mussels sampled

in the fall, the question then becomes: Is the correlation size-related or

season-related? To approach that question we attempted to collect A.

imbecillis individuals across a broad spectrum of shell lengths for every

sample month. Scatter diagrams were then constructed for each month

(Fig. 2 and 3). Thus, in Pond 27 (Fig. 2) it is quite apparent that the

positive correlations are indeed related to host shell length. The lack of

correlation between host length and number of mites present in Pond 14

(Fig. 3) is not easily explained, but the seasonal factor has been

removed because of a wide-ranging distribution of host lengths for each

month (except for October and November).

An understanding of the growth rates and anticipated life

expectancy of A. imbecillis individuals reveals why shell length may not

be a good indicator of mites present. Harmon (1987) has rather

convincingly argued that 70-79 mm will likely be the dominant size class

of mussels in a population at McClintic Wildlife Station. Small A.

imbecillis individuals grow rapidly, approaching their maximum shell

length of ~80 mm after 3 to 4 years. Because a 75-mm-long individual

could be in its fourth growing season, or in its eighth or ninth, it could

be argued that shell length alone provides insufficient information for

Unionicola formosa in West Virginia Mussels 41

inferences about numbers of mites present. That argument is strength-

ened by our data, which suggest that host density should also be

considered an important factor in determining the number of mites

present per host mussel (Fig. 1A). Dimock (1985) noted that age of host

may be correlated with number of mites present because increase in age

would allow for increased exposure time to invasive stages of the mite.

That is a reasonable conclusion even though no one has convincingly

demonstrated how to age members of this mussel species — at least

beyond the third growing season — with any degree of confidence.

Availability of oviposition sites, as suggested by Mitchell (1965), might

be a good estimator of number of mites present relative to host size.

Certainly some inventive measure of weight, or of gill area, could be

devised to test Mitchell’s hypothesis. One feels tempted to assess the

influence of mantle cavity volume, and perhaps host tissue response to

mite infections as well, although the latter measure may prove ex-

ceedingly difficult to describe.

Over the past 3 years we have never been able to collect mussels

from McClintic ponds in the winter months (December through

February). Our lack of data for March and April is an unfortunate

omission. That oversight, coupled with an unexplained population crash

of A. imbecillis in Pond 27 that began in mid- August of 1986, further

restricted our ability to draw definitive conclusions regarding seasonal

influences on adult mite infections. Nevertheless, a couple of compari-

sons can be made. Gordon et al. (1979), whose study period covered the

same months as ours, reported no seasonal differences in either

prevalence or intensity of U. formosa infestations in A. cataracta. Our

findings were basically similar, i.e. the prevalence was identical for every

month sampled (100%) and differences in mean intensities were

statistically insignificant for male mites in both ponds and for female

mites in Pond 27 (Fig. 1A-C). On the other hand, Dimock (1985) noted

seasonal variation for U. formosa in A. imbecillis'. Adult females were

most numerous in the winter and least so in late spring and summer.

Although we detected significant differences between monthly means for

female mites in Pond 14 (F = 2.504, 83 df; P - 0.0282), there was no

seasonal trend. Means for August and September, for example, were

widely separated (Fig. 1A). Thus, our knowledge of seasonal influence

on populations of U. formosa cannot be presented as a simple

generalization.

ACKNOWLEDGMENTS. — We thank Dr. Malcolm F. Vidrine,

Louisiana State University at Eunice, for his helpful suggestions during

the course of our study and for his subsequent review of the manuscript.

42 James E. Joy and Jeffrey W. Hively

LITERATURE CITED

Anderson, R. M., and R. M. May. 1978. Regulation and stability of host-

parasite population interactions. J. Anim. Ecol. 47:219-247 and 249-267.

Dimock, R. V., Jr. 1983. In defense of the harem: intraspecific agression by

male water mites (Acari: Unionicolidae). Ann. Entomol. Soc. Am.

76:463-465.

Dimock, R. V., Jr. 1985. Population dynamics of Unionicola formosa (Acari:

Unionicolidae), a water mite with a harem. Am. Midi. Nat. 114:168-179.

Gordon, M. J., B. K. Swan, and C. G. Paterson. 1979. The biology of

Unionicola formosa (Dana and Whelpley): a water mite parasitic irt the

unionid bivalve, Anodonta cataracta (Say), in a New Brunswick lake. Can.

J. Zool. 57:1748-1756.

Harmon, J. 1987. Density, length frequency distribution, growth rate, and

condition of the freshwater mussel, Anodonta imbecillis (Say, 1829), in six

ponds at the Clifton F. McClintic Public Hunting and Fishing Area,

Mason County, West Virginia. M.S. thesis, Marshall Univ., Huntington,

W.Va.

Hevers, J. 1980. Biologisch-okologische Untersuchungen zum Entwicklungszy-

kulus der in Deutschland auftretenden Unionicola- Arten (Hydrachnellae,

Acari). Arch. Hydrobiol. Suppl. 57:324-373.

Humes, A. G., and H. A. Jamnback. 1950. Najadicola ingens (Koenike), a

water-mite parasitic in fresh-water clams. Psyche 57:77-87.

Lanciani, C. A. 1975. Parasite induced alterations in host reproduction and

survival. Ecology 56:689-695.

May, R. M. 1983. Parasitic infections as regulators of animal populations.

Am. Sci. 71:36-45.

Mitchell, R. 1965. Population regulation of a water mite parasitic on unionid

mussels. J. Parasitol. 51:990-996.

Vidrine, M. F. 1980. Systematics and coevolution of unionicolid water-mites

and their unionid mussel hosts in the eastern United States. Ph.D.

dissertation, Univ. Southwest. Louisiana, Lafayette.

Vidrine, M. F. 1986. Five new species in the subgenus Parasitatax (Acari:

Unionicolidae: Unionicola ) from North America and Asia, with a re-

evaluation of related species. Int. J. Acarol. 12:141-153.

Accepted 12 April 1989

Reproduction in the Hispid Cotton Rat,

Sigmodon hispidus Say and Ord (Rodentia: Muridae),

in Southeastern Virginia

Robert K. Rose and Michael H. Mitchell

Department of Biological Sciences

Old Dominion University

Norfolk, Virginia 23529

ABSTRACT. — The hispid cotton rat, Sigmodon hispidus Say and

Ord, a species of the southwestern United States that has been moving

northward and eastward in this century, was first observed in Virginia

in 1940. In this study of the cotton rat in southeastern Virginia, most

males were reproductively competent from February through November,

embryos were recorded from March through October, and litter sizes

were comparable to those from other locations except Kansas. Also

unlike the cotton rat in Kansas, animals grew at substantial rates

during the winter in Virginia. The hispid cotton rat seems to have

adjusted its breeding season in Virginia by the cessation of breeding

early in autumn, which permits the last young of the season to attain

nearly adult size before winter arrives. Both young and adults are able

to maintain and even increase their autumnal body mass throughout

the winter. Timing and length of the breeding season and the patterns

of body growth suggest that the hispid cotton rat is well adapted to

winter, and hence to persistence of the species, in southeastern Virginia.

The hispid cotton rat, Sigmodon hispidus Say and Ord, is a

cricetine rodent that has dramatically expanded its distribution in the

central and southeastern states in historic times (Genoways and Schlitter

1967). First recorded in Virginia from Mecklenberg Co. in 1940 (Patton

1941), it moved northward in the lower Piedmont into Amelia Co.

(Lewis 1944) and then north of the James River in central Virginia

(Pagels 1977). The current distribution is believed to extend from

Virginia Beach westward to points north of Richmond and southwestward

through Halifax Co., or approximately throughout the southeastern

one-third of Virginia.

Because the hispid cotton rat has tropical affinities (Hall 1981,

Zimmerman 1970), it is surprising that the species has been able to

extend its range to the present northern limit of its distribution and to

cope with winters in such states as Kansas, Tennessee, and Virginia.

Furthermore, we expected to observe that this rodent has a shorter

breeding season in those marginal populations than in Texas or Mexico,

which are closer to the center of distribution for the species. In fact,

Brimleyana 16:43-59, July 1990

43

44

Robert K. Rose and Michael H. Mitchell

some features of the expected pattern have been reported in Kansas

(McClenaghan and Gaines 1978), Oklahoma (Goertz 1965), and Tennessee

(Dunaway and Kay 1964), although the details vary somewhat from

location to location. Kilgore (1970) examined the possibility that northern

populations might have larger litters than central populations as a way

of compensating for increased winter mortality; he found significantly

larger litter sizes in Kansas than in Texas.

The primary objective of our study was to examine details of

reproduction and patterns of body growth in a population of cotton rats

at the northern limit of the species distribution on the East Coast. Using

monthly samples of live-caught cotton rats that were necropsied and

examined for evidence of reproduction, we learned that cotton rats in

Virginia suspended breeding from early November through late March,

had litters no larger than those in central populations (Texas), and had

a larger weight gain in males than in females during the winter.

MATERIALS AND METHODS

From October 1983 to November 1984, cotton rats were obtained

using Fitch and Sherman live traps baited with chicken scratch feed (a

mixture of wheat, millet, and cracked corn). Although not always

attained, the goal was a sample of 30 animals per month. No animals

were taken in January or August. The 250-ha study area, an old field in

Portsmouth, Va., was dominated by grasses, Panicum spp. and

Andropogon spp.; a spikerush, Juncus effusus; and, at the margins,

young sweet gum trees, Liquidambar styraciflua. Other common species

of plants found in the study area were trumpet creeper, Campsis

radicans; cane, Arudinaria gigantea ; saltbush, Iva frutescens\ goldenrods,

Solidago spp.; and giant ragweed, Ambrosia artemisiifolia. Less common

were black oak, Quercus nigra\ grape, Vitis rotundifolia; loblolly pine,

Pinus taeda\ smooth sumac, Rhus copallina\ dogfennel, Eupatorium

capillifolium; blackberries, Rubus spp.; briers, Smilax spp.; willows,

Salix spp.; and cattails, Typha angustif olia . Traps were moved from

place to place to prevent excessive depletion of the cotton rat at a local

site.

All animals > 50 g (lower limit of potential breeders) were killed

with chloroform in the laboratory and frozen until necropsy, when the

following information was recorded for each: (1) body mass (g), (2)

overall body length (mm), and (3) length of tail (mm).

Additional data were recorded for females: (1) number of placental

scars, (2) number of embryos, (3) uterine mass (uterus + embryos), (4)

number of corpora lutea, and (5) parity class. The parity classes were

defined as nulliparous females without embryos or placental scars (also

Hispid Cotton Rat

45

lacking well-developed mammary glands and nipples); primiparous

females with one set of placental scars and corpora albicantia or with

embryos and corpora lutea (but not placental scars or corpora albicantia);

and multiparous females with more than one set of placental scars or

with embryos, corpora albicantia, and placental scars. In the analysis of

body mass, the mass of the uterus was subtracted so that pregnancy

would not confound the results.

Additional data were recorded for males: (1) testes position (scrotal

or abdominal), (2) paired testicular mass (mg), and (3) condition of

epididymal tubules (looped or convoluted). Males were considered to be

breeding if the epididymal tubules were convoluted (Jameson 1950).

Data are presented as x ± SE.

RESULTS

Females

The most reliable indicator of female reproductive state is pregnancy.

Of 148 females, 48% were pregnant. However, no pregnant females were

collected from November through February (Fig. 1). The level of

breeding in females was high from March through October, when the

average pregnancy rate was 68.7% (including October of both 1983 and

1984).

Using the Chi-square test, we found no differences (x^ - 0.76, 2 df,

P > 0.50) in the proportions of females that were pregnant in the April-

May, June-July, and September-October bimonthly periods (not sampled

in August). Thus, as measured by pregnancy, females bred at a uniform

rate during these months of peak activity.

Overall, litter size averaged 5.00 ± 0.284 SE. However, during the

peak breeding months of April to October, females averaged 5.18 ±

0.274 embryos per female. There was significant variation in litter size {-

embryo counts) throughout the months of the breeding seson (ANOVA:

F - 30.46, df = 4,62, P < 0.005) with largest litters (a = 7.83 ± 0.984) in

May. In contrast, females in April averaged only 4. 18 + 0.652 embryos

per female, and the two pregnant females in October 1983 had one and

two embryos, unusually small litters for the cotton rat.

The pregnant females were divided into primiparous (those in their

first reproductive experience) and multiparous (experienced breeders)

groups to determine whether a difference in litter size was attributable

to reproductive experience. Although there was a trend toward larger

litters in multiparous females, there was no significant difference between

the litter sizes of primiparous (x = 4.76 ± 0.378 SE) and multiparous (x

= 5.72 ± 0.371) females during the “peak” breeding months (/ s = 1.74, df

= 65, 0. 1 > P > 0.05). Thus, season had greater influence on litter size

than age of the female.

46

Robert K. Rose and Michael H. Mitchell

ONDFMAMJ J SON

1983 1984

Fig. 1. Monthly percentages of females (>50 g) that were pregnant. Sample

sizes are given above each bar. N = 148. (No collections in January and August.)

100

80

60

40

20

0

Hispid Cotton Rat

47

0 N D F

1983

A M

J J

1984

S 0

N

2. Monthly percentages of males (>50 g) that were breeding based on the

ence of convoluted cauda epididymides in the testes. Sample sizes are given

/e each bar. N = 152. (No collections in January and August.)

48 Robert K. Rose and Michael H. Mitchell

Prenatal Mortality

Preimplantation mortality, which occurs before the embryo has

implanted in the uterine wall, can be estimated by comparing the

number of ovulation sites with the number of embryos. After ovulation,

the remnant of each ovarian follicle is retained. It quickly enlarges into

a corpus luteum, a structure 2. 0-2. 5 mm in diameter, which can easily be

seen and counted. If all ova are fertilized and the resulting embryos are

successfully implanted, the number of corpora lutea corresponds exactly

to the number of embryos. However, if there are, for example, seven

corpora lutea but only six embryos, then one ovum has been lost to

preimplantation mortality.

There are two potential obstacles to making accurate estimates of

preimplantation mortality. The corpus luteum enlarges quickly as it

produces progesterone to maintain the thick wall of the pregnant uterus.

However, embryos do not appear as bulges in the uterus until day 10 in

the 27-day gestation period (Meyer and Meyer 1944). Thus, for a few

days the enlarged corpora lutea indicate pregnancy but no embryos are

evident. A second problem is twinning, the production of two embryos

from the same ovum. In this study, at least four females were judged to

be pregnant (i.e. had enlarged corpora lutea) though no embryos were

seen, and there was one case of probable twinning. When these females

were eliminated from the analysis, preimplantation mortality averaged

5.7% of 371 ova.

Males

Reproductive potential (fertility) in males is most reliably indicated

by the presence of convolutions in the cauda epididymides, which

Jameson (1950) found to be highly correlated with the presence of

sperm in the tubules. Relative testicular mass (the ratio of weight of

testes to weight of animal) is a fair predictor of maturity, because the

testes grow rapidly in late winter prior to the onset of the breeding

season. We used both of these indicators of male breeding capability.

Using convoluted cauda epididymides as a criterion, we found that

males were fertile longer than females (Fig. 2), from February (73%

fertile) to November (33% fertile); from March through July, all males

were fertile. According to this criterion, the breeding season of males

begins about one month earlier and ends about one month later than

that of females.

As is typical of males of many temperate-zone mammals, testes

undergo a dramatic regression in late autumn. In S. hispidus, the mass

of the paired testes of a 120-g male might be 2,000 mg at the height of

the breeding season, compared with only 80 mg after testicular regression.

With regression, the cauda epididymides lose their convolutions and

become looped. Such males are no longer fertile.

170

150

130

110

90

70

50

30

10

0

Fig

Hispid Cotton Rat

49

0 N D F

A M J J SON

1983

1984

3.

Mean testicular mass per 10 g of body weight for each month of study,

sizes as given in Fig. 2.)

50

Robert K. Rose and Michael H. Mitchell

1983 1984

Fig. 4. Mean monthly body masses of male and female cotton rats. (Sample

sizes as given in Fig. 1 and 2.)

Hispid Cotton Rat

51

1983

1984

Fig. 5. Mean monthly body lengths (total length minus tail length) of male and

female cotton rats. (Sample sizes as given in Fig. 1 and 2.)

52

Robert K. Rose and Michael H. Mitchell

Because not all males are of the same size and because testicular

mass is approximately proportional to body mass (Keller and Krebs

1970), we computed the testicular mass per 10 g of body mass (Fig. 3).

This assessment of male reproduction closely parallels the breeding

season of males based on cauda epididymides (Fig. 2). The testes grew

rapidly in late winter (February) so that overwintered males were fully

mature by March. The somewhat lower fertility rates of males in late

summer (September) probably were a result of an increasing proportion

of young males included in the samples. However, later decreases in

fertility (November and December in Fig. 2 and 3) were primarily a

result of testicular regression in adult males.

Dynamics of Body Size

In some parts of the United States, populations of cotton rats have

substantial winter mortality (e.g. Dunaway and Kaye 1964, Sauer 1985).

Slade et al. (1984) showed that cotton rats surviving the winter in

eastern Kansas tend to weigh nearly the same regardless of age; large

animals lose mass and young animals entering the winter grow slowly,

so that by spring most animals are approximately the same mass. Severe

mortality and weight loss in the winter make an evaluation of body

weight dynamics particularly important in S. hispidus at the northern

limit of its distribution, such as in southeastern Virginia.

Of course, chance plays a role in determining the average mass of a

sample of field-caught cotton rats, particularly during periods when

young animals are entering the trappable population. However, in this

study, those effects are minimized because juveniles and small subadults

(<50 g) were not collected for necropsy. Overall, males (x = 101.48 =t

2.027 g) were significantly heavier than females (jt = 94.26 ± 1.872 g).

Body mass differences were smallest (Fig. 4) at the end of the breeding

season (October and November). Males were much larger (20-30 g) than

females throughout the winter in this study.

Body length (Fig. 5) showed similar trends, with males averaging

141.11 ± 1.557 mm and females 137.63 ± 1.441mm. Males had roughly

linear growth in body length throughout the late autumn and winter,

and the decline in mean length was probably a result of the recruitment

of spring-born animals into the trappable populations. Body lengths of

males and females were most similar in October and November, a

pattern also seen with body mass (Fig. 4).

DISCUSSION

Mammals seem to adjust the breeding rate to the mortality rate at a

given location (Sadleir 1969). Mammals, particularly small mammals,

can increase reproduction by one or more of the following means:

Hispid Cotton Rat

53

becoming sexually mature at an earlier age, increasing litter size, or

increasing the number of litters per year (extending the breeding season

or decreasing the interval between litters). Decreasing the time between

litters is achieved by a short gestation period and rapid postnatal

development, so that the interval between conception and weaning is

minimal. In the most rapidly breeding individuals, mating often occurs

within 24 to 48 hours after parturition; consequently, a lactating female

frequently is pregnant with the next litter. In general, small mammals in

the tropics have small litters and long breeding seasons (Sadleir 1969).

However, in temperate locations, small mammals tend to compensate

for shorter breeding seasons and increased mortality by producing

larger litters (e.g. in Peromyscus\ Smith and McGinnis 1968). Of the

seven species of New World Sigmodon, only S. hispidus has a widespread

and expanding distribution in temperate North America, making it a

candidate species to examine for evidence of adjustments in its breeding

biology in response to the harsher winter conditions endured by

populations colonizing northern locations.

Pregnancy Rate

Maximum rates of pregnancy were achieved early and sustained

throughout the breeding season in Virginia. The observed pregnancy

rate often exceeded the theoretically observable maximum pregnancy

rate, such as in April when 95% of 21 females were pregnant (Fig. 1).

Because bulges in the uterus cannot be detected during the first 9 days

of pregnancy, embryos can be counted only for 18 days of the 27-day

gestation period. During the peak breeding season, mating usually

occurs within 24 hours of parturition, resulting in a 28-day interval

between litters. Because embryos can be seen only for 18 days of these

28 days, the theoretical maximum pregnancy rate that can be observed

is 18/28, or 64.3%, which is the detectable pregnancy rate if all females

are pregnant all of the time. The higher rate in April likely is a result of

synchronous breeding at the start of the reproductive season. Breeding

synchrony in small mammals diminishes progressively from the start of

the breeding season and disappears after the second litters are born, in

part because of increasing variation in litter interval among overwintered

females but mostly as a result of spring-born females entering the

breeding population (at 45-60 days of age for cotton rats).

Nothing is known of the actual litter intervals of cotton rats in

natural populations, but longer post-partum mating intervals would

lower the maximum observable pregnancy rate below the 64.3% value.

The observed pregnancy rate during the breeding season (68.7%) slightly

exceeded the theoretical value; that can be explained by sampling error

or, more likely, by changes in the behavior of pregnant females.

54

Robert K. Rose and Michael H. Mitchell

Randolph et al. (1977) found that the fat accumulated during the last

half of pregnancy was used during lactation, when energy demands

outstripped the female’s speed in processing food. It is plausible that

females in the later stages of pregnancy would be increasingly attracted

to the high-energy food source (mixed seeds) that was used as bait.

Although Dunaway and Kaye (1964) did not calculate monthly

pregnancy rates, they did observe low levels and apparently sporadic

breeding throughout what they judged to be a relatively mild Tennessee

winter. In Oklahoma, Goertz (1965) found no pregnant females during a

severe winter, but he did find pregnant females during November,

December, and February of a milder winter. Goertz reported highest

pregnancy rates during May to September. Haines (1961) recorded no

embryos in Texas cotton rats from October to February, but he did

record corpora lutea throughout the year. Haines observed the highest

pregnancy rates between February and July, with low rates after

September and the lowest rates in December. Thus, the breeding season

seems to be somewhat earlier in Texas compared with Tennessee,

Oklahoma, or Virginia.

In duration, methods, and analysis, our study most closely parallels

that of McClenaghan and Gaines (1978), conducted near Lawrence,

Kan. They found no breeding from November through March, which is

similar to what we observed in the Virginia population. The pregnancy

rate in Kansas was low (30%) in April, highest (over 80%) in May, and

generally greater than 70% from June through October. Overall, the

patterns of breeding in Virginia and Kansas were similar for both sexes.

Litter Size

Within a species, litter size is affected by several interacting factors,

including age, parity, body weight, and nutritional state (Sadleir 1969).

In a recent exhaustive review, Cameron and McClure (1988) examined

the patterns of breeding in female Sigmodon hispidus by evaluating

published and unpublished laboratory and field data. Using a stepwise

multiple regression analysis on data from 18 studies, Cameron and

McClure (1988: table 2) examined the patterns of geographic variation

in litter size and the effects of body size on litter size. By finding

latitude, longitude, and body length to be significantly associated with

mean litter size, their analysis “confirmed the existence of both north-

south and east-west variation in litter size.” Largest litters were reported

for the large females of the north-central states.

A further analysis “indicated that latitudinal and longitudinal

variation in litter size were due primarily to differences among subspecies”

(Cameron and McClure 1988). Specifically, S. hispidus texianus, which

had the largest litters at 7.20 ± 0.23 SE, averaged 8.35 ± 0.35 in Kansas

Hispid Cotton Rat

55

but only 5.10 ± 0.37 in coastal Texas (Houston). Nutrition may also

contribute to these differences within this subspecies (Cameron and

McClure 1988; table 6). Although litters of S. hispidus virginianus , the

subspecies in Virginia, were significantly larger than those of Mexican

and Central American subspecies, they were significantly smaller than

those of S. h. texianus (Cameron and McClure 1988).

The litter size of 5.00 for Virginia cotton rats lies in the range of

values reported from other studies (Cameron and McClure 1988: table

2), although on the low side for “northern populations.” Populations

from Tennessee averaged 6.1 embryos per litter (Dunaway and Kaye

1961), from Oklahoma 6.0 (Goertz 1965), from western Kansas 6.7

(Fleharty and Choate 1973), and from eastern Kansas 9.0 (McClenaghan

and Gaines 1978). Furthermore, laboratory animals derived from

Houston, Kansas, and Tennessee populations and raised by McClure at

Indiana University remained significantly different in average litter size

even after 16-28 generations and 8-12 years in the laboratory (Cameron

and McClure 1988: figure 2). Thus, the determination of litter size in

Sigmodon hispidus is complex, involving both genetic and environmental

factors.

Although Lawrence, Kan., and Portsmouth, Va., are both near 37°

N latitude, the Kansas winters are longer and colder (average 2° C), in

the absence of moderating oceanic effects. In coastal Virginia, snow falls

only once or twice a year and periods of freezing weather rarely last

more than a few days. Despite the more moderate conditions in Virginia

(Cameron and McClure 1988: table 4), the Virginia cotton rats did not

breed longer than the Kansas cotton rats, and Virginia litter sizes as well

as body sizes were significantly smaller. However, female cotton rats in

Virginia were pregnant at nearly maximum levels throughout the breeding

season (Fig. 1), and there was a trend (0. 1 > P> 0.05) for multiparous

females to have larger litters than primiparous females. Thus, differences

in age of onset of breeding and in longevity (neither of which was

measured in these studies) may be important in affecting geographic

differences in the dynamics of these populations.

Male Breeding

The breeding season of males began in February and lasted to

November (Fig. 2). Based on the breeding criterion of convolutions in

the cauda epididymides, 73.3% of males were in breeding condition in

February and 100% were fertile from March through June. McClenaghan

and Gaines (1978), who also used epididymal convolutions to determine

breeding condition in males, did not find 100% breeding in any month.

Their highest monthly rates were just under 90% in June and August;

and in all other months during the breeding season except May, fertility

56

Robert K. Rose and Michael H. Mitchell

rates were less than 60%. Haines (1961) measured spermatogenesis and

found that the production of sperm remained high from February

through October (which was the breeding season for males in our

study). Dunaway and Kaye (1964), who assessed male reproduction

based on live-caught animals, observed that just under 100% of males

had descended testes (i.e. were mature) during the June-September

period. They noted a decline in the percentages of mature males from

October to December, but by January, the proportion of males with

descended testes again began to increase.

Testicular mass is closely related to reproductive condition in males

(Haines 1961). Our results agree with those of McClenaghan and Gaines

(1978), who found testicular mass to be highest from June through

September, also the peak breeding months for males as determined by

convolutions of the cauda epididymides in the present study. In Kansas,

smallest testicular masses were recorded for December, but we found

the smallest testicular masses in the October to December period, with

dramatic monthly increases from December through April. Testicular

mass remained high from April through July and then declined sharply

(Fig. 3), probably at first because of the recruitment of young males in

the trappable population and later also because of testicular regression

of adult males. Goertz (1965) reported large testicular masses from

February through September and low values in the remaining months.

In Texas, Haines (1961) reported spermatogenesis in males with the

largest testicular masses; he provided perhaps the best available

information on the relationship between these two variables. He found

the largest average testicular mass per 10 g of body mass during the

period from February through August, after which testicular mass

declined until November. McClenaghan and Gaines (1978) and Haines

(1961) found that testicular regression resulted in a reduction to about

1 / 30th of the maximum testicular mass, compared with a value of about

1 /26th in our study.

Breeding Season

During the breeding season, 68.7% of females were pregnant, and

in most of the same months all males were judged to be fertile. The

breeding season in males started one month earlier and ended one

month later than the breeding season in females. That pattern is

common in mammals (Sadleir 1969), and it is interpreted as adaptive in

that the energy costs for breeding in females are greater than those in

males. As a result of the earlier onset of fertility in males, mature males

are ready to copulate and produce fertile matings when females undergo

the first estrous cycle of the spring.

The breeding season in Virginia closely paralleled that found by

McClenaghan and Gaines (1978) in Kansas. Both locations are at or

57

Hispid Cotton Rat

near the northern limit of distribution for hispid cotton rats; therefore,

it is not surprising to find some similarities. Differences also were noted.

Although the breeding season in Virginia started a month earlier in both

sexes, it lasted until October for females in both Kansas and Virginia.

Another similarity in the breeding of Sigmodon in these two studies is

the percentage of breeding females; the rate was relatively low during

the first month in which pregnancies occurred and then rose sharply to

near the maximal rate in the following month. In McClenaghan and

Gaines (1978), this trend must be inferred because their study ended in

April. However, if the high May level at the start of their study can be

extrapolated to the preceding April, a large increase in breeding level

occurred at the same time in Kansas and Virginia.

In Texas (Haines 1961), the breeding season began in the same

month as in Virginia, but it ended one month sooner. This is an

unexpected result if we assume that the breeding season has been

shortened at more northerly locations because of the constraints of

winter on the energy budgets of mammals. We would expect the

breeding season to be longer at more southerly locations. Goertz (1965),

who found pregnant females in some winter months, believed that

breeding was possible in Oklahoma under the favorable conditions of

mild winters.

Effects of Body Size

Several factors affect patterns of body size in Sigmodon hispidus,

including sex, latitude, subspecies, and nutrition. In some, but not all,

populations males are larger than females, and northern populations

tend to have larger skeletal sizes and, in some seasons, higher fat

content. Sigmodon h. texianus is significantly larger than all other

subspecies (Cameron and McClure 1988). McClenaghan (1977) found

generally larger skull and skeletal variables for cotton rats from northern

populations (Kansas and Virginia) than from southern localities (Mexico),

but Kansas and Virginia populations differed in only one skeletal

feature. However, the seasonal pattern of body growth of Virginia

cotton rats differs from that of Kansas cotton rats, in which Slade et al.

(1984) found that the large adult animals lost weight over the winter. By

contrast, in Virginia, the males in particular gained body mass steadily

throughout the winter months (Fig. 4). This pattern of winter increase is

evident but less well defined for body length (Fig. 5), although males did

increase in length nearly every month from October to May or June.

Mean body mass for females was low in October (79.7 ± 6.77 g)

but it rose sharply to 98.6 ± 10.19 g in November. The mean values for

December and February were low, indicating that females were not

gaining weight during this time. Because there was no breeding during

these months, the lower mean mass in winter cannot be a result of the

58

Robert K. Rose and Michael H. Mitchell

recruitment of young, lightweight animals; therefore, we can assume

that females either lose mass or fail to gain significant mass during the

winter months. From February through July, the monthly mean mass

of females rose steadily, indicating a real increase in body mass during

this time. Then, when the young of the year finally entered the trappable

population, the monthly mean mass of females (and males) declined.

Patterns of body length were similar to those of body mass.

Although both length and mass are measures of body size, length may

be a more reliable index of body growth in S. hispidus because (1)

animals lose mass but not length during starvation or during winter, (2)

both sexes divert resources away from growth and towards reproduction

during the breeding season, and (3) females store up body fat during

pregnancy in preparation for the greater energy demands during lactation

(Randolph et al. 1977).

In conclusion, cotton rats in southeastern Virginia seem to be well

adapted to the northern limit of their present distribution on the East

Coast; their March-to-October breeding season and the sustained growth

of overwintering individuals suggest high survival rates of both young

and adults during the winter months. The modest litter size may

indicate that, unlike Kansas populations, Virginia populations have not

been selected for larger litter sizes to compensate for winter mortality.

ACKNOWLEDGMENTS.— We thank Patricia Hopkins, Sean Priest,

and David Wade for field work and other assistance with this project.

LITERATURE CITED

Cameron, G. N., and P. A. McClure. 1988. Geographic variation in life history

traits of the hispid cotton rat ( Sigmodon hispidus). In Evolution of Life

Histories of Mammals: Theory and Patterns (M. Boyce, editor). Yale Univ.

Press.

Dunaway, P. B., and S. V. Kaye. 1961. Studies of small mammal populations

on the radioactive White Oak Lake bed. Trans. 26th N. Am. Resource

Conf.: 167-1 85.

Dunaway, P. B., and S. V. Kaye. 1964. Weights of cotton rats in relation to

season, breeding, and environmental radioactive contamination. Am. Midi.

Nat. 71:141-155.

Fleharty, E. D., and J. R. Choate. 1973. Bioenergetic strategies of the cotton

rat, Sigmodon hispidus. J. Mammal. 54:680-692.

Genoways, H. H., and D. A. Schlitter. 1967. Northward dispersal of the hispid

cotton rat in Nebraska and Missouri. Trans., Kans. Acad. Sci. 69:356-357.

Goertz, J. W. 1965. Reproductive variation in cotton rats. Am. Midi. Nat.

74:329-340.

Hispid Cotton Rat

59

Haines, H. 1961. Seasonal changes in the reproductive organs of the cotton

rat, Sigmodon hispidus. Texas J. Sci. 8:219-230.

Hall, E. R. 1981. The Mammals of North America. John Wiley and Sons, New

York.

Jameson, E. W., Jr. 1950. Determining fecundity in male small mammals. J.

Mammal. 31:433-436.

Keller, B. L., and C. J. Krebs. 1970. Microtus population biology: III.

Reproductive changes in fluctuating populations of M. ochrogaster and M.

pennsylvanicus in southern Indiana, 1965-1967. Ecol. Monogr. 40:263-294.

Kilgore, D. L., Jr. 1970. The effects of northward dispersal on growth rate of

young, size of young at birth, and litter size in Sigmodon hispidus. Am.

Midi. Nat. 84:510-520.

Lewis, J. B. 1944. Cotton rat in lower piedmont Virginia. J. Mammal.

25:195-196.

McClenaghan, L. R., Jr. 1977. Genic variability, morphological variation and

reproduction in central and marginal populations of Sigmodon hispidus.

Ph.D. dissertation, Univ. Kansas, Lawrence.

McClenaghan, L. R., Jr., and M. S. Gaines. 1978. Reproduction in marginal

populations of the hispid cotton rat {Sigmodon hispidus) in northeastern

Kansas. Occas. Pap., Univ. Kan. Mus. Nat. Hist. 74:1-16.

Meyer, B. J., and R. K. Meyer. 1944. Growth and reproduction of the cotton

rat, Sigmodon hispidus hispidus, under laboratory conditions. J. Mammal.

25:107-129.

Pagels, J. F. 1977. Distribution and habitat of cotton rat {Sigmodon hispidus )

in central Virginia. Va. J. Sci. 28:133-135.

Patton, C. P. 1941. The eastern cotton rat in Virginia. J. Mammal. 22:91.

Randolph, P. A., J. C. Randolph, K. Mattingly, and M. M. Foster. 1977.

Energy costs of reproduction in the cotton rat, Sigmodon hispidus. Ecology

58:31-45.

Sadleir, R. M. F. S. 1969. The ecology of reproduction in wild and domestic

mammals. Methuen and Co., Ltd.

Sauer, J. R. 1985. Mortality associated with severe weather in a northern

population of cotton rats. Am. Midi. Nat. 113:188-189.

Slade, N. A., J. R. Sauer, and G. E. Glass. 1984. Seasonal variation in field-

determined growth rates of the hispid cotton rat {Sigmodon hispidus ). J.

Mammal. 65:263-270.

Smith, M. H., and J. T. McGinnis. 1968. Relationships of latitude, altitude,

and body size to litter size and mean annual production of offspring in

Peromyscus . Res. Pop. Ecol. (Japan) 10:115-126.

Zimmerman, E. G. 1970. Karyology, systematics and chromosomal evolution

in the rodent genus, Sigmodon. Mich. State Univ., Mus. Publ., Biol. Ser.

4(9):385-454.

Accepted 14 April 1989

60

.

•• * . • • ' ■ . ■ : '*3

Occurrence of the Milliped Auturus erythropygos erythropygos

(Brandt) in Virginia (Polydesmida: Platyrhacidae)

Rowland M. Shelley

North Carolina State Museum of Natural Sciences

P. O. Box 27647, Raleigh, North Carolina 27611

The milliped Auturus erythropygos erythropygos (Brandt) is known

from scattered localities ranging from the Fall Zone of northern North

Carolina to the northern Coastal Plain of Georgia and westward into

the Kings Mountain inselberg and the western Piedmont Plateau of

North Carolina (Shelley 1978, 1982, Filka and Shelley 1980). The

northernmost site, 8.3 km (5.2 miles) WSW of Gaston, Northampton

Co., N.C, is the only one north of the Roanoke River, and because it is

only 3.2 km (2 miles) S of the Virginia border, I (Shelley 1978, 1982)

predicted discovery in that state.

A Virginia locality can now be confirmed. I collected two males

and one female on 1 1 May 1988, in Brunswick Co., 9.6 km (6 miles) SE

of Lawrenceville, along Va. Hwy. 670 just south of the Meherrin River,

about 19 km (11.9 miles) N of the North Carolina state line and due

north of the Northampton Co. site. The specimens were encountered in

typical habitat for American platyrhacids, under bark and in a moist,

rotting oak stump at the base of a wooded slope. They conform to the

anatomical illustrations and description of A. e. erythropygos (Shelley

1982). Efforts to find the milliped in adjacent counties to the north, east,

and west in May and September, 1988, were unsuccessful; therefore, the

extent of the population north of the Roanoke River is unknown. Both

the Brunswick and the Northampton samples were obtained in May,

but those from western North Carolina were recorded in April, July,

August, October, and December of other years (Shelley 1982), which

suggests that the milliped should be collectable in Virginia during the

same months. The Brunswick Co. locality is the northernmost for the

species, genus, tribe, and family along the Atlantic Coast and the only

known occurrence of the Platyrhacidae in eastern Virginia. Euryurus

leachii (Gray) occurs in the westernmost counties near West Virginia,

Kentucky, and Tennessee (Hoffman 1978).

The Virginia specimens of A. e. erythropygos are housed in the

North Carolina State Museum of Natural Sciences invertebrate collection,

catalog number A4897. The known distribution of the Platyrhacidae in

Virginia is depicted in Fig. 1.

Brimleyana 16:61-62, July 1990

61

62

Rowland M. Shelley

Fig. 1. Distribution of the milliped family Platyrhacidae in Virginia. Dots,

Euryurus leachii ; square, Auturus erythropygos erythropygos.

LITERATURE CITED

Filka, M. E., and R. M. Shelley. 1980. The milliped fauna of the Kings

Mountain Region of North Carolina (Arthropoda: Diplopoda). Brim-

leyana 4: 1-42.

Hoffman, R. L. 1978. North American millipeds of the genus Euryurus

(Polydesmida: Platyrhacidae). Trans. Am. Entomol. Soc. 104:37-68.

Shelley, R. M. 1978. Millipeds of the eastern piedmont region of North

Carolina, U.S.A. (Diplopoda). J. Nat. Hist. 12:37-79.

Shelley, R. M. 1982. Revision of the milliped genus Auturus (Polydesmida:

Platyrhacidae). Canadian J. Zool. 60:3249-3267.

Accepted 30 May 1989

Kleptoparasitism of a River Otter, Lutra canadensis ,

by a Bobcat, Felis rufus, in South Carolina

(Mammalia: Carnivora)

James F. Bergan1

Savannah River Ecology Laboratory

Drawer E, Aiken, South Carolina 29801

ABSTRACT. — In January 1985, a bobcat, Felis rufus, was observed

robbing a river otter, Lutra canadensis, of an American coot, Fulica

americana. Observations of interactions between these two carnivores

are practically nonexistent. This observation indicates that bobcats can

assume an aggressive kleptoparasitic mode of behavior toward river

otters.

Bobcats, Felis rufus, are opportunistic in prey selection, usually

choosing abundant and easily captured prey (Pollack 1951, Beasom and

Moore 1977, King et al. 1983). Carrion is also consumed (McCord

1974). In spite of this general foraging strategy, there are no published

reports describing interspecific prey-stealing behavior (i.e. kleptoparasit-

ism) by the bobcat. I observed a case of kleptoparasitism involving a

bobcat and a river otter, Lutra canadensis. The incident occurred on

Par Pond, a 1,120-ha cooling reservoir located at the Savannah River

Plant, Barnwell Co., S.C.

At 1645 hours on 21 January 1985, while observing waterfowl with

a spotting scope from the reservoir dam, I observed a river otter

swimming. The otter climbed onto a 1-m ice shelf along the shoreline.

Approximately 50 American coots {Fulica americana ), 15 pied-billed

grebes {Podilymbus podiceps), 5 buffleheads {Bucephala albeola), and

10 lesser scaup {Ay thy a affinis) were feeding in the vicinity of the otter.

The otter remained on the shore approximately 1 minute, returned to

the water, and swam toward the feeding aggregation of waterfowl about

45 m away. After approaching within 10 m of several coots, the otter

submerged; it resurfaced in 35 seconds and began swimming to shore

with a coot in its mouth. The coot’s wing was flapping as the otter

climbed onto the ice shelf. The otter then plucked the breast feathers

and fed upon the breast. Feeding continued as the abdominal cavity was

opened and a portion of the viscera consumed. At this point, the otter

appeared startled, quickly looked about, and jumped into the water. A

bobcat came into view and appeared to emit a snarl as the otter swam

'Present address: Department of Range and Wildlife Management, Texas Tech University,

Lubbock, TX 79409.

Brimleyana 16:63-65, July 1990

63

64

James F. Bergan

from the shoreline. The bobcat picked up the coot and quickly made its

way into the thick vegetation that bordered the shoreline. Upon later

inspection of the site of the interaction, all that remained were coot

feathers and some entrails.

Weather conditions were uncommonly harsh in the area on z\

January, with winds out of the northwest at 8-10 km/ hr and an air

temperature of-5°C. The incident took place approximately 75 m from

my location on the dam.

This incident is unique in the literature not only because of the

kleptoparasitism by a bobcat but also because the bobcat displaced an

atypical competitor, a river otter, from a prey item. A report of a

bobcat killing a young river otter is the only previously documented

description of a direct interaction between these two mammalian carni-

vores (Young 1958).

Several incidents of river otter predation upon aquatic birds have

been documented. Meyerriecks (1963) reported a river otter preying

upon a common gallinule ( Gallinula chloropus ) in Florida. In Alaska,

Quinlan (1983) found that river otters killed over 75% of the immature

fork-tailed storm-petrels ( Oceanodroma furcata) and Leach’s storm-

petrels ( O . leucorhoa) in a single colony. Cahn (1937) observed a river

otter’s unsuccessful attempt to catch a coot in Ontario, Canada. Most

studies of river otter food habits have found that birds make up a minor

portion of the diet (Loranger 1981, Knudsen and Hale 1968, Stenson et

al. 1984).

ACKNOWLEDGMENTS. — Helpful comments were provided by

John J. Mayer and Loren M. Smith. Work was funded by the Savannah

River Ecology Laboratory (DOE Contract DE-AC09-76SROO819, Univer-

sity of Georgia Institute of Ecology) and Texas Tech University (College

of Agricultural Sciences Manuscript T-9-575).

LITERATURE CITED

Beasom, S. L., and R. A. Moore. 1977. Bobcat food habit response to a change

in prey abundance. Southwest. Nat. 21:451-457.

Cahn, A. R. 1937. The mammals of the Quetico Provincial Park of Ontario.

J. Mammal. 18:19-30.

King, A. M., R. A. Lancia, D. Miller, D. K. Woodward, and J. D. Hair. 1983.

Winter food habits of bobcats in North Carolina. Brimleyana 9: 1 1 1-122.

Knudsen, G. J., and J. B. Hale. 1968. Food habits of otters in the Great

Lakes region. J. Wildl. Manage. 32:89-93.

Loranger, A. J. 1981. Late fall and early winter foods of the river otter ( Lutra

canadensis) in Massachusetts, 1976-1978. Pages' 599-605 in Worldwide

Furbearer Conference Proceedings. Vol. 1. August 3-1 1, 1980, Frostburg,

Md. (J. A. Chapman and D. Pursley, editors).

River Otter Kleptoparasitism

65

McCord, C. M. 1974. Selection of winter habitat by bobcats ( Lynx rufus) on the

Quabbin Reservation, Massachusetts. J. Mammal. 55:428-437.

Meyerriecks A. J. 1963. Florida otter preys on common gallinule. J. Mam-

mal. 44:425-426.

Pollack, E. M. 1951. Food habits of bobcats in the New England states. J.

Wildl. Manage. 15:209-213.

Quinlan, S. E. 1983. Avian and river otter predation in a storm-petrel colony. J.

Wildl. Manage. 47:1036-1043.

Stenson, G. B., G. A. Badgero, and H. D. Fisher. 1984. Food habits of the river

otter Lutra canadensis in the marine environment of British Columbia.

Can. J. Zool. 62:88-91.

Young, S. P. 1958. The Bobcat of North America. Stackpole Co., Harrisburg,

Pa., and Wildl. Manage. Inst., Washington, D.C.

Accepted 25 August 1989

66

A DISTRIBUTIONAL SURVEY

OF NORTH CAROLINA MAMMALS

by

David S. Lee, John B. Funderburg, Jr., and Mary K. Clark

This book lists all the mammals of North Carolina and offers

species accounts and range maps for all of the non-marine species.

Introductory chapters describe the plant communities of the state as

they relate to mammal distribution and discuss local zoogeographic

patterns.

1982 72 pages Softbound ISBN 0-917134-04-4

Price: $5, postpaid. North Carolina residents add 5% sales tax. Please make

checks payable in U.S. currency to NCDA Museum Extension Fund.

Send to MAMMAL BOOK, N.C. State Museum of Natural Sciences,

P.O. Box 27647, Raleigh, NC 27611.

Spring Movement Patterns of Two Radio-tagged

Male Spotted Turtles

Jeff Lovich

Savannah River Ecology Laboratory

Drawer E, Aiken, South Carolina 29801

ABSTRACT. — Spring movements of male spotted turtles ( Clemmys

guttata ) on the upper Coastal Plain of South Carolina may be extensive.

Movement in 24 hours for two turtles equipped with small radio

transmitters ranged from 0 to 423 m. The typical activity pattern

involved a series of movements throughout each pond that was occupied,

followed by overland travel to the next nearest body of water. Back-

tracking to previously occupied habitats was observed only once.

These movements may reflect mate-search activity and support the

concept of “transient behavior.”

Males turtles of many species appear to move greater distances

than females at certain times of the year (Chelazzi and Francisci 1979,

Morreale et al. 1984, Parker 1984, Berry 1986, Gibbons 1986). This

difference is thought to be representative of divergent reproductive

strategies between the sexes, but other explanations for movement are

possible, including seasonal migrations to and from overwintering sites

and departure from an unsuitable habitat (Gibbons 1986). In general,

the reproductive strategy hypothesis predicts that male reproductive

success is dependent on the number of mating opportunities available

(Trivers 1972, Williams 1975, Maynard Smith 1978). Under this assump-

tion, a male turtle’s reproductive success could benefit by increasing

movements in search of females during the breeding season (Morreale et

al. 1984). Occasionally, movements are extensive and some males may

act as “transients,” moving regularly throughout the active season

without recrossing areas previously traversed (Kiester et al. 1982, Parker

1984). Observations of the spotted turtle, Clemmys guttata (Schneider),

made during a study of its seasonal activity patterns in South Carolina

(Lovich 1988), support the concept of extensive movements by males

during the breeding season. The purpose of this note is to report those

observations.

The study site was located along Risher Road on the Savannah

River Plant in Barnwell Co., S.C. This area is characterized by scattered,

shallow, ephemeral marshes, and cypress-tupelo ponds separated by

pine plantations and clearcuts. Two adult male C. guttata , designated

ACJ and ACI, were collected on 2 March 1987 and equipped with small

Brimleyana 16:67-71, July 1990

67

68

Jeff Lovich

radio transmitters. The plastron lengths of the turtles were 93 mm and

99 mm, respectively. A transmitter was attached to the posterior portion

of each carapace and accounted for 10% or less of the animal’s mass.

The turtles were released at the respective points of capture on 6 March

and located daily until 15 April. From 15 to 23 April, positions were

determined every 48 hours. Both turtles were periodically brought back

to the laboratory during the study period for battery replacement or

transmitter-package repair. In spite of the downtime, 35 observations of

movement were obtained for each turtle. Every attempt was made to

minimize disturbance to the turtles during tracking. After each turtle

was located, the straight-line distance to the previous point of capture

was measured with a meter tape or determined from aerial photographs.

A summary of major movement patterns is shown in Fig. 1.

Movements in 24 hours ranged from 0 to 423 m. Twenty-two percent of

all daily changes in location were greater than 100 m. Rates of movement

reached 20.7 m/hour with a mean of 2.7 m/hour. The distances moved

between captures did not differ significantly between the two turtles

(Mann-Whitney U = 617; df = 1; P = 0.80). ACJ moved a total of

2,750 m and ACI 1,843 m. The greatest straight-line distance achieved

from original point of capture was more than 1,000 m for each turtle.

During these movements, each animal occupied three separate aquatic

habitats (marshes and ponds) (Fig. 1). Known time spent in each habitat

ranged from 4 to 20 days. The typical activity pattern involved a series

of movements through an aquatic habitat followed by overland

movements (up to 2 days) to the next nearest body of water. Return by

an individual to a previously occupied aquatic habitat was observed

only once.

Detailed descriptions of short-term movements have not previously

been reported for this species. However, Ernst (1968) found that a small

proportion of C. guttata in a Pennsylvania population returned to the

original point of capture 4 to 64 days after being moved 805 m upstream

from his study site. Netting (1936) reported movement of C. guttata

from an upland hibernation site to a small swamp, but provided no

further data. Ernst (1976) found that normal daily movements (based on

hand recaptures) were rarely more than 20 m, but males were occasionally

captured up to 250 m from water during the mating season. Seasonal

microhabitat selection was reported by Ward et al. (1976), but daily

movement data were not provided. The results of the present study,

although provisional, suggest transient behavior (as defined by Kiester

et al. 1982, Parker 1984) as well as oriented long-distance movement in

male C. guttata. Because these movements occurred during the mating

season (Ernst 1976), they may be a reflection of mate-search activity, as

suggested by Morreale et al. (1984). It is not likely that these short-term

Movement Patterns of Spotted Turtles

69

Fig. 1. Map of study area showing major aquatic habitats (temporary marshes

and ponds) along Risher Road. Paths represent major movement patterns and

do not include every capture point. Starting points are designated with a star.

movements were in response to major changes in the aquatic habitats

since water levels remained relatively stable during the period of study.

Additionally, no correlations between environmental temperatures and

turtle movements were detected.

Comparative data are not available for females, but I predict that

they will show greater site fidelity than males. Ernst (1970) reported

similar home-range size in male and female C. guttata , but his results

70

Jeff Lovich

may be an artifact of long-term hand recapture techniques. Kiester et al.

(1982) considered radiotelemetry to be a basic prerequisite for determining

transient behavior in turtle populations.

The possible importance of transient males in maintaining gene

flow between adjacent turtle populations was suggested by Kiester et al.

(1982). In fact, genetic exchange among populations of freshwater

turtles in adjacent aquatic habitats has been demonstrated electrophoretically

(Scribner et al. 1984). The phenotypic homogeneity exhibited by the

widely distributed spotted turtle (Ernst and Barbour 1972) may be a

result of this behavior.

ACKNOWLEDGEMENTS.— I thank Jim Knight, Marie Fulmer,

Aileen Adriano, Steve Gotte, and Tony Mills for field assistance. Jim

Knight, Whit Gibbons, Mara McDonald, and two anonymous reviewers

provided useful criticisms of an earlier version of the manuscript.

Special thanks are extended to Justin Congdon and Robert Fischer for

showing me how to use radiotelemetry effectively. This research was

supported by Contract DE-AC09-76SROO819 between the U.S. Department

of Energy and the University of Georgia.

LITERATURE CITED

Berry, K. H. 1986. Desert tortoise (Gopherus agassizii) relocation: implications

of social behavior and movements. Herpetologica 42: 113-125.

Chelazzi, G., and F. Francisci. 1979. Movement patterns and homing

behavior of Testudo hermanni Gmelin (Reptilia Testudinidae). Monitore

Zool. Ital. 13:105-127.

Ernst, C. H. 1968. Homing ability in the spotted turtle, Clemmys guttata

(Schneider). Herpetologica 24:77-78.

Ernst, C. H. 1970. Home range of the spotted turtle, Clemmys guttata

(Schneider). Copeia 1970:391-393.

Ernst, C. H. 1976. Ecology of the spotted turtle, Clemmys guttata (Reptilia,

Testudines, Testudinidae), in southeastern Pennsylvania. J. Herpetol. 10:25-33.

Ernst, C. H., and R. W. Barbour. 1972. Turtles of the United States. Univ.

Press Kentucky, Lexington.

Gibbons, J. W. 1986. Movement patterns among turtle populations: applicability

to management of the desert tortoise. Herpetologica 42:104-1 13.

Kiester, A. R., C. W. Schwartz, and E. R. Schwartz. 1982. Promotion of gene

flow by transient individuals in an otherwise sedentary population of box

turtles ( Terrapene Carolina triunguis ). Evolution 36:617-619.

Lovich, J. E. 1988. Geographic variation in the seasonal activity cycle of

spotted turtles, Clemmys guttata. J. Herpetol. 22:482-485.

Maynard Smith, J. 1978. The ecology of sex. Pages 159-179 in Behavioral

Ecology: An Evolutionary Approach (J. R. Krebs and N. B. Davies,

editors). Sinauer Associates, Inc., Sunderland, Mass.

Movement Patterns of Spotted Turtles

71

Morreale, S. J., J. W. Gibbons, and J. D. Congdon. 1984. Significance of

activity and movements in the yellow-bellied slider turtle ( Pseudemys

scripta). Can. J. Zool. 62:1038-1042.

Netting, M. G. 1936. Hibernation and migration of the spotted turtle,

Clemmys guttata (Schneider). Copeia 1936:112.

Parker, W. S. 1984. Immigration and dispersal of slider turtles Pseudemys

scripta in Mississippi farm ponds. Am. Midi. Nat. 1 12: 280-293.

Scribner, K. T., M. H. Smith, and J. W. Gibbons. 1984. Genetic differentiation

among local populations of the yellow-bellied slider turtle ( Pseudemys

scripta). Herpetologica 40:382-387.

Trivers, R. L. 1972. Parental investment and sexual selection. Pages 136-179 in

Sexual Selection and the Descent of Man (B. Campbell, editor). Aldine-

Atherton, Inc., Chicago.

Ward, F. P., C. J. Hohmann, J. F. Ulrich, and S. E. Hill. 1976. Seasonal

microhabitat selections of spotted turtles ( Clemmys guttata ) in Maryland

elucidated by radioisotope tracking. Herpetologica 32:60-64.

Williams, G. C. 1975. Sex and Evolution. Princeton Univ. Press, Princeton,

N.J.

Accepted 28 August 1989

72

ENDANGERED, THREATENED, AND

RARE FAUNA OF NORTH CAROLINA

PART I.

A RE-EVALUATION OF THE MAMMALS

Edited by Mary Kay Clark

This book is a report prepared by a committee appointed in 1985

by the North Carolina State Museum of Natural Sciences to re-evaluate

the list of mammals presented in Endangered and Threatened Plants

and Animals of North Carolina (John E. Cooper, Sarah S. Robinson,

and John B. Funderburg, editors. N.C. State Mus. Nat. Hist., Raleigh,

1977), which is now out of print. Committee members were Mary Kay

Clark, David A. Adams, William F. Adams, Carl W. Betsill, John B.

Funderburg, Roger A. Powell, Wm. David Webster, and Peter D.

Weigh The report treats 21 species listed in the following status

categories: Endangered (5), Threatened (1), Vulnerable (6), and

Undetermined (9). Most species accounts discuss the animal’s physical

characteristics, range, habitat, life history and ecology, special sig-

nificance, and status (including the rationale for the evaluation and

recommendations for protection) and provide a range map and an

illustration of the animal’s external characters. Ruth Brunstetter and

Renaldo Kuhler illustrated the book. An introductory section contributed

by Ms. Clark discusses the changes in status that occurred in the decade

between 1975 and 1985. It also mentions efforts to protect marine

mammals and includes a checklist of the cetaceans known from North

Carolina.

1987 52 pages Softbound ISBN 0-917134-14-1

Price: $5 postpaid. North Carolina residents add 5% sales tax. Please make

checks payable in U.S. currency to NCDA Museum Extension Fund.

Send order to: ETR MAMMALS, N.C. State Museum of Natural Sciences,

P.O. Box 27647, Raleigh, NC 27611.

New Records of the Distribution

and the Intestinal Parasites of the

Endangered Northern Flying Squirrel, Glaucomys sabrinus

(Mammalia: Sciuridae), in Virginia

John F. Pagels

Department of Biology

Virginia Commonwealth University

Richmond, Virginia 23284

Ralph P. Eckerlin

Natural Sciences Division

Northern Virginia Community College

Annandale, Virginia 22003

John R. Baker

Virginia Department of Game and Inland Fisheries

Marion, Virginia 24354

AND

Michael L. Fies

Virginia Department of Game and Inland Fisheries

Charlottesville, Virginia 22901

ABSTRACT. — Three specimens of Glaucomys sabrinus are reported

from localities in the mountains of Grayson and Highland counties,

Virginia. Thirty-three other G. sabrinus were captured and released at

or near these sites. Only one specimen had been previously recorded in

Virginia, in Smyth Co. Five sites where we captured G. sabrinus had a

mean elevation of 1,354 m (1,097-1,615 m), and the southernmost sites

were at the greatest elevation. The typical habitat was a mixed forest

of red spruce or other northern-type conifers and northern hardwoods.

Three species of intestinal nematodes were recovered from the

squirrels: Citellinema bifurcatum , Strongyloides robustus, and Syphacia

thompsoni.

The northern flying squirrel, Glaucomys sabrinus (Shaw), has an

extensive distribution in the northern United States and Canada. In the

southern Appalachian Mountains it is extremely rare and is considered

a relict of the ice ages. It is known from only a few scattered populations

in West Virginia, Tennessee, North Carolina, and Virginia (Wells-

Gosling and Heaney 1984). In 1985, both subspecies of the northern

flying squirrel that occur in the southern Appalachians, G. s. fuscus

Brimleyana 16:73-78, July 1990

73

74

Pagels, Eckerlin, Baker, and Fies

Miller and G. s. coloratus Handley, were listed as federally endangered

by the U.S. Fish and Wildlife Service. Weigl (1987) summarized factors

considered as major threats to survival of these subspecies. Among these

are loss of habitat, competition with the southern flying squirrel, G.

volans (L.), and a parasitic nematode ( Strongyloides sp.), a form that is

harbored by the southern flying squirrel without apparent harm but that

may be lethal or debilitating when transferred to G. sabrinus.

The present report on G. sabrinus in Virginia updates distributional

data and provides information on its intestinal parasites. Some

ecological data are included herein. Payne et al. (1989) provide a more

extensive description of plant communities associated with G. sabrinus

in the southern Appalachians, including Virginia.

The only previously recorded specimen of G. sabrinus in Virginia

was trapped in 1959 by Handley at 1,615 m on Whitetop Mountain in

Smyth Co. (Handley 1980). The animal was captured in a snap trap

attached to a red spruce bole in a mixed forest of red spruce ( Picea

rubens), yellow birch ( Betula lutea), and sugar maple (Acer saccharum)

within a few hundred meters of an almost pure stand of red spruce.

In December 1985, one of us (J.R.B.) found the remains of a

female northern flying squirrel along the headwaters of a creek at 1,478

m in Grayson Co., 7.6 km E of the Smyth Co. site. The remains were

under the edge of a large rock and partially covered with leaves. Tracks

of a mink, Mustela vison Schreber, leading to and from the spot were

evident in the snow around the carcass and indicated predation as the

cause of death. The vegetation was a mixed forest of Fraser fir (Abies

fraseri ) and red spruce, with American beech (Fagus grandifolia ), yellow

birch, and red maple (Acer rubrum) the most prevalent canopy

hardwoods and with rhododendron an important understory component.

In April 1986, two specimens were taken about 336 km NE of the

Smyth Co. site, in the Laurel Fork area of Highland Co. These

specimens were also taken along a headwater stream. They were

accidentally trapped in Museum Special mouse traps during a study of

the fleas of Virginia by one of the authors (R.P.E.). These specimens

were captured in a northern hardwood forest dominated by yellow

birch, American beech, and sugar maple. An almost pure stand of

young red spruce was within 10 m of the capture site. Mountain laurel

(Kalmia latifolia ) was the dominant understory plant. This locality, at

1,158 m, is about 19 km E of Cheat Bridge, the nearest locality from

which G. 5. fuscus has been taken in West Virginia (C. W. Stihler,

personal communication). Several irregularly oriented mountain ridges

separate the two sites.

In addition to the specimen records, 33 northern flying squirrels

were captured and released at 4 of 22 nest-box locations in Virginia

75

Northern Flying Squirrel in Virginia

during an ongoing study that was initiated in 1985 (Fies and Pagels

1988). The locations included the Grayson site (19 captures) where the

specimen apparently killed by a mink was found and the Smyth Co. site

(10 captures) of Handley’s (1980) first record in 1959. The other sites

were in Highland Co., approximately 5 km S (1 capture) and 20 km S (3

captures), respectively, of the site where the snap-trapped specimens

described above were taken. These sites, at 1,097 m and 1,219 m, were

also mixed forests characterized by red spruce and various northern

hardwood species.

In West Virginia, where a nest-box and live-trapping study of G.

sabrinus has been under way for several years (Stihler et al. 1987), of

more than 60 captures, G. sabrinus was taken only in spruce or spruce-

northern hardwood forests (K. B. Knight and C. W. Stihler, personal

communication). However, as we found at one of our Highland Co.

sites where a stand of red spruce was 10 m away, red spruce does not

necessarily characterize the point of capture. Weigl (1987) observed,

“Although often associated with spruce-fir forests, this form is more

commonly captured in adjacent stands of mature hardwoods . . . .” It is

curious that no spruce or fir trees were present within miles of the

capture site of a specimen taken in 1935 in Tennessee (Handley 1953).

Payne et al. (1989) found red spruce in their transects at all 13 sites

where G. sabrinus was collected. Nearly all data indicate that both

northern-type conifers and northern hardwoods are necessary com-

ponents of the habitat of G. sabrinus.

The specimens from the Highland Co. site were examined for

parasites. Both of the squirrels harbored nematode parasites. The large

female (VCU 4629) had Syphacia thompsoni (3 males, 10 females) in the

cecum, Strongyloides robustus (31 females) in the anterior 10 cm of the

small intestine, and a single male Citellinema bifurcatum also in the

anterior small intestine. The smaller female (VCU 4630) had a female S.

robustus in the anterior 10 cm of the small intestine and 11 male S.

thompsoni in the cecum.

Only the formalin-fixed intestinal tract of the Grayson Co. specimen

(VCU 4615) was available for parasitological study. Two female S.

robustus were found in the anterior small intestine, and two male S.

thompsoni in the cecum. Unsporulated coccidean oocysts that averaged

21.8 pm by 14.3 pm were present but uncommon. By gross examination

slight hyperemia of the intestine was noted in the area surrounding the

31 individuals of S. robustus.

»•

All of these nematode species have been reported from sciurid hosts

in Virginia previously (Davidson 1976, Eckerlin 1985, Parker 1968,

1971, Price 1928) and from Glaucomys sabrinus elsewhere in the United

States (Eckerlin 1974, Rausch and Tiner 1948), but they have not

76

Pagels, Eckerlin, Baker, and Fies

previously been recorded from this host in Virginia. Voucher specimens

of Citellinema bifurcatum, Strongyloides robustus, and Syphacia

thompsoni have been deposited in the Helminth Collection of the U.S.

National Museum of Natural History with accession numbers 79566,

79567, and 79568, respectively.

The presence of only male S. thompsoni in the two squirrels that

had this species deserves comment. Although life cycle data for this

nematode are not available, it is generally agreed that male oxyurids do

not survive more than a short time after the females have been fertilized.

Consequently, females usually outnumber males, but we found 2 males

and 1 1 males in the absence of females. A unisexual infection of 32 male

S. thompsoni was reported from the red squirrel, Tamiasciurus

hudsonicus (Erxleben), in Wisconsin (Tiner and Rausch 1949). It was

suggested that the unusual sex ratio could be attributed to the occurrence

in an unnatural host. Glaucomys volans is the type host for S.

thompsoni, but the high prevalence in G. sabrinus suggests that the

latter species also is a suitable host. Clearly, more data are needed to

clarify the relationship between S. thompsoni and G. sabrinus .

The Highland Co. specimens (both skin and skull), VCU 4629,

female, and VCU 4630, female, measured respectively: total length 260,

250; tail vertebrae 120, 110; hindfoot 35, 36; ear 24, 26. The only

standard measurements available for the Grayson County specimen are

length of tail vertebrae 139 mm and hindfoot 38 mm. The partial

museum skin is deposited in the Virginia Commonwealth University

Mammal Collection (VCU 4615, female).

That G. sabrinus was only recently taken in the Laurel Fork area in

Highland Co., which has been sampled periodically over many years,

suggests that the squirrel is rare. However, it also suggests that the

apparent rarity may at least in part reflect the difficulty in trapping this

animal or the inadequacy of collecting techniques that have been used.

Weigl (1978) found that G. sabrinus and G. volans occasionally occur in

the same woodlot although perhaps only temporarily. Further, Weigl

(1978) found in captive populations that G. volans was able to control

nests more often than G. sabrinus could and that it was much more

aggressive than G. sabrinus in defending home area. One of us (J.F.P.)

handled and ear-tagged numerous G. volans at the Highland Co. sites

before and after the captures of the northern flying squirrels. C. O.

Handley, Jr. (personal communication) observed numerous individuals

of G. volans at one of the sites as much as 50 years ago. The species

have coexisted at Stuart Knob in West Virginia for at least 36 years

(Stihler et al. 1987). However, again perhaps suggestive of its great

rarity, low trappability, or temporary coinhabitance of the sites, we

77

Northern Flying Squirrel in Virginia

have had no recaptures or new captures of G. sabrinus at the Highland

Co. locations to date. Finally, it has also been noted that G. volans

carries a parasitic nematode of the genus Strongyloides , and it has been

suggested that this nematode is lethal to G. sabrinus (Weigl 1977). Each

of the three G. sabrinus that we were able to examine and report here

harbored S. robustus, and it appeared that each was a healthy adult

animal until its death. In West Virginia, Stihler et al. (1987) reported

that “the 1 northern flying squirrel for which parasitological data are

available was infected with Strongyloides, but there was no evidence

this was debilitating.” Certainly, these data are limited and more study

is needed on all aspects of the biology of G. sabrinus , including its

interactions with G. volans, if we are to develop guidelines for its

recovery.

ACKNOWLEDGMENTS. — This paper, in part, is a result of a

study on the northern flying squirrel by the Virginia Department of

Game and Inland Fisheries (VDGIF) and the U.S. Forest Service. The

studies of the senior author have also been supported, in part, by funds

from the VDGIF Nongame Wildlife and Endangered Species Program.

The field assistance of numerous VDGIF biologists and game managers,

as well as several U.S. Forest Service personnel on the George

Washington and Thomas Jefferson National Forests, is greatly appre-

ciated. The authors are also grateful to Kenneth B. Knight and Craig

W. Stihler and the numerous individuals who have provided helpful

information or comments on this manuscript, including Walter Bulmer,

Jack A. Cranford, Charles O. Handley, Jr., Judy Jacobs, Karen

Terwilliger, J. Lewis Payne, Donald R. Young, and Peter D. Weigl.

LITERATURE CITED

Davidson, W. R. 1976. Endoparasites of selected populations of gray

squirrels ( Sciurus carolinensis) in the southeastern United States. Proc.

Helminthological Soc. Wash. 43:211-217.

Eckerlin, R. P. 1974. Studies on the life cycle of Strongyloides robustus

Chandler, 1942, and a survey of the helminths of Connecticut sciurids.

Ph.D. thesis, Univ. Connecticut.

Eckerlin, R. P. 1985. Parasites of suburban gray squirrels, Sciurus

carolinensis , in northern Virginia. Va. J. Sci. 36:107.

Fies, M. L., and J. F. Pagels. 1988. Northern flying squirrel investigations.

Pages 75-80 in Virginia Nongame and Endangered Wildlife Investi-

gations, Annual Report 1987-1988, Va. Dept. Game and Inland Fish.

Handley, C. O., Jr. 1953. A new flying squirrel from the southern

Appalachian Mountains. Proc. Biol. Soc. Wash. 66:191-194.

78

Pagels, Eckerlin, Baker, and Fies

Handley, C. O., Jr. 1980. Northern flying squirrel, Glaucomys sabrinus

fuscus Miller. Pages 513-516 in Endangered and Threatened Plants and

Animals of Virginia, D. W. Linzey, editor. Center for Environ. Studies,

Va. Polytech. Inst, and State Univ., Blacksburg.

Parker, J. C. 1968. Parasites of the gray squirrel in Virginia. J. Parasitol.

54:633-634.

Parker, J. C. 1971. Protozoan, helminth and arthropod parasites of the gray

squirrel in southwestern Virginia. Ph.D. dissertation, Virginia Polytechnic

Institute and State Univ.

Payne, J. L., D. R. Young, and J. F. Pagels. 1989. Plant community

characteristics associated with the endangered northern flying squirrel,

Glaucomys sabrinus, in the southern Appalachians. Am. Midi. Nat.

121:285-292.

Price, E. W. 1928. Two new nematode worms from rodents. Proc. U.S. Nat.

Mus., No. 2749. 74(Art.4): 1-5.

Rausch, R., and J. D. Tiner. 1948. Studies on the parasitic helminths of the

North Central States. I. Helminths of Sciuridae. Am. Midi. Nat.

39:728-747.

Stihler, C. W., K. B. Knight, and V. K. Urban. 1987. The northern flying

squirrel in West Virginia. Pages 176-183 in Proc. Third Southeast.

Nongame and Endangered Species Symp., R. R. Odom, K. A.

Riddleberger, and J. C. Ozier, editors. Univ. Georgia, Athens.

Tiner, J. D., and R. Rausch. 1949. Syphacia thompsoni (Nematoda:

Oxyuridae) from the red squirrel. J. Mammal. 30:202-203.

Weigl, P. D. 1977. Glaucomys sabrinus coloratus Handley. Pages 398-400 in

Endangered and Threatened Plants and Animals of North Carolina, J. E.

Cooper, S. S. Robinson, and J. B. Funderburg, editors. N.C. State Mus.

Nat. Hist., Raleigh.

Weigl, P. D. 1978. Resource overlap, interspecific interactions and the

distribution of the flying squirrels, Glaucomys volans and G. sabrinus.

Am. Midi. Nat. 100:83-96.

Weigl, P. D. 1987. Glaucomys sabrinus coloratus Handley. Pages 12-15 in

Endangered, Threatened, and Rare Fauna of North Carolina, Part I. A

Re-evaluation of the Mammals, M. K. Clark, editor. Occas. Pap. N.C. Biol.

Surv., N.C. State Mus. Nat. Sci., Raleigh.

Wells-Gosling, N., and L. R. Heaney. 1984. Glaucomys sabrinus. Mammalian

Species, No. 229: 1-8.

Accepted 28 August 1989

Age Estimates for a Population of American Toads,

Bufo americanus (Salientia: Bufonidae),

in Northern Virginia

Heather J. Kalb1 and George R. Zug

Department of Vertebrate Zoology

National Museum of Natural History

Smithsonian Institution

Washington, D.C. 20560

ABSTRACT. — Age estimation by skeletochronology was made on

adult female and male Bufo americanus from a small population in

northern Virginia. This breeding population consisted predominantly

of males 3-4 years old and females 4-5 years old. Size and age were not

closely correlated, i.e. the larger toads were not necessarily older.

Large males did not appear to have an advantage over the small in

mating; average SVL of amplectic males (64.5 mm) was similar to that

of calling males (64.8 mm). First breeding occurred in late March or

early Arpil and appeared dependent upon water temperatures greater

than 1 1°C in breeding ponds.

An old technique for the estimation of age in amphibians and

reptiles, skeletochronology, has recently gained renewed interest. The

use of bone layers to estimate the ages of frogs or snakes has occurred

irregularly since the early 1900s; however, the technique received

considerable criticism because of the researchers’ inability to demonstrate

that each bone layer represents the same time period. This criticism has

been addressed by several European herpetologists (e.g. Castanet 1985),

who have used known-aged or captive-raised animals to test the annual

deposition of a single bone layer in elements of the appendicular

skeleton.

Hemelaar and van Gelder (1980) studied two populations of the

European toad, Bufo bufo (L.) and demonstrated that a single periosteal

layer was deposited each year. Their experimental demonstration was

simple, but elegant. They marked toads so that each could be recognized

individually and then removed bone samples in successive years. This

procedure showed that an extra layer was present in the second year. In

an attempt to repeat their experiment, Zug initiated a mark and

recapture study of an American toad ( Bufo americanus Holbrook)

population in a suburban park in northern Virginia. Owing to the

transitory nature of the population and an irregular sampling regime, a

Hemelaar/ van Gelder-type study could not be repeated, although we

'Present address: Department of Biology, Texas A&M University, College

Station, TX 77843

Brimleyana 16:79-86, July 1990

79

80

Heather J. Kalb and George R. Zug

were able to estimate the ages of some marked individuals. In this

report, we summarize our skeletochronological analyses and natural

history observations.

MATERIALS AND METHODS

Each spring from 1983 through 1988, the toads breeding in Eakin

Park (Fairfax Co., Va.) have been censused. In 1983-85, the toads were

captured, measured, weighed, and individually marked. The toads were

marked uniquely by removing the distal two phalanges using the 1-2-4-7

coding technique (Ferner 1979). The toe tips were immediately preserved

in 10% buffered formalin and 3-12 months later prepared histologically

(see Zug and Rand 1987). The first digit (1 and 10) of the hand was not

used because injury to the “thumb” could have affected the reproductive

activities of the males. The numbers of individuals marked were: 1983,

19; 1984, 40; 1985, 13. In subsequent years, all toads were counted and

their location and behavior were noted.

Two age estimates can be derived from the histological slides: (1)

the number of periosteal layers observed, complete and incomplete, and

(2) the sum of the number of complete layers observed and the estimated

number of layers partially and completely resorbed. The estimate of

resorbed layers is based on Hemelaar’s (1985: figures 2,3) graphic

technique for the estimation of periosteal growth in B. bufo. Bufo

americanus and B. bufo are equivalent-sized species; therefore, they are

assumed to have similar growth rates. The use of Hemelaar’s technique

and the potential difficulties with the assumptions are discussed in later

sections.

Each toad’s snout-vent length was measured with dial calipers to

the nearest 0.1 mm (though measuring a live, struggling toad is, at best,

accurate to 1.0 mm). Weights were taken on Pesola spring scales to the

nearest 1 g. Air and water temperatures were recorded with a Muller and

Weber quick reading cloacal thermometer to the nearest 0.2° C.

Histological sections were examined independently by the two

authors. Heather Kalb (HK) reviewed all the slides; selected the best

section of the phalanx for each individual; and measured the diameters

of the central marrow cavity, the endosteal ring, and each periosteal

layer (Mark of Skeletal Growth, MSG) as delineated by a LAG (Line of

Arrested Growth), across the dorsoventral axis of the phalanx. Because

all toads were caught in early spring and had not begun a new growth

cycle, the outside diameter of the phalanx represents the last LAG. The

number of measured MSGs equals the total number of incomplete and

complete periosteal layers present in each phalanx. George Zug (GZ)

counted the total number of MSGs on the sections identified by HK,

81

American Toad in Northern Virginia

but without knowledge of HK’s counts and measurements. These data

serve to estimate the age of the individual toads based on the standard

skeletochronological assumption that each MSG (= one growth cycle)

represents one year in the life of a toad.

RESULTS

Age Estimates and Population Structure

The initial goal of the mark and recapture study was to repeat

Hemelaar/van Gelder’s confirmation of the addition of one new MSG

each year. Only one marked toad (a male) was caught in a subsequent

year. This toad showed two MSGs in 1984 and three MSGs in 1985.

Because of the poor recapture rate, marking was discontinued after

1985.

Comparison of MSG counts. Independent counts permit the

evaluation of the precision of data collection and the “legibility” of the

MSGs in this sample. Of the total sample (N = 69), 61 phalangeal

sections could be read by both observers; 46 counts (75%) were identical,

13 differed by ±1 and two by ±2. The differences were examined by the

Wilcoxon matched-pairs signed-rank test (z = 3.26, P - 0.0006). This

nonparametric test evaluates the null hypothesis that the differences

between the two observers’ counts are random. The z value rejects that

hypotheses, because when different, GZ’s counts are less than HK’s

counts with only one exception.

Age estimates. All subsequent discussions of age are based on HK’s

counts and measurements. Examination of the phalangeal histology

suggested that the first year’s periosteal growth layer (MSG) had been

resorbed in many of the Eakin Park toads. To correct our age estimates

for the lost MSGs, we needed to determine the number of MSGs lost

from each toad’s phalanx. Hemelaar’s graphic technique (1985: figures

2,3) offered a mechanism for such a determination. We selected a

maximum diameter of 220 jum for the first LAG/ MSG (= first year’s

periosteal growth), because 12 toads showed minimum LAG diameters

<220 and toads with resorption/endosteal core diameters <220 had

minimum MSG diameters >230 jtim. By establishing a minimum first-

year MSG diameter, the minimum MSG diameter data for each toad

could be assigned to a year class (Fig. 1). The assignment of each toad’s

minimum MSG to a year class locks the larger MSG diameters for that

toad into subsequent year classes. We then summed the diameters for

each age class and calculated MSG diameter means and standard

deviations for each class (Fig. 1). The estimated age for each toad is,

thus, the number of MSG diameters measured plus the estimated

82

Heather J. Kalb and George R. Zug

5

4

1

200 300 400 500

MSG Diameter (jn m)

Fig. 1. The relationship between MSG (marks of skeletal growth) diameters

and age in Bufo americanus from Eakin Park, Fairfax Co., Va.

number of MSGs lost (i.e. 0, 1, or 2 years in this data set). Because a

LAG forms each winter and periosteal growth begins anew each spring,

we accept the hatching of tadpoles as age 0 and each subsequent spring

as a “birthdate”; therefore a toad with two MSGs and captured in

spring 1988 hatched and metamorphosed in 1986.

Population structure. The following interpretations are based on

the total of the number of MSGs seen and the number of MSGs

estimated to have been lost. In 1983, Zug marked 19 adult toads; age

estimates are available for 16 toads. Males ranged from 2 to 4 years and

females from 4 to 5 years (Fig. 2). In 1984, 53 adults were seen, 41

marked, and ages estimated for 31; the age range is the same as for 1983

toads. Fifteen toads were seen and marked in 1985; ages estimated for

14 with a similar age distribution as in 1983-84.

Estimated age and body size. There is no significant correlation

between age and body size (snout-vent length; SVL). In males (N = 46),

the correlation coefficient is low (r = 0.20, P> 0.05), and when all toads

with estimated age (N = 61) are examined, the correlation is marginally

better (r = 0.45, P < 0.01). In contrast, the correlation (r = 0.92) is

significant for the association of weight and length (wt = -103.89 + 2.08

SVL) for all toads captured.

Age and size in amplectic toads. Since all males captured were

calling and the females were moving toward the male choruses or in

American Toad in Northern Virginia

83

JG

rc

3

■o

■

>

■o

c

o

k.

<D

n

E

3

z

5

Age (yr)

Fig. 2. Age (estimated) distributions for the Eakin Park, Fairfax Co., Va.,

population of breeding toads from 1983 through 1985. Males are plotted above

the horizontal line, females (stippled) below the line.

amplexus, we assumed that our sample contained all adults. Male (N =

51) SVL averaged 64.8 mm (51.1-74.3, SD = 4.76), and female (N = 10)

SVL averaged 82.3 mm (71.4-92.6, SD = 7.30). The males in amplexus

(N = 9) had a mean of 64.5 mm SVL (51.5-71.4, SD = 6.51); the females

in amplexus (N = 9) had a mean of 80.6 mm SVL (71.4-92.6, SD = 7.58).

In 1983, the four amplectic males were three (N = 3) and four years old;

the five amplectic males in 1984 were 2 (N = 2), 3 (N = 2), and 4 years

old. No amplexus was observed in 1985. The females were 4-5 years in

1983-84.

Physical Environment and Breeding

Dates of first breeding. First breeding, i.e. presence of amplectic

pairs, was observed on 8 April 1983, 5 April 1984, and 28 March 1987.

On 29 March 1985, males were calling and several females were observed

84

Heather J. Kalb and George R. Zug

entering the breeding pond, but no amplexus occurred while GZ was

present. In 1986, the first major chorus did not occur until 21 April; no

females were observed on that date. No major choruses occurred in

1988, although a few males called for the first time on 3 April; because

no Bufo egg strings were located from April through mid-May, it

appears that breeding did not occur.

Air and water temperatures. The means and ranges of temperatures

observed during breeding were: air 12.8°, 10.0-19.5°C; water 13.2°,

11.0-17.0°C; N = 10. These data include temperatures recorded sub-

sequent to first breeding as well as at first breeding.

Breeding sites. The study area is a suburban park that encompasses

a small permanent stream and its valley, abutted on the north and south

by housing developments. The park includes a mix of hardwood forest,

cattail marsh, and mowed grass tracts. Three aquatic sites are generally

available for toad spawning: (west to east) a temporary woodland pond

(ca. 20 m, maximum length), a cattail marsh (ca. 70 m) with lagoon-like

extensions into adjacent forest, and a permanent man-made pond (ca.

25 m) at the edge of a recreational field. In 1983, an amplectic pair was

found in the flooded grass area adjacent to the woodland pond. In 1984,

amplectic pairs occurred in the marsh and subsequently in the woodland

pond. First and subsequent breeding activities were observed in the

woodland pond in 1985. The 1986 chorus occurred in the permanent

pond. First and second breeding activity in 1987 occurred in the marsh,

third in the permanent pond, and fourth in the pond and marsh. A few

males were calling for the first time on 3 April 1988 in the permanent

pond; during subsequent visits an occasional male was heard in various

localities, but no breeding was observed in 1988.

DISCUSSION

The single recaptured toad with an additional MSG after 1 year is

not confirmation of the “annual deposition of a periosteal layer”

assumption. It does, however, indicate that the Eakin Park B. ameri-

canus produce one MSG each year, and our age estimates are based on

that assumption.

The blind protocol for determining the number of MSGs permits

an evaluation of the legibility of the growth layers and the accuracy of

the observers. The agreement of our observations for 75% of the toads

and a difference of one MSG in 21% of the toads indicate a moderately

high reliability of the counts. The Wilcoxon sign test shows a

conservative bias for the GZ counts, but it does not invalidate the use of

the counts for estimating age. To avoid inconsistency, only the HK data

were used in estimating ages.

American Toad in Northern Virginia

85

We made a number of other assumptions in our skeletochron-

ological analysis, and these assumptions weaken the precision and

reliability of our age estimates. Procedurally, it is desirable to record

MSG counts and measurements from the same region in the same bone.

We used penultimate phalanges in most cases, and those phalanges

derived from the second through fourth digits of the left or right hand.

Using phalanges from different digits and occasionally the ultimate

phalanx should not alter the number of visible MSGs, but it will

increase the variation in MSG diameters. Further, the maximum

diameter of the first MSG was selected in part on the basis of Hemelaar’s

data (1985) on B. bufo, a different though similar-sized species of toad.

Even though the preceding factors may have increased the variation of

our MSG measurements, the range of diameters for each year class (Fig.

1) is comparable to Hemelaar’s (1985: figures 2,3) data. Therefore we

believe that the resulting age estimates are reasonably accurate, and if

used as a unit rather than individually, they provide a reliable picture of

the age structure of this American toad population.

The only other age structure data for B. americanus derives from a

population in central Illinois (Acker et al. 1986), nearly identical in

latitude to Eakin Park. In the Illinois population, males 2-3 years old

and females 3-4 years old were the most numerous age classes. This

pattern does not match that of the Eakin Park population, although the

age range is nearly identical for the two populations. The most striking

difference between the two populations is the absence of 2-year-old

females at Eakin Park and the presence of few 4-year-old and no 5-year-

old males in Illinois. Although not wishing to overinterpret these data,

they do suggest that the average lifespan for male American toads

seldom exceeds 4 years.

Size and age are not correlated in either the Illinois or the Eakin

Park population. Both populations have females significantly larger

than the males, although the Illinois females are smaller (65-84 mm

SVL). The males of the two populations are approximately equal in size

(Illinois, 55-75 mm SVL).

The average size of Eakin Park ampletic males was nearly identical

to the average for the entire male population. Indeed, the smallest male

captured was in amplexus; thus, these data suggest that nonrandom

mating by larger toads is not operating in the Eakin Park population.

Although large male B. americanus have a mating advantage in some

populations (e.g. Gatz 1981), this advantage is not evident or clear-cut

in all populations (e.g. Kruse 1981, Wilbur et al. 1978). Our samples

are, however, too small to evaluate nonrandom mating critically in the

Eakin Park population.

86

Heather J. Kalb and George R. Zug

The Eakin Park breeding population is fairly mobile. It is not tied

tightly to a single site, but has shifted between four major open-water

areas during the 6 years of observations. Perhaps this is typical of

American toad populations; however, we have been unable to find

comparable data in the literature.

Breeding activity in B. americanus was associated with the first

warm rain in late March or early April. The controlling factor seemed

to be water temperature. Males did not call if the water temperature fell

below 11°C, and cold fronts following a spring rain stopped chorus

activity.

ACKNOWLEDGMENTS.— We wish to thank H. Wimer for the

histological preparations and J. Mitchell and C. Pague for reviewing the

manuscript. We also thank Mr. Gilman C. Aldridge and the Fairfax

County Park Authority for permission (permit SC-37) to study the

toads in Eakin Park.

LITERATURE CITED

Acker, P. M., K. C. Kruse, and E. B. Krehbiel. 1986. Aging Bufo

americanus by skeletochronology. J. Herpetol. 20(4):570-574.

Castanet, Jacques. 1985. La squelettochronologie chez les reptiles. I. Resultats

experimentaux sur la signification des marques de croissance squelettiques

chez les lezards et les tortues. Ann. Sci. Nat., Zool., Paris, ser. 13, 7:23-40.

Ferner, J. W. 1979. A Review of Marking Techniques for Amphibians and

Reptiles. Herpetol. Circ., No. 9.

Gatz, A. J., Jr. 1981. Non-random mating by size in American toads, Bufo

americanus. Anim. Behav. 29:1004-1012.

Hemelaar, A. 1985. An improved method to estimate the number of year

rings resorbed in phalanges of Bufo bufo (L.) and its application to

populations from different latitudes and altitudes. Amphibia-Reptilia

6:323-341.

Hemelaar, A. S. M., and J. J. van Gelder. 1980. Annual growth rings in

phalanges of Bufo bufo (Anura, Amphibia) from the Netherlands and their

use for age determination. Netherland J. Zool. 30:129-135.

Kruse, K. C. 1981. Mating succes, fertilization potential, and male body size

in the American toad ( Bufo americanus). Herpetologica 37:228-233.

Wilbur, H. M., D. I. Rubenstein, and L. Fairchild. 1978. Sexual selection

in toads: the roles of female choice and male body size. Evolution

32:264-270.

Zug, G. R., and A. S. Rand. 1987. Estimation of age in nesting female Iguana

iguana: testing skeletochronology in a tropical lizard. Amphibia-Reptilia

8:237-250.

Accepted 30 August 1989

Small Mammals in the Great Dismal Swamp

of Virginia and North Carolina

Robert K. Rose, Roger K. Everton, Jean F. Stankavich,

and John W. Walke

Department of Biological Sciences

Old Dominion University, Norfolk, Virginia 23529

ABSTRACT. — Small mammals were surveyed in a range of habitats

in the Great Dismal Swamp of Virginia and North Carolina. The

survey is based on three chronologically overlapping studies, each

lasting 15-18 months and for which the results have been reported

separately. A different trapping method was used in each of the three

studies: nest boxes, Fitch live traps, or pitfall traps. Only two species

of mammals, both arboreal, were taken in nest boxes, compared with

10 and 9 species in Fitch live traps and pitfall traps, respectively. The

Fitch live traps had a much higher catch rate per 1,000 trap-nights

than either of the other methods. However, pitfall traps were more

efficient at catching Sorex longirostris fisheri and Synaptomys cooperi

helaletes, two mammals that were previously believed to be rare.

Although the catch rates were comparable in nonforested habitats and

in forest, more individuals and more species were obtained in the

former. At least 5 of the 12 collected species do not occur in the

forests. These studies added Sigmodon hispidus to the mammals

known from the Dismal Swamp, and the results suggest that Peromyscus

gossypinus no longer occurs in the swamp.

The Great Dismal Swamp, which lies close to the Atlantic Ocean in

southeastern Virginia and northeastern North Carolina, is a wooded

swampland that is flooded annually from December through March or

April. The soils, which range from sandy through deep peat, are

saturated throughout the winter, but in years of extreme drought, fires

sometimes burn deeply into the organic soils and also destroy large

areas of forest. These physical factors, flooding and fires, and attempts

to control them, have had marked effects on the past and present biota

of the swamp.

The Dismal Swamp encompassed a diversity of habitats before

human attempts to change it. Where the soils burned deeply, bald

cypress trees, Taxodium distichum, often flourished when the normal

hydroperiod returned. Where hot, shallow fires occurred, the regeneration

of dense stands of Atlantic white cedar, Chamaecy paris thyoides , some-

times resulted, and other conditions favored the development of large

stands of cane, Arundinaria gigantea, the only native American bamboo.

Brimleyana 16:87-101, July 1990

87

88

Rose, Everton, Stankavich, and Walke

Slightly elevated “mesic islands” supported oaks and even beeches, trees

that are typical of the upland habitats in the region. Thus, the Dismal

Swamp that developed after the recession of the Wisconsin glacier from

the region 8-10 thousand years ago was a swampland of vegetational

diversity, a diversity maintained by a regime of flood and fire.

The flooding cycle is in part a result of the unusual geology of the

swamp, which includes an escarpment on the western boundary and

underlying impervious clays that prevent the rapid loss of rainfall to an

underground aquifer. Thus, water moves slowly eastward toward the

old duneline that forms the eastern boundary. In the winter months,

reduced evapotranspiration and moderate rainfall combine with a high

water table to inundate many sections of the swamp.

Since the Colonial Period, many land developers have attempted to

exploit the swamp. Even George Washington participated in a scheme

to drain and clear the swamp and convert it to farmland. Invariably

those efforts failed, and always the swamp returned to forested swamp-

land. However, the lowered water table resulting from the construction

of ditches and the suppression of fires in this century has changed the

character of the forest. Today, the Dismal Swamp forests are pre-

dominantly black gum, Nyssa sylvatica , water gum, Nyssa aquatica, and

red maple, Acer rubrum, with scattered patches of bald cypress and

Atlantic white cedar. The formerly extensive areas of cane and evergreen

shrub thicket have been greatly reduced (Musselman et al. 1977). Stands

of cane now are virtually absent except where preserved or maintained

by human activity, such as the 3- to 5-year mowing treatment under a

powerline. As a result, the swamp is moving slowly and inexorably in

the direction of domination by maple and gum trees, and towards less

vegetational diversity. We must assume that this will not favor the

biota, including the mammals, which has become adapted to life in a

physically harsh and biologically variable environment. In an effort to

conserve this distinctive swamp forest, the Union Camp Corporation

donated nearly 19,000 ha of land within the Great Dismal Swamp to the

Nature Conservancy in 1973, which in turn deeded the land to the U.S.

Fish and Wildlife Service (USFWS). In 1974, the USFWS created the

Great Dismal Swamp National Wildlife Refuge, which has grown

through other donations and purchases to its present size of more than

45,000 ha, about three-fourths of which is located in Virginia.

In their efforts to attract buyers, early land developers often greatly

exaggerated the numbers and kinds of wildlife in their descriptions of

the swamp (Handley 1979). The first accounts of what actually was

present were written in a U.S. Geological Survey annual report (Shaler

1890). Shortly afterwards, a major collecting effort was made by the

Small Mammals in Great Dismal Swamp 89

U.S. Department of Agriculture’s Bureau of Biological Surveys, directed

by C. Hart Merriam. Between 1895 and 1898, teams of investigators

studied and collected in the swamp for a total of 23 weeks. As a result,

several new taxa of mammals were described from the swamp, including

a southeastern shrew, Sorex longirostris fisheri Merriam, and a short-

tailed shrew, Blarina brevicauda telmalestes Merriam, each of which is

much larger than its nearby upland subspecies; a Pleistocene relict

population of southern bog lemming, Synaptomys cooperi helaletes

Merriam; and a distinctive muskrat, Ondatra zibethicus macrodon

(Merriam). A meadow vole, Microtus pennsylvanicus nigrans Rhoads,

was described from the North Carolina section of the swamp (Rhoads

and Young 1897). Thus, from the first investigations it was clear that

there were several unusual mammals in the Great Dismal Swamp.

(Although named as distinct species, these mammals have since been

relegated to subspecies status, as shown here.)

The few attempts to study Dismal Swamp mammals in this century

have been summarized by Handley (1979), who had access to the

unpublished data and field notes of government surveys conducted in

the Dismal Swamp. The early studies (1895-1906 period) indicated that

the small-mammal fauna was dominated by forest-dwelling species

[white-footed mouse, Peromyscus leucopus leucopus (Rafinesque); cotton

mouse, Peromyscus gossypinus gossypinus (LeConte); golden mouse,

Ochrotomys nuttalli nuttalli (Harlan); and B. brevicauda], with other

rodents and shrews contributing to a total of 12 species (Handley 1979).

Handley speculated, as others had done, that some of the species may

have disappeared as a result of the changes in the water level and the

vegetation within the swamp.

Bre idling (1980, see also Breidling et al. 1983) trapped briefly on, and

measured the food production of, small plots in four forest types in the

swamp. The only other previous study was conducted in late winter and

spring of 1980, when Rose (1981a) set lines of pitfall traps under a

powerline in the northwestern section of the Dismal Swamp in an effort

to catch S. c. helaletes and S. I fisheri. Within a short time he had

caught as many S. 1. fisheri as had previously been taken in the swamp,

and rediscovered S. c. helaletes, which had not been reported in this

century (Rose 1981b). This short study (Rose 1981a) provided the

preliminary information for a 12-month project funded by the USFWS’s

Office of Endangered Species, which sought to determine the status of

S. /. fisheri and S. c. helaletes and to determine the critical habitats for

these taxa. The grant provided support for the following studies: (1)

Dismal Swamp forest mammals, in which nest boxes were used to

evaluate arboreal small mammals (Walke 1984, Rose and Walke 1988);

90

Rose, Everton, Stankavich, and Walke

(2) the demography of mammals living in an opening and along an

ecotone within the forest, in which live traps were used (Stankavich

1984); and (3) the distribution and habitats of small mammals in the

Dismal Swamp (Everton 1985), in which pitfall traps were used. Together

those studies form the basis for this paper. Our objectives were to

determine the status of the two rare species and to learn more about the

distribution and abundance of the small mammals of the Dismal

Swamp.

MATERIALS AND METHODS

Each study involved 15-18 months of field work, conducted during

the period October 1980 through February 1982, during which time the

region was in a severe drought.

Walke (1984) tested the idea of Breidling et al. (1983) that forest

mammals are present in low numbers because of the poor quality and

unpredictability of the food supply, by the use of four large grids (on

1.96 ha, with 8X8 sites at 20-m intervals), each with the 64 tree-

mounted nest boxes designed to be suitable for use by arboreal P.

leucopus and O. nuttalli. In the two experimental grids, 100 g of mixed

seeds and lab blocks was added to each nest box whenever it was

examined. The two control grids had nest boxes that provided shelter

and hay for building nests, but did not have supplemental food. Nest

boxes were examined at biweekly intervals (later at weekly intervals

when activity levels increased) to catch animals and to evaluate evidence

of their activity (presence of nests, food caches, and scats).

Because her live-trapping study was conducted during a drought,

Stankavich (1984) studied small mammals in what might be considered

ephemeral habitats. Fitch live traps (one per station) were set at 7.6-m

intervals in two rectangular grids (0.38 and 0.40 ha) under a 40-m-wide

110-kv powerline located in the northwest corner of the swamp. These

grids were placed between the pairs of grids of nest boxes in an effort to

monitor the movements of small mammals from one habitat to another.

The two grids differed in amount of flooding and in composition of

vegetation, with one dominated by cane and the other by thick herbaceous

vegetation, primarily Panicum grasses and spikerush, Juncus effusus ;

sections of the latter grid remained flooded even in the drought.

Trapping was conducted for 2 days every 2 weeks from October 1980

through February 1982. (On frequent visits since, the second grid has

been totally flooded, some sections to 1-m depths.)

Everton (1985) used 0.25-ha grids, each consisting of a 5-by-5 plot

with a water-filled no. 10 tin can, sunk so that the lip was flush with the

ground surface, as a pitfall trap at each station. Pitfall traps were

Small Mammals in Great Dismal Swamp

91

Fig. 1. Map showing the location of the live-trap and nest -box study grids (*)

and the 13 pitfall study grids (•) in the Great Dismal Swamp of Virginia and

North Carolina. The boundary encloses the current Great Dismal Swamp

National Wildlife Refuge. The inset at upper left shows the location of the

map area in eastern Virginia and North Carolina.

chosen because of their proven effectiveness in catching shrews, lemmings,

and other species that are difficult to catch with conventional snap or

live traps. Ten grids were placed in a range of nonforested habitats, and

three grids were set in mature forests. Locations of the study grids in the

Dismal Swamp are shown in Fig. 1.

RESULTS

A total of 359 small mammals were taken during the 18 months

encompassed by the three studies (Table 1). Live and pitfall trapping

yielded similar results, both in numbers of individuals (155 and 159) and

in numbers of species (10 and 9). The nest boxes yielded 45 individuals

of two arboreal species. In the three studies collectively, the three most

numerous species were B. brevicauda; the eastern harvest mouse,

Reithrodontomys humulis humulis (Audubon and Bachman); and P. /.

leucopus. Five or fewer specimens were taken of each of the following:

92

Rose, Everton, Stankavich, and Walke

the least shrew, Cryptotis parva parva (Say); the marsh rice rat,

Oryzomys palustris palustris (Harlan); the hispid cotton rat, Sigmodon

hispidus virginianus Gardner; the woodland vole, Microtus pinetorum

scalopsoides (Audubon and Bachman); and the house mouse, Mus

musculus L. Between 26 and 44 individuals each were trapped of S. 1.

fisheri, O. n. nuttalli , M. p. nigrans, and S. c. helaletes. Peromyscus g.

gossypinus , one of four most common small mammals in the early

studies, was absent.

Combining the data from the three studies, 301 mammals of 12

species were taken from 12 nonforested study grids (Table 2), compared

with 58 specimens of four species from seven forested sites. Trapping

efforts in the two habitat types were not comparable; 79.5% of the

64,653 trap-nights were conducted on nonforested grids. Nevertheless,

the catch rates of 5.857 and 4.376 individuals per 1,000 trap-nights were

similar. The nonforested sites included wet grassland dominated by M.

p. nigrans , dry grassland dominated by R. h. humulis , and young pine

plantations and regenerating forest up to 15 years old dominated by S.

1. fisheri , B. brevicauda, and S. c. helaletes. Habitats with young trees

and plentiful grasses frequently yielded the greatest numbers of individuals

and species. The mature forests were mostly red maple and black gum,

which predominate throughout the Dismal Swamp, but one forest site

also had numerous loblolly pines, Pinus taeda, indicating drier conditions

and a slightly higher elevation.

Although live and pitfall trapping yielded comparable numbers of

individuals (Table 1), the capture efficiencies of these methods differed

substantially. Expressed as a catch rate per 1,000 trap-nights, live

trapping was more than three times as efficient (11.962 vs. 3.454) as

pitfall trapping (Table 3). Capture efficiencies were comparable only for

B. brevicauda and M. p. scalopsoides , although the sample size is

exceedingly small for the latter species. The only other species taken by

pitfall trapping at even half of the catch rate of live trapping was S. c.

helaletes. Sorex l. fisheri and C. p. parva were taken only with pitfall

traps, whereas O. p. palustris, S. h. virginianus, and M. musculus (one

or two individuals of each) were live-trapped only. Interestingly, for the

arboreal mice, the catch rate using nest boxes was two or three times

greater than that using pitfall traps (Table 3), and for P. 1. leucopus, the

nest box was slightly less than half as efficient as the live trap (0.45 vs.

1.133 per 1,000 trap-nights).

These studies nearly double the amount of information about small

mammals in the Dismal Swamp (Table 4). The earliest studies (sum-

marized in Handley 1979) focused heavily on forested sites, so the

finding that P. leucopus and O. nuttalli (both arboreal) and the litter-

dwelling B. brevicauda were the most common mammals is not surprising.

93

Small Mammals in Great Dismal Swamp

Table 1. Numbers of individual small mammals taken in a range of habitats

using three different trapping methods during concurrent studies in the Great

Dismal Swamp of Virginia and North Carolina.

Live Pitfall Nest-box

aPercent refers to the proportion of that species to the total individuals (359)

taken in the study.

94

Rose, Everton, Stankavich, and Walke

Table 2. Number and percent (of individuals within a species) of small

mammals taken in 12 nonforested and 7 forested study grids in the Dismal

Swamp.

Nonforest habitat Forest habitat

(51,399)a (13,254)a

aNumbers in parentheses are the total number of trap-nights in that habitat.

The studies of Handley (8 days in 1953), Breidling (on four forest plots

for 1 week in each of three seasons in 1979 and 1980), and Rose (two

study sites over 2 months in 1980) were brief or restricted to a few sites.

By contrast, the current studies lasted 15-18 months each, and together

evaluated the mammals on 19 study grids. Our studies recorded one new

species for the swamp, S. h. virginianus . Two individuals were recaptured

Small Mammals in Great Dismal Swamp

95

Table 3. Comparison of trapping efficiencies for three methods of trapping

small mammals, expressed as the number of new individuals taken per 1,000

trap-nights.

several times over a 2-month period on the driest live-trap grid. For the

first time in this century, O. p. palustris (one specimen in live trap) and

M. p. scalopsoides (four specimens in live and pitfall traps) were

collected. No specimens of P. g. gossypinus were collected in these

studies, and only two have been collected in this century [in 1933 by

Dice (1940)]. We can conclude that its numbers and distribution have

declined, and perhaps it is now absent from the swamp forests. The

largest apparent increases in numbers were for the shrews, because

pitfall traps were used, and R. h. humulis, most of which were taken in

live traps.

96

Rose, Everton, Stankavich, and Walke

Table 4. A comparison of the results of small mammal studies conducted in the

Dismal Swamp, based on Handley (1979), recent studies [Handley’s 1953 in

Handley (1979), Breidling et al. 1983, Rose 1981a], and the present studies.3

Percent of

o •

Dice (1940) caught four Peromyscus leucopus and two P. gossypinus near

Lake Drummond in 1933.

DISCUSSION

The results of our three studies substantially advance our under-

standing of the distribution and abundance of Dismal Swamp mammals.

One species, S. h. virginianus, was recorded in the swamp for the first

time, and P. g. gossypinus probably is now absent. Thus, the total

remains at 12 species of small mammals, as in the 1895-1906 period

(Table 4). However, we now have information about mammals in

nonforested habitats as well as large sample sizes for several species.

Small Mammals in Great Dismal Swamp

97

The arboreal P. 1. leucopus and O. n. nuttalli are common today, as

in the past, and so is B. brevicauda (Table 4). The species showing the

largest numerical increases in collections conducted during this decade,

including the results of Rose (1981a), were S. 1. fisheri, S. c. helaletes,

and R. h. humulis (Table 4). That substantially larger numbers were

recorded has two causes: the use of different trapping methods and the

greater sampling effort in nonforested habitats (Table 2). Early studies

relied heavily on snap or break-back traps. In our research, all 44 S. I

fisheri were taken with pitfall traps (Table 1), an expected result because

this shrew is rarely collected by any other means (Rose 1980), and most

R. h. humulis (93%) were taken in live traps. These two methods yielded

all S. c. helaletes (Table 1).

The Dismal Swamp southern bog lemming, S. c. helaletes , a

distinctive relict subspecies, remains enigmatic as a study subject. We

noted the cuttings and green dropping of this species at the start of the

study on one live-trap grid, but we did not catch any S. c. helaletes until

the tenth month of trapping, after which we caught 1 1 in the span of a

few weeks on that grid. Pitfall trapping yielded S. c. helaletes from

nearly half of the nonforested grids, and we determined that it sometimes

was common. The same can be said of the Dismal Swamp southeastern

shrew, S. 1. fisheri ; it was found on more than half of the pitfall grids

and it, too, was locally abundant, especially in habitats in early succession.

Thus, we determined that these two supposedly rare species, whose

status was a particular objective of the pitfall trapping, were widespread

and sometimes common. However, because the upland subspecies of the

southeastern shrew, Sorex longirostris longirostris Bachman, is found

nearby, S. /. fisheri has been listed by the USFWS as threatened (FR

51,287: 26 September 1986). That decision was made because the drying

conditions created by ditching and draining may favor the movement of

the S. I longirostris into the Dismal Swamp, thereby potentially resulting

in interbreeding and perhaps genetic swamping of the restricted and less

common S. 1. fisheri. On the other hand, S. c. helaletes has never been

Federally listed, because it is widespread (1,000-km2 area), colonizes

early-successional stages and persists there until the forest matures, is

locally abundant, and is isolated by 300 km from the nearest conspecific

subspecies. Thus, although S. c. helaletes was believed by some investiga-

tors to be extinct, it apparently is thriving.

The second reason that we were able to collect these three species in

numbers indicating that they are common is that the live- and pitfall-

trapping studies focused on nonforested habitats (Table 2). Overall, 91%

of S. /. fisheri and 100% of R. h. humulis and of S. c. helaletes were

taken from nonforested habitats. These habitats ranged from fields with

purely herbaceous vegetation to natural or planted stands of trees up to

98

Rose, Everton, Stankavich, and Walke

15 years old. As long as grasses remained in the understory, S. c.

helaletes persisted. Sorex /. fisheri persisted even in mature forests with

no grasses, but at lower densities than in early serai stages. Reithro-

dontomys h. humulis were restricted to early serai stages, i.e. those with

few saplings or shrubs, where they attained densities as great as 25/ ha

on live-trapping grids (Stankavich 1984). Trapping in areas dominated

by herbaceous vegetation no doubt contributed to the relatively large

number of M. p. nigrans compared with previous studies (Table 4).

Clark et al. (1985), working in and near Carolina Bays and pocosins in

North Carolina, also reported 3-5 times higher capture success when

trapping on edges or in fields compared with the interior of pocosins.

Five species (C. parva, O. palustris, S. hispidus, M. pinetorum, and

M. musculus ), each represented by one to five specimens, were found

only in nonforested habitats (Table 2). Except for S. hispidus , all had

been collected in the past, usually in low numbers, and should be

considered as minor species in the Dismal Swamp. For example, C.

parva is most abundant in the region in dry oldfield habitats (Rose

1983, Everton 1985), habitats that are absent in the Dismal Swamp.

Although little is known of the ecology of O. palustris , it is highly

aquatic and therefore well adapted to live in swamps. The decline in

numbers of O. palustris (Table 4) may be more apparent than real, or it

could indicate a loss of habitat. Sigmodon hispidus , first reported from

Virginia (Mecklenburg Co. in 1940) by Patton (1941), has been expanding

its range throughout the Midwest and East. In Virginia, it has crossed

the James River near Richmond (Pagels 1977), but its northward path is

blocked in eastern Virginia by the Chesapeake Bay. As a species that is

well adapted to the dry grassland of the Southwest, S. hispidus probably

is poorly adapted to conditions of long-term inundation of its habitat,

particularly if winters are relatively cold. Furthermore, the species is

found primarily in habitat dominated by grasses and other herbaceous

vegetation, and it seems not to tolerate much woody vegetation in its

habitat. Although patches of suitable habitat may be produced by fires

or clearcutting, that habitat will probably occur in remote sections of

the swamp, where it is separated from the closest source populations of

S. hispidus by large expanses of unsuitable cover.

The woodland vole, M. pinetorum , also called the pine vole,

usually is associated with the edge of forest and oldfield. Although M.

pinetorum sometimes is common in well-drained upland forests in the

region, it apparently is not common in the seasonally flooded forests of

the Dismal Swamp. Finally, M. musculus , introduced to North America

from Europe during colonial times, usually is a commensal of man or is

restricted to disturbed areas such as recently plowed fields, croplands,

Small Mammals in Great Dismal Swamp 99

or the earliest successional stages. In general, M. musculus does not

coexist with native mammals once the latter become well established.

Because there are no buildings or croplands, there is today relatively

little disturbed habitat in the Dismal Swamp, except for that resulting

from an occasional fire or blowdown. Hence, there is little opportunity

for M. musculus to flourish. Except for S. hispidus, which was not

found prior to these studies, the five species that we found in lowest

number (1-5) also were present, but rare to uncommon, in the early

studies of the Dismal Swamp (Table 4).

The numbers of different individuals taken by the three methods

differed substantially in these studies (Table 3). Higher catch rates for

almost all species were obtained in live traps compared with pitfall

traps. The exception was B. brevicauda, for which the rates were

comparable. One thing we learned in the pitfall-trapping study was that

most of the animals were taken in the first 2-4 weeks. Catch rates

dropped off sharply thereafter. On grids established midway in the

study, we placed plastic snap-on lids on the pitfall traps after a month

of trapping, and weeks later reopened them. These grids had higher

catch rates, i.e. yielded more animals over fewer weeks of trapping. Had

we used this technique throughout the study, the catch rates for pitfall

trapping would have been substantially higher.

Besides yielding moderate catch rates, live traps are also useful

because individuals can be trapped repeatedly and marked to obtain

information on growth, reproduction, and density. The primary advantage

of pitfall traps is that some species, particularly S. longirostris, rarely

are taken by any other means. An additional advantage is that, unlike

live traps, pitfall traps can be checked at irregular intervals, e.g. weekly

or biweekly, which permits a large amount of information to be obtained

in relation to the time spent tending the traps. Especially for locations

deep inside the Dismal Swamp, pitfall traps are useful even though the

catch rate is lower than for live traps (and based on Wiener and Smith

1972, much lower than it would be for snap traps).

The relatively high number of O. nuttalli (22, Table 1) and the

catch rate for this form in the live-trap grids were surprising, particularly

because other studies in the swamp have shown it to be less common

than Peromyscus . We believe our success resulted from the habitat

sampled, because the powerline right of way provided a large amount of

ecotone, which seems to be ideal for O. nuttalli (Layne 1958). All but

one of the O. nuttalli in the nest-box study also were taken at the

ecotone. These results reinforce Dueser and Shugart’s (1978) suggestion

that O. nuttalli is a habitat specialist and requires the complex vegetational

structure provided along the edge of a forest. In the forest proper,

100

Rose, Everton, Stankavich, and Walke

however, P. 1. leucopus remained most common, as seen in the nest-box

study (Table 1). In the live-trap study, most of the P. /. leucopus also

were trapped at the edges of the grids, i.e. in the ecotone.

In conclusion, these studies showed the supposedly rare Dismal

Swamp subspecies of S. /. fisheri and S. c. helaletes to be widespread

and locally abundant. However, S. I fisheri is affected by interbreeding

with a nearby upland race and now is listed as threatened by the

USFWS. Our studies nearly double the amount of information for

small mammals in the Dismal Swamp, documenting one additional

species {S. hispidus ) and one probable loss ( P . gossypinus) in this

century. The slightly higher catch rate (= abundance) and greater numbers

of species from nonforested habitats suggest that any management plan

that creates clearings or other vegetational heterogeneity will promote

the diversity and abundance of small mammals in the Dismal Swamp.

Fortunately, the management plan recently developed for use in the

Great Dismal Swamp National Wildlife Refuge calls for the implementa-

tion of such management measures.

ACKNOWLEDGMENTS. — We acknowledge the support of these

studies by a grant (No. 81-602) to Old Dominion University (R. Rose,

Principal Investigator) from the USFWS Office of Endangered Species,

and the cooperation of the Great Dismal Swamp National Wildlife

Refuge office, particularly M. Keith Garrett, Wildlife Biologist. We also

thank Sharon Everton, Dave Harrelson, Dave Stankavich, Karen

Terwilliger, Janet Walke, and Tom Wilcox for assistance.

LITERATURE CITED

Breidling, F. E. 1980. An evaluation of small rodent populations in four

Dismal Swamp plant communities. Master’s thesis, Old Dominion Univ.,

Norfolk, Va.

Breidling, F. E., F. P. Day, Jr., and R. K. Rose. 1983. An evaluation of small

rodents in four Dismal Swamp plant communities. Va. J. Sci. 34:14-28.

Clark, M. K., D. S. Lee, and J. B. Funderburg, Jr. 1985. The mammal fauna

of Carolina bays and pocosins, and associated communities in North

Carolina: an overview. Brimleyana 11:1-38.

Dice, L. R. 1940. Relationships between the wood-mouse and the cotton-

mouse in eastern Virginia. J. Mammal. 21:14-23.

Dueser, R. D., and H. H. Shugart, Jr. 1978. Microhabitats in a forest-floor

small mammal fauna. Ecology 59:89-98.

Everton, R. K. 1985. The relationship between habitat structure and small

mammal communities in southeastern Virginia and northeastern North Carolina.

Master’s thesis, Old Dominion Univ., Norfolk, Va.

Small Mammals in Great Dismal Swamp

101

Handley, C. O., Jr. 1979. Mammals of the Dismal Swamp: a historical

account. Pages 297-357 in The Great Dismal Swamp (P. W. Kirk, Jr.,

editor). Univ. Press Virginia, Charlottesville.

Layne, J. N. 1958. Notes on mammals of southern Illinois. Am. Midi. Nat.

60:219-254.

Musselman, L. J., D. L. Nickrent, and G. F. Levy. 1977. A contribution

towards a vascular flora of the Great Dismal Swamp. Rhodora 79:240-268.

Pagels, J. F. 1977. Distribution and habitat of cotton rat ( Sigmodon hispidus)

in central Virgina. Va. J. Sci. 28:133-135.

Patton, C. P. 1941. The eastern cotton rat in Virginia. J. Mammal. 22:91.

Rhoads, S. N., and R. T. Young. 1897. Notes on a collection of small

mammals from northeastern North Carolina. Proc. Acad. Nat. Sci., Philadelphia

(1897):303-312.

Rose, R. K. 1980. The southeastern shrew, Sorex longirostris, in southern

Indiana. J. Mammal. 61:162-164.

Rose, R. K. 1981a. Small mammals in openings in Virginia’s Dismal Swamp.

Brimleyana 6:45-50.

Rose, R. K. 1981b. Synaptomys not extinct in the Dismal Swamp. J.

Mammal. 60:844-845.

Rose, R. K. 1983. A study of two rare mammals endemic to the Virginia/

North Carolina Dismal Swamp. Final report, USFWS, Contract No.

14-16-000581-033.

Rose, R. K., and J. W. Walke. 1988. Seasonal use of nest boxes by

Peromyscus and Ochrotomys in the Dismal Swamp of Virginia. Am. Midi.

Nat. 120:258-267.

Shaler, N. S. 1890. General account of the fresh-water morasses of the United

States, with a description of the Dismal Swamp District of Virginia and

North Carolina. U.S. Geol. Survey, Ann. Rep. 1 0( 1):26 1-339. [Original not

seen; from Handley (1979).]

Stankavich, J. F. 1984. Demographic analysis and microhabitat relationships

of a small mammal community in clearings of the Great Dismal Swamp.

Master’s thesis. Old Dominion Univ., Norfolk, Va.

Walke, J. W. 1984. Activity levels of arboreal rodents Peromyscus and

Ochrotomys evaluated with nest cans in seasonally flooded forests. Master’s

thesis, Old Dominion Univ., Norfolk, Va.

Wiener, J. G., and M. H. Smith. 1972. Relative efficiences of four small

mammal traps. J. Mammal. 53:868-873.

Accepted 7 September 1989

102

THE SEASIDE SPARROW,

ITS BIOLOGY AND MANAGEMENT

Edited by

Thomas L. Quay, John B. Funderburg, Jr., David S. Lee,

Eloise F. Potter, and Chandler S. Robbins

The proceedings of a symposium held at Raleigh, North Carolina,

in October 1981, this book presents the keynote address of F. Eugene

Hester, Deputy Director of the U.S. Fish and Wildlife Service, a

bibliography of publications on the Seaside Sparrow, and 16 major

papers on the species. Authors include Arthur W. Cooper, Oliver L.

Austin, Jr., Herbert W. Kale, II, William Post, Harold W. Werner,

Glen E. Woolfenden, Mary Victoria McDonald, Jon S. Greenlaw,

Michael F. Delany, James A. Mosher, Thomas L. Merriam, James A.

Kushlan, Oron L. Bass, Jr., Dale L. Taylor, Thomas A. Webber, and

George F. Gee. A full-color frontispiece by John Henry Dick illustrates

the nine races of the Seaside Sparrow, and a recording prepared by J.

W. Hardy supplements two papers on vocalizations.

“The Seaside Sparrow, with its extensive but exceedingly narrow

breeding range in the coastal salt marshes, is a fascinating species. All

the authors emphasize that the salt marsh habitat is at peril. . . . The

collection is well worth reading.” — George A. Hall, Wilson Bulletin.

1983 174 pages Softbound ISBN 0-91 7134-05-2

Price: $15, postpaid, North Carolina residents add 5% sales tax. Please make

checks payable in U.S. currency to NCDA Museum Extension Fund.

Send to SEASIDE SPARROW, N.C. State Museum of Natural Sciences,

P.O. Box 27647, Raleigh, NC 27611.

Recent Changes in the Freshwater Molluscan Fauna

of the Greenfield Lake Basin,

North Carolina

William F. Adams

Environmental Resources Branch

United States Army Corps of Engineers

P.O. Box 1890, Wilmington, North Carolina 28402

ABSTRACT. — The molluscan fauna of the Greenfield Lake basin has

undergone significant changes in recent years. In surveys conducted

from January 1984 through October 1987, 16 species of mollusks were

found in the basin. Twelve of those species were recorded from the

basin for the first time; 15 species previously recorded in the basin

were not found. The most noticeable change was the almost total

elimination of the Unionidae. Changes in the molluscan fauna of the

Greenfield Lake basin have probably been caused by a combination of

factors and events. Those agents of change are still at work today and

undoubtedly keep the molluscan population of the lake from ever

establishing any type of equilibrium. Partial winter drawdowns are

believed to be the most damaging aspect of the water management

presently undertaken in the watershed. Pollution from non-point

sources has probably also played a role in changing the fauna.

Mollusks in Greenfield Lake, in Wilmington, New Hanover Co.,

N.C., were frequently sampled by malacologists during the early part of

the twentieth century. Records of those collections are scattered in the

literature, but a comprehensive survey of the fauna has never been

undertaken.

Because of its prominence in the literature of freshwater malacology

on the south Atlantic slope, any changes in the molluscan fauna of

Greenfield Lake are of general interest. Because of the available data on

prior species occurrence, a comprehensive survey was undertaken to

assess what changes have taken place in the composition of the lake’s

molluscan fauna.

Many of the taxa cited in this paper are in need of systematic

revision. Resolving taxonomic problems was not a purpose of this

study, but the subject has to be addressed because of the abundance of

synonyms for some species and the revisions that have taken place since

the time of the first Greenfield Lake records. A synonymy section has

therefore been included to deal with this problem to the level necessary

Brimleyana 16:103-117, July 1990

103

104

William F. Adams

for identification of species. Gender problems in the species names used

by previous authors have not been rectified.

HISTORY OF GREENFIELD LAKE

Greenfield Lake was created prior to 1750 by the impoundment of

a small low-lying stream located south of Wilmington to provide a

source of fresh water for a nearby rice plantation and for the operation

of grist and saw mills. The rice plantation, lake, and surrounding lands

were known collectively as “Greenfields” (Moore 1968). The surface

elevation is 1.85 m above mean sea level, the surface area is 75 ha, and

the drainage basin is approximately 1,100 ha. Maximum depth is

approximately 2.5 m.

The City of Wilmington purchased the lake in 1925 for use as a

public park, and a circumferential road approximately 8 km long was

constructed as a public works project during the Great Depression

(Moore 1968). In 1935, fallen trees were removed from the lake and the

surrounding area was landscaped (Moore 1975). Removal of marsh

grasses and additional snags took place in 1945 for mosquito control

(Appleberry 1945).

Today, Greenfield Lake is well within Wilmington, and its watershed

is almost completely developed. Land use consists primarily of residential

and commercial areas with open space in the form of a golf course and

small, isolated tracts of woodlands. Lands immediately adjacent to the

lake are still used as a public park, which is a major recreation center

for Wilmington.

Urbanization of the watershed has resulted in a decreased detention

time for storm waters and, consequently, an increased introduction of

nutrients, pesticides, and metals into Greefield Lake. Breaks in sewer

lines on the bottom of the lake and sewage overflow from nearby

manholes during storms have also added large pulses of nutrients. The

latter are now a major problem and in recent years have contributed to

severe algal blooms ( Lyngbya ). Control has been attempted through

algacides (chelated copper and dichlobenil); introductions of the exotic

fishes tilapia, Tilapia aurea (Steindachner), and grass carp, Ctenopharyngo-

don idella Valenciennes; and partial drawdowns during the winter

months to desiccate and freeze the algae. Central portions of the lake

have been dredged in attempts to remove nutrient-laden sediments and

deepen the lake beyond the photic zone.

All of the tributaries leading into Greenfield Lake have been

channelized to improve drainage in the basin. Maintenance of the

channels is periodic and is confined to portions upstream of Lakeshore

Drive, the perimeter road. The outlet creek below the dam has been

Freshwater Molluscan Fauna

105

channelized to its confluence with the Cape Fear River and has changed

from a freshwater system to an intermittently brackish one.

HISTORICAL COLLECTIONS

Documentation of the historical occurrence of species in the lake

was taken from published literature. Records from unpublished collections

in museums and universities were not sought. In summarizing her

private collection of North Carolina freshwater mollusks plus those in

the Academy of Natural Sciences, Philadelphia, and in the Museum of

Zoology of the University of Michigan, Dawley (1965) provided several

Greenfield Lake citations of both bivalves and gastropods. However,

this work was not exhaustive, as major mollusk collections, such as

those at Harvard University and the National Museum of Natural

History, were not searched. Early in this century, Bartsch searched for

Planorbis magnificus (Pilsbry, 1903), which had been first described

from “the lower Cape Fear River.” He found that snail to be fairly

abundant in the lake; discovered and described another new planorbid,

Planorbis eucosmius (Bartsch, 1908); and noted but did not describe

two other apparently new mollusks. Bartsch’s (1908) discussion of the

habitat of P. magnificus provides the only available description of the

aquatic macrophyte community in Greenfield Lake early in this century.

Rehder (1949) reported Campeloma rufum (Haldeman, 1841) from the

lake in his discussion of land and freshwater snails collected during his

travels through the region. Walter (1954) reported Pseudo sue cine a

columella (Say, 1817) from the lake in his discussion of the range of the

species but did not disclose the source of the record. Porter (1985)

reported forms of Campeloma believed to be C. geniculum (Conrad,

1834) from Greenfield Lake and mentioned other species of gastropods

reported by Dawley (1965).

Records of Bivalvia from Greenfield Lake deal principally with the

Unionidae and were summarized by Johnson (1970), who recorded eight

species from the lake or the creek below the spillway: Lampsilis radiata

(Gmelin, 1791), Villosa delumbis (Conrad, 1834), Villosa vibex (Conrad,

1834), Anodonta couperiana (Lea, 1842), Anodonta imbecillis (Say,

1829), Uniomerus tetralasmus (Say, 1831), Elliptio complanata (Lightfoot,

1786), and Elliptio lanceolata (Lea, 1828). Elliptio fisherianus (Lea,

1838) was recorded from the lake by Bailey (1940), and Morrison (1972)

mentioned that Anodonta imbecillis, A. couperiana, and A. teres (Conrad,

1834) occur there. Porter (1985) reported Anodonta cataracta (Say,

1817) and, in citing correspondence with J. P. E. Morrison, also

reported E. fisherianus, A. teres, Villosa vaughaniana (Lea, 1838), and

Lampsilis ochracea (Say, 1817) from the lake. Heard (1963, 1965)

106

William F. Adams

reported Greenfield Lake specimens of Musculium transversum (Say,

1829) and Eupera cubensis (Prime, 1865). Walter (1954) also reported E.

cubensis from the lake in his discussion of the range of the species but

did not disclose the source of the record.

All published records from the Greenfield Lake system are summarized

in Table 1. Names are presented as published without regard to subsequent

taxonomic revisions or the possibilities of misidentifications, which are

discussed later.

SYNONYMY

Most of the collections cited in Table 1 are old, and the names have

undergone substantial taxonomic revision. The following brief discussion

of synonymy brings these historical records into a modern taxonomic

framework. Species names are still considered valid if they appear in

Table 1 and are not discussed below.

Gastropoda

Clench (1962) synonomized the three forms of Campeloma rufum

(Haldeman, 1841) with three separate species. Campeloma rufum was

synonymized with C. crassulum Rafinesque, 1819, a species of the Great

Lakes-St. Lawrence and the Mississippi drainages. Campeloma r.

meridionale (Pilsbry, 1916) was synonomized with C. limum (Anthony,

1860), and Campeloma r. geniculiforme (Pilsbry, 1916) was synonymized

with C. geniculum. Which form Rehder (1949) collected in Greenfield

Lake is unknown, and the ranges of C. geniculum and C. limum

reported by Burch (1982) make either species a possibility.

The genus Gillia Stimpson, 1865, has not been revised. Gillia

crenata (Haldeman, 1840) has not been synonymized with any other

forms; however, modern taxonomic keys make no reference to this

species. Specimens of G. crenata in the National Museum of Natural

History are being treated as G. altilus (Lea, 1841) (A. G. Gerberich,

personal communication).

Although its taxonomic status is still in doubt, Planorbis eucosmius

(Bartsch, 1908) has been transferred to the genus Helisoma Swainson,

1840 (Burch 1982). Burch (1982) speculated that P. eucosmius may

simply be a juvenile form of //. anceps anceps (Menke, 1830), whereas

Fuller (1977) assigned P. eucosmius to the South American genus

Taphius H. & A. Adams, 1855. Proper taxonomic placement will not be

possible until additional specimens are acquired for soft-tissue analysis.

Planorbis magnificus (Pilsbry, 1903) was transferred to Planorbella

Haldeman, 1842, by Baker (1945) and the species epithet emended to

the feminine magnifica. Planorbis Muller, 1774, refers to Palearctic and

Ethiopian forms (Burch 1982).

Freshwater Molluscan Fauna

107

Table 1. Published records of freshwater mollusks from the Greenfield Lake

drainage basin.

108

William F. Adams

The genus Liogyrus Gill, 1863, is now considered a subgenus of

Amnicola Gould, 1841 (Burch 1982).

Bivalvia

Johnson (1970) synonymized Elliptio fisheriana (Lea, 1838) with E.

lanceolata (Lea, 1828). Davis (1984) determined that E. fisheriana was

distinct from E. lanceolata and that E. folliculata (Lea, 1838) and E.

producta (Conrad, 1836), which had also been synonymized with E.

lanceolata by Johnson (1970), were also distinct from it and from each

other. Johnson (1984) synonymized E. producta with E. angustata (Lea,

1831). The ranges of these species are uncertain, and which species

occurred historically in Greenfield Lake is unknown.

Uniomerus tetralasmus (Say, 1831) is still a valid species, but south

Atlantic drainage Uniomerus are now considered to be U. obesus (Lea,

183 1) (Johnson 1984).

Johnson (1970) synonymized Anodonta teres with A. cataracta and

Villosa vaughaniana with V. delumbus. However, Anodonta teres and

Villosa vaughaniana are still regarded as distinct by some researchers

(e.g. Porter 1985, Turgeon et al. 1988).

The genus Limosina Clession, 1872, is synonymous with Eupera

Bourguignat, 1877 (Heard 1965). Eupera cubensis (Prime, 1865) and

Musculium transversum (Say, 1829) are still considered to be valid

names.

METHODS

All of the major tributaries, the nearshore lake bottom out to a

distance of about 10 m, and the creek downstream of the dam were

sampled between January 1984 and October 1987. Eighteen stations

were investigated and are shown in Fig. 1. During January 1984 and

again in January 1986, the lake was drained for several weeks, and the

substrate for 6-30 m from the bank was exposed. This permitted

thorough searches, although some mollusks may have retreated with the

receding water. Because movements of unionids during such events have

been shown to be random (Samad and Stanley 1986), the species

obtained are considered to be representative. The central portions of the

lake could not be sampled, as soft sediments and deep water made

collecting by the methods used impossible. Summer collections of

gastropods were made by sweeping a fine-mesh net through floating

aquatic macrophytes, by raking mats of submerged aquatic macrophytes,

and by hand. The creek below the dam was sampled by hand and by

raking during low tides in summer 1985 and during the winter drawdown.

Approximately 40 man-hours were spent in collecting.

Freshwater Molluscan Fauna

109

Fig. 1. Location of collection stations in the Greenfield Lake drainage basin.

110

William F. Adams

Salinity measurements in the creek below the dam were taken

during and after the 1986 drawdown. All measurements were taken

from surface waters and determined by using a refractometer.

Identifications of Unionidae were made using Johnson (1970).

Sphaeriidae were determined from Herrington (1962) and Burch (1975),

and the nomenclature follows Burch (1975). Gastropods, excepting the

Ancylidae, were identified using Burch (1982), and his nomenclature is

used. Identification and nomenclature of the Ancylidae follow Basch

(1963).

RESULTS

Sixteen species of mollusks, 10 gastropods and 6 bivalves, were

collected from Greenfield Lake, its tributaries, and the creek below the

spillway (Table 2). Twelve species were recorded from the lake or the

drainage basin for the first time, and 15 species previously reported were

not relocated.

Campeloma decisum occurred in only one tributary. Planorbella

trivolvis and Helisoma anceps were relatively common throughout the

lake in the nearshore area. Pseudosuccinea columella was found in quiet

waters of the finger portions of the lake but was absent from the main

body of the lake. Because of the fragility of the shell of P. columella ,

occasional waves may render the main body of the lake unsuitable

habitat for that species. Menetus dilatatus and Gyraulus deflectus were

common in the upper end of the lake. Gyraulus parvus was found only

in the golf course ponds of Station 16. Ferrissia fragilis was common

on leaf litter and trash throughout the lake.

Anodonta is the only unionid genus now occurring in the basin.

Anodonta species are normally associated with lentic habitats, and all

live specimens were obtained from the upper portions of the lake.

Anodonta cataracta may be common in the central portions of the lake;

numerous valves of this species were found at the interior base of the

dam during the drawdown period. Three forms of A. cataracta occur in

the lake: a form with dorsal and ventral margins roughly parallel, a

form with a broadly rounded ventral margin, and a form with a concave

ventral margin.

Only one valve of Andonta imbecillus was found during the present

study, in the fall of 1987 on a spoil pile resulting from rechannelization

of the tributary at Station 13. Despite extensive searches of other spoil

piles, no additional specimens could be located, and none were found

during drawdowns. Because this tributary was at least temporarily

disrupted, the status of A. imbecillus cannot be determined. It is at best

extremely rare, and perhaps it is extirpated from the Greenfield Lake

basin.

Freshwater Molluscan Fauna

111

Table 2. Species of freshwater mollusks collected from Greenfield Lake with

collection stations.

Sphaeriid clams were numerically the most abundant bivalve mollusks

in the lake system. High densities of Sphaerium occidentale and Musculium

transversum were discovered under algal mats in nearshore areas during

the January 1984 drawdown, but numbers were much reduced in the

summer of 1987.

Only Rangia cuneata (Gray, in Sowerby 1831) and Polymesoda

caroliniana (Bose, 1802), both brackish-water species, occur in the creek

below the spillway. Salinity appears to be the limiting factor for

freshwater species in this area as tides bring brackish water from the

Cape Fear River into the creek. Surface salinity measurements taken

112

William F. Adams

during February 1986 indicate a range of 0 parts per thousand (ppt) to 5

ppt when the lake was being reimpounded after a drawdown and there

was no water being released from the lake. Concentrations up to 3 ppt

were observed during normal summer releases from the lake. Hopkins

et al. (1973) note that R. cuneata is restricted to areas where salinity is

below 15 ppt most of the time and may occupy portions of creeks and

tidal rivers where salinities are continuously below 1 ppt for extended

periods.

DISCUSSION

Distributions of many species in the lake appear to be spotty based

on the collection station records (Table 2). Differences between stations

may be caused by microhabitat conditions, by disturbance histories, or

by differing efficiency of collecting at each station. Species of mollusks

not recorded in this survey may yet be found to exist in the lake.

Misidentifications in Historical Records

The record of Laevape x diaphanus by Dawley (1965) may be based

on a misidentification of L. fuscus\ she did not mention the source of

the specimen or the determination. Laevape x diaphanus typically inhabits

rock bottoms in slowly flowing waters, and I have collected it on debris

in swamp streams near Wilmington. However, L. fuscus is fairly common

in southeastern North Carolina, and it would be more likely than L.

diaphanus to have occurred in the lacustrine habitats provided by the

lake.

Campeloma geniculum may also have been misidentified, because it

is difficult to separate from C. decisum. In discussing current taxonomy

of the genus Campeloma , Clench (1962) correctly remarked that “few

genera among our North American freshwater mollusks are in a more

confused state.” Bailey (1940) states that Pilsbry examined specimens of

C. geniculum from the lake and referred to them as “a rather unusually

rounded form of the species.” That would imply that the specimens may

have been C. decisum, which is separated from C. geniculum by its

more rounded shoulders (Burch 1982). Rehder (1949) collected C.

rufum (which would now be synonymized with either C. geniculum or

C. limum , see Synonymy above) from Lake Waccamaw, approximately

50 km W of the study area, as well as from Greenfield Lake. Campeloma

that had characteristics of both C. geniculum and C. decisum were

noted by Porter (1985) in his collections from Lake Waccamaw, and

one specimen that had characteristics of C. geniculum was found during

the present survey. Campeloma geniculum is generally expected to have

a more southern range, but a population is certainly possible in south-

eastern North Carolina.

Freshwater Molluscan Fauna

113

Changes in the Molluscan Fauna

All of the species recorded during this survey are common natives

and were probably present in the lake in prior years but were simply not

mentioned in previously published accounts.

Neither Planorbella magnifica nor Helisoma eucosmius has survived

in Greenfield Lake, as Fuller (1977) conjectured. Until recently both of

these species were considered extinct by some authors (Opler 1976,

Imlay 1977, Palmer 1985), but P. magnifica has recently been found in

Orton Pond, approximately 40 km S of Greenfield Lake (Adams and

Gerberich 1988). Helisoma eucosmius has not been located at Orton

Pond, but the habitat appears to be suitable for it.

Campeloma geniculum was not positively identified from the lake.

One specimen of what appeared to be this species was collected at

Station 15, but owing to the difficulties associated with separating it

from C. decisum and the fact that only one was encountered, the

specimen was counted as C. decisum.

Gillia altilus and members of the genera Amnicola and Laevapex

were not found in this study. Suitable habitat for these species has

probably been eliminated by changes in the lake.

By comparing Tables 1 and 2, it can be seen that virtually all of the

gastropods collected during this study represent new species records for

the lake. Of the species represented in Table 2, only the identifications

of the Physidae are uncertain. The species Physella hendersoni and P.

heterostropha could be expected to occur in Greenfield Lake. These

species are difficult to separate based on shell characters, and Burch

(1982) provides only illustrations. Physella heterostropha tends to be

more robust than P. hendersoni, and it was on this character that the

species were separated. “Robustness,” however, is a very weak taxonomic

character.

Changes in the molluscan fauna of the lake are most dramatically

shown in the virtual elimination of the Unionidae. Only two species,

Anodonta cataracta (= A. teres ) and A. imbecillus, are found in the

Greenfield Lake system today compared with 10 that were recorded

historically. The loss of Anodonta couperiana from the lake, previously

suspected by Shelley (1987), has been confirmed. Eupera cubensis, a

sphaeriid also previously reported from the lake, could not be relocated.

When the decline of the Unionidae began is impossible to place.

However, since Bailey collected Elliptio fisheriana in 1940 and Morrison

mentioned three species of Anodonta in the lake in 1972, it would

appear that this is a recent and rapid phenomenon.

The two brackish-water species, Rangia cuneata and Polymesoda

caroliniana, and the freshwater sphaeriid, Sphaerium occidental , are

the only new records of bivalves from the Greenfield Lake basin.

1 14

William F. Adams

Possible Causes of Change

Degraded water quality has been a persistent problem in Greenfield

Lake in recent years, and its decline has been brought about primarily

by the urbanization of the watershed. Elevated levels of nutrients,

pesticides, and metals and the occurrences of algal blooms have been

documented or suspected to have adverse effects on mollusks (Havlik

1987, Havlik and Marking 1987) and have probably done so in the lake.

Several fish kills that occurred in the lake during the study period were

attributed by local authorities to low levels of dissolved oxygen resulting

from excess nutrients. The copper-based algacides used to control

Lyngbia may be adversely affecting the entire benthic macroinvertebrate

community. Havlik and Marking (1987) report that copper sulfate is

toxic to freshwater bivalves at concentrations of 2 to 18.7 mg/ liter in

acute exposures and as low as 25 parts per billion in long-term exposures.

Hanson and Stefan (1984) studied the effects of long-term copper

sulfate application on lakes in Minnesota and found that the normal

functioning of the ecosystems were severely disrupted. Long-term effects

that were discovered included copper accumulation in lake sediments,

changes in species composition from game fishes to rough fishes,

disappearance of macrophytes, and severe reductions in benthic

macroinvertebrates.

The water quality of tributaries of Greenfield Lake may have

changed over time, because of improved drainage in the upper portions

of its basin and the removal of groundwater from the underlying aquifer

by residential wells. Bartsch (1908) states that Greenfield Lake was

spring fed at the time of his collection, but much of the present

freshwater input comes from runoff. In addition, groundwater in the

vicinity of the lake may be polluted. If so, it may take many years for

pollutants entering the lake to be purged, even if similar pollutants from

overland runoff are curbed.

I suspect that the factor most damaging to the unionid populations

has been winter-season partial drawdowns. They have a twofold purpose:

to permit removal of nearshore trash and debris and to kill the exposed

algae mats through cold temperatures and desiccation. These drawdowns

have had, and continue to have, a profound effect on the mollusks of

the lake ecosystem, because all mollusks occurring in this exposed area

are subjected to desiccation and to nighttime temperatures that are

frequently well below freezing. Anodonta cataracta killed by exposure

were observed in several locations. Long (1983) noted significant mortality

of A. cataracta, A. imbecillis, and Lampsilis radiata in a Maryland

reservoir when summer water levels were drawn down rapidly. Samad

and Stanley (1986) found that Elliptio complanata and L. radiata were

Freshwater Molluscan Fauna

115

almost totally eliminated by drawdowns of a lake in Maine. Libois and

Hallet-Libois (1987) found that thin-shelled unionids such as Anodonta

suffered high mortality during a 3-week drawdown of the River Meuse

in Belgium. From those studies, and the observations gathered here, it

appears clear that as long as drawdowns are used as a management

measure in Greefield Lake, the unionid population cannot recover.

Not all of the effects of drawdowns are obvious and direct. Lake

drawdowns drastically reduce the amount of nearshore habitat available

for the fish community and were observed to result in substantial

mortality of fish through strandings, increased predation by wading

birds and gulls, and cold shock. This loss to the fishery is directly

related to the health of the unionid population, because the Unionidae

rely on a fish host for the glochidial stage of their life cycle. A reduction

in the species diversity or abundance of fishes will, therefore, reduce the

number of glochidial hosts available.

Drawdowns also reduce the cross-sectional area of the impounded

portions of tributaries. This reduction in cross-section, with tributary

inflow remaining the same, causes an increase in water velocities in

normally lacustrine areas, and that results in a massive shifting and

redistribution of the bottom sediments. Several Anodonta cataracta

were observed with trails in the sand behind them, apparently attempting

to reestablish themselves. While shifting substrates can be a normal

occurrence in lotic environments, the lacustrine organisms of the lake

probably have difficulty coping with such changes.

During reimpoundment, which takes from 2 to 3 weeks, virtually

no fresh water is released from the dam. Therefore, undiluted waters

from the Cape Fear River reach the base of the dam with each high tide,

and all fish and benthic organisms in the outlet stream are exposed to

abnormally high salinities. Uniomerus tetralasmus and Villosa vibex

were recorded below the spillway by Johnson (1970), and Dawley (1965)

reported Gillia altilis from that area. Habitat for those species has been

altered by increased salinity levels.

As discussed previously, all of the tributaries of Greenfiled Lake

have been channelized to enhance drainage. These creek channels are

periodically maintained by dragline, which results in almost total removal

of all benthic organisms and aquatic vegetation. The consequences of

such actions for molluscan populations are undoubtedly profound;

Greenfield Lake and its tributaries will be repopulated only from

undisturbed waters upstream or downstream of the maintenance area or

from outside sources. Because channels run from the lake all the way to

the headwaters, recolonization is presumed to be very slow.

William F. Adams

1 16

ACKNOWLEDGEMENTS.— I thank Dr. Rowland M. Shelley for

his critical review of the manuscript.

LITERATURE CITED

Adams, W. F., and A. G. Gerberich. 1988. Rediscovery of Planorbella

magnified (Pilsbry) in southeastern North Carolina. Nautilus 102:125-126.

Appleberry, E. L. 1945. Wilmington, N.C. [Habitat destruction during nesting

season.] Chat 9:59-60.

Bailey, J. L. 1940. Wilmington, North Carolina records. Nautilus 54:69

Baker, F. C. 1945. The Molluscan Family Planorbidae. Univ. Illinois Press,

Urbana.

Bartsch, P. 1908. Notes on the fresh-water mollusk Planorbis magnificus and

descriptions of two new forms of the same genus from the southern states.

Proc. U.S. Natl. Mus. 33:697-700.

Basch, P. F. 1963. A review of the freshwater limpet snails of North America

(Mollusca: Pulmonata). Bull. Mus. Comp. Zool. 129:399-461.

Burch, J. B. 1975. Freshwater Sphaeriacean Clams (Mollusca: Pelecypoda) of

North America. Malacological Publications, Hamburg, Mich.

Burch, J. B. 1982. Freshwater Snails (Mollusca: Gastropoda) of North

America. Environ. Prot. Agency, EPA-600/ 3-82-026.

Clench, W. J. 1962. A catalogue of the Viviparidae of North America with

notes on the distribution of Viviparus georgianus Lea. Occas. Pap. Mollusks

2(27):261 -287.

Davis, G. M. 1984. Genetic relationships among some North American

Unionidae (Bivalvia): sibling species, convergence, and cladistic relationships.

Malacologia 25:629-648.

Dawley, C. 1965. Checklist of freshwater mollusks of North Carolina.

Sterkiana 19:35-39.

Fuller, S. L. H. 1977. Freshwater and terrestrial mollusks. Pages 143-194 in

Endangered and Threatened Plants and Animals of North Carolina (J. E.

Cooper, S. S. Robinson, and J. B. Funderburg, editors). N.C. State Mus.

Nat. Hist., Raleigh.

Hanson, M. J., and H. G. Stefan. 1984. Side effects of 58 years of copper

sulfate treatment of the Fairmont Lakes, Minnesota. Water Resour. Bull.

20:889-900.

Havlik, M. E. 1987. Probable causes and considerations of the naiad mollusk

die-off in the upper Mississippi River. In Proceedings of the Workshop on

Die-offs of Freshwater Mussels in the United States (R. J. Neves, editor).

Va. Polytech. Inst, and State Univ., Blacksburg.

Havlik, M. E., and L. L. Marking. 1987. Effects of Contaminants on Naiad

Mollusks (Unionidae): a Review. Resour. Publ. 164, U.S. Fish Wildl. Serv.,

Washington, D.C.

Heard, W. H. 1963. Survey of the Sphaeriidae (Mollusca: Pelecypoda) of the

southern United States. Proc. La. Acad. Sci. 26:101-120.

Heard, W. H. 1965. Recent Eupera (Pelecypoda: Sphaeriidae) in the United

States. Am. Midi. Nat. 74:309-317.

Freshwater Molluscan Fauna

117

Herrington, H. B. 1962. A revision of the Sphaeriidae of North America

(Mollusca: Pelecypoda). Univ. Michigan, Misc. Publ. Mus. Zool. 118:1-74,

pis. 1-7.

Hopkins, S. H., J. W. Anderson, and K. Horvath. 1973. The brackish water

clam Rangia cuneata as indicator of ecological effects of salinity changes in

coastal waters. U.S. Army Engineer Waterways Exp. Sta., Vicksburg, Miss.

Imlay, M. J. 1977. Competing for survival. Water Spectrum 9:7-14.

Johnson, R. I. 1970. The systematics and zoogeography of the Unionidae

(Mollusca: Bivalvia) of the southern Atlantic slope. Bull. Mus. Comp.

Zool. 140:263-449.

Johnson, R. I. 1984. A new mussel, Lampsilis ( Lampsilis ) fullerkati (Bivalvia:

Unionidae) from Lake Waccamaw, Columbus County, North Carolina,

with a list of the other unionid species of the Waccamaw River system.

Occas. Pap. Mollusks 4(63):305-3 19.

Libois, R. M., and C. Hallet-Libois. 1987. The unionid mussels (Mollusca,

Bivalvia) of the Belgian Upper River Meuse: an assessment of the impact of

hydraulic works on the river water self-purification. Biol. Conserv. 42:1 15-132.

Long, G. A. 1983. The unionids (Bivalvia) of Loch Raven Reservoir,

Maryland. Nautilus 97:1 14-1 16.

Moore, L. T. 1968. Stories Old and New of the Cape Fear Region. Wilmington

Printing Co., Wilmington, N.C.

Moore, L. T. 1975. Greenfield marks its 50th year. Wilmington Star-News, 13

April 1975.

Morrison, J. P. E. 1972. Sympatric species of Elliptio in North Carolina. Bull.

Am. Malacol. Union 1971:38-39.

Opler, P. A. 1976. The parade of passing species: a survey of extinctions in the

U.S. Sci. Teacher 43(9):30-34.

Palmer, S. 1985. Some extinct molluscs of the U.S. A. Atala 13(1): 1-7.

Pilsbry, H. A. 1903. The greatest American Planorbis. Nautilus 17:75-76.

Porter, H. A. 1985. Rare and endangered fauna of Lake Waccamaw, North

Carolina watershed system. Vol. 1 and 2. N.C. Endangered Species Restor-

ation: Job Title No. VI-7. N.C. Wildl. Resour. Comm., Raleigh.

Rehder, H. A. 1949. Some land and freshwater mollusks from the coastal

region of Virgina and North and South Carolina. Nautilus 62: 121-126.

Samad, F., and J. G. Stanley. 1986. Loss of freshwater shellfish after a water

drawdown in Lake Sebasticook, Maine. J. Freshwater Ecol. 3:519-523.

Shelley, R. M. 1987. Unionid mollusks from the Cape Fear River basin, North

Carolina, with a comparison of the faunas of the Neuse, Tar, and Cape

Fear drainages (Bivalvia: Unionacea). Brimleyana 13:67-89.

Turgeon, D. D., A. E. Bogan, E. V. Coan, W. K. Emerson, W. G. Lyons, W. L.

Pratt, C. F. E. Roper, A. Scheltema, F. G. Thompson, and J. D. Williams.

Common and Scientific Names of Aquatic Invertebrates From the United

and Canada: Mollusks. Am. Fish. Soc. Special Publ. 16.

Walter, W. M. 1954. Mollusks of the upper Neuse River, North Carolina.

Ph.D. dissertation, Duke Univ., Durham.

Accepted 11 September 1989

118

AUTUMN LAND-BIRD MIGRATION

ON THE BARRIER ISLANDS OF NORTHEASTERN

NORTH CAROLINA

by

Paul W. Sykes, Jr.

For three consecutive years Sykes investigated the autumn migration

of land birds in the Bodie Island and Pea Island area of coastal North

Carolina. During a 102-day period in 1965, he recorded 110,482

individual birds of 148 species. He was able to correlate major influxes

of migratory species with specific weather patterns. His data show

seasonal peaks of southward movement for the land-bird species that

pass along the North Carolina coast in large numbers. In addition,

Sykes recorded five species native to the western United States. Three of

these vagrants provided the first reports of Swainson’s Hawk, Sage

Thrasher, and Western Meadowlark for North Carolina.

1986 49 pages Softbound ISBN 0-917134-12-5

Price: $5 postpaid. North Carolina residents add 5% sales tax. Please make

checks payable in U.S. currency to NCDA Museum Extension Fund.

Send order to: LAND-BIRD MIGRATION, N.C. State Museum of Natural

Sciences, P.O. Box 27647, Raleigh, NC 2761 1.

John White and the Earliest (1585-1587) Illustrations

of North American Reptiles

Hobart M. Smith, Michael J. Preston,1 Rozella B. Smith,2

and Eugene F. Irey1’2

Department of Environmental, Population and Organismic Biology

University of Colorado, Boulder, Colorado 80309

ABSTRACT. — Five paintings of reptiles executed by John White

between 1585 and 1587 are the earliest known depictions of North

American herpetozoa. They represent the box turtle, Terrapene Carolina ;

the diamondback terrapin, Malaclemys terrapin ; the loggerhead, Caretta

caretta; a West Indian iguana, Cyclura ( cychlural ); and a crocodilian

( Crocodylus acutusl). Black-and-white reproductions of the paintings

are provided, as well as a brief account of White’s role in the history of

North American exploration.

It is well recognized (e.g. Smith and Smith 1973) that Francisco

Hernandez was the earliest naturalist to depict American reptiles and

amphibians, through his travels in Mexico from 1570 to 1577. Lamenta-

bly, his manuscript of 15 folio volumes with many illustrations was

destroyed by fire in 1671, but not before it had been extensively revised,

abridged, amended, and plagiarized by several authors. An abridged

version appeared under Hernandez’s name in 1628, augmented in 1649

(see Hulton and Quinn 1964), and the earliest annotated excerpts

appeared under the authorship of Francisco Ximenez in 1615. Only

black-and-white copies of the original illustrations have survived. The

next illustrations of North American species of reptiles and amphibians,

by a naturalist, did not appear until 1743, in Catesby’s famed monograph

(cf. Adler 1979).

Between the dates of Hernandez’s and Catesby’s works, and indeed

only shortly after the earlier of the two, in the period from 1585 to 1587,

John White prepared numerous watercolor paintings, drawings, and

maps while living on Roanoke Island, N.C., or while traveling in its

vicinity and in the West Indies. Of his work dating from 1577 to 1590,

about 75 paintings remain, including one of uncertain origin; 62 depict

animals, plants, Indians, and geography of the areas he visited in the

New World. All of these are replicas White made of the originals, which

are now all lost (Quinn 1955). Five of the 62 paintings depict reptiles;

others of biological interest include 6 of plants, 13 of fishes, 6 of birds,

and 7 of various invertebrates. White’s many other paintings, known

•Department of English, University of Colorado, Boulder, CO 80309-0226.

2 Deceased.

Brimleyana 16:1 19-131, July 1990

119

120

Smith, Preston, Smith, and Irey

through copies made by others, were not preserved. In their time, all

were widely known, widely admired, and usually poorly imitated. His

Indian paintings, copied notably by DeBry, were for centuries the

primary basis for European concepts of native Americans.

The first generally available reproduction of the entire collection of

75 paintings appeared in color, copied from tinted photostats, not the

originals, in a single volume by Lorant (1946). An elaborate, handsome

analysis and reproduction followed 18 years later in a two-volume work

published by, and drawing exhaustively upon the resources of, the

British Museum (Hulton and Quinn 1964). Hulton (1965, 1984) also

reproduced the White paintings and drawings, and Cumming, Skelton,

and Quinn (1972), in a beautifully illustrated book on North American

explorations, reproduced four of the five reptile paintings by White, two

in color.

All the works cited above included most or all of at least the

American drawings, including the reptiles. However, none of the reptile

paintings had been reproduced for general public access prior to 1946,

and none of the three works in which they subsequently appeared had

been prominently noted by herpetologists. In order to bring White’s

contributions more generally to the attention of herpetologists, we here

reproduce in monochrome all five reptile paintings, with the coopera-

tion of the Trustees of the British Museum, where all of White’s known

extant replicas are located.

Although White’s reptile paintings have never been given much

attention by herpetologists, they were first discovered as early seven-

teenth-century copies (by a near descendant of White’s) in a portfolio

acquired in 1709 or shortly thereafter by Sir Hans Sloane, and now in

the British Museum. Sloane had numerous copies made of these copies,

and in turn, Catesby in 1731-1743 copied seven of Sloane ’s copies of

White’s paintings in his “Natural History,” among them the “iguana.”

Lorant (personal communication) sought the assistance of the authori-

ties in the British Museum (Natural History) in identifying the reptile

paintings, and for the Hulton-Quinn volumes Doris M. Cochran of the

U.S. National Museum of Natural History and J. C. Battersby of the

British Museum (Natural History) furnished expert comments on identifica-

tions, expanding on the identifications detailed in Quinn (1955). Howard

H. Peckham, Helen T. Gaige, and Carl Hubbs prepared a locally

distributed pamphlet for a meeting in Ann Arbor, Mich., of the

American Society of Ichthyologists and Herpetologists in 1946, bearing

a reproduction of White’s sea turtle painting as its frontispiece, for a

display of rare books in the Clements Library of the University of

Michigan. We are not aware of any other herpetological attention,

John White’s North American Reptiles

Th

JL land ' 'SU w thr fStu.j^es cJweme a due a/T other Tor/ s

Fig. 1. Terrapene Carolina ( l)carolina (L.), as depicted by John White.

Reproduced from a watercolor of 187-mm greatest straight-line object dimension.

although the box turtle painting is reproduced in color in Borland

(1975).

The Cochran identifications of the three turtles are unimpeachable

(Hulton and Quinn 1964). One painting clearly depicts Terrapene

Carolina (L.), presumably T. c. Carolina (Fig. 1); another represents

Malaclemys terrapin (Schoepff). presumably M. t. centrata (Latreille)

(Fig. 2); and a sea turtle (Fig. 3) is readily identifiable as Caretta caretta

caretta (L.). All three species occur widely on the Atlantic coast and

could easily have been taken near Roanoke Island, where White spent

most of his time. The subspecies of all three turtles are here suggested

on the basis of geographic probability, not depiction of subspecific

characters. Cochran noted (Hulton and Quinn 1964) that the Caretta is

shown with too long a tail and with a pattern too regular and contrasty,

and that the Malaclemys is improperly shown with six fingers and toes

on the right side. Two species are shown with the wrong number of

marginals (too many for Terrapene , too few for Malaclemys).

122

Smith, Preston, Smith, and Irey

Fig. 2 Malaclemys terrapin (tycentrata (Latreille), as depicted by John White

Reproduced from a watercolor of 243-mm greatest straight-line object dimension

John White’s North American Reptiles

123

Fig. 3. Caretta caretta caretta (L.), as depicted by John White. Reproduced

from a watercolor of 221 -mm greatest straight-line object dimension.

124

Smith, Preston, Smith, and Irey

The single lizard illustrated (Fig. 4) is more of a problem. It

is not a temperate North American species, and it is obviously one of the

large, herbivorous iguanines. Because White’s voyages took him into the

northern West Indies, where Cyclura is the only abundant large iguanine,

there is little doubt that some member of that genus is represented. In

Lorant’s book the species is identified as Cyclura carinata, which occurs

(Schwartz and Henderson 1988) only on Booby Cay, Bahama Islands,

and on Turks and Caicos islands. Cochran and Battersby (Hulton and

Quinn 1964) could only conclude that some Cyclura species was

represented, admitting some resemblance to C. carinata. However, Dr.

Albert Schwartz, to whom we turned for help in identifying the species,

explicitly eliminated C. carinata from reasonable consideration because

of its low dorsal crest scales. He also regarded C. cornuta (with

conspicuous spiny verticils on tail), C. ricordi (conspicuous spiny verticils

on tail), and C. rileyi (low crest scales) as equally improbable subjects.

The most likely species, he suggests, is C. cychlura , which not only

agrees structurally but occurs widely in the Bahamas, on Andros (the

largest island of that group) as well as on others. Because the journals of

White’s travels describe repeated landings in the Bahamas, it seems very

likely that C. cychlura (Cuvier) is indeed the species depicted. Cyclura

nubila of Cuba and the Cayman Islands and C. collei of Jamaica,

though structurally in agreement, are less likely candidates on the basis

of probable infrequency of visits by White. However, identification

cannot be certain, for White did travel extensively in the West Indies

(Hakluyt 1589), and furthermore could well have seen specimens of

virtually any species transported by native traders or by explorers such

as Francis Drake, with whom he frequently associated.

The remaining illustration (Fig. 5) represents a crocodilian and is

labeled “Allagatto” on the painting, but it is identified as Crocodylus

acutus (Cuvier) (" Crocodilus americanus” ) in Lorant. Cochran (Hulton

and Quinn 1964) regarded the drawing as impossible to identify, showing

features of both Alligator and Crocodylus , but she concluded that it

most likely represents Alligator mississippiensis (Daudin). She also

noted that the inscription indicating a length of 3 feet, 4 inches, and an

age of 1 month, must be in error; the age would certainly be more than

2 years at that size. In White’s time the alligator occured commonly in

the vicinity of Roanoke Island, as well as northward into Virginia and

southward throughout Florida. White could have seen specimens of it

anywhere in his travels in what are now parts of Virginia and North

Carolina. However, he also explored the northern West Indies, where

the alligator does not occur but the crocodile was common. Thus, we

agree that either species could easily have been illustrated. Although his

John White’s North American Reptiles 125

chances of exposure to the alligator were obviously much more numerous

than chances of exposure to the crocodile, the features shown certainly

more closely conform with those of a crocodile than an alligator. The

label “Allagatto” is just as likely a corruption of the Spanish “el

lagarto,” applied to the crocodile in the West Indies, as of “alligator”

(also a corruption from Spanish), applied to Alligator. The light color

and narrow straight jaws are particularly significant. We conclude that

the crocodile significantly influenced the depiction, even if it is a

composite. The crocodile does occur in Florida, but only at the extreme

southern tip; White did not reach any part of Florida, his two paintings

of Floridians having been copied from the work of Jacques Le Moyne

de Morgues.

There is also the possibility, brought to our attention by Dr. Adler,

that White could have been influenced in his depiction of the crocodilian

by illustrations widely circulated in Europe by that time of the similar

Crocodylus niloticus of Egypt. In view of the reasonably close accuracy

of his other paintings, we assume that White did not have a chance to

examine any crocodilian very closely, else his depiction would have been

more faithful to the subject. Hence, the influence of extraneous

impressions, as of C. niloticus illustrations of his era, should not be

excluded as a possibility in the apparent absence of close observations

of the American species.

The preceding identifications were adopted by Hulton (1984) from

a preliminary version of this article that he kindly reviewed in 1981.

Despite the flaws now evident in White’s herpetological paintings

when compared with modern illustrations, in the context of his era his

drawings are remarkably superior, surpassing any others executed for

several succeeding generations, including the works of Catesby (1731-

43) and Bartram (1791). His stature as a natural history artist is

unequalled and merits as much honor as is commonly awarded, for

example, to Audubon in a much later era. Because all of his paintings

that now survive are replicas of originals now lost, it is likely that some

fidelity to the originals has been lost.

White’s five paintings are the earliest known for North American

and West Indian reptiles, and the earliest now in existence for any

reptiles of the western hemisphere. They are not the earliest published

illustrations, however, because they were not reproduced for the general

public until Lorant included them in his 1946 work.

Although mass-reproduced with accuracy only in the past 35 years

or so, some of White’s painting were redrawn many times in works in

the preceding centuries, beginning with the Sloane (1709-14?) portfolio

of 112 leaves of drawings (now only 110). In the Sloane volume the

126

Smith, Preston, Smith, and Irey

Fig. 4 Cyclura (l)cychlura (Cuvier), as depicted by John White. Reproduced

from a watercolor of 200-mm greatest straight-line object dimension.

John White’s North American Reptiles

127

I

is

N

s

i

W$d

Fig. 5 ( l)Crocodylus acutus (Cuvier), as depicted by John White. Reproduced

from a watercolor of 218-mm greatest straight-line object dimension.

128

Smith, Preston, Smith, and Irey

Cyclura and crocodilian, as well as a skink and a snake, were present

among the 44 drawings adapted from originals by White that are now

lost. Catesby copied his “iguana” from the Cyclura (Hulton and Quinn

1964). The monochrome reproductions of these two additional reptiles

(Hulton and Quinn 1964) are too inaccurate (whether originally so or

from unfaithful copying) to identify satisfactorily. Cochran (in Hulton

and Quinn 1964) regarded the snake as “probably” a Lampropeltis,

whereas we suggest that Nerodia is a more likely model. She thought

the skink was “probably” a mature male Eumeces fasciatus (L.), although

noting that E. inexpectatus and E. laticeps also occur in the vicinity of

Roanoke Island and could possibly have been represented in the painting.

There are only two other redrawn versions of White’s herpetological

paintings of early date, both noted in Hulton and Quinn (1964). One is

a 1589 work by Walter Bigges (“A summarie and true discourse of Sir

Francis Drake’s West Indian Voyage . . .”), showing an “iguana” and an

“alligator or crocodile” drawn by Baptista Boazio from White’s figures,

here reproduced as Fig. 4 and 5. A “turtle” also shown in that volume is

too crudely stylized to be identifiable; its uncertain source is apparently

not White. The second is John Mountgomery’s manuscript of 1588-1589

(“A treatice concerning the navie of England . . .”), in which White’s

Caretta is shown quite clearly in a corner of a large naval panorama.

According to Hulton (1984), this is “the earliest known copy of a John

White drawing.”

The only other herpetological subject matter in Hulton and Quinn’s

illustrations is some stylized snakes decorating the bodies of some Piet

warriors illustrated in manuscripts by Lucas de Heere, about 1575

(Hulton and Quinn 1964).

Although numerous non-herpetological paintings by White were

republished in subsequent years, not until the 1930s were the reptiles

reproduced faithfully in color. One unique, complete set of all of

White’s paintings was copied by a Miss Bessie Barclay for the Newport

News public schools, and five other sets consisting of tinted photostats

were prepared under the direction of Mrs. Sonia Tregaskis (Hulton and

Quinn 1964). One of the Tregaskis sets is at the University of Michigan

and was the source of the reproduction in the pamphlet by Peckham,

Gaige, and Hubbs (1946), and in the book by Lorant (1946), which

finally brought White’s work to rank-and-file accessibility. Hulton and

Quinn (1964) pursued their definitive work with the conviction that

Lorant’s book, “though useful, was unfortunately marred by the entirely

unreliable quality of the plates and text alike” — an opinion universally

shared by academic reviewers.

The history of John White is shrouded with uncertainties (Quinn

1955). He was an Englishman of sufficient stature to be appointed by

John White’s North American Reptiles 129

Sir Walter Raleigh as governor of a group of 1 13 men and women sent

in 1587 as a second attempt to establish a colony on Roanoke Island,

“Virginia” (now North Carolina). White, along with the eminent scientist

Thomas Harriot, had gone with the earlier group in 1585, stayed with

them on the island for a year, and returned (as did the rest of the

colonists) with Sir Francis Drake to England in 1586. The second group

that White accompanied in 1587 did not fare particularly well; only a

little more than a month after their arrival on Roanoke (22 July 1587),

White departed again (27 August 1587) for England to procure supplies

for the colonists. In that interval, on 18 August 1587, his daughter

Eleanor, wedded to Ananias Dare, gave birth to the first child born of

English parents in America — Virginia Dare. A second child (name and

sex unknown) was born on Roanoke Island to Dyonis and Margery

Harvey, just a week or so later.

Unfortunately, war with Spain was then imminent, and hence

supplies, ships, and personnel were diffcult to obtain. Not until 1590,

three years after his departure to obtain succor for the colonists, was it

possible for White to return to Roanoke Island, where little trace of the

colonists, including White’s daughter and granddaughter, was found.

Thus ensued the mystery of the “Lost Colony,” famous in the history of

early English settlement in America. Quinn (1985) concluded that most

migrated to southeastern Virginia, where they lived peacefully with a

friendly Indian tribe until about 1607, when they were massacred by

Powhatan’s tribe. The rest, a very small group, remained for a time on

Roanoke Island, but ultimately moved to nearby Croatoan Island to

await White, presumably living with Indians there, but their fate is

unknown.

White retired into virtual obscurity in Ireland shortly after the

unsuccessful relief mission returned to England. His career as an

administrator and governmental leader was ignominious, but not so his

visual records of paintings, drawings, and maps, executed less than a

quarter of a century before the first successful English colonization took

place in Virginia.

John White’s role as herpetologist was a small facet of the large

part he played in early American history and of the contribution he

made to the image of America among educated Europeans. He was a

gifted artist of the first rank in his time. Although direct credit was long

in coming, since his copiers gained far greater fame than he, his

paintings created images of native North America that linger today on

both sides of the Atlantic, to a considerable extent as a result of his

collaboration with Harriot (1588). White’s Lost Colony has fostered

legends that continue to stimulate the imaginations of those who have

concern for colonial American history.

130

Smith, Preston, Smith, and Irey

ACKNOWLEDGMENTS.— We are much indebted to Dr. Kraig

Adler of Cornell University, an eminent historian of herpetology, for

counsel and encouragement; to Dr. Albert Schwartz, of Miami-Dade

Community College, a leading authority on West Indian herpetology,

for careful appraisal of the proper identification of the Cyclura depicted

by White; to the trustees of the British Museum for the privilege of

reproducing selected paintings by White; to Dr. John E. Cooper,

formerly of the North Carolina State Museum of Natural Sciences, for

extensive aid and encouragement; and through Dr. Cooper for the

honor of much historical guidance from Dr. David B. Quinn, a leading

historian on the early settlement of North America, and from Paul H.

Hulton, an eminent authority on early English art history.

LITERATURE CITED

Adler, K. 1979. A brief history of herpetology in North America before 1900.

Soc. Study Amphib. Reptiles Herpetol. Circ. 8:1-40.

Bartram, W. 1791. Travels through North & South Carolina, Georgia, East &

West Florida, the Cherokee Country, the Extensive Territories of the

Muscogulges, or Creek Confederacy, and the Country of the Chactaws; Contain-

ing an Account of the Soil and Natural Productions of Those Regions,

Together with Observations on the Manners of the Indians. James and

Johnson, Philadelphia.

Borland, H. G. 1975. The History of Wildlife in America. Natl. Wildl.

Federation, Washington, D.C.

Catesby, M. 1731-43. The Natural History of Carolina, Florida, and the

Bahama Islands. 2 vol. Privately printed by author, London.

Cumming, W. P., R. A. Skelton, and D. B. Quinn. 1972. The Discovery of

North America. American Heritage, New York.

Hakluyt, R. (fil.). 1589. The Principall Navigations, Voiages and Discoveries

of the English Nation, Made by Sea or Ouer Land, to the Most Remote

and Farthest Distant Quarters of the Earth at any Time Within the

Compasse of These 1500 years; Divided into Three Seuerall Parts, According

to the Positions of the Regions Whereunto They Were Directed. Barker,

London. [Reprinted 1965, Hakluyt Soc., London. 2 vol.]

Harriot (also spelled Hariot), T. 1588. A Briefe and True Report of the New

Found Land of Virginia. Robinson, London. [Reprinted several times,

notably in Quinn (1955, Vol. 1).]

Hulton, P. H. 1965. The watercolor drawings of John White from the British

Museum. U.S. National Gallery of Art, Washington.

Hulton, P. H. 1984. America 1585: The Complete Drawings of John White.

Univ. North Carolina Press, Chapel Hill.

Hulton, P. H., and D. B. Quinn. 1964. The American Drawings of John

White, 1577-1590, with Drawings of European and Oriental Subjects. Vol.

1. A Catalogue Raisonne and a Study of the Artist. Vol. 2. Reproductions

of the originals in Colour Facsimile and of Derivatives in Monochrome.

British Museum, London.

John White’s North American Reptiles

131

Lorant, S. 1946. The New World: The first Pictures of America Made by John

White and Jacques Le Moyne and Engraved by Theodore De Bry with

Contemporary Narratives of the Huguenot Settlement in Florida 1562-

1565 and the Virginia Colony 1585-1590. Duell, Sloan & Pierce, New York.

Peckham, H. H., H. T. Gaige, and C. L. Hubbs. 1946. Ichthyologia et

herpetologia americana. Bull. W. L. Clements Library, Univ. Mich. 25:1-22.

Quinn, D. B. 1955. The Roanoke Voyages. 2 vol. Hakluyt Soc., London.

Quinn, D. B. 1985. Set Fair for Roanoke: Voyages and Colonies, 1584-1606.

Univ. North Carolina Press, Chapel Hill.

Schwartz, A., and R. W. Henderson. 1988. West Indian amphibians and

reptiles: a check-list. Milwaukee Public Mus. Contr. Biol. Geol., No. 74.

Sloane, H. 1709-14? The original draughts of ye habits, towns customs &c: of

the West Indians and of the plants birds fishes &c found in Groenland,

Virginia, Guiana &c by Mr John White who was a painter & accompanied

Sir Walter Raleighe in his voyage. British Museum, London. Formerly Ms.

5270, now Paintings and Drawings 199. a. 3., 112 leaves.

Smith, H. M., and R. B. Smith. 1973. Synopsis of the Herpetofauna of

Mexico. Vol. II. Analysis of the Literature Exclusive of the Axolotl. Eric

Lundberg, Augusta, W.V.

Accepted 11 September 1989

132

POTENTIAL EFFECTS OF OIL SPILLS ON SEABIRDS

AND SELECTED OTHER OCEANIC VERTEBRATES

OFF THE NORTH CAROLINA COAST

by

David S. Lee and Mary C. Socci

Based primarily on data gathered offshore during the past 15 years by the

staff of the North Carolina State Museum of Natural Sciences, this book

presents information regarding distribution and susceptibility to oil pollution

for 25 oceanic species: 14 birds, 6 mammals, and 5 turtles. An overlay can be

placed on range maps to demonstrate the proximity of species occurrence to the

oil lease sites off Cape Hatteras.

1989 64 pages Softbound ISBN 0-917134-18-4

Price: $8, postpaid. North Carolina residents add 5% sales tax. Please make checks

payable to U.S. currency to NCDA Museum Extension Fund.

Send order to: OIL SPILL BOOK, N.C. State Museum of Natural Sciences,

P.O. Box 27647, Raleigh, NC 27611.

Genetic Patterns and Population Structure

in Appalachian Trechus of the vandykei Group

(Coleoptera: Carabidae)

Thomas C. Kane

Department of Biological Sciences

University of Cincinnati, Cincinnati, Ohio 45221

Thomas C. Barr, Jr.

School of Biological Sciences

University of Kentucky, Lexington, Kentucky 40506

AND

Gail E. Stratton

Department of Biology

Albion College, Albion, Michigan 49224

ABSTRACT. — The genus Trechus is diverse and widespread in the

southern Appalachian region. A majority of its species are alpine

endemics, altitudinally restricted to elevations above 1,350 m. Five

taxonomic subgroups of the vandykei species group of Trechus were

examined electrophoretically to assess patterns of differentiation within

and between taxa. Genetic differentiation within subgroups is slight to

moderate, suggesting that gene flow between local populations is

maintained or has only recently been interrupted. Differentiation

between subgroups is moderate to very great, indicating complete

genetic isolation at present. Varying degrees of affinity between sub-

groups are consistent with the hypothesis that speciation has resulted

from lineage vicariance caused by fluctuating Pleistocene climates. The

vandykei group belongs to the endemic southern Appalachian subgenus

Microtrechus, which probably originated southwest of a lowland

dispersal barrier, the Asheville basin and the French Broad River

valley. Electrophoretic data indicate that the vandykei subgroup has

dispersed northeast of this lowland relatively recently. Affinities of this

subgroup with isolates in the Great Smoky and Unicoi mountains,

rather than the pisgahensis subgroup in the Great Balsam Mountains

immediately south of the Asheville basin, suggest that this lowland has

been a dispersal barrier throughout the Pleistocene and earlier.

Dispersal of Microtrechus species east of the Asheville basin and of

Trechus, s. str., species west of the lowland probably occurred across

the narrow French Broad River gorge and the mountain chains along

the North Carolina-Tennessee border.

More than three-fourths of North American species of the carabid

beetle genus Trechus occur in the southern Appalachians. In this region

Brimleyana 16:133-150, July 1990

133

134

T. C. Kane, T. C. Barr, Jr., and G. E. Stratton

Trechus exceeds all other carabid genera in taxonomic diversity (Barr

1985a). Nearly 55 Trechus taxa are known from the region at present; a

great majority of them are alpine endemics isolated at elevations above

1,350 m. Much of the diversity in this genus is a result of lineage

vicariance associated with fluctuating climatic regimes during the

Pleistocene. Presumably ancestral Trechus species were more continuously

distributed at lower elevations during colder, wetter climates associated

with glacial maxima. However, the warmer, drier interglacial climates

made lowlands inhospitable to most trechines and resulted in vertical

contraction and fragmentation of ranges. The present insular pattern of

distribution of alpine Trechus taxa is a direct result of the recent

climatic regime (Barr 1962, 1979, 1985a).

Evolution in Appalachian Trechus has also been influenced by the

Asheville basin, a major lowland drained by the French Broad and

Pigeon rivers. The two subgenera represented in the region, Trechus, s.

str. (males with two protarsomeres enlarged, dentate, and setose beneath),

and Microtrechus (males with only one protarsomere so modified), are

essentially separated by the Asheville basin and the French Broad River.

Microtrechus , endemic to the Unaka mountain province, appears to

have evolved west of the basin, in isolation from subgenus Trechus, s.

str., to the east (Barr 1962, 1979). However, occurrence of a limited

number of Microtrechus species east of the French Broad River and a

few Trechus, s. str., species west of the river suggests that the barrier has

recently been breached (Barr 1985a). In general, the area southwest of

Asheville exhibits greater endemicity and diversity in many groups of

carabids, and carabids in the mountains northeast of Asheville are

taxonomically much closer to carabid species and genera in the mountains

of western Virginia and eastern West Virginia (see Barr 1969 for

summary). The eyeless, wingless, edaphobitic species of Arianops

(Coleoptera: Pselaphidae) also show much the same pattern (Barr 1974;

for a detailed discussion of the evolutionary impact of the Asheville

lowland, see Barr 1985a).

Morphological differences between closely related isolates of Trechus

are often subtle, involving quite minor, though consistent, characters.

The taxonomy of the isolates belonging to the vandykei species group of

Microtrechus has proven especially difficult. Twelve upland isolates — all

more or less morphologically distinct and strictly allopatric — are known

from western North Carolina and eastern Tennessee (Fig. 1. shows all

known localities except Joanna Bald, in the Snowbird Mountains).

Beetles in this group are quite small (total length means <3 mm) and

characteristically inhabit the superficial layers of moist or wet litter in

the forest floor and carpets of loose, wet, fluffy mosses. Local populations

135

Appalachian Trechus of the vandykei Group

are often quite abundant, exhibiting densities of about 20 to 80 individuals

per square meter. The vandykei group is represented in the Black-Great

Craggy and Bald mountains east and north of Asheville, and to the west

in the Great Smoky, Newfound, Unicoi, Cheoah, Snowbird, Tusquitee-

Valley River, and Great Balsam ranges. An apparent relict population

occurs on Whiteside Mountain, where the Cowees meet the Blue Ridge

escarpment. Nevertheless, some curious distributional gaps exist: Pop-

ulations assignable to this group have not been found elsewhere in the

Cowees, on the Toxaway Mountain spur off the Great Balsams, in the

Plott Balsams (between the Smokies and Great Balsams), nor in the

Nantahalas, despite special efforts to collect them there. Discontinuity is

also found in the chain of higher peaks along the Tennessee-North

Carolina border, with the vandykei group represented on Sandymush

Bald and Camp Creek Bald, but not on Tennessee Bluff in between.

Trechus taxa assignable to other species groups are relatively abundant

in these areas where the vandykei group is absent.

Although extensive collecting over the past quarter century has well

established the altitudinal restriction of these isolates, such negative

evidence does not totally preclude the possibility of some limited gene

flow across lowlands. In fact, one specimen of T. bowlingi has been

taken in Greenbrier Cove, in the Great Smokies, and one specimen of T.

tusquitee was taken near Old Road Gap along the north approach to

Tusquitee Bald, both specimens near elevations of about 900 m. Also,

several altitudinally restricted species, including T. bowlingi , occur at

about 950 to 1,050 m on the north (Tennessee) side of the Great

Smokies, where cool microclimates prevail as cold air flows down from

the crest through deep ravines. However, the rarity of specimens at

lower elevations indicates only that minimal gene flow across lowlands

is possible, not that it is significant.

In this study we examine a suite of closely related Trechus taxa (the

vandykei species group) using the technique of gel electrophoresis. This

technique permits us to look at another set of characters, enzymatic

proteins, thus providing data complementary to biogeographic and

morphological considerations. Electrophoretic data are easily quantified

and permit insight into relative degrees of biochemical differentiation.

The underlying genetic basis of electrophoretic variation can usually be

inferred, allowing us to determine whether or not limited gene flow

exists between geographically and altitudinally isolated populations.

Although most of the isolates sampled are restricted to single peaks,

three are more widely distributed in major, continuous uplands — T.

vandykei in the Black and Great Craggy mountains, T. bowlingi in the

Great Smoky Mountains, and T pisgahensis in the Great Balsam

136

T. C. Kane, T. C. Barr, Jr., and G. E. Stratton

Mountains, including Pisgah Ledge. We sampled three populations of

vandykei and four populations each of bowlingi and pisgahensis ; thus

differentiation between local populations within the same continuous

upland can serve as a baseline for comparing taxa on isolated peaks.

The isolates within the vandykei species group can be arranged in

five subgroups, each bearing one of the five available trivial names of

taxa assigned to the species group. Although all 12 isolates differ to a

greater or lesser extent in minor morphological characters and will be

the subject of a subsequent taxonomic paper (Barr, in preparation), we

propose no new names in this paper. The subgroups are as follows.

1) vandykei subgroup — (a) T. vandykei Jeannel (1931), Black and

Great Craggy mountains and adjacent Blue Ridge, Yancey Co. and

MacDowell Co., N.C.; three other isolates on (b) Camp Creek Bald,

Greene Co., Tenn., and Madison Co., N.C.; (c) Big Bald, Unicoi Co.,

Tenn., and Yancey Co., N.C.; and (d) Unaka Mountain, Unicoi Co.,

Tenn., and Mitchell Co., N.C.

2) bowlingi subgroup — one isolate, T. bowlingi Barr (1962), widespread

in the eastern two-thirds of the Great Smoky Mountains National Park,

Tenn. and N.C.

3) tusquitee subgroup — (a) T. tusquitee Barr (1979), Tusquitee

Bald, Macon Co., Clay Co., and Cherokee Co., N.C.; two other isolates

on (b) Joanna Bald (Snowbird Mountains), Graham Co. and Cherokee

Co., N.C.; and (c) Cheoah Bald, Graham Co. and Swain Co., N.C.

4) haoe subgroup — one isolate, T. haoe Barr (1962), known only

from Haoe Lead above Joyce Kilmer Memorial Forest, Unicoi Mountains,

Graham Co., N.C.

5) pisgahensis subgroup — (a) T. pisgahensis Barr (1979), widely

distributed in the Great Balsam Mountains and their eastern arm,

Pisgah Ledge, in Buncombe Co., Haywood Co., Jackson Co., and Transylvania

Co., N.C.; two other isolates on (b) Sandymush Bald, Newfound

Mountains, Haywood Co. and Madison Co., N.C., and (c) Whiteside

Mountain, Jackson Co. and Macon Co., N.C.

METHODS

A total of 19 populations representing 11 of the 12 known isolates

of the vandykei group were sampled during the summer in 1982, 1983,

and 1984 for electrophoretic analysis. [We were unable to recollect the

Joanna Bald population discovered in 1960 (Barr 1962), although other

Trechus (Microtrechus) species belonging to the nebulosus group were

found there in some abundance.] Beetles were collected by sifting forest-

floor litter or moss from a hardware-cloth basket into a plastic dishpan,

where they could be removed with an aspirator or by hand. The most

137

Appalachian Trechus of the vandykei Group

productive microhabitat proved to be damp, but not excessively wet,

litter in Rhododendron thickets. Beetles were transported to the laboratory

in 4-ounce glass or plastic jars cooled in an ice chest. During the

summer of 1982, specimens were maintained alive in a refrigerator at

the Highlands Biological Station prior to electrophoresis, which was

also conducted at the Station. Beetles collected at the end of the

summer in 1982 and all 1983 and 1984 collections were returned alive to

the University of Cincinnati; after identification and sexing they were

placed individually into 400-jul microcentrifuge tubes and frozen at

-80° C until used for electrophoresis.

Electrophoresis was conducted on vertical polyacrylamide slab gels

with a Hoefer Scientific SE600 system. Single beetles were ground in

approximately 40^1 of grinding buffer (0.01M Tris-HCl, pH 7.0,

containing 0.001M EDTA, 1% Triton-X, and 25% sucrose). Initial

screening dictated that only a single sample could be obtained from

each specimen because of the beetles’ small size, low enzyme activity, or

both. Ten enzymatic systems were surveyed: alkaline phosphatase (ALP),

carbonic anhydrase (CAH), esterase (EST), hexokinase (HEX), malate

dehydrogenase (MDH), mannose phosphate isomerase (MPI), phospho-

glucose isomerase (PGI), phosphoglucomutase (PGM), superoxide

dismutase (SOD), and xanthine dehydrogenase (XDH). Only five systems,

to include eight presumptive loci, could be consistently scored in all

individuals of all taxa: CAH (1 locus), EST (2 loci), MDH (2 loci), PGI

(1 locus), and SOD (2 loci). Staining techniques for these systems were

adapted from Brewer (1970), Harris and Hopkinson (1976), and Shaw

and Prasad (1970).

All data analysis was accomplished with a FORTRAN-77 version

of the BIOSYS-1 Program developed by Swofford and Selander (1981).

This program contains routines for population genetic analysis as well

as procedures for the production of phenograms and other types of

phylogenetic analyses.

RESULTS

For practical purposes, the 19 populations sampled were assigned

to the five subgroups whose limits and distribution are described in the

introductory section. Locations of the sampling sites and the abbreviations

employed for them throughout this paper are presented in Table 1, and

their relative geographic configuration is shown in Fig. 1. The sampled

vandykei group populations (except those of Sandymush Bald and

Cheoah Bald) coexist with larger Trechus species of other groups; see

Barr (1985a) for details of the various species guilds to which the

vandykei-g roup isolates belong.

Table 1. Location of sampling sites for Trechus beetles in North Carolina and

Tennessee.

1) vandykei subgroup

Buncombe Co. and Yancey Co., at

juncture of Black and Great Craggy

mountains

Yancey Co./Tenn.: Unicoi Co., near

summit

Madison Co./Tenn.: Greene Co.,

near summit

Yancey Co., near summit

Buncombe Co., 3 km W of Craggy

Dome

Mitchell Co./Tenn.: Unicoi Co., 10

km E of Erwin

Swain Co./Tenn.: Sevier Co.,

Clingmans Dome Road

Haywood Co., eastern Great Smokies

Sevier Co., 4 km W of Mount Guyot

Sevier Co., at US 441 bridge over

Walker Prong

Graham and Swain counties, 5 km S

of Stecoah, just west of summit

Cherokee Co., Clay Co., and Macon

Co., immediately south and east of

summit

Graham Co., ridge north of Joyce

Kilmer Memorial Forest

Haywood Co., just north of gap, 2

km W of Devils Courthouse

Haywood Co., just north of gap, 4

km WNW of Beech Gap

Jackson Co., west side, 4 km SE of

Balsam

Henderson Co. and Transylvania

Co., 0.5 km E of summit

Haywood Co. and Madison Co., 5

km E of Cove, at summit

Jackson Co., 4 km SW of Cashiers,

summit

Appalachian Trechus of the vandykei Group

139

Fig. 1. Map of the Unaka mountain region, showing locations of Trechus

vandykei species-group populations examined in this study. Subgroup designa-

tions of populations are as follows: bowlingi — CG Collins Gap, HE Heintooga

Overlook; RC Ramsay Cascades, WP Walker Prong; haoe — HA Haoe Lead;

pisgahensis — BG Beech Gap, BP Bearpen Gap, DG Deep Gap, MP Mount

Pisgah, SM Sandymush Bald, WH Whiteside Mountain; tusquitee — CH Cheoah

Bald, TU Tusquitee Bald; vandykei — BA Balsam Gap, BB Big Bald, CC Camp

Creek Bald, MM Mount Mitchell, SN Snowball Mountain, UN Unaka Mountain.

Six of the eight loci examined were monomorphic, with the same

electromorph fixed in all populations of all taxa examined (Table 2).

The remaining two loci, CAH and PGI (Table 2, part B), were variable

within and/or between populations. Four of the five subgroups showed

very similar patterns of genetic variability (Table 3), with all local

populations variable at both the CAH and PGI loci; i.e. average

polymorphism (P) = 0.25, and average heterozygosity (H) ranged between

0.059 and 0.1 10 for these subgroups. The pisgahensis subgroup exhibited

less genetic variation. Only the PGI locus was variable in this subgroup,

and only in three of the six populations studied, such that P = 0.08 and

H = 0.009 (Table 3).

Table 2. Electromorph frequencies at eight electrophoretic loci for 19 populations of the Trechus vandykei species group.

o

c3

o

u

oo —

in — —

in (N

o o

m oc

<N OO

— o

o o

so

o

o'

O n O in

<N <N in (N

o it m

o’ o’ o'

Z o

X

U

o

m

oa

co

so

m

OO (N

in it

O' o

o o

in o o o o

(N 00 0 00 00

in O (N O

o o o' o

2

O'

r-

< in

CO it

<Lt

o

53

-s:

<

X

m

CS

— O'

O' o

oo —

o> o

in — o'

ci r- cs

rs r~~-

c> o

m o \o t

m o oo —

m t rs

O 0)0)

— r-

m r~-

o’

O m

r- —

o o

I

£

oo

o

so

CM

o

2

C/3

O so

O —

O

Tj- vO

it m

oo —

O C>

»5

ST

-s:

a

S?

5S,

CL.

S

o

Q

o

o it

O m

O

vO it

(N r~

O' o

o o’

o

ni

o

Cl oo

QQ —

O m

o —

o

CQ

o

(N

o r-

o —

o

— O'

Tf <n

o^ o

o o'

X .

< c<fflUQWlL a Oc<qquQwU.

U c 0.

Monomorphic loci (6) with the same electromorph fixed in all populations: EST-1; EST-2; M DEI-1; M DEI-2; SOD-1; SOD-2. Monomorphic and/

polymorphic loci (2): CAH, PGI.

Appalachian Trechus of the vandykei Group

141

Table 3. Genetic variability at eight enzymatic loci for five subgroups of the

Trechus vandykei species group.

142

T. C. Kane, T. C. Barr, Jr., and G. E. Stratton

Genotype frequencies at variable loci in each population were

tested statistically for fit to expectations under Hardy-Weinberg Equili-

brium (HWE). Three statistical tests were used to test correspondence of

the data to HWE: Chi-square goodness of fit with Levene’s (1949)

correction for small sample size, an exact probability test analogous to

Fisher’s exact test, and — in cases where three or more electromorphs

occurred in a population — a Chi-square test with pooling. Because

sample sizes were relatively small (i.e. almost always < 50 per population),

and the Chi-square test is less reliable when expected values for some

classes are small (Sokal and Rohlf 1981), we chose to reject the null

hypothesis of HEW only when all tests applied indicated statistically

significant ( P < 0.05) deviation from HWE. Sixteen populations were

variable at the PGI locus, and in all cases at least one test was not

significant ( P > 0.05), suggesting that none deviates from HWE.

Two populations of the bowlingi subgroup (CG and WP; Fig. 1)

were found to deviate significantly from HWE expectations at the CAH

locus by all three tests. For the remaining 1 1 variable populations,

however, HWE at the CAH locus could not be rejected by one or more

statistical tests. Two separate collections, one in 1983 and the other in

1984, were made at both the CG and WP sites. Further, the two

collections were made at slightly different, though proximate, microhabitats

within each site. In contrast, the HE and RC bowlingi locations were

sampled on a single date and at single sites in each location. Thus

deviation from HWE at the CAH locus for the CG and WP locations

may reflect temporal and / or microspatial heterogeneity in gene frequency

in these populations. However, the fact that no such genetic heterogeneity

is evident at the PGI locus for these two locations argues against this

contention. Further, because 27 of the 29 cases of variable loci in

populations meet our conservative requirements for fit to HWE, it is

difficult to ascribe much significance to these two exceptions from

available data.

Variation among populations within subgroups was examined using

F-statistics and a Chi-square contingency analysis (Workman and

Niswander 1970) (Table 4). Significant heterogeneity in gene frequency

( P < 0.05) occurs among populations in the bowlingi, tusquitee, and

vandykei subgroups at both the CAH and PGI loci. The six pisgahensis-

subgroup populations, which are monomorphic for the same allele at

the CAH locus, show significant heterogeneity in allele frequency at the

PGI locus. For those subgroups in which significant heterogeneity in

allele frequencies was observed, genetic differentiation can be described

as slight (Fsx < 0.05) to moderate (0.05 < Fsx <0.15) (Table 4). Rogers’

Genetic Similarity (S) calculated over all loci produces values of

143

Appalachian Trechus of the vandykei Group

Table 4. F-statistics and heterogeneity Chi-square values for four subgroups of

the Trechus vandykei species group.

o •

F - correlation between uniting gametes relative to the gametes of the total

IT population.

°F s = average correlation over subdivisions of uniting gametes relative to those

of their own subdivision.

p

Fst = correlation of random gametes within subdivisions relative to gametes of

the total population.

d ** = P < 0.01; *** = P < 0.005.

approximately 0.90 or greater for comparisons between local populations

within subgroups (Table 5).

Differentiation between subgroups is substantial in some cases.

Intersubgroup genetic similarities (Table 5) range from values that are

not very different from infrasubgroup similarities ( haoe vs. vandykei, S

= 0.921) to values suggesting more distant affinity (pisgahensis vs.

vandykei, S = 0.776). Clustering of Rogers’ Distance values for the 19

populations using UPGMA (Sneath and Sokal 1973) produces five

clearcut groupings (Fig. 2.). The six populations of the pisgahensis

subgroup form the most distinct cluster. Thus, this subgroup, which

appears to have lower genetic variability than the other four, is also the

most biochemically dissimilar of the five subgroups (i.e., S = 0.826 for

all between-subgroup comparisons). Closer affinities are observed between

the cluster containing the two tusquitee- subgroup populations and the

cluster of the four bowlingi populations. The six populations in the

fourth cluster include all of the populations of the vandykei subgroup,

Table 5. Rogers (1972) coefficients of genetic similarity (S) for comparisons of five subgroups of the Trechus vandykei speci

group.3

144

T. C. Kane, T. C. Barr, Jr., and G. E. Stratton

t A

•A*

C

<3

>

© ©

3

m

Ob

o

© ©

2.1

<u *->

JD &

B I

3 CXi

►7 °

bO

<N

bO

O

ctj

Values shown are averages of pairwise comparisons of appropriate populations. Values in parentheses are the ranges

of values appropriate to each comparison.

Appalachian Trechus of the vandykei Group

145

P/SGANENS/S

r-H

TUSQU/TEE

BOWLING!

VANDYKE!

HAOE

BG

BP

DG

WH

*MP

■SM

■CH

TU

CG

•RC

HE

WP

BA

UN

MM

•SN

• CC

* BB

HA

— i

0.00

0.15

0.10

0.05

ROGERS DISTANCE

Fig. 2. UPGMA dendrogram of 19 populations of the Trechus vankydei species

group, generated from Rogers’ Distance values for eight biochemical loci.

and the single haoe population has its closest affinities with this

subgroup.

Genetic differentiation can also be described using Nei’s (1977) gene

diversity. In this analysis the total gene diversity observed (Ht) can be

apportioned among the various levels of a specified hierarchy (Table 6).

We recognize four hierarchical levels of gene diversity: (1) within

colonies (populations), He; (2) among populations within subgroups,

Dcs; (3) between subgroups within regions, Dsr; and (4) between

regions, Drt, where the two regions are the areas northeast and southwest

of the Asheville basin, respectively. Approximately 4-5% of total gene

diversity is attributable to differences between infrasubgroup populations

(Dcs). More than 50% of gene diversity is a result of differentiation

between subgroups (Dsr and Drt). Differentiation between subgroups

within regions (Dsr) appears to be as great as (PGI) or greater than

(CAH) differentiation between regions (Drt) (Table 6).

Affinities between various subgroups can also be seen in the

geographic patterns of certain diagnostic electromorphs at the two

variable loci (Table 2). Electromorph “C” at the PGI locus is restricted

to populations in the northeast region, where it occurs in all but the CC

(westernmost) population of the vandykei subgroup. Conversely,

electromorph “D” at the PGI locus is present in only two vandykei

146

T. C. Kane, T. C. Barr, Jr., and G. E. Stratton

subgroup populations (CC, BB), where it occurs at low frequency.

However, electromorph “D” occurs in all populations in the southwest

region and is generally the most frequent PGI electromorph in these

populations. If dispersal from southwest to northeast took place along

the mountains on the Tennessee-North Carolina border, it is significant

that the “D” electromorph occurs in populations on the first two major

summits (CC and BB) east of the French Broad River and not elsewhere.

At the CAH locus the “D” and “F” electromorphs, which are the only

CAH variants present in the northeast populations, are either rare (“D”)

or absent (“F”) in southwest populations, with one exception. The

exception is the single peripheral population of T. haoe (HA), in which

both electromorphs occur at frequencies comparable to those seen in

vandykei-subgrouy populations of the northeast region (Table 2). The

pisgahensis subgroup can be distinguished from all other subgroups by

the presence of a “null” allele at the CAH locus. This enzyme system,

which stains intensely and consistently in the other 13 populations

sampled, fails to yield any zones of staining in the six pisgahensis-

subgroup populations examined.

DISCUSSION

Analysis of the electrophoretic data in this study suggests that

altitudinal isolation of subgroups in the vandykei species group of

Trechus is complete. The very great genetic differentiation between

subgroups (FSJ = 0.533), the fixation of a unique allele in one subgroup

(CAH locus in pisgahensis ), and the restriction of alleles to geographic

regions (e.g. PGI electromorph “C” to the northeast region) indicate an

absence of gene flow between subgroups and regions at present. The

pisgahensis subgroup appears to have been isolated for the longest

period of time. Not only does it have a unique allele fixed at the CAH

locus, but it also shows a marked reduction in heterozygosity, which is

consistent with long-term isolation. This result is evidence that the

Asheville basin, which lies between the ranges of the pisgahensis and

vandykei subgroups, may have been an effective dispersal barrier to

these beetles even during glacial maxima; consequently, the earlier

taxonomic treatment of vandykei and pisgahensis as geographic races of

the same biological species (Barr 1979) requires modification.

The degree of differentiation between populations within subgroups

provides a basis for interpreting divergence between subgroups. For

those subgroups whose component populations are isolated by lowland

barriers (i.e. pisgahensis, vandykei, tusquitee), differentiation within the

subgroup is only slight to moderate (FSJ < 0.15). Assuming that gene

flow is no longer maintained between such isolates, these data suggest

that the break in gene flow has been relatively recent, possibly as recent

Appalachian Trechus of the vandykei Group

147

Table 6. Analysis of gene diversity at two electrophoretic loci in 19 populations

of five subgroups of the Trechus vandykei species groupa.

Number

of

aHT = total gene diversity. Hc = gene diversity within colonies (local populations).

Dcs = gene diversity among populations within subgroups. DSR = gene

diversity between subgroups within regions. D_x = gene diversity between

K. I

regions.

as post-Wisconsian. In fact, differentiation among populations of bowlingi

is as great as that observed within the three “multi-isolate” subgroups

mentioned above. The bowlingi populations, however, are sampled

from an abundant species continuously distributed over the upland area

in the eastern two-thirds of the Great Smoky Mountains, the most

extensive mountain range in the region. Furthermore, bowlingi is

sympatric and usually syntopic with different assemblages of other,

more narrowly distributed Trechus species in different parts of its range

(Barr 1962, 1979). Thus, differentiation in bowlingi could result from

longer, stepwise pathways of gene flow, but it may also reflect local

adaptation to a broader spectrum of microhabitat heterogeneity

throughout its more extensive geographic range.

Biochemical, biogeographical, and morphological affinities (Barr

1979) between (a) bowlingi and the tusquitee subgroup and (b) haoe and

the vandykei subgroup suggest a relatively recent common ancestor in

each case. Other relationships between subgroups are vague, suggesting

that the associated speciation events occurred in the more distant past.

The present data do permit some speculation as to the route of dispersal

of the vandykei species group from the southwest to the northeast

region. As previously noted, the taxonomic hypothesis that vandykei

and pisgahensis belong to the same biological species (Barr 1979) is

rendered untenable on two counts: (1) the electrophoretic data indicate

that pisgahensis (and its morphologically related Sandymush and

Whiteside isolates) is the most distinct biochemically and least variable

of the five vandykei- group clades, suggesting more distant affinity with

148

T. C. Kane, T. C. Barr, Jr., and G. E. Stratton

vandykei and longer-term isolation than had previously been postulated;

and (2) the vandykei subgroup has a much stronger affinity with haoe,

the similarity being most striking at the CAH locus, where both

subgroups not only have the same two electromorphs but have them in

very similar frequencies. Trechus haoe and T. vandykei are readily

differentiated by morphological characters (Barr 1962); furthermore,

they occupy the western and eastern extremes of the geographic region

tracked by the entire vandykei group.

It appears that the vandykei subgroup dispersed across the French

Broad River valley by a route farther north than previously suggested,

probably along the chain of high mountains on the Tennessee-North

Carolina border, then south into the Blacks, Great Craggies, and

adjacent high Blue Ridge. The electrophoretic data clearly are in accord

with biogeographical and morphological data supporting this hypothesis;

the only representative of Microtrechus in the northeast region other

than the vandykei subgroup isolates is T. inexpectatus (Barr 1985b),

described from Camp Creek Bald (CC), the first major peak encountered

as one proceeds eastward along the Tennessee-North Carolina border

from the French Broad River valley. Going in the opposite direction,

from northeast to southwest across the French Broad, the first major

peak encountered is Tennessee Bluff (Cocke Co., Tenn., and Madison

Co., N.C.); two Trechus s. str. taxa occur on Tennessee Bluff, an

undescribed species related to T. scopulosus (Barr 1962, 1979) and the

only subspecies (undescribed) of the abundant, widely distributed

(northwest North Carolina, northeast Tennessee, southwest Virginia,

eastern West Virginia) polytypic T. ( T .) hydropicus known from the

southwest region (Barr, in preparation). The hypothesis of an earlier,

more remote separation of subgenus Microtrechus from Trechus s. str.

during a period of isolation in the Unaka region west of the Asheville

basin is thus supported not only by “center of origin” considerations but

by additional biochemical and biogeographical data.

More difficult to explain are the close affinities between T. haoe

(Unicoi Mountains) and the vandykei subgroup, given that, at present,

the range of T. bowlingi in the Great Smoky Mountains intervenes. One

possible but speculative scenario (Barr 1985a) for evolution of the

vandykei group includes the following sequence:

1) An ancestral species diverged from Microtrechus by occupying

the niche of small predator in superficial litter; among southern

Appalachian Trechus spp., only the vandykei group isolates occupy that

niche at present (Barr 1985a).

2) The ancestral species split into populations in the two major

mountain ranges of the southwest region, the Smokies (GSM) and the

Great Balsams-Pisgah Ledge (GBM).

149

Appalachian Trechus of the vandykei Group

3) The GBM population dispersed northward and southward, but

only relict populations survive in the Newfound Mountains to the north

and on Whiteside Mountain to the south. The GSM population,

however, dispersed to the northeast along the Tennessee-North Carolina

border and southwest into the Cheoah, Snowbird, Tusquitee-Valley

River, and Unicoi mountains.

4) In the Smokies, the extensive upland area and microhabitat

heterogeneity (patchiness) favored colonization and speciation in other

species groups of Trechus ; the vandykei- group isolate was subject to

strong selection pressure for niche divergence to permit coexistence.

Meanwhile, colonies dispersing to the northeast and southwest evolved

less rapidly; consequently, they are more similar to each other than to

bowlingi , the species that now occupies the Smokies.

5) An early offshoot of the GSM population eventually colonized

the mountains southwest of the Smokies — Tusquitee Bald, Joanna Bald,

and Cheoah Bald. Finding only one, two, or no competing Trechus

species, they diverged less rapidly than the GSM ancestor. Possibly they

represent an evolutionary stage intermediate to the beetles that dispersed

northeastward and the present-day, more derivative species now found

in the Smokies, T. bowlingi.

6) Meanwhile, the original GSM ancestral type crossed the French

Broad River valley and successively colonized CC, BB, the Blacks-Great

Craggies to the south, and finally Unaka Mountain to the northeast. A

relict population of the southwesterly dispersal survives in the Unicoi

Mountains as T. haoe. The two peripheral subgroups have evolved less

(at least biochemically) than the parental populations they left behind in

the Smokies.

ACKNOWLEDGMENTS .— We thank William Badaracca, Judith

Barr, and Thomas C. Barr III for assistance with the field collecting. A

significant portion of this research was conducted at the Highlands

Biological Laboratory, Highlands, N.C. We acknowledge the cooperation

of the staff at Great Smoky Mountains National Park, Gatlinburg,

Tenn., and Wolf Laurel Resort, Mars Hill, N.C., for permission to

collect in areas under their supervison. This study was supported in part

by National Science Foundation grants DEB-8202273 (TCK) and DEB-

8202339 (TCB) and by a grant from the Highlands Biological Foundation

(TCB).

LITERATURE CITED

Barr, T. C., Jr. 1962. The genus Trechus (Coleoptera: Carabidae: Trechini) in

the southern Appalachians. Coleopt. Bull. 16:65-92.

150 T. C. Kane, T. C. Barr, Jr., and G. E. Stratton

Barr, T. C., Jr. 1969. Evolution of the Carabidae (Coleoptera) in the southern

Appalachians. Pages 67-92 in The Distributional History of the Biota of the

Southern Appalachians. Part 1: Invertebrates (P. C. Holt, editor). Res.

Div. Mon. 1, Va. Polytech. Inst., Blacksburg.

Barr, T. C., Jr. 1979. Revision of Appalachian Trechus (Coleoptera: Carabidae).

Brimleyana 2:29-75.

Barr, T. C., Jr. 1985a. Pattern and process in speciation of trechine beetles in

eastern North America (Coleoptera: Carabidae). Pages 350-407 in Taxonomy,

Phylogeny, and Zoogeography of Beetles and Ants (G. E. Ball, editor). Dr.

W. Junk Publ., Dordrecht.

Barr, T. C., Jr. 1985b. New trechine beetles (Coleoptera: Carabidae) from the

Appalachian region. Brimleyana 11:119-134.

Brewer, G. J. 1970. An Introduction to Isozyme Techniques. Academic Press,

New York.

Harris, H., and D. A. Hopkinson. 1976. Handbook of Enzyme Electrophoresis

in Human Genetics. North Holland, Amsterdam.

Jeannel, R. 1931. Revision des Trechinae de l’Amerique du Nord. Arch. zool.

exp. et gen. 71:403-499.

Levene, H. 1949. On a matching problem arising in genetics. Ann. Math. Stat.

20:91-94.

Nei, M. 1977. F-statistics and analysis of gene diversity in subdivided

populations. Ann. Human Gen. 41:225-233.

Shaw, C. R., and R. Prasad. 1970. Starch gel electrophoresis of enzymes — a

compilation. Biochem. Gen. 4:297-320.

Sneath, P. H. A., and R. R. Sokal. 1973. Numerical Taxonomy. W. H.

Freeman, San Francisco.

Sokal, R. R., and F. J. Rohlf. 1981. Biometry. 2nd ed. W. H. Freeman, San

Francisco.

Swofford, D. L., and R. B. Selander. 1981. BIOSYS-1: A FORTRAN

program for the comprehensive analysis of electrophoretic data in population

genetics and systematics. J. Heredity 72:281-283.

Workman, P. L., and J. D. Niswander. 1970. Population studies on south-

western Indian tribes. II. Local genetic differentiation in the Papago. Am.

J. Human Gen. 22:24-49.

Accepted 18 September 1989

BRIMLEYANA

A Journal of Zoology of the Southeastern United States

151

INVITATION FOR PAPERS

The editorial staff of Brimleyana welcomes contributions. Research papers

based on studies conducted in the southeastern United States are invited. Areas

of interest include systematics, evolution, zoogeography, ecology, behavior, and

paleozoology. Brief communications, as well as articles of more typical length,

will be considered. Page charges will be waived for those who lack institutional

or grant support. Inquiries should be directed to:

Editor, Brimleyana

North Carolina State Museum of Natural Sciences

P.O. Box 27647

Raleigh, NC 27611

DATE OF MAILING

Brimleyana No. 15 was mailed on 17 February 1989.

ERRATA

Brimleyana No. 15:

Page 114, lines 18 and 20: The genus for these two flycatchers should be

spelled Empidonax.

Page 120, line 19: Monarch genus should be spelled Danaus.

INFORMATION FOR CONTRIBUTORS

Submit original and two copies of manuscripts to Editor, Brimleyana, North Carolina

State Museum of Natural Sciences, P.O. Box 27647, Raleigh, NC 27611. In the case of

multiple authorship, indicate correspondent. Manuscripts submitted for publication in this

journal should not currently be submitted elsewhere.

Preparation of manuscript. Adhere generally to the Council of Biology Editors Style

Manual, Fifth Edition. Use medium-weight bond paper, 8!4 * 11 inches, and leave at least

an inch margin on all sides. Double-space all typewritten material, including tables and

literature cited.

The first page will contain only the title, the name(s) and address(es) of the author(s),

and necessary footnotes to those. Where appropriate, the title will indicate at least two

higher categories to which taxa belong. Example: Studies of the genus Hobbseus Fitzpa-

trick and Payne (Decapoda: Cambaridae).

A brief informative abstract on a separate sheet follows the title page. Indicative

abstracts are not acceptable. Footnotes, other than those in tables (see below), will be used

only when absolutely necessary, and will be numbered consecutively throughout the

paper.

Individuality of writing style and text organization are encouraged, but most papers are

best presented in the form of Introduction, Materials and Methods, Results, Discussion,

and Literature Cited, with those headings centered and all capitals. Headings plus

subheadings must be kept to a total of three levels.

Scientific names in taxonomic papers will include the author when first used in the text.

Descriptions of new taxa must be in accordance with the requirements of established

international codes. Presentations of etymologies are desirable.

Form for literature cited:

Woodall, W. R., Jr., and J. B. Wallace. 1972. The benthic fauna in four small southern

Appalachian streams. Am. Midi. Nat. 88:393-407.

Crocker, D. W., and D. W. Barr. 1968. Handbook of the Crayfishes of Ontario. Univ.

Ontario Press, Toronto.

Authors, not the editor, are responsible for verifying references.

Form for citing references in text: parenthetical (Woodall and Wallace 1972). 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. Footnotes in each table will be indicated by consecutive superscript lower-case

letters. Indicate lightly in pencil in the margin of the original manuscript where tables and

illustrations 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‘/2 * 1 1 inches). They should be proportioned to make optimal use of the

page format of the journal. Plates must be prepared and presented as they are to appear,

and should not be presented as collections of individual items for arrangement by the

editors. Do not mount individual photographs. The author’s name, title of the manuscript,

figure number, and the notation “Top” should be penciled lightly on the back of each

illustration. Lettering on original drawings and maps should be of adequate size to permit

proper reduction when needed. Do not type on illustrations; legends should be typed,

double-spaced, on separate sheets.

Page charges, reprints, and proofs. A charge of $30 per page 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 at time of submittal. This

will not affect acceptance for normal publication. Contributors who pay full page costs

will be furnished 100 free reprints. Reprint order forms will be sent with page proofs and

are to be returned to the Managing Editor. On papers with more than one author, it is the

responsibility of the correspondent to assure that other authors have an opportunity to

obtain reprints. Both galley and page proofs are to be corrected, signed, and returned to

the Managing Editor within seven days. Please notify the Managing Editor if additional

time is required for proofreading.

BRIMLEYANA No. 16, JULY 1990

CONTENTS

First Record of the Rock Vole, Microtus chrotorrhinus (Miller) (Rodentia: Cricetidae), in

Virginia. John F. Pagels 1

Occurrence of a Northern Cicada, Okanagana rimosa (Homoptera: Cicadidae), in the Southern

Appalachians. E. E. Brown and J. D. Brown 5

Distribution and Ecology of the Blackside Dace, Phoxinus cumberlandensis (Osteichthyes:

Cyprinidae). Christopher J. O’Bara 9

Oviposition, Larval Development, and Metamorphosis in the Wood Frog, Rana sylvatica

(Anura: Ranidae), in Georgia. Carlos D. Camp, Charles E. Condee, and

D. Glenn Lovell 17

Seasonal Diet of the Margined Madtom, Noturus insignis (Osteichthyes: Ictaluridae), in a North

Carolina Piedmont Stream. Robert P. Creed, Jr., and Seth R. Reice 23

Population Dynamics of Adult Unionicola formosa (Acari: Hydracarina), a Parasite of

Anodonta imbecillis (Mollusca: Bivalvia), in West Virginia. James E. Joy and

Jeffrey W. Hively 33

Reproduction in the Hispid Cotton Rat, Sigmodon hispidus Say and Ord (Rodentia: Muridae),

in Southeastern Virginia. Robert K. Rose and Michael H. Mitchell 43

Occurrence of the Milliped Auturus erythropygos erythropygos (Brandt) in Virginia

(Polydesmida: Platyrhacidae). Rowland M. Shelley 61

Kleptoparasitism of a River Otter, Lutra canadensis, by a Bobcat, Felis rufus, in South Carolina

(Mammalia: Carnivora). James F. Bergan 63

Spring Movement Patterns of Two Radio-tagged Male Spotted Turtles. Jeff Lovich 67

New Records of the Distribution and the Intestinal Parasites of the Endangered Northern Flying

Squirrel, Glaucomys sabrinus (Mammalia: Sciuridae), in Virginia. John F. Pagels, Ralph P.

Eckerlin, John R. Baker, and Michael L. Fies 73

Age Estimates for a Population of American Toads, Bufo americanus (Salientia: Bufonidae), in

Northern Virginia. Heather J. Kalb and George R. Zug 79

Small Mammals in the Great Dismal Swamp of Virginia and North Carolina. Robert K. Rose,

Roger K. Everton, Jean F. Stankavich, and John W. Walke 87

Recent Changes in the Freshwater Molluscan Fauna of the Greenfield Lake Basin, North

Carolina. William F. Adams 103

John White and the Earliest (1585-1587) Illustrations of North American Reptiles. Hobart M.

Smith, Michael J. Preston, Rozella B. Smith, and Eugene F. Irey 119

Genetic Patterns and Population Structure in Appalachian Trechus of the vandykei Group

(Coleoptera: Carabidae). Thomas C. Kane, Thomas C. Barr, Jr., and Gail E. Stratton 133

Miscellany 151

200334