|
of the
FLORIDA STATE MUSEUM
Biological Sciences
Volume 28 1982 Numbers 1-4
COLLECTED PAPERS
IN MEMORY OF
HOWARD W. CAMPBELL
UNIVERSITY OF FLORIDA
GAINESVILLE
Numbers of the BULLETIN OF THE FLORIDA STATE MUSEUM, BIOLOGICAL
SCIENCES, are published at irregular intervals. Volumes contain about 300 pages and are
not necessarily completed in any one calendar ear.
J. C. Dickinson. Jr.. Editor
Rhoda J. Bryant. Managing Editor
Consultants for this issue:
John \\. Hermanson
G. Roy Horst
Stephen R. Humphrey
J. Larry Landers
Margaret K. Lantgxorthy
Bruce J. MacFadden
Rochelle A. Marrinan
Brian K. McNab
Gary S. Morgan
Henry Setzer
Fred G. Thompson
Elizabeth S. \\inmz
C(onilnilcation' concernim.I pirelhae or \(.hda f nt *it pibli(atio in ai! Ia:i r!: ,
should be addressed to: Manaoinc Editor. Bulletin. I 'lrida State \MueMiim,. L:.lter. 1
Florida: (;amiesn ille. Fl[ 32(11. LB A
Cop\riglht b\ the Florida State Museuinn o thle 'lll\erst\ it Florida
BULLETIN OF THE FLORIDA STATE MUSEUM
BIOLOGICAL SCIENCES
1982
Volume 28
No. CONTENTS Page
Foreword ii
Bibliography of Howard Wallace Campbell v
1 The Helicina umbonata Complex in the West Indies
(Gastropoda, Prosobranchia, Helicinidae). Fred G.
Thompson. 22 October 1982. $1.75 1
2 Microtus pennsylvanicus (Rodentia: Muridae) in Flor-
ida: A Pleistocene Relict in a Coastal Saltmarsh.
Charles A. Woods, William Post, and C. William
Kilpatrick. 22 October 1982. $2.10 25
3 Distribution and Evolution of Florida's Troglobitic
Crayfishes. Richard Franz and David S. Lee. 22
October 1982. $1.90 53
4 Human Predation on the Gopher Tortoise (Gopherus
polyphemus) in North-central Florida. Robert W.
Taylor, Jr. 22 October 1982. $1.75 79
J. C. Dickinson, Jr., Editor
Rhoda J. Bryant, Managing Editor
FLORIDA STATE MUSEUM
UNIVERSITY OF FLORIDA
GAINESVILLE, FLORIDA
FOREWORD
This special issue of the Bulletin
honors H. W. Campbell, who died of
Cancer in Gainesville, Florida, on 10
December 1981. He was a valued col-
league and longtime friend of the
Florida State Museum.
Howard Wallace Campbell was
born in Baltimore, Maryland, on 23
October 1935. Early in his life he ac-
quired the nickname "Duke," an ap-
pellation that in later life seemed
suited to his tall and lean frame though
Sm somewhat out of place for his gentle
personality. As a young boy, Duke de-
veloped an avid interest in natural
S i history and was soon combing the
hills, forests, abandoned fields, and
marshes around Baltimore in pursuit
of reptiles and amphibians. He was only sixteen when he joined the
Natural History Society of Maryland, but he rapidly became one of the
most active members of the organization. Duke participated in many
Society outings and soon was leading field trips to various parts of the
eastern United States. It was on one of those trips that he first became
fascinated with Florida's reptile fauna. His broad interest in nature,
coupled with his enthusiasm and expanding knowledge of the subject,
gained him the respect of fellow members, many of whom he influenced
into further pursuit of careers and hobbies in biology.
Duke attended the University of Florida as an undergraduate student
majoring in zoology. During his spare time he earned much of his pocket
money by collecting live reptiles and amphibians for friends in Baltimore
who were compiling a photographic inventory of the North American
herpetofauna. And it was difficult to be in the field very long with Duke
without becoming envious of his collecting abilities. Friendly competition
to see who could collect the most species or specimens in an afternoon kept
Duke in beer and many of us buying. Duke first became involved with the
Florida State Museum in 1957 through regular use of the museum's collec-
tions for a course in herpetology. His strong interest in the rich and diverse
relictual biota of the southeastern United States developed during those
years and continued throughout his life. Duke graduated from Florida in
1958 with a Bachelor of Science degree.
During a two year stint in the U.S. Army, Duke was stationed at Ft.
Benning, Georgia, which allowed him to return to the University of
Florida when on leave to maintain contact with colleagues.
Duke continued his graduate studies at the University of California at
Los Angeles, where he earned a Master of Science degree in 1963 and a
Ph.D. in 1967. At UCLA, Duke studied physiological ecology and neuro-
biology to gain a better understanding of the evolutionary adaptations of
various species.
After graduation, Duke accepted a Postdoctoral Fellowship at
Washington University's Center for the Biology of Natural Systems in St.
Louis and conducted research on reptile and amphibian ecology at its
Middle American Research Unit in Panama. A year later he returned to
the University of Florida as a Postdoctoral Fellow in the Department of
Neuroscience, College of Medicine (1968-1969), and then became Interim
Assistant Curator of Herpetology, Florida State Museum (1969-1970),
and Interim Assistant Professor of Zoology, College of Arts and Sciences
(1970-1972). He then left academia to work for a year as a biological con-
sultant with an environmental engineering firm in Pennsylvania, a posi-
tion that involved him in environmental impact studies in this country
and in the middle east. In 1973, Duke accepted a position as Zoologist in
the Office of Endangered Species and International Activities, U.S. Fish
and Wildlife Service, in Washington, D.C. As the first herpetologist on
that staff, he was expected to develop recovery plans for the reptiles and
amphibians on the U.S. Endangered Species List. His ability to work with
others and his broad background in zoology made him extremely effective
in his work, but he was not happy laboring in the huge Washington
bureaucracy so far removed from the southeastern habitats and fauna
that he loved.
In 1974, Duke transferred to the position of Chief Zoologist in the
National Fish and Wildlife Laboratory, U.S. Fish and Wildlife Service, at
the Field Station in Gainesville, Florida. He remained in this position for
the balance of his career. In 1980, Duke also received an appointment as
Adjunct Professor in the School of Forest Resources and Conservation,
University of Florida.
Duke was well-liked and respected by all who knew him. His informal
attitude and quiet manner set other people at ease and promoted coopera-
tion. He possessed a quick and penetrating insight into people. As senior
biologist at the Gainesville Laboratory, Duke initiated many inter-
discliplinary research projects and did not hesitate to involve colleagues
and students in the work. It was often his responsibility to report the
results of environmental impact assessments on southeastern habitats and
wildlife. Thousands of data filled pages resulted from his efforts, and
though they helped to decide various issues at public hearings and to
develop government conservation policy, few were ever released by the
bureaucracy for publication in scientific journals. Results he could not
share through publication, Duke freely passed verbally to colleagues and
students. He was generous with his time and assistance. Consequently, his
published scientific contributions are but a small measure of his ac-
complishments as a researcher.
Duke traveled extensively in his role as a wildlife biologist. The U.S.
Fish and Wildlife Service involved him in international projects in Latin
America and the Caribbean, in Australia and Papua New Guinea, and in
southern Asia, as well as in the United States. He adapted readily to local
cultures, enjoying each and quickly winning friends wherever he worked.
Duke was more than just a professional wildlife biologist, he was also
an avid conservationist, contributing time and effort to many environ-
mental causes. From 1979 to 1981, he served as chairman of the Crocodile
Specialist Group of the Species Survival Commission (SSC), International
Union for the Conservation of Nature and Natural Resources. In that
capacity he helped draft conservation programs for this group of reptiles
in nations around the world. He also served as consultant on Sirenia to the
SSC Marine Mammal Specialist Group. He had a natural ability to find a
middle course between rampant exploitation and total prohibition, a
course that conserved the living resource while allowing its controlled
utilization. For this he earned the respect of both sides in many a wildlife
issue. And it is in his capacity as a biologist and conservationist of interna-
tional standing that Duke will be remembered.
Duke was long a supporter of the Florida State Museum. As an under-
graduate he often contributed specimens to the museum's research collec-
tions, while at the same time good-naturedly poking fun at the student
assistants working in the Herp Range by calling them "Splashers," a
reference to the sound of specimens being dropped into preservative. His
later studies for the U.S. Fish and Wildlife Service yielded tens of
thousands of voucher specimens to the museum's systematics collections,
including the first large, statistically valid samples taken from many
Florida habitats.
While a student at the University of California, Duke married Linda
Kathlene French on 9 April 1965. He was a close and devoted husband
and father. He is survived by his wife, Kathy, and their two children,
Mariel Lee and Colin James; and by Sabra, a daughter from a previous
marriage.
This issue of the Bulletin expresses our admiration for Duke and our
sense of loss at his passing.
F. Wayne King
Director
Florida State Museum
BIBLIOGRAPHY OF HOWARD WALLACE CAMPBELL
1958. (with J. Alan Holman). An interesting feeding activity of Microhyla carolinensis.
Herpetologica 14: 205.
1960. (with R.S. Simmons). Notes on the eggs and young of Eumneces callicephalus Bocourt.
Ibid. 17: 212-213.
1960. The bog turtle in Maryland. Maryland Naturalist 30: 15-16.
1962. An extension of the range of Haldea valeriae in Florida. Copeia 1962(2): 438-439.
1962. (with R. S. Simmons). Notes on the eggs and larvae of Rhyjacosiredon altimironi
Duges. Herpetologica 18: 131-133.
1962. (with R. S. Simmons). Notes on some reptiles and amphibians from the West Coast of
Mexico. Bull. So. California Acad. Sci. 61: 193-203.
1965. (with T. R. Howell). Herpetological records from \1...,.,L' Herpetologica 21:
130-140.
1967. (with W. E. Evans). Sound production in two species of tortoises. Ibid. 23: 204-209.
1967. (with N. Suga). Frequency sensitivity of single auditory neurons in the gecko. Coh'ou/yx
cariegatus. Science 157: 88-90.
1967. The turtle in modern research. J. Internatl. Turtle Tort. Soc. 1: 1-3.
1967. The acoustic behavior of turtles. Ibid. 1: 13-14. 44.
1968. (with Peter H. Hartline). Hearing and vibration detection in snakes. Proc. Internatl.
Union Physiol. Sci. 7: 183 (abstra.)
1969. The effects of temperature on the auditory sensitivity of lizards. Physiol. Zool. 42(2):
183-210.
1969. (with Peter H. Hartline). Auditory and vibratory responses in the midbrain of snakes.
Science 163: 1221-1223.
1969. Adaptation and perception. J. Internatl. Turtle Tort. Soc. 3(1): 6-9.
1969. The unsung chicken turtle. Ibid. 3(5): 4 pp.
1970. (with S. R. Telford). Ecological observations on an all female population of the lizard
Lepidophyma flacimiaculatium in Panama. Copeia 1970(2): 379-381.
1970. Prey selection in naive Elaphe obsoleta (Squamata: Serpentes): a reappraisal.
Psychon. Sci. 21(5): 300-301.
1970. Social and vocal behavior of Mexican crocodilians. Amer. Philos. Soc. Yrbk.: 301
(abstr.)
1971. (with S. R. Telford). Observations on two species of the Hyla rubra group (Anura:
Hylidae) in Panama. J. Herpetology 5(1-2): 52-55.
1971. Observations on the thermal activity of some tropical lizards of the genus Anolis
(Iguanidae). Caribbean J. Sci. 11(1-2): 17-20.
1971. Subspecies, handle or hobble? Bull. Maryland Herp. Soc. 7(4): 87-90.
1972. (with K. Campbell). Subdivision, turtle style: behavior and resource partitioning. J.
Internatl. Turtle Tort. Soc. 6(2): 16-19, 35.
1972. Ecological or phylogenetic interpretations of crocodilian nesting behaviours. Nature
238: 404-405.
1972. (with W. E. Evans). Observations on the vocal behavior of chelonians. Herpetologica
28(3): 277-280.
1972. (with S. P. Christman). Dangerous land snakes of southeast Asia. Pp. 27-84 in G. V.
Pickwell and W. E. Evans (eds.) Handbook of dangerous animals for field personnel.
Naval Undersea Center TP 324, San Diego.
1972. (with S. P. Christman and W. E. Evans). Crocodiles of southeast Asia. Ibid. pp. 85-96.
1973. Ecological observations on Anolis lionotus and Anolis poecilopus (Reptilia, Sauria) in
Panama. Amer. Mus. Novit. (2516): 1-29.
1973. Preliminary report: Status in, ... 'r...r., of Morelet's crocodile in Mexico. Zoologica
57(3): 135-136.
1973. Observations on the acoustic behavior of crocodilians. Zoologica: 1-11.
1973. Endangered species lists: need for consistency. HISS News-J. 1(3): 83-84.
1974. Turtles at the brink: our endangered species. Bull. Maryland Herp. Soc. 10(1): 1-7.
1974. Anelytropsis and A. papillosus. Pp. 156.1-156.2 in Catalogue of North American
Amphibians and Reptiles. Soc. Study Amphib. Rep.
1974. (with J. Jackson and K. Campbell). The feeding behavior of crocodilians: validity of
the evidence from stomach contents. J. Herpetology 8(4): 378-381.
1975. (Review of) State regulations for collecting reptiles and amphibians in the fifty United
States by Adrian F. Czajka and Max A. Nickerson. Copeia 1975(2): 386.
1976. The Florida manatee and related species. Plaster Jacket No. 25.
1976. (with J. A. Powell) Endangered species: the Florida manatee. Florida Nat., April:
15-20.
1976. (with J. Jackson and W. Ingram). The dorsal pigmentation pattern of snakes as an an-
tipredator strategy: a multivariate approach. American Nat. 110(976): 1029-1053.
1977. Manatees. In Making Aquatic Weeds Useful: Some Perspectives for Developing
Countries. Nat. Acad. Sci., pp. viii, 1-174.
1977. The Florida manatee, Trichechus manatus. Pp. xvi, 1-444 (396-397) in J. E. Cooper,
S. S. Robinson and J. B. Funderburg (eds.). Endangered and Threatened Plants and
Animals of North Carolina. North Carolina State Mus. Nat. Hist.; Raleigh.
1978. (with F. G. Thompson). Observations on a captive Mona Island Boa, Epicrates
monensis monensis Zenneck. Bull. Maryland Herp. Soc. 14(2): 98-99.
1978. (with A. B. Irvine). The feeding ecology of the Florida manatee. Aquaculture
12(1977): 249-251.
1978. (with A. B. Irvine). Aerial census of the West Indian Manatee, Trichechus manatus,
in the Southeastern United States. J. Mamm. 59(3): 613-617.
1978. (with D. Gicca). Resena preliminary del estado actual y distribution del manati
(Trichechus manatus) en Mexico. An. Inst. Biol. Univ. Nac. Auton. Mexico, 49 (Ser.
Zool): 257-264.
1978. (with B. A. Irvine). Aerial census of the West Indian manatee, Trichechus manatus,
in the southeastern United States. Pp. 17-22 in R. L. Brownell and K. Ralls (eds.).
The West Indian Manatee in Florida. Special Publication, Florida Dept. Nat. Res.
1978. (with A. B. Irvine). Manatee mortality during the unusually cold winter of
1976-1977. Ibid.: 86-92.
1978. (with A. B. Irvine and D. K. Odell). Manatee mortality in the southeastern United
States from 1974 through 1977. Ibid.: 67-76.
1978. Status reports for the following species. In W. Hillstead (ed.). Endangered verte-
brates of the southeastern United States. Tall Timbers Research Station and South-
eastern Division of Wildlife Society.: Diamondback terrapin, Florida pine snake,
short-tailed snake, Miami black-headed snake.
1978. (with M. Fogarty). Status reports for the following species. In J. N. Layne (ed.). En-
dangered vertebrates of Florida. Florida Audubon Society: Short-tailed snake,
Stilosoma extenuatum; Miami black-headed snake, Tantilla oolitica; American
alligator, Alligator mississipiensis.
1979. (with S. Busack). Laboratory care. Pp. 109-125 in M. Harless and M. Morlock (eds.),
Turtles: Perspectives and Research. Wiley. Interscience, Inc.,: New York.
In
Press: (with S. P. Christman). The systematic status of Phyllorhynchus decurtatus porelli
Powers and Banta (Reptilia: Colubridae). J. Herpetology.
(with S. P. Christman). Techniques for herpetofaunal community analysis. Proc.
Symp. Herpetol. Communities. Lawrence, Kansas.
(with S. P. Christman). The herpetological components of Florida sandhill and sand
pine scrub Association. Ibid.
(with N. Scott). A history of research on herpetological communities. Ibid.
(with B. Means). Effects of prescribed burning on amphibians and reptiles. Proc.
Symp. Effects Prescribed Burning Southern Forests.
REPORTS
1971. Reptiles and amphibians of Amelia Island. Pp. 152-180 in Amelia Island, Nassau
County, Florida: a preliminary ecological inventory. Jack McCormick & Associates.
1971. Annotated list of the reptiles and amphibians of Amelia Island. Ibid.: 280-311.
1973. (with D. Zumeta). Wildlife. In Preliminary Ecological Inventory for a Proposed
Refinery in Gloucester Co., N.J. Jake McCormick & Associates. 187 p.
1976. (with S. Christman). The amphibians and reptiles of the Proposed Cross Florida
Barge Canal Route. Cross Florida Barge Canal Restudy Report, U.S. Army Corps of
Engineers. 249 p.
1976. The manatee populations of the proposed Cross Florida Barge Canal route. Cross
Florida Barge Canal Restudy Report, U.S. Army Corps of Engineers. 19 p.
1977. (with S. Christman). Impact of proposed phosphate strip mining in the Osceola Na-
tional Forest, Florida. Distributed by Office of Biological Services, USFWS. 431 p.
1977. Compilation of the biology of the American crocodile, Crocodylus acutis. USFWS
Wildlife Species Data Base Series.
1977. Compilation of the biology of the short tailed snake, Stilosoma extenuatum. Ibid.
1977. Compilation of the biology of the Miami black-headed snake, Tantilla oolitica. Ibid.
1977. Compilation of the biology of the West Indian manatee, Trichechus manatus. Ibid.
1977. Feasibility Study: Restoration of the Atlantic Ridley turtle (Lepidochelys kempi) as a
breeding species on the Padre Island National Seashore, Texas. NPS Report, pp. 1-24.
1978. Compilation of the biology of the American alligator, Alligator mississippiensis.
USFWS Wildlife Species Data Base Series.
1978. St. Marks National Wildlife Refuge: Forestry Management and non-game wildlife.
USFWS. 75 p.
1978. (with S. Christman). An annotated bibliography of the fish and wildlife resources of
Galveston Bay, Texas. Distributed by Office of Biological Services, USFWS. 600 p.
1978. (with S. Christman and H. Kochman). A list of the vertebrate species of the world.
USFWS in conjunction with the I.U.C.N. 283 p. (computer print-out).
INVITED PAPERS
1973. The role of natural areas in the preservation of genetic diversity. In Symposium:
Towards a National System of Ecological Preserves-The Genetic, Systematic and
Ecological Basis of Natural Area Preservation. American Soc. of Zoologists, Houston,
Texas, December.
1975. Toward a recovery plan for the American alligator. Big Reptiles Panel, Audubon
Society, New Orleans, La., April.
1976. Biology and conservation of the manatee. Oceans Festival '76. New York Aquarium,
New York.
1977. Herpetological community organization in the Florida "sandhill" habitat. Symposium
on Herpetological Communities, Lawrence, Kansas, August.
1977. Techniques for herpetofaunal community analysis. Ibid.
THE HELICINA UMBONATA COMPLEX
IN THE WEST INDIES
(GASTROPODA, PROSOBRANCHIA, HELICINIDAE)
FRED C. THOMPSON'
ABSTRACT: The umbonata species group is placed in Helicina (s.s.) because of characteristics
of the radula and embryonic shell sculpture. The group consists of five taxa: H. u. umbonata
Shuttleworth, H. u. pitheca, new subspecies, H. liobasis new species, H. dominicensis Pfeiffer,
and H. rhips new species. The latter three species occur on Hispaniola, H. u. umbonata on
Puerto Rico, and tH. u. pitheca on Mona Island.
Foraging specialization and habitat selection are discussed as the basis for evolution
within the Helicinidae.
SUMARIO: El grupo de species umbonata es incluida en el genero Helicina (s. s.) por las
caracteristicas de la radula y la escultura de la concha embri6nica. El grupo consiste de cinco
taxa: H. u. imbonata Shuttleworth, H. u. pitheca n. sub sp., H. liohasis n. sp., 1H. domini-
censis Pfeiffer y H. rhips n. sp., Las tres ultimas species ocurren en la Isla Hispaniola, H. u.
umbonata en Puerto Rico y H. u. pitheca en la Isla Mona.
La especializacion de forage y la selecci6n de habitat son discutidas como bases para la
evoluci6n dentro de la familiar Helicinidae.
TABLE OF CONTENTS
P R O L O G U E .......... .................. .. .. ... ..... . ... 2
ACKNOW LED GEM ENTS . ... ......................................... 2
INTROD UCTION . . .. .................. . . 2
HISTORICAL ACCOUNT .............. ..................................... 2
ADAPTIVE RADIATION WITHIN THE HELICINIDAE .......................... ..... 5
GENERIC RELATIONSHIPS OF THE UMBONATA SPECIES GROUP ............. ...... 5
Characters of the um bonata group ...................................... 7
Comparisons with Helicina .. .. ................. 9
Com prisons w ith Lucidella (s. s.) ..................................... 9
Comparisons with Lucidella (Poenia) ............................. ..... 9
Comparisons with Alcadia ...... . .............. 9
Generic relationships .. .. . ....... .. ............ 10
ECOLOGICAL DEPLOYMENT OF THE UMBONATA SPECIES GROUP ........... ........ .. 11
SPECIFIC R ELATIONSH IPS .................................................. 11
SYSTEMATICS OF THE UMBONATA SPECIES GROUP ......................... 12
Helicina u. umbonata SHUTTLEWORTH .............. ...................12
Helicina umbonata pitheca NEW SUBSPECIES ................... ............... 15
H elicina liobasis NEW SPECIES .. ................................ . .. 17
H elicina dom inicensis PFEIFFER .. ........................................... 19
H elicina rhips NEW SPECIES ................................................. 21
L ITERAT U RE C ITE D ....................................................... 23
'The author is Associate Curator in Malacology, Florida State Museum, University of Florida, Gainesville, FL 32611.
THOMPSON, FRED G. 1982. The Helicina umbonata complex in the West Indies. Bull.
Florida State Mus., Biol. Sci. 28(1): 1-23.
BULLETIN FLORIDA STATE MUSEUM
PROLOGUE
In May 1974 Howard W. Campbell initiated a study to assess the
potential environmental impact of proposed economic developments on
Mona Island. A survey of the land snail fauna was made by Campbell and
the author. The molluscan fauna had been summarized (Clench 1951),
but some taxonomic problems remained to be resolved. One was the
generic relationships of the helicinid land snail known as Lucidella um-
bonata (Shuttleworth). The following year a similar trip was organized
by Campbell to Saona Island and the adjacent mainland of Hispaniola.
Another species of helicinid was discovered that is closely related to um-
bonata. Additional field work in Puerto Rico and Hispaniola yielded
critical material that resolves the systematics of umbonata and related
species. This paper is dedicated to Dr. Howard W. Campbell, who was
instrumental in much of my field work in Middle America and the United
States. He shall be remembered for his companionship on many field trips
and the close friendship I enjoyed since we first met.
ACKNOWLEDGEMENTS
Field work relevant to this study on Puerto Rico and Mona Island was supported, in part,
by contracts 14-16-0088-785 and 85911-1400-906-01 from the U.S. Fish and Wildlife Service.
Field work on Hispaniola was supported, in part, by National Geographic Society Research
Grant No. 1541. I am grateful to the officials of both organizations for the support they have
rendered to me. Figures 8-29 were made on a HITACHI S-415-A Scanning Electron Micro-
scope in the Department of Zoology, University of Florida. The following individuals assisted
me in the field: Steven P. Christman, Ronald I. Crombie, Richard Franz, Roy W. McDiarmid,
and, of course, Howard W. Campbell.
INTRODUCTION
HISTORICAL ACCOUNT
This paper concerns a group of moderate-sized operculate land snails
of the family Helicinidae that occurs on Puerto Rico, Hispaniola (Fig. 7),
and small islands in between. Two are known species. Two others and a
subspecies are described as new. Over the years the known species have
been shuffled among various genera on the basis of limited data on the
shell and operculum. The discovery of the new taxa on Hispaniola and
Mona Island requires a reexamination of the others in order to resolve
questions about systematic relationships. New data based on embryonic
shell sculpture and the radula demonstrate the close relationships among
the species and that their generic affinities are different than previously
thought.
The two known species are Helicina umbonata Shuttleworth and
Helicina dominicensis Pfeiffer 1851. Both were described on the basis of
shell and opercular characteristics. Details of the embryonic shell sculp-
ture were not noted, because such data were not utilized in helicinid sys-
tematics until now. The radula of neither species had been described.
Vol. 28, No. 1
THOMPSON: HELECINA UMBONATA COMPLEX
S 1-2,4-6
equal 2 mm.
____I 1-2,4-6
FIGURES 1-6. -Opercula of Helicina. 1 -H. u. umbonata Sh. (UF 23139); 2- H. u. pitheca n.
ssp. (UF 24064); 3-H. rhips n. sp., paratype (UF 23234); 4-H. dominicensis Pfr. (UF
23549); 5-H. liobasis n. sp., paratype (UF 25264); 6-H. liobasis n. sp. (UF 25266). Scales
equal 2 mm.
H. umbonata is a common Puerto Rican snail that has long been
familiar to malacologists. Its shell was figured twice (Wagner 1911, van
der Schalie 1948), and it is well represented in museum collections. On no
occasion has it been misidentified in the literature, nor has any other
species been mistaken for it. Wagner (1911:355), in his monograph of the
Helicinidae, placed umbonata in Lucidella because of similar shape and
postembryonic shell sculpture to the Jamaican L. aureola (Ferrusac), the
type species of Lucidella. Most subsequent authors followed Wagner.
Baker (1962:17) placed umbonata in Lucidella, subgenus Poenia without
comment. The present study demonstrates the close relationship of um-
bonata to Helicina (s. s.) and not to Lucidella (s. 1.).
H. dominicensis is less well known. It was mentioned only thrice since
its original description. Wagner (1911:340) placed dominicensis in
Alcadia on the basis of its operculum. This study demonstrates that
dominicensis is not related to Alcadia but to Helicina (s. s.).
1982
A T L A N T I C
0 C E A N
BANI SAONA ISLAND
H. liobasis
SH. rhips
SH. dominicensis
C A R I B B E A N S E A
7207 6-90
FIGURE 7.-Distribution of the Helicina umbonata group on Hispaniola.
H
OTI
0
Cj2
H
H
cj~
80
THOMPSON: HELECINA UMBONATA COMPLEX
ADAPTIVE RADIATION WITHIN
THE HELICINIDAE
Helicinids are anatomically conservative (Baker 1925, 1926; Thomp-
son 1968, 1980). Complex reproductive organs and pallial structures
evolved as adaptations for a terrestrial existence, but once they evolved
very little differentiation of these organs and structures occurred with fur-
ther radiation of subfamilies and genera. Subsequent radiation centered
about trophic specialization and habitat selection (Thompson 1980).
Thus, the most significant morphological features reflecting phylogeny
within the family are foraging structures that evolved around trophic
specialization and external structures that are adaptable to habitat selec-
tion. The former involves the radula. The latter involves the shell and
operculum.
Wagner (1911) proposed a classification based on the postembryonic
shell and the operculum. Both organs have intimate contact with the en-
vironment and are readily affected by evolutionary pressures associated
with habitat selection. These pressures fluctuate constantly and are
repetitive in different places and at different times. Wagner's classifica-
tion does not take into account the phenomenon of convergence. Mor-
phological structures highly adaptable to external environmental pres-
sures provide innumerable opportunities for convergence and have
limited application to systematics (Caen 1964).
Shell features produced during embryonic development within the
egg capsule are conservative. They are the consequences of reproductive
specializations centering around nidification. They show little variations
in species or groups of closely related species living in diverse habitats.
The radula of helicinids is also a conservative morphological system
and is a useful indicator of phyletic relationships (Baker 1922). The
radular structure of Helicina is a generalized state for the family and is
relatively primitive in its evolutionary grade. Variations from the
Helicina grade are derived conditions (Thompson 1980). Characteristics
of the radula coupled with embryonic shell features offer great advantage
over other structures in showing supraspecific relationships.
GENERIC RELATIONSHIPS OF THE
umbonata SPECIES GROUP
In order to determine the systematic affinities of the umbonata species
group it was necessary to examine the embryonic sculpture and radula of
those genera to which umbonata and dominicensis have been referred.
The embryonic shells of a large number of tropical American species were
examined. Five are discussed because of their relevance to the umbonata
group: Helicina neritella Lamarck, Lucidella aureola (Ferrussac), L.
(Poenia) depressa (Gray), L. (Poenia) lirata (Pfeiffer) and Alcadia major
BULLETIN FLORIDA STATE MUSEUM
a8
i,.- '
Vol. 28, No. 1
THOMPSON: HELECINA UMBONATA COMPLEX
(Gray). All but lirata are the type species for their respective genera or
subgenera. Baker (1922) gives data on the radula of many species of
helicinids which allow comparisons with the umbonata group. The
radula of two Lucidella are figured in this paper, L. aureola and L. (P.)
lirata.
The helicinid radula contains three fields of teeth, the centrals,
laterals, and marginals. The central and lateral fields are specialized
within different subfamilies and genera for different trophic roles. The
marginal fields show less variation. The central field contains seven longi-
tudinal rows of teeth, the rachidian or R-central bordered on each side by
the A-central, B-central, and C-central (Figs. 8, 9). The lateral field con-
sists of a complex of two partially fused teeth on each side of the central
field, the cusp-bearing comb-lateral and the reinforcing accessory plate
(Figs. 9, 11). The marginal field has a large number of slender teeth with
accuminate cusps (Fig. 10).
CHARACTERISTICS OF THE UMBONATA GROUP.- The umbonata group in-
cludes four species of medium-sized (5-8mm) globose or dome-shaped
snails with about 5-6 whorls. All have rugose spiral sculpture on the
postembryonic whorls (Fig. 12). The protoconch, or embryonic shell, is
about 0.7 mm in diameter and is sculptured with a dense mesh of spirally
arranged pits (Figs. 16-18), which are readily distinguishable at 50X with
a dissecting microscope.
The operculum is yellowish-orange and narrowly auriculate with a
narrow apex hooked to the left (Figs. 1-6). The membranous base extends
beyond the edge of the calcareous plate. The calcareous plate has a thin
and narrow ridge along the columellar margin. The shape of the apex and
the columellar ridge are significant specific characters.
The radula of H. u. umbonata (Fig. 12) and H. u. pitheca (Figs. 8-11)
are described. The central teeth of the radula have projecting cusp-
bearing faces (Figs. 8, 9, 11). The A-central has a large pentagonal-shaped
base with the face raised on a pedicel like a shelf, and projects laterally
with 3-4 cusps. The B-central is erect with a comb-shaped face bearing 4
cusps. The C-central narrows toward the face and bears 3 small cusps.
The comb-lateral has 6-7 heavy subequal cusps. The accessory plate has a
short base and a narrow laterally projecting appendix; the reflection in-
vests only the outer tip of the comb-lateral without covering any cusps
(Fig. 11). The marginal field has about 26 slender, sickle-shaped teeth on
FIGURES 8-15. -Helicinid radulae. 8 -Helicina umbonata pitheca n. ssp., central field (UF
22896), x 375; 9-Same, central and lateral field, x 400; 10-Same, marginal teeth, x
500; 11 Same, lateral field, x 500; 12 H. u. umbonata Pfr., central and lateral field (UF
35133); 13-Lucidella (Poenia) lirata (Pfr.), central and lateral field (UF 22988), x 500;
14-L. aureola (Adams), hemisection (UF 22987), x 175; 15-Same, central and lateral
field, x 375.
BULLETIN FLORIDA STATE MUSEUM
Vol. 28. No. 1
THOMPSON: HELECINA UMBONATA COMPLEX
each side; the innermost teeth have 4 heavy cusps. The outermost teeth
have 9-10 very delicate cusps (Fig. 10).
COMPARISONS WITH HELICINA. The protoconch of H. neritella is
about 0.9 mm in diameter, and has a mesh of spirally arranged pits (Fig.
19) similar to umbonata. The radula is described by Baker (1922:52-53;
pl. 3, fig. 6; pl. 4, fig. 17). H. neritella and umbonata are alike. They dif-
fer only in the number of cusps on the comb-lateral. H. neritella has eight
cusps.
COMPARISONS WITH LUCIDELLA (s. s.). -The protoconch of L. aureola
is smaller than umbonata (0.5 mm). It is sculptured with very minute pits
that are spaced farther apart than their diameters (Figs. 24, 25). The in-
tervening space does not form a mesh, and the pits are not arranged in
discrete spiral series. The radula, specialized for scraping fine food par-
ticles, is figured in Baker (1922) and in this paper (Figs. 4, 5). The
A-central is long, narrow, and oblique; its cutting edge is long, narrow,
and lacks denticles. The B-central is shorter and has about 3 rounded
knob-like vestigial cusps along the outer edge of the blade. The C-central
is spatulate with about 3 poorly defined cusps. The comb lateral has 9
nearly equal cusps. The accessory plate has a very long base bearing a
short lateral appendix. The marginals have 2 cusps on the innermost teeth
and 4 on the outermost.
COMPARISONS WITH LUCIDELLA (POENIA). -The protoconch of L. (P.)
depressus (Fig. 29) and L. (P.) lirata (Figs. 27, 28) are illustrated. The
subgenus is similar to Lucidella (s. s.) in the small size of the protoconch
(0.4 mm). The pitted sculpture is sparser and finer than in Lucidella and
is most conspicuous along the periphery. The radula is described and
figured by Baker (1922:55; pl. 3, fig. 5; pl. 5, fig. 21) and here (Fig. 13).
As in Lucidella (s. s.) it is specialized for scraping fine substances. The
A-central is narrow, elongate, and oblique with a hood-shaped face. The
B-central is comb-shaped with about 6 small cusps. The comb-lateral has
about 7 cusps. The accessory plate has a long base and a short triangular
appendix (Baker, 1922: pl. 5, fig. 21). The inner marginal teeth have 3
cusps. The outer marginals have 4 cusps.
COMPARISONS WITH ALCADIA. The protoconch of A. major (Fig. 26) is
large (1.0 mm) and is sculptured with coarse, irregularly spaced radial
threads. The inner curvature bears a few irregular longitudinal striations.
FIGURES 16-23. -Electron micrographs of Helicina shells. 16- H. u. umbonata Sh., em-
bryonic sculpture (UF 25249), x 60; 17 H. liobasis n. sp., embryonic sculpture (UF 25247),
x 50; 18 -H. rhips n. sp., embryonic sculpture (UF 22989), x 50; 19 -H. neritella, em-
bryonic sculpture (UF 22992), x 50; 20 -H. u. umbonata Sh., periostracal bristles on last
whorl (UF 25249), x 40; 21 -Same showing postembryonic sculpture and descent of aper-
ture, x 6.5; 22- H. liobasis n. sp., showing smooth base (UF 25254), x 6.5; 23- H. u. um-
bonata Sh., showing sculpture on base (UF 25249), x 5.5.
1982
BULLETIN FLORIDA STATE MUSEUM
FIGUREs 24-29.-Electron micrographs of helicinid embryonic sculpture. 24-Lucidella
aureola (Adams) (UF 22991), x 50; 25-Same, x 250; 26-Alcadia major (Gray) (UF
22980a), x 25; 27 -Lucidella (Poenia) lirata (Pfr.) (UF 22990), x 50; 28-Same, x 250;
29 -Lucidella (Poenia) depressa (Gray) (UF 35167), x 250.
Pitted sculpture is absent. The radula of A. major is like that of the
Vianinae, not like the Helicininae. Other generic units placed in Alcadia
as subgenera have Helicininae radula, but a protoconch similar to
Alcadia.
GENERIC RELATIONSHIPS. -The umbonata species group is placed in
Helicina (s. s.) because of similarities in embryonic shell sculpture and
Vol. 28, No. 1
THOMPSON: HELECINA UMBONATA COMPLEX
radular tooth structures. Like Helicina neritella the radula is specialized
for gouging coarse vegetal food particles whereas the radula of Lucidella
is specialized for scraping fine substances. Relationships to Lucidella
(s. 1.) and Alcadia are not tenable in light of the data presented above.
The umbonata group is unique among West Indian Helicina because of
the rugose postembryonic spiral sculpture. Some mainland subgenera
placed in Helicina have rugose spiral cords, but their embryonic shell
sculptures differ. The status of the subgeneric groups associated with
Helicina needs to be reexamined.
ECOLOGICAL DEPLOYMENT OF THE
umbonata SPECIES GRouP
The umbonata group lives in submesic and xeric habitats, although
occasionally populations of H. u. umbonata occur in mesic stations. The
species are adapted to exist in dry habitats through modification of the
coloration, sculpture, and microhabitat selection.
Two species, H. umbonata and H. liobasis are arboreal, foraging and
aestivating on trees and shrubs close to the ground. The shells are whitish
with conspicuously colored bands and blotches. The markings disrupt the
appearance of the light colored shell on a variegated background. Live
snails on shrubs are inconspicuous because of the background produced
by limbs, vines, leaves, and flowers. The rugose spiral sculpture adds to
the visual disruption. Both species also have periostracal bristles (Fig. 20)
which usually are worn from adult shells. Presumably these bristles add to
visual disruption of the juveniles.
Two other species, H. dominicensis and H. rhips, have darker colored
shells in which the color pattern is not as distinct. Dark-colored, poorly
patterned shells are typical of ground-dwelling land snails. These two
species live under terrestrial bromeliads and Agave sp. This habitat is pro-
tective from desiccation by rosettes of leaves around the base of the plants
tightly oppressed against the ground. Presumably the snails forage on but
not upon the plants. The matted sculpture of the postembryonic shell
adds to the inconspicuous appearance of the snail among dead plant
debris. The two species are ecologically segregated further. H. domini-
censis lives under Bromelia pinguin on alluvial substrates. H. rhips lives
under Agave cf. americana on limestone substrates.
SPECIFIC RELATIONSHIPS
The interspecific relationships of the umbonata group are indicated by
the shell and opercular structures, and are related to the ecological
deployment of the four species. Two subgroups are represented. One con-
sists of two arboreal, brightly colored snails with strong spiral sculpture,
1982
BULLETIN FLORIDA STATE MUSEUM
H. umbonata and H. liobasis. The other consists of two terrestrial, dull
colored snails with matted sculpture, H. dominicensis and H. rhips. The
first species, H. umbonata, occurs east of Hispaniola. The other three
occur on Hispaniola. The Hispaniolan species are alike in having an
auriculate-shaped operculum with a pointed apex strongly hooked to the
left (Figs. 3-6). H. umbonata has a blunt apex weakly hooked to the left
(Figs. 1-2).
Two interpretations of phylogeny can be made. One is that H. um-
bonata and a Hispaniolan prototype diverged morphologically through
modification of the operculum and became segregated geographically on
Puerto Rico and Hispaniola. Subsequently the Hispaniolan prototype
diverged in sculpture, coloration, and habitat deployment to form one
lineage leading to H. liobasis and the other leading to H. dominicensis
and H. rhips. The latter lineage diverged further through adaptive selec-
tion for different plant associations. This interpretation assumes that
within the group an arboreal habitat is primitive and a terrestrial habitat
is secondarily derived.
A second interpretation of phylogeny is that an Hispaniolan prototype
diverged to give rise to two lineages through adaptive selection and
modifications of color and sculpture. An arboreal lineage diverged fur-
ther through minor modifications of sculpture and opercular shape
because of geographic isolation, with H. liobasis on Hispaniola and H.
umbonata on Puerto Rico and nearby islands. A terrestrial lineage leading
to H. dominicensis and H. rhips diverged further through adaptive selec-
tion for different plant associations. This interpretation permits the
Hispaniolan prototype to be either arboreal or terrestrial. Usually among
land snails an arboreal existence is a derived condition. A terrestrial pro-
totype is favored, because within the Helicinidae the vast majority of the
species are terrestrial. The second interpretation of phylogeny is accepted
because it requires the fewest assumptions and is consistent with other
patterns of ecological deployment in land snails.
SYSTEMATICS OF THE UMBONATA SPECIES GROUP
Helicina u. umbonata SHUTTLEWORTH
Helicina umbonata Shuttleworth 1854; Diagnosen Neuer Mollusken,
6:93. -Pfeiffer 1858; Monogr. Pneumo. Viv.: 187. -Dall and Simp-
son 1901; Bull. U.S. Fish Comm. (1900) 1:447.
Lucidella umbonata (Shuttleworth), Wagner 1911; in Martini and Chem-
nitz Syst. Conch.-Cab. (Helicinidae): 340; pl. 67:21-24. -van der
Schalie 1948; Misc. Publ. Mus. Zool. Univ. Mich. (70):23; pl. 1, fig.
6; map 3. -Boss and Jacobson 1974; Occ. Pap. Moll. 4:37.
Lucidella (Poenia) umbonata (Shuttleworth) Baker 1962; Nautilus 76:17.
Vol. 28, No. 1
THOMPSON: HELECINA UMBONATA COMPLEX
FIGUREs 30-35. -Helicina u. umbonata Sh. 30-32, 35-10 km SE of Guanica, Puerto Rico
(UF 35139); 33. 34-Cerro de las Cuevas, SE of Juana Diaz, Puerto Rico, 450 m alt. (UF
25008).
SHELL (FIGS. 30-35).-Medium-sized, about 5-6 mm in diameter.
Depressed dome-shaped, about 0.64-0.75 times as high as wide. Whorls
5.0-5.4. Periphery angulate or carinate; angulation usually accentuated
by rugose beaded spiral cords along periphery; occasional specimens
rounded peripherally with rugose cords producing angulation. Base of
shell flattened around umbilical callus. Protoconch with relatively rugose
pits spirally arranged (Fig. 16). Sculpture on subsequent whorls consisting
of relatively rugose spiral cords and weaker, irregular axial striations
(Fig. 21). The cords on the base become weaker, but are continuous into
the aperture (Figs. 23, 25); apical whorls with about 3-10 cords. Cords
separated by deep, sharp grooves that are about the same depth as the
suture and obscure the suture. Aperture 0.71-0.90 times as high as wide.
Outer lip sharp-edged, but internally thickened and very weakly reflected
laterally and basally. Columella arched forward, terminating as a low,
poorly defined denticle. Color pattern almost always poorly defined. A
peripheral band occurs very infrequently. More often the spire contains
irregular yellow to rust-colored blotches, spots, and streaks that tend to be
BULLETIN FLORIDA STATE MUSEUM
spirally arranged. Interior of aperture light orange. The ground color
varies from pinkish-white to light brown. Specimens from mesic habitats
are darker with a poorly defined pattern of lighter yellow markings on the
spire (Fig. 33, 34). Specimens from xeric habitats tend to be lighter with
orange- or rust-colored markings (Fig. 30-32).
OPERCULUM (FIG. 1). -Calcareous plate broadly auriculate in shape;
apex bluntly rounded, slightly hooked to left; widest at about the middle;
evenly rounded below. Columellar margin with a thick, sharp ridge that
is highest below and becomes lower toward the apex. Outer surface
covered with minute granules.
Measurements in mm based on two population samples, one from a
xeric habitat and the other from a mesic habitat, are representative.
XERIC MESIC
(UF 35139, 25 specimens) (UF 35140, 23 specimens)
Shell width 4.4-5.5 4.9-6.3
Shell height 3.2-3.9 3.2-4.5
Aperture width 1.5-1.9 1.8-2.3
Aperture height 1.9-2.4 2.2-2.8
Whorls 5.0-5.4 4.9-5.4
Aside from slight differences in size and color associated with habitat,
no other significant variations are observed.
DISTRIBUTION. This subspecies is confined to a narrow coastal belt of
calcareous terrain along the western half of Puerto Rico. The specimens
examined were collected from within the area reported by van der Schalie
(1948:23; map 3) and add little distribution information, except for the
specimens from Cerro de las Cuevas. This station is about 50 km east of
Ponce. The snail is found most commonly in xeric habitats, but occa-
sionally is found in mesic forests. It is arboreal and aestivates on shrubs,
vines, and herbaceous plants within a few meters of the ground. During
wet weather it also climbs on limestone rocks (Baker 1962:62).
SPECIMENS EXAMINED. -PUERTO RICO: 2 km SSE Guanica (UF
35138); 7 km SE Guanica (UF 35141); 10 km SE Guanica (UF 25249-50;
35139); south slope Cerro de las Cuevas, 3.5 km N, 4.7 km E Juana Diaz,
450 m (UF 25008, 25251, 35138, 35140, 35141, 35165).
REMARKS. Helicina umbonata is most closely related to H. liobasis,
which occurs on the eastern end of Hispaniola and which is geographically
most proximal to the two subspecies of H. umbonata. H. umbonata and
H. liobasis have similar spiral sculpture on the apex of the shell. The spiral
cords are not hatched by conspicuous axial threads, thus differing from
the sculpture of H. dominicensis and H. rhips. H. umbonata differs from
H. liobasis by its sculptured base, depressed shape, relatively wide aper-
Vol. 28, No. 1
THOMPSON: HELECINA UMBONATA COMPLEX
ture, and the shape of its operculum. These characters are discussed in
greater detail under H. liobasis. The enlarged beaded peripheral cords
characteristic of H. u. umbonata are unique within the species group. Its
operculum differs from Hispaniolan species in that its widest point lies
above the middle, the columellar ridge is thicker and more rounded, and
the apex is rounded.
H. umbonata includes two subspecies. The nominate subspecies is
confined to Puerto Rico. The other is confined to Mona Island and Monito
Island, which lie in Mona Passage between Puerto Rico and Hispaniola.
Differences between the two subspecies are discussed below.
Helicina umbonata pitheca NEW SUBSPECIES
Lucidella umbonata (Shuttleworth) Clench 1950; Jour. de Conchyl.
80:271 [Mona Island]. -Thompson 1976; Nautilus 90:152 [Monito
Island].
SHELL (FIGS. 36-43). -Medium-sized, about 5.1-6.5 mm in diameter.
Depressed conical, about 0.7-0.8 times as high as wide; spire low, convex
with a protruding apex, about 0.42-0.56 times height of shell. Whorls
5.2-5.7; protoconch with 1.3-1.5 whorls, sculptured as in H. u. um-
bonata. Following whorls sculptured with low spiral lirations that be-
come obsolete on last whorl; last 1/4-1/2 whorl nearly smooth except for
incremental striations. Suture distinct, not obscured by spiral sculpture.
Spiral lirations on base continuing into aperture, diminishing near
umbilical callus, which is granular and bordered behind by a strong spiral
ridge that continues to base of columella as a slightly projecting, blunt
denticle. Body whorl rounded at periphery and inflated on base. Aperture
broadly auriculate, almost as high as wide; width of aperture 0.41-0.45
times width of shell; height 0.44-0.58 times height of shell. Outer lip
sharp-edged, weakly reflected, with a low internal callus. Aperture slowly
descending below periphery along last quarter whorl. Color yellowish-
white with an orange supraperipheral spiral band that is wavy along its
dorsal edge and may be broken into a spiral row of spots and blotches,
especially on upper whorls. Occasional specimens white without orange
markings. Protoconch yellow. Interior of aperture orange. Outer lip and
columella white.
OPERCULUM (FIG. 2). -Similar to H. u. umbonata except that the
columellar ridge is lower, wider, and not as sharp-edged, and the widest
point is above the midline.
Measurements in mm of 12 specimens from a mesic habitat, selected
from the type lot, are given below holotypee in parentheses). Measure-
ments of 10 specimens from a xeric habitat are also given. These two
BULLETIN FLORIDA STATE MUSEUM
FIGUREs 36-43. -Helicina umbonata pitheca n. ssp. 36-39 -Holotype (UF 35168); 40-43-
Paratypes (UF 35169).
samples encompass the variation usually encountered, except for occa-
sional gerontic specimens.
MESIC (UF 35169) XERIC (UF 24964)
Shell width 5.5-6.5 (6.3) 5.1-5.8
Shell height 4.0-5.0 (4.8) 3.8-4.3
Aperture width 2.0-2.4 (2.7) 2.0-2.3
Aperture height 2.3-2.9 (2.2) 2.1-2.5
Whorls 5.3-5.7 (5.7) 5.2-5.5
TYPE LOCALITY. -MONA ISLAND, Sardinero. HOLOTYPE: (UF
35168); collected 21 May 1974 by Howard W. Campbell and Fred G.
Thompson. PARATYPES: UF 35169(89); same data as the holotype.
DISTMBUTION. Endemic on Mona Island (UF collections, 17 stations)
and nearby Monito Island (Thompson, 1976; UF 35158). They appear to
be ubiquitous on Mona Island. Live snails usually are found on plants
within a few meters of the ground. The snail aestivates on trees, shrubs,
and herbaceous plants.
The specimens from Monito Island are somewhat intermediate in
character between umbonata and pitheca. They have typical pitheca
sculpture, color, and descent of the aperture.
REMARKS -This is a well-differentiated subspecies that is readily dis-
tinguished from H. u. umbonata by its color, sculpture, and shape. All
Vol. 28, No. 1
THOMPSON: HELECINA UMBONATA COMPLEX
samples of H. u. pitheca are dominated by specimens with a distinct
supraperipheral orange band. Some specimens have a spiral series of spots
or lack markings on the last whorl. H. u. umbonata has a poorly defined
color pattern that seldom forms a peripheral band, although the blotches
and spots tend to be spirally arranged. In H. u. umbonata the spiral lira-
tions are quite heavy and obscure the suture, whereas H. u. pitheca has
weaker spiral lirations that do not obscure the suture. The H. u. umbonata
lirations are stronger along the periphery, accentuating the peripheral
angle, and the sculpture continues undiminished to the aperture. H. u.
pitheca has a more inflated body whorl with a rounded periphery and a
weakly inflated base, and the aperture descends slightly below the
periphery. In H. u. umbonata the aperture descends farther along the last
1/10 whorl (Figs. 21, 43).
ETYMOLOGY. -From the Greek pithekos, meaning monkey, in this
case a patronym for Mona Island; mona, from modern Spanish, meaning
monkey.
Helicina liobasis NEW SPECIES
SHELL (FIGs. 44-49).-Medium-sized, about 5 mm in width. Dome-
shaped, about 0.78-0.90 times as high as wide; spire convex, raised, about
0.42-0.53 times height of shell. Whorls 4.6-5.1 (5.1 in holotype). Last
whorl rounded, descending to aperture below peripheral band. Proto-
conch with 1.1 whorls; sculptured with fine pits arranged in spiral series
(Fig. 17). Subsequent whorls sculptured with about 3-6 low, wide spiral
cords that lie above the periphery; cords crossed by variable axial stria-
tions which become obsolete on the periphery. Base smooth, or nearly so
(Figs. 22, 45); some specimens have occasional faint spiral and axial stria-
tions. Whorls inflated; last 1/4 whorl descending below periphery to aper-
ture. Umbilical callus blue-gray in live specimens, finely granular. Color
pattern consisting of rust-colored spiral bands on a cream-white back-
ground. Lower band distinct, lying along periphery. Upper band poorly
defined, lying along middle of shoulder and corresponding part of earlier
whorls; separated from suture by a narrow white zone. In some specimens
upper band is broken into several narrow bands that lie in grooves be-
tween spiral cords. Frequently upper band has superimposed rusty
blotches that extend to suture. Protoconch yellow. Lip and columella
white. Interior of aperture light orange. Aperture half-moon shaped,
oblique and sigmoid in lateral profile; about 0.47-0.58 times height of
shell and 0.42-0.46 times width of shell. Lip sharp-edged; internally
thickened above and increasingly thicker to base. Outer lip very slightly
reflected, bordered behind by a very shallow groove. Columella oblique
and curved forward, ending in a very weak basal denticle that is con-
nected to posterior edge of umbilical callus by a weak buttress.
1982
BULLETIN FLORIDA STATE MUSEUM
FIGUREs 44-49. -Helicina liobasi, n. sp. 44-46 -Holotype (UF 25263); 47-49 Paratypes
(UF 25264).
OPERCULUM (FIGs. 5-6). -Calcareous plate widest at middle; base
bluntly rounded. Apex pointed, strongly hooked to left. Columellar ridge
sharp, narrow, highest near base, not reflected laterally.
Measurements in mm of 19 specimens from the type lot are as follow
holotypee in parentheses): shell width 4.8-5.4 (5.2), shell height 3.8-4.7
(4.3), aperture width 1.9-2.4 (2.4), aperture height 2.0-2.5 (2.4).
The shell of H. liobasis shows little variation aside from size and
relative height. The type lot nearly encompasses the full range of varia-
tion found in other populations. The series from 4 km SE Punta Cana con-
tains a gerontic shell 6.6 mm in diameter. The series from Punta Palmillos
has a preponderance of relatively depressed shells, but the total variation
within this series is within the variation of the type lot.
DISTRIBUTION (FIG. 7). -Found along the extreme eastern tip of
Hispaniola and offshore islands in Departamento ( = Dpto.) de Altagracia,
Dominican Republic. It appears to be restricted to a narrow zone along
the coast. Numerous field collections were made at localities inland from
Vol. 28, No. 1
THOMPSON: HELECINA UMBONATA COMPLEX
El Macao, Punta Cana, and Boca de Yuma. No specimens were found ex-
cept at coastal localities.
H. liobasis, like H. umbonata, is arboreal. Live specimens were col-
lected on Saona Island, Punta Palmillos, and south of El Macao crawling
on coconut palms and shrubs. Other live specimens were found
aestivating on tree trunks and shrub stems.
TYPE LOCALITY. -DOMINICAN REPUBLIC, ALTAGRACIA DPro.,
north coast of Isla Saona opposite Isla Catalinita. The area consists of a
low submesic open forest on a limestone terrain with occasional small
clearings. The lower vegetation was heavily browsed by goats. Snail shells
occurred under limestone slabs in forest floor litter. Occasional live
specimens were found on tree trunks. HOLOTYPE: UF 25265; collected
18 February 1975 by Howard W. Campbell and Fred G. Thompson.
PARATYPES: UF 25264 (59), UF 25247 (3), UF 25254 (1); same data as
the holotype.
OTHER LOCALITIES.-DOMINICAN REPUBLIC: ALTAGRACIA DPTo.:
1 km SW El Macao (UF 25268, 8); Punta Cana (UF 25276, 1); 4 km SE
Punta Cana (UF 25269, 2); Punta Palmillas (UF 25266, 23); Isla
Catalinita (UF 25267, 7).
REMARKS. Helicina liobasis is similar to H. dominicensis and H. rhips
in its dome-shaped shell, its narrow pointed operculum, and its color pat-
tern. It differs from H. umbonata in these as well as other characteristics,
which are discussed under that species. H. liobasis differs from the His-
paniolan species by its weaker columella denticle, its sculpture, and its
peristome. The sculpture of H. liobasis consists of much wider spiral cords
and lacks the axial threads that give a matted pattern to the shells of H.
dominicensis and H. rhips. The outer peristome of H. liobasis is hardly
reflected. What little reflection occurs is accentuated by a weak postlabial
groove. In other Hispaniolan species, the outer and basal lips are con-
spicuously reflected.
ETYMOLOGY. -Liobasis, from the Greek, leios meaning smooth and
basis, meaning base, alluding to the lack of sculpture on the base of the
shell.
Helicina dominicensis PFEIFFER
Helicina dominicensis Pfeiffer 1851; Proc. Zool. Soc. (London): 149.-
1852; Monogr. Pneumon. Viv. 1:352. -Crosse 1891; Jour. de Con-
chyl. 39:185.
Alcadia (Analcadia) dominicensis (Pfeiffer) Wagner 1911; in Martini and
Chemnitz, Syst. Conchyl. Cab., Helicinidae: 355; pl. 68, figs. 1-3.
SHELL (FIGs. 50-53). -Medium-sized, about 5-6 mm wide; conico-
1982
BULLETIN FLORIDA STATE MUSEUM
FIGUREs 50-53. -Helicina dominicensis Pfr. (UF 23183).
globose, about 0.85-0.92 times as high as wide. Spire raised, straight-
sided, forming an obtuse cone with a blunt apex. Body whorl rotund,
uniformly rounded. Suture deeply impressed. Whorls 5.0-5.3. Protoconch
with about 1.3 whorls, typically sculptured for group. Whorls regularly
increasing in size, shouldered along suture, regularly descending; last
1/10 whorl descending slightly below periphery to aperture. Whorls
sculptured with numerous low spiral cords that are crossed by finer, lower
radial threads forming a matted surface; spiral cords slightly finer on base
of shell than above. Umbilical callus small, about 1/6 diameter of shell,
white, granular, with a weak indentation at base of columella. Ground
color yellowish-white with a broad spiral dull yellow zone around
periphery and on spire. Protoconch light yellow. Interior of aperture light
orange. Peristome white. Aperture half-moon shaped; about as high as
wide; about 0.49-0.52 times height of shell. Outer lip straight-edged
above, weakly reflected laterally and below. Columella concave internally
and arched forward below, forming a blunt denticle.
OPERCULUM (FIG. 4). -Calcareous plate acutely pointed above and
bluntly rounded below. Apex sharp, moderately hooked to left. Widest
point of plate at middle. Columellar ridge thick and sharp-crested below,
very low along apical half; ridge not projecting sideways beyond margin
of basal plate.
Measurements in mm based upon 11 adults are as follow: shell width
5.2-5.8, shell height 4.6-5.3, aperture width 2.4-2.7, aperture height
2.4-2.7.
DISTRIBUTION (FIG. 7). -Endemic to Hispaniola where it has a
moderate distribution in the Azua basin and occurs as far east as the area
near Bani. I found it at only five localities. Its paucity in collections may
be due to its apparent habitat preference. I found it only in clusters of the
terrestrial bromeliad, Bromelia pinguin, which grows in dense thickets on
alluvium in xeric areas, and it was not found on limestone outcrops within
Vol. 28, No. 1
THOMPSON: HELECINA UMBONATA COMPLEX
the same area. This habitat was infrequently searched for snails because
relatively few species occur there compared to the much richer fauna
associated with limestone. Subsequent search in the bromeliad habitat
may show that the species is more widely distributed.
TYPE LOCALITY.-Pfeiffer (1851:149) described this species from
specimens collected by August Salle, and gave Haiti ( = Hispaniola) as the
type locality. Crosse (1891:39) recorded it from Las Charcas, Dominican
Republic, also from specimens collected by Salle who visited there in
1849. Apparently Sall6 did not collect it elsewhere, nor has it been
reported from elsewhere. Therefore, the type locality is restricted herein
to Las Charcas, Departamento de Azua, Dominican Republic.
SPECIMENS EXAMINED. -DOMINICAN REPUBLIC; AzUA DTro: 1 km
SE Peralto, 510 m (UF 23549.3); 1 km N Estebania, 170 m (UF 23547.1);
2 km N Las Charcas, 80 m (UF 28183.8); 6 km ESE Las Charcas (UF
23063.3). PERAVIA DPTo: 6 km WNW Bani (UF 23548.2).
REMARKS. H. dominicensis is very similar to H. rhips in shape, color,
and sculpture. The two differ in size, whorl count, and aspects of the
operculum, as is discussed under the latter species.
Helicina rhips NEW SPECIES
SHELL (FIGs. 54-57). -Large, about 7-8 mm in diameter. Conico-
globose, about 0.84-0.92 times as high as wide; spiral convex, raised
about 0.45-0.50 height of shell. Whorls 5.5-5.9 in adults. Protoconch con-
sisting of about 1.1 whorls that are sculptured uniformly with small pits
arranged in spiral series (Fig. 18). Subsequent whorls with relatively
strong spiral chords crossed by finer axial threads creating a latticed pat-
tern; about 3-4 chords on second whorl; about 20-25 chords between
suture and lip on last whorl. Chords uniform in size over last whorl. Um-
bilical callus coarsely granular, about 1/5 diameter of base, light yellow.
Color of shell dull cream white to light pink with diffuse light yellow to
light orange bands on body whorl; upper band lying about 1/3 of distance
from suture to lip insertion; lower band lying just above periphery and lip
insertion. Bands coalescing on upper whorls forming a brighter yellow or
orange apex. Peristome and columella white. Interior of aperture dark
yellow to orange. Aperture semilunar, about as high as wide, about
0.50-0.55 times height of shell and about 0.44-0.48 times width of shell;
descending slightly and inserting on body whorl just below periphery. Lip
moderately reflected at periphery and along base, not reflected above.
Lip oblique and sigmoid in lateral profile. Columella arched and curved
forward, ending as a blunt denticle that is connected to posterior edge of
umbilical callus by a narrow buttress.
OPERCULUM (FIG. 3). Calcareous plate acutely pointed above, bluntly
1982
BULLETIN FLORIDA STATE MUSEUM
FIGUREs 54-57. -Helicina rhips n. sp. 54-56 -Holotype (UF 23064); 57-Paratype (UF
23234).
rounded below, and widest at midpoint. Apex more strongly hooked to
left than in H. dominicensis. Columellar ridge sharp-edged, and project-
ing sideways beyond margin of basal plate.
Measurements in mm based on 25 mature specimens are as follow
holotypee in parenthesis): shell width 7.0-8.4 (8.2), shell height 6.2-7.5
(7.0), aperture width 3.3-4.0 (3.7), aperture height 3.3-3.8 (3.7).
DISTRIBUTION (FIG. 7). -Endemic to Hispaniola where it is known
only from a low limestone mountain ridge that extends from Bani north-
ward for about 8 km. The ridge is covered with a dense submesic thorn
forest. The snail was not found in adjacent areas where numerous collec-
tions were made. No live specimens were found. Dead shells were always
associated with Agave cf. americana.
TYPE LOCALITY. -DOMINICAN REPUBLIC: PERAVIA DPTo., 6 km
north of Bani, 100 m elevation. The type locality is along the east side of a
limestone mountain ridge about 0.5 km west of the Rio Bani. The area at
the base of the ridge consists of a tallus slope of huge limestone slabs and
boulders. Specimens were found only along the base of the rock wall
forming the ridge, not on the tallus slope. HOLOTYPE: UF 23064, col-
lected 23 January 1980 by Fred G. Thompson. PARATYPES: UF 23234
(23), same data as the holotype; UF 25246 (14), 2 km NE Bani, 100 m; UF
23190 (2), 1 km N Bani, 75 m.
REMARKS.-Helicina rhips is closely related to H. dominicensis and ap-
pears to be a giant replica of the latter. The two species differ in size,
whorl-count, and the structure of the opercula. They are considered
species for the following reasons. In other species of helicinids such size
discrepancies do not occur, even over large geographic areas. In addition
to its larger size, H. rhips has about a half whorl more than does H.
dominicensis. Usually in helicinids subspecies differences in size are not
accompanied by such differences in whorl count. The reflected col-
umellar ridge on the operculum is unique within the umbonata species
group, and is a degree of differentiation not commonly found among
closely-related species. Finally, no trends toward intergradation occur.
Vol. 28, No. 1
THOMPSON: HELECINA UMBONATA COMPLEX
H. dominicensis occupies a relatively large geographic area within the
Azua basin and is monomorphic in size, whorl-count, and opercular
structure throughout the area.
ETYMOLOGY. -Rhips, from the Greek rhips, a wickerwork, net, or
screen in allusion to the sculpture of the shell.
LITERATURE CITED
Baker, H. B. 1922. Notes on the radula of the Helicinidae. Proc. Acad. Nat. Sci. Philadelphia
76: 29-67; pls. 3-7.
1925. Anatomy of Hendersonia (Helicina occulta): A primitive land snail. Proc.
Acad. Nat. Sci. Philadelphia 77: 273-303: pls. 7-10.
1926. Anatomical notes on American Helicinidae. Proc. Acad. Nat. Sci. Phila-
delphia 78: 35-56; pl. 4-8.
. . 1962. Puerto Rican land operculates. Nautilus 76: 16-22.
Caen, A. J. 1964. The perfection of animals. In J. E. Carthy and E. L. Duddington (eds.).
Viewpoints in Biology (3): 36-63; Butterworths, London.
Clench, Win. J. 1951. Land shells of Mona Island. Puerto Rico. Jour. de Conchyl. 90:
269-276.
Thompson, F. G. 1978. Ceochasma, a remarkable new land snail from Colima, Mexico
(Gastropoda, Prosobranchia, Helicinidae). Proc. Biol. Soc. Washington 81: 45-52.
1980. Proserpinoid land snails and their relationships within the Archeogastro-
poda. Malacologia 20: 1-33.
Van der Schalie, H. 1948. The land and freshwater mollusks of Puerto Rico. Misc. Pub. Mus.
Zool. Univ. Michigan (70): 1-134; pls. 1-14.
Wagner, A. 1911. Die Familie der Helicinidae: In Martini und Chemnitz, Syst. Conchyl.-
Cab., Nurnberg: 1-391; pls. 1-70.
MICROTUS PENNSYLVANICUS
(RODENTIA: MURIDAE) IN FLORIDA:
A PLEISTOCENE RELICT IN A COASTAL SALTMARSH
CHARLES A. WOODS, WILLIAM POST, AND
C. WILLIAM KILPATRICK1
ABSTRACT: licrottus pcntisylvanoicus exists as a disjunct relictual population in a tidal salt
marsh located on Waccasassa Bay on the Florida Gulf coast. This form, described as a new
subspecies, inhabits Distichlis flats and other areas of low vegetation in the salt marsh, while
Oryzomys is more common in areas of Juncus. The saltmarsh vole was less abundant than
Oryzomys during the study period. In comparison with other forms of Microtuis pennsyl-
canicus, the saltmarsh vole is larger in body size, darker in coloration, and has a larger hind
foot. The current distribution of the new subspecies of Microturs appears to be relictual from a
more widespread range in the Gulf Coastal Plain during the last 10,000 years. The range of
the vole presumably was restricted by changes in vegetation and rising sea levels during the
last 8000 years. The vole is in a grassland habitat isolated by unsuitable forest habitat and is
500 km south of the nearest population of Al. p. pennsylvanicus in Georgia. Analysis of blood
proteins indicates a lack of genetic diversity. a condition that prevails in other insular popula-
tions of Microtrus pennsylcanicusi. The saltmarsh vole survives in low numbers under very
harsh ecological conditions and is vulnerable to extirpation by natural disasters associated
with wind and water.
SUMA1UO: Microtuns pennsilvanicus existe como una poblacion relicta disgregada en el pantano
salino de marea ubicado en la Bahia de Waccasassa, en la costa del Golfo de Florida. Esta for-
ma, descrita como una nueva subespecie, habitat pastizales de Distichlis y otras areas de vege-
tacion baja en el pantano salino, mientras que Oryzoinys es mas comun en areas de Juncus. El
raton del pantano salino fue menos abundante que Oryzomys durante el period de studio.
En comparacion a otras formas de Mlicrotus pennsylcanicus, el rat6n del pantano salino es de
cuerpo mas grande, coloracion mas oscura y pata posterior mas grande. La present distribu-
cion de la nueva subespecie de Microtus parece ser el relicto de una poblaci6n de ratones dis-
tribuida en area mas amplia en las planicies costeras del Golfo durante los ultimos 10,000
anos. El area de distribucion de este raton presumiblemente ha sido restringida por cambios
en la vegetaci6n y niveles ascendentes del mar durante los ultimos 8,000 anos. Este raton se
encuentra en on habitat de pastizal aislado por bosque no apropiado como habitat y situado a
500 kmi. al Sur de la poblacion mais cercana de M. p. pentnsylvanicus en Georgia. El anilisis de
proteinas de la sangre indica la falta de diversidad genetic, condition que prevalece en otras
poblaciones insulares de Microtus pennsylvanicuis. El rat6n del pantano salino sobrevive en
limitados numieros bajo condiciones ecologicas muy desfavorables y es vulnerable a ser elimi-
nado por desastres naturales asociados a vientos y agua.
'Dr \\ niods is (Curaitor ... ..1 a ind Chairiman of tihe Dl.partlment of Natural Sc'lencs., Floridali State MSiiscuni,
i'n] ritt oil f FInrida. ( .. II 32611: Dr Pont is Adjunct Curator in (....1, I i Florida State Museii m. D)r.
Knliatrti k is Ass oiiati Proftss.or ofl Zooiloni., Uniiersnt\ of V5\'tcriln Burlirngton, I, ,
WOODS, C. A., W. POST, AND C. W. KILPATRICK. 1982. Microtus pein.sylvanicus
(Rodentia: Muridae) in Florida: A Pleistocene Relict In a Coastal Saltinarsh. Bull. Florida
State Mus., Biol. Sci. 28(2):25-52.
BULLETIN FLORIDA STATE MUSEUM
TABLE OF CONTENTS
INTRODUCTION . . . ....... ..... ...... 26
ACKNOW LEDGEM ENTS ........ . .................. .. . . ..... 27
MATERIALS AND METHODS . .. ................. ... .. ...... 27
ST U D Y A R E A ......... ..................................................... 28
R E SU L T S ........... . .. . ...... .. . ... .. .. ....... 30
ECOLOGICAL OBSERVATIONS . . . ......... . ........... 30
TAXONOMIC STATUS ...... . ................... . .. 36
Microtus pennsylvanicus dukecampbelli NEW SUBSPECIES ............. ......... 38
D ISC U SSIO N ......... .. .. ... .................. ...... ... ... 41
LATE PLEISTOCENE AND HOLOCENE DISTRIBUTION ......... . .... ...... 41
M ORPHOLOGICAL TRENDS .................. ... .......... .. ..... 45
SURVIVAL IN A TIDAL SALTMARSH ENVIRONMENT . . . ....... ......... 47
LITERAT U RE C ITED ... ... ...................... .......... ......... 49
INTRODUCTION
While studying the seaside sparrow Ammospiza maritima in a coastal
salt marsh on the west coast of Florida, we discovered the presence of
Microtus pennsylvanicus, a small mammal previously unrecorded as an
extant form in Florida. The current known range of the meadow vole in
eastern North America extends as far south as Newton County in the mid-
Piedmont area of Georgia, where the subspecies M. p. pennsylvanicus is
found in wet meadows (the preferred habitat) and dry upland fields
(Odum 1948, Golley 1962). Along the Atlantic coastal plain it occurs as
far south as Charleston, South Carolina (Hall 1981). The Florida popula-
tion of Microtus is 500 km south of both the above populations in the
eastern United States. Recent analyses of the mammals of the lower
coastal plain (Okefenokee Swamp) of Georgia did not reveal the presence
of Microtus pennsylvanicus (Laerm et al. 1980), and the species has not
been reported in Florida in recent times in spite of a long history of mam-
mal surveys beginning in the late nineteenth century. The meadow vole is
known to occur in coastal saltmarsh habitats in several parts of its range
(Hall 1981, Strickland, pers. comm., Wetzel, pers. comm.) It is more fre-
quently found in Hudsonian, Canadian, and Upper Austral life zones
(Martin 1968). Near the southern limits of its range the meadow vole ap-
pears to be restricted to the lower Austral zone in Georgia and the upper
Austral zone as a disjunct western population in the state of Chihuahua,
Mexico (Anderson 1961, Bradley and Cockrum 1968, Anderson and Hub-
bard 1971). The ecological factors that limit the southern distribution of
Microtus pennsylvanicus are unknown, but Martin (1968) speculated that
they were related to temperature.
Vol. 28, No. 2
WOODS ET AL.: M. PENNSYLVANICUS
ACKNOWLEDGEMENTS
We wish to acknowledge the following individuals and institutions for making specimens
available: Joshua Laerm, Museum of Natural History of the University of Georgia (UGA);
Gordon L. Kirkland, Jr., Shippensburg State College Vertebrate Museum (SSC); Ralph M.
Wetzel, University of Connecticut Museum of Natural History (UCM). For assistance with
field work under difficult conditions we thank Greg Gutche, John Hermanson, Gregory Mur-
ray, Kathryn Winnet-Murray, Rick Sullivan, Laurie Wilkins, Kenneth Wilkins, and Lois
Wood. Anne Pratt assisted in the analyses of fossil Microtus. The assistance of Margaret K.
Langworthy in the field, as well as in the analysis of the data. is gratefully acknowledged. We
also thank John Hermanson, Roy Horst, Stephen Humphrey, Gary Morgan, and Henry W.
Setzer for reviewing the manuscript. Financial support for the project was from the Florida
Game and Fresh Water Fish Commission, the Florida State Museum, University of Florida
(UF), Florida Audubon Society, and Tall Timbers Research Station.
MATERIALS AND METHODS
Animals were trapped with Sherman live traps. Two traps were set at each station: one on
the ground and one up in the vegetation. To determine the comparative number of rodents on
the study area in 1980 we operated 297 traps during a six-night period (March 5, 6. 9, 16. 17,
24). To make the trap array conform to the shape of a peninsula, the grid was distorted so that
the traps on the E-W grid lines were at 12.5 m intervals, while on the N-S lines they were
placed at 25 m intervals. The size of the trapping area was 3.75 ha. Traps were prebaited for
three days and thereafter were set at twilight and checked at dawn. Subsequent trapping
periods in the area were carried out in April, May, July, September, and November, 1980,
and January, February, and April, 1981. For estimates of small mammal densities from
mark-recapture data we employed Jolly's stochastic model (Jolly 1965).
Vegetation samples were taken following the methods of Wiens (1969). For habitat
analysis we randomly placed a 1 m2 quadrant along 18 randomly located transects, each 71 m
long. The corners of this square served as subpoints at which we vertically positioned a thin
metal rod. We tallied the number of contacts that vegetation made with the rod. From this
information we were able to determine the numerically dominant vegetation at each random
subpoint. In addition, we determined the dominant cover type at each trapping station by the
same method.
Microtus were tagged with ear tags, sexed, weighed, and released. Six adult specimens
were collected for systematic analysis and are now in the collections of the Florida State
Museum. Because of the apparent small size of the population, only six specimens were col-
lected in an effort to insure that a breeding population survived in the area.
For blood protein analysis, blood was collected in the field from the suborbital canthal
sinus of 14 individuals that were subsequently released. Hemolysate and serum samples were
prepared according to the methods of Selander et al. (1971). Horizontal starch gel electro-
phoresis was used to fractionate samples employing a 13 % concentration of hydrolyzed starch
(Sigma Chemical Company). Buffer systems were those described by Selander et al. (1971)
with one exception, that used to fractionate albumins described by Jensen and Rasmussen
(1971, 1972). Enzymes and nonenzymatic proteins were identified by the stains described by
Selander et al. (1971).
For chromosomal studies voles were injected intraperitonically with a 0.004% Velban
(Eli Lilly and Company) solution. A cell suspension was obtained from the bone marrow of
the hind limbs and processed according to the technique outlined by Lee (1969). A karyogram
of mitotic chromosomes was constructed from a photomicrograph.
The specimens collected on the site were made into standard museum study skins and
complete skeletons. For taxonomic analysis standard field measurements and ten additional
measurements were taken following the format in Snyder (1954). In the analysis, only adult
BULLETIN FLORIDA STATE MUSEUM
animals were compared (Snyder's age groups 10-12). Males and females of the Florida
population were grouped together in the analysis because of the small sample size. The
Florida population was compared with 14 known adult specimens of M. p. pennsylvanicus
from Georgia (sexes combined), which were borrowed from the Museum of Natural History,
University of Georgia. A survey of major museums in the United States indicated that no
other specimens of Microtus from Georgia were known. The Florida population was also
compared with 14 specimens (sexes combined) of M. p. nigrans from a tidal salt marsh in
Maryland, 20 specimens (sexes combined) of M. p. pennsylvanicus from western Connecticut,
and 20 specimens (sexes combined) of M. p. pennsylvanicus from a tidal salt marsh in eastern
Connecticut.
Other comparisons in the analysis were based on published measurements of other forms
of Microtus pennsylvanicus. In the case of M. p. pennsylvanicus from Pennsylvania (Snyder
1954), only data on males were available for comparison. In this case statistical comparisons
are significant only for characters shown by Snyder to be independent of sex. It was not possi-
ble to control for the influence of the season of collection on the means of the samples. Snyder
(1954), however, demonstrated that in the populations of M. p. pennsylvanicus he studied
there was no clear-cut relationship between the means and the date of the collection.
Dental variation was evaluated through the use of camera lucida drawings of tooth rows.
These drawings were traced onto semi-transparent mylar sheets and superimposed for com-
parisons of symmetry and form. Patterns were then transferred into tabular form which
enabled calculation of the normal pattern for a population as well as percent variation from
the norm.
STUDY AREA
The study site is on Waccasassa Bay, an extension of the Gulf of Mexico, in Gulf Ham-
mock, Levy County, Florida (Fig. 1). The marsh is bounded on the southwest by a creek, on
the south by Waccasassa Bay, and inland by the extensive lowland hardwood forest charac-
teristic of the upper west coast of Florida. The shoreline is gradual, dotted with numerous
oyster bars and small islands and cut by many tidal creeks. The topography of the study site is
flat, except for creeks up to 1 m deep. In our measurements, tides averaged 0.8 m and salinity
19.03 1.09 (32 stations on 24 April). At low tide extensive mud banks, riddled with burrows
of crabs (Uca and Sesarma), are exposed. Vegetation covers 94.1% of the study area, and
tidal creeks cover 5.9% Figure 2 is a photograph of the main study site at low tide. Vascular
plant species, in order of percent relative cover, are: smooth cordgrass (Spartina alterniflora)
37.8%,, black rush (Juncus roemerianus) 26.3 %, seashore saltgrass (Distichlis spicata) 23.2 %,
perennial glasswort (Salicornia virginica) 8.1%, saltwort (Batits maritima) 3.2%, sea oxeye
(Borrichia frutescens) 1%, Virginia dropseed (Sporobolus virginicus) 1%, sea lavender
(Limonium carolinianum) 1%, Christmas berry (Lycium carolinianum) 1%; key grass
(Monanthochloe littoralis) 1%, saltmarsh aster (Aster tenuifolius) 1%, and wrack deposits
1 %. The relative elevation of the major plant communities is indicated by water depths (cm)
during spring flood tides, where greater water depths signify lower elevations: S. alterniflora
(N = 11) 25.6 7.2, J. roemerianus (N = 11) 20.9 + 5.3, B. maritima (N =3) 16.5, D. spicata
(N =31) 15.3 + 3.3, and S. virginica (N = 6) 13.8 1.2.
Because the area is so low, flooding is a major factor on the study site. Extreme tides
occur approximately twice a month and inundate the entire area. Unpredictable high water
levels occur on the study site when a spring tide coincides with high southerly or westerly
winds or rain. In the study area two such unpredictable floods occurred in 1979 and one in
1980. The effects of hurricanes are also of potential major importance. The National Oceanic
and Atmospheric Administration (Fernald 1981) lists 21 major hurricanes in Florida between
1900 and 1975. Hurricane Alma in May 1970 passed directly over Levy County as it came
ashore from the Gulf, and other hurricanes passed nearby in the Gulf of Mexico in 1966,
1972, and 1979 (Fernald 1981).
Vol. 28, No. 2
1982 WOODS ET AL.: M. PENNSYLVANICUS 29
VJ
FIGtiRE 1. -Distribution of fossil localities of Microtus pennsyhlanic.us in Florida.
BULLETIN FLORIDA STATE MUSEUM
FIcuRE 2.-Photograph taken on 5 September 1980 of the eastern site at low tide. In the
foreground is suitable Microtis habitat containing Spartina and Distichlis. In the background
is a band of Juncus. The inland coastal forest is on the horizon.
RESULTS
ECOLOGICAL OBSERVATIONS
Ecological data were collected in the Waccasassa Bay area between 22
and 26 June 1979 on a site (hereinafter referred to as the "western site")
within 25 m of a maritime live oak forest. The site was abandoned because
of disturbances by raccoons. Between 21 February and 15 June 1980
ecological data were collected on a nearby site farther into the marsh,
which will be referred to as the "eastern site." In the combined areas of
the western and eastern sites we trapped four species of rodents. In order
of abundance these were Oryzomys palustris, Microtus pennsylvanicus,
Sigmodon hispidus, and Peromyscus gossypinus (Table 1). The two brief
study periods produced very different results, and these differences con-
tinued in the long-term general study on the western site. The western site
was closer to drier areas, such as the scrub fringe and forest, and also con-
tained more Juncus (black rush). In transect studies from the forest to the
sea the scrub fringe was most frequented by Sigmodon and the forest by
Vol. 28, No. 2
WOODS ET AL.: M. PENNSYLVANICUS
TABLE 1. -Summary of small mammal trapping- Waccasassa Bay area.
19791 19802
All All
New Captures New Captures
Oryzomys palustris 55 69 53 211
Sigmodon hispidus 10 29 2 7
Peromyscus gossypinus 4 5 0 0
Microtu.s pennsylvanicus 0 0 9 36
'Three days' trapping: 22. 24. 26 June
'Six days' trapping. 5, 6, 9, 16, 17. 24 March
Peromyscus. Oryzomys was more common in Juncus than in other areas
of the salt marsh.
Table 2 summarizes the trapping success on both areas for all species
and microhabitats. Our results differ markedly from data collected by
Harris (1953) on the abundance of small mammals in a tidal salt marsh in
Dorchester County, Maryland, where 121 Microtus were trapped in con-
trast to 78 Oryzomys. An analysis of the percentage of small mammals
preyed upon by predators in the Maryland marshes yielded similar pro-
portions: fox scats had 50% Microtus, 5% Oryzomys; barn owl pellets
72% Microtus, 27% Oryzomys; and marsh hawk pellets 73% Microtus,
15% Oryzomys (Harris 1953). Our analysis of seven marsh hawk pellets
from the eastern site yielded 1 Microtus, 1 Oryzomys, and 5 passerines.
In our study in Florida the main plant associations sampled for mam-
mals were Juncus (27% of trap placements) and Distichlis (52%). The
location of these two microhabitats differs (Fig. 3; Post 1981). Juncus oc-
cupies slightly lower, but usually more landward sections of the marsh.
The Distichlis flats are at higher elevations but closer to the bay. The
average depth of spring tide floodwater in Juncus was 21 cm vs 15 cm in
Distichlis. Comparing the number of Oryzomys captures in these two
vegetation types (Table 3,) shows that significantly more Oryzomys were
caught in Juncus than in Distichlis (X2 corr., 1 df = 223; P .001). Because
of limited data, no such comparison is possible for Microtus, but a com-
parison of the overall distribution of Oryzomys and Microtus in the four
vegetation types shows that they differed significantly (Table 4). Propor-
tionately more Microtus than Oryzomys were captured on the Distichlis
flats. Microtus, unlike Oryzomys, avoids areas dominated by S. alterni-
flora. The Microtus were also caught less frequently in Juncus than were
Oryzomys. That no Microtus were trapped in 1979 on the western site
may be related to the dominance of Juncus and absence of Distichlis in
that area.
In New York salt marshes, Microtus is more often associated with
shorter grass (Spartina patens, Distichlis spicata) than with taller grass
(Phragmites, S. alterniflora) (Post and Greenlaw, unpubl. data). In Mary-
1982
Table 2. -Small mammal habitat selection and trapping success. Z
1979- Western Site 1980-Eastern Site Total O
Traps S. P. 0. Traps S. M. 0. Traps S. P. M. O.
Available hispidus gossypinus pialstris Available his pidus pennsylvanicns palustris Available hivpidus gossypinus penmsylvanicus palustris >
Juncus 165 12 3 46 452 0 7 79 617 12 3 7 125
S. alterniflora 47 3 0 10 189 0 0 33 236 2 0 7 43
Distichlis 0 876 2 21 74 876 2 5 21 74
Salicornia 61 3 0 14 136 5 6 12 117 8 0 6 26
Wrack 0 46 0 2 3 46 0 0 2 3
Scrub edge 16 11 2 0 16 11 2 0 0
Unknown 8 1 0 0 8 1 0 0 0
TOTAL 297 29 5 70 1699 7 36 211 1996 36 5 36 271
Ammospiza maritima
Peromyscus gossypinus
Sigmodon hispidus
Microtus pennsylvanicu
Oryzomys palustris
Cj~
Forest i-Juncus I- Distichlis Jcunucsl Spartina
VEGETATION
FIGURE 3. -Vegetation zones in a salt marsh in WVaceasassa Bay, Levy County, Florida.
BULLETIN FLORIDA STATE MUSEUM
TABLE 3. -Comparative numbers of Oryzomys palustris captured in different microhabitats.
Microhabitat
Juncus Distichlis Total
No. of Traps
with Oryzomys 79 74 153
No. of traps
empty' 366 779 1145
Total traps 445 835 1298
'Traps containing other animals or improperly set, were not included in this category.
TABLE 4. Comparison of number of Oryzomys palustris and Microtus pennsylvanicus caught
in different microhabitats.
Number of:
Microhabitat' 0. palustris M. pennsylvanicus Total
Juncus 79 7 86
S. alterniflora 33 0 33
Distichlis 74 21 95
Salicornia 15 8 23
Total 201 36 237
'The overall distribution of the two species in the various microhabitats is significantly different (X2 = 26.22; P .001).
land marshes characterized by salt grasses, S. patens, S. alterniflora, and
D. spicata, Microtus was more abundant than Oryzomys (Harris 1953).
Microtus may avoid Juncus because its stems are more coarse and taller
and may be less attractive as food. The more arboreal Oryzomys would be
better adapted to use the tall, smooth stems of Juncus while foraging.
No estimates of Oryzomys density in salt marshes are published, but in
some other habitats density reaches 19/ha (Smith and Vrieze 1979). No
population estimate can be made for the western site because of raccoon
disturbances. Comparative numbers of rodents captured on this site are:
55 Oryzomys, 10 Sigmodon, and 4 Peromyscus (Table 1). The latter two
species were captured within 25 m of the forest. Oryzomys was the sole
occupant of the lower elevations of the salt marsh on the western site. On
the eastern site population density was estimated at 8/ha (95 % confidence
interval: 7/ha-10/ha). Too few of the other species were captured to
estimate their populations, but in comparison with 53 Oryzomys captured
during the period 5-24 March on the eastern site we captured 9 Microtus
and 2 Sigmodon.
Getz (1961) calculated that the densities of M. pennsylvanicus in a
freshwater marsh in Michigan varied between 16/ha in late February and
64/ha in July. It is clear from our results that densities of Microtus on the
study plot at the time of our study were far below those observed by Getz.
Our data are too few to allow estimation of home range sizes, but
Vol. 28, No. 2
WOODS ET AL.: M. PENNSYLVANICUS
distances between successive captures allow some comparisons between
sexes and species (Table 5). The average distance between captures for
Oryzomys was 66 m, while the same value for Microtus was 32 m. One
adult male Oryzomys moved 251 m. In another case a subadult male
crossed a 60 m tidal inlet to reach an island 175 m from his original cap-
ture point. These data correspond to the estimates made by Negus et al.
(1961) for home ranges of Oryzomys in coastal scrub: 0.33 ha for males
and 0.21 ha for females. If these areas are considered circular, the di-
ameter of the male's home range is 69 m, and the female's is 52 m, which
agrees closely with our average intercapture distances of 71 m for males
and 60 m for females.
Our data on the average distance between captures for Florida Micro-
tus can be compared with other published data. Harris (1953) calculated
the average maximum distance between capture points for M. p. nigrans
in a Maryland salt marsh to be 38.5 m. This is similar to our figure of 32
m. If we transcribe these figures into the area of a circle to get a gen-
eralized average home range, then it is possible to make a comparison
with other published data. For the Maryland population the generalized
home range will be 1164 m2, and for our Florida data it will be 804 m2.
Getz (1961) calculated a maximum home range size for both males and
females from a freshwater marsh in Michigan to be approximately 800
m2. He found that the home range size for male M. pennsylvanicus in the
marsh was significantly larger than for females, and that home range size
for both sexes was inversely proportional to population density. The
population density of both male and female Microtus at maximum home
range size was approximately 16 individuals/ha. In a study of M. pennsyl-
vanicus from an old field in Virginia, Madison (1980) used radiotelemetry
to demonstrate daily ranges for male voles of 192 m2 (69 m2 for females) at
an average density of 155 individuals/ha.
From 21 February to 24 March 1980, an equal proportion of male and
female Oryzomys were captured: 46 males and 44 females. In the same
period we caught 5 male and 9 female Microtus.
For Oryzomys we noted signs of breeding activity in early March,
when we first captured females with perforated vaginas and males with
scrotal testes. The appearance of smaller Oryzomys in the population also
indicated recruitment of young. The proportion of juveniles (animals
weighing less than 30 g) in the population was 4% in February, 11% in
March, and 21 % in June. We found three litters of Oryzomys, each with
three young. These data (Table 6) indicate that reproductive activity in
Oryzomys increases between February and June and reaches a highpoint
in June. In contrast, Microtus were sexually active in late February when
trapping began. A specimen (UF 12136) collected on 27 February 1980
had six embryos, as did another (UF 12001) collected on 9 March 1980. A
1982
BULLETIN FLORIDA STATE MUSEUM
TABLE 5. -Movements of rodents, 1980.
Species Individuals Distances between successive captures
Sex (Captures) X SD SE Var. Range 95% CI
Oryzomys palustris
males 14 (48) 71.5 66.83 9.65 4466 12-263
females 13 (38) 60.03 48.81 7.92 2382 0-251
both 26 (86) 66.43 59.49 6.42 3539 0-263 53.67-79.19
Microtus pennsylvanicus
males 3 (10) 38.8 25.66 (8.11) 658 0-80
females 6 (13) 26.0 17.7 (4.92) 314 0-56
both 9 (23) 31.6 21.97 4.58 483 0-80 22.12-41.08
subadult weighing 30 g was captured on 29 May 1980. In Microtus repro-
ductive activity appears to extend throughout the study period, but is
most active in February and March. These data are consistent with Har-
ris' (1953) study of a saltmarsh community in Maryland. He found the
peak of reproductive activity for Microtus in March with a lesser peak in
June, while no Oryzomys were reproductively active in March but 100%
were in June and July.
Of the two rodents occupying lower marsh elevations, Oryzomys
weighed significantly less than Microtus (Table 7). Sigmodon, which only
occupied patches of Borrichia, in the highest marsh elevations, weighed
74.4 g. The variability of Oryzomys weights increased from February to
June, reflecting the addition of juveniles to the population.
TAXONOMIC STATUS
Our examination of the dental morphology of the small vole popula-
tion from Florida indicates that while the pattern of the enamel reentrant
folds was extremely variable, a consistent feature was the presence of a
fifth rounded posterior loop on M2 characteristic of Microtus pennsyl-
vanicus (Hall 1981). Chromosomal analysis of a single female Microtus
(UF 12136) from the Florida population revealed a diploid number of 46.
The karyotype (Fig. 4) consists of 4 pairs of biarmed chromosomes and 19
pairs of acrocentric chromosomes. This pattern is identical to that reported
in M. pennsylvanicus by Matthey (1952) and Hsu and Benirschke (1967).
The chromosomal pattern is clearly different from that of other south-
eastern microtines, such as M. ochrogaster with 2n = 54 (Matthey 1955)
and M. pinetorum with 2n = 62 (Matthey 1955, Beck and Mahan 1978).
Blood proteins (Table 8) from 14 Microtus from the Florida popula-
tion encoded by nine structural loci were analyzed by horizontal starch
gel electrophoresis (Selander et al. 1971). This population demonstrated a
mean coefficient of genetic identity of 0.943 with northern populations
(Vermont, Pennsylvania) of M. pennsylvanicus and 0.556 with M. ochro-
Vol. 28, No. 2
TABLE 6. -Indications of breeding activity among rodents trapped at Cedar Keys, 1980.
Oryzomys palustris Microtus pennsylvanicus
No. : No. : No. Total No. : No. : No. Total
Date Examined Scrotal Perforate (%) Examined Scrotal Perforate (%)
24 Feb 31 0 0 0 5 0 2 2 (40)
5 Mar 18 0 0 0 3 1 1 2 (66)
6 Mar 28 1 2 3 (10) 6 1 4 5 (83)
9 Mar 28 1 0 1 ( 3) 3 1 1 2 (66)
16 Mar 19 2 3 5 (26) 3 1 0 1 (33)
17 Mar 27 5 4 9 (33) 7 2 4 6 (85)
24 Mar 22 3 2 5 (22) 5 3 0 3 (60)
29 May 23 2 5 7 (30) 7 0 0 0
15 Jun 55 13 16 29 (52) 7 1 3 4 (57)
TABLE 7. -Weights (g) of small mammals captured in salt marsh on Waccasassa Bay ( SD).
Oryzomys paluslris Microtus pcmusylvanicn's
Males Females Both Sexes Males Females Both Sexes
Date N ( 5) N ( ) N ( ) N (x) N (x) N (x)
February 27 (52.4 2.5) 22 (38.3 1.7) 49 (46.1 1.9)
March 62 (48.5 1.7) 52 (37.4 1.2) 114 (43.4 1.2)
February-
March 13 (56.9 4.3) 17 (56.9 2.4) 30 (56.9 2.3)
May 10 (59.4 6.1) 12 (54.1 +4.6) 22 (56.5+3.7)
June 25 (52.6 4.4) 29 (46.1 2.9) 54 (49.1 + 2.6)
May-June 9 (55.7 6.3) 6 (63.2 3.0) 15 (58.7 +4.0)
Total, 1980 100 (46.5 1.8) 45 (57.5 2.9)
BULLETIN FLORIDA STATE MUSEUM
X X
f a 0- 00 AA o0 e
FIGURE 4.- Karyotype of female Microtus pennsylvanicus (UF 12136) from a salt marsh in
Waccasassa Bay, Levy County, Florida.
gaster. Electrophoretic analysis of blood proteins strongly indicate that
the affinities of the Florida population of Microtus are with M. pennsyl-
vanicus. The chromosomal and electrophoretic data therefore confirm the
conclusion based on morphological evidence that the Florida population
is a new subspecies of M. pennsylvanicus.
Microtus pennsylvanicus dukecampbelli NEW SUBSPECIES
HOLOTYPE. -Male, adult, skin and skeleton; UF 12005; collected 28
April 1980 by C. A. Woods.
TYPE LOCALITY. -Island Field Marsh in Waccasassa Bay, Levy County,
Florida. Because of the apparent extreme rarity of this form, the precise
type locality will not be published, but is available as part of the perma-
nent records of the Florida State Museum, University of Florida.
PARATYPES. -UF 12001 9 12136 9, 12202 o", 15087 9, all skins with
complete skeletons. Other referred material includes UF 10910, the com-
plete skeleton of an adult 9, and UF 15242, fur, teeth, and skeletal
fragments from a marsh hawk pellet.
DISTRIBUTION. Known only from the type locality.
DIAGNOSIS. -Entire animal and skull large; ears short, hind foot long;
color dark black-brown on the back grading to dark gray on the belly.
COMPARISONS (TABLE 9). From M. p. pennsylvanicus, the subspecies
to the north at the southern extent of its range in Georgia (Newton, Gwin-
nett, Clark, and Union counties), M. p. dukecampbelli differs in being
significantly (two-tailed t-test at 95% level) larger in total length, tail
Vol. 28, No. 2
WOODS ET AL.: M. PENNSYLVANICUS
TABLE 8. Allelic frequency and genetic variability of representative populations of Microtus.
Alleles are designated according to proportional electrophoretic mobility relative
to the fast migrating allele. Frequencies of alleles at polymorphic loci are given in
parentheses: all other loci are monomorphic.
.Microtus pennsylcanicus
populations
Florida Pennsvlvania Vermont Microtuis
Allele n = 14 n =20 n = 12 ochrogaster
Hemoglobin
Alphao0" 1.00 1.00 1.00 1.00
Beta'00 .25 .27
Beta"7 1.00 .75 .73 1.00
Albumin""' 1.00 1.00 1.00
Albumin98 1.00
6-Phosphogluconate
dehydrogenase1"" 1.00 1.00 1.00 1.00
Lactate
dehydrogenase-2100 1.00 1.00 1.00 1.00
Esterase- 1o"' .53 .25 -
Esterase-1" .08 -
Esterase-1"9 1.00 .38 .75 1.00
Esterase-168 .01 -
Esterase-51"" .23 .25
Esterase-5" 1.00 .77 .75
Esterase-50 -1- 1.00
Esterase-810" 1.00
Esterase-875 1.00 1.00 1.00
Esterase-910" 1.00
Esterase-985 1.00 1.00 1.00 -
length, hind foot length, greatest length of skull, condylozygomatic
length, length of incisive foramen, length of diastema, cranial breadth,
and zygomatic breadth. It also differs in having a shorter rostrum and is
much darker in coloration. M. p. dukecampbelli differs from M. p. penn-
sylvanicus at the mid-range area in Pennsylvania (Goin 1943, Snyder
1954) in all the above characters, except that tail length is not significantly
different and rostrum is longer rather than shorter (i.e. the rostrum in
Pennsylvania mice is exceptionally short while it is exceptionally long in
Georgia mice). In addition, M. p. dukecampbelli has smaller ears, is
heavier in body weight, has longer nasals, and a higher cranium. Com-
pared with M. p. nigrans, the saltmarsh-dwelling form from coastal
Maryland (Somerset County), M. p. dukecampbelli is larger in all charac-
ters measured except total length, tail length, hind foot length (which is
larger, but not significantly so), and in the lengths of the ear, rostrum,
and interorbital breadth. In M. p. dukecampbelli the ear and length of
the rostrum are shorter (i.e. the Maryland form has a long rostrum). M. p.
dukecampbelli is darker in coloration than M. p. nigrans.
TABLE 9. -External body and skull measurements (in mm) and weights (in g) of Microtus pennsylvanicus dukecampbelli, M. p. pennsylvanicus (known
Georgia specimens), M. p. pennsylvanicus (from Pennsylvania after Snyder 1954) and M. p. nigrans (from Maryland)
Greatest Condylo- Incisive Inter-
Total Tail Hind length Zygomatic Nasal foramen Diastema Rostrum Cranial orbital Zygomatic Cranial
length length foot Ear Weight Skull length length length length length breadth breadth breadth height ]
M. p. dukecampbelli (c 9) .
N 5 5 5 5 5 6 6 6 6 6 6 6 6 6 6 Z
X 186.10 48.80 23.36 13.00 61.20 31.17 24.65 8.28 5.69 9,22 6.52 11.71 3.60 16.89 11.66 I
S 7.65 4.27 1.07 1.87 14.66 0.52 0.28 0.16 0.18 0.22 0.29 0.15 0.09 0.37 0.23 r,
S2 46.84 14.56 0.92 2.80 171.85 0.22 0.07 0.02 0.03 0.04 0.07 0.02 0.01 0.11 0.04 0
Range 178-198 43-55 22-25 11-16 43-80 30.7-32.0 24.3-25.1 8.1-8.5 5.5-6.0 9.0-9.6 6.2-7.0 11.6-12.0 3.5-4.0 16.5-17.3 11.2-11.8 -
M. p. pennsylvanicus (Georgia, c 9)
N 14 14 14 7 14 4 4 6 8 8 4 3 8 5 0 >
X 153.79 35.57 20.71 14.14 48.32 28.50 21.34 8.02 4.72 8.37 7.16 10.98 3.66 15.36 -
S 10.38 4.42 1.82 1.95 13.52 1.72 1.43 0.54 0.23 0.50 0.36 0.47 0.23 0.69 -
S2 100.03 18.10 3.06 3.27 169.70 2.23 1.53 0.25 0.04 0.22 0.10 0.15 0.05 0.38 -
Range 128-173 27-43 18-25 12-18 29-75 26.8-30.5 19.8-23.1 7.4-8.7 4.4-5.1 7.7-9.3 6.8-7.5 10.5-11.3 3.4-4.1 14.4-16.1 -
M. p. pennsylvanicus (Pennsylvania, ca Cr)
N 38 38 60 38 22 34 34 30 37 37 34 52 58 33 23 r
X 167.40 44.40 20.60 14.00 44.19 27.35 21.31 7.70 5.21 8.34 6.04 11.03 3.66 15.23 10.07 "
S 7.40 5.70 0.80 1.00 6.29 0.63 0.52 0.42 0.28 0.27 0.23 0.29 0.12 0.47 0.27
S2 54.76 32.49 0.64 1.00 39.16 0.40 0.27 0.18 0.08 0.07 0.05 0.08 0.01 0.22 0.07
Range -
M. p. nigrans (Maryland, c 9)
N 13 13 13 13 13 13 13 13 13 13 13 13 13 13 12
X 174.92 44.46 22.31 14.62 50.73 29.42 22.22 7.62 5.30 8.76 7.20 11.20 3.70 16.22 11.10
S 17.54 6.04 1.18 0.77 8.18 0.77 0.61 0.34 0.42 0.31 0.44 0.24 0.16 0.40 0.43
Sy 284.07 33.63 1.29 0.54 61.75 0.55 0.34 0.11 0.16 0.09 0.17 0.05 0.02 0.15 0.17
Range 146-203 33-56 21-25 13-16 37-62 28.4-30.9 21.7-23.3 7.2-8.3 4.5-6.1 8.2-9.3 6.6-7.8 10.8-11.6 3.4-4.0 15.7-17.0 10.4-11.7
00
z
WOODS ET AL.: M. PENNSYLVANICUS
REMARKS. -A comparison of M. p. dukecampbelli with other forms of
Microtus pennsylvanicus from a variety of locations, habitats, and sub-
species indicates that it is most distinct from the Georgia population.
Even though the populations of mice from Georgia are considered part of
the same subspecies (M. p. pennsylvanicus) as the Pennsylvania forms ex-
amined by Snyder (1954), the length of the rostrum in the two forms dif-
fers dramatically (see Comparisons above and Table 9). This raises the
possibility that the Georgia forms are also a distinct subpopulation (sub-
species) from the more typical M. p. pennsylvanicus to the north. The
criteria for using subspecies and populations are discussed by Snyder
(1954) and Anderson (1959). The known specimens from Georgia are in
poor condition and limited in number. Therefore the documentation of
morphological and geographical variability of Microtus pennsylvanicus in
Georgia requires additional specimens.
Comparisons with M. p. pennsylvanicus from both Georgia and Penn-
sylvania show that the Florida form differs in most characters and is
distinct from both no matter what their relations to each other might be.
The Florida form is most similar in morphology to M. p. nigrans, which
also inhabits coastal salt marshes (Handley and Patten 1947, Gordon Kirk-
land, pers. comm.). These similarities include large body size, long hind
feet, and dark coloration. The Florida form is also similar to insular forms
such as M. p. copelandi from Grand Manan Island, New Brunswick, and
M. p. magdalenensis from Magdalen Island, Quebec (Youngman 1967). It
shares with insular forms large body size, as well as cranial characters
such as long nasals and broad braincase.
ETYMOLOGY. -Named in memory of zoologist Howard W. "Duke"
Campbell, Jr., of Gainesville, Florida, who died 10 December 1981.
DISCUSSION
LATE PLEISTOCENE AND HOLOCENE DISTRIBUTION
The first report of Microtus pennsylvanicus from Florida was by Arata
(1961), who described and illustrated the dentitions of two specimens (UF
3917, 3918) from the late Pleistocene Devil's Den fauna (near Williston,
Levy County). The specimens were collected in a richly stratified bone
bed approximately 9 m under water in a collapsed sinkhole. Additional
Pleistocene records from Florida have been reported by Martin (1968) and
Martin and Webb (1974) from Devil's Den (UF 12816), as well as froin:
Arredondo IA (UF 3586), a limestone chimney 10 km SW Gainesville
(Alachua County), and Withlacoochee River 7A (UF 13090), a limestone
pocket on the bank of the Withlacoochee River 13 km W Dunnellon
(Citrus County). Other Florida specimens of M. pennsylvanicus in the
BULLETIN FLORIDA STATE MUSEUM
Florida State Museum collections are from Waccasassa River 2B (UF
16333) from near Gulf Hammock (Levy County), as well as additional
specimens from Devil's Den (UF 12818) and the Withlacoochee River
(UF 28000).
The locations of these four fossil sites are plotted in Figure 1. The ages
of these deposits are not well known in all cases. Arredondo IA, for exam-
ple, may be Illinoian, Sangamonian, or Wisconsinan (see Martin 1968).
The Withlacoochee River 7A site is considered to be between 11,000 and
30,000 years old. The Devil's Den fauna is known to date around 8000
years before present (B.P.) (Martin and Webb 1974).
The small mammal fauna in southeastern North America from 10,000
B.P. to the present is not well known and the nature of the climate during
the Holocene is controversial. The method of deducing the climate of an
area from the presence or absence of key faunal elements, such as "steppe"
species and "boreal" species, can also be hazardous because species fre-
quently respond to different factors in different environments (Whittaker
1967). The Devil's Den fauna in Florida, for example, includes several
species that are considered by Martin and Webb (1974) as "northern"
forms, such as Microtus pennsylvanicus, the gray bat Myotis grisescens,
and the muskrat Ondatra zibethicus. Whether these forms were in Florida
because the climate was cool, or for some other reason, is difficult to
deduce from faunal evidence alone. Other possible reasons are increased
precipitation, the "Accordian Effect" (Dansereau 1957, Blair 1958), or a
general reduction of environmental gradients that allowed for more
diverse faunas and floras (Graham 1976).
Graham (1976) demonstrated that microtine rodent species diversity
(number of vole species present per local fauna) was greater during the
late Wisconsinan than it is at the present time in the same areas. This may
indicate a more "equable" climate during late glacial and early post-
glacial times, when temperature and moisture gradients were present but
reduced. In Florida and southern Georgia, Watts (1971, 1980) contended
that strong southwest winds further modified the climate and resulted in
warm dry conditions. He based his conclusions on evidence from pollen
samples gathered at Lake Louise on the Georgia-Florida border, approxi-
mately 180 km north of the Devil's Den site and from Sheelar and Mud
Lakes near Gainesville, Florida. These data on climate and vegetation are
from areas that are geographically close to the known range of Microtus
pennsylvanicus and from approximately the same time-5000-10,000
B.P. Watts (1980) estimated that 8510 B.P. in central and northern
Florida there was an oak savanna habitat with broad patches of bluestem
prairie characterized by various grasses, as well as Ambrosia, Iva, and
Artemsia. Such a climate and habitat would be suitable for voles, and it is
probable that Microtus pennsylvanicus was broadly distributed along
Vol. 28, No. 2
WOODS ET AL.: M. PENNSYLVANICUS
water courses and wet areas in patches of bluestem prairie and savanna in
northern and central Florida that existed until approximately 5000 B.P.
Between 6710 and 5000 B.P. areas of pine expanded and eliminated the
prairie plants. Black gum, sweet gum, ironwood, and other trees and
vines became established by 5000 B.P., eliminating suitable upland
habitat for grassland voles (Watts 1980).
Sea levels were much lower 10,000 years B.P., and suitable lowland
savanna and prairie habitats extended far to the west into what is now the
Gulf of Mexico. Blackwelder et al. (1979) estimated that at 10,000 B.P.
sea levels were approximately 25 m lower than present levels in the Gulf
coast area, which would extend suitable terrestrial habitats 100 km to the
west of the present coastline along Waccasassa Bay and expose a broad
area to the west along the Gulf coast. The area of expanded habitat would
be most extensive between Clearwater and the mouth of the Appalachicola
River to the north (Fernald 1981). Waccasassa Bay is in the center of this
once extensive area that has been inundated in the last 10,000 years and is
consistent with the location of the relict population of Microtus. The im-
portance of this exposed area of habitat along the Gulf coast on Florida
mammals was discussed by Webb (1977), who referred to it as the Gulf
Coast Savanna Corridor.
M. pennsylvanicus is also known from the late Wisconsinan of Loui-
siana, where a molar was reported from Kimball Creek in West Feliciana
Parish (Martin 1968). These spotty records indicate that M. penn-
sylvanicus probably had an extensive range in the Gulf coast area during
the late Pleistocene. Although the current distribution of M. pennsyl-
vanicus (Hall 1981) indicates that the closest population of meadow voles
is far to the north, it is not clear that the Florida saltmarsh vole is
necessarily a relict of northern populations. Frey et al. (1975) did not find
any M. pennsylvanicus in their samples of late Pleistocene mammals in
Georgia estuaries. Therefore M. p. dukecampbelli may be a relict of the
more westerly distributed population of M. pennsylvanicus in the Gulf
Coast Savanna and adjacent patches of bluestem prairie in northcentral
Florida. The distribution of the vole would have been severely restricted
by rising waters to the west and the expanding pine forest inland.
Archaeological evidence indicates that M. pennsylvanicus was not
widely distributed in Florida after 5000 B.P., and that the ranges of other
forms, such as Myotis grisescens and Ondatra zibethicus, were also be-
coming restricted. The muskrat remained in north Florida until approxi-
mately 3150 B.P. (Bullen 1958). Together with the data on sea levels and
vegetation, this information indicates that the vole must have had a
severely limited range for the last several thousand years. Dalquest et al.
(1969) and Graham (1976) indicated that in Edwards County, Texas
(Schulze Cave), the modern faunal composition was reached by 3800 B.P.
BULLETIN FLORIDA STATE MUSEUM
That date closely agrees with the conclusions of Bryson et al. (1970) and
Webb and Bryson (1972) for the time modern faunal assemblages became
established in other parts of the eastern United States. Therefore, M. p.
dukecampbelli probably was disjunct from M. pennsylvanicus to the
north and west for at least 5000 years, and probably longer if Watts
(1980) is correct about the late Quaternary vegetative history of the south-
eastern United States.
The presence of a disjunct relictual population of Microtus pennsyl-
vanicus in Levy County therefore is not a surprise when one examines the
late Pleistocene distribution of this form (Fig. 1; Martin 1968). The small
relict population from Florida is analogous to the relict populations of M.
ochrogaster ludovicianus in the Gulf coast area of Texas and Louisiana
(now probably extinct), M. montanus arizonensis in Arizona and New
Mexico, and M. pennsylvanicus chihuahuensis in northern Mexico. These
forms are undoubtedly relicts of populations more widely distributed dur-
ing the Wisconsinan glacial period of the Pleistocene. Anderson and Hub-
bard (1971) noted that M. p. chihuahuensis, the relict population from
Chihuahua, Mexico (Bradley and Cockrum 1968), is larger and darker in
coloration than is M. p. modestus to the north. They also noted that cer-
tain southern, relict, marsh-dwelling, and small isolated populations of
M. montanus are characterized by large size and dark pelage. Anderson
(1959) stated that in M. montanus the largest and blackest individuals are
in the southwestern part of the geographic range, while in the northern
part of the range the mice are smaller and grayer.
In all of the southern relictual populations discussed here the popula-
tions are similar to M. p. dukecampbelli in morphology and habitat.
Their disjunct status is certainly the result of changing climates after the
Wisconsinan glacial event. In the case of M. p. chihuahuensis there are
known fossil sites containing M. pennsylvanicus in southern New Mexico
(Harris et al. 1973), as well as other isolated relict populations between
Chihuahua and the current range of M. p. modestus in New Mexico
(Anderson 1961, Anderson and Hubbard 1971). In both Florida and the
southwest known dates for the fossils are between 8000 and 11,000 B.P.
(Harris et al. 1973, Martin and Webb 1974). In both areas the mice
probably existed in considerably broader distributions well into the
Holocene, and the true influence of the unique habitat on the morphology
of the populations did not become a major factor until the last 4000 years
(Bullen 1958, Guilday, pers. comm.). The degree of morphological
distinctiveness of M. p. dukecampbelli from M. p. pennsylvanicus is
similar to that achieved by M. p. breweri (Muskeget Island, Massachu-
setts), M. p. rufidorsum (Martha's Vineyard Island), and M. nesophilus
(Gull Island, Suffolk County, New York). Martha's Vineyard and Nan-
tucket islands were separated from Cape Cod less than 5000 years ago,
Vol. 28, No. 2
WOODS ET AL.: M. PENNSYLVANICUS
and Muskeget Island became isolated from Nantucket as recently as 2000
to 3000 years ago (Youngman 1967).
MORPHOLOGICAL TRENDS
The morphological differences between M. p. dukecampbelli and M.
p. pennsylvanicus to the north are difficult to explain. The large size alone
could be the result of an "island effect," (Case 1978). The disjunct distri-
bution of M. p. dukecampbelli in the narrow band of coastal salt marsh
between the Gulf and lowland forest is, in effect, insular. It has been well
demonstrated that Microtus gets larger when isolated on islands (Young-
man 1967). However, the influence of the salt marsh (semi-aquatic
habitat) also must be taken into consideration. While M. p. dukecampbelli
is significantly distinct in most characters from M. p. nigrans, which is
restricted to a similar coastal salt marsh in Maryland, it shares with M. p.
nigrans similarities in body size, tail length, hind foot length, and colora-
tion.
Von Bloeker (1932) demonstrated that saltmarsh-dwelling races of
three small mammals in California were characterized by dark colora-
tion. Grinnell (1913) noted a direct relationship between saltmarsh in-
habitation and the quantity of pigmentation in the pellage of small mam-
mals. The blackest forms occur in marshes with the least salt (von Bloeker
1932, Grinnell 1932), however, and so the dark pigmentation must be
related to something other than salt alone. Most salt- and freshwater-
marsh soils are dark in coloration, suggesting a relationship between ex-
tremely dark marsh soils and the dark pellage of the small mammals in-
habiting the marshes.
To identify trends that might be associated with the saltmarsh habitat,
we compared a population M. p. pennsylvanicus from interior Connecti-
cut (Mansfield Depot, Tolland County) with one from a coastal marsh
habitat (Hammonasset State Park, New Haven County). Measurements
(in mm) of 20 individuals from each population were compared: inland-
total length 152.6, tail length 41.8, hind foot length 21.7, ear length 13.9;
salt marsh- total length 156.7, tail length 41.7, hind foot length 20.3, ear
length 13.5. There was no significant difference between these two popu-
lations in coloration, total length, tail length, or ear length. The hind foot
length was significantly (two-tailed t-test at 99 % level) smaller in the salt-
marsh population, the opposite trend found in M. p. nigrans and M. p.
dukecampbelli. Therefore, in Microtus pennsylvanicus trends in colora-
tion, body size, and hind foot length discussed above do not appear to be
linked in a simple fashion to the saltmarsh environment alone.
As discussed earlier, other forms of Microtus that are disjunct relictual
populations to the south of the main distribution of the taxon are also
1982
BULLETIN FLORIDA STATE MUSEUM
large and dark. The relictual population of M. p. chihuahuensis is located
7.5 km SE Galeana, Ojo de los Reyes, in an area of Chihuahua, Mexico,
that is currently extremely dry and unsuitable for voles. The specific area
of distribution, however, is characterized by numerous spring runs, and
the Microtus live in sedge mats along the spring run (Bradley and
Cockrum 1968, Guilday, pers. comm.). As most other populations found
in southern refugia are also living in pockets of moist vegetation sur-
rounded by otherwise unsuitable habitat, they all share a similar environ-
ment. Natural selection associated with this environment may have
resulted in a significant amount of parallelism in such morphological
characteristics as coloration, body size, and the size of the hind foot. This
"environment" includes the evolutionary factors associated with "insular
effect" as well as "aquatic effect" in influencing the final morphology of
the relict population.
The serum proteins of the Florida saltmarsh vole have been examined
for nine loci. Northern mainland populations of M. pennsylvanicus have a
mean heterozygosity for these nine loci of 0.17 (= 17%), while the
Florida vole has no polymorphism (=0%) at any loci examined. The
absence of genetic variation within this population is consistent with the
reduction of genetic variability observed among insular populations of
mammals (Kilpatrick 1981), and has been specifically demonstrated for
M. pennsylvanicus and M. breweri (Fivush et al. 1975, Kilpatrick 1981).
The lack of genetic and chromosomal differentiation is also consistent
with data on isolated populations of Sigmodon hispidus of the Colorado
River Valley (McClenaghan 1980).
Even though M. p. dukecampbelli lacks genetic variation, as
measured by variation in serum proteins, it does exhibit dental variability.
Variation in the pattern of reentrant folds in Microtus has been noted by
many authors (Miller 1896, Goin 1943, Guilday and Bender 1960,
Guthrie 1965, Corbet 1975). One frequently observed anomaly is the
presence of six triangles instead of five on the first lower molar. Miller
(1896) found this anomaly in 4 (8.3%) of the 285 specimens of M. penn-
sylvanicus he examined. We found the variant in 4 (28.6%) of the 14
specimens of M. p. pennsylvanicus from Georgia and 2 (14%) of the 14
specimens from Maryland. In two specimens from Georgia one side had
six triangles while the other side had five. All specimens of M. p.
dukecampbelli examined had six triangles on the first lower molar.
However, one specimen (UF 12001) had six triangles' on one side and
seven on the other. Other features in the pattern of reentrant folds of the
remaining molar teeth were extremely variable. Therefore, it must be
kept in mind that the pattern of reentrant folds in Microtus, especially of
the first lower molars, is normally variable.
Although all the M. p. dukecampbelli examined had at least six
Vol. 28, No. 2
WOODS ET AL.: M. PENNSYLVANICUS
triangles on the first lower molar, we are hesitant to consider this
character diagnostic of the subspecies. The small sample size of our
population, as well as normal dental variation that is characteristic of
Microtus, cause us to doubt the stability of this character. In addition,
Corbet (1975) found that a similar dental characteristic in a disjunct
population of Clethrionomys glareolus was ephemeral. When Corbet
reexamined the population he described 15 years earlier, he found that
the dental variant he discussed had returned to the normal condition. The
high percentage of six triangles on the first lower molar of M. p. dukecamp-
belli could be the result of "bottleneck effect" as a consequence of periodic
reductions in population size to a few individuals. An alternative explana-
tion is that small population size results in inbreeding, increasing the fre-
quency of a polygenic character. We favor the latter hypothesis in the
light of the finding of Nei et al. (1975) of the inability of an evolutionary
bottleneck to significantly alter the genetic structure of populations.
Guthrie's (1965, 1971) theory that the high level of variation observed
in the dental morphology of Microtus pennsylvanicus is the result of rapid
evolution may be applicable to M. p. dukecampbelli, for it has diverged
in morphology from standard populations of M. p. pennsylvanicus and
may be subjected to strong selective pressures. The continued high level of
dental variation in spite of an apparent reduction in genetic variability in
M. p. dukecampbelli would be a result of the breakdown of balanced
heterozygous linkage groups if this vole is part of a rapidly evolving
population (Guthrie 1965). Corbet (1963) showed that two populations of
Clethrionomys glareolus living in adjacent habitats having different
vegetation types differed greatly in tooth complexity. Guthrie (1971) con-
cluded that dental morphology is directly tied up with the demands of the
diet, and teeth are sensitive indicators of dietary shifts. Therefore, the
combination of increased tooth complexity and a high level of dental
variability could be the result of intense natural selection related to the
diet of M. p. dukecampbelli in a salt marsh characterized by abrasive
plant materials.
The lack of genetic variability in association with the observed dental
variability is difficult to explain, however, and raises the probability that
some dental variability could result from mechanical factors such as tooth-
wear. Guthrie (1971) was not able to demonstrate a high level of heredi-
tability in the pattern of reentrant folds in the areas of molar teeth that
are most variable. Great care should be taken in using subtle differences
in dental patterns as taxonomic characters.
SURVIVAL IN A TIDAL SALTMARSH ENVIRONMENT
The means by which the Florida saltmarsh vole survives tidal flooding
are unknown. Tides at the study site frequently cover the marsh with up
1982
BULLETIN FLORIDA STATE MUSEUM
to 25 cm of water, and occasional natural disasters have flooded the area
with wind-driven water. Harris (1953) posed a similar question for M. p.
nigrans living in tidal marshes in Maryland. One significant difference be-
tween Maryland and Florida is the presence of large populations of musk-
rats (Ondatra zibethicus macrodon) in Maryland marshes. Harris (1953)
demonstrated that Microtus frequently use Ondatra "houses" (feeding
shelters, abandoned houses, as well as occupied houses). The usage
became especially frequent at times of high water. After the marsh had
been covered by 5 cm of water for several days, vole sign was found at 25
percent of muskrat houses, whereas when the marsh was covered with 30
cm of water sign was recorded at 72 percent of 60 muskrat houses ex-
amined. Muskrat houses therefore are important in the survival of Micro-
tus in Maryland and may partially explain why Microtus is more common
in tidal marshes in Maryland than it is in Waccasassa Bay.
The habitat of the California vole, Microtus californicus, is subjected
to daily tidal submergence during the winter months, and that species
therefore faces the same problem of survival faced by voles in Florida.
Fisler (1961) investigated how small mammals survive under these ex-
treme conditions and observed that Microtus californicus climbs into
vegetation and remains well concealed just above the rising water level.
The voles sit in a humped-back position with their tails tucked under their
bodies, and for the most part are completely dry. He also observed mice
sitting on boards and under the bark of floating logs. When the water
became higher than the vegetation itself, the mice were able to swim.
Fisler observed that even when forced to swim, the animals stayed within
their home ranges. They are able to do this even under difficult condi-
tions, and Fisler reported examples of Microtus swimming through rough
water and against strong tidal currents. He reported observing one
Microtus swim a 12.3 m slough, diving to a depth of more than 1.3 m and
swimming a distance of about 6.8 m under water. Therefore, in spite of a
high mortality rate that occurs when the occasional high tide inundates
all the marsh vegetation, Microtus californicus were capable of remaining
on their home ranges through the winter high tides and successfully main-
tained the adult population in the marshes during this critical period
(Fisler 1961).
In Florida, survival during high tides and at times of extreme inunda-
tion from wind-driven water off the shallow Gulf might be accomplished
in several ways: (1) voles might move to high land adjacent to the marshes
or onto the occasional islands in the marshes; (2) they might climb up into
the tops of vegetation to stay above the water level; (3) they might have an
unusual ability to withstand becoming watersoaked and surviv-
ing long periods under these conditions while swimming about; and (4)
they might have an extensive distribution throughout the saltmarsh area
Vol. 28, No. 2
WOODS ET AL.: M. PENNSYLVANICUS
and depend on frequent recolonization of marginal habitats most severely
affected by tides and storms. All of these factors must be at least partially
true. Harris (1953) found vole sign on nearby islands in Maryland and
also found voles clinging to branches of the high tide bush (Ivafrutescens).
In addition, Microtus pennsylvanicus is well known for its ability to with-
stand adverse conditions related to water and cold. The total distribution
of Microtus in the area of Waccasassa Bay is unknown at this time because
of the extreme difficulty of trapping in the area and because the vole is (or
was for the trapping period) so rare. During the entire study period, 31
separate Microtus were tagged. However the study area is in a suitable
position to be sheltered from wind-driven levels of high water, and may
represent a refugium from which Microtus spreads out into the more ex-
posed areas of marsh. Unlike the situation in Maryland, however, where
the distribution of Microtus includes areas of suitable habitat in interior
sections which also could serve as sources of recolonization following
natural disasters, the Florida saltmarsh vole seems to be totally restricted
to the vulnerable saltmarsh habitat.
It is clear from the above discussion that Microtus pennsylvanicus sur-
vives in Florida salt marshes under much more severe conditions and
restrictions than it does in Maryland and faces many of the same problems
as Microtus californicus. It must survive by a variety of mechanisms, but
the two most likely conditions are its ability to climb the saltmarsh vegeta-
tion to get above high water levels (even swimming from plant to plant
when the need arises) and the unique, somewhat sheltered position of the
study site in Waccasassa Bay. However the population is extremely
vulnerable to the effects of tides and weather. The very existence and con-
tinued survival of M. p. dukecampbelli as a relict population under such
conditions for such a long period of time is an evolutionary and ecological
paradox.
LITERATURE CITED
Anderson, S. 1959. Distribution, variation and relationships of the montage vole, Microtus
montanus. Univ. Kansas Mus. Nat. Hist. 9: 415-511.
1961. A relict population of Microtus pennsylvanicus in southwestern New
Mexico. Am. Mus. Nat. Hist. Novit. 2034: 1-3.
and J. P. Hubbard. 1971. Notes on geographic variation of Microtus pennsyl-
vanicus (Mammalia, Rodentia) in New Mexico and Chihuahua. Am. Mus. Nat. Hist.
Novit. 2460: 1-8.
Arata, A. A. 1961. Meadow vole (Microtus pennsylvanicus) from the Quaternary of Florida.
Quart. J. Fla. Acad. Sci. 24: 117-121.
Beck, M. L., and J. T. Mahan. 1978. The chromosomes of Microtus pinetorum. J. Hered.
69: 343-344.
Blackwelder, B. W., 0. H. Pilkey, and J. D. Howard. 1979. Late Wisconsinan sea levels
1982
50 BULLETIN FLORIDA STATE MUSEUM Vol. 28, No. 2
on the southeast U.S. Atlantic Shelf based on in-place shoreline indicators. Science
204: 618-620.
Blair, W. F. 1958. Distribution pattern of vertebrates in the southern United States in rela-
tion to past and present environments. Pp. 433-468 in Zoogeography, C. L. Hubbs (ed.).
Amer. Assoc. Adv. Sci., Washington, D. C. 509 p.
Bradley, W. G., and E. L. Cockrum. 1968. A new subspecies of the meadow vole (Microtus
pennsylvanicus) from northwestern Chihuahua, Mexico. Am. Mus. Nat. Hist. Novit.
2325: 1-7.
Bryson, R. A., D. A. Baerreis, and W. M. Wendland. 1970. The character of late glacial and
postglacial climatic changes. Pp. 53-74 in Pleistocene and Recent environments of the
central Great Plains, D. Wakefield and J. K. Jones, Jr. (eds.). Univ. Press Kansas. Spec.
Publ. 3. 433 p.
Bullen, R. P. 1958. Six sites near the Chattahoochee River in the Jim Woodruff Reservation
area. Florida Bull. Bur. Am. Ethnol. 169: 316-357.
Case, T. J. 1978. A general explanation for insular body size trends in terrestrial vertebrates.
Ecology 59: 1-18.
Corbet, C. B. 1963. An isolated population of bank vole Clethrionomys glareolus with aber-
rant dental pattern. Proc. Zool. Soc. London 140: 315-319.
1975. Examples of short- and long-term changes of dental pattern in Scottish
voles (Rodentia, Microtinae). Mammal Rev. 5: 17-21.
Dalquest, W. W., E. Roth, and F. Judd. 1969. The mammal fauna of Schulze Cave, Edwards
County, Texas. Bull. Florida State Mus., Biol. Sci. 13(4): 205-276.
Dansereau, P. 1957. Biogeography-An ecological perspective. Ronald Press, New York.
394 p.
Fernald, E. A. (ed.). 1981. Atlas of Florida. Florida State Univ. Foundation, Tallahassee.
276 p.
Fisler, G. F. 1961. Behavior of salt-marsh Microtus during winter high tides. J. Mamm. 42:
37-43.
Fivush, B., R. Parker, and R. H. Tamarin. 1975. Karyotype of the beach vole, Microtus
breweri, an endemic island species. J. Mamm. 56: 272-273.
Frey, R. W., M. R. Voorhies, and J. D. Howard. 1975. Estuaries of the Georgia coast,
U.S.A.: Sedimentology and Biology. VIII Fossil and Recent skeletal remains in Georgia
estuaries. Senckenbergiana Maritima 7: 257-295.
Getz, L. L. 1961. Home ranges, territory, and movement of the meadow vole. J. Mamm.
42: 24-36.
Goin, 0. B. 1943. A study of individual variation in Microtus pennsylvanicus pennsylvanicus.
J. Mamm. 24: 212-223.
Golley, F. B. 1962. Mammals of Georgia. Univ. Georgia Press, Athens. 218 p.
Graham, R. W. 1976. Late Wisconsin mammalian faunas and environmental gradients of the
eastern United States. Paleobiology 2: 343-350.
Grinnell, J. 1913. The species of the mammalian genus Sorex of west-central California, with
a note on the vertebrate palustrine faunas of the region. Univ. California Publ. Zool.
10: 179-195.
1932. A relic shrew from central California. Univ. California Publ. Zool. 38:
389-390.
Guilday, J. E., and M. S. Bender. 1960. Late Pleistocene records of the yellow-cheeked vole,
Microtus xanthognathus (Leach). Ann. Carnegie Mus. 35: 315-329.
Guthrie, R. D. 1965. Variability in characters undergoing rapid evolution, an analysis of
Microtws molars. Evolution 19: 214-233.
1971. Factors regulating the evolution of microtine tooth complexity. Zeits.
Saugetierkunde 36: 37-54.
Hall, E. R. 1981. The mammals of North America. 2nd Edition. Vol. II. Wiley, New York.
Pp. 601-1181 + 1-90.
WOODS ET AL.: M. PENNSYLVANICUS
Handle, C. 0., Jr., and C. P. Patten. 1947. Wild mammals of Virginia. Virginia Comm.
Game Inland Fish., Richmond. 220 p.
Harris, A. H., R. A. Smartt, and W. R. Smartt. 1973. Cryptotis parva from the Pleistocene of
New Mexico. J. Mamm. 54: 512-513.
Harris, V. T. 1953. Ecological relationships of meadow voles and rice rats in tidal marshes.
J. Mamm. 34: 479-487.
Hsu, T. C., and K. Benirschke (eds.). 1967. An atlas of mammalian chromosomes. Springer-
Verlag, New York. Vol. I, Folio 15.
Jensen, J. N., and D. I. Rasmussen. 1971. Serum albumins in natural populations of Pero-
imyscus. J. Mamm. 52: 508-514.
and 1972. Erratum in serum albumins in natural populations of
Peromyscus. J. Mamm. 53: 413.
Jolly, G. M. 1965. Explicit estimates from capture-recapture data with both death and immi-
gration-stochastic model. Biometrika 52: 225-247.
Kilpatrick, C. W. 1981. Genetic structure of insular populations. Pp. 28-59 in Mammalian
population genetics, M. H. Smith and J. Joule (eds.). Univ. Georgia Press, Athens. 380 p.
Laerm, J. A., B. J. Freeman, L. J. Vitt. J. M. Meyers. and L. Logan. 1980. Vertebrates of
the Okefenokee Swamp. Brimleyana 4: 47-73.
Lee, M. R. 1969. A widely applicable technic for direct processing of bone marrow for
chromosomes of vertebrates. Stain Technol. 44: 155-158.
Madison, D. M. 1980. Space use and social structure in meadow voles, Microtus pennsyl-
vanicus. Behav. Ecol. Sociobiol. 7: 65-71.
Martin, R. A. 1968. Late Pleistocene distribution of Microtuos pennsylvanicu.s. J. Mamm. 49:
265-271.
and S. D. Webb. 1974. Late Pleistocene mammals from Devil's Den Fauna,
Levy County. Pp. 114-145 in Pleistocene mammals of Florida, S. D. Webb (ed.). Univ.
Presses Florida, Gainesville. 270 p.
Matthey, R. 1952. Chromosomes de Muridae (Microtinae et Cricetinae). Chromosoma 5:
113-138.
1955. Nouveaus documents sur les chromosomes des Muridae. Problems de
cytologie compare et de taxonomie chez les Microtinae. Rev. Suisse Zool. 62: 163-206.
McClenaghan, L. R., Jr., 1980. The genetic structure of an isolated population of Sigmodon
hispidus from the lower Colorado River valley. J. Mamm. 61: 304-307.
Miller, G. S., Jr. 1896. Genera and subgenera of voles and lemmings. North American
Fauna 12: 1-78 + 3 pl.
Negus, N. C., E. Could, and R. K. Chipman. 1961. Ecology of the rice rat Oryzomys
palustris (Harlan), on Breton Island, Gulf of Mexico, with a critique of social stress
theory. Tulane Stud. Zool. 8: 93-123.
Nei, M., T. Maruyama, and R. Chakraborty. 1975. The bottleneck effect and genetic
variability in populations. Evolution 29: 1-10.
Odum, E. P. 1948. Microtus from the Piedmont of Georgia. J. Mamm. 29: 74.
Post, W. 1981. The influence of rice rats Oryzomys palustris on the habitat use of the seaside
sparrow Ammospiza maritima. Behav. Ecol. Sociobiol. 9: 35-40.
Selander, R. K., M. H. Smith, S. Y. Yans, W. E. Johnson, and J. B. Gentry. 1971. Bio-
chemical polymorphism and systematics in the genus Peromyscus. I. Variation in the
old-field mouse (P. r.,,11,, .. ,, polionotus). Stud. Genet., Univ. Texas 7103: 50-90.
Smith, A. T., and J. M. Vrieze. 1979. Population structure of Everglades rodents: Responses
to a patchy environment. J. Mamm. 60: 778-794.
Snyder, D. P. 1954. Skull variation in the meadow vole (Microtus pennsylvanicus) in Penn-
sylvania. Ann. Carnegie Mus. 33: 201-234.
Von Bloeker, J. C., Jr. 1932. Three new mammals from salt marsh areas in southern Cali-
fornia. Proc. Biol. Soc. Wash. 45: 131-138.
52 BULLETIN FLORIDA STATE MUSEUM Vol. 28, No. 2
Watts, W. A. 1971. Postglacial and interglacial vegetation history of southern Georgia and
central Florida. Ecology 52: 676-690.
____ 1980. The late Quaternary vegetation history of the southeastern United States.
Ann. Rev. Ecol. Syst. 11: 387-409.
Webb, S. D. 1977. Evolution of savanna vertebrates in the New World. Part 1: North
America. Ann. Rev. Ecol. Syst. 8: 355-380.
Webb, T., III, and R. A. Bryson. 1972. Late and postglacial climatic changes in the northern
mid-west USA: Quantitative estimates derived from fossil pollen spectra by multivariate
statistical analysis. Quaternary Res. 2: 70-115.
Whittaker, R. H. 1967. Gradient analysis of vegetation. Biol. Rev. 42: 207-264.
Wiens, J. A. 1969. An approach to the study of ecological relationships among grassland
birds. AOU Monog. No. 8.
Youngman, P. M. 1967. Insular populations of the meadow vole, Microtus pennsylvanicus,
from northeastern North America, with descriptions of two new subspecies. J. Mamm.
48: 579-588.
DISTRIBUTION AND EVOLUTION OF FLORIDA'S
TROGLOBITIC CRAYFISHES
RICHARD FRANZ AND DAVID S. LEE'
ABSTRACT: The current knowledge of Florida's troglobitic crayfish fauna is discussed, inter-
pretations of distributional and ecological patterns are reviewed, and an explanation of their
evolutionary history is attempted.
These crayfishes are restricted to certain geological formations that have light to nonexis-
tent plastic overburdens. Areas with moderate to heavy accumulations over the carbonate
rocks lack these crustaceans. The Crystal River Formation, a group of highly soluble Eocene
limestones, is the most important geological element influencing the distribution of most
Florida cave-dwelling crayfishes. Members of the Procambarus lucifugus complex (with the
possible exception of an undescribed species from Lake County), Procambarus pallidus,
Troglocambarus maclanei, and Cambarus cryptodytes, are apparently confined to this for-
mation. The remaining species are confined to other limestones (Procambarus acherontis and
Procambarus species from Lake County in the Hawthorne Formation, Procambarus miller
in the Miami Oolite, Procambarus horsti and Procambarus orcinus in the St. Marks Forma-
tion).
Field observations suggest that available sources of food energy dictate which species
groups inhabit particular cave systems. Species complexes that are most restricted to en-
vironments which provide large accumulations of organic detritus become ecologically and
geographically isolated from other populations and exhibit increased speciation. For example,
members of the Procambarus lucifugus complex seem to have demanding energy budgets and
occur only in localized karst areas exhibiting mature features and high energy input.
Dependency on constant energy supplies provided by large sinkhole entrances and/or bat
roosts probably limits their dispersal ability. In contrast, members of the Procambarus
pallidus complex inhabit systems with limited energy inputs, such as springs and sinkholes
with small openings. Troglocambarus maclanei occurs syntopically with members of the Pro-
cambarus lucifugus and Procambarus pallidus complexes and may be capable of interstital
movements from one cave to another. This may help to explain its extensive distribution.
SUMARIO: El corriente conocimiento de la fauna del astaco troglobitico de Florida es tratado;
las interpretaciones de los models ecologicos y de distribucion son revisados, y se intent dar
una explicacion de su historic evolutiva.
Estos astacos estan restringidos a algunas formaciones geol6gicas que son muy poco o nada
clasticas. Las areas con acumulaciones moderadas a pesadas sobre las rocas carb6nicas
carecen de estos astacos. La formaci6n de Crystal River, un grupo de piedras calizas del
Eoceno, es el element geol6gico mas important que influye en la distribucion de la mayoria
de los astacos habitantes de las cuevas en Florida. Los miembros del complejo de Procambarus
lucifugus (con la possible excepcion de una especie que no esta descrita para el Lago County),
Procambarus pallidus, Troglocambarus maclanei, y Cambarus cryptodytes paracen estar
confinados a esta formaci6n. Las species restantes estan confinadas a otras piedras calizas
IRichard Franz is an Associatc in Natural Sciences. Florida State Museiun. Unietrsia of Florida, Gaines ille 32611; Dalid
S, Lee is Curator of Birds and Manmmals, North Carolina State Munseum. Raleigh 27607,
FRANZ, R., AND D.S. LEE. 1982. Distribution and evolution of Florida's troglobitic cray-
fishes, Bull. Florida State Mus., Biol. Sci. 28(3):53-78.
BULLETIN FLORIDA STATE MUSEUM
(Procambarus acherontis y la especie de Procambarus del Lago County en la formaci6n
Hawthorne; Procambarus miller en el oolito de Miami, Procambarus horsti y Procambarus
orcinus en la formaci6n St. Marks).
Las observaciones de campo indican que la fuente de energia alimenticia disponible deter-
mina cuales grupos de species viven en sistemas particulares de cuevas. Los complejos de
species mis restringidas a grandes acumulaciones de detrito organico llegan e estar aisladas
ecol6gica y geogrificamente de las otras poblaciones y exhiben frecuente especiaci6n. Por
ejemplo, miembros del complejo Procambarus lucifugus parencen tener ciertas demands de
energia por lo que occurred solamente en areas localizadas de Karst, exhibiendo estructuras
maduras y gran gasto de energia. La dependencia de un constant suplemento de energia pro-
vista por grandes entradas a hundimientos y/o perchas de murcielagos problemente limita su
poder de dispersion.
En cambio, miembros del complejo de Procambarus pallidus habitan en sistemas con
limitada energia, como manantiales y hundimientos con pequefias aberturas. Troglocam-
barus maclanei ocurre sint6picamente con miembros de los complejos de Procambarus
lucifugus y Procambarus pallidus y pueden moverse intersticialmente de una cueva a otra.
Esto puede ayudar a explicar su extensa distribuci6n.
TABLE OF CONTENTS
INTRODUCTION ......... ............................................ 54
ACKNOWLEDGEMENTS .................. ................ .......... 55
FLORIDA'S TROGLOBITIC CRAYFISH FAUNA ................................ 55
DISTRIBUTIONAL PATTERNS .................................. .. ......... .. 62
SUGGESTED EVOLUTIONARY HISTORY OF CERTAIN
TRO GLOBITIC SPECIES ......................................... . . 71
LITERATU RE C ITED . ........... ........................................ 76
INTRODUCTION
Florida has the most diverse troglobitic crayfish fauna in the world. In
this paper we will attempt to document their zoogeographic and
ecological relationships and explain the reasons for the presence of this
unusually rich fauna. The information presented here is based partly on
existing literature, which is mostly of a taxonomic-distributional nature,
and on our 17 years of field observations.
Our interest in the problems regarding the distributional patterns in
Florida's troglobitic crayfishes was brought into sharp focus in 1964 while
collecting Procambarus in northcentral Alachua County. We found that
two caves, less than one kilometer apart, were inhabited by different
species. These species, Procambarus pallidus and P. lucifugus, occurred in
different densities, were distributed distinctly within the caves, and their
populations were composed of proportionately different age classes. It is
fortunate that our investigations started in these particular caves, for as
our interest and area of field activities expanded we found ourselves com-
paring later observations to these two contrasting populations.
Through the efforts of Horton H. Hobbs, Jr., the Florida troglobitic
crayfish fauna is now well known. Prior to his monumental works on
Florida crayfishes in the late 1930's, only one troglobitic species, Cam-
Vol. 28, No. 3
FRANZ AND LEE: TROGLOBITIC CRAYFISHES
barus (= Procambarus) acherontis Lonnberg, was known from the state.
Between 1940 and 1942, Hobbs described five additional cave
forms- Cambarus (= Procambarus) lucifugus lucifugus, Cambarus
( = Procambarus) lucifugus alachua, Cambarus ( = Procambarus)
pallidus, Troglocambarus maclanei, and Cambarus cryptodytes. In 1971
and 1972, three additional cave species were described from the penin-
sula: Procambarus miller Hobbs (1971), Procambarus orcinus Hobbs and
Means (1972), and Procambarus horsti Hobbs and Means (1972). These
discoveries brought about renewed interest in sinkhole and spring
crayfishes. As a result, three other troglobitic species were recently
discovered. Of these, Procambarus erythrops Relyea and Sutton (1975)
and Procambarus franzi Hobbs and Lee (1976) are formally named; the
other must await the discovery of Form I males before it can be described.
In this paper we summarize previously published information, present
additional data gathered by us and others, and provide an analysis of the
accumulated distributional and ecological data.
ACKNOWLEDGEMENTS
Much of the information presented in this report was obtained through the generosity of
others. Cavers and cave divers, particularly Paul Heinerth, Brian Houha, the late Bill Hurst,
Buford Pruitt, and Paul Smith, provided invaluable observations and occasional specimens.
Mr. Heinerth enthusiastically collected numerous specimens and detailed field notes from im-
portant and often dangerous cave systems. We also wish to thank: Patricia Ashton, Ray
Ashton, Jon Baskin, Lea Franz, Michael Frazier, Barbara Lee, A.T. Leitheuser, and
numerous other friends who assisted us in the field; land owners, particularly Mr. and Mrs.
Marion Bishop of Newberry and Mr. Junior Kelly of Bell, for their hospitality while working
on their properties; Thomas Scott, Bureau of Geology in Tallashassee, Florida, for well-core
data and general help with local geology; Rhoda J. Bryant, Carter R. Gilbert, Bruce J. Mac-
Fadden, Fred G. Thompson, and S. David Webb, Florida State Museum, for guidance in the
preparation of the manuscript; Helen S. Bates, Nancy R. Halliday, Eugene Hanfling, and M.
Glen Rogers, Florida State Museum, for help in the preparation of the figures; and William
Young, Altamonte Springs, and Walter Wood, U.S. Geological Survey, for data on Procam-
barus acherontis. We are grateful to John E. and Martha R. Cooper, North Carolina State
Museum, for their enthusiasm and guidance of our crayfish studies. We especially thank
Horton H. Hobbs, Jr., for openly sharing his ideas with us, providing identifications, and en-
couraging our activities from their earliest stages.
Field work was, in part, sponsored through a contract from the U.S. Fish and Wildlife
Service (#SFWB 115344) to study the impact of phosphate mining on the Osceola National
Forest and the upper Suwannee River. We wish to acknowledge the late Howard W.
Campbell, former Chief of the Gainesville Lab. and Steven P. Christman, Project Leader,
of the U.S. Fish and Wildlife Service, for their assistance and support during this phase of
the project.
FLORIDA'S TROGLOBITIC CRAYFISH FAUNA
Eleven troglobitic crayfishes belonging to the genera Cambarus, Pro-
cambarus, and Troglocambarus are reported from the state, representing
the most diverse, known troglobitic crayfish fauna. The general distribu-
tion of each species is shown in Figures 1-4. For convenience we have
1982
BULLETIN FLORIDA STATE MUSEUM
FIGURE 1. Distribution of (1) Procambarus lucifugus alachua. (2) Procambarus lucifugus
lucifugus. Unnumbered populations are considered intergrades.
organized them into six groups (in part following Hobbs et al. 1977). The
following list shows the species included in each group, attempts to briefly
illustrate their relationships with other crayfishes, discusses their general
distributions, and gives a list of additional localities not found in Hobbs et
al. (1977).
GROUP A.-The genus Cambarus is represented by Cambarus
(Jugicambarus) cryptodytes Hobbs. Hobbs and Barr (1960) and Hobbs et
al. (1977) considered this a relict species, with its closest relatives occur-
ring in Tennessee and the Ozark Mountain areas.
Cambarus cryptodytes is the only Florida troglobitic crayfish known
to occur outside the state. This species has been observed at 15 localities in
Jackson County, Florida, and one in Decatur County, Georgia (Fig. 5).
Its range may extend farther northward in Georgia along the Flint River
to the vicinity of Albany, Dougherty County, but no specimens are
available from this area. This presumption is based on the occurrence of
Vol. 28, No. 3
FRANZ AND LEE: TROGLOBITIC CRAYFISHES
FIGURE 2. Distribution of (1) Cambarus cryptodytes. (2) Procaimbarus crythrops. (3) Pro-
camnbarus franzi. (4) Procaimbarus species.
the cave salamander Haideotriton wallacei at Albany (Carr 1939). This
salamander is associated with C. cryptodytes at practically all known
localities in Jackson and Decatur counties.
In Florida the distribution of C. cryptodytes lies entirely within the
northern two-thirds of Jackson County in a physiographic unit, the
Marianna River Valley Lowlands, of the Coastal Plain Province (as shown
in Moore 1955). All populations occur within the Chipola River drainage,
except the Graceville site (type .locality) which is in the Holmes Creek
basin of the Choctawhatchee River drainage. All but the Graceville
population were found in caves; the series from Graceville was taken from
an open well (Hobbs et al. 1977).
ADDITIONAL RECORDS: JACKSON COUNTY-Cave in the Woods (Sec. 22,
T.5N, R.11W), Hole in Wall (Sec. 5, T.4N, R.9W), Twin Cave (Sec. 6,
T.4N, R.9W).
GROUP B.-One species, Procambarus (Lonnbergius) acherontis
1982
BULLETIN FLORIDA STATE MUSEUM
00
FIGURE 3. Distribution of (1) Troglocambarus maclanei, (2) Procambarus acherontis, (3)
Procambarus miller.
described by Lonnberg in 1894, is apparently unrelated to any other liv-
ing species of the genus (Hobbs 1972; Hobbs et al. 1977). It is known from
one site in Orange County and three in Seminole County. The Altamonte
Springs specimen (Form II male, carapace length [CL] 24 mm) was ex-
amined by R. Franz (December 1977); however, the specimen was re-
tained by Mr. William Young, Young's Well Drilling, Altamonte Springs,
Florida.
ADDITIONAL RECORDS: ORANGE COUNTY-Long Lake well (Sec. 36,
T.21S, R.28E); SEMINOLE COUNTY-Altamonte Springs well (Sec. 13,
T.21S, R.29E).
GROUP C.-This group contains one species, Procambarus (Leconti-
cambarus) miller Hobbs, which is closely related to the epigean species,
Procambarus alleni (Faxon). It is the only known troglobite in this
subgenus (Hobbs, 1971) and is only recorded from the type locality, a well
Vol. 28, No. 3
FRANZ AND LEE: TROGLOBITIC CRAYFISHES
FIGuRE 4. Distribution of (1) Procambarus orcinus, (2) Procambarus horsti, (3) Procam-
barus pallidus. Locality 3 with arrow is a questionable record for Procambarus pallidus
from Eickelburger Cave, Marion County.
in Miami, Dade Cbunty. No additional information is available beyond
the original description.
GROUP D.-Five members of the P. lucifugus complex of the subgenus
Ortmannicus (Pictus Group, as defined by Hobbs, 1942) make up this
group. Species in the complex appear closely related to Procambarus pic-
tus Hobbs, a stream-inhabiting species endemic to the Black Creek
drainage of northeastern Florida (Hobbs et al., 1977; Franz and Franz,
1979). The group includes Procambarus lucifugus lucifugus (Hobbs), P.
lucifugus alachua (Hobbs), P. erythrops Relyea and Sutton, P. franzi
Hobbs and Lee, and one undescribed species, Procambarus sp. (Hobbs
and Lee, 1976; Hobbs et al., 1977; Hobbs, pers. comm.). The five
members of the lucifugus complex occur mostly in small, well-defined
geographical areas with no overlap in their ranges (Figs. 1-2).
1982
BULLETIN FLORIDA STATE MUSEUM
ALA.
SMalone
HOLMES Graceille Compbellton
COUNTY *
Greenwood GA.
.. . .. . ---------- 0 01
WASHINGTON i- .
COUNTY ..... Cottondale r-,o -
Marianna --------
El Alford
Sneads
GADSDEN
COUNTY
0 5 10 15 20
-- ---- -------.. -__ __.__._-_ _._ 0 2 4 6 8 10 12 i
BAY COUNTY CALHOUN COUNTY M, le,
FIGURE 5. Specific records (solid circles) of Cambarus cryptodytes in Jackson County show-
ing its relationship with the Marianna Limestone (darker band) and the Crystal River Forma-
tion (lighter band). Circles may represent more than one site. Geological features are from
Moore (1955).
Procambarus lucifugus, as presently understood, has the largest range
of any member in the lucifugus complex, extending from western Levy,
Gilchrist, and Alachua counties, southward through Marion County to
Citrus and Hernando counties. This seemingly large range, however, is
not continuous, but consists of five apparently isolated populations.
Populations 1 and 2, known from three caves near Bell in western
Gilchrist County and one cave in Levy County, appear to be isolated from
the population in adjacent Alachua County. After examination of several
females, Hobbs (pers. comm.) believes them to be more similar to the
nominate subspecies than to the adjacent P. 1. alachua populations.
Resolution of their affinities must await the collection of Form I males. In
Alachua County population 3 (or Procambarus lucifugus alachua) is con-
fined to a small area extending a few kilometers north and south of a line
drawn between Gainesville and Newberry. Population 4 lies in
southwestern Marion County, approximately 50 km south of the Alachua
population. Hobbs (1942) suggested that this Marion population is in-
termediate between that in Alachua County and those in Citrus and Her-
nando counties. His taxonomic decision was tentative, awaiting the
collection of Form I males from the intermediate area (Hobbs, 1942). Col-
Vol. 28, No. 3
FRANZ AND LEE: TROGLOBITIC CRAYFISHES
elections made by us at Sunday Sink and Ocala Caverns include many
Form I males, and examination of these by Hobbs bears out his original
conclusion. The intermediate population is completely allopatric and lies
between P. 1. alachua and P. 1. lucifugus. Population 5 (or P. 1. lucifugus)
occurs in Citrus and Hernando counties, west of the Withlacoochee River
and southwest of the Marion County population. Our knowledge of this
population is limited. The original accounts were provided by Hobbs
(1940, 1942), and we can only add that Paul Heinerth (pers. comm.)
observed small crayfishes, presumably this form, in a sinkhole near
Weekiwachee Springs, Hernando County.
In addition to Procambarus lucifugus, four other species belong to the
P. lucifugus complex. Procambarus erythrops is endemic to a small area in
southern Suwannee County. Only four sites have been located, although
several dozen sinks in the area have been examined by us for crayfish. This
population lies in a small karsted pocket north of the Santa Fe River, east
of the Suwannee River, and west of the Ichetucknee Springs Run. Thus,
P. erythrops is 14 km north of the Gilchrist population of P. lucifugus.
Procambarus franzi occurs in northern Marion County between two
populations of P. lucifugus, and Hobbs and Lee (1976) suggested a close
kinship between the two species. The two known sites are separated by 7
km. Only two specimens of Procambarus sp. are available at this time,
both from Alexander Springs, Lake County (see Relyea and Sutton, 1976,
for the history of their discovery).
ADDITIONAL RECORDS: Procambarus franzi: MARION COUNTY-Hell Hole
(Sec. 6, T.12S, R.21E). Procambarus lucifugus: ALACHUA COUNTY-Tusk
Cave (Sec. 27, T.9S, R.18E); GILCHRIST COUNTY-Kelly Sinks (Sec. 34,
T.8S, R.14E), Old Walker Farm Sinks (Sec. 35, T.8S, R.14E); LEVY
COUNTY-Manatee Springs (Sec. 26, T.11S, R.13E); MARION COUNTY-
Ocala Caverns (Sec. 23, T.16S, R.22E), Redding Catacombs (Sec. 20,
T.16S, R.22E).
GROUP E.-Three species, also in the subgenus Ortmannicus of the
Pictus Group, comprise this group. Hobbs et al. (1977) believed that
members of the Procambarus pallidus complex are closely related to Pro-
cambarus lepidodactylus Hobbs, a stream-inhabiting species that occurs
in northeastern South Carolina and adjacent North Carolina. The two
species, Procambarus horsti Hobbs and Means and Procambarus orcinus
Hobbs and Means, are more closely related to each other than to Procam-
barus pallidus (Hobbs and Means, 1972). The three together, however,
form a natural group (Hobbs, pers. comm.).
Procambarus horsti and Procambarus orcinus are completely
allopatric to each other and to all other troglobitic crayfishes and occur in
the area south of Tallahassee (in Jefferson, Leon, and Wakulla counties).
P. orcinus is known from five localities within the Woodville Karst, while
1982
BULLETIN FLORIDA STATE MUSEUM
P. horsti has been found in only one site, Big Blue Spring, on the Wacissa
River. The range of P. pallidus includes western Alachua County, one
cave just south of the Alachua-Levy county line, in Levy County, and the
karst following the Santa Fe and upper Suwannee rivers. Its range
overlaps with members of the lucifugus complex and with Troglocam-
barus maclanei in Alachua and Suwannee counties (Fig. 6). It has also
been reported from one cave in southwestern Marion County (listed by
Hobbs et al., 1977), but we question this record (Fig. 4). The specimen
was supposedly collected by Richard Warren in Eickelburger Cave, a site
which has since been destroyed by quarrying activities (Florida
Speleological Society, pers. comm.). Hobbs (pers. comm.) has recently re-
examined this specimen at our request and has found it indeed to be Pro-
cambarus pallidus. We question the record because it is 40 km south of
the closest known locality (Archer Caves), and because no additional
specimens have been collected from caves close to the Eickelburger Cave
locality, even though there has been extensive sampling in that area. We
assume that the specimen was collected elsewhere and was inadvertently
placed with the Eickelburger Cave samples.
ADDITIONAL RECORDS: Procambarus pallidus: ALACHUA COUNTY-Ala-
chua Sink (Sec. 10, T.8S, R.18E), 32 Foot Cave (Sec. 18, T.10S,R.19E);
GILCHRIST COUNTY-Devil's Eye Spring (Sec. 34, T.7S, R.16E), Jinnie
Spring (Sec. 34, T.7S, R.16E); LAFAYETTE COUNTY-Troy Springs (Sec.
34, T.5S, R.13E); LEVY COUNTY Archer Caves (Sec. ?, T.11S, R.17E);
MADISON COUNTY- Thunderhole (Sec. 10, T.1N, R.10E); SUWANNEE
COUNTY-Little River Spring (Sec. 1, T.6S, R.13E).
GROUP F.- Troglocambarus maclanei Hobbs is a highly specialized
troglobitic species derived from the Pictus Group and thought to be
closely related to Procambarus ancylus of the seminolae subgroup (Hobbs
et al., 1977). P. ancylus, a lentic species, presently occurs in southeastern
North Carolina and adjacent South Carolina (Hobbs, 1974).
Troglocambarus maclanei is known from seven localities in Alachua,
Citrus, Marion, and Suwannee counties (Fig. 3). Its range spans both the
Santa Fe and southern Withlacoochee River basins and completely
overlaps the range of the lucifugus complex and a portion of the pallidus
complex. There is no morphological evidence of differentiation.
ADDITIONAL RECORDS: MARION COUNTY-Sunday Sink (Sec. 24, T.6S,
R.22E), Orange Lake Cave (Sec. 34, T.12S, R.21E).
DISTRIBUTIONAL PATTERNS
The distributional patterns of Florida troglobitic crayfishes have been
Vol. 28, No. 3
FRANZ AND LEE: TROGLOBITIC CRAYFISHES
FIGURE 6. Specific records for cave crayfishes in the Suwannee River basin, showing their
relationship to the 15 and 30m contours. Solid circles = Procambarus pallidis, hollow
squares = Procambarus lucifugus (population 1), solid triangles = Procambarus franzi,
hollow triangles = Procambarus erythrops, hollow circles = Procambarus lucifugus
alachua (population 2).
the subject of portions of several papers (Hobbs 1942, 1958; Hobbs et al.
1977; Caine 1974; Relyea et al. 1976). However, we can now make addi-
tional interpretations, based on particular geological features and certain
ecological factors.
EFFECTS OF OVERBURDEN
An important distributional phenomenon is the lack of troglobitic
crayfishes in areas where moderate to heavy accumulations of uncon-
solidated sediments overlie the limestone (Fig. 7). This affects all of the
troglobitic species by preventing the penetration of organic matter to the
aquifer. Sediments, even coarse sands, filter detritus, and only through
cave entrances and sinkholes is trophic input possible. Subterranean areas
covered with thick accumulations of clastic overburden are thus severely
energy limited.
BULLETIN FLORIDA STATE MUSEUM
FIGURE 7. Distribution of moderate to heavy accumulations of plastic materials overlying
carbonate rocks (hatched areas on map).
PHYSIOGRAPHIC FEATURES
Local distributions of certain crayfishes can be related to the escarp-
ment of the Northern Highlands, low hills composed of residual plastics
and rivers. The escarpment of the Northern Highlands runs roughly
northwest to southeast, from near Tallahassee to Gainesville, and approx-
imates the 30 m contour. It is a narrow band of rolling hills composed of
plastics and characterized by extensive sinkhole activity. Streams flowing
west from the Northern Highlands are quickly swallowed by these
sinkholes (Williams et al. 1977). The most spectacular example in the
northern peninsula is the disappearance of the Santa Fe River into a large
sinkhole at Oleno State Park and its resurgence some 5 km away. West of
the escarpment are karsted regions (e.g. western Alachua County and
southern Columbia and Suwannee counties) inhabited by cave crayfishes
(Fig. 6). In western Alachua County, Procambarus pallidus and P.
Vol. 28, No. 3
FRANZ AND LEE: TROGLOBITIC CRAYFISHES
lucifugus alachua appear to inhabit different portions of this karst plain
with respect to the escarpment. P. pallidus occurs primarily in "newly"
developing cave systems along the escarpment (Warrens Cave, Fort Clark
Church, and sites along the Santa Fe and Suwannee rivers), although it is
found in other systems, including three sites where it occasionally has
been known to occur with P. 1. alachua. P. 1. alachua, on the other hand,
appears to be primarily restricted to more mature systems below the
escarpment.
Southwest of the Northern Highlands are several series of low hills
that are apparently remnants of a once more extensive highland. They
resemble the Northern Highlands escarpment in that they are composed
of similar sediments and have similar sinkhole activity. Procambarus
franzi is associated with a set of hills south of Orange Lake in northern
Marion County, and the intergrade population 3 of P. lucifugus with the
hills of southern Marion County.
Caine (1974) attempted to show a correlation between the distribution
of Florida troglobitic crayfishes and elevation, in particular the 15 m con-
tour. However, after plotting crayfish localities, we found them to be be-
tween 3 m elevation at Manatee Springs and 46 m at Fort Clark Church.
Rivers may play a role in the distribution of Florida's troglobitic
crayfishes, although it is not a direct relationship as implied in Relyea
(1976). Rivers are the erosional agents that expose additional limestone. A
species like P. pallidus prefers caves in the "newly" emerging karst, as
seen in its distribution along the Suwannee and Santa Fe River basins
(Fig. 6).
GEOLOGICAL FORMATIONS
Florida troglobitic crayfishes are confined to carbonate rocks, but not
all aquifers in carbonate rocks are inhabited. We can only speculate as to
the factors controlling this apparent selectivity.
Of the geological facies influencing the distribution of these
crayfishes, the Crystal River Formation, a series of highly soluble Eocene
limestones forming the upper portion of the Ocala Group, is the most im-
portant (Fig. 8). Procambarus lucifugus, P. erythrops, P. franzi, P. pal-
lidus, and Troglocambarus maclanei are almost completely confined
to it. However, under special conditions P. pallidus is occasionally found
associated with additional limestones. At Thunder Hole, Suwanacoochee
Springs, and Peacock Slough, P. pallidus is associated in part with the
Suwannee limestone. These three cave systems lie along contacts between
the Suwannee Limestone and the Crystal River Formation. Cambarus
cryptodytes mostly occurs in caves lying near the contact between the
Marianna Limestone and the Crystal River Formation, although the
1982
BULLETIN FLORIDA STATE MUSEUM
SCrystal River Fm.
St. Marks Fm.
Suwannee Limestone
as W, o
FIGuRE 8. Distribution of certain carbonate rocks in Florida. Solid circles on insert map are
crayfish localities. Geological features based on the map by Vernon and Puri (1964).
Graceville and Geromes Cave records are wholly within the latter forma-
tion (Fig. 5). P. acherontis and P. sp. are known from sites in the
Hawthorne Formation, although they may ultimately be found associated
with the Ocala Group (Crystal River and Williston formations). P.
acherontis exists in the Palm Springs-Lake Brantley-Altamonte Springs
region, a karsted area in southwestern Seminole County and adjacent
Orange County. Wells drilled in the Palm Springs-Lake Brantley area
show profiles consisting of varying amounts of Plio-Pleistocene terrace
Vol. 28, No. 3
FRANZ AND LEE: TROGLOBITIC CRAYFISHES
deposits covering 15-18 m of Hawthorne Formation, underlain with
limestone of the Williston and Inglis formations (lower members of the
Ocala Group). The Crystal River Formation is available here as thin beds
in some wells, but is otherwise absent. Water from Palm Springs issues
from two intersecting joints, each less than a meter wide in the
Hawthorne. Crayfishes brought up from the Geological Survey's well at
Long Lake and a private well at Altamonte Springs probably originated
in the Ocala Group, since both of these wells are over 30 m deep. Procam-
barus orcinus and P. horsti are associated with the St. Marks Formation.
The five known sites for P. orcinus are clustered about the town of Wood-
ville, an area referred to as the Woodville Karst. P. horsti is known from a
single locality, Big Blue Spring on the Wacissa River, which is developed
as a deep shaft penetrating the St. Marks at the surface and the Suwannee
Limestone below. The karst area around Woodville is extensive and
separated from that at Big Blue Spring by a wide gap composed of Suwan-
nee Limestone (Fig. 8). In the area about Big Blue Spring, the St. Marks
occurs as a thin, crescent-shaped band lying adjacent to the Cody Scarp.
We hypothesize that at one time the St. Marks Formation was continuous
but, because much of the region was eroded by the Wacissa River, the for-
mation has been divided. With this division we believe that a widespread
subterranean crayfish population became fragmented, eventually dif-
ferentiating into the two present-day species. Hobbs and Means (1972)
suggested a very close relationship between these forms, and their
distribution supports this hypothesis. P. miller is known from one site in
the Miami oolite, but its range remains a mystery.
Although certain geological formations that appear suitable for oc-
cupation by troglobitic crayfishes lie at the surface, none is known from
these formations. The most obvious are the Suwannee and Tampa
limestones. Large areas of the state have exposures of these formations
(Fig. 8), and solution features such as caves, sinks, and springs are com-
mon to them. The lack of troglobitic crayfishes here is intriguing. Banks
(1976) showed these limestones to be generally harder than the Crystal
River and Williston formations, with zones of extremely hard (or dense)
dolomite, breccia, shell, and clay beds. These hard zones may limit
solubility and the extent to which cave systems can develop, and may
restrict sinkhole formation. The latter would prevent access by crayfishes
into the heart of the geological formation; the former would discourage
the downward movement of detritus, and without food the crayfishes
would be unable to survive. Other factors, like the chemical properties of
the waters in the uninhabited aquifers of these rocks, would be worthy
topics of investigation.
BULLETIN FLORIDA STATE MUSEUM
ECOLOGICAL FACTORS
Caine (1978) presented the only ecological data available for a few
species of Florida troglobitic crayfishes. This information has been incor-
porated into our interpretation. As we have no personal field experiences
with either Procambarus acherontis (Group B) or P. miller (Group C),
our discussion is limited to those in other groups, but we suspect that
many of our observations may also apply to these two species.
The distribution of Florida's cave crayfishes within cave systems is
clearly related to the availability of food. This food consists of either plant
or animal debris washed in through sinks from surface ecosystems or
guano under bat roosts.
Two bats, Myotis austroriparius and M. grisescens, use Florida caves
in sufficient numbers to accumulate guano deposits. Estimates of nursery
colonies of M. austroriparius at Sunday Sink and Orange Lake Cave in-
dicate populations of 14,000 and 17,000, respectively (Zinn, 1977; pers.
comm). The bats are present for several months in late spring and summer
before the young can fly, after which the caves are usually vacated. At
Sunday Sink, the crayfish, estimated at more than 1000 in number, con-
tinue to congregate under the bat roost for approximately two months
after the bats have left. They eventually disperse into deep recesses of the
cave, suggesting that bat guano may be a short term resource. These bats
normally limit their nursery areas to portions of the cave directly over
water, and, since suitable maternity caves are limited, it is reasonable to
assume that the dependence of certain crayfish populations on bat activity
may have been continuous for a long period.
Crayfish distributions fall into two basic patterns with respect to the
dispersal of organic food matter. First, in caves where there is con-
siderable subterranean water movement, incoming food is carried great
distances from the input source, and populations of crayfishes in these
caves appear to be widely, and more or less uniformly, distributed
throughout the aquifer. Second, in areas where water movement and the
energy source is more static, populations are clustered around the limited
organic source.
In the Florida panhandle, Cambarus cryptodytes is known from one
well and 14 caves. Although small to moderate numbers of these
crayfishes concentrate around accumulated organic debris, individuals
can be expected throughout the aquifer. Here the karst is characterized by
shallow meandering cave systems occasionally penetrating below the
water table. Since lateral transport of detritus is much more extensive
than in the peninsular karst, crayfishes are not concentrated at the input
areas but are more widely distributed. Seasonal flooding (usually in
spring) causes ground water to flood otherwise dry caves, and Cambarus
cryptodytes appears to take advantage of rising waters to temporarily
Vol. 28, No. 3
FRANZ AND LEE: TROGLOBITIC CRAYFISHES
move into recently flooded areas for exploitation of newly submerged
food resources. In many caves there is a slow but regular flow of water
throughout most of the year. This movement provides a dependable
method of dispersal for organic matter.
Cambarus cryptodytes does not appear to thrive in aquatic systems
rich in nutrients. In Geromes Cave, where there is an extremely large
guano deposit and little lateral water movement, we only encountered
two Cambarus cryptodytes in 20 visits. However, we found moderate
numbers of a surface-dwelling species, Procambarus paeninsulanus. This
is the only site in Florida where both troglobitic and epigean species were
found together. In Bat Cave there is also an extremely rich supply of
organic material, and Cambarus cryptodytes has never been found here.
Procambarus pallidus, and apparently Procambarus horsti and Pro-
cambarus orcinus (Group E), occur under conditions similar to those
associated with Cambarus cryptodytes. Relatively few individuals (< 50)
are found at energy input areas (small sinks, solution pipes, and springs).
To our knowledge, no population of Group E species has been found
under bat roosts. These crayfishes seem to be unaffected by light and com-
monly occur in "blue hole" sinks, and occasionally in the open, outside of
spring mouths. Within cave systems, Procambarus pallidus is typically
found where potential food accumulates in small, often isolated pockets.
Occasionally, these pockets are in protected areas within major conduits
where strong water currents occur.
Members of the Procambarus lucifugus complex (Group D) occur
most often in collapse sinks and under bat roosts, where large amounts of
organic material are available (Fig. 9). Probably because of the large food
resources, this group maintains the largest populations of the Florida
troglobitic species. The Procambarus erythrops population at Sims Sink is
estimated at 400-500 individuals, and the intergrade Procambarus
lucifugus population at Sunday Sink at more than 1000. In most caves we
studied, the largest populations were in more stagnant portions where
much detritus accumulated and the energy source remained relatively
stable.
Based on these observations we conclude that Procambarus pallidus is
the most competitive and the most successful colonizer of low energy
systems. There are few cases where members of the palliduss" and "lu-
cifugus" groups occur together, even though their overall geographic
distributions indicate that there is considerable opportunity to do so. We
know of only three cases of reported syntopic occurrence: P. pallidus out-
numbers P. lucifugus in Squirrel Chimney and P. lucifugus outnumbers
P. pallidus in Goat and Hog sinks. In each case the rarer species is known
only from one or two individuals that we assume represent vagrants. This,
along with the distribution of Troglocambarus, Crangonyx, and Caeci-
BULLETIN FLORIDA STATE MUSEUM
A. SINK WITH LARGE DRAINAGE BASIN
B. SINK WITH SMALL DRAINAGE BASIN
C. CAVE D. SPRING
LARGE ORGANIC DEPOSITS
SMALL ORGANIC DEPOSITS
PARTICULATE DEPOSITS
BAT GUANO
BAT ROOST
FIGURE 9. Types of openings to aquifers inhabited by Florida cave crayfishes, showing the
distributions of organic debris. Type A and C entrances represent examples of high energy in-
put sources; Type B and D, low energy input sources. When water levels occur close to the
surface in Type A and B sinkholes, the ambient light causes the water to appear blue. Under
these conditions, the sinkholes are locally referred to as "blue holes."
dotea in north-central Florida, suggests that subterranean routes for dis-
persal are open but that the amount of incoming energy clearly dictates
which of these species groups will be successful in colonizing a given sys-
tem. The presence of populations of the cave-dwelling fish, yellow bull-
heads, Ictalurus natilis (Relyea and Sutton 1973), and Notropis harper
(Marshall 1947, Hubbs 1956), in north-central Florida caves, often many
kilometers from the nearest epigean habitats, suggests that much of the
aquatic subterranean system is interconnected.
The distribution of Troglocambarus maclanei, a small species, seems
to be related to the size of organic particles, and it is usually found in
zones of lowest energy concentrations (Fig. 9). Hobbs (1942) described
Vol. 28, No. 3
FRANZ AND LEE: TROGLOBITIC CRAYFISHES
their "peculiar habit of clinging ventral-side-up to the submerged
ceilings" of the flooded cave passage in Squirrel Chimney. Sutton (pers.
comm.) observed that Troglocambarus maclanei in Sims Sink were con-
centrated on the lower portions of the detrital cone where the particle size
was small. He also noted that there was little overlap of the areas in which
this crayfish and the larger Procambarus erythrops occurred. The two
species seem to be distributed along an energy gradient that occurs out-
ward from these energy islands. There is apparently little competition
between Troglocambarus maclanei and the larger species since, at each
Troglocambarus maclanei locality, there occurs a member of the
lucifugus or pallidus complex. Conversely it is exceptional to find any of
the other species living sympatrically.
Troglocambarus appears to exist at much lower densities than any of
the species discussed above, although it is not clear whether this is simply
because of their low visibility or whether it is indicative of actual
numbers.
Caine (1978) and Dickson and Franz (1980) demonstrated that certain
hypogean Florida crayfishes (including members of both the pallidus and
the lucifugus groups) have lower metabolic rates than related surface
species. In addition, data on whole body respiration rates (Baskin, un-
publ. manuscript) and on respiration rates, ATP turnover, and adenylate
energy charge in excised gills suggests that P. pallidus exhibits a greater
physiological adjustment to low energy cave environments than do P.
erythrops, P. franzi, and P. lucifugus, which parallels the observed habi-
tat differences between these two groups. No similar data are available
for Troglocambarus maclanei.
SUGGESTED EVOLUTIONARY HISTORY
OF CERTAIN TROGLOBITIC SPECIES
Hobbs (1942) originally visualized two separate invasions of crayfishes
into Florida aquifers, and attempted to relate them to peninsular
submergence during the Pleistocene. He later inferred a third invasion
with the discovery of Procambarus miller in South Florida (Hobbs 1971).
Mohr and Poulson (1966) speculated that north-central Florida cave
crayfish evolved primarily through avoidance of competition with the
larger and more recent invaders dominating areas of food abundance.
Caine (1974) discussed cave crayfish distributions in relation to max-
imum sea level stands and suggested that all cave Procambarus, except P.
miller and P. acherontis, invaded subterranean habitats in the Pleisto-
cene. He considered P. miller to be recently derived, stating that "this
species may have been isolated after the aquifer was lowered in southern
Florida in the 1920's." Caine suggested that P. acherontis arose after
1982
BULLETIN FLORIDA STATE MUSEUM
flooding above the 15 m level in the early Pleistocene. We believe most
of these statements to be incorrect.
Recent reinterpretation of Florida's Tertiary and Quaternary history
and the correlation of certain geological features with present troglobitic
crayfish distributions allow us to reevaluate the evolutionary history of
these crustaceans. One of the major conflicts of earlier attempts at ex-
planations was the concept of drastic eustatic fluctuations during the
Pleistocene. Newer interpretations indicate that rises in sea levels during
this period probably never exceeded 20 meters above present m.s.l. and
maybe much less (E.S. Deevey, pers. comm.; Alt and Brooks 1965; Alt
1968). Geological and paleontological evidence also suggests the presence
of at least a short, truncated Florida peninsula since at least the Pliocene
(S.D. Webb, pers. comm.). These two concepts provide us with a persis-
tent land mass, and a large span of time in which crayfishes could enter
Florida and invade spelean habitats.
The ancestral forms essentially entered the particular geological for-
mations where their modern counterparts exist today (i.e. Crystal River,
Hawthorne and St. Marks formations). We should also point out that
many present sinks are probably ancient. Webb (1974) documents that
some have existed at their present locations and remained open for at least
2,000,000 years. Other sinkholes having terrestrial Miocene and even late
Oligocene fossils associated with them suggest that sinkholes have
characterized Florida for millions of years.
Remains of bats that are indistinguishable from present day Myotis
austroriparius are known from several Northcentral Florida fossil
deposits of middle to late Pleistocene age (Martin 1974, Webb 1974). This
implies that many of the bat/crayfish colonies now present in this area
may have been more or less continuous at many sites. Because of this we
suggest the strong possibility of extremely protracted histories of par-
ticular crayfish colonies.
Hobbs (1958) postulated "that the Propictus Stock had gained a
foothold in at least some streams of the southeast not later than the
Pliocene-probably much earlier-and that their migrations from one
river system to another were largely dependent upon stream piracy." We
suspect that at least some members of this stock entered Florida shortly
after the establishment of the peninsula. Only a small group of this once
important faunal element exists in Florida today, represented by Procam-
barus pictus (Hobbs), P. young Hobbs, P. seminolae Hobbs and the
troglobitic members of the P. pallidus and P. lucifugus complexes and
Troglocambarus maclanei. The first two species show a relictual distribu-
tion, being endemic to a few isolated creek systems associated with old
landforms. P. pictus is on the southeastern flank of the Northern
Highlands and P. young on the southern escarpment of the Ap-
Vol. 28, No. 3
FRANZ AND LEE: TROGLOBITIC CRAYFISHES
palachicola Highlands. P. seminolae, a flatwoods species, is perhaps a
later migrant to Florida, but at least one member of this subgroup must
have reached Florida earlier to account for the existence of T. maclanei.
The drainage pattern of the ancestral peninsula was probably one of
short streams draining flatwoods and flowing directly into the sea. The
occurrence of fossil streams like those of the Love Bone Bed and
McGeehee Farm sites in western Alachua County, which date from the
late Miocene (latest Clarendonian to early Hemphillian) (Webb et al.
1981), helps substantiate this pattern. The coalescing of several headwater
streams probably gave rise to the precursor of the Suwannee River, at a
somewhat later time. Sediment analysis and bone orientation studies at
Love Bone Bed reveal that this ancient stream had a sufficient current to
support a stream fauna, including stream-dwelling crayfishes. Unfor-
tunately crayfishes are rarely fossilized, but some plate-like carbonate
structures collected at the Love Site suspiciously resemble crayfish
gastroliths. These stream faunas probably moved from one creek system
to another through stream piracy. Thus, P. pictus could have arrived in
Black Creek (suggested as a Pliocene relict by Burgess and Franz 1978)
and P. young in the Wetappo-St. Marks systems in this way. The elimina-
tion of other surface pictus-like populations from intermediate areas can
be explained by the disappearance of the small stream systems along the
western margin of the ancestral peninsula. This occurred as the large ex-
panses of karst west of the Northern Highlands and south of the Ap-
palachicola Highlands were exposed. Streams flowing off these highlands
were captured by sinks, as they are today; the Santa Fe, Aucilla, and
Chipola rivers disappear into sinks for varying distances. We see this as an
excellent opportunity for surface crayfishes to gain entrance to the
aquifer, and propose that multiple invasions of the aquifers account for
the number of modern troglobitic species and their varying degrees of
specialization.
There are numerous documented accounts of recent "overnight" cap-
tures of central Florida lakes by sinks during periods when lowered water
tables were unable to provide support for time-eroded limestone
(Lakeland-Bartow Ridge in Polk County). These captures would certainly
pour tremendous quantities of organic debris into the aquifer and perhaps
provide enough food to sponsor colonizing surface crayfishes through
generations of early adaptations.
It is unfortunate that it is not possible to reconstruct the exact se-
quence of invasion of crayfishes into subterrean habitats. Mohr and
Poulson (1966) suggested that the most specialized species entered the
system first and adapted into its present form because of competitive
pressure from larger species, which later invaded the underground
habitats. Although this may well be the case, it could be argued with
1982
BULLETIN FLORIDA STATE MUSEUM
equal vigor that the earliest invader would already be specialized enough
to have a competitive edge over later invaders. The amount of time
available for evolution and development is certainly sufficient for almost
any sequence of events.
It is probable that two ancestral forms may have entered at different
points and at different times. This may account for the existence of the
multispecies in the P. pallidus and P. lucifugus complexes. The modern
range of P. lucifugus consists of five allopatric populations. We believe
that these populations were continuous with one another, and that cer-
tain barriers arose to effectively impede gene flow. Some of these barriers
can be tentatively identified. The western Gilchrist population appears to
be isolated from that in adjacent Alachua County by a system of north-
south trending ridges and the floodplain of an extinct river. The combina-
tion of deep terrace sands and dense floodplain sediments is probably
enough to discourage the entry of detritus into, and the movement of
crayfish within, the aquifer. No obvious barriers separate the Alachua
population from the intermediate one in southwestern Marion County.
However, a second closely related species, P. franzi, fills the gap. This
species may have competitively excluded P. lucifugus in that area by exer-
ting some biological influence that caused or initiated its extirpation. The
barrier between the intermediate and the Citrus-Hernando populations,
the Withlacoochee River, is more obvious. This river removed the Crystal
River Formation from a wide band, exposing the Williston and Inglis for-
mations, which are less soluble lower members of the Ocala Group.
It is perhaps worth speculating on the origins of the north-central
Florida cave crayfish populations in view of their ecological distributions.
Troglocambarus maclanei is the most widespread Florida cave crayfish
species, yet exhibits no morphological differentiation (or variation). Since
this species is ostensibly the least dependent on north-central Florida
energy islands, it has apparently retained a more or less constant gene
flow between colonies by following interstitial routes. P. pallidus, which
has an intermediate dependency on energy islands, has colonized a
relatively large area, and pallidus-like stocks have given rise to two addi-
tional species. The P. lucifugus complex, composed of the largest group of
taxonomically distinct populations, has the greatest dependency on
energy islands and therefore experiences less gene interchange than do
other north-central Florida crayfishes. Thus it is reasonable to assume
that the degree of ecological dependence on energy islands has, in varying
degrees, dictated the ability to disperse, limited the gene flow, and
resulted in the diverse speciation of central Florida's cave crayfishes.
At least three different life styles evolved. P. pallidus, and possibly P.
horsti and P. orcinus, had more efficient metabolisms and were able to oc-
cupy caves, springs and solution tube sinkholes where only limited
Vol. 28, No. 3
FRANZ AND LEE: TROGLOBITIC CRAYFISHES
amounts of detrital energy were available. Members of the P. lucifugus
complex, on the other hand, capitalized on the greater amounts of energy
associated with collapse sinks and bat caves. The peculiar feeding
modifications and life style of Troglocambarus maclanei may have
developed to allow coexistence with the larger species, as suggested by
Mohr and Poulson (1966) or developed independently in caves with
modest energy availability. Changing sea levels, undulating water tables,
and shifts in major areas of energy input may have secondarily forced
species to coexist, and at times could have either united or further isolated
closely related stocks.
P. miller is known from a single well in southeast Florida. It is ap-
parently derived from a parental stock of P. alleni, an epigean species that
is common in south Florida. We have observed P. alleni living in water-
filled limestone sinks on Big Pine Key, Monroe County, Florida. Caine
(1974) theorized that the ancestral form burrowed into subterranean
habitats or entered through exposures of oolitic limestone. The seasonal
fluctuation of surface waters in south Florida would certainly provide
numerous opportunities for crayfishes to move into karst areas and then
become at least seasonally confined to solution pools and sinks. Caine
(1974) suggested that P. miller is not as old as other troglobitic cam-
barids, and that the species may have evolved after the aquifer was
lowered in south Florida in the 1920's. That this form retains some black
pigment and facets in the eye was his only apparent evidence of this (yet
various populations of the lucifugus complex also have eye spots),
although he was certainly led to a conclusion of recent evolution based on
the supposed submergence of south Florida in the Pleistocene.
Although geologic evidence is lacking, zoogeographic patterns in-
dicate that the most recent submergence of south Florida may not have
occurred as recently as believed. Telford (1966) surmised that an endemic
species of snake, Tantilla oolitica, in south Florida may have lived on high
ground in Dade County during the Peorian interglacial. A diverse tropical
element, with no apparent endemics is limited to a narrow strip of south
coastal Florida and the Keys (see Neill 1957), which includes the area in-
habited by P. miller. Also known from this general area are numerous
recognized races and/or disjunct populations of terrestrial vertebrates. All
this suggests that some type of refugium persisted in this region for a con-
siderable period. Although the exact age of the area is not clear, it cer-
tainly existed for a long enough period to suggest that P. miller could
have appeared prior to the 1920's.
In contrast to the energy island species assemblages of troglobitic
crayfishes in peninsula Florida, a single, apparently wide ranging species,
Cambarus cryptodytes, occurs in west Florida and southwest Georgia.
We attribute this distribution and the lack of defined populations to the
1982
BULLETIN FLORIDA STATE MUSEUM
ample lateral transport of detritus throughout the aquifer. Although in-
dividuals will congregate around available food sources, they freely
disperse once the food source is exhausted.
In summary, the specious nature of Florida's troglobitic crayfish
fauna resulted from the invasion of the aquifer by at least 6 different sur-
face crayfishes, possibly beginning as early as the late Miocene or
Pliocene. They probably gained access to the subterranean environment
when surface streams were captured by sinkholes. Multiple invasions of
single species at different times and in different areas could account for
the occurrences of multi-species in both the Procambarus lucifugus and
P. pallidus complexes. Once underground, certain crayfishes, possibly re-
sulting from sympatry, engaged in occupying different parts of the in-
habitable cave environment. Some species (e.g. members of the P. lucifu-
gus complex) inhabited the energy rich areas (i.e. in large sinks and under
bat roosts), while others were not as dependent on these accumulations.
The low energy species (e.g. members of the P. pallidus complex and
Troglocambarus maclanei) had greater dispersal abilities, and today oc-
cupy the greatest geographic ranges.
LITERATURE CITED
Alt, D. 1968. Patterns of post-Miocene eustatic fluctuation of sea level. Paleogeog.,
Paleoclim., Paleoecol. 5: 87-94.
and H.K. Brooks. 1965. Age of Florida marine terraces. J. Geol. 73:
406-411.
Banks, J.E. 1976. Middle tertiary lithostratigraphic units in the geologic column of central
Florida. Pp. 63-65 in Mid-Tertiary carbonates: Citrus, Levy, Marion Counties, West-
Central Florida. Southeastern Geol. Soc. Field Trip Guide Book No. 18.
Burgess, G.H., and R. Franz. 1978. Zoogeography of the aquatic fauna of the St. Johns River
system with comments on adjacent peninsular faunas. Amer. Midl. Naturalist. 100(1):
160-170.
Caine, E.A. 1974. Zoogeography of the Floridian troglobitic crayfishes, genus Procambarus.
Amer. Midl. Naturalist 92(2): 487-492.
1978. Comparative ecology of epigean and hypogean crayfish (Crustacea:
Cambaridae) from northwestern Florida. Amer. Midl. Naturalist 99(2): 315-329.
Carr, A.F. 1939. Haideotriton wallacei, a new subterranean salamander from Georgia. Occ.
Pap. Boston Soc. Nat. Hist. 8: 333-336.
Dickson, G.W., and R. Franz. 1980. Respiration rates, ATP turnover and adenylate energy
charge in excised gills of surface and cave crayfish. Comp. Biochem. Physiol. 65A:
375-379.,
Franz, R., and L.M. Franz. 1979. Distribution, habitat preference, and status of populations
of the Black Creek crayfish, Procambarus (Ortmannicus) pictus (Decapoda: Cambari-
dae). Florida Scientist 42(1): 13-17.
Hendry, C.W., and C.R. Sproul. 1966. Geology and ground water resources of Leon
County, Florida. Florida Geol. Surv., Bull. No. 47: 178 p.
Hobbs, H.H., Jr. 1940. Seven new crayfishes of the genus Cambarus from Florida, with notes
on other species. Proc. U.S. Nat. Mus. 89(3097): 387-423.
Vol. 28, No. 3
1982 FRANZ AND LEE: TROGLOBITIC CRAYFISHES 77
1942. The crayfishes of Florida. Univ. Florida Press. Biol. Sci. Ser. 3(2):
179 p.
1958. The evolutionary history of the pictus group of the crayfish genus
Procambarus. Quart. J. Florida Acad. Sci. 21: 71-91.
1971. A new troglobitic crayfish from Florida. Quart. J. Florida Acad. Sci.
34(2): 114-124.
1972. The subgenera of the genus Procambarus (Decapoda: Astacidae).
Smithsonian Contrib. Zool. 117: 1-22.
1974. A checklist of North and Middle American crayfishes (Decapoda:
Astacidae and Cambaridae). Smithsonian Contrib. Zool. 166: 1-161.
and T.C. Barr. Jr. 1960. The origins and affinities of the troglobitic
crayfishes of North America (Decapoda. Astacidae). I. The genus Cambarus. Amer.
Midl. Naturalist 64(1): 12-33.
1 H.H. Hobbs. III. and M.A. Daniel. 1977. A review of the troglobitic
decapod crustaceans of the Americas. Smithsonian Contrib. Zool. 244: 1-183.
and D.S. Lee. 1976. A news troglobitic crayfish (Decapoda, Cambaridae)
from peninsular Florida. Proc. Biol. Soc. Washington 89(32): 383-392.
and D.B. Means. 1972. Two new troglobitic crayfishes (Decapoda,
Astacidae) from Florida. Proc. Biol. Soc. Washington 84: 393-410.
Hubbs, C.L. 1956. Preliminary analysis of the American cyprinid fishes, seven new. referred
to as the genus Hybopsis. subgenus Erimystax. Occ. Pap. Mus. Zool. Unix. Mich.
578: 1-8.
Lee. D.S. 1969. Notes on the feeding behavior of cave-dwelling bullfrogs. Herpetologica
25(3): 211-212.
Lonnberg, E. 1894. Cambarids from Florida. a new blind species. Zool. Anz. 17: 125-127.
Marshall. N. 1947. The spring run and cave habitats of Erimystax harper (Fowler). Ecology
28(1): 68-75.
Martin, R.A. 1974. Fossil mammals from the Coleman IIA fauna. Sumter County. Pp. 35-99
in S.D. Webb (ed.). Pleistocene Mammals of Florida. Univ. Presses Florida. Gainesville.
270 p.
Mohr, C.E., and T.L. Poulson. 1966. The Life of the Cave. McGraw-Hill Book Co., New
York. 232 p.
Moore, W.E. 1955. Geology of Jackson County, Florida. Florida Geol. Surs. Bull. No. 37:
101 p.
Neill, W.T. 1957. Historical biogeography of present-day Florida. Bull. Florida State Mus.,
Biol. Sci. 2(7): 175-220.
Relyea, K. 1976. Aquatic cavernicoles in Florida. Florida Naturalist 49(4): 6, 25, 30.
D. Blody, and K. Bankowski. 1976. A Florida troglobitic crayfish: Biogeo-
graphic implications. Florida Scientist 39(2): 71-72.
and B. Sutton. 1973(1974). Ecological data for a Florida troglobitic cray-
fish. Florida Scientist 36(2-4): 234-235.
and B. Sutton. 1974. Cave dwelling yellow bullheads in Florida. Florida
Scientist 36(1): 31-34.
and B. Sutton. 1975. A new troglobitic crayfish of the genus Procambarus
from Florida (Decapoda: Astacidae). Tulane Stud. Zool. Bot. 19: 8-16.
Telford. S.R., Jr. 1966. Variation among the southeastern crowned snakes, genus Tantilla.
Bull. Florida State Mus.. Biol. Sci. 10(7): 261-304.
Vernon, R.O.. and H.S. Puri. 1964. Geologic map of Florida. Florida Dix. Geol.: Map Ser.
No. 18.
Warren. B.D. 1961. The obligative cavernicoles of Florida. Florida Speoleol. Soc. Spec. Pap.
1: 1-10.
Webb. S.D. 1974. Chronology of Florida Pleistocene mammals. Pp. 5-31 in S.D. \Webb (ed.).
Pleistocene Mammals of Florida. Univ. Presses Florida. Gainesville. 270 p.
78 BULLETIN FLORIDA STATE MUSEUM Vol. 28, No. 3
__ B.J. MacFadden, and J.A. Baskin. 1981. Geology and paleontology of the
Love Bone Bed from the late Miocene of Florida. Am. J. Sci. 281: 513-544.
Williams, K.E., D. Nicol, and A.F. Randazzo. 1977. The geology of the western part of
Alachua County, Florida. Florida Bur. Geol. Rept. Invest. No. 85: 98 p.
Yon, J.W., Jr. 1966. Geology of Jefferson County, Florida. Florida Ceol. Surv. Bull. No. 48:
119 p.
Zinn, T.L. 1977. Community ecology of Florida bats with emphasis on Myotis
austroriparius. M.S. thesis, Univ. Florida, Gainesville. 87 p.
HUMAN PREDATION ON THE GOPHER TORTOISE
(GOPHER US POLYPHEMUS)
IN NORTH-CENTRAL FLORIDA
ROBERT W. TAYLOR, JR.1
ABSTRACT: Human predation on Gopherus polyphemus was investigated through personal
observation of and participation in the process, hunter interviews, examination of butchered
animals, and laboratory dissections of tortoises. Information was gathered about the hunters'
methods, the results of their efforts, and the effects on local tortoise populations. Light preda-
tion probably does not have a strong adverse effect on a population. Intensive exploitation,
however, may seriously affect the viability of populations of this species because of the tor-
toise's extremely low reproductive rate and the difficulty in replacing lost individuals. Ap-
proximately equal numbers of each sex are taken, and the size distribution of butchered
animals reflects that of typical colonies. Predation on the gopher tortoise by man is
widespread, despite the fact that the small edible portion of each animal leads to the rela-
tively high cost of obtaining the flesh. The use of G. polyphemus as a food item is perpetu-
ated by the culture of certain groups and their traditional exploitation of the species.
RESUMEN: La predacion humana sobre tortugas Gopherus polyphemus fue ii.'. .r.J.., por
observaci6n personal, entrevistas con los cazadores, examen de ejemplares colectados y
posteriormente disecados en el laboratorio. Se obtuvo information acerca de los metodos v
resultados de caza asi como de los efectos de 6sta sobre las poblaciones locales de tortuga. Es
probable que la predaci6n leave no afecte a la poblaci6n de tortugas de manera muy
desfavorable. Sin embargo, la intensive explotaci6n podria afectar seriamente la viabilidad de
las poblaciones de esta especie debido a su baja tasa de reproduccion y por la dificultad de
reemplazo de aquellos individuos predados. Se tomaron nfimeros aproximadamente iguales
de ambos sexos y, la distribucion de tamano en ejemplares colectados refleja igualdad al de las
colonies tipicas. La predaci6n de la tortuga por el hombre es extensive a pesar del alto costo
con relacion a la parte comestible. El uso do G. polyphemus como alimento es perpetuado por
la cultural de ciertos grupos y por la traditional explotacion de esta especie.
TABLE OF CONTENTS
INTRODUCTION .. . 80
STATUS OF Copherus polyphe i us . ............. . . . 80
ABCIAEOLOGICAL PERSPECTIVE . ............... ... 81
MATERIALS AND METHODS .. ................ . . .......... 82
RESU LTS ...... . .... .. . .. 83
D ISC U SSIO N . .. . .......... . . . 99
CHANGES IN POPULATION SIZE ......... . . 99
BASES OF HUMAN PREDATION . . . . 100
SUMMARY AND CONCLUSIONS ... ...... . 102
LITERATURE CITED .. ........ . ... ... .. 103
'Thi author is the Dtirctor of The Nture (Center. l rt \l\rrs, Flrida
TAYLOR, ROBERT W., JR. 1982. Human predation on the gopher tortoise (Gopherus
polyphemus) in north-central Florida. Bull. Florida State Mus.. Biol. Sci. 28(4):79-102.
BULLETIN FLORIDA STATE MUSEUM
INTRODUCTION
Numerous references to the use of the gopher tortoise (Gopherus
polyphemus) as food by humans are found in the herpetological
literature. Early North American herpetologists included brief statements
about the capture or utilization of these animals (Daudin 1802; Holbrook
1836; LeConte 1836; Agassiz 1857). In this century, many authors have
made reference to the exploitation of this species by man, and some have
also commented upon the possible effects of this predation on tortoise
populations (Fisher 1917; Roosevelt 1917; Hallinan 1923; Carr 1952;
Oliver 1955; Auffenberg 1969, 1978; Ernst and Barbour 1972; Auffenberg
and Franz in press). Many of these accounts, however, are based upon
second-hand or anecdotal information and do not go beyond stating the
basic fact that gopher tortoises are captured and eaten by man. Several
popular articles have contained more detailed information about the pro-
cedure involved in capturing and preparing gophers for the table (Ander-
son 1949; Trowbridge 1952; Alberson 1953; Hutt 1967; Thomas 1978).
Even these somewhat more detailed accounts, however, are still primarily
concerned with the capture of the tortoise and its conversion into a meal.
This study is an initial attempt to take a comprehensive look at human
predation on G. polyphemus in the Alachua County, Florida area in
order to show more clearly how and to what extent the tortoises are uti-
lized as a food source and the possible effects of this predation. The
specific objectives of the investigation are to describe the human costs and
benefits associated with collecting tortoises for food, and to evaluate the
possible ecological effects upon local tortoise populations and the species
as a whole. Additionally, an effort is made to define and examine the
motivation of those individuals who look upon G. polyphemus only (or
primarily) as an exploitable resource.
STATUS OF Gopherus polyphemus
The gopher tortoise is classified as a game animal by the Florida Game
and Fresh Water Fish Commission. The 1981-1982 summary of hunting
rules and regulations published by that agency sets a possession limit of
five tortoises per person. This means, for example, that a family of four
persons could legally take 20 gopher tortoises each day for their own use.
There are no restrictions on the size or sex of tortoises that nay be taken.
Since 1980 a closed season has been in effect from 1 April to 30 June each
year. The sale or purchase of any gopher turtle is prohibited at all times.
Incongruously, G. polyphemus is also listed as a "Species of Special
Concern" by the same state agency, due to its drastic decline in numbers
during recent years and the uncertainty of its status. Auffenberg and
Franz (in press) have documented the reduction in numbers over the spe-
Vol. 28, No. 4
TAYLOR: HUMAN PREDATION-ON G. POLYPHEMUS
cies' range, examined the causes of this decline, and predicted continuing
population decreases.
The conflicting policies at the highest level of state wildlife manage-
ment can only foster misunderstanding and disinterest among Floridians
concerning the status of the gopher tortoise within the state and wl at
management policies are appropriate. Fortunately, the Game and Fresh
Water Fish Commission has recently undertaken a research program to
investigate exploited reptiles and amphibians, with initial emphasis on G.
polyphemus (Tommy Hines, pers. comm.). The results of that research
program should provide a basis for the promulgation of appropriate
management regulations based on sound biological data, which have
been lacking up to this time.
ARCHAEOLOGICAL PERSPECTIVE
Gopherus polyphemus has been consistently exploited by the in-
habitants of Florida for 4000 + years. Bonnie McEwan of the Department
of Anthropology, University of Florida, has compiled information about
this exploitation based on material in the Zooarchaeology collection of the
Florida State Museum and has generously allowed me to present it here.
Based primarily on the contents of refuse deposits (middens), ar-
chaeological excavations in Florida reveal gopher tortoise remains in 75 %
of the sites examined. This value is probably conservative since the highly
domed shape of the carapace (which is the skeletal element most often
found) could lead to its use in other activity areas of a given site. For ex-
ample, if used as a vessel or rattle those shells would for the most part not
be included in these data.
A substantially higher proportion of coastal sites have been excavated
than inland occupation areas. Despite this, when present, G. polyphemus
averages 3.7% of the faunal assemblage based on number of individuals.
Additional data from sites where these turtles would have been more
easily accessible (along the central ridge of the state) would undoubtedly
increase this value.
The level of utilization of the gopher tortoise by aboriginal peoples
reflects the relative importance of terrestrial and aquatic resources in
their societies. Data from the Palmer site in west-central Florida, which
has been analyzed for three distinct time periods spanning 3000 years, in-
dicates that a reduction in utilization of this species follows the develop-
ment of a sophisticated fishing technology or more likely closer access to
aquatic habitats through time. The trend toward a greater reliance on
freshwater and marine resources, and the consequent de-emphasis of ter-
restrial fauna exploitation is observed in most of the archaeological sites
analyzed. Not until the time of European contact and the subsequent
1982
82 BULLETIN FLORIDA STATE MUSEUM Vol. 28, No. 4
reliance on domesticated animals is the terrestrial fauna represented to
the same degree as the aquatic in coastal sites.
Sea turtles are represented to a greater degree in sites where they are
found than are gopher tortoises, averaging 10.1% where they occur based
on number of individuals. This greater intensity of use of Cheloniidae
probably occurred because of the breeding or feeding congregations of
these species and the opportunity to gather large numbers of them at the
coastal sites, rather than the result of dietary preference.
The exploitation of Gopherus polyphemus as a food resource 4000
years ago was due to several factors that are still in effect today. Gopher
tortoises represent a readily available food source which oftentimes re-
quires only minimal procurement materials and skill. The location of
their burrows is obvious, and their movements are often indicated by
well-worn trails nearby. Their docile nature and slow movements do not
serve them well during encounters with humans. Along with the oppor-
tunistic nature of man, these characteristics have facilitated human ex-
ploitation of the gopher tortoise in the past and continue to be important
today.
MATERIALS AND METHODS
Data were gathered between September 1978 and May 1980 in the following ways:
INTERVIEWS. -Thirty-three interviews involving a total of 41 persons were conducted in
the Alachua County, Florida, area. All the persons interviewed used Gopherus polyphemus as
an occasional or regular food item. Most of the individuals interviewed were contacted
through previous informants or by being approached by the author. A few individuals were
interviewed after the discovery during other conversations that they caught and ate gophers.
Information gathered included where and how often tortoises were hunted, how the tortoises
were caught, cleaned, and cooked, which parts of the animal were utilized as food, informa-
tion about the sale and purchase of tortoises, and any other pertinent data.
COLLECTION TRIPS. Between September 1978 and April 1980, 32 gopher collecting trips
took place. During 21 of these the author was in the company of one or more experienced
gopher pullers. During these trips, data were gathered on capture methods and efficiency,
numbers and characteristics of tortoises collected, and hook location and damage. It was dur-
ing these trips that the essence of the "rural gopher puller" was delineated. Following these ex-
cursions the methods used to clean and cook the animals were observed. Although illegal, tor-
toises are regularly bought and sold by local residents. Price ranges for various sized tortoises
were noted. The sale of individuals was observed on several occasions.
DISSECTIONS. -In order to accurately estimate the amount of edible flesh obtainable from
a tortoise, dissections of 36 individuals were performed. These animals had been used in
another research project and had been sacrificed. All edible portions of each tortoise were
removed and weighed. This information was then used along with size data from the tortoises
involved to determine the relationship between an animal's size and the proportion of that in-
dividual that could be eaten. A comparison was also made between the cost of tortoise flesh
and that of commercially available domesticated animals.
REFUSE SHELLS. -Groups of discarded, cleaned tortoise shells were sometimes found in
rural regions near housing areas, or at certain locations in the forests where trash and garbage
TAYLOR: HUMAN PREDATION-ON G. POLYPHEMUS
are deposited (casual dumps). Whenever possible, these discarded shells were salvaged and
returned to the lab for analysis in order to determine the size and sex of the individuals from
which they came. Information about cleaning methods was also derived from examination of
these remains. The sex of each individual was determined based on overall shell morphology
after a subjective evaluation with respect to the following characters: plastral concavity,
degree of anterior gular projection, degree of xiphyplastral thickening, extent of curvature of
posterior carapacial margin, and shell thickness. Except for shell thickness, these variables are
essentially those of McRae et al. (1981). An estimate of the distribution within size and sex
classes was later made to assess the impact of capture techniques upon the species.
RESULTS
interviews with persons who utilize Gopherus polyphemus as a food
resource revealed a wide range in the frequency of their hunting ac-
tivities. Average intervals between trips varied from a week to several
months, but no individual claimed to adhere to any sort of regular
schedule. Some persons would go every day if time permitted or if hunting
was unusually good. Under most circumstances the physical effort re-
quired to capture tortoises is substantial, thereby reducing the number of
trips that would otherwise be made by some persons and deterring others
altogether. The latter is often the case with elderly individuals who, in
their younger days, regularly hunted tortoises.
To successfully hunt tortoises, a person must first identify and have ac-
cess to a site that is populated by the animals. Knowledge of possible
hunting sites is gained either through personal experience, by word of
mouth, or by actively searching. Active searching involves patrolling a
possible area (usually in an automobile) and making exploratory trips into
wooded areas, pastures, old fields, or other suitable habitats. Some per-
sons are more respectful than others of land owners' rights and request
permission before venturing onto private property. This is particularly
true if a fence must be crossed. Many property owners are quite willing to
allow gopher hunting on their land, especially those who have cattle or
horses, or those with hayfields that must be cultivated and harvested with
wheeled vehicles. The reason for refusal of a request to hunt tortoises on
private property usually has nothing to do with concern for the tortoises'
welfare, but is simply distrust of strangers on the property. A contributing
factor in this regard may be that most landowners are white, while my ex-
perience indicates that most gopher-pullers are black. The problem of ob-
taining permission is avoided entirely on public land. In the Alachua
County area, the largest tract of public land containing large gopher tor-
toise populations is the Ocala National Forest (primarily in Marion
County) approximately 80 kilometers away. This area is hunted regularly
by residents of Alachua and Putnam counties.
Once access to an area has been gained, the problem becomes one of
actually securing the gophers. This is accomplished by one of three
1982
BULLETIN FLORIDA STATE MUSEUM
general methods. The easiest way is simply to pick up individuals that are
found away from their burrows. This procedure is almost a completely
chance event, and surely accounts for a very small percentage of the total
number of tortoises taken. Many more are probably run over or picked up
by motorists while the animals are crossing the roads than are found by
persons hunting them. Fisher (1917) and Ditmars (1946) reported making
use of the animals' tracks in the sand to capture some individuals, and the
tortoises' habit of following well defined trails through sufficiently dense
vegetation is well known (Ernst and Barbour 1972). However, these aids
are normally of little use, except possibly in an area with an unusually
high tortoise density. During approximately 160 hours spent in the field
collecting gophers over the past two years, I encountered only four tor-
toises away from their burrows.
The second general method of capture involves the use of some sort of
trap. The most common type of trap consists of a five gallon bucket placed
in a hole dug immediately outside the burrow mouth, and often covered
with paper or vegetation. As the tortoise enters or leaves the burrow it
must cross the top of the bucket and will fall in. Agassiz (1857) reported
that this method was effective, and I have captured many individuals in
this way. A less commonly used type of trap (it was described to me only
once) consists of a snare placed at the burrow mouth. A loop of heavy
monofilament fishing line is supported in such a way that the tortoise's
head must pass through it as he enters or leaves the burrow. One end of
the line is secured to nearby sturdy vegetation or a stick deployed as an an-
chor. The loop closes around the tortoise's neck and holds the animal until
removed by the trapper. Both of these trapping methods can easily lead to
the death of the tortoises, even if the animals were not going to be killed
anyway, as is sometimes the case with farmers desiring only to remove the
animals from near their crops. A tortoise left in a bucket trap through
midday during the Florida summer will almost certainly die because of
heat stress, unless the day happens to be overcast. Likewise, the snare
around the gopher's neck will get progressively tighter as the animal
struggles to escape, leading to asphyxiation unless it is removed shortly
after capture.
The capture method that accounts for the vast majority of gophers
taken for human consumption involves the use of a gopher pulling "hook,"
and is described by Fisher (1917), Hallinan (1923), Anderson (1949),
Alberson (1953), Hutt (1967), and Thomas (1978). These hooks vary in
construction, but all consist of a long, flexible shaft, to the end of which is
attached a sturdy bent metal piece. The long, flexible body of the hook
allows it to be inserted into the deep, curved tortoise burrow, while the
bent metal piece snags onto some part of the animal (usually the shell).
The tortoise may then be physically pulled up the length of the burrow
Vol. 28, No. 4
TAYLOR: HUMAN PREDATION-ON G. POLYPHEMUS
and out the mouth. Many materials have been used as the main body of
the gopher hook, including garden hoses, small diameter concrete
reinforcing rods, appropriately shaped lumber products, and wild
grapevines. Today, however, these hooks are made almost exclusively
from a large diameter (6-7 mm) "wire" that is a structural component of
modern box-spring bedding units. These wires support the periphery of
the upper side of the box-spring unit, and are from 5.8 to 7.1 meters long,
depending on the size of the bedding unit from which they are taken.
These wires perfectly satisfy the required combination of flexibility
needed to follow the sometimes highly curved gopher burrow, and stiff-
ness needed to be pushed in and worked from outside the mouth. The tip
of the wire is heated and bent around to create the required hook on the
end. A wooden handle is attached to the other end of the wire to provide a
secure handhold for the operator. The completed hook is usually 6.1 to
7.6 meters long. When not in use, the flexibility of the wire allows it to be
coiled up into a circle about one meter in diameter, which makes storage
or transportation in the trunk of a car relatively easy (Figs. 1 and 2).
To "pull" the gopher from its burrow the tip of the hook is inserted
into the mouth and gradually worked down until the tortoise is encoun-
tered. With experience the operator can usually discriminate between a
gopher and a root, rock, or other hard object. If a tortoise is felt, the tip of
the wire is maneuvered back and forth, in and out ("fished"), until it
E 1.Tip of gopher pulling hook.
FIGURE 1. -Tip of gopher pulling hook.
86 BULLETIN FLORIDA STATE MUSEUM Vol. 28, No. 4
' .' . .
. ..
: '-i *' " 7
FIGURE 2. -Gopher pulling hook coiled up for transportation. Yardstick is for size compari-
son.
becomes hooked on some part of the gopher's body. At that point, the
operator moves away from the burrow, pulling the hook from the hole,
and dragging the tortoise to the surface (Fig. 3).
The holding power of a tortoise inside the burrow is considerable.
Many get hooked under the rear of the carapace while they are facing
downward, as they would be if they had entered the burrow and not
turned around. Upon being hooked these individuals often extend their
front limbs and dig into the sides of their burrow. It is not the strength of
their limbs that is overcome by the constant pressure being exerted by the
human at the other end of the hook, but the side walls of the burrow,
which give way under such stress. Extraordinary force is often required to
pull the gopher out of the burrow (Fig. 3). The strength of two men is
sometimes insufficient to dislodge a tortoise from its position.
During the 32 gopher collecting trips, I observed 130 tortoises pulled
from their burrows (X = 4.1 tortoises/trip). The number of captures
varied greatly however, depending upon whether I was alone and
therefore attempting to pull the tortoises myself (X = i.0 tortoise/trip),
or relying on the abilities of an experienced puller who was present
(X = 6.0 tortoises/trip). The time invested in these 32 trips varied from
about two hours to most of the day, with a typical trip requiring four to
six hours.
TAYLOR: HUMAN PREDATION-ON G. POLYPHEMUS
FIGURE 3.-Local hunter pulling gopher tortoise from burrow. Note extreme force that is
required.
PULLING SUCCESS RATE. -During 12 collecting trips the number of ac-
tive burrows of adults in the site was either counted exactly or estimated
(six trips in each case). The number of tortoises collected from these sites
was then used to calculate the pulling success rate for each site. The mean
pulling success rate for the six sites for which systematic burrow counts
were made was 21% (29, 27, 22, 17, 17, and 13 percent). The mean pull-
ing success rate for the six sites for which the burrow count was estimated
was 20% (26, 26, 24, 17, 13, and 12 percent). The success rate as it was
calculated was slightly higher than it would have been if every burrow in
the area (inactive and of juveniles) had been included in the total burrow
count.
The number of tortoises pulled in one area can vary widely, depend-
ing upon whether they have been hunted there before, and the time and
effort a hunter is willing to expend. Failure to capture any gophers may
result if few are present to begin with, the hunter is unskilled, or the bur-
rows are too long or crooked. At the other extreme, one hunter told me,
one day after the fact, that he pulled 32 in 2.5 hours. An average number
would surely lie somewhere in between. By working for a few hours in an
area with a good number of tortoises a skilled hunter could expect to pull
5 to 15 animals. I have seen one person take from 8 to 13 gophers on 7
separate occasions.
1982
BULLETIN FLORIDA STATE MUSEUM
The legal possession limit and closed season in Florida are for the most
part misunderstood or ignored by gopher hunters. When asked about the
regulations governing the taking of tortoises, most individuals could not
correctly cite them, (although most knew that there were regulations of
some kind), even if they have been hunting the animals for many years.
LOCATION OF HOOK. The location of the hook on the body was
recorded for 106 of the 130 tortoises included in the study as each in-
dividual was pulled from the burrow (Table 1). The hook locations for the
other 24 tortoises were unknown. Most of these were individuals that
became separated from the hook immediately after being pulled through
the burrow mouth and before the hook site could be noted. The great ma-
jority (77.4 %) of individuals for which the hook location was known were
pulled with the hook either under the rear of the carapace (40.6%) or at
the left or right axillary notch (35.8%). No significant difference in hook
location between males and females was found. Body size, however, in-
fluenced what part of the body would be hooked. Of 40 individuals
classified as large (greater than 270 mm total length), 24 (60%) were
hooked under the rear edge of the carapace, while not a single small tor-
toise (less than 240 mm total length) was captured in that way. Con-
versely, only 7 (17.5%) large individuals were hooked at the axillary
notch, whereas this location accounted for 12 of 21 captures (57.1%)
among small tortoises. Frequencies of hook locations for medium-sized
animals were approximately intermediate between the other two groups.
The number of captures in which the hook was not at the rear of the
carapace or axillary notch was low (n = 24, 22.6% of known), and there
was no obvious effect of tortoise size on captures at these locations.
HOOK DAMAGE. Of the 130 tortoises seen pulled, 27 (21.8%) suffered
some kind of damage as a result of the pulling hook. Of those injuries, 15
(55.6%) involved damage to the tortoise's shell, while the remaining 12
were less serious wounds to the skin or flesh (Table 2). Three individuals
suffered damage to the shell in two (n = 2) or three (n = 1) places, but
in Table 2 are counted only as one injury each.
The axillary and inguinal notches of the bridge were particularly
prone to damage by the pulling hook. Although comprising only 38.7 % of
the known hook locations (Table 1), they accounted for 59.3% of the in-
juries suffered by the tortoises (Table 2). Hooking at the front of the
carapace resulted in an even greater chance of injury. Only 10 individuals
were pulled with the hook at the location, but there was an equal number
of tortoises with injuries in that area.
The size of the tortoise greatly influenced the incidence of damage to
the animal during the pulling process. Table 3 contains the number of in-
dividuals and percent of the total injured that fall into each size category.
Clearly, smaller individuals experienced a substantially higher rate of in-
Vol. 28, No. 4
TABLE 1. -Locations of pulling hook for tortoises taken from burrows, by sex and body size.
% % Sex Body Size
Location of hook n of total of known Male Female Unknown Sm Med Lg Unknown
Rear of carapace 43 33.1 40.6 17 20 6 0 12 24 7
Axillary-notch 39 30.0 36.8 16 17 6 12 15 7 5
Front legs 10 7.7 9.4 1 4 5 0 3 3 4
Front of carapace 10 7.7 9.4 2 5 3 4 2 2 2
Inguinal notch 2 1.5 1.9 2 0 0 0 2 0 0
Xiphiplastron 1 0.7 0.9 1 0 0 0 0 1 0
Flesh of tail 1 0.7 0.9 1 0 0 0 0 1 0
Unknown 24 18.5 6 11 7 5 7 2 10
Totals 130 46 57 27 21 41 40 28
BULLETIN FLORIDA STATE MUSEUM
TABLE 2.-Injuries suffered by tortoises while being pulled from burrows.
%
Type of injury Location n of total
Broken shell Anterior edge of carapace 7 25.9
Axillary notch 5 18.5
Anterior carapace and plastron 2 7.4
Anterior and posterior carapace, 1 3.7
and posterior plastron
Punctured skin Axillary notch 10 37.0
Inguinal notch 1 3.7
Torn flesh Tail 1 3.7
27
jury than did larger individuals. This is undoubtedly due to the much
thinner and less ossified (and therefore weaker) shells of the smaller
animals.
Additionally, the percentage of females injured during the pulling
process was greater than that of males. Of 22 individuals injured for
which the sex was known, 15 were females while only 7 were males (Table
3). When considering only severe damage (i.e. broken shell), the dif-
ference between the sexes was even more striking. Of the 15 individuals
which suffered injury to the shell, 12 were females (7 small and 5
medium) while only 2 were males (1 small and 1 medium). The sex of one
was unknown. This differential susceptibility to injury resulted from a
difference in shell thickness (and strength) between the sexes. Identical
circular plugs were cut from the shells of 36 freshly killed tortoises (16
males and 20 females) at the level of the third pleural bone, were air
dried, and the scute removed. The weights of these bone plugs were com-
pared by sex using a paired observation t-test. Fourteen pairs of observa-
tions were obtained in which the male and female value of total length
differed by no more than 8 mm. The plug weights of these 14 pairs were
then tested, and showed a significant difference at P<0.05 (df = 12).
This difference indicates that males have stronger shells than females and
explains why the latter are injured more often during pulling.
Two additional factors influence the frequency of damage suffered by
tortoises: the relative sharpness of the very tip of the gopher hook, and the
length of this terminal portion. During manufacture of the apparatus
most persons take care to file the end to a rounded point and not to have
the bent portion too long. If either of these precautions is not taken the
chance of puncturing or slicing into the shell or soft tissues is greatly in-
creased. Also, if a root or other obstruction is encountered while pulling
the tortoise from the burrow, a greater force is exerted on the animal's
Vol. 28, No. 4
|
