1

\Ui 770M

HARVARD UNIVERSITY

v

«

LIBRARY

OF THE

Museum of Comparative Zoology

-MA -N(jew Or|e

VOLUME 2
1954-1955

TULANE UNIVERSITY
NEW ORLEANS

Jilt 1 9 «*■

TULANE STUDIES IN ZOOLOGY is devoted primarily to the
zoology of the area bordering the Gulf of Mexico and the Caribbean
Sea. Each number is issued separately and deals with an individual
study. As volumes are completed, title pages and tables of contents
are distributed to institutions exchanging the entire series.

Manuscripts submitted for publication are evaluated by the editor and
by an editorial committee selected for each paper. Contributors need
not be members of the Tulane University faculty.

MEMBERS OF THE EDITORIAL COMMITTEES FOR
PAPERS PUBLISHED IN THIS VOLUME

Reeve M. Bailey, University of Michigan

Arthur N. Bragg, University of Oklahoma

Fred R. Cagle, Tulane University

R. V. Chamberlain, University of Utah

Ralph W. Dexter, Kent State University

Llewellyn T. Evans, Long Island Biological Association

Gordon Gunter, University of Texas

Robert A. Hefner, Miami University

Richard L. Hoffman, Cornell University

Carl L. Hubbs, Scripps Institution of Oceanography

Arthur G. Humes, Boston University

Ernest A. Lachner, U. S. National Museum

H. F. Loomis, U. S. Department of Agriculture

G. Robert Lunz, Bears Bluff Laboratories

J. G. Mackin, Texas A. & M. College

Ernst Mayr, Harvard University

Malcolm R. Miller, Stanford University

Walter G. Moore, Loyola University

J. Bennett Olson, Children's Hospital Society, Los Angeles

Karl P. Schmidt, Chicago Natural History Museum

Royal D. Suttkus, Tulane University

Harry C. Yeatman, University of the South

CONTENTS OF VOLUME 2

NUMBER PAGE

1. A NEW EULINMADIA FROM THE RICE FIELDS OF
ARKANSAS WITH A KEY TO THE AMERICAN SPE-
CIES OF THE GENUS (CONCHOSTRACA, LIMNADII-
D A E )

N. T. Mattox 1

2. STUDIES IN THE ECOLOGY OF THE NARROW-
MOUTHED TOAD, MICROHYLA CAROLINENSIS CARO-
LINENSIS _

Paul K. Anderson 13

3. A NEW SPECIES OF DIAPTOMUS FROM LOUISIANA
AND TEXAS WITH NOTES ON THE SUBGENUS
LEPTODIAPTOMUS (COPEPODA, CALANOIDA) ....

Mildred Stratton Wilson 49

4. THREE NEW SPECIES AND NEW RECORDS OF
SOUTHERN MILLIPEDS

Nell B. Causey 61

5. NOTROPIS BAILEY I, A NEW CYPRINID FISH FROM
THE PASCAGOULA AND MOBILE BAY DRAINAGES
OF MISSISSIPPI AND ALABAMA

Royal D. Suttkus and Edward C. Raney 69

6. A RECONSIDERATION OF THE RACER, COLUBER
CONSTRICTOR, IN EASTERN UNITED STATES _

Walter Auffenberg- 87

7. NOTROPIS HYPSILEPIS, A NEW CYPRINID FISH
FROM THE APALACHICOLA RIVER SYSTEM OF
GEORGIA AND ALABAMA

Royal D. Suttkus and Edward C. Raney 159

8. THE FISHES OF TWO GULF COASTAL MARSH
AREAS OF FLORIDA

John D. Kilby 173

Printed in the U.S.A.

at New Orleans, by

Hauser Printing Co., Inc.

O -jv n~i\tuj &ile_etu.

^©aasfs sip ©©as

SS9 ^(D®IL(D@^

Volume 2, Number 1

September 14, 1954

cVvu t&£iix <*U^ cv~ir

,

A NEW EULIMNADIA FROM THE RICE FIELDS OF ARKANSAS
WITH A KEY TO THE AMERICAN SPECIES OF THE GENUS

(CONCHOSTRACA, LlMNADIIDAE)

N. T. MATTOX

DEPARTMENT OF ZOOLOGY, ALLAN HANCOCK FOUNDATION,
LOS ANGELES, CALIFORNIA.

LI
SEP 2 4 1954

N '5

TULANE UNIVERSITY
NEW ORLEANS

TULANE STUDIES IN ZOOLOGY is devoted primarily to the
zoology of the area bordering the Gulf of Mexico and the Caribbean
Sea. Each number is issued separately and deals with an individual
study. As volumes are completed, title pages and tables of contents
are distributed to institutions exchanging the entire series.

Manuscripts submitted for publication are evaluated by the editor
and by an editorial committee selected for each paper. Contributors
need not be members of the Tulane University faculty.

EDITORIAL COMMITTEE FOR THIS NUMBER

J. G. Mackin, Professor of Marine Biology, Marine Laboratory
of Texas A & M College, Galveston, Texas.

Ralph W. Dexter, Professor of Biology, Kent State University,
Kent, Ohio.

Walter G. Moore, Associate Professor of Biology, Loyola Uni-
versity, New Orleans, Louisiana.

Manuscripts should be submitted on good paper, as original type-
written copy, double-spaced, and carefully corrected.

Separate numbers may be purchased by individuals, but subscriptions
are not accepted. Authors may obtain copies for personal use at cost.

Address all communications concerning exchanges, manuscripts, edi-
torial matters, and orders for individual numbers to the editor. Re-
mittances should be made payable to Tulane University.

When citing this series authors are requested to use the following
abbreviations: Tulane Stud, Zool.

Price for this number: $0.25.

Assistant to the Editor:
Carol L. Freret

George Henry Perm, Editor,
c/o Department of Zoology,
Tulane University,
New Orleans, U. S. A.

.. '. .

:•

SEP 2

4 7954

>

A NEW EUL1MNAD1A FROM THE RICE FIELDS OF ARKANSAS
WITH A KEY TO THE AMERICAN SPECIES OF THE GENUS

(CONCHOSTRACA, LlMNADIIDAE)

N. T. MATTOX

Department of Zoology, Allan Hancock Foundation,

Los Angeles, California. 1

During the course of studying mosquito larvae in the rice fields near
Stuttgart, Arkansas, Dr. F. E. Whitehead, Professor of Entomology at
the University of Arkansas, encountered numerous specimens of a con-
chostracan phyllopod which were forwarded to the writer for deter-
mination. This, a new species, represents the second of the genus
Eulimnadia to be collected in these rice fields by Dr. Whitehead. The
other species, Es alineata Mattox, was collected near Stuttgart in August
1948 and again in June 1949. This presents an interesting ecological
situation with two so different species occurring in the same region
and in the same ecological niche. The present species is radically dif-
ferent from alineata. The name Eulimnadia oryzae is here proposed for
this new species. The specific name refers to Oryza, the generic name
of the rice plant.

EULIMNADIA ORYZAE, sp. nov.

Male. — With the characters of the genus (Packard, 1883, and Daday,
1926). The bivalved shell is practically colorless, transparent, and
elongate-oval (fig. 1). The dorsal margin is only slightly rounded with
the highest elevation at the umbo approximately one-third the length
from the anterior margin. The ventral margin is regularly rounded, the
posterior end is more elongate than the anterior. The length of the
adult shell averages 6.8 mm with an average height of 4.2 mm. The
variations in the type lot are 7.5 by 4.5 mm to 6.3 by 3.8 mm. The
lines of growth vary in number from 10 to 12, and are usually crowded.

The head possesses the typical pyriform frontal organ which is lo-
cated near the base of the eye protuberance (fig. 6). The front is
moderately concave, the greatest concavity just above the ocellus; the
rostrum is not greatly extended. The head is held in a position such as
to place the eye protuberance directly forward; the head gives the im-
pression of being quite small, much smaller than that of the female.
The scape of the second antennae is very elongate, one half the length
of the scape extends beyond the tip of the rostrum. The flagella of the
second antennae are variously spined (not figured) and have nine seg-
ments each, the two flagella are of approximately the same length.
The first antennae extend to the third segment of the second antennae
flagella. The first antennae possess 12 to 15 dorsal sensory papillae.

The body bears 18 pairs of swimming appendages; the first and
second are modified into the typical gnathopods. The two segments
of the digitiform sixth endite of the first gnathopod are unequal, the

1 Allan Hancock Foundation Contribution No. 136.

.- i

4 Tulane Studies in Zoology Vol. 2

distal segment is approximately three times the length of the basal
segment ( f ig 4 ) . The distal segment of the more elongate sixth endite
of the second gnathopod is approximately twice the length of the basal
segment ( fig. 5 ) . There is a very conspicuous notch at the base of the
"thumb" of the fourth endite on both the first and second gnathopod.
The posterior 9 or 10 body segments bear mid-dorsal spines.

The telson is elongate, rhomboid in outline, the ventral margin only
slightly more than one half the entire length of the dorsal spined mar-
gin (fig. 3). The dorsal ridges of the telson bear from 14 to 20 spines;
15 is the average number. The terminal telson claws (cercopods) are
elongate and tapered, extending posteriorly and are longer than the
dorsal margin of the telson. The dorsal forked filament arises between
the third and fourth pair of telson spines.

Female. — The female shell (fig. 2) is similar to the male but much
more ovate in outline. The dorsal margin is much more convex than
in the male with the greatest height approximately one-third the dis-
tance from the anterior end. There is a slight concavity on the posterior
portion of the dorsal margin. The average size of the adult female shell
is 8.0 mm in length by 5.4 mm in height. The number of lines of
growth averages 11, with a variation of 9 to 12 in the adult (egg
bearing) females.

The front of the head is evenly and shallowly concave with a less
pronounced rostrum than the male (fig. 7). In general outline the
head of the female is much larger and more quadrate than the male.
The second antennae are similar to those of the male. The first an-
tennae are much shorter than in the male extending only to the first
segment of the second antennal flagella. The occipital notch, as in the
male, is very conspicuous.

The 18 pairs of swimming legs are unspecialized except for the
elongated ovigerous epipodites of the ninth and tenth pairs. This
epipodite is 2.5 times longer than the exopodite on the tenth pair (fig.
8). The dorsal body spines and the telson are similar to those of the
male.

Type locality. — Rice fields at Stuttgart, Arkansas County, Arkansas.

Types. — Holotype, male, U.S.N.M. No. 96032; paratypes, both sexes,
U.S.N.M. No. 96033; and in writer's collection.

Remarks. — Eulimnadia oryzae seems to be closely related to the east-
ern species E. ventricosa Mattox (1953), which is known from Vir-
ginia, Maryland, and Georgia. The shell of male oryzae is more elongate
in outline and more attenuated posteriorly. The shell of females is not
as full or elevated as in ventricosa; the length-height ratio of 1.5:1 as
against 1.3: 1 for ventricosa. The concavity on the posterior slope of ory-
zae is more pronounced than in ventricosa. The absence of color in the
shell is distinctive for oryzae. The scape of the second antennae is longer
in oryzae; the first antennae are proportionately shorter; the front is
less concave; the telson is more elongate, and the cercopods longer in
oryzae than in ventricosa. The basal segment of the sixth endite on the

No. 1

Mattox: New Eulimnadia from Arkansas

Figures 1-8. Eulimnadia oryzae, sp. nov. 1. Shell of male; 2. Shell of
female; 3. Telson; 4. First gnathopod of male; 5. Second gna-
thopod of male; 6. Profile of head of male; 7. Head of female;
8. Tenth swimming leg of female with egg on ovigerous epipo-
<dite. Scale lines for figures 4 and 5 equal 0.5 mm; all others
are 1.0 mm.

Tulane Studies in Zoology Vol. 2

first gnathopod of oryzae is proportionately shorter; the terminal seg-
ment of the sixth endite of the second gnathopod is proportionately
longer than in ventricosa. The epipodite of the tenth leg of the female
in oryzae is 2.5 longer than the exopodite, in ventricosa this proportion
is less than 2.0. The number of growth lines on the shell and the
number of telson spines are similar in oryzae and ventricosa, characters
which separate these species from other North American forms. E.
alineata, which has been taken from the rice fields in this same region,
is smaller in size and is distinctive in lacking growth lines on the shell.

This species is based on one collection containing seven male and 33
female individuals taken on June 30, 1953 by Prof. F. E. Whitehead.
Dr. Whitehead has indicated that these animals were so abundant that
he believed there were actually thousands of them in a single cubic foot.
The ecological notes, provided by Dr. Whitehead, in regard to the
occurrence of these animals are worthy of note.

The rice fields typically are plowed in early spring and drilled to
rice in April or May. Three to five weeks later, or when the rice is
four to eight inches tall, the fields are flooded. After three to four
weeks the fields are drained for the control of the rice water weevil.
When the soil has become dry they are again flooded and kept under
water until a week or ten days before harvest. The phyllopods occur
in both the first and second flooding, but it is two or three weeks fol-
lowing the second flooding that they are most frequently extremely
numerous. This is another indication of the rapid rate of development
found in this group. E. diversa Mattox (1937) was shown to mature
in 14 days.

Another interesting observation made by Dr. Whitehead indicates an
unexpected resistance of these animals to insecticides. During field ex-
periments a number of chlorinated hydrocarbon insecticides gave almost
complete control of the mosquito larvae and produced a high mortality
among other aquatic invertebrates present. However, without definite
quantitative data it was observed that even at the higher strengths of
insecticides there was an extremely low mortality among the phyllopods.
These observations suggest some unknown physiological resistance pos-
sessed by Eulimnadia and possibly by other members of the order. To
the best of the writer's knowledge the Conchostraca have not been ex-
tensively studied so far as physiological properties are concerned.

The North American species of EULIMNADIA

The genus Eulimnadia was proposed by Packard ( 1874) in order to
separate this genus from Limnadia Brongniart 1820, the other member
of the family Limnadiidae. This family is comprised of conchostracan
phyllopods, those with a bivalve shell, possessing a pyriform, pedicu-
lated frontal organ on the mid dorsal surface of the head. The 18
pairs of trunk appendages, the conspicuous ventral spine on the telson
at the base of the cercopods, and the second antennae flagella with 9
segments are the distinctive characters of the Eulimnadia^ These charac-

No. 1 Mattox: New Eulimnadia from Arkansas 7

ters were established by Daday in his monograph of the conchostracans
in 1926.

The various Eulimnadia, as well as most of the conchostracans, usu-
ally appear in temporary pools during the warm months of the year.
Some collections have been made in very small pools of rain water
that appeared and dried up within two weeks. Many of the species of
this genus are known only from their type localities indicating a very
restricted range. Other species seem to be more widely dispersed. Mem-
bers of this genus do not attain large size; all are less than 10 mm in
shell length; most average 5 or 6 mm.

The morphological characters most used in separating the species
are: the lines of growth on the shell; the size and shape of the shell;
the telson spination; the form of the head; the male gnathopods, and
the relative lengths of the antennae. The following is a key to the
known species of the genus Eulimnadia found in North America. This
key is intended only for sexually mature individuals. Maturity is
judged on the basis of the presence of gravid females carrying egg
masses on the modified epipodites of the ninth and tenth legs.

Key to the Genus EULIMNADIA of North America

1. Shell with lines of growth 2

Shell with no lines of growth; 4.2 by
2.6 mm average; 9 to 12 telson
spines; forked filament of telson be-
tween spines 3 and 4. Eulimnadia alineata Mattox 1953

Known only from rice fields at Stutt-
gart, Arkansas.

2. Shell with one to four lines of growth;

elongate, not strongly convex dorsally 3

Shell of females with 5 to 12 growth
lines; usually ovate and convex dor-
sally 7

3. Telson with 9 or 10 dorsal spines 4

Telson with 12 to 16 dorsal spines 5

4. Shell 5 to 6 mm long by 3 to 4 mm wide;

one to four growth lines; rostrum
rounded; forked filament arises be-
tween telson spines 3 and 4

Eulimnadia antillarum (Baird 1852)

Has been taken in Louisiana and
Mexico.

Shell 4.3 by 2.5 mm average; one to four
growth lines; rostrum pointed; first
antennae extend to fifth segment of
second antennae in male; forked fila-
ment arises between telson spines 2
and 3. Eulimnadia francesae Mattox 1953

From pool's on Bear Island, Potomac
River, Montgomery Co., Maryland.

8 Tulane Studies in Zoology Vol. 2

5. Shell normally with two growth lines;

shell of males averages 4.2 by 2.5
mm; front of head slightly convex;
first antennae of male extend to
fourth segment of second antennae;
forked filament of telson between
spines 3 and 4. Eulimnadia diversa Mattox 1937

Has been found only at Urbana,
Illinois.

Shell with three or four lines of growth ;

6 to 7.5 mm in length 6

6. Telson with 12 dorsal spines; forked

filament arising between telson
spines 1 and 2; rostrum not pointed
and inflected; shell size average 6.2
by 3.8 mm; first antennae of male do
not extend beyond scape of the sec-
ond antennae. Eulimnadia agassizii Packard 1874

Known only from Penikese Island,
Massachusetts.

Telson with 16 dorsal spines, forked fila-
ment arising between spines 6 and 7 ;
rostrum strongly pointed and in-
flected; males shell size average 7.3
by 4.3 mm; first antennae of male
extend to fourth segment of second
antennae. Eulimnadia inflecta Mattox 1939

Found at Prophetsown, Illinois, and
Athens, Ohio.

7. Female with five lines of growth; ma-

ture size less than 8 mm in length 8

Lines of growth 7 to 12; mature size

more than 8 mm 9

8. Shell size averages 5 by 3 mm; 7 to 9

telson spines; male first antennae
slightly longer than those of female;
rostrum of male extended and sharp-
ly pointed. Eulimnadia antlei Mackin 1940

Taken at three localities in Okla-
homa.

Shell size averages 7 by 4 mm; tel-
son with 16 to 20 spines; male first
antennae extend to third segment
of second antennae, in female short-
er; rostrum rounded Eulimnadia texana Packard 1871

Widely distributed in Texas, Kansas,
Nebraska, and Oklahoma.

9. Male first antennae extend beyond scape

of second antennae; forked filament
of telson arises between spines 3
and 4. 10

No. 1 Mattox: New Eulimnadia from Arkansas 9

Male first antennae extend only to end
of scape of second antennae; forked
filament of telson arising between
spines 5 and 6; 14 pairs of telson
spines; average of 10 growth lines;
shell size averages 8.5 by 6 mm
Eulimnadia stoning tonensis Berry 1926

Has been found only at Stonington,
Connecticut.

10. Male first antennae extend to second

segment of second antennae; 14 to
20 pairs of telson spines; average
size of shell 8.1 by 5.5 mm; females
have an average of seven growth
lines; rostrum of male not attenuated
anteriorly. Eulimnadia thompsoni Mattox 1939

Found in pools on an island in the
Illinois River at Ottawa, Illinois.

Male first antennae extend to third seg-
ment of second antennae; 10 to 12
growth lines. 11

11. The male rostrum is attenuated to a

sharp point, slightly inflected, front
concave; shell very ventricose; shell
length-height ratio 1.3:1; telson cer-
copod length same as dorsal margin
of telson; 14 to 16 dorsal telson
spines; second antennal scape extends
to tip of rostrum. Eulimnadia ventricosa Mattox 1953

Taken at several' localities in Mary-
land, Virginia, and Georgia.

The male rostrum pointed but not
greatly inflected; front slightly con-
cave near rostral base; the occipital
notch is conspicuous; growth lines
crowded; shell length-height ratio
1.5:1; telson cercopods longer than
dorsal' telson margin ; 14 to 20 dorsal
telson spines; telson elongate; second
antenna! scape extends one-half
length beyond rostrum. Eulimnadia oryzae sp. nov.

From ricefields at Stuttgart, Arkan-
sas.

10 • Tulane Studies in Zoology Vol. 2

References Cited

Daday, Eugene 1926. Monographie systematique des phyllopodes con-
chostraces, Troiseme partie. Ann. Sci. Nat. Zool., Ser. 10, 11: 1-81.

Mattox, N. T. 1937. Studies on the life history of a new species of
fairy shrimp, Eulimnadia diversa. Trans. Amer. Micros. Soc, 61
(2) : 249-255.

1953a. A new conchostracan phyllopod, Eulimnadia

alineata, from Arkansas. Amer. Midi. Nat., 49 (1) : 210-213.

1953b. Two new species of Eulimnadia from Mary-
land and Virginia (Crustacea: Conchostraca). Jour. Wash. Acad.
Sci., 48 (2) : 57-60.

Packard, A. S. 1883. A monograph on the phyllopod Crustacea of
North America. 12th. Ann. Rept. U. S. Geol. and Geogr. Surv*
Terr., 1878 (Hayden Survey), Part I: 295-457.

/VQ-|\J/e/^

ts_^ l^£V"0

3-1

V192L&S93

a

TO®ais

Volume 2, Number 2

November 8, 1954

STUDIES IN THE ECOLOGY OF THE NARROW-MOUTHED

TOAD,

M1CROHYLA CAROLINENSIS CAROLINENSIS

PAUL K. ANDERSON,
ORANGE, MASSACHUSETTS

*ua. CGMP. ZGOL
LIBRAS)

YOV 1 9 1954

mm

UNIVERSITY

;

TULANE UNIVERSITY
NEW ORLEANS

TULANE STUDIES IN ZOOLOGY is devoted primarily to the
zoology of the area bordering the Gulf of Mexico and the Caribbean
Sea. Each number is issued separately and deals with an individual
study. As volumes are completed, title pages and tables of contents
are distributed to institutions exchanging the entire series.

Manuscripts submitted for publication are evaluated by the editor
and by an editorial committee selected for each paper. Contributors
need not be members of the Tulane University faculty.

EDITORIAL COMMITTEE FOR THIS NUMBER

Arthur N. Bragg, Professor of Zoology, University of Okla-
homa, Norman, Oklahoma.

Malcolm R. Miller, Assistant Professor of Anatomy, Stanford
University, Stanford, California.

Llewellyn T Evans, Research Associate, Long Island Biologi-
cal Association, Cold Spring Harbor, New York.

Manuscripts should be submitted on good paper, as original type-
written copy, double-spaced, and carefully corrected.

Separate numbers may be purchased by individuals, but subscriptions
are not accepted. Authors may obtain copies for personal use at cost.

Address all communications concerning exchanges, manuscripts, edi-
torial matters, and orders for individual numbers to the editor. Re
mittances should be made payable to Tulane University.

When citing this series authors are requested to use the following
abbreviations: Tulane Stud* Zool.

Price for this number: $0.50.

Assistants to the Editor:
Carol L. Freret
Donald W. Tinkle

George Henry Penn, Editor,
c/o Department of Zoology,
Tulane University,
New Orleans, U. S. A.

STUDIES IN THE ECOLOGY OF THE NARROW-MOUTHED

TOAD,

MICROHYLA CAROLINENSIS CAROLINENSIS1

PAUL K. ANDERSON,2

Orange, Massachusetts

Microbyla has received little attention from students of the Am-
phibia. The most extensive material is that of Wright (1932), Bragg
(1943, 1950a, 1950b), and Freiburg (1951). This study contributes
to our knowledge of the subspecies Microhyla carolinensis carolinensis
and to our general understanding of anuran behavior.

The nomenclatural history has been outlined by Hecht and Matalas
(1946) who concluded that M. olivacea and M. mazatlanensis should
be considered as subspecies of M. carolinensis* Their view is followed
in this paper but further evaluation of the systematics of these popu-
lations is required.

Methods

The detailed procedures for both the field and laboratory researches
are presented in pertinent sections of this report; only the general
approach and sources of material are considered here.

Field studies during an eighteen-month period gave the author an
opportunity to become thoroughly familiar with the species. Knowl-
edge gained from seven weeks of observation in Louisiana and Florida
was especially valuable.

Population samples were taken at several places near New Orleans,
and from pine forests near Angie, Washington Parish, Louisiana. The
New Orleans area material came from four localities: levees along
the Bonnet Carre Spillway ( St. Charles Parish, La. ) ; levees along the
Mississippi River near the Huey P. Long Bridge (Jefferson Parish,
La. ) ; abandoned fields at Camp Plauche ( near Harahan, Jefferson
Parish, La. ) ; and, cypress swamps in Tulane University's Sarpy Wild-
life Refuge (near Norco, St. Charles Parish, La.).

From the examination of 579 individuals (fresh and preserved)
data were gathered for study of population composition, reproductive
cycles, growth, and food habits. Snout-vent length measurements
were made with a vernier caliper. Secondary sex characters were
noted. The abdominal cavity was opened and the gonads were ex-
amined. Stomach contents of 203 individuals were analyzed.

Breeding behavior was observed in the field and laboratory. Ma-

1 From a thesis submitted in partial fulfillment of the requirements
for the degree of Master of Science of the Graduate School of Tulane

University.

2 The work was done under the direction of Dr. Fred R. Cagle, to
whom the author wishes to express his gratitude for the inspiration
and criticism that made the study possible. Dr. Harley N. Gould, of
Newcomb College (now emeritus), also gave generous aid in inter-
pretation of the cytological material.

£00L

tav

19

1954

;SiTY

16 Tulane Studies in Zoology VoL 2

cerated fresh pituitary glands were injected intraperitoneal^ to stimu-
late breeding activity in the laboratory.

Environment
Environmental Factors

Field observations suggest that there are two environmental factors
which limit, more than any others, the distribution of this species in
Louisiana. These two factors, moisture and cover, are closely related.
The importance of moisture to these frogs and their extreme suscepti-
bility to drying were demonstrated in trapping studies at Camp
Plauche.

Traps used in these studies were constructed from 1/16" mesh
wire screening. The area was an abandoned and partially overgrown
dump, providing abundant cover in the form of piles of tar paper,
plaster, tin cans, and scattered boards and timbers. There was a
sparse to dense covering of low grasses with an upper stratum of
dense weeds including such forms as ragweed {Ambrosia spp.) and
goldenrod (Solidago spp.). In the immediate area of the traps there
was a stand of low willow (Salix nigra) which shaded some of them.
Despite the shade it was necessary to check these traps within two
hours after sunrise in order to prevent death of almost all individuals
from desiccation. Even on cloudy days death resulted within four or
five hours if the animals were not removed from their exposed position.

Desiccation makes these frogs vulnerable to predators, especially
small ants (Iridomyrmex humilis) .

Another apparent effect of moisture was the variation in the ef-
fectiveness of hand collecting methods at various periods during the
year in the Camp Plauche area. In the moist spring months Microhyla
was abundant under surface cover such as logs, old cardboard, and
other debris. As the weather became drier, specimens became more
and more difficult to discover in this manner. All cover had been
carefully replaced and no animals had been removed from the area.
By August and Stepember several hours of intensive search failed to
reveal a single specimen. Possibly there had been a migration of
the population from the immediate area to contiguous places, but
thorough search failed to reveal individuals. Sporadic trapping dur-
ing this period produced few frogs. This indicates that at least a
few individuals remained and sought out retreats at some depth in
the earth. By the following spring the accessible population had
returned to its original level.

Many other factors influence the distribution of the frogs. Sub-
strate in the habitat may range from permanently or temporarily
flooded swamp or marsh to dry sandy pinelands. The amount of
litter, duff, and humus differs greatly. That the substrate does affect
population density was indicated by studies of pineland populations.
Anderson, Liner and Etheridge (1952) found that the size of Micro-
hyla populations in a Louisiana pineland area fluctuated with the
depth of the organic layers of the soil.

No. 2 Anderson: Ecology of Microhyla 17

In all habitats there seems to be a distinction between simple over-
head cover, as characterized by boards, paper, tin cans, logs, or stones,
and more complicated covers, where various crevices, crannies, or
burrows are used. In cypress swamps near New Orleans Microhyla
is often found buried in rotten logs and stumps, usually in relatively
loose, soft wood. One individual was taken in a dead stump eight
feet above the ground. Stumps have also provided the greatest con-
centrations thus far discovered. On March 15, 1950, a low hollow
stump with an inside diameter averaging thirty inches produced
twenty-six juveniles in a stratum six inches thick. Crevices between
leaf bases of palmetto plants (Sabal minor) are used by Microhyla
in swamps and in open thickets at the edge of the marshes. These
log and palmetto covers often may be completely isolated by water
of considerable depth during part of the year. Piles of hay or straw,
and other decaying vegetation, such as the muskrat houses in marshes,
are also favored cover types. All of these covers have at least two
common factors, moisture and a loose structure which permits easy
penetration.

Under certain conditions a short semi-burrow is constructed by the
frog. Typically these are found where there is relatively loose over-
head cover and the soil is granular. Under these circumstances a
pear-shaped depression is hollowed out so that only the top of the
animal's head is visible. This behavior was commonly shown by
individuals on open levees.

Location of suitable breeding sites is also a factor that can exercise
a limiting effect upon any anuran species. The wide adaptability of
the present species, as shown in detail in a following section on
breeding behavior, makes it unlikely that this factor can be of major
importance anywhere in Louisiana.

Habitats
The foregoing remarks upon certain of the ecological requirements
provide a groundwork for discussion of the ecological distribution of
the species in Louisiana. Plant associations provide the best key to
this distribution. The habitats favored by Microhyla may be divided
into forest communities and non-forest or field and marsh communities.

Microhyla occurs in four major forest types: cypress-gum swamp,
bottomland hardwood, live-oak ridges, and pine-oak uplands. The
bottomland hardwoods and live-oak ridges have received little study.

A cypress-gum swamp in the Sarpy Wildlife Refuge has been
studied intensively. The area is covered with a second growth cypress-
gum forest interspersed with pure stands of cypress (Taxodium
distichum) and of cut grass (Zizaniopsis miliacea) marsh. There
are numerous "borrow pits" along the roads, and several sloughs and
bayous partially covered by duckweed (Lemna spp.), water hyacinth
(Eichhornia crassipes) and water fern (Azolla caroliniana) . The
water level fluctuates over a three foot interval. During a fourteen-
month period the forest floor was uncovered for only about four

18 Tulane Studies in Zoology Vol. 2

months in late spring and summer. When the water is at a depth
of only six or eight inches, there are numerous stumps, logs and
hummocks exposed. The most productive collecting sites are float-
ing or partly sunken logs, stumps, and the bases of palmetto plants.

The pine-oak community is essentially a dry environment in the
areas studied. Longleaf pine (Pinus palustris) and post oak (Quercus
stellata) are major forest trees. Along small streams and in moist
depressions there are cypress-gum associations. Frequent burning
removes much of the litter and the soil often has no considerable
organic layer. Despite this, there appear to be large amphibian and
reptile populations.

The non-forest habitats differ markedly from the forest types. They
are abandoned fields, grassy levees, short grass areas (such as pastures,
lawns, and golf courses), and marshes. Pure palmetto stands seem
best considered as an intermediate type as they most often occur along
the boundary zone between forest and marsh.

In abandoned fields and grassy levees the cover is about the same
although the vegetation is different. The levees are typically covered
with low dense grass, but may be grazed and thus belong in the short
grass group. Abandoned fields are characterized by a mixture of
grasses, weeds, and shrubs.

In urban areas, such as lawns and golf courses, the evaluation of
the cover is difficult. There are always hedgerows or gardens where
some cover exists, but thorough examination of such areas is difficult
and no definite conclusion as to cover preference can be made.

Marshland communities are somewhat analagous to the cypress
swamps in respect to water relationships, but have very different
cover facilities. Such ridges as may be present may support live-oak
{Quercus virginiana) or willow. Dr. Cagle tells me that where
muskrats are present the frogs have been found in number in the
lodges. Any hummock of organic material above the surface will
often have one or more individuals buried in it.

Behavior

Diet Activity Cycle

Captive animals in outdoor terraria were observed at hourly or
half-hourly intervals through the late afternoon and evening. The
frogs burrowed in rotten wood during the day and moved to the
surface at early dusk. Most individuals watched from concealment
until it became almost fully dark and then began to move about in
the open. Activity was typically much reduced by 10 to 11 P.M.
Usually only one or two individuals out of six to ten in the terrarium
were visible after midnight, whereas all were visible at dusk.

Field observations in Audubon Park, New Orleans, substantiated
this pattern. Individuals were found moving about on the short grass
of the golf course on many occasions at night. The peak of abundance
was usually reached at the time full darkness began.

No. 2 Anderson: Ecology of Microbyla 19

Movement

Locomotion on land is usually described as a scramble or crawl.
This describes the escape behavior when disturbed. Locomotion of
undisturbed animals may be described as a walk. Individuals have
been observed in terraria and in the field to walk slowly and rather
deliberately with the body well off the ground. When disturbed in
the open where there is no cover readily available they will hop, cover-
ing distances up to twelve inches. They do not land well when
hopping and usually fall at an angle so that they roll and have to
regain their balance before starting the next hop, often in a different
direction. Such erratic movement could have escape value.

In the water they are buoyant and swim well. Pope's statement
(1919) that they do not dive is incorrect. Calling males and pairs
in amplexus usually dive immediately when disturbed and individuals
of both sexes floating in flood waters in the Bonnet Carre Spillway
dived readily to escape capture.

Trapping yielded data on movement. The traps used at Camp
Plauche were constructed of fine wire screen stapled to wooden
frames. The frame consisted of one-inch strips nailed together to
form a square, ten inches on a side. Between two square frames a
single heavy piece of board, one or two inches wide and eighteen
inches long was nailed. Screening was then attached to this frame
to form a rectangular cage. The sides and one end were closed and
a funnel of the same screening was set into the open end. This
funnel projected inward for about eight inches and had an opening
about four inches in diameter at the inner end. Drift fences eight
to ten inches in height were constructed of asphalt siding stapled
to stakes at intervals of about two feet. The angle of the fences
leading into the mouth of the trap was about forty-five degrees.
Before the fence was put down, the ground was scraped bare and, if
possible, a shallow trench three inches wide was dug. The lower
edge of the fence was bent and laid flat to the ground in the trench
to form an apron on the inner side of the V, earth and debris were
scraped over it and packed down to discourage animals from burrow-
ing beneath the fence. Since it was hoped to study breeding move-
ments, as well as normal activity, with these traps, they were placed
around a small temporary pond. Five traps were used at the start
of the study. All traps faced outward and were connected with an
unbroken line of fence so that the pond was entirely fenced off. The
line of the traps and the fence formed a clover-leaf around the pond
with the traps at the angles between the leaflets.

Towards the end of the trapping period, which lasted from April
19 to May 22, two outgoing traps were added. They did not operate
efficiently since there was no apron on the pond side of the fence
and escape beneath it was probably not difficult.

All individuals found alive in the traps were tattooed with a
number on the under surface of the thigh, sexed, measured, and re-
leased on the inner side of the fence. Since one of the returns had

20 Tulane Studies in Zoology Vol. 2

a faded and illegible number, it was feared that the markings were
not holding and that many of the supposedly new individuals caught
in the traps might be recaptures. For this reason six frogs were
tattooed and kept in rotten wood in a moist terrarium for a period
of two months and their markings were periodically examined. Al-
though there was considerable fading, all but one of the numbers
were clearly legible. The one illegible mark had faded, but there was
no possibility of overlooking the fact that the individual had been
marked. Therefore it is assumed that all unmarked individuals taken
in the traps were being caught for the first time.

Trapping results show that considerable movement must have oc-
curred (fig. 1). The first ten days of trap operations preceded the
opening of the breeding season. In this period the average nightly
catch was 2.7 individuals (1.0 males, 0.7 females, 0.7 juveniles).
After the start of the breeding season there were twenty-three days
of trap operation with an average nightly catch of 3.1 individuals
(2.0 males, 0.45 females, and 0.5 juveniles). Nine trapped speci-
mens were found dead and dried. These were included in the average
nightly catch but could not be sexed or aged and thus cause a slight
discrepancy between the two sets of figures. There were only five
recoveries of marked individuals. All of these were less than six
days after release and most were taken by lifting logs or other cover
inside the fence. One was taken six days after its release inside the
fence in an outgoing trap.

It appears that a large population was present and that there was
considerable individual movement, both before and during the breed-
ing season. An increase in total movement during the breeding sea-
son seems to be indicated by the trapping data. This increase occurs
entirely among the males, and there is an apparent decrease in move-
ment among females and juveniles.

The relative difference in movement of the sexes is strikingly
demonstrated by examining the three large population samples which
were based on movement and comparing them with samples taken
by hand collecting. The sex ratios of all three samples are strongly
in favor of males. Of 69 adults trapped at Camp Plauche, 52 were
males and 17 were females (fig. 1). A sample taken from a pipe-
line ditch (Anderson, Liner, and Etheridge, 1952) during the breed-
ing season consists of 147 males and 80 females. A February sample
from an aggregation caused by rising flood waters has less disparity
between the sexes (94 males and 55 females). The sex ratio was
close to 1:1 in samples taken by hand collecting. Of 57 individuals
taken in March, 29 were males and 28 females. Twenty-two of these
came from the cypress swamps of the Sarpy Wildlife Refuge and 35
from the levee near the Huey P. Long Bridge. The swamp sample
was made up entirely of immature individuals and consisted of 11
males and 11 females. The levee sample had 18 males (7 adults and
11 juveniles) and 17 females (13 adults and 4 juveniles).

These contrasting data in regard to sex ratios present a problem

No. 2

Anderson: Ecology of Microhyla

21

a^

i i i

->

B B-Bn-rzB

B B

.'UVENILES

fffl ■ n ■ m n

I i I i I I i l I I I I I I I I i I I I I I i i i I i I ) I ) i i i

20 21 22 23 24 25 26 27 28 29 30 I 2 3 4 5 6 7 6 9 10 II 12 13 14 15 16 17 18 19 20 21 22

Figure 1. Summary of trapping data at Camp Plauche. Breeding
activity (indicated by arrow) continued from April 29th through-
out the study period. Each rectangle indicates one individual.
Rainfall occurred on January 23-24, 27-30, and May 13-15.

in judgment. Either the movement is equal and the population truly
predominantely male or else the sexes are about equal in numbers and
the males are much more active. The case for greater activity on
the part of the males seems the better one.

It is apparent this movement is related to the amount of moisture
and the intensity of breeding drives. Temperature obviously also
controls the activity. Samples taken in cold weather show a much
greater percentage of empty stomachs than do those taken in warmer
weather, indicating that periods of cold (50°F, or less) may prevent
feeding over considerable intervals. In the New Orleans area field
observations show that inactivity due to cold is sporadic and that the
animals are generally active throughout the year.

Food Habits

Dickerson (1906) stated that Microhyla are generally eaters of ants.

Stomachs of 203 specimens were examined. These represent a
year-round sampling from all habitats, though cypress-gum, short-
grass and pineland habitats are best represented. Most food items
are relatively undamaged in the stomach and can be identified readily.
Food organisms were identified to order or family. A count was
made of the individuals of each form present (Table 1). When the
organisms were broken, heads were counted as whole individuals.
Food items were sorted into piles and the percent of the total volume
estimated for each form (Table 2). Stomachs are described as "full,"
"partly full," and "almost empty" (Tables 2, 3). Despite their
smaller number the 40 "full" stomachs are probably the most repre-
sentative indication of food habits. Organisms preyed upon by
Microhyla range from 0.4 to 6.3 mm in length, averaging about 3 mm.

22 Tulane Studies in Zoology Vol. 2

TABLE 1.
Number of Individuals of Food Organisms in Full Stomachs

Food Organisms Mean Maximum

Mollusca (snails) 1.5 2

Crustacea (Isopoda) — 1

Arachnida (spiders) 2.0 4

Arachnida (mites) 1.2 2

Collembola 13.3 21

Isoptera 30.6 78

Coleoptera

Carabidae 1.5 2

Curculionidae 1.6 4

Elateridae 1.2 2

Erotylidae — 2

Scydmaenidae 2.0 2

Staphylinidae 2.9 8

Unidentified 1.4 3

Total Coleoptera 3.5 9

Lepidoptera — 1

Hymenoptera

Formicidae 65.6 248

Unidentified 1.0 1

Unidentified insects 2JS 7

Ants, termites, and small beetles are the principal items. Some of
the beetles ( scydmaenids, clavigerids, and staphylinids ) are com-
monly associated with ants. Many of the organisms taken are secre-
tive, suggesting that the frogs may feed while under cover as well as
during the period of surface activity.

Food remains leave the stomach by late afternoon. When the
material enters the large intestine, it is formed into fecal pellets 3 to
5 mm long and 2 to 4 mm in diameter.

Predation

No instance of natural predation has been observed in the field.
Wright (1932) reported that Micro by la had been eaten by a garter
snake {Tbamnophis sirtalis). Anderson (1942) reported an instance
of predation by the copperhead. Blair's (1936) report of Microbyla
dwelling in tarantula burrows indicates that they may live unscathed
in close association with formidable predators.

The secretion of Microbyla carolinensis has been thought to give
it some protection from predators. If a bit of the mucous secretion
touches one's eye, a violent burning sensation may persist for more
than an hour. This secretion was also found to be irritating to the
mucous membranes of the mouth and throat.

The functions of this skin secretion in protection of the animal

from attack by ants was observed in the field on April 5, 1951. The

following is an excerpt from field notes of that date.

"Two individuals were found under a post which was
swarming with small brown ants. When a frog was
placed in the aroused colony numerous ants immediately

No. 2 Anderson: Ecology of Microhyla 23

attacked it. The frog sat very quietly and much ma-
terial was secreted by the skin glands. The ants rapidly
became entangled in this thick layer and their attention
was diverted into attempts to escape rather than con-
tinued attack upon the frog. After a short while the
frog ducked under a few lumps of dirt. When uncov-
ered it was found free of ants. The frog was replaced
among the ants and the procedure repeated. The next
time the frog sought cover it could be seen that the
mucous and the entangled ants were brushed off when
the animal passed any object. (Captain Joseph Jones,
U.S.A.F. identified the ants as Iridomyrmex pruino-
sus.) ."
Field notes also show that this secretion may be toxic to other
amphibians.

"Thirty plus Microhyla in small bottle with four Hyla
cinerea and one H. squirella. Upon reaching the labora-
tory H. cinerea and the H. squirella appeared dead.
When washed in several rinses of water one of the H.
cinerea recovered somewhat. It was placed in an aquar-
ium and by noon the next day was stronger and had
moved a little, but it weakened rapidly when handled.
The tongue was partly extruded through the side of the

TABLE 2.
Average Percent by Volume of Stomach Contents

Food Organisms

Full Stomachs

Partly Full Stomachs

(Percent)

(Percent)

Mollusca (snails)

15.5

54.0

Crustacea (Isopoda)

22.5

—

Chilopoda

22.5

—

Arachnida (spiders)

9.3

9.1

Arachnida (mites)

1.3

6.3

Collembola

15.0

14.0

Isoptera

67.8

73.9

Hemiptera

3.7

11.4

Dermaptera

8.0

9.0

Coleoptera

Carabidae

1.5

—

Clavigeridae

1.0

—

Curculionidae

15.5

19.8

Elateridae

10.0

18.2

Erotylidae

—

20.2

Scydmaenidae

3.5

5.0

Unidentified

7.5

17.0

Total Coleoptera

15.8

21.6

Lepidoptera

10.0

30.0

Hymenoptera

Formicidae

69.0

58.2

Unidentified

5.5

—

Unidentified insects

17.6

19.2

Plant fragments

9.7

18.1

Sand

97.6

80.8

Unidentified material

41.7

38.2

24

lulane Studies in Zoology

VoL 2

mouth and the mouth contained much slime and some
blood. The next day the animal was found dead. Sev-
eral of the Microhyla showed ill effects, but recovered
shortly."
Some predation under abnormal conditions was found. Microhyla
in the traps, handicapped by dehydration, were attacked and killed by
small ants. Numerous ants would usually first be trapped in the
frog's secretion and killed. On one occasion the headless body of a
frog was found in a trap. It had apparently been killed and partially
eaten by some small mammal.

Voice

In addition to the breeding call, which will be discussed in detail

later, two other sounds are produced. Both are apparently protest

notes. The first and most common one is a faint clicking produced

when the individual is handled. The second, a chirp or squawk

TABLE 3.
Frequencies of Occurrence of Food Organisms

Food
Organisms

All
Stomachs
(203)

Full
Stomachs
(40)

Partly
Full
Stomachs
(50)

Almost
Empty *
Stomachs
(65)

Mollusca (snails)

4

2

2

—

Crustacea (Isopoda)

3

2

—

1

Chilopoda

2

2

—

—

Arachnida (spiders)

16

8

4

2

Arachnida (mites)

27

4

13

10

Collembola

7

3

3

1

Isoptera

12

6

3

1

Hemiptera

14

7

5

2

Dermaptera

2

1

1

—

Coleoptera

Carabidae

2

2

—

—

Clavigeridae

1

1

—

—

Curculionidae

12

6

6

—

Elate ridae

8

5

2

1

Erotylidae

1

—

1

—

Scydmaenidae

5

4

1

—

Staphylinidae

18

12

5

1

Unidentified

35

16

11

8

Total Coleoptera

58

23

20

10

Lepidoptera

2

1

1

—

Hymenoptera

Formicidae

114

29

43

36

Unidentified

4

2

1

1

Unidentified insects

19

7

4

8

Shed skin

10

—

—

—

Plant fragments

36

9

13

10

Sand

26

5

9

8

Unidentified

25

6

13

6

* Other stomachs : Empty — 32, Not noted — 10.

No. 2 Anderson: Ecology of Microhyla 25

sometimes given when the frog is roughly handled, apparently also
replaces the clicking noise when the animal is in water. These sounds
are made by both sexes.

Reproduction
Gonadal Cycle
The reproductive behavior of most North American anurans has
been assumed to be strongly cyclic. This conception is based on the
sharply delineated breeding seasons of many frogs and toads inhabit-
ing the northeastern portions of the continent. Bragg (1950a) on
the basis of his field observations in Oklahoma, has strongly challenged
the application of this view to several species. The question of cyclic
versus non-cyclic patterns indicates a need for study of the gonadal
cycles of many forms. The study of Glass and Rugh (1944) of the
normal cycle of spermatogenesis of Rana pipiens in Vermont and
the study of Fischer and Richards (1950) on the ovarian cycle of
Acris crepitans in Oklahoma are the best sources of comparative
information.

Sections of 44 testes and 31 ovaries were studied. Some ovaries
and testes for cytological preparations were taken from fresh speci-
mens and fixed in Bouin's, while others were taken from specimens
which had been fixed in formaldehyde, washed in a deformalizing
solution and placed in 50 percent alcohol, according to standard
museum practice. There did not seem to be any great difference in
the results obtained by these two methods. The gonads were em-
bedded in paraffin and sectioned at eight microns. Mayer's haema-
toxylin or Galigher's haematoxylin was used, the latter usually prov-
ing more satisfactory.

Male Reproductive Cycle. — The appearance of the seminiferous
tubules of the anuran testis has certain dissimilarities from that of
the mammalian testis. The various stages of spermatogenesis are
not evenly dispersed around the wall of the tubule, but are found in
clusters which project out into the lumen and are distinct from other
clusters. Rugh (1951) has suggested that each of these larger
clusters may come from a single spermatogonium. There is no
zonation inward from the tubule wall, but a linear arrangement of
clusters along the wall. Sertoli nuclei are scattered close to the wall
and do not occur in clusters. Spermatogonia occur singly or in small
clusters.

The first step in interpretation of testes slides is identification
of the several stages of spermatogenesis. Rugh (1951) describes
the stages in some detail as found in Rana pipiens. His descriptions
of spermatozoa, spermatids and secondary spermatocytes are readily
applicable to the present species, but difficulty was encountered in
identifying primary spermatocytes on the basis of his illustrations and
descriptions. Rugh stated that nuclei of primary spermatocytes are
irregularly spherical, very large, and never numerous. These char-
acters readily fit nuclei which appear to belong to Sertoli cells, but

26 Tulane Studies in Zoology Vol. 2

leave unidentified another type which is a major component of the
seminiferous tubule wall. Because of this difficulty it is believed
that a description of the stages as interpreted in this study is necessary.
Spermatozoa are easily identified. The heads are small, darkly
stained, and bullet-like in shape. They may occur in clusters in the
Sertoli cells or scattered in the lumen of the tubules. All cells which
have elongated nuclei are classed as spermatozoa.

Spermatids are small in size and usually consist of dense, darkly
stained, circular nuclei, each surrounded by a ring of clear cytoplasm.
These cells occur in clusters which extend into the lumen from the
tubule wall and each cluster contains a large number of individual
cells (mean 41.9). They usually differ from the secondary sperm-
atocytes only in size and in the larger number of cells in a cluster.

Secondary spermatocytes occur in clusters containing a small num-
ber (mean 20.9) of cells larger than those found in spermatid clusters.
Each cell consists of a medium-sized, dense, darkly stained nucleus
with an outer ring of clear cytoplasm.

Primary spermatocytes occur in clusters which contain still fewer
individuals (mean 13.5) and do not project very far into the lumen.
The nuclei are large and granular, but are not dense or darkly staining.
Spermatogonia are much like the primary spermatocytes in appear-
ance, but are smaller and often occur singly rather than in clusters.
Since the change to primary spermatocytes is gradual, no exact
criterion for separation is possible.

Sertoli cells have very large, pale granular nuclei. These may be
irregular or regular and rounded in outline. They always occur singly,
close to the wall of the tubule, and are never found in clusters. There
are only a few (two to eight) visible in each tubule. They are often
associated with clumps of spermatozoa which are obviously in Sertoli
cells.

Breeding activity of Microbyla in Louisiana occurs from the latter
part of April until the middle of September. Testes of mature frogs
taken throughout the year show all the maturation stages of sperm-
atogenesis, but the relative numbers and the relative amount of space
allocated to each stage varies greatly (fig. 2).

Testes of mature individuals taken in October are characterized by
the presence of small numbers of spermatozoa scattered about in the
lumen of the tubules. A few spermatozoa also are in Sertoli cells, but
nowhere do they occur in dense masses. Primary spermatocytes are
the major component of the tubule wall.

Some November individuals show a condition similar to that found
in October. In others spermatozoa have begun to increase in number,
appearing in concentrations in Sertoli cells. There is a marked de-
crease in the number of primary spermatocytes. These trends con-
tinue and by January the massed spermatozoa in the Sertoli cells are
so numerous as to cause a reduction in the size of the lumen. The
primary spermatocytes have lost their place as the major component
of the tubule wall and are found only in small scattered clusters.

No. 2

Anderson: Ecology of Microhyla

27

/ y

1 / V /

VA

fflHBMKSIH frVn

OCT NOV DEC JAN FEB MAR APR MAY JUNE JULY AUG SEPT

Figure 2. Spermatogenesis in Microhyla c. carolinensis. Solid line
= spermatozoa ; dashed line = spermatids; long dash-short dash
line = primary spermatocytes; dotted line = secondary sperma-
tocytes; wide solid black line at base of figure = breeding season.

In February the massing of the spermatozoa in Sertoli cells reaches
its peak. Massed spermatozoa nearly fill the lumen of the tubule and
often extend outward to the tubule wall. Primary and secondary
spermatocytes and spermatids occur in small clusters. The accumu-
lation of the spermatozoa in the Sertoli cells seems to be a slow,
gradual process in the months after the close of the breeding season.
During this period large numbers of secondary spermatocytes and
spermatids do not occur.

The massed condition of the spermatozoa persists through the early
part of the breeding season. With the initiation of breeding, however,
there is a gradual thinning out of the spermatozoa in the Sertoli cells
and some appearance of massing in the lumen of the tubule.

For two months preceding the breeding season and during the first
two months of the latter period a peak in general activity is reached.
This is marked by the massing of the spermatozoa and by a great
increase in the numbers of secondary spermatocytes and spermatids
(fig. 2). This increase begins in March and the number of secondary
spermatocytes and spermatids remains high through May and June.

In June there is a slow reduction in the number of secondary
spermatocytes and spermatids, paralleled by a thinning out of the
spermatozoa. There appears to be a very slight increase in the num-
ber of primary spermatocytes at this time. These trends continue
through July and August.

In summary, the male gonadal cycle may be divided into two major
phases. The first is the slow build up of spermatozoa following the
cessation of breeding activity. This phase culminates in February
with the massed spermatozoa filling the tubules. The second phase

28 Tulane Studies in Zoology Vol. 2

is initiated in March with the appearance of numerous secondary
spermatocytes and spermatids. During the early part of the breeding
season there is a height of spermatogenic activity characterized by
the draining of mature spermatozoa from the supplies in the Sertoli
cells and a partial replacement of these from the developing secondary
spermatocytes and spermatids. With the end of the breeding season
the supply of the latter stages is no longer maintained and the sperm-
atozoa are thus drained to a very low level, preparatory to the start
of the first phase once again.

Secondary Sexual Characters and the Male Cycle. — Sexual dimorph-
ism in Microhyla occurs in size, coloration, presence of vocal sacs,
and presence of dermal modifications in the male. The size relation-
ship will be discussed at another point in this paper, but the modi-
fication of skin pigmentation and structure show some cyclic fluctu-
ations and will be considered here.

The black throat does not show marked cyclic changes and is char-
acteristic of more than 85 percent of the males 22 mm or more in
snout- vent length throughout the year. The throat of the male is
usually abruptly darker than the abdomen and pectoral region. Some
difficulty may be encountered in applying this character to strongly
pigmented individuals. There is a slight intensification of this dark
throat color during the breeding period.

Two secondary sex characteristics may be observed only in the
breeding season. One, the presence of chin tubercles, was first de-
scribed by Mittleman (1950). Tubercles may occur in a single or
double row along the ventral surface of the lower jaw and may range
from one to 28 in number. They are conical and project approxi-
mately 0.07 mm, as measured with an ocular micrometer scale. In
the June sample of 129 males displaying a black throat, 114 showed
tubercles. The second characteristic is the wrinkling or roughness
of the skin of the whole throat area of the male. Wrinkling may be
caused by the development of the vocal sac. There may also be small
rough patches and occasionally transverse bands of tiny tubercles.
There is great variation in these throat characteristics.

Female Reproductive Cycle. — The present study utilizes data de-
rived from the study of gonad sections and the measurement of gross
features of the female gonads. Examination of the gross changes in
the ovary proved especially informative. This organ consists of a
double-walled sack, normally divided into three to five lobes. The
oogonia and the developing oocytes lie between the inner and outer
walls of the sack. As the oocytes grow they bulge into the lumen of
the ovary and are closely covered by the very thin inner membrane
which forms most of the follicular wall. During their growth the
oocytes pass through three distinct stages. Analysis of the occurrence
of these stages furnishes a basis for determining the reproductive cycle.

The first of the oocyte stages is characterized by a clear, trans-
parent cytoplasm having a whitish, granular-appearing nucleus. These
cells have a diameter of from 0.08 to 0.45 mm, falling between 0.11

No. 2 Anderson: Ecology of Microhyla 29

and 0.34 mm in most cases. These are present in all ovaries through-
out the year. This clear stage is the only one present in all frogs
less than 19 mm in length and in some up to a length of 22 mm.

Oocytes belonging to the second and succeeding stage of develop-
ment become opaque and milky-white. They range in size from 0.23
to 0.77 mm and are found throughout the year in ovaries of most
frogs over 20 mm in snout-vent length.

Oocytes belonging to the third and final stage are pigmented.
These range in diameter from 0.69 to 1.49 mm. The smaller ones
are pigmented throughout and are a dark brown color. The larger
ones show a fully developed differentiation of the animal and vegetal
poles, the former being a very dark brown, the latter white. The
pigmented oocytes are present in individuals of more than 23 mm
in length from January until breeding takes place. They are absent
in all ovaries from late September until January.

In addition to developing oocytes, the walls and lumen of the ovary
may contain degenerating pigmented oocytes. Some of these may be
quite large and retain their shape while others are indicated only by
small amorphous concentrations of dark pigment in the ovarian wall.
These are believed to be oocytes which failed to ovulate or were
caught in the ovarian tissues after ovulation. Presence of these
oocytes suggests previous breeding, though they do not occur in all
females which have had previous breeding seasons.

The ovaries of 142 frogs were examined. The ovarian wall was
broken and the fragments studied under a dissecting microscope.
The presence of various stages of oocyte development was noted and
a sample of each class measured with an ocular micrometer. Since
a random sample of each of the various stages could not be taken,
ten or more oocytes of each class were measured representing the
extremes and the apparent general trend. As the ovaries enlarge with
the growth of the pigmented oocytes, many of the cells become
angular. It is necessary therefore to search for oocytes which are still
nearly spherical in making measurements.

The various oocyte types as outlined above also show certain cyto-
logical differences. The nuclei of the small oocytes, belonging to
the clear variety, are regular in outline and have a few dark-staining,
rounded nucleoli near the edge. The cytoplasm of these small oocytes
may sometimes show large, radially oriented vacuoles. These vacuoles
are reduced in size with the growth of the oocytes and eventually
disappear entirely. They may contain small, dark-staining bodies and
it seems possible that these have some relation to the yolk nuclei
reported in Bufo by King (1908) and in Acris by Fisher and Richards
(1950). The nuclei of the white oocytes show the greatest cyclic
change. There is increasing irregularity of the nuclear margin, an
increase in the number of nucleoli and a movement of the latter
toward the center of the nucleus as the breeding season approaches.
These trends are continued as the oocytes become pigmented and a
highly irregular germinal vesicle, or nucleus, results. This has a

30 Tulane Studies in Zoology Vol. 2

dense scattering of small nucleoli at its center. The pigment in these
maturing oocytes is laid down mostly around the outer rim of the
cytoplasm.

The oviduct of the female and the fat bodies in both sexes display
variation in correlation with that in the gonads. The oviduct is a
white tube, wavy in its anterior part and much convoluted in the
posterior glandular portion. It fluctuates greatly in size, being rela-
tively slim throughout the post-breeding period in the fall and winter
and enlarging greatly as the breeding season approaches. The fat
bodies exhibit a reversed cycle. This occurs to some extent in both
sexes but is clearly marked only in the female. The fat bodies are
enlarged during the fall and winter and are reduced in size and
finally almost disappear with the maximum development of the
ovaries at the beginning of the breeding season. It is possible that
they store materials used in the developing oocytes.

The results of the oocyte measurements give a clear picture of the
female reproductive cycle. Females which have recently deposited
their eggs have, with the exception of a few pigmented eggs that
were not ovulated, only clear and white oocytes in the ovary. The
ovary appears as a white, angular structure, molded about the postero-
dorsal surface of the digestive tract and the posterior end of the body
cavity. In section the ovaries also show oocytes of two classes. The
larger ones have a palely staining cytoplasm which seems to indicate
some yolk accumulation. The margins of the nuclei are irregular.
The nucleoli are near the edge of the nucleus in some cells, or scattered
inward to the centers in others. The smaller oocytes have a less
variable, more darkly stained nucleus. The nuclear margins are
regular or slightly scalloped and the nucleoli are near the surface.
The smallest oocytes often have numerous large vacuoles arranged
radially around the outer edge of the cytoplasm. The most noticeable
features of these sections is the presence of collapsed follicular sacs.

There were no mature individuals available from the months of
September and October, but the November and December samples
furnish information which makes an evaluation of the period possible.
Apparently there is little growth in oocyte size during the months of
early fall and winter. The most notable feature is the disappearance
of the collapsed follicles. There is considerable variation in the size
of the oocytes. The larger ones have irregularly margined nuclei, and
the nucleoli are scattered throughout the nucleus. Some small oocytes
show vacuolated cytoplasm and others do not.

In January the first pigmented oocytes appear. From this time on
they increase rapidly in number. Growth of pigmented oocytes fol-
lows a smooth curve with the exception of the March sample which
was taken in 1951 (the other samples were taken in 1949-50). The
apparent explanation for this discrepancy is that the 1949-50 winter
and spring were exceptionally mild, while the 1950-51 winter was
cold. If this effect is actually due to climatic conditions, it indicates
considerable plasticity in the female reproductive cycle. This tardy

No. 2 Anderson: Ecology of Microbyla 31

development was apparently compensated for rapidly, for the breed-
ing season began two weeks earlier in 1951 than it did in 1950.

During the period of development of pigmented oocytes there
appears to be a decrease in the size of the white oocytes. This may
be explained on the assumption that the larger white oocytes are
becoming pigmented, and fall in another class, leaving only smaller
cells behind and thus reducing the average size.

As a result of the great growth of the pigmented follicles the ovaries
increase rapidly in size, almost surrounding the digestive tract. The
white and clear oocytes are crowded into the corners between the
maturing pigmented oocytes. All classes of oocytes are now angular
in shape due to pressure from others nearby. With this ovarian
enlargement there occurs the great enlargement of the oviducts,
especially the glandular portions.

Discussion. — The gonadal cycle of male Microbyla contrasts sharply
with that found for Rana pipiens in Vermont by Glass and Rugh
(1944). The differences are understandable when the breeding sea-
sons and activity periods of the two animals are considered. Rana
pipiens is, in Vermont, a form with a restricted period of activity
and a long hibernation; physiological processes are at their height
during the warm months of the year and are slow during the long
periods of hibernation. In the spring there is a short, definite, breed-
ing period, preceding the seasonal optimum of environmental condi-
tions, which gives the tadpoles a favorable period for growth and
transformation and gives the adult a chance to prepare for the coming
year. In Microbyla in southern Louisiana there is no period of
hibernation, and there is no need for a rapid build up of reserves or
an early preparation for the coming breeding season of the next year.
There is also no problem of time for larval development.

Another interesting problem arising in relation to reproductive
cycles is the matter of the timing of individual activity during an
extended breeding season. Why one female deposits her eggs in
April and another does so in September is not known. There seemed
to be little variation in the degree of preparation for breeding among
the females of the June sample. If the time of individual breeding
is fixed by inheritance, interesting evolutionary implications arise,
such as reproductive isolation resulting in fragmentation of sympatric
populations.

The female reproductive cycle described here agrees with that given
by Fisher and Richards (1950) for Acris in some respects, and differs
considerably in others. Fisher and Richards do not emphasize the
external morphology of the oocytes and thus comparison is difficult.
They report a range of stages found in each ovary and this is inter-
preted as an indication of differential rates of development. In
Microbyla it seems apparent that this variation cannot be attributed
to differential rates but rather indicates that the ovary contains, at
any given time, oocytes destined for use in at least two and possibly
three breeding seasons. This situation apparently differs sharply

32 Tulane Studies in Zoology Vol. 2

from the ovarian cycle of Rana pipiens, for Rugh (1951) reports
that after ovulation "the ovary has been freed of mature eggs and
contains only oogonia with no pigment and little if any yolk." Al-
though oogonia may be found in most sections of Microhyla ovaries,
they are never common, while white and clear oocytes are always
abundant. The pattern of nuclear change with oocyte growth seems
to be identical with that reported by Fisher and Richards (1950) and
Rugh (1951). However, Fisher and Richards (1950) seem to imply
that such a phenomenon as the movement of the nucleoli toward the
center of the nucleus occurs in all oocytes in a given ovary at once.
This is not true of Microhyla for the nuclear condition is correlated
with the gross morphology and small, clear-type oocytes always have
few nucleoli and these are always near the nuclear border.

Reproductive Behavior

Field observations of the reproductive behavior of Microhyla are
difficult because of the shyness of calling males and the dense vege-
tation in most breeding sites. However, techniques for field study
of certain aspects have been worked out, and with practice it becomes
easier to discover the males in their calling positions. Two sets of
field observations, one at Century, Florida on July 15, 1950 and the
other at New Orleans on July 28, 1950 provided much valuable
material.

In addition to field observations, specimens injected with whole
pituitary gland after the method standardized by Rugh (1935) were
studied in the laboratory. Glands used were from ranid frogs of
three species. Average doses in terms of glands from mature indi-
viduals of these donor species are as follows: Rana pipiens, two; Rana
clamitans, one-half; Rana catesbeiana, one-third. Injections were made
with a 1.5 ml syringe and #22 needle. Care was taken not to inject
more than 0.05 ml of water into any individual. Individuals were
placed in a large container with shallow water and some material
upon which they could climb about. When amplexus began they
were transferred to small aquaria for observation.

Information on other questions of reproductive behavior was ob-
tained by examination of ovaries and oviducts.

It has been demonstrated above that a definite gonadal cycle
exists. This cycle is considered as the primary factor in control of
reproductive behavior. It may be assumed to be under hormonal
control but has shown variations apparently attributable to climatic
conditions.

Environmental Factors. — Weather conditions have been shown to
influence the degree of general activity as well as breeding activity.
Most authors are agreed that rain stimulates breeding behavior. Thus
Wright (1932) states: "This subterranean species normally breeds
and mates at night. Heavy, warm rains of one to four inches may
precipitate mating in daylight and rarely one captures a croaking
male at mid-day." Barbour ( 1941 ) reported a large chorus develop-
ing after a cloudburst. Bragg (1950a) on the other hand, stated that

No. 2 Anderson: Ecology of Microbyla 33

breeding of M. c. carolinensis in Oklahoma is not much influenced
by rainfall. In this respect it should be noted that most authors, in
describing time of breeding of this and other frogs, do not distinguish
between occurrence of calling individuals and actual mating. There
is no evidence that the two are necessarily synchronous. The notes
of the writer show that choruses, like movement, are often initiated
by rain. The first breeding of the 1950 season did not take place
until after two days of rain. Until that time there had been no call-
ing or other breeding activity. On the second night just after the
rain had ceased large choruses were found at several localities and
pairs were found in amplexus. Shorter rains as little as four days
previously had had no effect and it seems obvious that this prolonged
and very heavy rainfall was needed to initiate breeding.

Light, with its apparent inhibiting effect upon movement, also
influences the breeding behavior. Daylight choruses are common at
the height of the breeding season even in very dry weather, but no
evidence of diurnal amplexus or egg deposition has been noticed.

Great adaptability is shown in regard to breeding sites. Males
have been found calling, and tadpoles seen, in wheel ruts, ditches,
small permanent ponds and margins of cypress lakes. These can be
classified as being deep water or shallow water sites. Deep water
sites are those with water one to six feet deep. Breeding activity
takes place in such areas only when a dense mat of floating vege-
tation such as water hyacinth (Eichhornia crassipes) or alligator
weed (Alternanthera philoxeroides) is present. Shallow water sites
are less than one foot in depth and some have a maximum depth of
only one or two inches. These sites may be temporary or permanent
waters. Wright (1932) recorded a great mortality of eggs due to
drying of temporary sites in the Okefinokee swamp.

Vegetation in shallow water may be dense or non-existent, but call-
ing males have been seen using sticks, lumps of earth, crevices in
the earth or in logs, loose bark, sparse short grasses and large clumps
of tall grasses, alligator weed and water hyacinth as cover. This
reaction is apparently based upon two factors: the shyness of the
animal which keeps it partially under cover, and the need of some
sort of support for maintenance of proper calling position.

The bottom of breeding ponds ranges from mud to decomposing
organic matter to hard clay.

Observations made near New Orleans on July 28, 1950 suggest that
there may be some sort of location sense in regard to traditional breed-
ing sites. A location in Jefferson Parish was used heavily all through
the spring. It consisted of a weedy pond, about 2.5 feet in maximum
depth and 60 feet in diameter, with a wide belt of alligator weed
around the shore. Acris called from the short and sparse weed nearer
the center of the pond and Microbyla from the dense, tall weed along
the edge. At some time between June 1st and July 20th an abandoned
field stretching behind this pond was cleared with a bulldozer and
the pond filled in. On July 28 after a rainy day Microbyla were

34

Tulane Studies in Zoology

Vol. 2

found calling from puddles and crannies between lumps of earth.
There was no living vegetation but cover was furnished in some pools
by root stubs or bits of floating board. There was no apparent dif-
ference between any two parts of the field of approximately five
acres. Despite this apparent complete similarity of possible breeding
places over the whole area, the large chorus was found only in the
immediate area of the former pond.

As there is a daily cycle of normal activity, there is also a cycle
of breeding activity. It has been stated that during periods of intense
breeding activity many individuals may remain in the breeding area
and call throughout the day. In these and in all other cases there is
a movement of individuals to the breeding sites at dusk. Observa-
tions at a small artificial pool in Audubon Park, New Orleans, show
that the peak of this movement is just at full darkness. The males
precede the females in moving to the breeding sites, the peak of
male movement preceding that of the females by ten to twenty
minutes.

Voice.— The call is usually described as a sheep-like bleat. Burt
(1932) noted that in Louisiana and Mississippi he heard two calls,
one "goat-like" and the other a "soft murmur." The writer has never
been able to hear anything resembling this latter call either in field
or laboratory. Some variations in the normal call take place when
amplexus begins. The normal call has a duration of about 1.4 seconds
at 25 °C (water temperature).

The calling posture is constant as reported by various authors
(Wright 1932, Barbour 1941). The body is usually in a vertical or
near vertical position, immersed in water up to the axilla and sup-
ported against some object by the forelimbs.

While calling, the male does not move about, but retains his position
without change for long periods of time. Of six individuals marked
by placing tagged sticks near them at 8:45 P.M. (Century, Florida
July 16, 1950), all retained positions at 9:00 P.M., four retained
position at 9:30 P.M. and at 10:00 P.M. Since eggs were found
in the regions occupied by the two that left their positions, it is likely
that they had mated successfully. Again, near New Orleans on July

e-

e-
-e-

e-

■e-

-e-

■e-

-e-

-e-

*e-

c

D

E

-0-

■e-

■e-

-e-

-e

-e-

^e-

■e

-e-

f Le

e-

Figure 3. Calling patterns, July 28, 1950. Each letter and corre-
sponding line represents a single individual. Vertical lines indi-
cate pauses between series of calls.

No. 2

Anderson: Ecology of Microhyla

35

TABLE 4.

Calling Patterns in a Small Chorus

38 series of calls (total studied)

34 —

2 or more in

dividuals

20 —

3 or more in

dividuals

14 —

4 or more individuals

5 —

5 or more individuals

3 —

6 individuals

Calling frequencies

Individuals

A B C D

E

F

Total Calls

29 25 16 23

11

10

Started series 9 3 0

3

6

Called alone

0 0 0 0

3

1

Frequency of calling sequences

Combinations of

Combinations of

Combinations of

two individuals

three individual's

fou

r individuals

(30 possible comb.)

(240 possible comb.)

(420 possible comb.)

AB — 20

ABC — 9

ABCD — 3

AD— 4

ABD — 2

ABDC — 1

BC — 15

ADB — 2

ABDF — 1

BD— 2

ADE — 1

BDFE — 1

CD— 4

BCD — 4

BCDE — 1

DA— 9

BDC — 1

CDEG — 1

DB— 3

BDF — 1

DABC — 4

DC— 1

DAB — 6

DABD — 1

DE— 2

DFE — 1

EABC — 1

DF— 2

CDE — 1

EDAD — 1

ED— 1

EAB — 1

FDAB — 1

FA— 6

EDA — 1

FEDA — 1

FD— 1

FAB — 2

FADE — 1

FE— 3

FAD — 1
FDA — 1
FED — 1

FABC — 2

28th six individuals under observation maintained their positions
with very little change for two and one-half hours. One remained
in position and called for three hours before beginning to move about.
The calling of a group around a small pool was studied in both
the Florida and New Orleans areas. A predominant pattern was
perceived in the Florida chorus. A technique applied at a New
Orleans site permitted the observer to record and analyze this pattern.
A number of parallel lines were drawn, each representing an indi-
vidual and designated by a letter. Each time an individual called a
dot was placed on the line representing that individual. Since the
calls of the individuals in a small aggregation are in bursts with
intervals of silence between, each series is written as one would
write a bar of music. Figure 3 gives an example of calling records
obtained in this way and Table 4 presents the analysis of the data
obtained for 38 series of calls around one pool. There were six males
calling around this pool which was about three feet in diameter.

36 Tulane Studies in Zoology Vol. 2

They could all be observed easily during this period.

When the data are summarized relationships between calling indi-
viduals are shown. Of the 38 series of calls, 34 have two or more
individuals represented (Table 4). With 6 individuals calling there
were 30 possible combinations of two. Of these possible sequences

15 occurred a total of 75 times. On the basis of equal chance and
calling activity for each individual it would be expected that each
of 30 combinations would have occurred slightly less than three
times as an average. Actually one sequence (AB) occurred 20 times,
another (BC) occurred 15 times, and a third (DA) occurred nine
times. If those series which had three individuals calling are con-
sidered there are 120 possibilities, but only 16 occurred a total of
35 times. The chances that any sequence should occur more than
once is thus small, yet the pattern ABC occurred nine times, the
pattern DAB occurred six times, and the pattern BCD occurred four
times. Considering series of four individuals there are 240 possible
combinations; 14 series occurred a total of 20 times, and DABC
occurred four times, ABCD three times, and FABC twice. There
is thus a definite pattern, revolving around four individuals: A, B, C,
and D. Examining the status of the remaining two individuals, E
and F, it is seen in Table 4 that they are the only ones which called
alone, that is, without being part of a series. It is further noted that
they called less than the other individuals, but usually either called
alone or began the series. Turning back to the original four it is
seen that A and B called the most, but that B was not aggressive,
rarely initiating a series of calls, while A and D were both aggressive
in this sense. In contrast, C never started a series. Carrying these
relationships further it is found that B followed A in 20 of the
25 series in which it took part, and that C followed B in 15 out of

16 series in which it took part.

The significance of these relations is not understood. A, apparently
the dominant member of the group, remained in position and kept
calling the longest, but did not succeed in attracting a female. Two
other members of the group (exactly which ones could not be deter-
mined at the time) did go into amplexus some time after calling
observations were made.

There seems to be no restriction upon calling densities. Calling
individuals may be as close as one inch apart (three or four indi-
viduals at times) or as far as fifteen or twenty feet.

In the Jefferson Parish area studies on calling pattern there were
numerous small complexes similar to the one studied. An attempt
was made to apply the same method to groups as was applied to
individuals and thus determine if there were relations between them
also. The results did not indicate any significant patterns.

Sex Recognition. — The phenomena of sex recognition are obscure.
On several occasions it has been possible to detect a faint vibration
when males were handled, similar to that found in Bufo, but a func-

No. 2 Anderson: Ecology of Microbyla 37

tion in sex recognition is not indicated by observations of mating
behavior.

There are very few errors in recognition and only in one instance
(in the laboratory) has a male been found clasping or attempting to
clasp another male. Both individuals in this one case had received
heavy doses of pituitary gland. The males have never been observed
to clasp inanimate objects as male Bufo do.

Initiation of amplexus has been observed in water in laboratory
aquaria. The female floats in a position characteristic of the species,
with hind legs widely spread. The male, appearing to move about
at random, will, when behind her, swim into clasping position and
work up her back until the normal amplexus can be achieved. This
procedure is not believed to illustrate the normal courtship behavior.

The following excerpt is from field notes taken at the Jefferson
Parish area, July 28th, 1950. These notes were taken at the small
pool described above where the studies on calling pattern had just
been made. The observer sat about two feet from the pool. To his
left was individual A in calling position behind a floating board.
Directly across the pool (about 20 inches wide at this point) was
individual B. Directly in front of the observer on the near bank
were three individuals called in the notes the "trio". They were
about one inch apart in a triangular formation. They were partly
hidden by some sticks and other debris and behind them were at
least one, possibly two other frogs which were hidden completely.
The notes read:

"During 9:00 to 9:30 interval the moon becomes
brighter as the sky clears. The inter-group intervals
seem to get longer and the intra-group calling density
may be less. Individual seen moving across middle of
pool, perhaps middle one of trio which has gone — it dives
to avoid light and a small individual simultaneously
appears on land near third of trio — it moves past in
front of and almost touching the latter, which calls with
increased frequency. An individual appears again in
the middle of the pool briefly and the new individual
near the trio vanishes.

"Three grunts seem to come from near B's position
and an individual seen scrambling up the bank from that
area — B does not react. Another individual seen about
one inch from B. Two more grunts heard, but position
not definite, one has a croaklike quality — location estab-
lished for these sounds — a pair in amplexus on board
near A.

"Nine minutes later a large individual comes down
bank near the trio. The latter (two and possibly three
are still present) begin to call rapidly. "A" speeds up
calling in response. All three of trio members in origi-
nal position, the newcomer moves between them and dis-
appears. Three minutes later a pair in amplexus ap-
pears in that area."

These notes illustrate the difficulty of observation. They also illus-
trate several points. One is that calling males respond to the approach
of another individual, presumably a female, by increasing the calling

38 Tulane Studies in Zoology Vol. 2

rate, but do not move from position to attempt amplexus, although
the individual may pass very close. There appears to be two possi-
bilities : ( 1 ) the wandering, non-calling individuals are females, and
some action on the part of the female other than appearance or near-
ness and movement must serve as a release mechanism for amplexus;
(2) these individuals are actually males and their sex is recognized
by other males without amplexus or other contact. Observations
indicate that most of the wandering individuals in a chorus are females.

The grunts and croaks noted at the approximate time of occurrence
of amplexus have also been observed in the laboratory when terraria
were set up in such a way that the males could call normally and the
females move to them.

Noble and Aronson ( 1942 ) found that female Rana pipiens were
not receptive to amplexus until ovulation had taken place. In Micro-
byla ovulation may sometimes begin either just before or after
amplexus has started. Dissection of a female taken in amplexus and
preserved forty-five minutes later revealed that ovulation had just
begun. Approximately twenty percent of the eggs had reached the
oviduct. Examination of thirteen other females taken in amplexus
showed ovulation still in process in six individuals and complete in
seven. Two unmated females found moving about in the same
chorus had nearly completed ovulation. Observations in both field
and laboratory indicate that egg deposition took place in from one
and one-half to two hours after amplexus began. Thus ovulation is
rapid and the females become receptive to males at or before its
commencement. This pattern of rapid ovulation and early receptive-
ness might be of value to the species in arid climates, permitting the
animals to take advantage of sudden rains for breeding purposes.

Amplexus is semi-pectoral or axillary depending on size relation-
ships. It may be inguinal in cases where the female is not receptive,
or in cases where a male grasps a male. In amplexus, as observed in
pituitary injected individuals in the laboratory, the male rests on
top of the female, clasping her strongly in a semi-pectoral embrace.
The arms of the male pass just behind those of the female. The
thumbs dig into the pectoral area just posterior and ventral to the
insertion of the female's forearms. The palms are turned out and
forward. There is considerable size disparity and the snout of the
male just reaches the line between the female's eyes. The back of the
male is just above the water surface, the thighs are flexed, with the
shanks pointing dorso-laterally and the feet laterally and anteriorly.
The body of the female is at an angle of about forty-five degrees to
the surface of the water. Her thighs are partially extended, the shanks
point postero-laterally, the feet laterally and slightly anterior.

Egg Deposition. — At the beginning of amplexus the male and female
cooperate in swimming movements. As the time for egg deposition
nears, the male ceases swimming actions and cocks his hind legs
tightly in against his body.

Egg deposition can be anticipated by noting certain features of the

No. 2 Anderson: Ecology of Microbyla 39

behavior of the clasping pair. About twenty minutes before the time
of deposition the female becomes restless and moves about. The
interval between initiation of this behavior and egg deposition was
the same in the field on July 28 as in the laboratory. In the field the
female traveled about one foot overland to a deeper pool and back
again carrying the male on her back. At about this time there is a
change in the swimming movements of the female. She now swims
with very short, jerky, simultaneous movements of the hind limbs.

At the start of egg deposition the female may raise the hind limbs
and rear part of the body to the surface of the water and bring the
extended hind limbs forward until they are at right angles to the
midline on the same plane as the water surface. This produces a
swaybacked position. The feet are then moved in short forward
strokes, moving the animal backward one to two inches. The strokes
then shorten further and the legs move ventrally. (This preparatory
stage was omitted in those individuals seen in the field, but was
exhibited by two of three pairs in the laboratory.) The female then
extends the legs tensely in a ventro-lateral position with the toes bent
downward, bringing the cloaca upward above the water surface. As
she does this the male slides forward so that his cloacal opening is
about one-eight inch anterior to that of the female. He brings his
feet together so that they curve around her buttocks on either side
of the cloaca with the ankle joint at about the cloaca level. In this
position the eggs appear to boil from the cloaca of the female. About
thirty eggs are laid in each such ejaculation.

No information has been obtained on the duration of the clasp
after egg deposition is complete. In the laboratory the clasping
position apparently may be retained indefinitely if the pair are kept
in the water. If they have an opportunity to leave the water, it may
be broken relatively soon and exit from the water may possibly stimu-
late the male to release his grip.

Discussion. — Bragg (1950b) has asserted that two major breeding
patterns are discernible in North American anurans. These he desig-
nated as mesic and xeric. The mesic pattern, as he has defined it, is
basically cyclic, little influenced by rainfall, and with physiological
control vested in hormonal and other internal factors. In comparison,
he defined the xeric pattern as non-cyclic, with the breeding inde-
pendent of internal control to a great extent and initiated by rainfall.
He recognized also, a third, or intermediate pattern. Microbyla c.
carolinensis was considered by him to be a mesic form and Microbyla
c. olivacea a xeric form. His basis for this separation was the belief
that the breeding of M. c. carolinensis was more restricted as to sea-
son and was little influenced by rainfall. Unfortunately he had not
had full opportunity to observe this form and make comparison. The
cyclic gonadal pattern of carolinensis is certainly influenced to some
degree by rainfall, particularly as to initiation of the breeding season.
There is some breeding activity independent of rainfall but this
activity is much reduced. The gonadal cycle appears to be adaptable

40 Tulane Studies in Zoology Vol. 2

to either a mesic or xeric behavior pattern and an existing pattern
might be merely an effect of the local climate.

One factor in the breeding behavior which Bragg postulated for
forms showing xeric behavior patterns is supported by material
gathered here. This is the possibility that ovulation may not actually
begin until after amplexus has started and the fact that it is quite rapid.

The social aspects of the reproductive behavior are so little studied
that there can be little evaluation of them at this time. The signifi-
cance to be given to the difference and similarities between the actions
of egg deposition in Microhyla and in Rana pipiens as described by
Noble and Aronson (1942) is also uncertain.

Reproductive Potential

The sample taken from the ditch at Angie in June was used for
study of the reproductive potential of the species. In utilizing these
data it is assumed that the sample is representative of the breeding
females of the whole population as regards to both composition and
the condition of unspent females. This means that it is also assumed
that while males exhibit more movement than females, mature
females of all size groups, in the equivalent physiological condition,
move equally. (Although the breeding season had been underway
for some time, no spent females were found in the sample.)

A total ovarian count was made upon each of 66 females with
mature oocytes. The compliment of eggs per female ranged from 152
to 1,089. The mean number of eggs was 510 and the total was
33,657. An analyses by year classes shows that the three-year-olds
bear the burden of the reproduction (Table 5).

The fact that so few two-year old individuals are recorded is diffi-
cult to explain. In view of the indications noted below that some of
these may not breed until their third year, it seems possible also that
many may not breed until late in the season. Periodic samples at
the breeding ponds should prove useful in answering this question.

Growth and Development

The development of the embryo and the tadpole is not reported in
this paper, although they have received some study. Deckert (1914)
and Ryder (1891) have published short notes. Wright (1932) re-
ported considerable material, and Orton (1946) discussed the growth
and color pattern of the tadpole.

Both embryonic development and growth of the tadpole through
metamorphosis are rapid, as would be expected in a species that often
breeds in temporary rain pools.

Snout-vent measurements were made on 633 frogs. Preserved,
freshly killed, and live individuals were used. The percentage of
error in measurement of each type was calculated. This percentage
was small in all methods (preserved, 1.2; freshly-killed 3.8; alive, 2.1).
Measurements of a test series of individuals measured in all three con-
ditions were used to determine if corrective factors were necessary
for fresh and live material. The corrective factor of 0.989 indicated

No. 2

Anderson: Ecology of Microhyla

41

TABLE 5.
Reproductive Potential: Analysis by

Year Classes

Age

(yrs.)

No. of Mean No.
Size (mm) Specimens of Eggs

Total
Eggs

Percent of

Total
Production

2
3

4

23.0 — 24.4 6 377.8
24.5 — 26.9 44 466.3
27.0— — 16 680.8

2,267
20,516
10,894

6.7
60.9
32.4

for fresh material as contrasted with preserved material is small enough
so that in this study it may be ignored. It was found necessary to
apply a corrective factor of 0.914 to measurements of live individuals.

The samples were found to be affected by the collecting situations
studied. One was forced movement, caused by rising flood waters
in the Bonnet Carre Spillway. The second was a combination of
general feeding and breeding movement bringing the animals into
traps at the Camp Plauche area and the third was apparently primarily
breeding movement which caused the frogs to become trapped in a
pipeline ditch near Angie, Louisiana.

Snout- vent frequencies were plotted for each month (fig. 4). A
class interval of 0.7 mm seemed best suited for both large and small
samples.

Study of these frequency distributions show that no sample is fully
representative of the population. Year classes can be delineated
only approximately and an accurate growth rate cannot be calculated.
Only a rough outline of growth can be given.

Individuals transform at about 10.0 mm from May until October.
Since some may thus have four months start on others of the same
year class in their growth, the difficulties of delimiting the year classes
are not surprising. The effect of this seems to be diminished some-
what in older individuals. A tentative separation into year groups
for the months of May and June shows a progressively smaller range
per group as size increases. It appears that most individuals are
13.0 to 19.0 mm in their first spring following transformation, 21.0
to 24.0 mm the second spring, 24.5 to 26.9 mm the third spring, and
from 27 to 35 mm in succeeding years. This arrangement is sub-
stantiated by attainment of sexual maturity in the 21 to 24 mm size
class. These estimates are found to coincide roughly with Wright's
(1932) estimates of the early growth of the species in the Okefinokee.
He found the transformation size to range from 8.5 to 12.0 mm with
a mean of 10.8 mm. He then gave the year classes for the following
spring as 15.0 to 20.0 mm for the first year group. However, he
thinks they are 20 to 25 or 27 mm in the second spring and 27 to
36 mm in the third. The break at 26 to 27 mm in the curves given
here is slight, but it occurs often enough to be significant. When
the June samples are plotted with an 0.5 mm interval this break is
greatly emphasized. Wright's classification would indicate that growth
of maximum sized individuals is 8 mm from transformation until the
first spring, 7 mm in the second year, and 9 mm in the third. This

42

Tulane Studies in Zoology

Vol. 2

maintenance of an extremely rapid growth seems improbable in the
light of the general understanding of the growth of reptiles and
amphibians.

The size groups are demonstrated best if we combine the samples
from October through February. Four classes indicate a maximum
life span of five or more years (fig. 4).

12.6- 14.2

133 I4»

SNOUT-VENT LENGTH IN MILLIMETERS

Figure 4. Size frequencies in population samples, sexes combined.
Open rectangles = males ; solid rectangles = females.

No. 2 Anderson: Ecology of Microhyla 43

Individuals attaining sexual maturity by their second spring range
in size from 21 to 24 mm. Several criteria are used to indicate the
attainment of maturity. The criteria used for males are ( 1 ) the
appearance of secondary sex characters, (2) the presence of mature
spermatozoa, as indicated by smears of testes of both fresh and pre-
served individuals, and (3) the development of the gonads as indi-
cated by the study of histological sections. The secondary sex char-
acters of the female are largely negative and the examination of the
ovary is the only criterion.

A distinction must be made between the determination of the
smallest sexually mature size class at the breeding season and the actual
time at which the individuals take up the adult gonadal cycle. A
discussion of the relative importance of the two dates is academic,
for both are important.

Analysis of the June sample shows that the secondary sex char-
acters first develop in males of 21.0 to 22.9 mm snout-vent length.
The blackening of the throat seems to occur first and the tubercles
to appear at a slightly larger size (from 21.5 to 23.9 mm). Mature
spermatozoa were found in smears of testes of individuals 21.0 mm
or larger, but were sometimes absent in individuals of 23.0 to 23.4 mm.

Smears of testes taken throughout the year show that mature and
active spermatozoa often are present before there is any external indi-
cation of sex. Spermatozoa have been found in smears of testes of
individuals as small as 18.4 mm. In histological sections a very few
spermatozoa were found in the testes of an individual measuring only
15.5 mm in snout-vent length. This individual was taken in October,
showing that spermatozoa may be present during the year in which
the individual transformed, a full year and a half before the first
breeding activity is expected. Further study of sections of testes of
immature individuals allows an estimation of the time at which the
condition of the gonad approximates and begins to follow the seasonal
cycle of adult individuals. The immature testis is characterized by
very small tubules, each containing only a few cells. The arrange-
ment of the cells in the tubules is very loose. They are not in clumps
as found in mature testes. These characteristics are retained through
the first winter and then are gradually lost so that by the middle or
end of the second summer the mature condition is approximated and
the annual cycle begins.

Study of June females demonstrates that pigmented eggs first occur
at 23.0 mm but are lacking in some individuals up to 24.5 mm. In
all monthly samples, comparison of the minimum size in which white
oocytes are found and the maximum size in which only clear ones
are found shows a considerable overlap. These facts indicate that
some females may not deposit eggs until their third year.

There are insufficient specimens in the critical early fall period to
allow an estimation of the time at which the young females develop
the adult cyclic pattern.

Sexual dimorphism in size has been remarked upon by Wright

44 Tulane Studies in Zoology Vol. 2

TABLE 6.

Sexual Dimorphism in

Size

Males

Females

Percent of

Percent of

Size Groups (mm)

No.

Total males

No.

Total females

22.2 — 23.7

62

16.6

27

10.4

23.8 — 25.3

75

20.1

44

16.9

25.4 — 26.9

85

22.8

64

24.6

27.0 — 28.5

38

10.2

32

12.3

28.6 — 30.1

19

5.1

19

7.3

30.2 — 31.7

15

4.0

4

1.5

31.8— —

1

.3

6

2.3

(1932). This is manifest in the larger maximum size of females and
by the relative percentages of male and female individuals in the
larger size groups. The maximum size for males is 32.2 mm and
the maximum for females 34.3 mm. A relatively larger percentage
of females in the samples fall in the larger size groups (Table 6).
This dimorphism is slight.

Shedding was observed in six females used in breeding experiments.
Shed skins are found occasionally in the stomachs and may be identi-
fied by the outlines of toes or fingers (Table 3).

Populations

Some information about populations can be derived from the larger
samples. The trapping data from Camp Plauche also gives a little
information about the density and composition of a population.

In the discussion of movement, reference was made to the sex ratios
found in various population samples. The relatively large percentage
of males was considered to be due, in part at least, to greater move-
ment by that sex. In support of this it was pointed out that samples
collected by hand showed a less unbalanced relation.

The sex ratios are more nearly even in the smaller size classes. This
could be attributed to an actual difference in the ratios in juvenile
and adult populations or explained on the assumption that the rela-
tively greater movement of the males does not express itself fully in
immature individuals. It is impossible to tell whether either one, or
both, of these factors contribute to the observed result.

The continuous catch of unmarked individuals at Camp Plauche
indicates a large population, but since most recoveries were not made
in the traps it is not possible to attempt a population estimate.

Interesting variations in size composition are seen in the different
samples (fig. 4). A large percentage of immature individuals oc-
curred only in the October-November and March-April samples. Both
of these were hand collected samples. The proportion of immature
individuals is considerably higher in the October-November sample
than in any other. This is to be expected at the end of the breeding
season but an additional significance is also suggested. This sample
was collected in the cypress swamps of the Sarpy Wildlife Refuge

No. 2 Anderson: Ecology of Microbyla 45

and corroborates field observations which indicate a vastly different
size composition in these swamp populations as compared with those
from the abandoned fields and levees. Further comparative samples
of both breeding and non-breeding populations must be taken before
this situation can be interpreted. Some observations in the field sug-
gest an ecological segregation of different size groups in the swamps.
The most abnormal of the samples from the point of view of com-
position is the June sample. This is found to be almost entirely
composed of mature individuals and indicates a great difference in
the movement of adult and juvenile individuals during the breeding
season, rather than a true picture of the population.

Conclusions

1. Microbyla carolinensis carolinensis is highly adaptable in regard
to habitat requirement. Moisture seems to be the most critical habitat
requirement.

2. There is considerable nocturnal movement by all individuals.
The males wander more than the females and juveniles.

3. A wide range of small organisms is taken as food. The major
food items are ants, beetles, and termites.

4. There is little evidence of predation.

5. The reproductive behavior is cyclic. The gonadal cycle is, how-
ever, very different from that of Rana pipiens in the northeast. This
difference seems to be correlated with seasonal activity as controlled
by temperature.

6. Very little modification of the gonadal cycle, if any, would be
required to permit breeding activity along the pattern which Bragg
(1950a) designated as non-cyclic. The determining factor is probably
environmental.

7. The breeding behavior is complex and is in need of further
study. The phenomena of sex recognition found in other North
American anurans were not observed in this subspecies.

8. The calling patterns of small groups indicate some sort of com-
plex social relationship between calling individuals.

9. Population composition varies with the habitat. Further studies
in the population dynamics of this subspecies are needed.

References Cited

Anderson, Paul 1942. Amphibians and reptiles of Jackson county,

Missouri. Bull. Chicago Acad. Sci., 6: 203-220.
Anderson, P. K., E. A. Liner and R. E. Etheridge 1952. Amphibian

and reptile populations of a Louisiana pineland area. Ecology,

33: 274-278.
Barbour, Roger W. 1941. Gastrophyrne carolinensis in Kentucky.

Copeia, 1941: 262.
Blair, W. Frank 1936. A note on the ecology of Microhyla olivacea.

Copeia, 1936: 115.
Bragg, Arthur N. 1943. Observations on the ecology and natural his-
tory of Anura, XV. The hylids and microhylids in Oklahoma.

Great Basin Nat., 4: 62-80.

46 Tulane Studies in Zoology Vol. 2

1950a. Observations on Microhyla. Wassmann

Jour. Biol, 8: 113-118.

1950b. Observations on the ecology and natural

history of Anura, XVII. Adaptations and distribution in accord-
ance with habits in Oklahoma. In Researches on the Amphibia
of Oklahoma. Univ. Okla. Press, Norman.

Burt, Charles E. 1932. Records of amphibians and reptiles from
the eastern and central United States, 1931. Amer. Midi. Nat.,
13: 75-85.

Deckert, Richard F. 1914. Further notes on the Salientia of Jack-
sonville, Florida. Copeia, No. 9: [1-3].

Dickerson, Mary C. 1906. The Frog Book. Doubleday, Doran and
Co., New York.

Fisher, Helen T. and A. Richards 1950. The annual ovarian cycle
of Acris crepitans Baird. In Researches on the Amphibia of
Oklahoma. Univ. Okla. Press, Norman.

Freiburg, Richard E. 1951. An ecological study of the narrow-
mouthed toad (Microhyla) in northeastern Kansas. Trans. Kans.
Acad. Set, 54: 374-386.

Glass, Frieda M. and Roberts Rugh 1944. Seasonal' study of the
normal and pituitary stimulated frog (Rana pipiens) , I. Testis
and thumb pad. Jour. Morph., 74: 409-427.

Goin, Coleman J. 1949. The peep order in peepers; a swamp water
serenade. Quar. Jour. Fla. Acad. Sci., 11: 59-61.

Hecht, Max K. and B. L. Matalas 1946. A review of the middle
American toads of the genus Microhyla. Amer. Mus. Novitates,
No. 1315: 1-21.

King, Helen Dean 1908. The oogenesis of Bufo lentiginosus. Jour.
Morph., 19: 369-438.

Mittleman, M. B. 1950. Miscellaneous notes on some amphibians
and reptiles from the southeastern United States. Herpetologica,
6: 20-24.

Noble, G. K. and L. Aronson 1942. Sexual behavior of Anura, I.
The normal mating pattern of Rana pipiens. Bull. Amer. Mus.
Nat. Hist., 80: 127-142.

Orton, G. L. 1946. Larval development of the eastern narrow-
mouthed frog, Microhyla carolinensis (Holbrook), in Louisiana.
Ann. Carnegie Mus., 30: 241-246.

Parker, H. W. 1934. A Monograph of the Frogs of the Family
Microhylidae. British Mus. (Nat. Hist.), London.

Ryder, J. A. 1891. Notes on the development of Engystoma. Amer.
Nat., 25: 838-840.

Rugh, Roberts 1935. Pituitary-induced sexual reaction in the Anura.
Biol. Bull, 68: 74-81.

1951. The Frog. Its Reproduction and Develop-
ment. Blakiston and Co., Philadelphia.

Wright, Albert Hazen 1932. Life-Histories of the Frogs of Oke-
finokee Swamp, Georgia. Macmillan Co., New York, N. Y.

and Anna Allen Wright 1949. Handbook of

Frogs and Toads of the United States and Canada. Comstock
Publ. Co., Ithaca, N. Y.

u

Volume 2, Number 3

November 29, 1954

A NEW SPECIES OF DIAPTOMUS FROM LOUISIANA AND

TEXAS WITH NOTES ON THE SUBGENUS

LEPTODIAPTOMUS

(COPEPODA, CALANOIDA)

MILDRED STRATTON WILSON,

ARCTIC HEALTH RESEARCH CENTER, U. S. PUBLIC
HEALTH SERVICE, ANCHORAGE, ALASKA

?. zqol

DEC 1 7 1954

/*/■ |

TULANE UNIVERSITY
NEW ORLEANS

TULANE STUDIES IN ZOOLOGY is devoted primarily to the
zoology of the area bordering the Gulf of Mexico and the Caribbean
Sea. Each number is issued separately and deals with an individual
study. As volumes are completed, title pages and tables of contents
are distributed to institutions exchanging the entire series.

Manuscripts submitted for publication are evaluated by the editor
and by an editorial committee selected for each paper. Contributors
need not be members of the Tulane University faculty.

EDITORIAL COMMITTEE FOR THIS NUMBER

Arthur G. Humes, Associate Professor of Biology, Boston Uni-
versity, Boston, Massachusetts.

Harry C. Yeatman, Assistant Professor of Biology, University
of the South, Sewanee, Tennessee.

J. Bennet Olson, Senior Research Associate, Children's Hospi-
tal Society, Los Angeles, California.

Manuscripts should be submitted on good paper, as original type-
written copy, double-spaced, and carefully corrected.

Separate numbers may be purchased by individuals, but subscriptions
are not accepted. Authors may obtain copies for personal use at cost.

Address all communications concerning exchanges, manuscripts, edi-
torial matters, and orders for individual numbers to the editor. Re-
mittances should be made payable to Tulane University.

When citing this series authors are requested to use the following
abbreviations: Tulane Stud. Zool.

Price for this number: $0.25.

George Henry Perm, Editor,
c/o Department of Zoology,
Tulane University,
New Orleans, U. S. A.
Assistants to the Editor:

Carol L. Freret

Donald W. Tinkle

OEC \ 7 '1954

A NEW SPECIES OF D1APTOMUS FROM LOUISIANA AND
TEXAS WITH NOTES ON THE SUBGENUS

LEPTODIAPTOMUS

(COPEPODA, CALANOIDA)

MILDRED STRATTON WILSON,

Arctic Health Research Center, U. S. Public

Health Service, Anchorage, Alaska

The new species of Diaptomus for which the diagnosis is given
herein has been known to me for some time from slides of dissected
appendages in the S. F. Light accession in the United States National
Museum. These appendages were from copepods collected in south-
eastern Texas by Dr. Edward S. Deevey. Since no more material
was available it was impossible to characterize the species adequately,
although it was apparent that it was related to the signicauda group
of the subgenus Leptodiaptomus. A collection recently made in
Louisiana by Dr. Walter G. Moore of Loyola University contained
specimens agreeing with this form in the structure of the appendages,
and verifying its position in the signicauda group by the presence
in the female of a distal process on the right side of the genital
segment.

The species is named for Dr. Moore whose interest in the fresh-
water invertebrates of Louisiana has previously revealed two other
new species of Diaptomus (M. S. Wilson and Moore 1953a, b) and
added much valuable material to study collections.

DIAPTOMUS (LEPTODIAPTOMUS) MOOREI, sp. nov.
Specimens Examined. — Type lot: twenty-one 2 2 (two ovigerous),
five $ $ , shallow, muddy roadside pond, Louisiana highway 20, 12
miles south of Natchitoches, Natchitoches Parish, Louisiana, April
10, 1953, W. G. Moore. Associated with D. pallidus Herrick. Holo-
type 2, United States National Museum catalog number 96023; allo-
type $ , number 96024.

One 2, non-ovigerous; ditch pond, U. S. highway 71, south of
Lebeau, St. Landry Parish, Louisiana, April 5, 1951, W. G. Moore.
Associated with D. virginiensis Marsh, D. louisianensis M. S. Wilson
and Moore and Osphranticum labronectum S. A. Forbes.

Slide from Light collection, United States National Museum, $
leg 5 and right antennule; 2 leg 5. Pond in swampy area ("Gum
Swamp"), Walker County, Texas, May 11, 1940, E. S. Deevey. Asso-
ciated with D. clavipes Schacht and Osphranticum labronectum.

Diagnosis. — With these characters of the subgenus Leptodiaptomus:
one seta on segments 11 and 13-19 of female and left male anten-
nules. Right antennule male, segment 14 without spinous process.
Maxilliped slender, setae of distal portion not clawlike; four setae on
distal lobe of basal segment. Leg 2 with Schmeil's organ on endopod
segment 2 of both sexes. Leg 5 of female with third segment of
exopod not developed and two setae on endopod. Leg 5 of male:

52 Tulane Studies in Zoology Vol. 2

both processes of left exopod 2 short and digitiform, with rounded
ends, the inner placed medially; pads of left exopod 2 both medial
and well defined, without deep constriction between them; lateral
spine of right exopod 2 not inserted on the same plane as that of
the segment, strongly directed backwards; left basipod 2 with pro-
nounced crosswise grooving of medial inner margin; claw of right
leg at least as long as exopod.

Length, $ 1.27-1.32 mm; $ 1.15 mm. Greatest width of meta-
some in both sexes in segments 2 and 3. Segment 5 of female
(lateral view) not swollen dorsally. Metasomal wings (fig. 1) only
a little asymmetrical, not expanded laterally, inner part rounded but
not enlarged into prominent lobes, distal outer corner produced
posteriorly; sensilla minute. Urosome female (figs. 1, 2) two-
segmented; genital segment with small cuticular lobe at distal outer
portion of right side; proximal part with prominent lateral protru-
sions, that of left side larger and more rounded than that of right;
anal segment and caudal rami subequal in length; caudal rami ciliate
on inner margin. Urosome male (fig. 4) with segments 3 and 4
a little asymmetrical; segments 2, 3 and 4 with groups of spinules
on dorsal, posterior surfaces. Spermatophore elongate; when attached,
reaching beyond caudal rami of female (fig. 2). Ovisacs containing
16 and 17 small eggs respectively.

Antennules of female reaching beyond end of caudal rami. Right
antennule of male (fig. 5) with spine of segment 8 not enlarged,
that on 13 longer than that on 11 and reaching a little beyond seg-
ment 14. Proportions of spines to the segmental width and to one
another:

Segment

10

11

13

Segment width

12

12

16

Spine length

10

12

29

Process of segment 23 (fig. 6) not reaching to middle of segment
24, the apex strongly outcurved.

Leg 5, female (fig. 3). Basal sensillum minute, on enlarged papil-
liform protrusion. Length of exopod segments 1 and 2 subequal;
exopod with two lateral unequal setae. Endopod nearly as long as
inner margin of exopod 1, the inner prolongation sharply pointed,
the subequal setae a little less than half the length of the endopod
(9:19).

Leg 5, male (figs. 7, 8). Left leg reaching to end of right exopod
1. Basal sensilla of both legs elongate and slender, on enlarged
papilliform protrusions. Second basipod segments of both right and
left legs with prominent inner protrusions of proximal portion. That
of right basipod an upwardly projecting membranous process; the
left an upward projection with sclerotized margin, the area of the
usual inner groove posterior to this projection swollen marginally.
Right exopod segment 1 with large hyaline lamella, subrectangular
in shape, attached to midportion of segment and projecting mesiad.

No. 3

Wilson: A New Diaptomus

53

Figures 1-8. Diaptomus moorei, sp. nov. Female: 1. metasomal
wings and urosome with detail distal process of genital segment,
dorsal view; 2. same, lateral view, specimen with attached sperma-
tophore; 3. leg 5, with detail endopod setae. Male: 4. metasomal
wings and urosome, dorsal view; 5. right antennule, spines of seg-
ments 8-16; 6. same, apical segments; 7. leg 5, posterior view;
8. same, detail basipods and right exopod 1.

54 Tulane Studies in Zoology Vol. 2

Lateral spine of right exopod 2 strongly directed posteriorly, distally
placed at a point about 89 percent of total length of segment; its
length greater than the width of the segment. Claw a little longer
than exopod, evenly curved, hardly tapered beyond the slightly
swollen proximal portion so that its width is not greatly reduced
throughout; tip blunt. Left exopod segments subequal to one an-
other; processes of segment 2 subequal, the distal a little less than
one-fourth length of segment. Right endopod not enlarged, a little
shorter than first exopod segment; the left reaching just beyond the
middle of second exopod segment.

Notes on the Subgenus Leptodiaptomus

The subgenus Leptodiaptomus is the largest in number of species
and the most widely spread group of diaptomids in North America.
It is also represented in eastern Asia by angustilobus Sars (1898),
and by tyrrelli Poppe which was reported from Kamchatka by
Smirnov (1931). Since other leptodiaptomids occur in Alaska and
northern Canada {pribilojensis, nudus, sicilis, ashlandi) it is possible
that further investigation in the little known area of northern Asia
may reveal their presence there.

A new key to the North American calanoid species has been pre-
pared for the forthcoming revised edition of Ward and Whipple's
Fresh-Water Biology. This includes synonyms, summary of distri-
bution, and new illustrations of both sexes of all the known species
of Leptodiaptomus. It is hoped that this will considerably clarify
the confusion that has existed in the literature on this group. In this
present paper, therefore, only the taxonomy of moorei will be dis-
cussed.

Some points, however, should be made on the nomenclature of the
subgenus. As pointed out elsewhere (M. S. Wilson, 1951), Psychro-
diaptomus Kiefer (1938) proposed for angustilobus, is a synonym
of Leptodiaptomus Light (1938). In addition, the name Eutricho-
diaptomus Light (1939) should also be placed in its synonymy. This
subgenus was proposed for ashlandi, but I fail to find in comparative
study of all the species of the group, any structural differences of
subgeneric value separating it from the other species of Lepto-
diaptomus.

Also, there is in my mind a question as to whether the subgenus
Pelorodiaptomus Light (1939) proposed for trybomi, should be main-
tained. This species has never been found since the original col-
lection and accurate knowledge of much of its detail is still lacking.
The only specimens available in North America are in the Schacht
slide collection in the Illinois State Natural History Survey. These
are from the type lot and consist of two females and one male. These
slides have been studied and new drawings of the female prepared
for the key in Ward and Whipple. The male fifth leg was so
mounted that the relationship of pertinent parts could not be entirely
clarified to my satisfaction. However, none of the characters ob-

No. 3 Wilson: A New Diaptomus 55

served in this study indicates any sharp departure from basic lepto-
diaptomid structure. One of the most striking differences, but one
not of subgeneric value, is the exceptional length of the sensilla
(sensory spines) of the basal segments in the fifth legs of both sexes.
Oddly, this distinctive character in the female has never been noted
in literature. In this, as well as all other characters, trybomi appears
to present an extreme or bizarre development of the structures found
in leptodiaptomid species. The problem cannot be fully clarified
without additional unmounted specimens, and until such time as they
are available for study, the status of Pelorodiaptomus is uncertain.
Examination of the type material shows, however, that the species is
distinct, not synonymous with judayi nor "a chimaera based upon
poorly preserved material" as suggested by Kincaid (1953).

Comparative studies of leptodiaptomid species reveal within the
subgenus some natural groups which are held together by combina-
tions of several minor structural distinctions. Diaptomus moorei is
referable to one of these groups which includes signicauda. Lilljeborg,
novamexicanus Herrick and nudus Marsh. The characters indicating
the close relationship of these species are:

( 1 ) Presence in the female of a distal process on the right side
of the genital segment, which is also characterized by rather promi-
nent lateral protrusions of the area bearing the sensilla.

(2) Endopod of the female fifth leg not swollen or rounded
distally, but instead produced into a sharply pointed projection beyond
the insertion of the setae.

(3) The process of the twenty-third segment of the male right
antennule very short (not reaching beyond middle of segment 24)
and strongly outcurved at the tip.

(4) Spine of segment 13 of male right antennule longer than
that of segment 11. Segments 15 and 16 without cuticular spinous
processes, except for a minute process usually present on segment 15
of signicauda.

( 5 ) Claw of right fifth leg of male not tapered, but rather evenly
broadened throughout its length, with slightly constricted, rounded
or blunt end.

(6) Second basipod segment of male right leg with the inner
proximal portion bulging upward, or with hyaline membranous
extension.

In this group, the females of signicauda and moorei resemble one
another closely. The metasomal wings are well developed in all, but
those of novamexicanus and nudus are individually quite distinctive
with prominent inner lobes on the left side and pronounced asym-
metry; those of signicauda and moorei are very similar to one an-
other, lacking well developed inner lobes and being only slightly
asymmetrical. D. signicauda and nudus may show, in lateral view, a
more or less pronounced gibbosity of the fifth metasomal segment
not noted in moorei, but this is a variable character and its absence

56 Tulane Studies in Zoology Vol. 2

would not necessarily serve to distinguish moorei from signicauda.
The process of the genital segment is very short in all the specimens
of moorei, thus differing from the much more developed process
observed in signicauda, but the possible extent of variation is not
known. Considerable care should be taken therefore in separating
females of these two species.

Two other leptodiaptomids, judayi and spinicornis, have on the
genital segment a distal process similar in structure to that of this
group. The immediate relationship of these two species does not
appear to be with the signicauda group since the males differ in
several characters.

The males of the signicauda group are distinct from one another
in several characters. The great length of the spine of segment 13
of the right antennule as shown for moorei (fig. 5) is also found in
nudus; it is shorter in the other two species, though always longer
than that on segment 11. D. novamexicanus is strikingly individual
in the structure of the process of the twenty-third segment of the
male right antennule. Ds nudus has an individual character in the
presence of small cuticular lobes on the ventral inner side of the
caudal rami of the male. There is a striking difference in the species
of the group in the position of the lateral spine of the second exopod
segment of the right fifth leg, that of moorei being most distally
placed. The approximate comparative position of this spine in the
four species is shown by the following percentage figures which
indicate the point of placement of the spine from the base of the
segment, in relation to the total length of the segment:

nudus 43%

novamexicanus 65 %

signicauda 11%

moorei 89%

The fifth leg of the male of Diaptomus moorei differs from the
other species in the position of this spine; in the enlarged inner pro-
trusion of the left second basipod segment; and in the subrectangular
shape and medial position of the lamellar process of the right first
exopod segment. In identification of leptodiaptomids, care should
be taken not to confuse this species with siciloides or connexus, two
species closely related to one another, which also have a somewhat
squarish or rectangular lamella on the right first exopod segment,
and a similar process on the twenty-third segment of the right anten-
nule. These species can be separated from moorei by the position
of the lamella; that of siciloides and connexus is placed in the distal
half of the segment, that of moorei is centrally placed. These species
are further separable from moorei in both sexes by most of the char-
acters which distinguish the signicauda group.

The differentiation of the signicauda group and of leptodiaptomid
species which might be confused with them is shown by the following
keys.

No. 3 Wilson: A New Diaptomus 57

Females
(Species of Diaptomus (Leptodiaptomus) having distal process on
right side of genital segment)

1. Genital segment rounded but without

distinct protrusions spinicornis Light

Genital segment with distinct lateral

protrusions . 2

2. Left metasomal wing with prominent

inner lobe extending posteriorly be-
yond the outer portion of the wing 3

Left metasomal wing without such

prominent inner lobe. 4

3. Left wing conspicuously larger than

the right. .novamexicanus Herrick

Left wing not much larger than the

right nudus Marsh

4. Right wing, outer portion not posteri-

orly produced. judayi Marsh

Right wing, outer portion posteriorly

produced . 5

5. Genital segment, process reaching to

near middle of anal segment or be-
yond. signicauda Lill'jeborg

Process not reaching to middle of last

segment. moorei sp. nov.

Males

(Species in which the process of segment 23 of right antennule

does not reach beyond middle of segment 24, and has outcurved or

hooklike tip)

1. Right antennule, spine of segment 11
subequal to or longer than that of
segment 13. 2

Right antennule, spine of segment 11

shorter than that of segment 13 3

2. Right antennule, spine of segment 8

usually not enlarged (about same
length as that on segment 12) ;
metasome in dorsal view with great-
est width at about the middle siciloides Lilljeborg

Spine of segment 8 enlarged (at least
twice the length of that on segment
12) ; metasome in dorsal view with
greatest width in cephalic segment connexus Light

3. Right antennule, segment 23, base of

process starting at middle of seg-
ment. novamexicanus Herrick

This process entirely apical 4

4. Leg 5, right exopod 2, lateral spine

above middle of segment nudus Marsh

58 Tulane Studies in Zoology Vol. 2

This spine below middle of segment 5

5. Leg 5, right exopod 1 with rounded,
distall'y directed hyaline process on
inner margin. signicauda Lilljeborg

This segment with subrectangular,

medially directed process moorei sp. nov.

Certain synonymy among these species should be noted here. No
types of D. novamexicanus are in existence, but it appears impossible
in the study of specimens from over a wide geographical range to
recognize more than one species in which the process of the male
right antennule is so peculiarly modified. The names washingtonen-
sis Marsh (1907) and garciai Osorio Tafall (1942) are therefore
apparently synonyms of novamexicanus as has already been suggested
by Kincaid (1953).

Some other synonyms in Diaptomus (Leptodiaptomus) proposed
by Kincaid are not tenable. As pointed out above, judayi is not a
synonym of trybomi. Likewise, connexus is not identical with
siciloides, though it is not yet clear what the type of taxonomic rela-
tionship between the two species is. The fifth leg of the males are
indeed very similar as pointed out by Kincaid, but the body form of
both sexes (see Key above) and the metasomal wings and the uro-
some of the females are distinct in the two forms. The differences
in the females are as great as any that commonly exist between other
allied species of leptodiaptomids. These characters were not variable
in any of the several collections of both forms that have been
examined (including the type lot of connexus) so that the exact
relationship of the two may not be determinable on the basis of
anatomical studies alone. Attention should be given not only to
their variations, but to their distribution, associations and ecology,
particularly in the geographical areas in which they co-occur. Much
of significance applicable to the taxonomy of the genus may be
learned from study of cases such as these, so it is important that
their differences not be ignored or obscured as they would be if
accepted as synonymous without critical appraisal. Therefore, I have
retained the two names and separated them as species in the key in
the forthcoming revised edition of Ward and Whipple's Fresh-Water
Biology, in which figures illustrating the differences are included
(M. S. Wilson, In Press).

Diaptomus siciloides has a very broad distribution over the North
American continent, having been reported from most areas except
the extreme north and the east coast. In Louisiana, it is the species
listed by Moore (1950) as Diaptomus sp. from Lake Providence.
Diaptomus connexus was originally described from Kern County,
California by Light (1938) and additional collections in the Light
accession in the National Museum indicate that it has a much wider
distribution in the southwestern portion of the continent. It has
been identified by Dr. Light from other California localities, from
Nevada, Colorado, New Mexico and Lower California. Both species

No. 3 Wilson: A New Diaptomus 59

have been found in lakes, reservoirs and ponds. A Mexican form
named cuauhtemoci by Osorio Tafall (1941) is also involved in the
synonymy of one of these species. It may be referable to one or the
other, but examination of whole topotypic specimens would be neces-
sary to determine the exact identity.

Of the species of the signicauda group, signicauda and novamexi-
canus are known to be very common in the western part of the
continent, and particularly so in the Rocky and Cascade mountains
and the Sierra Nevadas. D. signicauda has been reported from Iowa
(Stromsten, 1920) which is outside its apparent normal geographic
range. D, novamexicanus is perhaps more widely distributed than
signicauda, being common at sea level in California, and extending
into Mexico. The record of this species from the Yucatan Peninsula
by C. B. Wilson (1936) would appear to include another species
since the figures given for the fifth legs of both sexes are neither
novamexicanus nor referable to the subgenus Leptodiaptomus. The
process of the male right antennule is similar to that of novamexicanus
and may have been drawn from that species.

These two species have not yet been reported or collected from the
far northern areas of the continent, but nudus, which also occurs in
the Rocky Mountains of both the United States and Canada, may
have an Arctic-alpine distribution pattern. It was found recently
in a collection made for me from a tundra pond at Bethel, on the
western Bering coast of Alaska.

The occurrence of the new species moorei in Louisiana and Texas,
suggests that it should be looked for in the Mississippi Valley, other
southern states of the Gulf of Mexico area and in Mexico and Cen-
tral America.

References Cited

Kiefer, F. 1938. Freilebende Susswassercopepoden von den Nord-
kurilen. Bull. Biogeogr. Soc. Japan, 8(4) : 75-94.

Kincaid, TREVOR 1953. A Contribution to the Taxonomy and Distri-
bution of the American Fresh-water Calanoid Crustacea. Cal-
liostoma Co., Seattle, 73 pp.

Light, S. F. 1938. New subgenera and species of diaptomid copepods
from the inland waters of California and Nevada. Univ. Calif.
Publ. Zool, 43(3) : 67-78.

1939. New American subgenera of Diaptomus

Westwood (Copepoda, Calanoida). Trans. Amer. Micros. Soc,
58(4) : 473-484.

Marsh, C. Dwight 1907. A revision of the North American species
of Diaptomus. Trans. Wise. Acad. Sci., Arts, Lttrs., 15(2):
381-516.

Moore, Walter G. 1950. Limnological studies of Louisiana lakes.
I. Lake Providence. Ecology, 31(1): 86-99.

Sars, G. O. 1898. The Cladocera, Copepoda and Ostracoda of the
Jana Expedition. Ann. Mus. Zool. (Acad. Imp. Sci.) St. Peters-
burg, 3: 324-359.

Smirnov, S. S. 1931. Ein Beitrag zur Copepoden-Fauna des Amur-
Gebietes. Archiv Hydrobiol, 23: 618-638.

60 Tulane Studies in Zoology Vol. 2

Stromsten, Frank A. 1920. Copepoda of the Okoboji region. Proc.
Iowa Acad. Sci., 27: 269-270.

Tafall, B. F. Osorio 1941. Diaptomus cuauhtemoci no v. sp. de la
Mesa central de Mexico. Ciencia (Mexico), 2: 296-298.

1942. Un nuevo "Diaptomus" del Mexico central

(Copepoda, Diaptomidae). Rev. Brasil Biol., 2(2): 147-154.

Wilson, Charles Branch 1936. Copepods from the cenotes and
caves of the Yucatan Peninsula, with notes on cladocerans.
Carnegie Inst. Wash. Publ. 457: 77-88.

Wilson, Mildred Stratton 1951. A new subgenus of Diaptomus
(Copepoda: Calanoida), including an Asiatic species and a new
species from Alaska. Jour Wash. Acad. Sci., 41(5) : 168-179.

The Calanoida. In: Ward and Whipple, Fresh-
Water Biology, Revised Ed. (In Press.)

Wilson, Mildred Stratton and Walter G. Moore 1953a. New rec-
ords of Diaptomus sanguineus and allied species from Louisiana,
with the description of a new species (Crustacea: Copepoda).
Jour. Wash. Acad. Sci., 43(4) : 121-127.

1953b. Diagnosis of a new species

of diaptomid copepod from Louisiana. Trans. Amer. Micros.
Soc., 72(3): 292-295.

* n - / 1 [© w ov fea **>-sj

a

ip ®® as

Volume 2, Number 4

December 28, 1954

THREE NEW SPECIES AND NEW RECORDS OF SOUTHERN

MILLIPEDS

NELL B. CAUSEY,

FAYETTEVILLE, ARKANSAS

MBS. COMP. ZiOL
UBMH

JAN l 2 19!

sjty

TULANE UNIVERSITY
NEW ORLEANS

TULANE STUDIES IN ZOOLOGY is devoted primarily to the-
zoology of the area bordering the Gulf of Mexico and the Caribbean
Sea. Each number is issued separately and deals with an individual
study. As volumes are completed, title pages and tables of contents
are distributed to institutions exchanging the entire series.

Manuscripts submitted for publication are evaluated by the editor
and by an editorial committee selected for each paper. Contributors
need not be members of the Tulane University faculty.

EDITORIAL COMMITTEE FOR THIS NUMBER

R. V. Chamberlin, Emeritus Professor of Zoology, University
of Utah, Salt Lake City, Utah.

H. F. Loomis, Senior Agronomist, U. S. Plant Introduction
Garden, Coconut Grove, Florida.

Robert A. Hefner, Professor of Zoology, Miami University,
Oxford, Ohio.

Richard L. Hoffman, Department of Entomology, Cornell
University, Ithaca, New York.

Manuscripts should be submitted on good paper, as original type-
written copy, double-spaced, and carefully corrected.

Separate numbers may be purchased by individuals, but subscriptions
are not accepted. Authors may obtain copies for personal use at cost.
Address all communications concerning exchanges, manuscripts, edi-
torial matters, and orders for individual numbers to the editor. Re-
mittances should be made payable to Tulane University.

When citing this series authors are requested to use the following ;
abbreviations: Tulane Stud. Zool.

Price for this number: $0.25.

George Henry Penn, Editor,

c/o Department of Zoology,
Tulane University,
New Orleans, U. S. A.

Assistants to the Editor:
Carol L. Freret
Donald W. Tinkle

mud. ivrir. l

JAfJ

THREE NEW SPECIES AND NEW RECORDS OF SOUTHERN

MILLIPEDS
NELL B. CAUSEY,

Fayetteville, Arkansas

Unless stated otherwise, the specimens mentioned in this paper were
collected by the author with the assistance of Dr. David Causey and
are now in the author's collection; holotypes will be deposited in the
American Museum of Natural History.

Order Nematophora
Suborder Lysiopetalida

Three genera of lysiopetaloid millipeds are represented in the fauna
of the Gulf States. The best known is Abacion Rafinesque (1820)
of which several species occur in the states bound by a line from
central Texas to North Dakota on the west and the Atlantic coast on
the east. Three species of Delophon Chamberlin (1943) are known
from North Carolina, Georgia, and Mississippi, but Texophon
Chamberlin (1946) is known only from the generotype from Laguna
Madre, Cameron Co., Texas. In species of Delophon and Texophon
the body length is 25 mm or less, making them subject to easy con-
fusion with larval specimens of the larger and ubiquitous species of
Abacion*

The following key to these genera and to the three species of
Delophon is based mainly on the gonopods of adult males.

1. Terminal branches of telopodite very

short. Texophon

Terminal branches of telopodite long,

conspicuous. 2

2. Terminal branches of telopodite vari-

ously curved, directed laterad. Cox-
ite of gonopodal segment never
sheath-like. Abacion

Terminal branches of telopodite almost
straight, retrorse. Coxite on gonop-
odal segment forms a long, partial
sheath around the shaft of the
telopodite. (Delophon) 3

3. Terminal region of telopodite divided

into two sharply pointed branches.
About 34 ocelli in 7 or 8 transverse
series. Delophon serrulatum, sp. nov.

Terminal region of telopodite divided
into three sharply pointed branches.
Ocelli in 8 or 9 transverse series. 4

4. About 58 ocelli in each eye

Delophon georgianum Chamberlin (1943).

About 43 or 45 ocelli in each eye

Delophon carolinum Hoffman (1950).

Family DORYPETALIDAE
Genus Delophon
Delophon Chamberlin, 1943, Bull. Univ. Utah, Biol. Ser., 8(2): 13.

64 Tulane Studies in Zoology Vol. 2

DELOPHON SERRULATUM, sp. nov.

Diagnosis. — Distinguished from the other two species of the genus
by the reduced number of ocelli and by the presence of two rather
than three acute terminal branches on the telopodite of the male
gonopods.

Type locality. — Mississippi: Baldwin County, 9 miles west of Loxley
at the junction of U. S. Highway No. 90 and Mississippi State Route
No. 89; Jan. 5, 1954; 1 male.

Male holotype. — Body slender, cylindric, about 26.5 mm long, 1.2
mm wide; 48 segments, the last three legless. Medium brown above;
legs and venter light brown. Eyes triangular, about 34 ocelli in seven
or eight transverse series. Segments 2 and 3 of antennae subequal
and slightly longer than segment 4, which is longer than 5. Forehead
moderately rounded. Segment 6 conspicuously larger than segments
1 through 5 and slightly larger than 7 and the following segments.
Collum with about 20 longitudinal crests, all ending on the caudal
margin and of uniform height but not of uniform length. On the
anterior segments both the primary and the secondary crests extend
the full length of the metazonite and are subequal in height; on the
middle segments the secondary crests diminish in both height and
length, and on the posterior segments they fade out. The transition
from ten to fourteen dorsal crests between the repugnatorial pores
takes place on segment 12. Pore crests fusiform and scarcely more
prominent than the adjacent primary crests. Last tergite short, with-
out crests, but with six tuberculate setae near the hind margin and
two long, projecting papillate hairs. Setae at end of primary crests
not more than one-fourth the length of the crests. Legs 2, 3, and
4 each with a velutinous pad on the ventral surface of the distal half
of the last article.

In situ the two gonopods are parallel and contiguous as far distad
as the terminal branches, which are bent sharply in a retrorse direc-
tion. The long, ventro-caudally arched shaft of the telopodite is
partially enclosed by the coxal sheath, or coxite, which extends as far
ventrad as the terminal branches. The two terminal branches (figs.
1, 2, bl, bm) are almost straight and sharply pointed; the lateral
branch is broader and slightly longer than the medial branch. The
seminal canal branch (fig. I, be) is short, blunt, and arises from the
medial branch. The coxal sheath is minutely rugose, and its apical
region, which is less expanded along the mesial margin than in
georgianum, is thin and irregularly serrulate (fig. 3, cs).

Ecology, — The collection site is a hilly, thin, mixed woodland.
Myriapods were scarce; other than this specimen, a few lithoboid
centipeds and larvae of Abacion sp. and Cleidogona sp. were found.

Suborder Chordeumida
Family Cleidogonidae
Genus Cleidogona
Cleidogona Cook and Collins 1896, Ann. New York Acad. Sci.,
9(1-3): 41-42.

No. 4

Causey: Three New Millipeds

65

Cleidogona aspera Causey

Cleidogona aspera Causey, 1951, Jour. Washington Acad. Sci.,
41(2):78, figs. 1-4.

New record. — Louisiana: Lincoln Par., Ruston; Oct., 1951; W. J.
Harman; 3 males, 5 females.

Figures 1-10. Delophon serrulatiim, sp. nov. : 1. distal region of
telopodite of left gonopod, lateral view ; 2. same, caudal view ; 3. basal
region of left gonopod, mesial view. Cleidogona sublettei, sp. nov.:
4. ventral branch of left gonopod, ventral view; 5. left gonopod,
lateral view. Cleidogona arkansana, sp. nov.: 6. ninth leg; 7. base
of eleventh leg; 8. gonopods, ventral view; 9. left gonopod, lateral
view. Orthoporus omatus (Girard) : 10. ventral margin of collum,
right side, female. Abbreviations: be = seminal canal branch, bl=-
lateral branch, bm = mesial branch, co = coxa, cs = coxal sheath,
d = dorsal branch, p = lateral process, s = cut middle surface of
sternum, t = shaft of telopodite, v = ventral branch.

66 Tulane Studies in Zoology Vol. 2

CLEIDOGONA SUBLETTEI, sp. nov.

Diagnosis. — The male gonopods of this species closely resemble
those of C. mississippiana Chamberlin (1942). The two species are
best separated by the absence in C. sublettei of a subterminal spine
from the dorsal surface of the ventral branch.

Type locality. — Louisiana: Natchitoches Parish, Grand Ecore; Nov.
13, 1953; James E. Sublette, 3 males.

Male holotype. — Length about 20 mm. Brown above and laterad,
with the usual areolated buff maculae; venter cream; tarsi brown,
remainder of legs cream. Antennae and vertex of head brown;
ocelli black, forming a triangular patch, in series of 1, 7, 6, 5, 4, 3, 2.

The second articles of the ninth leg are as in mississippiana, but
the coxae differ in that the excavation is broader, as in C. laminata
Cook and Collins (1895) and C. aspera Causey (1951), and the
distal surface is granular. The coxae of the tenth and eleventh legs
have each the usual short, cylindrical process through which the coxal
gland opens; the eleventh coxae have each a small, flattened, capitate
process on the caudal surface. The sternal process at the base of the
twelfth legs is as in C. unita Causey (1951).

The ventral branches of the gonopods are attenuated, subparallel,
relatively simple, and slightly arched, so the apices rest on the base
of the eleventh legs; the laterally directed subterminal process (figs.
4, 5, v, p) is a spine rather than a lamella, as it is in mississippiana.
The dorsal branch is straight, with the apical region broad and
flattened (fig. 5, v) .

CLEIDOGONA ARKANSANA, sp. nov.

Diagnosis. — The male gonopods of this species are nearest those of
C. fustis Cook and Collins (1895). The two species are easily sepa-
rated by the apex of the ventral branch, which in fustis is deeply
bifid and abruptly twisted, and in arkansana is shallowly bifid and
not twisted.

Type locality. — Arkansas: Dallas Co., four miles east of Princeton
at a picnic site on Arkansas State Route No. 8; Jan. 7, 1954; one
male, one female.

Male holotype. — Length about 15 mm. Light brown above and
laterad, the usual areolated maculae scarcely visible. Antennae and
vertex of head brown; ocelli black, forming a triangular patch, in
series of 1,7,6,5,4 (3), 3, 2.

The seventh legs have each a mammillate process as large as the
processes of the tenth legs on the caudal surface of the coxae. The
ninth legs (fig. 6) have the first articles relatively simple, but the
second articles are produced ventrad in an unusual way. The coxae
of the tenth and eleventh legs have each the usual short, cylindrical
process through which the coxal gland opens; the eleventh coxae
(fig. 7) have each an additional, smaller, cylindrical process on the
caudal surface. The sternal process at the bases of the twelfth legs
is as in C. unita Causey (1951).

The ventral branches of the gonopods (fig. 8, v) are slightly

No. 4 Causey: Three New Miilipeds 67

bowed laterad, bringing the apices almost together; most of the space
between these branches is filled by a thin lamella. The apex of the
dorsal branch is truncated, shallowly bifid, and slightly curved laterad
about the apex of the ventral branches. The arched dorsal branch
(fig. 9) is abruptly bent at the apex.
Female allotype. — Somatic characters agree with those of the male.

Order Polydesmida

Family EURYMERODESMIDAE
Genus Eurymerodesmus
Eurymerodesmus Brolemann 1900, Proc. Biol. Soc. Wash., 33:97-98.

Eurymerodesmus varius (McNeill )

Polydesmus varius McNeill, 1887, Proc. U. S. Natl. Mus., 10-323-
324, pi. 11, figs. 1, 2 (female, Escambia Co., Fla.).

Eurymerodesmus minimus Loomis, 1943, Jour. Washington Acad.
Sci., 33 (10): 320, fig. 2 (male, Marianna, Jackson Co., Fla.).

New records. — Florida: Escambia Co., Pensacola and Cantonment;
Jan. 5, 1954; under live oaks, adults and larvae of both sexes. Ala-
bama: Mobile Co., Spring Hill; March, 1954; one female; C. E.
Valentine.

The specimens of E. varius in my collection correspond so closely
to Loomis' excellent description of his E. minimus that there can be
little doubt that the latter must be regarded as a synonym of E. varius.

Family XYSTODESMIDAE
Genus Apheloria
Apheloria Chamberlin 1921, Canad. Ent., 53:232.

Apheloria reducta Chamberlin

Apheloria reducta Chamberlin, 1939, Bull. Univ. Utah, Biol. Ser.,
5(3) : 11, pi. 4, fig. 35 (male, Imboden, Lawrence Co., Ark.).

Neiv record. — Oklahoma: McCurtain Co.; July 20, 1954; one male,
width 10.8 mm, length about 41 mm; W. J. Harman.

This species has heretofore been known only from northern and
western Arkansas and southern Missouri.

Order Spirostreptida
Family SPIROSTREPTIDAE

Genus Orthoporus

Orthoporus Attems 1914, Zoologica, 25(5/6): 132.

Orthoporus ornatus (Girard)

Julus ornatus Girard, 1853- In Marcy, Report on Red River
Exped., p. 274 ("collected on the 27th of June").

Spirobolus ornatus, Wood, 1865, Trans. Amer. Philos. Soc, 13:208-
209; Bollman, 1887, Ann. New York Acad. Sci., 4:43.

Netv record. — Texas: Randall Co., Palo Duro Canyon State Park;
July, 1953; 1 female; H. H. litis.

Girard's references to the longer antennae, the shape of the eyes,
and the upper lip indicate that his Julus ornatus is an orthoporid, not

68 Tulane Studies in Zoology Vol. 2

a spirobolid, but there is little of specific value in his description.
His collection date fixes the type locality as either within or very
near the present Palo Duro Canyon State Park, Randall County, Texas.
A resident of that area reported to me in 1953 that "these big
thousand-leggers have been migrating across our road in great num-
bers all summer."

Female topotype. — Width 11 mm, length about 185 mm, 70 seg-
ments. Most segments with three annuli, of which the most anterior
is tan and visible only on the dorsal surface of many segments, the
middle is widest and almost black, and the caudal is narrow and dark
red. Legs almost black. The blue color observed by Girard and not
observed in this specimen is usually associated with a recently molted
condition. Eyes composed of seven transverse series of from four-
teen to three ocelli. Both corners of collum (fig. 10) rounded, the
anterior slightly produced, and the ventral margin slightly concave
between the corners. Collum with four impressed lines branching
from the marginal sulcus to the caudal margin; the upper limit of
the marginal sulcus is the height of the caudal row of ocelli. Body
surface finely and evenly punctuate. Segmental sutures light, only
slightly indented toward the pores, which are located about one-fourth
the distance from the segmental sutures to the caudal margin of the
segments. Caudal one-third of last tergite abruptly depressed and
more coarsely punctuate than the remainder of the surface; apex of
tergite slightly obtuse and well exceeded by the thin margins of the
anal valves.

The large size and the shape of the lateral margins of the collum
should serve to distinguish females of O. ornatus from other or-
thoporids in adjacent areas of Texas and Oklahoma.

References Cited

Causey, Nell B. 1951. New cleidogonid millipeds. Jour. Wash. Acad.
Sci., 41 (2) : 78-83.

Chamberlin, R. V. 1942. New southern millipeds. Bull. Univ. Utah,
Biol. Ser., 6 (8) : 1-19.

1943. On some genera and species of American

millipeds. Ibid., 8 (2) : 1-20.

1946. Texophon, a new genus in the diplopod fam-
ily Lysiopetalidae. Ent. News, 57 (4) : 97-99.

Cook, 0. F. and G. N. Collins 1895. The Craspedosomatidae of
North America. Ann. New York Acad. Sci., 9: 1-100.

Hoffman1, Richard L. 1950. Records and descriptions of diplopods
from the southern Appalachians. Jour. Elisha Mitchell Sci. Soc,
66 (1) : 11-33.

Hoffman, Richard L. and Ralph E. Crabill, Jr. 1953. C. S. Ra-
finesque as the real father of American myriapodology : an
analysis of his hitherto unrecognized species. Florida Ent., 36
(2) : 73-82.

Loomis, H. F. 1937. Crested millipeds of the family Lysiopetalidae
in North America, with descriptions of new genera and species.
Proc. U. S. Nat. Mus., 84 (3006) : 97-135.

Verhoeff, K. W. 1932. Unterordnung Lysiopetaloidea. In Bronns
Tierreich, Klasse Diplopoda, pp. 1500-1522.

^

V

_I

Volume 2, Number 5

January 12, 1955

NOTROPIS BAILEY I, A NEW CYPRINID FISH FROM THE

PASCAGOULA AND MOBILE BAY DRAINAGES OF

MISSISSIPPI AND ALABAMA

ROYAL D. SUTTKUS,

DEPARTMENT OF ZOOLOGY, TULANE UNIVERSITY,

NEW ORLEANS, LOUISIANA

and

EDWARD C. RANEY,

DEPARTMENT OF CONSERVATION, CORNELL UNIVERSITY,
ITHACA, NEW YORK

MUS

. CSf

(P. ZIQL

*

JAN

25

1955

TULANE UNIVERSITY
NEW ORLEANS

TULANE STUDIES IN ZOOLOGY is devoted primarily to the
zoology of the area bordering the Gulf of Mexico and the Caribbean
Sea. Each number is issued separately and deals with an individual
study. As volumes are completed, title pages and tables of contents
are distributed to institutions exchanging the entire series.

Manuscripts submitted for publication are evaluated by the editor
and by an editorial committee selected for each paper. Contributors
need not be members of the Tulane University faculty.

EDITORIAL COMMITTEE FOR THIS NUMBER

Carl L. Hubbs, Professor of Biology, Scripps Institution of
Oceanography, La Jolla, California.

Reeve M. Bailey, Curator of Fishes, Museum of Zoology, Uni-
versity of Michigan, Ann Arbor, Michigan.

Ernest A. Lachner, Associate Curator of Fishes, United States
National Museum, Washington, D. C.

Manuscripts should be submitted on good paper, as original type-
written copy, double-spaced, and carefully corrected.

Separate numbers may be purchased by individuals, but subscriptions
are not accepted. Authors may obtain copies for personal use at cost.
Address all communications concerning exchanges, manuscripts, edi-
torial matters, and orders for individual numbers to the editor. Re-
mittances should be made payable to Tulane University.

When citing this series authors are requested to use the following
abbreviations: Tulane Stud. Zool.

Price for this number: $0.30.

Assistants to the Editor:
Carol L. Freret
Donald W. Tinkle

George Henry Penn, Editor,

c/o Department of Zoology,
Tulane University,
New Orleans, U. S. A.

MUS.

. ZiOL

JAN

2 5 19

1 i

il

NOTROPIS BAILEY I, A NEW CYPRINID FISH FROM THE
PASCAGOULA AND MOBILE BAY DRAINAGES OF

MISSISSIPPI AND ALABAMA

ROYAL D. SUTTKUS,

Department of Zoology, Tulane University,
New Orleans, Louisiana

and

EDWARD C. RANEY,i

Department of Conservation, Cornell University,

Ithaca, New York

Intensive ichthyological field studies in southeastern United States
in recent years have revealed several new forms of Notropis, the larg-
est genus of North American Cyprinidae. One is described below
and also included are studies of other closely related forms, especially
Notropis lutipinnis (Jordan and Brayton) and Notropis chrosomus
(Jordan), which are little known. The habits and ecology of these
small fishes are suggested as interesting and worthwhile studies for
regional ichthyologists.

Material of the new species has been collected by the authors and
others listed below to whom we give hearty thanks for personal as-
sistance in the field or for the loan of specimens. They are Bancroft
Cooper, Ralph L. Chermock, Charles F. Cole, Robert H. Gibbs, Her-
bert D. Gibson, Charles D. Hancock, P. S. Handwerk, L. James Kezer,
H. T. Rainwater, C. Richard Robins, Merlin G. Suttkus and William
S. Woolcott. Vladimir Walters has made available tooth and anal
ray counts for Notropis lutipinnis. Helen J. Illick has kindly sup-
plied counts for the cephalic lateral line pores from an unpublished
study of these structures in North American Cyprinidae. We are
grateful to Reeve M. Bailey, Ernest A. Lachner, and Carl L. Hubbs
for a critical review of the manuscript and other favors. This form
is named for Dr. Reeve M. Bailey, Curator of Fishes, Museum of
Zoology, University of Michigan, who has contributed much to the
knowledge of North American freshwater fishes and who has been
ever helpful in our studies. Counts and measurements were taken
as described by Hubbs and Lagler (1947: 8-15).

NOTROPIS BAILEYI, sp. nov.
Figs. 1 and 2, Map 1

The type material consists of 222 specimens from 21 to 63 mm in
standard length taken at seven localities in the Alabama River System.
Other material examined consists of 338 specimens, 19-55 mm from
19 localities in the Tombigbee and Pascagoula drainages. Below in
parentheses are indicated the number of specimens and the range of
standard length in millimeters, e.g. (4, 35-56). In addition to stand-

1 Aid for field studies and collections of fishes now at Cornell was
obtained through awards from the Cornell University Faculty
Research Grants Committee.

72 Tulane Studies in Zoology Vol. 2

ard compass directions, with the following "of" deleted, these ab-
breviations are used: Co. = County, Cr. =: Creek, mi. = mile or
miles, R. = River, trib. = tributary (of), ANSP = The Academy of
Natural Sciences of Philadelphia, CU = Cornell University, TU =.
Tulane University, UMMZ = University of Michigan, Museum of
Zoology and USNM = United States National Museum.

Material. — Holotype, CU 28224, an adult female 59-6 mm in
standard length, taken in the Alabama River system in a trib. Sawack-
lahatchee Cr., 1.7 mi. W. Society Hill on U. S. Route 80, Macon Co.,
Alabama, on June 12, 1949, by Royal D. Suttkus, Charles F. Cole and
Robert H. Gibbs. Designated as paratypes (CU 16020) are 66
specimens from 29 to 63 mm in standard length which bear the same
data as the holotype.

Other paratypes from the Alabama River system of Alabama are
as follows: TU 3195 (4, 33-56), Pine Barren Cr, 2.5 mi. S. Forest
Home, Butler Co, June 2, 1951; TU 2637 (30, 30-43), trib. Alabama
R, 17.3 mi. S.S.W. Camden, Monroe Co, June 3, 1951; TU 3430
(75, 29-48), Pursley Cr, trib. Alabama R, 3.4 mi. S.W. Camden,
Wilcox Co, June 3, 1951; TU 2965 (4, 21-46), Gravel Cr, trib.
Pursley Cr, 6.3 mi. S.W. Camden, Wilcox Co, June 3, 1951; TU
3066 (41, 31-52), trib. Pursley Cr, 1.8 mi. E. Camden, Wilcox Co,
June 3, 1951; USNM 163568 (1, 31), 5 mi. W. Attalla, Etowah Co,
September 13, 1947; UMMZ 105526 (2, 59-61), branch 4 mi. W.
Auburn. Wire Road, [Lee Co.], June 1938; UMMZ 105554 (73,
33-53), creek, 4 mi. W. Auburn, Wire Road, [Lee Co.], March 9,
1939; UMMZ 124018 (1, 27), Big Swamp, 5 mi. S.E. Tuskegee,
[Macon Co.], September 13, 1937; UMMZ 124024 (1, 61), Long
Branch, 1 mi. W. Marvyn (Columbus Road), [Lee Co.], September
20, 1937; UMMZ 124094 (41, 18-48), Loblockee Cr, N. Auburn,
[Lee Co.], September 19, 1938; UMMZ 124117 (45, 27-55), Olin
Cr, (=Odom Cr, S. Auburn), [Lee Co.], September 26, 1938;
UMMZ 146517 (5, 42-47), Beaver Cr, 5 mi. W. Autaugaville,
Autauga Co, June 13, 1942; UMMZ 146533 (16, 34-55), trib.
Beaver Cr, 5 mi. W. Autaugaville, Autauga Co, March 20, 1942;
UMMZ 146544 (97, 30-56), Beaver Cr, 5 mi. W. Autaugaville,
Autauga Co, December 20, 1942.

Additional specimens examined are as follows: Black Warrior
drainage, Alabama, CU 13991 (2, 36-37), Black Warrior R, 10.1
mi. N.E. Eutaw, Greene Co, March 31, 1948; CU 21165 (2, 26-30),
3.5 mi. N.W. Black Warrior R. bridge at Demopolis, Greene Co,
March 10, 1951; CU 21166 (26, 19-46), Woods Cr. on Romulus
Rd, Tuscaloosa Co, March 10, 1951; CU 13999 (6, 28-45), trib.
Hurricane Cr, 1.6 mi. W. Vance, Tuscaloosa Co, March 31, 1948;
CU 13785 (38, 19-51), trib. Black Warrior R. at Cottondale, Tus-
caloosa Co, March 31, 1948; CU 19264 (20, 24-45), Hurricane Cr,
6 mi. E. Tuscaloosa, Tuscaloosa Co, October 21, 1950; CU 19259
(60, 28-54), Cottondale Cr. near Hurricane Cr, Tuscaloosa Co, May
19, 1950; UMMZ 166361 (60, 21-45), creek, 5 mi. N. Tuscaloosa,

No. 5 Suttkus and Raney: A New Cyprinid 73

detour US Hwy. 45, [Tuscaloosa Co.], May 4, 1952.

Tombigbee drainage, Alabama, CU 15544 (5, 32-45), trib. Ala-
muchee Cr., 2.2 mi. N.E. York, Sumter Co., March 31, 1948; CU
25705 (38, 29-41), Williams Cr., just N. Hamilton, Marion Co.,
August 17, 1953; UMMZ 113914 (42, 27-42), headwater of Lux-
apallila Cr., Va mi. S.E. Winfield, Fayette Co., June 4, 1932; UMMZ
135892 (2, 38-42), Vaughn Cr., 4 mi. N.E. Mount Sterling, Choctaw
Co., September 4, 1942; UMMZ 141152 (1, 46), lake at Winfield
(Winfield is on Luxapallila Cr.), [Marion Co.], 1943(?); UMMZ
166383 (275, 25-59), Luxapallila Cr., 2 mi. W. Winfield, US Hwy.
43, [Marion Co.], May 4, 1952.

Tombigbee drainage, Mississippi, UMMZ 113872 (3, 42-52),
Ponta Cr., trib. Sucarnoochee Cr., 12 mi. S. Electric Mills, US Hwy.
45, Lauderdale Co., June 3, 1932.

Pascagoula drainage system, Mississippi, Covington Co.: CU 16261
(15, 27-50), trib. Okatoma Cr., 1.7 mi. E. Collins, June 14, 1949;
CU 12578 (43, 22-55), trib. Bowie Cr, 10.9 mi. W. Collins, August
7, 1948; Clarke Co.: UMMZ 157782 (11, 23-57), small creek from
US Hwy. 11 to mouth in Chickasawhay R, 1 mi. N. Enterprise,
August 17, 1939; Wayne Co.: (all distances from Waynesboro Post
Office, abbreviated W.P.O.) TU 8052 (2, 29-31), Patton Cr, 2 mi.
W.P.O, Feb. 1954; TU 8031 (15, 26-47), Meadow Cr, 2.5 mi.
W.P.O, Feb. 1954; TU 8047 (9, 21-44), Tatum Cr, 3 mi. W.P.O,
March 17, 1954; TU 8073 (4, 32-48), 3.5 mi. W.P.O, Feb. 24,
1954; TU 8038 (15, 25-46), Limestone Cr, 4 mi. W.P.O, March 18,
1954; TU 8056 (25, 34-47), 3-5 mi. W.P.O, Feb. 18, 1954; TU
8063 (8, 31-48), Limestone Cr, 4 mi. W.P.O, March 24, 1954; TU
8066 (5, 34-50), Taylor Cr, 1.5 mi. W.P.O, March 9, 1954.

Diagnosis. — A species of Notropis with 2, 4 — 4, 2 teeth and 7 anal
rays. Other fin rays: dorsal 8; pectoral 14, occasionally 15; pelvic 8.
Scales: predorsal rows 15, rarely 16; above lateral line to dorsal
origin 5 or 6; below lateral line to anal origin 4 or 3; in lateral line
36, sometimes 37; around body before dorsal 23 to 30, usually 24 to
28; around caudal peduncle 12, sometimes 13. Body moderately
deep and compressed. Dorsal origin directly above pelvic origin.
Mouth inclined, moderately large; upper jaw much longer than eye
length. Lateral line complete. Anterior lateral line scales not ele-
vated. Middorsal streak dark, continuous and encircles dorsal fin
base. Moderate to strong dark lateral band on body. Prominent dark
basicaudal spot. Size small, to 60 mm standard length. Apparently
allied to but sharply differentiated from Notropis chrosomus and
Notropis lutipinnis, both of which have 8 anal rays.

Description. — Besides the pertinent characters given above, counts
and measurements are given in Tables 1-3 and include comparisons
with Notropis chrosomus and Notropis lutipinnis. Measurements of
the holotype are detailed in Table 1 and its fin ray and scale counts
in Table 3 are the usual counts unless indicated otherwise by an
asterisk. Additional characters are illustrated in Figures 1-4. Other

74

Tulane Studies in Zoology

Vol. 2

TABLE 1.

Measurements of Notropis in Thousandths of Standard Length.

For each Character is given the Range of Variation and

Below (in Parentheses) the Mean.

Species

baileyi

lutipinnis

chrosomus

Stream system

Alabama

Chattahoochee

Alabama

Catalog number, CU

28224

10020

28226

17743

No. and sex of

Holotyj

ie Paratype*

specimens measured

5

5 $,4 $

2 $,S£

5 $.5 $

Standard length

60.3

42.5-62.9

43.5-54.5

42.0-62.9

Dorsal origin to

506

492-518

518-544

496-531

snout tip

(510)

(533)

(518)

Dorsal origin to

532

514-535

474-531

500-533

caudal base

(525)

(508)

(517)

Dorsal origin to

318

304-323

314 346

312-343

occiput

(315)

(330)

(332)

Pelvic insertion to

520

491-520

460-506

480-518

snout tip

(504)

(490)

(497)

Anal origin to caudal

331

328-350

262-370

312-355

base

(338)

(344)

(335)

Body depth

255

228-261

220-252

217-252

(247)

(237)

(233)

Body width

159

140 162

137-167

151-183

(152)

(151)

(169)

Dorsal origin to

149

146-165

148-167

133-157

lateral line

(158)

(156)

(146)

Pelvic insertion to

119

107-135

088-119

097-128

lateral line

(115)

(106)

(114)

Caudal peduncle

210

210-236

190-239

218-253

length

(225)

(220)

(234)

Caudal peduncle

106

097-116

097-107

087-102

depth

(104)

(101)

(096)

Head length

253

244-258

267-288

218-238

(251)

(275)

(228)

Head depth

168

156-172

166-178

156-171

(168)

(173)

(174)

Head width

143

127-145

120-144

122-141

(136)

(136)

(131)

Interorbital,

084

084-088

082-094

080-088

least fleshy

(085)

(089)

(083)

Snout length

074

074-083

085-093

068-078

(080)

(088)

(074)

Eye length

074

066-074

063-074

057-068

(070)

(067)

(062)

Upper jaw

086

083-090

090-105

065-080

length

(086)

(100)

(073)

Suborbital,

032

030-036

028-034

028 033

least width

(032)

(031)

(030)

Dorsal fin,

214

210-239

218-247

183-231

depressed length

(219)

(231)

(204)

Anal fin,

194

177 194

218-232

179-200

depressed length

(185)

(225)

(187)

Caudal fin length

243

243-294

239-275

231-271

from mid-base to tip

(269)

(257)

(249)

of longest ray

Pectoral fin

173

173 209

201-223

176-217

length

(186)

(214)

(196)

Pelvic fin

153

151-172

152-178

147-176

length

(160)

(168)

(161)

* The mean includes the measurement of the holotype.

No. 5 Suttkus and Raney: A New Cyprinid 75

descriptive features follow: The fusiform body is rather deep and
somewhat compressed. Among the black banded species of Notropis
with 2, 4 — 4, 2 teeth it most closely resembles lutipinnis but differs
from it in many characters including the anal ray count (8 in luti-
pinnis). The mouth is terminal; the snout is slightly blunt. The
head is subtriangular. The dorsum rises rather uniformly from the
region of the nostrils and levels off only slightly before the dorsal
fin origin. Posteriorly it descends gradually and ascends again just
before the procurrent caudal rays. The ventral outline anteriorly is
only slightly less elevated than the dorsal. It curves abruptly upward
beginning at the pelvic base and levels off midway on the caudal
peduncle, which is rather stout. The lateral line is complete; an-
teriorly it begins only slightly below the upper margin of opercular
opening and descends gradually to level off about one-third of the
distance to the dorsal origin. It continues horizontally, slightly below
the ventral edge of the dark lateral band, to a point below the pos-
terior dorsal fin base. Here it rises gradually and continuously
reaching the upper edge of the dark lateral band at about the narrow-
est part of the caudal peduncle only to descend again near the caudal
base. In lutipinnis the course of the lateral line is similar to that
of baileyi except that it does not dip as far ventrally on the anterior
half nor reach as far dorsally on the posterior half of the lateral band.
In chrosomus the lateral line is included within the lateral band for
its entire length, but in its anterior half is closer to the ventral border
of the band, while the reverse is true on the posterior half of the body.

In the following description of the lateral line canal and pores on
the head, the number of specimens counted is given in parentheses.
In counting pores the anteriormost is number one. In N. baileyi, the
supratemporal canal is always incomplete, with two pores on each
side (7) or with 3 on the left and 2 on the right (1). N. chrosomus
is essentially the same: incomplete, with 2 pores on each side (6)
or with 2 on the left and 3 on the right ( 1 ) . N. lutipinnis differs
from both in having the supratemporal usually complete with only
3 pores (4); one had an incomplete canal with 2 pores on each side.

The supraorbital canal is complete in the three species under con-
sideration. N. baileyi usually has a pore count of 8 (6) or 9 (2);
a line projected vertically from the posterior margin of the eye falls
between pores six and seven when the count is 8 and between seven
and eight when the count is 9. N. lutipinnis is similar with 8 (3)
or 9 (2); a line from the posterior margin of the eye falls on or
very close to the next to last pore. N. chrosomus differs in the small
sample counted in having 8(2),9(4),orl0(l). A vertical from
the posterior margin of the eye falls between pores six and seven
when the count is 8, between pores seven and eight when count is
9, and between eight and nine when 10.

The infraorbital canal is complete in baileyi and lutipinnis. In
baileyi the count is 15 (4), 14 (3), or 13 ( 1). A vertical projected
from the anterior margin of the eye falls between pores five and six

76

Tulane Studies in Zoology

Vol.

TABLE 2.
Anal Ray Counts in Three Species of Notropis

Species and drainage

6

7

8

9

Num-
ber

Mean

baileyi

Alabama R.

1

65

1

—

67

7.00

Black Warrior R.

—

137

18

—

155

7.12

Tombigbee R.

—

41

2

—

43

7.05

Pascagoula R.

—

58

—

—

58

7.00

Totals

1

301

21

—

323

7.06

lutipinnis

Savannah, Altamaha
and Chattahooche
rivers

1

166

7

174

8.03

chrosomus

Alabama R.

—

4

75

12

91

8.09

and four and five respectively. In lutipinnis, the count is lower, 14
(4) or 13 (1); a line from the anterior margin of the eye falls
between pores five and six. The infraorbital in chrosomus is com-
plete in four out of five specimens studied and shows the greatest
variation in pore count with 13 (3), 14 (1), 15 (1), or 16 (1).
One specimen with a pore count of 16 is incomplete with a break
between pores twelve and thirteen in the temporal region. A line
projected from the anterior margin of the eye falls between pores four
and five.

The preoperculomandibular canal is complete in the three species.
In baileyi, the pore count is usually 10 (4), but occasionally 9 (2)
or 11 (2). In lutipinnis the count is 9 (3), 8 (1), or 10 (1).
In chrosomus the number is 10 (4), 11 (1), or 12 (2). In chroso-
mus all pores, but especially those of the preoperculomandibular
series are much larger than in either baileyi or lutipinnis. Also luti-
pinnis usually differs from chrosomus in having the preoperculo-
mandibular pores more widely separated and has a particularly wide
space between pores four and five.

Dorsal fin straight on posterior border; when depressed the an-
terior rays exceed the posterior rays. Anal fin slightly falcate; when
depressed the posterior rays exceed the anterior rays.

The pharyngeal arch is fairly strong. The teeth in the main row
are pointed and hooked at the tip. The serrulate grinding surface is
well developed especially on the middle two teeth. The teeth in

No. 5

Suttkus and Raney: A New Cyprinid

11

Figures 1-4. 1. (top) Notropis baileiji, paratype, adult female, 56
mm in standard length, from the type locality, a trib. Sawacklahat-
chee Cr., 1.7 mi. W. Society Hill, Tallapoosa River drainage, Macon
Co., Alabama; 2. (second) Notropis baileyi, adult male, 48 mm from
a trib. Black Warrior River at east limits of Cottondale, Tuscaloosa
Co., Alabama; 3. (third) Notropis lutipinnis, a mature female, 58
mm from Bushy Creek, 9 mi. S. Wrens, Savannah River drainage,
Jefferson Co., Georgia; 4. (bottom) Notropis chrosomus,
female, 58 mm from a trib. Oostanaula River, 6.3 mi. S.W.
Coosa River drainage, Gordon Co., Georgia. (Photographs
lass M. Payne)

a mature

Calhoun,

by Doug-

No. 5 Suttkus and Raney: A Neiv Cyprinid 79

the lesser row are about half as high as those in main row which are
pointed and hooked slightly; the grinding surface is moderately de-
veloped. The gill rakers on the anterior gill arch number 4 to 6.
They are short subtriangular in shape and only the two near the
center of the arch are pointed. The nape is scaled. The breast is
scaleless forward of a transverse line between the posterior margin
of the pelvic bases.

As has been noted repeatedly for other species of Notropis slight
but consistent color or pattern differences are present. Most of these
are included in Table 3. Typically, bailey i is a black banded species.
The band is prominent, is intensified posteriorly and if at all is only
weakly marked by more intense spots above and below each lateral
line pore. It continues on the head behind the eye but is less intense
immediately before the eye and fades considerably on the anterior
part of the snout. A sharp light stripe borders the dark band above.
A prominent and diagnostic feature, except in some specimens from
the Black Warrior system, is the series of dark ovals located on the
first full scale row below the dark lateral band ( fig. 1 ) . A prominent
quadrate basicaudal blotch is continuous with and somewhat darker
than the band immediately in front. The blotch continues on to
the 4 to 6 median caudal rays a distance about equal to the length of
the spot on the caudal peduncle. The dark edging on the midcaudal
rays continues to the fork but is weaker posteriorly. The prominent
middorsal dark streak is continuous, although not intensified, around
the base of the dorsal fin and although present behind the dorsal
base is only about two-thirds as wide. It narrows somewhat about
half way between the occiput and dorsal origin. The midventral
dark band on the caudal peduncle is limited to the area at the base
and immediately behind the depressed anal fin. The top of the head
is dark. A light preorbital bar extends forward to include the area
about the nostrils. A wedge-shaped dark blotch lies above the nos-
trils. A diamond-shaped dark bar crosses the head just anterior to
the hind margin of the eyes. The heart-shaped mark on the rear of
the head is more or less completely divided by a transverse light line
and is outlined on its posterior lateral margin by another light bar.
A dark subtriangular blotch overlays the temporal region anterior to
a dark transverse occipital bar which continues laterally and down-
ward behind the opercle to the pectoral fin base. The lips have
scattered melanophores. The chin is peppered and has a lighter area
near the center. The upper oral breathing valve bears a few mel-
anophores.

The fins are generally light. The first pectoral ray is darkish and
the second ray bears a few melanophores. The caudal fin rays,
especially those in the center, are finely margined with black. The
dorsal rays are margined with black and there are scattered mel-
anophores on the membranes near the base of the fin. The anal and
pelvic fins are white.

The dark lines which margin the scale pockets although not pro-

80

Tulane Studies in Zoology

Vol. 2

©
o

CO

•2
g.

o
*~

©
fa

o

05

fa

o
fa
S
w w

J fa

« i
s «

fa
o

o

CO

SI

Hi

s

o
O

•2

u

d>

•5

03

01

o3

>

JS

0>

u

T3

c/J

X

-u

3

O

BS°

o

OJ

*3

T3

0)

3

O 73

,«H

O

bo-*3 o3
.g+»43

•r- p d)

^ ^ H

3

CT3

£ h
O o3

ovS-a

n

-3

73

O)

C

ShJS

oi

•PH

■JJ

-3

-M

0 +?

o

eg

0)

s

bt

1*1

>.

3

co

o

43
hi)

<H

01

•—t

M

■— 1

VI

01

3

"3

«

2~
O

£

o

42

'a

3

3
u

CO

o

a
a>

O)

03

T3

_S

43

o>

43
03

-3
3
o>
+->
X

H

o>
X
o
03
o>

Pi

o

-t->

43

bo

o
55

43
60

T3

o>

OS
u

M

o>

&

o3

CO

o>

4 J. 3>

- 3 60

£ S fe

s o> o>

«4H -

C Bj 03

S3 0J«H >>

w ^ 01

o

a

-t-> «*H

c

.

0

eg

G

«H

<D

O

a

O*

M

r/l

-u

01

(/J

43

i — i

+j

CO

a

o

a

_ !

"3

s3

-3

»i

2

o

x

c

o

a.

>.

o

CJ

3

0)

Q S

S

Ol

c
o

a*

C3
U
-t->

c
o>
>

-a
o>

>

%

o»
co

>
co

S

o]
43 0>

■a-S

cBf- l

. --. oO

4«!
O

oj

C

o

4rf
u
03

s

o>
a,

03
co -o

4J 03

U Oi
03 43
0)

W o

4^
o>

Ol

a
a

s

T3
0>

03
Ol

o>
o,
o

a)
&,

o
o>

T3
0)
CU

o>

U
Oi
-l->
03

U
03
T3 0)

«H "

w o>

f-l
o>

o

S3

o
o

-a

h S3
&43

»J 03

c

0)
CO

43

<1

c

0)
co

C
a>
co
Ol

-a
_c

43
Ol
43

u

03
43

U
03

s^

3^
O o

6 w

s >>

.03 03

S^
o

P Ol

,a> to
«H c3

43^

tf +»

•$«

CO

43 U

boo

•^43

9,
Ol
Ol
T3

03 m
S S

l§

^ 03

43^

60 as

3 oi

i—l co

03

3
03

03

S
oi

2^

W c3

03

34*!
03^

43430

4rf

03

Q

CO

o

Oh

13
S3
03

OJ s

-4-> o3

■a I

03 f-i

C5

B

c

03

01

CO

03
43
-u

3
o
43

03

>>
T3
O

M

60^
3 "

o

03

a

CO

-2

<1

Oi ^3
e ™
ST3

9 3
co pj

53 O

03 -w

C0T3
P»> 01

03 C

!h Ol

8|.

01

OJ S

■S o

03 co

03 C
3 03

O* _
43 43

s .-a

CO S

3
Oi

CO

01

03 O

u o

OJ

h 3
03 03
T3 43

O
O,
CO

ro3

3
03
01

• 1— »
co
03

PQ

No. 5 Suttkus and Raney: A New Cyprinid 81

C S g 1

H3 W Q>

^ o rt

■ 1

-^ tO t>

o3 -^ 5 —

03

cU
'to

#

£

o O d "3

to sv^ e

00

©

00
1

■*j ••"_ cu

3
o
o

00

lO

> cu 3 cu
^ jj 03 U >>

U

to

co

-2 ° +3

03 g-i> 'ft

n o o o S?

co

-t-»

-(J

+->

c3

Fh

r

u

h

CU

o

©

o

> CJ2 3 £

!> 03 o3;3 Xfl

>H

.3

X

J3

o

co

CO

CO

03
CJ

g.3

— a

oo eo

t-

CO CM

l— *

1 1

C5

>>

CO

t~ CM

00 <M

^

1— H

'3

>>

>> >

1—1

r- -* *—*

io

03

3

S3
to

s
_o

•>*

sual
sual

T—j

'to

03

'to

3

3 3

03

W

03

cu

o

cu

O

cj
o

o

CJ

o

o

oo" -#

eo cm
i i

■*

^

co~

„

^

CO CM

00 «H

00

i-H

m

CO

CO CM

tO o

c^

o cu

S F^ CU

•~12 s

,„ 3 o

2 « cu "2

.2 > 3 .5

O — js

5 to *h .-«

*H & jj CU

35 -.B 3 £

*3 — v5>

cu « 5s"

££ U -hi CO

s* cu

co

CU ,-(

.3
03

#

to

cu

CO

o

to

1

1

to
to

3
O

lO

o

cu

CJ

00

OS

1—1

,3

CJ

3

Ofl

§

c

03
+3

OX)
3
o

&0

3
o

1-3

00

00

o r* co £: . eo

5— _cu

■° w b So -2 .2 .2 ^ ^ «, " .2 3 2

. C CU T3 T3

C r— I "" tOtO jo

cu g c d « 3 03

* 00

c- <^

co 4

>,w

co

03 ~

.— i

*

11

to to

>> co

CO

, rare
often

4>
+->

o

, occa
-30, u

lO .

^

co co

rH U3

•^

CO CM

<M

1

o

CO

•o

rH

eo

tH

1—1

1

00

i

T)<

>,

CO

co

CM

r- 1

~

03

>>

>>

>>

>

o

en

3
to

3

1-^

03

3

W5

03

3

3

13

3

'm

>>

1— H

•*

CO

to

to

to

S

T-(

3

C
CU

3

3

3

CJ

8

03

CM

O

oo"

of

r-T

t>

kO

h

CO

o

o

eo

t

t-

CO
1

CO

T— 1

— 1

oo"

1—1

co

tjT

eo

CM

l-H

"J * q JS5 ±> 2 2

5ig 8 S IIS J ^ ^ ? 5? s

5*9.5 S J7? C.Ss c ^r »o - -cococm"

cu _c

0 TJ be , S3 „ aj

03 +» CU r1 m r^ 15

^ .s § "g § al a g ? §

C>» •'-03,OK1 03^"T3CU

SiH -3 cu ^ B 3 to "So «,-5 >» S ^ 3 cu * 13

^^~ ft03^_i _, J3 i— irn •!-• -_7 'i-i M »( ,

0

^.•C "5 ft ft^.rS« -S to 3 ^ jS _

p»»

•H S S'H ih ^,CU+J CJ fHEjjQiM^HW^

72 ^ Ow-u _, ^3 ■— i 03 •;-. .rt-p1 O—03

|t ^s ill s|.|°55g0 B- S s-g^-g

£!r3 +3 f.Ta . » 'tj b OX

£ -ft to S0^^^ ^'-i S° 'H

II P |i|fi!l!.^«!^sil-s^

Qcs Os-DiPqjsffiiiMo^cjft, co

82 Tulane Studies in Zoology Vol. 2

nounced cross the light stripe which lies above and parallels the dark
lateral band of the midside. The scale centers are somewhat lighter
than their borders. The breast and belly are white; occasionally the
dark of the peritoneum shows through.

In the field males from Black Warrior system taken on March
31 were observed to have a wide whitish or lead colored line along
the side. C. Richard Robins notes that specimens taken on August
12 had a metallic lateral band and that the caudal fin and the pectoral
fin base were orange. At breeding time it is brightly colored with
shades of red through yellow. Detailed color notes on specimens
contained in UMMZ 105554, 146533 and 166383 were furnished by
Reeve M. Bailey. Specimens in UMMZ 105554 had a deep-lying
light orange area above the steel blue-black lateral band which graded
to yellow over the back. The orange and the black were separated
superficially by a leaden-white line. The lateral band was also
bordered on its lower side by a silver white streak which petered out
toward the caudal but was broadened on the trunk and anteriorly
graded into the silvery side of the face. The mid-dorsal stripe and
scale borders above were deep moss-green. There was considerable
orange-red on the lower sides posteriorly. Bright red colored the
axil of the pectoral fin but not about the pelvics. A red line
bordered the dorsal margin of the lateral band on the side of the
head; was interrupted by the eye but reappeared as a watery rose
stripe around the snout. There was a wash of rose color on the fins
which was strongest on the areas immediately above and below the
basicaudal spot. Specimens in UMMZ 146533 were essentially the
same in color as above, however the sides anteriorly were bright red
which shaded into orange posteriorly. The dorso-lateral white line
was broadened anteriorly and obscured the lateral band. The red on
the sides faded dorsally but was strong on the breast, about the base
of the pelvics, and down to and posterior to the anal fin. The pecto-
ral axil was deep red as were the markings behind the eye and on
the snout. The body and fins of the males in UMMZ 166383 were
bright yellow instead of washed with red. The females lacked the
bright yellow on the body and fins. The sex products were easily
expressed from the breeding individuals.

Nuptial tubercles are well developed in both sexes. In the male
the top and sides of the head bear fine or moderate sized scattered
tubercles with rather coarse ones prominently interspersed. Both
types of tubercles are present on the opercle but only fine ones
appear on the cheek. Fine tubercles cover the underside of the head.
Large pointed ones occur on the lower jaw. Scattered fine tubercles
with a few interspersed large pointed ones are found on the snout.
N. lutipinnis differs in having only fine or moderately developed
tubercles on these parts. N. chrosomus has large coarse tubercles
scattered over the top of the head and snout, with a well developed
row on edge of the latter; scattered large tubercles and numerous
small ones on the opercles, cheek and preorbital area; and moderate

No. 5 Suttkus and Raney: A New Cyprinid 83

tubercles scattered on the dentary and underside of head. N. baileyi
has small tubercles on the edge of the scales and has 2 to 3 larger
ones in vertical row near the scale base. On the ventral scales there
are fewer tubercles posteriorly. In lutipinnis the scales are densely
tuberculate, with large ones on the margin. Tubercles are common
posteriorly and on the underside of the body. In chrosomus there
is a prominent row on the posterior margin of each scale. Those on
the second scale row below the lateral line are not as well developed.
Those on upper and lower pectoral rays are in one or two rows. In
lutipinnis a row of sharp tubercles line the rays on the upper side
of the pectoral fin. N. chrosomus is like lutipinnis but lacks tuber-
cles on the first ray. The pelvic and anal fins of baileyi and lutipinnis
have a row on the upper side of the rays and on either side of the
dorsal rays while these are absent in chrosomus. Tubercles are
abundant along the caudal rays of lutipinnis but are absent in baileyi
and chrosomus.

Sexual dimorphism. — In males of baileyi, the pelvic fins are longer
and nearly always reach to the origin of the anal fin, and in the
exceptions always extend posterior to the urogenital elevation. In
females the shorter pelvic fins barely, if at all, extend beyond the
distal end of urogenital elevation. In males the pectoral fins are
broader than in females. However they are not as broad propor-
tionally as in males of lutipinnis. In females the pectoral fins are
narrower and have more acute distal tips. In females the nuptial
tubercles are less developed than in males although some tubercles
on the head are nearly as large. There are none or only a few small
tubercles on the dorsal and lateral scales. Transverse rows of fine
tubercles line the belly scales. The head is finely tuberculate, with
a scattering of large tubercles.

In males of lutipinnis the pelvic fin nearly always reaches the
origin of the anal fin and commonly extends well beyond, whereas
in the female the pelvic fin may reach the posterior edge of the
urogenital elevation, but seldom if ever to the origin of the anal fin.
In lutipinnis the pectoral fin in the male has a rounded distal tip
and is much broader than in the female. Nuptial tubercles are less
developed in the female; some are found on the margins of dorsal
and lateral scales; none or only a few occur on the lower sides and
belly, and fine tubercles are scattered on the head. Some may be
seen on parts of all fins.

The sexual dimorphism is similar in chrosomus. The pelvic fins
are longer in the male and extend posteriorly beyond the urinogenital
elevation and often nearly to the origin of the anal fin, whereas in
the female the pelvics do not reach the former point. The pectoral
fin of the male is longer and more pointed. In females some large
tubercles are scattered over the head and anterior body, but are lack-
ing on the fins. Both sexes are brightly colored with red and orange
at spawning time. This is one of the most attractive species of
Notropis.

84 Tulane Studies in Zoology Vol. 2

Infraspecific variation. — The dark ovals typically found immedi-
ately below the dark lateral band are missing in adult specimens
taken in the Black Warrior River system (fig. 2). They also show
a higher average number of circumferential scales (26 to 30, usually
26 to 28) while those from the Pascagoula and Alabama River systems
range from 23 to 27, usually 24 to 26. Black Warrior specimens
also have larger and coarser nuptial tubercles.

Relationships. — Among the species of Notropis with 2, 4 — 4, 2
teeth and a dark lateral band, baileyi is similar to and probably re-
lated to lutipinnis which is not found in the same drainage. It is
also related to chrosomus which is found in the upper reaches of
the Alabama River system and which was taken in the same collec-
tion with baileyi. N. baileyi differs trenchantly from both in the
anal ray count, which is 8 in lutipinnis and chrosomus, and in many
details of pattern, proportions, fin ray and scale counts given for the
three species in comparative form in Tables 1 and 3-

On re-examination the types of N. norrisi Fowler (1945: 235, figs.
114 and 116), which were said in the original description to resemble
lutipinnis, prove to be Hybopsis harperi (Fowler) the types of
which were also studied. The teeth are 4 — 4 and the anal rays 8 in
the type, ANSP 70705, and in 5 paratypes, ANSP 70706-10. Ernest
A. Lachner has pointed out that the types of norrisi which are listed
as being collected from near Fitzgerald, Georgia, are more likely to
have been taken from the Suwannee River system than from the
"basin of the Altamaha River". This seems certainly to be true if
taken in or near Fitzgerald where the drainage system is to the south
via the Alapaha River system. A re-examination of the types of
N. williami Fowler (1945: 236, figs. 117, 118, 119) shows this to
be N. peter soni Fowler (types also examined), as was surmised by
Hubbs and Raney (1951: 9). The teeth of williami on re-examina-
tion are 2, 4—4, 2 in the "type", ANSP 70702, not 1, 4 — 4, 1 as
given in the original description. Fin ray counts on the three types
of williami are anal 7, dorsal 8, pectorals 13 and pelvics 8 (7 on one
side of one specimen). Scale counts on the type and largest para-
type follow: predorsal rows 16, 15; scales around body before dorsal
fin 23, 22; scales around caudal peduncle 12 and lateral line scales 35.

N. baileyi may be separated from petersoni, xaenocephalus, roseus
and an undescribed species from the Mobile Bay drainage, all of
which have a dark lateral band, 7 anal rays and 2, 4 — 4, 2 teeth by
the following characters: 1. middorsal stripe dark and continuous,
encircling the dorsal fin base and extending as far as the procurrent
caudal rays, rather than being poorly developed or absent posteriorly,
not encircling the dorsal base; 2. mouth large, with the upper jaw
much longer than the eye, rather than of equal length or smaller;
3. dark lateral band immediately behind the eye continuous and of
the same intensity and width as the band on the opercle, rather than
being broken immediately behind eye and represented by scattered
melanophores; 4. dark lateral band on body fairly uniform in density

No. 5

Suttkus and Raney: A New Cyprinid

85

throughout, with the lower margin entire, rather than being of vary-
ing density, especially on the anterior half, and with small spots
associated with the lateral line pores. As determined from an
examination of the types, baileyi is not the "wide mouth" variety of
"Notropis xaenocephalus" of Jordan (1877: 335).

Distribution and ecology. — Upper Coastal Plain and nearby Pied-

90

68

86

84

Map 1. Distribution of Notropis baileyi. (Circle indicates type

locality)

mont from the Leaf and Chickasawhay rivers of the Pascagoula drain-
age of Mississippi eastward through the Mobile Bay drainage of
Alabama where it has been taken in the Tombigbee, Black Warrior,
and Alabama rivers; in the Alabama basin it extends eastward as far
as the Tallapoosa River (Map 1). A single specimen from the
upper Piedmont, 5 mi. W. of Attalla in the Coosa River system
indicates the possibility of scattered populations throughout the
Piedmont.

All of our material comes from small wooded streams which were
seldom more than 15 feet and often only 5 to 10 feet wide with a
moderate flow. Bottom was almost always sand and gravel. The
water was white where noted and vegetation was usually absent.

References Cited

Fowler, Henry W. 1945. A study of the fishes of the southern Pied-
mont and Coastal Plain. Monogr. Acad. Nat. Sci. Phila., 7: i-vi,
1-408, figs. 1-313.

86 Tulane Studies in Zoology Vol. 2

Jordan, David S. 1877. A partial synopsis of the fishes of upper
Georgia. Ann. N. Y. Lyceum Nat. Hist., 9: 307-377.

Hubbs, Carl L. and Karl F. Lagler 1947. (and 2nd printing, 1949)
Fishes of the Great Lakes Region. Bull. Cranbrook Inst. Sci.,
26: i-xi, 1-186, many figs.

Hubbs, Carl L. and Edward C. Raney 1951. Status, subspecies and
variations of Notropis cummin gsae, a cyprinid fish of the south-
eastern United States. Occ. Pap. Mus. Zool. Univ. Mich., No.
535: 1-25.

Y,4-/W v'

^- &■ 7> S

;J

Volume 2, Number 6

February 25, 1955

A RECONSIDERATION OF THE RACER, COLUBER
CONSTRICTOR, IN EASTERN UNITED STATES

WALTER AUFFENBERG,

DEPARTMENT OF BIOLOGY, UNIVERSITY OF FLORIDA,
GAINESYrLLE,-FL0R7DT~i

Ma?.*. '3
NBKISfTY

TULANE UNIVERSITY
NEW ORLEANS

TULANE STUDIES IN ZOOLOGY is devoted primarily to the
zoology of the area bordering the Gulf of Mexico and the Caribbean
Sea. Each number is issued separately and deals with an individual
study. As volumes are completed, title pages and tables of contents
are distributed to institutions exchanging the entire series.

Manuscripts submitted for publication are evaluated by the editor
and by an editorial committee selected for each paper. Contributors
need not be members of the Tulane University faculty.

EDITORIAL COMMITTEE FOR THIS NUMBER

Ernst Mayr, Professor of Zoology, Harvard University, Cam-
bridge, Massachusetts

Karl P. Schmidt, Chief Curator of Zoology, Chicago Natural
History Museum, Chicago, Illinois

Fred R. Cagle, Professor of Zoology, Tulane University, New
Orleans, Louisiana

Manuscripts should be submitted on good paper, as original type-
written copy, double-spaced, and carefully corrected.

Separate numbers may be purchased by individuals, but subscriptions
are not accepted. Authors may obtain copies for personal use at cost.
Address all communications concerning exchanges, manuscripts, edi-
torial matters, and orders for individual numbers to the editor. Re-
mittances should be made payable to Tulane University.

When citing this series authors are requested to use the following
abbreviations: Tulane Stud. Zoo I.

Price for this number: $1.00.

Assistants to the Editor:
Carol L. Freret
Donald W. Tinkle

George Henry Penn, Editor,

c/o Department of Zoology,
Tulane University,
New Orleans, U. S. A.

; «A« 8 1955

A RECONSIDERATION OF THE RACER, COLUBER
CONSTRICTOR, IN EASTERN UNITED STATES

WALTER AUFFENBERG.i

Department of Biology, University of Florida,

Gainesville, Florida

A quarter of a century has passed since the publication of Orten-
burger's (1928) monograph on the whipsnakes and racers. During
the intervening years such factors as effect generic, specific and racial
validity, distributional problems and the evaluation of diagnostic char-
acters in these snakes have been quite generally discussed in many
short papers. Yet, the taxonomic status and relationships of the
various subspecies are still incompletely known. This seems especially
true of those forms which inhabit the coastal areas of southern United
States. Descriptions have been based on characters present in a few
specimens at hand, with little knowledge of the extent of variation,
either individual or geographic.

This study is primarily concerned with the morphological variations
existing in the eastern forms. Notes and comments are presented
on more western populations when they contribute to a better under-
standing of the variations which are found in the East.

In order to understand individual and geographic variation of south-
eastern forms practically all of the available material from the area
was studied. Thus a total of 1,560 specimens from both institutional
and private collections was examined. Since Florida was to be a
focal point in the study of variation, many more specimens were
examined from this state than any other (1,008). Of the remaining
552, 491 were from east of the Great Plains, and over two-thirds of
these from the southeastern states.

Efforts were made to examine as many living specimens as pos-
sible, and 377 became available for study. The results were especially
worthwhile for without live animals knowledge of certain colors
(which fade rapidly in preservatives) would not have been available.

Throughout the course of this paper the following abbreviations
are used when reference is made to specimens or to collections:
AMNH = American Museum of Natural History, New York, New
York; ANSP = Academy of Natural Sciences of Philadelphia, Phila-
delphia, Pennsylvania; CHM = Charleston Museum, Charleston, South
Carolina; CM = Carnegie Museum, Pittsburgh, Pennsylvania; CNHM
= Chicago Natural History Museum, Chicago, Illinois; UF = Depart-
ment of Biology, University of Florida, Gainesville, Florida; JWC =
John W. Crenshaw Collection, Gainesville, Florida; MCZ = Museum
of Comparative Zoology, Harvard University, Cambridge, Massachu-
setts; RA-WTN = Ross Allen-Wilfred T. Neill Collection, Silver
Springs, Florida; ST = Sam Telford Collection, Winter Haven, Flor-

1 A thesis presented to the Graduate Council of the University of
Florida in partial fulfillment of the requirements for the Degree of
Master of Science.

90 Tulane Studies in Zoology Vol. 2

ida; T = University of Texas, Austin, Texas; TU = Tulane Univer-
sity, New Orleans, Louisiana; UMMZ = University of Michigan
Museum of Zoology, Ann Arbor, Michigan; WA = Walter Auffen-
berg Collection, Gainesville, Florida.

Taxonomy
There are two major groups included within the genus Coluber,
the whipsnakes and racers. As now recognized there are eight sub-
species of the racer, Coluber constrictor? Together they are dis-
tributed throughout the United States, from coast to coast, and from
the Gulf of Mexico and Guatemala northward into southern Canada.
Although widely distributed, there are probably no true desert forms.
However, some populations occur in mountainous areas, between, or
even surrounded by truly arid conditions. They do not seem to
occur on the tops of high mountains.3 The range of Coluber con-
strictor mormon extends from the Pacific Coast to, and including
most of, the Rocky Mountains; Coluber c. flaviventris is found in
the Great Plains, from the eastern slopes of the Rocky Mountains to
the forested provinces of eastern North America; Coluber c. stejnegeri-
anus occurs from Guatemala northward to Nueces County, Texas,
along the Gulf Coast; while Coluber c. paludicolus is found in the
Everglades of southern Florida and on Cape Canaveral, Brevard
County, Florida. These four subspecies are light in color. They may
vary from green through brown to bluish-gray. The belly is always
light. Another form, Coluber c. antbicus, of western Louisiana and
eastern Texas, is spotted with yellow, buff, or bluish, on a blue,
black, or tan ground color. The remaining three races are slaty-black
above and below. The light-colored chin contrasts with the dark
belly or throat. These are distributed throughout eastern North
America. Coluber c. priapus is found in the Coastal Plain from North
Carolina southward throughout most of Florida, westward to the
Mississippi River, and thence northward to southern Illinois and
Indiana along the Mississippi and Ohio River Valleys. Coluber c.
constrictor is found from the eastern edge of the Mississippi Valley
to the Atlantic Coast in the northeast, and to the Fall Line in the
southeast. In the northeastern parts of its range it is found as far
northward as Halifax, Canada. Coluber c. helvigularis subsp. nov.
is found in the area of the Chipola and Appalachicola River Valleys
of western Florida (fig. 1).

2 Etheridge (1952) has shown that the supposedly diagnostic char-
acters of Coluber ortenburgeri fall within the variation of Coluber c.
stejnegerianus, and that the former should be regarded "at most" a
subspecies of constrictor. In view of the lack of any diagnostic char-
acter which would serve to separate the two forms I see no reason for
retaining ortenburgeri as a nominal' subspecies. The name should be
referred to the synonymy of Coluber c. stejnegerianus. Coluber c.
haasti of Bell (1952) is not regarded as a valid form for reasons to
be given later in this paper.

3 Brimley (1942) stated that C. c. constrictor is not found above
4,000 feet in North Carolina. Linsdale (1940) stated that C. c.
mormon is found between 4,000 and 6,000 feet elevation in Nevada.

No. 6

Auffenberg: Coluber constrictor

91

Figure 1. Distribution of the subspecies of Coluber constrictor.

92 Tulane Studies in Zoology Vol. 2

All of these races are known to intergrade with one another at the
mutual borders of their ranges with the exception of C. c. mormon
and C. c. flaviventris. Intermediate populations between these two
forms probably occur, although as yet none have been reported.
Intergradation (if it does exist) is to be expected in Montana, where
a record for C, c. mormon brings this racer's range within 130 miles
of that of C. c. flaviventris (Stuart, 1930:44). Intermediate popu-
lations will probably also be found to occur throughout, or near, the
Front Ranges of the Rocky Mountains.

Besides these eight recognized races there are several populations
of varying sizes which, judging from their morphological characters,,
may be incipient subspecies. Areas in which these populations occur
are listed here, although they will be discussed later in the paper.
(1) The Outer, or Lower Florida Keys: speckled individuals reminis-
cent of intergrade specimens between C. c. priapus and C. c. anthicus
are available from certain of these islands. However, they coexist
with typical, uniformly black specimens and probably represent noth-
ing more than mutant individuals. (2) The Mississippi Delta and
the associated mainland: C. c, flaviventris from this area seems to be
rather distinct from populations throughout the remainder of the
range of this subspecies. These differences mainly concern coloration
and scutellation. (3) Eastern Steppe Province: populations of C. c.
flaviventris from the states of Ohio, Michigan, Indiana and Illinois
differ from populations of the same subspecies from the Great Plains.
The name Coluber c. foxi Baird and Girard, has recently been revived
for this population (Schmidt, 1953:187). However, from the speci-
mens which I have examined I am forced to conclude at the present
time that a gradual cline exists between the Great Plains and Eastern
Steppe populations. Further work should be conducted in these
areas, especially Missouri and Iowa, to determine the nature and
extent of the area of intergradation and degree of homogeneity of
characters in the regions pertinent to an understanding of the situ-
ation. Until this is done it may be more advantageous to retain
C. c. foxi in the synonomy of flaviventris. This interpretation has
been followed throughout the present paper. (4) St. Johns River
Valley, Florida: specimens of C. c. priapus from Seminole County
to Lake Okeechobee along the lowlands of the St. Johns River are
intermediate between C. c. priapus and C. c. paludicolus. This popu-
lation is homogeneous and can be separated from either priapus or
paludicolus on the basis of the amount of white on the ventral sur-
face. However, they cannot be distinguished from intergrades be-
tween priapus and paludicolus from other areas where ranges of the
two subspecies meet. Preliminary investigation suggests that the St.
Johns River lowlands receive considerable gene-flow from Everglades
populations in other species of snakes as well, e.g. Liodytes alleni,
Thamnophis sauritus. Lampropeltis getidus, L, doliata, and Diadophis
punctatus. The two areas are separated by a very low stream divide,
and the ecological characters of both provinces are quite similar.

No. 6 Auffenberg: Coluber constrictor 93

The condition seems somewhat comparable to that between the
flaviventris — joxi — constrictor complex, and in this paper the lower
St. Johns River specimens are considered as representing an inter-
grade population. However, further study may show that they deserve
nominal status. At the present time no real advantage would be
gained by recognizing these specimens as a separate race, for the
implication would be that they had attained an evolutionary level
equal to that of the other accepted subspecies in North America. In
view of our present knowledge this hypothesis seems untenable.

Key to Adult Specimens

1. Dorsal surface and most of ventral sur-

face slaty black. 2

Dorsal surface lighter, being greenish,
olive, brownish, bluish, or spotted.
Ventral surface light, being whitish,
yellow or light slaty blue. 4

2. Enlarged basal hemipenial spines up to

2y2 times the length of the adjacent
proximal spines; in large adults
there is usually very little white on
the chin and throat. The suprala-
bials are usually almost completely
black in large adults. The iris is al-
ways brown or dark amber, never

bright red or orange Coluber constrictor constrictor

Enlarged basal hemipenial spines at
least 3 times the length of the adja-
cent proximal spines, and are usually
4 times the length of these spines;
in large adults from peninsular
Florida there is usually a consider-
able amount of white on the chin
and throat; the supralabials are usu-
ally half white and half black in
specimens from peninsular Florida,
darker on the Florida Keys and
throughout the remainder of the
range. The iris is usually brown,
but may be amber, red, or even
orange in specimens from central
peninsular Florida. '. . 3

3. Chin light tan or brown, or mottled with

light tan or brown; never all white
or blotched with black. Supralabials
and infralabials always with a light
tan or brownish suffusion. Iris in
adults always brown, never bright

red or orange. Coluber constrictor helvigularis

Chin white, or mottled with black, never
light brown or tan. Supralabials and
infralabials white, black, or with
both colors present, never distinctly
light tan or brown except near the
area of intergradation with paludi-
colus, or in isolated populations in
western peninsular Florida. Iris
brown, red, amber, or orange. Coluber constrictor priapus

94 Tulane Studies in Zoology Vol. 2

4. Dorsal surface spotted with white, yel-

low, buff or dark blue. Coluber constrictor' anthicus

Dorsal surface not spotted, unicolor. _. 5

5. Usually 7 supralabials, average adults

over three feet. 6

Usually 8 supralabials, average adults

under three feet. 7

6. Subcaudals less than 99; ventral color

usually some shade of yellow, but
may be grayish, bluish-white or
yellowish- white; iris usually brown,
but may be reddish in specimens
from southeastern Louisiana; prac-
tically never any darker clouding on
the belly; 1st supral'abial rarely in
contact with the loreal. Coluber constrictor flaviventris

Subcaudals more than 99; ventral color
whitish, never distinctly yellow; iris
always reddish, amber or orange,
rarely brown; usually with light
brownish-gray or powder blue cloudy
markings on the belly; 1st suprala-
bial commonly in contact with the
loreal. Coluber constrictor paludicolus

7. Infralabials usually eight; dorsal color

greenish. Coluber constrictor stejnegerianus

Infralabials usually nine; dorsal color

usually olive. Coluber constrictor mormon.

For the most part this study has concerned itself with the variations
existing in the four eastern races of Coluber constrictor: constrictor,
priapus, paludicolus, and helvigularis (new subspecies). Variation in
the portion of the range of the race flaviventris which is found east
of the Mississippi river is discussed in somewhat less detail. An
abbreviated synonymy and description of the four subspecies men-
tioned above follow:

Coluber constrictor constrictor Linnaeus

Northern Black Racer

Coluber constrictor (part) Linnaeus, 1758: 385 ("Habitat in America

septentrionale").
Bascanion constrictor Baird and Girard, 1853: 93.
Bascanion foxii (part) Baird and Girard, 1853: 96.
Coryphodon constrictor (part) Dumeril, Bibron and Dumeril, 1854:

183.
Masticophis flagelliformis testaceous (part) Jan, 1863: 65.
Coryphodon constrictor vertustus (part) Jan, 1867: fig. 3-
Bascanion flaviventris (part) Dumeril and Bocourt, 1870: 697.
Bascanion constrictor flaviventris (part) Yarrow and Henshaw, 1879,

App. L. of App. N. N.
Zamenis constrictor (part) Boulenger, 1893: 387.
Zamenis constrictor flaviventris (part) Ditmars, 1907: 285.
Zamenis constrictor constrictor (part) Ellis and Henderson, 1913: 103-

No. 6 Aujjenberg: Coluber constrictor 95

Coluber constrictor constrictor (part) Stejneger and Barbour, 1917:

79.
Coluber constrictor flaviventris (part) Ortenburger, 1928: 175.

Diagnosis. — A subspecies of Coluber constrictor in which the ven-
tral and dorsal coloration of the adults is slaty black; the enlarged
hemipenial spine is less than three times the length of the adjacent
proximal spine and the eyes are always brown or dark amber.

Holotype. — None assigned, but the type locality is probably in the
vicinity of Philadelphia (Dunn and Wood, 1939).

Description. — Dorsal surface in adults uniform slaty black, fading
to dark olive gray where it integrades with flaviventris. Scales with
a satiny luster. Top and sides of head slaty black like the remainder
of the body and tail. In full grown specimens the supralabials prac-
tically all black, but in some adults, or in smaller specimens, black
above and white below. Ventral coloration dark slaty black to dark
gray. Chin usually white and in large specimens more or less
spotted with black, sometimes practically all black. Iris usually dark
brown or dark amber.

Juvenile specimens possess a distinct pattern until after the first
year. Ground color slaty gray or brownish-gray. There are 48 to 71
saddle-shaped blotches of dark gray, dark brown or reddish-brown,
edged with darker gray or black, arranged in a single series down
the back, and extending down the sides to the second or third scale
rows.

On the side of the body, alternating with the dorsal saddles, are
faint darker spots one to two scales long and wide. They may occur
singly or in pairs, one spot being above the other. Small black
punctations may occur over the same area.

On the anterior half to two-thirds of the belly there are many small
semicircular or crescent-like gray, black, or brownish spots, usually
two on each scute, which are usually found on the lateral edges of the
ventral plates.

Supralabials cream or white, with a dark brown or black edge on
the posterior margin of each plate. Infralabials white or cream, and
unspotted.

Specimens in the transitional stage of color suffusion are usually
very dark gray dorsally, with faint indications of dorsal spots an-
teriorly. Top and sides of the head usually black. The snout may be
lighter, depending on stage of transition. Supralabials usually black
above and white below.

The ventral coloration of the juveniles is frequently retained though
the color of the dorsal surface may be typical of mature specimens.
At one stage of the transition, when the ventral coloration begins to
turn darker, the underside of the tail and the anterior third of the
belly remain light, while the middle portion of the belly slowly turns
darker. The darker coloration moves in from the lateral portions of
the ventral plates, and in some specimens there remains a light area,
appearing somewhat as a stripe extending down the center of the

96 Tulane Studies in Zoology Vol. 2

belly. Specimens which are further along in the transition usually
have uniform, slaty black bellies. Blotches or spots of black or very
dark gray are frequently found on the throat. The dark ventral spots
of the juveniles may be retained for a long time, as has been shown
by Pope (1944: 172), who mentioned seeing traces of juvenile belly
markings on a specimen of flaviventris that was 42 inches in length.

Dorsal scale rows normally 17-17-15. Reduction takes place by
loss of the third or fourth lateral scale row. Occasionally an addi-
tional vertebral reduction by loss of the seventh or eighth scale row.
Ventrals from 161-193 (mean 178.2 ±1.2), males 177.3 ±1.9, and
females 178.8±1.3; subcaudals from 72-101 (mean 96.2±1.5),
males 99-8±1.9, females 93.2± 1.7; supralabials usually 7, from 8-11;
preoculars normally 2, occasionally 1 or 3; postoculars 2, frequently
3; temporals 2-2-2, but quite variable. Anterior chin shields shorter
than the posterior pair; loreal present, usually longer than high, and
occasionally divided; body elongate, with a long and slender tail.

Large specimens are about five feet long. Ortenburger (1928: 196)
mentioned a large specimen 1660 mm in total length as the largest
that he had measured. Babbitt (1932: 27) reported a specimen
1854 mm long and has told me that he measured a specimen from
Simsbury, Conn., which was 1905 mm in length. Roddy (1923:
32) recorded one from Pennsylvania "over six feet". Another large
specimen which was 1886 mm in total length was reported by Atkin-
son (1901: 146).

The hemipenial characters are as follows: organ slightly bilobed,
with a single sulcus; spines number 89-130, in 6-9 rows, covering
about one-third of the length of the organ; the tip is smooth with a
raised margin formed by the edges of the distal calyces, these edges
being made up of several minute, flat, blunt, spine-like processes;
three large spines are set around the basal portion of the organ.
These large spines are usually from 2-2 Yj times longer than their
predecessors in the same row. The base of the organ is smooth, but
occasionally may be covered with very small spinules.

Maxillary teeth 14-16; dentary teeth 17-22; palatine teeth 13-15;
and pterygoid teeth 20-26.

Range. — By and large C. c. constrictor is an eastern humid forest
form. It ranges from Halifax, Canada, south through eastern Maine
to northern North Carolina. Further inland it is found north of
the Fall Line, throughout the Piedmont and Appalachian Mountain
and Plateau provinces. It occurs as far westward as the Mississippi
Valley in the states of Mississippi, Tennessee and Kentucky. The
range has been shown to be correlated with the unglaciated regions
of Ohio, occurring only in the eastern and south-central parts of that
state (Conant, 1938: 50).

Remarks. — Occasional light-bellied adults are found in various parts
of the range. They resemble intergrades between flaviventris and
constrictor from Ohio in that they are black dorsally and have light
gray or yellowish-gray ventral surfaces. Eckel and Paulmier (1902:

No. 6 Auffenberg: Coluber constrictor 97

371) reported such an individual from New York which was "yellow"
ventrally.

Coluber constrictor priapus Dunn and Wood
Southern Black Racer
Coluber constrictor Holbrook, 1843: 55.
Bascanion constrictor (part) Baird and Girard, 1853: 93.
Coryphodon constrictor Gunther, 1858: 108.
Zamenis constrictor (part) Boulenger, 1893: 387.
Coluber constrictor constrictor (part) Stejneger and Barbour, 1917:

79.
Coluber constrictor priapus Dunn and Wood, 1939: 4.
Coluber constrictor haasti Bell, 1952: 21.

Diagnosis. — A subspecies of Coluber constrictor in which the dorsal
and ventral coloration is mostly slaty black; the enlarged basal hemi-
penial spine is three or more times the length of its predecessor in
the same row. The iris is usually brown or dark amber, but may be
red in both juveniles and adults from Florida.

Holotype.—ANSP 16111, collected by George B. Wood. The type
locality is West Palm Beach, Palm Beach County, Florida.

Description. — For the most part very similar to Coluber c. con-
strictor. Dorsal surface of adults uniform black with a satiny luster.
Top and sides of head black. Snout brownish in many half grown
specimens or even adults. Supralabials usually quite dark, practically
always showing slight traces of white on lower portions of at least
the anterior plates. Large adults from Florida frequently have these
scales colored black above and white below. Ventral coloration dark
slaty gray to slaty black. In some specimens from the northern parts
of its range, as well as those from the Florida Keys, there is no white
on the throat, while in others from peninsular Florida the white may
extend a considerable distance on the throat and anterior parts of
the belly. In some specimens from central Florida this white color
may extend for nearly two-thirds the entire length of the venter.
Chin usually white, but may be spotted with black or very dark gray,
and is frequently so suffused by black that only a few traces of white
are discernible. Iris usually brown, but sometimes dark amber. In
Florida the iris may be reddish, or even orange.

The coloration of juvenile specimens is very similar to that of
constrictor, except that in central Florida the dorsal saddles may be
brownish-red, and the posterior ventral plates may show a distinct
reddish or even a pinkish tinge. There are 47-73 saddles along the
back. Small lateral spots alternate with the larger dorsal blotches.
Ventral spots one-fourth to one-eighth as wide as ventrals. These
markings usually brownish or even black in color, but may be reddish
or orange in specimens from Florida. The transitional coloration is
seemingly identical with that of constrictor.

Dorsal scale rows 17-17-15. Reduction occurs laterally on the third
or fourth scale rows. Occasional specimens are examined that also

98 Tulane Studies in Zoology Vol. 2

have a vertebral reduction by a loss of the seventh or eighth scale
rows. Ventrals , 154-199 (mean 178.0±0.58), males 177.2±0.87,
females 179.1 ±0^84; subcaudals 70-119 (mean 99.8±0.62), males
101.1 ±0.61, females 96.3rh0.98. Supralabials usually 7, but vary from
6-9; infralabials 6-12, usually 8; preoculars 2, rarely 3; postoculars 2,
occasionally 3; temporals 2-2-2, but quite variable. Anterior chin
shields shorter than posterior pair; loreal present, longer than high,
rarely divided and in specimens from Florida frequently in contact
with the first supralabial, rarely so in specimens from other parts of
its range.

At least the Florida specimens of this race seem smaller than most
examples of constrictor that I have examined. Whether or not this
is true of populations in the more western or northern parts of the
range I cannot say with any degree of certainty. Of the southeastern
specimens that I have examined, the largest was a male 1320 mm in
total length. Ditmars (1936: 94) has mentioned that individuals
from the southeastern states are not as large as those from more
northern states. This seems to be true, for the average length of
171 Florida specimens with a uniform coloration is 920 mm while
from New York 34 similarly-colored specimens average 1041 mm.

Hemipenial spines as in constrictor, except that the enlarged basal
spines are much longer, practically always three times the length of
the adjacent proximal spines of the same row. They vary from one
to 5lA times the length of these adjacent spines. Smaller spines
from 72-132, disposed in 6-10 rows, covering about one-third of
the organ.

Maxillary teeth 14-17; dentary teeth 16-20; palatine teeth 13-16;
and pterygoid teeth 20-30.

Range. — The range of C. c. priapus includes the southeastern
Coastal Plain from northeastern North Carolina, including all of
Florida except the Everglades, and associated regions, Cape Canaveral,
Brevard County, and the Chipola and Appalachicola River Valleys in
northwestern Florida. The populations on the Lower Florida Keys
are typical of this race. It also occurs along the Gulf Coastal Plain
to the Mississippi River, thence up its valley, including the north-
western part of Louisiana and a large portion of Arkansas, except the
Boston Mountains and adjacent areas, and northward to the vicinity
of St. Louis, Missouri. In the Ohio River Valley it is found in
southern Illinois and Indiana. It occurs in the states of Kentucky
and Tennessee where they border the valley of the Mississippi River.

Remarks. — Black-colored specimens from outside of the range as
just described have been reported. A specimen "slaty black above"
has been reported from Eddyville, Iowa by Guthrie (1926: 177).
Anderson (1942: 211) mentioned two specimens taken near Inde-
pendence, Missouri that were black dorsally and blue-gray ventrally.
These would seem to be intergrades, although the author noted that
other specimens from the area were colored like typical flaviventris.
Burt and Hoyle (1934: 34) reported a "blue-black" specimen taken

No. 6 Aufjenberg: Coluber constrictor 99

in Rogers County, Oklahoma, well within the range of flaviventris.
Branson (1904: 412) stated that although present in Kansas, black-
colored racers are rare in that area. Strecker (1908: 73) stated that
a black Coluber constrictor was taken in the Bosque Hills, north of
Waco, McLennon County, Texas, and in 1915 (p. 35) reported that
they were taken "as far south as Bosque and McLennon Counties".
Strecker and Williams (1927: 21) reported that a half-grown speci-
men from San Marcos, Texas was typical of the race constrictor,
being black in color. They added that similar specimens are found
in the hill districts of Travis County and other sections nearby. This
does not seem to have been mentioned in the literature since 1927,
and students familiar with this area do not seem acquainted with
this color variety.

The distribution of this race in the western portions of its range
is not well known. All of the specimens I have examined from
states west of the Mississippi River have a hemipenial spine char-
acteristic of the race priapus. The nearest typical specimen of C. c.
constrictor comes from east of the Mississippi River. On this basis
I have assumed that all reliable records of black-colored snakes from
west of the river refer to individuals genetically more similar to
specimens of priapus than to constrictor.

It will be noted that the range of this form combined with the
range of C. c. constrictor is not the same as that indicated by Orten-
burger (1928: 183 ff.), who included Arkansas within the range of
flaviventris. A number of the specimens examined from Arkansas
are black on both dorsal and ventral surfaces. For example, AMNH
64730 from Big Buffalo River, near Pruitt, Ark., is black above and
below, with a white chin. The supralabials are black above and
white below. AMNH 64731 from Little Rock, Pulaski County, is
black above and the anterior portion of the ventral surface is white
and heavily mottled with black. The supralabials are black above
and white below. I have included these and similar specimens in
the race priapus. Only one specimen is referable to flaviventris
among those which I have seen from Arkansas (CM 25025, from
McGehee, Desha County). This individual is grayish above and
very light-colored below. The supralabials are dark gray above and
yellow below.

The literature includes references to black-colored snakes from
Arkansas, usually identified as constrictor by the authors. Thus
Taylor restricted the range of flaviventris in Arkansas to the north-
western portion of the state and listed both Washington and Benton
Counties. On the other hand, Burt (1935: 312) reported flavi-
ventris from Boone, Carrol, Cleveland, Madison, Crawford, Pope,
Saline and Washington Counties, indicating a much more extensive
range. Such references might be indicative of interdigitation between
the races. Dellinger and Black (1932: 21) referred to this situation
in stating the following:

"Most students believe that there is a great deal of

100 Tulane Studies in Zoology Vol. 2

intergradation between the two subspecies of constrictor
found in Arkansas, and that the ranges of both forms
overlap to a considerable extent. For the most part
flaviventris is found in the hills and constrictor [pri-
apus] in the lowlands . . . but the separation is far
from distinct . . . [and] we believe the disposition of
these forms as made by Ortenburger is preferable."

It is unfortunate that none of these authors described or mentioned
intergrades from these areas. Intermediate specimens do exist, such
as those from the lower tier of counties in the southwestern corner
or Missouri. In USNM 99600 from Berryville, Carrol County, Arkan-
sas, the dorsal surface and supralabials are black. Ventrally it is
very light colored, indicating some genetic affinities with flaviventris,
and is best described as an intergrade. CM 25018 from 3.7 mi. west
of Wynne, Cross County, CM 25083, 2 miles west of Fayetteville,
Washington County and CM 25004 from Berryville, Carrol County
(all in Arkansas) are similar, being black dorsally and bluish-gray
ventrally. Until further study is conducted in this area (preferably
of an ecological nature) the best disposition of specimens from
Arkansas is to call all racers that are black both dorsally and ventrally,
and which possess white chins, priapus. All specimens that are black
or bluish-black dorsally and light gray or bluish-white ventrally should
be called intergrades, and all individuals bluish, brownish, or gray
dorsally and white, gray, or yellowish below, flaviventris.

Although Brown (1950) cited no records of black-colored racers
from Texas, priapus probably occurs there. Its range would seem to
embrace only the northeastern corner of the state. Strecker (1908:
73) mentioned specimens of constrictor (probably priapus) from
northeastern Texas. I have heard of three other specimens from
this area that are very dark, but which, unfortunately, I could not
examine. Whether or not black racers are found in eastern Okla-
homa is unknown. Ortenburger (1926, 1927, 1928 and 1930a and b)
and Smith and Leonard (1934) did not mention C. c. constrictor
in the state. Blair and Hubbell (1938) indicated that the Mississippi
Biotic Province extends into the southeastern corner of Oklahoma.
This is the warmest and most humid part of the state, and the plant
associations are typical of those of the Coastal Plain. A common
snake of the Coastal Plain, Farancia abacura, has recently been re-
corded from this region (Burger, 1948: 133). A Coluber constrictor
(T 14003), from 3 miles northeast of Procter, Oklahoma, is black
dorsally and bluish-white ventrally. Unfortunately, the specimen is
rather small and exact identification is difficult. However, its colora-
tion is reminiscent of young intergrade specimens between black
and light colored races in other portions of North America. An-
other specimen (T 6513) from 1 mile south of Nashola, Oklahoma,
resembles an intergrade specimen in color, although it seems closer
to priapus than to flaviventris, being black dorsally and bluish-black
ventrally. The supralabials are practically all black. Burt and Burt
(1929: 10) recorded a specimen of flaviventris from Wister, Le

No. 6 Auffenberg: Coluber constrictor 101

Flore County, Oklahoma in the extreme southeastern part of the
state that somewhat confuses the issue. If populations of black-
colored racers occur in both northeastern Texas and southeastern
Oklahoma these will undoubtedly be found to represent the subspecies
priapus.

As mentioned above, Ortenburger placed southern Illinois within
the range of flaviventris. This is difficult to understand in view of
his color descriptions of this race and of constrictor. One of the
specimens, the color of which was described as typical of flaviventris
is from Hillsboro, Illinois. It was "blackish mouse gray" on the
dorsal surface and lateral tenth of the ventral plates. Below it was
"light olive gray" on the median half of the belly. In describing the
typical coloration of constrictor a specimen was chosen from Marion
County, Florida. The dorsal surface was "blackish mouse gray",
covering the lateral fifth of the ventral plates. The middle portion
of each scute was "Hathi gray", and the anterior edges "light olive
gray". Obviously these two specimens were very similar. At most
the Illinois specimen was an intergrade. That southern Illinois con-
tains other areas with black-colored racers is indicated by numerous
references in the literature. Lueth (1941: 29) and Cagle (1941)
mentioned C. c. constrictor from extreme southern Illinois. In 1942
Cagle (p. 187) stated that most specimens from Union and Jack-
son Counties were intergrades between flaviventris and constrictor.
Blanchard (1924: 536) reported two specimens of C c. constrictor
from Johnson County, Illinois. Ditmars (1936: 59) mentioned con-
strictor from Swallow Rock, Jackson County, Illinois. USNM 94362
from the same locality is intermediate in that the belly is fairly
light. On the other hand, the supralabials and dorsal surface are
both black. This specimen, though similar to flaviventris in its
belly color, is closer to priapus {constrictor of earlier workers). Of
four adults from Olney, Illinois (USNM 13833, 14150a, b and c)
three are intermediate. One is a good example of the race priapus.
This specimen is black dorsally and ventrally. The chin is white
and this color extends posteriorly to the eighth ventral plate. The
supralabials in all are black above and white below.

Mittleman (1947: 481) has already called attention to the black-
colored racers of southern Indiana. Of five specimens from wheat-
land, three are referable to priapus. Two are intergrades. A speci-
men from New Albany is typical of the race priapus*

Coluber c. priapus also extends into extreme southeastern Missouri.
It probably occurs in a narrow band along the Mississippi River. I
have examined intermediates from St. Louis, Montgomery, Oregon,
Ozark, Jackson, Jefferson and Cape Girardeau Counties, Missouri.

Coluber constrictor haasti, recently described by Bell (1952: 21),
is referred to the synonymy of Coluber c. priapus, because the char-
acters given as diagnostic are not confined to the Florida Keys and
they are frequently, and in some cases characteristically observed in
other localities.

102 Tulane Studies in Zoology Vol. 2

Coluber constrictor paludicolus Auffenberg and Babbitt

Everglades Racer

Coluber constrictor constrictor Conant, 1930: 60.

Coluber constrictor priapus Allen and Slatten, 1945 : 26.

Coluber constrictor paludicolus Auffenberg and Babbitt, 1953: 44.

Diagnosis. — A light-colored racer, bluish-gray, greenish-gray, or
occasionally brownish-gray dorsally. Ventrally whitish or light powder
blue, with or without darker cloudy markings. Gulars frequently
tinted with brown and the iris is usually red, sometimes orange or
bright yellow. It differs from jlaviventris in a higher number of
subcaudals, a higher mean number of ventrals, the high frequency
of contact between the 1st supralabial and the loreal and in the
juveniles which have a smaller number of dorsal blotches.

Holotype.—USNM 131900, collected March 23, 1950, by Lewis
Hall Babbitt, three miles west of Kendall, Dade County, Florida.

Description. — Dorsal coloration usually bluish-gray, but varies from
greenish-gray to greenish-brown. A few are decidedly tan. Ventrally
yellowish-white to white, frequently bluish-white. Ventral markings
of adults, when present, may be in the form of cloudy blotches rang-
ing from powder blue to grayish-brown or light tan, and in some
specimens placed along the edges of the ventral plates so that there
is formed a distinct light median line. Posterior gulars and lateral
scales of the anterior part of neck usually tinged with light brown.
Small black punctations may be scattered on the ventral surface.
Top and sides of head usually olive. Snout frequently tan. Supra-
labials bluish-gray above, usually with a band or area of tan or
grayish-brown below this. The bottom third of each plate is white,
or in some cases yellowish. Chin white, but may be yellowish. Iris
usually red, frequently orange, amber or even bright yellow, rarely
brownish-red in adults.

Juveniles with 42 to 65 reddish-brown dorsal saddles, frequently
decidedly reddish, or even pink. The smaller alternating lateral spots
are reddish. Ground color grayish brown. A distinct pinkish or
reddish hue is present posteriorly. Ventral surface pinkish-white
anteriorly and decidedly pinkish posteriorly. Ventral spots usually
small, scattered and reddish to reddish-brown in color. Many speci-
mens have no ventral markings at all.

Dorsal scale rows usually 17-17-15, the reduction at the lateral third
or fourth scale rows. Occasional specimens have an additional verte-
bral reduction of the seventh or eighth scale rows. Head plates
normal in number. The 1st supralabial frequently in contact with
the loreal, which is rarely divided. Ventrals 177-190, mean 184.0±
0.36, males 183.7±0.56, females 187.0±1.07. Subcaudals 99-117,
mean 107.2±0.45, males 107.8±0.81, females 106.9±1.01. Total
length divided by the tail length varies from 3.2 to 3.8.

The largest specimen of this race examined is 1689 mm, while the
mean length of 22 uni-colored specimens is 1119.2 mm. Hemipenial
characters and dentition the same as in priapus.

No. 6

Auffenberg: Coluber constrictor

103

Range* — This subspecies is found in two isolated populations
which do not seem to be recognizably differentiated. One of these
is on Cape Canaveral, Brevard County, Florida. Intergrades between
this population and those of priapus on the adjacent mainland occur
on Merritts Island (fig. 2). The other, and more widely distributed
population, is found in the Everglades and adjacent areas of southern
Florida, from Cape Sable throughout the southern tip of the penin-

CQcape
canaveral

• PRIAPUS

O PALUD1C0LUS
© INTERGRADES

Figure 2. Distribution of Coluber c. paludicolus on Cape Canaveral,
Brevard County, Florida.

104

Tulane Studies in Zoology

Vol. 2

sula, westward to the Big Cypress Swamp in Collier County, north-
ward to the southeastern portion of the Okaloacoochee Slough in
Hendry County, eastward to central Palm Beach County, and south-
ward throughout most of Broward County with the exception of the
eastern portion along the Atlantic Coast. It intergrades with Coluber
c. priapus on both Upper and Lower Matecumbe Keys, Monroe
County, and possibly on Key Largo. Intergrades are also found

• COLUBER C. PRIAPUS

O COLUBER C. PALUDICOLUS

© INTERGRADES

PHYSIOGRAPHIC DIV

I.-EVERGLAOES

2.-BIG CYPRESS SWAMP

3.-SOUTHWESTERN MANGROVE SWAMP

4- MIAMI ROCK RIDGE

5.-ATLANTIC COAST STRIP

6.-FLATLANDS ...

Figure 3. Distribution of Coluber c. paludicolus and Coluber c. pri-
apus in southern Florida. For the most part, paludicolus is confined
to the Everglades, the southern portion of the Southwestern Man-
grove Swamp and the Miami Rock Ridge. It is also known from
Biscay ne Key in Miami Bay and Key Largo.

No. 6 Auffenberg: Coluber constrictor 105

around the periphery of the range as stated above, except, of course,
where it borders the sea (fig. 3). Intermediate specimens between
priapus and paludicolus appear practically identical with those be-
tween priapus and flaviventris, or between constrictor and flaviventris.
They are black dorsally and light slaty-black, gray, or light bluish-
white ventrally. The supralabials are black above and light gray to
white below.

Remarks. — Besides the two populations mentioned above, I have
examined three other specimens from peninsular Florida which are
colored like examples of this race. One is from "Key West, Monroe
County" (USNM 14417), while the other is from "Long Pine Key,
Monroe County" (MCZ 31797). These two specimens, though out-
side the range of paludicolus, probably have the correct locality data.
Influence from this race could be expected over the relatively short
distance to the mainland. However, USNM 83294, labeled "Gaines-
ville", Alachua County, Florida, undoubtedly has incorrect locality
data. Of close to 200 racers which I have examined from Alachua
County, none approached the coloration of paludicolus.

Coluber constrictor helvigularis, subsp. nov.
Brown-chinned Racer

Diagnosis. — A subspecies of Coluber constrictor recognized by the
large amount of very light brown or light tan on the supralabials,
infralabials, and especially the chin and throat of adults. Although
adult specimens of paludicolus normally have tan supralabials and
posterior gulars, these two races are immediately distinguishable on
the basis of their dorsal and ventral coloration. In helvigularis these
surfaces are slaty-black, while in paludicolus they are much lighter.

Holotype.— CM 21462, collected April 9, 1942, by Coleman J.
Goin, 8 miles west of Wewahitchka, Gulf County, Florida.

Description of holotype. — An adult female in which the top and
sides of the head as well as the dorsal surface is dark slaty-black.
Supralabials black above and light brown on the lower half of each
plate. Infralabials and throat entirely light brown. Chin very light
brown, with a few white blotches, each covering only a few scales.
There are traces of a slightly darker shade of brown on the belly as
well as on the underside of the tail wherever the slaty-black color,
which covers most of the belly and tail, is not evident.

Supralabials 7-7, infralabials 8-8; no contact between the loreal and
the first supralabial; postoculars 2-2; preoculars 2-2; temporals 2-2-2
on one side and 2-1-2-2 on the other; gulars 4. Scale rows 17-17-15.
Ventrals 173, anal divided, subcaudals 102.

Paratypes. — Two specimens are designated as paratypes: UF 1496-7,
Scott's Ferry, Calhoun County, Florida.

Variation. — Top and sides of the head and the dorsal surface of
the body of adults black. Ventrally, however, there may be varying
traces of very light brown intermingled with the more prevalent
black color. Chin light brown, but with a few white blotches. Throat

106 Tulane Studies in Zoology Vol. 2

either very light brown or white, the latter strongly contrasted with
the brown chin. Supralabials always light tan or light brown on the
lower half of each scale and black above. Snout either black or with
varying traces of very light brown. Infralabials also tan or light brown.

Supralabials 7-7; infralabials 8-8; loreal in contact with first supra-
labial in three specimens; preoculars 2-2; postoculars 2-2; gulars
ranging from 3-5, mean 4.0; temporals normally 2-2-2, but in some
specimens 2-1-2-2 or 2-1-2. Dorsal scale rows 17-17-15, with the
lateral reduction occurring by loss of the third or fourth scale row,
from the 99th-126th ventral. Ventrals 169-174, mean 171.6±2.32.
Subcaudals 88-102, mean 97.3 ±1.76. In two of the specimens the
last ventral before the anal plate is divided. The hemipenial spines
seem identical with those of priapus, including the enlarged proximal
spines.

Range. — The flood plains and adjacent pine flatwoods along the
Chipola and Appalachicola Rivers of northwest Florida. Specimens
are available from Gulf and Calhoun Counties (fig. 4).

Remarks. — A specimen in the Ross Allen-Wilfred T. Neill collec-
tion (no. 435) which bears the data Iron City, Seminole County,
Georgia, is typical of this race. The chin is heavily mottled with
light brown. Both the infra- and supralabials are suffused with this
color. No other specimens are available from the immediate locality.
However, seven specimens are available from the area near the junc-
tion of the Flint and Chattahoochee Rivers (UF 3728-34). Five of
them are from Florida and two from Georgia. Only two of these
specimens have the faintest indication of brown on the chin. Those
of the remaining specimens are pure white. In the collection of John
W. Crenshaw there are 18 adult specimens from Baker County,
Georgia. This locality is about 35 miles northeast of Iron City,
Seminole County. None show any tendency toward a brownish-tinted
chin. The only other specimen which is available from this section
of Georgia is one from Fort Benning, Chattahooche County (USNM
80948), which is on the Fall Line. It is an intergrade between
priapus and constrictor and has no brown on either the chin or labials.
In southeastern Alabama I have seen a specimen from Enterprise,
Coffee County (WA 2501). This locality is about 40 miles west of
the Apalachicola River. The specimen shows no tendencies towards
the coloration of helvigularis.

Intermediate specimens have been examined from Florida and ex-
treme southwestern Georgia. CM 21464 from 12.2 miles west of
Wewahitchka, Gulf County, has a light brownish tinted chin, with
considerable amount of white present. This specimen, though inter-
mediate, is closer to helvigularis, DBUF 2644 from Blue Springs,
Jackson County, is practically identical to priapus except that the
posterior gulars are light brownish. The supralabials also have con-
siderable brown on them. The same is true of two other specimens
from near Panama City, Bay County (USNM 81174). These last
three specimens are certainly closest to priapus, although the presence

No. 6

Auffenberg: Coluber constrictor

107

Figure 4. Distribution of Coluber c. priapus and Coluber c. hel-
vigularis, subsp. nov.

of a considerable amount of light tan coloration shows a notable
genetic influence from helvigularis populations farther eastward. All
of these localities are about 25 miles west of the Appalachicola River.
Two other specimens (WA 4001 and 4002) from 5-7 miles east of
Tyndall Air Base, Bay County, Florida, are intergrades but closer to

108 Tulane Studies in Zoology Vol. 2

helvigularis. A specimen which I have seen (no longer available)
from extreme western Leon County, near Bloxham, was typical of
the race priapus. UF 3731, from Sealy Springs, Seminole County,
Georgia, has very faint indications of light brown on the chin. The
same is true of UF 3729 from Chattahoochee, Gadson County, Florida.

Individual and Geographic Variation

CORRELATED VARIATION

Under this heading are discussed the variations in color and scutel-
lation which are correlated with geographic locality, sex, environment,
or ontogeny.

The term cline has been used to express a condition in which the
values of a variable character form a slope or gradient over a geo-
graphic area. With increasing knowledge of variation systematists
have come to recognize different types of clines. Although closely
related, two obvious types come to attention. These are the eco- and
geocliness In the first, a character gradient is present which con-
forms with an ecological gradient over the same area. In the second,
the character gradient simply seems correlated with distances or areas.
In most cases it is obviously difficult to divorce the ecocline from
the geocline.

Other than these two distinctions, character gradients may have
varying slopes when the values of the character being examined are
plotted against distance. Two general types of gradients seem to be
recognizable by this type of analysis. One of these is the external
cline. In this type the continuous sloping character-gradient is inter-
rupted and may be thought of as resembling a stair case, with the
separate subspecies corresponding with the treads, flat or very gently
sloping, and being united by steep slopes — the zones of intergradation.
The mean values for different characters of the various subspecies will
often form a gradient. This condition may also be called an inter-
group cline (see Huxley, 1943: 211). The other type is called an
internal cline, which does not show any steepening of the character
gradient, or cline, in certain areas, but rather a continuously sloping
character-gradient with no obvious interruptions (Huxley, loc. cit.).

A rather special type of cline, generally termed an ontocline, is also
shown in Coluber constrictor. It is an ontogenetic change that varies
as a character-gradient in degree or intensity with distance.

External Clines

Length of the hemipenial spines. — The characters associated with
the hemipenes of male snakes have long been used in taxonomic
studies of serpents. In general they have been used as characters of
generic, or specific consequence. However, Dunn and Wood ( 1939:
1 ff) have shown that proportional measurements of the proximal
spines may also serve as one of racial significance.

In Coluber constrictor the basal portion of the hemipenes is spinous.
These spines are disposed in definite rows, with a gradual proximal
increase in the size of the spines. At the basal end of certain of

No. 6

Auffenberg: Coluber constrictor

109

these rows there is also a very much enlarged spine. Dunn and Wood
{ibid.) have suggested a standardization of terms for the proportional
measurements of the enlarged basal spines and the smaller preceding
ones in the same rows. A basal spine that shows only a slight enlarge-
ment over the preceding one is called "not significantly enlarged".
If it is two times the length of the adjacent proximal spine it is
"significantly enlarged". It is called a "basal hook" if it is four times
the length of the next smaller spine in the same row (fig. 5).

I & XI

Figure 5. Variation of the enlarged basal spines on the hemipenis of
Coluber constrictor, a. Basal portion of the hemipenis of a typical
specimen of C. c. priapus (WA 3009, Alachua County, Florida).
b. Basal portion of the hemipenis of a typical specimen of C. c. con-
strictor (WA 3670, Atlanta, DeKalb County, Georgia).

Following an examination of several male black racers from eastern
North America, Dunn and Wood (ibicL) found that all of the
Florida specimens were provided with a "basal hook", while others
from New York, New Jersey, Pennsylvania and Virginia had basal
spines which were, at most, "significantly enlarged". In view of this
geographic variation, they proposed the name Coluber c. priapus as
a new subspecies, which was then known only from Florida. In the
present study an examination of many specimens from the eastern
half of North America has shown that this race is not restricted to
Florida, but has a distribution which is very similar to that of many
other Coastal Plain reptiles and amphibians.

The basal spines of this species may vary in two ways: (1) the
length of the spines (as shown by Dunn and Wood), and (2) the
number of basal spines which may be termed hooks. The values of
the former may be constant over large areas of North America, with

110 Tulane Studies in Zoology Vol. 2

zones of intergradation between. However, the mean lengths in
these same areas vary geographically as a gradient. In view of these
two characteristics I have included this type of variation under "ex-
ternal dines". The latter situation ( 2 ) has already been referred
to by Dunn and Wood (ibid.), who stated that some Florida speci-
mens have one hook, some have two, and others even have three.
This will be discussed under "internal clines".

Although the eastern limits of the range of flaviventris has been
shown to lie westward of that indicated by Ortenburger (1928:
183 ff), specimens on the eastern side of the Mississippi River have
always been referred to the race constrictor. In view of this it is
unusual that Dunn and Wood (ibid.), in writing of a male from
Samburg, Tennessee (ANSP 4489) that had one basal "hook", did
not question the restricted Floridian range of priapus* Mittleman
(1947: 479) has indicated that specimens of Coluber c. constrictor
from near the area of intergradation with C. c. flaviventris possess the
large basal hooks upon which priapus has been described. He raised
the question that if basal hooks occur among some specimens of
middle-western constrictor, the entire question of the correlation of
these structures with geographic dispersion should be re-examined.

All of the races of constrictor, with the exception of constrictor
constrictor, appear to have basal spines which may be termed "hooks".
In the black-colored racers there is at least one hook present in speci-
mens from southern Illinois and Indiana, western Kentucky and
Tennessee, extreme eastern Missouri, Arkansas, northeastern Louisiana
as well as the Florida Parishes, all of Mississippi except the north-
eastern corner, and all of Alabama, Georgia and South Carolina below
the Fall Line, throughout Florida, and in coastal North Carolina north
to the Cape Hatteras region. This, with the exception of the ranges
of paludicolus and helvigularis , is the distribution of priapus.

Practically all of the specimens that I have examined from the
Coastal Plain have a basal hook. However, the ratios varied from
one to five and a half. In view of such individual variation mean
ratios were calculated for various regions of the eastern United
States to obtain an estimate of broader geographic variation. These
ratios are given in Table 1. It is apparent that these data fall into
two groups which correspond to the ranges of the two races, priapus
and constrictor*

Along the Fall Line specimens are frequently taken in which the
spines are two and three quarters, three or three and a quarter times
the length of the smaller spines. I have called these specimens inter-
grades. In eastern Ohio, where color intermediates occur between
constrictor and flaviventris, the hemipenial spines do not seem to be
intermediate, but resemble those of constrictor.

The overall ratios of specimens for the southern portion of the
range (using individual ratios for calculation) was 3.7 ±0.67. The
corresponding value for northern specimens was 2.2±0.57. Hence
the two groups of specimens bear a difference in the mean values

No. 6 Auffenberg: Coluber constrictor 111

that is statistically significant (t = 5.73).

TABLE 1.

Geographic Variation in Mean Ratios of the Length

of the Proximal Hemipenial Spines

Area
Florida

Southern Georgia (below Fall Line)
Coastal South Carolina

Coastal North Carolina (excluding N.E. part)
Southern Alabama
Southern Mississippi
Eastern Louisiana
Arkansas (excluding N.W. part)
Western Ky., Tenn. & southern 111. & Ind.

Mean No. of

latio

Specimens

3.9

305

3.8

31

3.3

12

3.3

7

4.0

26

3.6

8

3.6

21

3.8

23

4.0

11

2.1

26

2.2

12

2.3

15

1.7

18

2.1

20

2.2

11

New York
Pennsylvania
Virginia
Eastern Ohio
Northern Georgia
Western North Carolina

Specimens from coastal North and South Carolina frequently have
basal spines which are about three times the length of the adjacent
proximal spine. There is no sharp break between the specimens
with small proximal spines which are only significantly enlarged, and
those with true basal hooks. The region of intergradation in the
coastal areas of these two states is wide and exhibits a more gentle
gradient than is found in the intergradation area along the Fall Line
of Georgia and Alabama.

Color. — It is obvious that the distinguishing character of priapus
is restricted in its value, not only in view of the fact that it is sex-
limited, but also in that it is accurately reliable in only rather large,
dead or preserved specimens. Coloration is one of the predominant
diagnostic characters used to distinguish the various races of Coluber
constrictor. It has been referred to time and again in relation to
southeastern populations of this species. Ortenburger (1928: 213)
was one of the first workers to mention the presence of a distinguish-
able coloration in specimens from the southeast: "In the southeast
the throat and chin are almost always immaculate blue-white, while
to the north of Georgia the chin and throat are white, but more or
less spotted with dark gray."

Ditmars (1936: 94; 1939: 191) mentioned this peculiarity and
gave some indication of a race confined to the southern portion of the
United States: "With specimens from South Carolina, Georgia and
Florida the white may cover the entire upper labials, thus indicating
a fairly well defined southern race."

The exact range of variation in this coloration has never been
examined.

A complication in the description of color in this species is that

112 Tulane Studies in Zoology Vol. 2

an ontogenetic color and pattern change occurs when the snake is
about one year old. The blotched pattern of the juvenile becomes
the uniform coloration of the adult through continuous and successive
stages of suffusion by a darker color. Color notes taken on specimens
from one area are not comparable to notes taken on specimens from
other areas, unless the specimens are of the same size or age. Un-
fortunately the size at darkening is not the same in every individual,
as has been shown by Kelly (1936: 43). In some cases the size at
color transition is somewhat correlated with geographic locality. An-
other difficulty is that in preservatives not only does the entire speci-
men become more or less darkened, but certain colors, such as those
of the iris and tongue, fade rapidly.

Adult dorsal color, — This is one of the most important criteria
used in distinguishing the races of Coluber constrictor. Based on
their dorsal coloration these subspecies fall into three main groups.
They include ( 1 ) the light, usually greenish, brownish or bluish
group, which is composed of the races C. c. mormon, stejnegerianus,
jlaviventris and paludicolus. For the most part they inhabit rather
open situations, under relatively little forest canopy. They are pri-
marily found in grasslands, and one subspecies, in marshes. Another
group (2) is formed of only one race, anthicus. It is spotted and
occurs in the pine forests of eastern Texas and western Louisiana
(fig. 16). The last group (3) is black and is composed of three
subspecies, constrictor, priapus and helvigularis. These are primarily
associated with the forested portions of eastern United States (fig.
16). It might be added that jlaviventris inhabits a great variety of
habitats and is also the most variable as regards coloration. Specimens
are lightest in the Great Plains and darkest in the northeastern por-
tions of its range where it has invaded the forests of the glaciated
regions of North America. This post-glacial eastern extension of
typical steppe fauna has been frequently discussed (Schmidt, 1938;
Ruthven, 1908; and Grobman, 1941).

The light-colored subspecies show perfect intergradation with
darker forms at their mutual borders, and for this reason adult color
is included under "external clines".

The dorsal color of the races priapus, helvigularis and constrictor
is nearly always slaty black. Ortenburger (1928: 212-213), stated
that ". . . specimens from southern Florida, and to a certain degree
those of New England are more intensely black than those from any
other part of the entire range". Undoubtedly he was writing of
specimens from the Florida Keys, for southern peninsular Florida is
inhabited by a very light race, paludicolus. This is exemplified,
not only by the lighter coloration of the dorsal surface, chin, supra-
labials and belly of adult specimens from this area, but by the lighter-
colored juveniles as well. This is opposed to the darkly-colored
juveniles and adults of northern parts of the state. The light color
of this Everglades population has been mentioned by Conant (1930:
60), Carr (1940: 81) and Allen and Slatten (1945: 25). Specimens

No. 6 Auffenberg: Coluber constrictor 113

from the Keys are black dorsally, but not appreciably darker than
specimens from Alabama, Mississippi, southern Georgia, or even
northern Florida.

The dorsal coloration of paludicolus is a bluish-gray, but in some
specimens is greenish-brown to greenish-gray. I have been told that
specimens from Cape Sable are tan. The only individual that I have
examined from this area is greenish-gray. Intergrades between this
race and priapus occur in Collier, Hendry, Palm Beach, Broward and
Monroe Counties ( fig. 3 ) .

There is a southward extension of the range of priapus down both
the east and west coasts of the peninsula. On the east its range seems
correlated with an area known as the Atlantic Coast Strip which is
mostly composed of deep sands and is higher than the Everglades.
Forests of pine cover most of the area, while in the dunes along the
sea the vegetation is decidedly scrubby. The race priapus also occurs
in the Pine Flatlands of the lower west coast. This area is gently
rolling, with many small ponds and sloughs. Open pine forests with
grasses and thickets of saw palmetto predominate the landscape. The
Florida Keys are for the most part covered with scrubby bushes and
trees. On the other hand, the Everglades is a shallow, praire-like
basin, that, before the digging of drainage canals, was about 80
percent marshland. Trees occur only along the edges of natural
sloughs or in "islands" on slightly higher ground. The conditions
of these two major habitats simulate those between the forests and
prairies of our middle west. The resemblance of the two prairie
situations, even though one is wet and the other is dry, and the
similarity of the color of the forms inhabiting both suggests that
approximately equivalent natural selective factors have acted on both
of these populations.

A series of seven specimens in the AMNH (6981-7) and a head
(USNM 44518) from, respectively, Canaveral Light and Canaveral,
Brevard County, Florida, are colored exactly like specimens of paludi-
colus from the Everglades. The color in all is greenish-gray to
greenish-brown. Three of them (AMNH 6983, 6985 and 6987) have
the light brownish-colored posterior gulars frequently observed in
paludicolus. Three adult intergrades between the Cape population
and priapus have been examined. USNM 11989, from Georgiana,
Brevard County, has a coloration which resembles that of the speci-
mens from Cape Canaveral. WA 3719, from seven miles east of
Titusville, Brevard County, is intermediate. In life it was black on
the anterior portion of the dorsal surface, and brownish-black pos-
teriorly. Ventrally it was light bluish-black, and, as frequently hap-
pens in intergrade specimens from the Big Cypress Swamp of Collier
County, there was a light powder blue line mid-ventrally. The
gulars and middle third of the supralabials were brownish. Six speci-
mens from Eau Gallie, Brevard County are typical of the race priapus,
except for one, AMNH 6919. This specimen is very dark above,
but quite light ventrally. The gulars are tinted with light brown.

114 Tulane Studies in Zoology Vol. 2

This is also considered an intermediate specimen (fig. 2).

There are two separate populations of paludicolus which do not
seem to differ in any color characters. The scale counts of these
two populations are also practically identical. An offshore bar extends
southward from Cape Canaveral for a considerable distance. One
intergrade specimen has been examined from this bar.

The dorsal color clines which are found in the area of intergrada-
tion between priapus and paludicolus are exactly the same as those
between jlaviventris and constrictor in Ohio, or between flaviventris
and priapus.

Specimens from intermediate areas are black dorsally and light
ventrally. The ventral color seems to change over a greater geo-
graphical distance than does the dorsal color. For example, in south-
ern Florida the change in ventral color from slaty black to the light
color of paludicolus takes place over approximately 55 miles, while
the change in dorsal color takes place over 30 miles at most. As
would be expected, the width of the intermediate zone is dependent
on the distance over which grasslands, or essentially prairie conditions,
change into a forested condition. As a result, the slope of the cline
is directly proportional to the proportion of prairie and forest con-
stituents over a given area. This is exemplified by the rather short
distance in which the dorsal color and topography change along the
edges of the Big Cypress Swamp and the Everglades, as compared to
the more gradual change in the dorsal color and topography in cen-
tral Ohio.

I have seen eight specimens from Upper and Lower Matecumbe
Keys, Monroe County, Florida, some 20 miles south of the mainland,
which are very unusual in that many of the dorsal scales have lighter
centers, so that they appear speckled. USNM 85303, from Key
West, Monroe County, Florida, also has dorsal scales which have
slightly lighter centers, but in addition, certain scales are distinctly
tinted, though faintly, with greenish, brown, or blue. The ground
color of all of these specimens is slaty black. They are reminiscent
of intergrades between anthicus and priapus which I have examined
from Louisiana and Arkansas, except that the speckling is not as
evident. Certain specimens of paludicolus. which I have seen have
scales whose centers are slightly darker than their edges.

Adult ventral color. — A resume of the general coloration of the
ventral surface as found in the various races would serve no real
purpose, since the belly color is seemingly correlated with the dorsal
coloration. That is, subspecies dark on the dorsal surface are also
darkly colored ventrally. It is only in the zones of intergradation
between light and dark-colored forms that one finds snakes that are
very dark above and light below. The contrast between the two
surfaces is considerable. Such specimens have been described in earlier
parts of the paper.

Recently Schmidt (1953: 187) revived the name Coluber c. foxi
(Baird and Girard) for those populations of Coluber constrictor in-

No. 6 Auffenberg: Coluber constrictor 115

habiting the Eastern Steppe Province of Ohio, Indiana, Illinois and
Michigan. Following an examination of many specimens from Ohio
and Michigan it is obvious that the adult ventral color of this popu-
lation and that of the Great Plains is quite different. Ortenburger
made reference to this difference (1928: 191), "The typical yellow
coloration is found in Kansas; the blue-gray, in the northern part of
the range". However, he failed to recognize two subspecies on this
basis alone. Specimens from the states of Iowa, Illinois, and Missouri
seem to indicate that the change in ventral coloration is gradual and
covers a great distance; that is, it seems to exhibit an internal, rather
than an external cline. Such a wide band of intergradation is incon-
gruent with the widths of zones of intergradation between other
subspecies of Cohtber constrictor in eastern United States.

The darker ventral surface of foxi is interpreted in the present
paper as an indication of past or present genetic influence from dark
bellied forms, such as constrictor or priapus. Specimens of flavi-
ventris from southern Louisiana, near the range of priapus, are very
similar to foxi in the coloration of their ventral surface. This cer-
tainly indicates an approach to priapus in this area. It has been
previously shown that the ventral coloration changes over a greater
area than does the dorsal coloration in areas of intergradation between
light and dark-colored forms.

Until a critical study is made of the nature and degree of inter-
gradation between the yellow and grayish-bellied forms, and the degree
of homogeneity in the two proposed nominal populations, the best
disposition seems to be to regard them as slightly differing popu-
lations of the same subspecies.

Occasional specimens are taken throughout the range of constrictor
and priapus that are black above and very light gray or bluish-white
below. These individuals appear like intergrades between priapus
and paludicolus or jlaviventris and constrictor. Eckel and Paulmier
(1902: 371) seem to have had such a specimen in mind when they
referred to specimens from New York which were "greenish-yellow"
ventrally, though black dorsally. This coloration seems most common
along the Atlantic Coast, yet still unusual in these areas. Only 4.3%
of the specimens from South Carolina are so colored, 1.9% in eastern
Georgia, 3.2% in northern Florida and 8.2% in central Florida.

Adult chin color. — The chin color of adult racers has been men-
tioned by a number of earlier workers as a possible diagnostic char-
acter of a southeastern race. One of the difficulties in describing
the coloration of racers lies in the fact that there is an ontogenetic
change in coloration. All juvenile racers have a white chin, usually
without markings of any kind. In the races jlaviventris, mormon,
stejnegerianus, anthicus, and paludicolus there is no appreciable color
change to adulthood, except possibly with an introduction of a yellow-
ish tint. However, in the black-colored racers of the eastern parts
of North America specimens have chins mottled with black, or brown,
as in the case of helvigularis. This mottled appearance is brought

116

Tulane Studies in Zoology

Vol. 2

about by growth and fusion of black blotches on the chin where
before there were none. In this manner some specimens attain prac-
tically solid black chins.

Not all specimens from one locality exhibit the same degree of
darkening on the chin. In twenty individuals from Key West, Mon-
roe County, Florida, practically every gradation between an immacu-
late white chin and one which is almost entirely black is represented.
The specimen with the least white on the chin that I have examined
from the Florida Keys is CMNH 33816, in which only the mental
scale is white. The darkest racer that I have examined is WA 2506
from 5 miles north of Florence, South Carolina. There is no pure
white on any part of this snake. The anterior portion of the chin
exhibits a few smoky-gray vermiculations.

Although Ortenburger (1928: 213) and Ditmars (1936: 94; 1939:
191) indicated that specimens from the southeast have less white on
the chin than specimens from more northern states, an evaluation of
individuals over 1000 mm in total length has indicated that the
number of specimens in the southeast with black-colored chins is
quite large. However, specimens from central Florida have a con-

B

Figure 6. The amount of white on the chin of three typical speci-
mens of Coluber constrictor from eastern United States. A. Typical
adult from New York State. B. Typical adult from South Carolina.
C. Typical adult from northern Florida. Any of these variations
may be found in any area of eastern North America, and the amount
of white on adult specimens is best measured by the mean number
of white ventrals (cf fig. 7).

No. 6 Auffenberg: Coluber constrictor 117

siderable amount of white on the chin and throat. This variation
is best measured by tabulating the number of ventrals on which white
occurs. When this is done for states along the Atlantic Coast the
distribution is typical of that of an external cline. There are two
areas in which the values, though distinct, are fairly constant, with a
steepened gradient in the zone of intergradation between. Figure 6
shows three typical specimens from the Atlantic coastal states, while
figure 7 shows the results of an analysis of the amount of white on
the chin and throat of specimens over 36 inches long from various
localities in eastern United States. Although a southeastern race
cannot be described on this basis alone, there is an indication of a
differentiated population occurring in the lowlands of the lower por-
tion of the St. Johns River in Florida. However, these specimens are
probably intergrades between priapus and paludicolus, even though
considerable homogeneity seems indicated (cf page 92).

Adult gular and supralabial color. — The lateral gulars of large speci-
mens of C. cs constrictor and C. c. priapus, except those from central
and south-central Florida, are usually spotted with black. Occasional
specimens have completely black gulars. In central Florida speci-
mens usually have much white on the chin and throat, and in these
the gulars are also white. In the region of the flood plains of the
Chipola and Appalachicola Rivers of western Florida the entire chin
and throat may be suffused with very light brown or tan. These
specimens have been designated as a new subspecies, C. c. helvi-
gularis. Brownish-tinted posterior gulars are also quite common in
specimens of C. c. paludicolus. Intergrades between the latter and
priapus frequently exhibit this coloration. Specimens from the west
coast of Florida rarely have brownish-tinted gulars. This color never
seems to be found in other specimens of priapus. Both paludicolus
and helvigularis have light brown or tan on the supralabials. In both,
though especially in the latter, the ventral surface is frequently clouded
with light brown or tan.

Specimens with unusually-colored gulars or chins are occasionally
found, such as a specimen from DeLand, Volusia County, Florida
(WA 3610) in which these parts were light orange in life. In speci-
mens of flaviventris , mormon, stejnegerianus, or anthicus the gular
scales are yellow, white, or yellowish-white.

The supralabials of full grown specimens of constrictor are usually
black with only traces of white on the lower parts of the first few
plates. This is also true of large examples of priapus from its entire
range, except individuals from central peninsular Florida. Specimens
from this area usually have supralabials which are black above and
white below. Half grown specimens of constrictor and priapus have
these scales similarly colored. Individuals are frequently taken in
south-central Florida from Lake Okeechobee south to the Big Cypress
Swamp of Collier County that have supralabials which are black
above, and immediately below this there is a narrow band of very
light tan or light brown. The bottom third is white. In this regard

118

Tulane Studies in Zoology

Vol. 2

NO. OF SPEC.
EXAMINED

0 2 4 6 8 10 12 14 16 18 20
MEAN NUMBER OF WHITE VENTRALS

Figure 7. The mean number of white ventrals on the throats of
specimens of Coluber constrictor from the Atlantic Coast area. The
distribution of mean values in this figure is a fairly good example
of an external cline in which the slope of the character-gradient is
rather level over two large areas, but with a rapid increase in slope
over a narrow portion of the range — the zones of character inter-
gradation. These zones occur in northern Florida and on the Upper
Keys. Although this figure would indicate the presence of another
nominal population in central Florida, the character, and its dis-
tribution in this region, points to a narrow zone of frequent gene
flow between two nominal populations inhabiting two rather similar
ecological situations, and separated by an obviously inefficient barrier
(see page 92).

these specimens show a genetic influence from populations of
paludicolus from farther south. In this race the upper portion of
each plate is dark gray or bluish-gray rather than black. Individuals
with light tan or brown on the supralabials are frequently taken in
the vicinity of Sarasota, Manatee County, in west-central Florida

No. 6 Auffenberg: Coluber constrictor 119

(fig. 16). The posterior gulars of specimens from this area are tan.
Individuals that possess a very small amount of light brown or tan
on the supralabials are also taken in the Gulf Hammock region of
western peninsular Florida. Full grown adults from northwestern
Florida, except those from within, or near the range of helvigularis ,
nearly always have black supralabials. Large specimens from this area
which have these plates half black and half white are much less
common than in specimens from peninsular Florida.

Along the zones of intergradation with light-colored racers of
middle North America the supralabials become dark blue or black
above and white below, gradually fading into the lighter colors char-
acteristic of the western subspecies. The supralabials of specimens
of anthicus from the eastern parts of its range are black above and
white below. In the western parts of its range the supralabials are
dark gray or bluish above and whitish below. In the middle portions
of the range of anthicus the supralabials are usually either all light
tan, or tan above and whitish below. The color of the supralabials
in jlaviventris may range from dark bluish-gray to light brown above.
Below they may vary from white to bright yellow. The latter color-
ation is especially prevalent in the Great Plains portion of its range,
while the former is most frequently found in the northeastern parts
of its range.

Temporal coloration. — There are certain other colors frequently
found on the heads of racers which may vary geographically. One
of the most obvious of these is the coloration of the temporal scales.
In most subspecies of Coluber constrictor these scales are slightly
darker than the supralabials, parietals, or lateral scales of the neck,
all of which surround the temporals. However, many specimens
from Michigan, Indiana, Ohio, Illinois, and especially southeastern
Louisiana have very darkly-colored temporals which, in many cases,
gives an impression of a vague, short, postocular stripe. This color
is usually a dark grayish-blue. Whether or not populations possess-
ing dark-colored temporals form steep gradients where they inter-
grade with populations having lighter-colored temporals is unknown.
However, the fact that this characteristic seems to be found in rather
definite areas may indicate that the zone of intergradation is narrow.

Internal Clines

Number of bemipenial spines. — It has been previously shown that
on mapping the proportionate lengths of the basal hemipenial spines
in specimens of Coluber constrictor an external cline is indicated;
i. e., a definite and relatively narrow area of intergradation exists
between populations having basal hemipenial hooks and those that
do not. However, the number of hooks also seems to vary geo-
graphically, but as a steady cline or gradient, with no obvious areas
of intergradation (steepening of the cline). Huxley (1943: 220)
also called this a continuous cline.

The maximum number of hooks found in any specimen is three.
This number occurs occasionally in all subspecies that normally possess

120 Tulane Studies in Zoology Vol. 2

at least one hook. Most of the specimens of C. c. mormon that were
examined by Dunn and Wood (1939: 1-4) had three hooks present.
In an examination of twelve specimens from California eight pos-
sessed three hooks, three had two and only one had one hook. In
15 specimens of C. c. stejnegerianus from Nueces County, Texas,
nine had three hooks, four had two and two had only a single hook.
In three specimens of C. c. anthicus none had three hooks, one had
two, and two had a single hook. Although C. c. jlaviventris fre-
quently has three hooks, this number seems much more common in
the southern portions of its range than in the northern parts. In
fact, a large number of specimens from Michigan do not have any
basal hooks at all, but have basal spines which are from three to
three and three-fourths times the length of the adjacent proximal
spines in the same rows. Of 211 adult black racers from Florida
(C. c. priapus) 72.0% have one hook, 14.6% have two hooks, 11.1%
have three hooks and 2.3% have no hooks. The evidence, though
admittedly scanty from extreme western United States, seems to in-
dicate that a large geocline is present in the number of basal hemi-
penial hooks as found on Coluber constrictor. Three hooks are most
frequently found in populations from the western states, one hook
being normal in eastern populations. Of 32 adult male specimens
from the states of Missouri and Arkansas, 34.3% have three hooks,
40.9% have two hooks, 21.7% have one hook and 3-1% have no
hooks, indicating an intermediate condition between the two ex-
tremes, both numerically and geographically.

Color of the young. — The transformation of the color and pattern
of the juvenile racers to that of the adult follows certain well defined
trends. On the dorsal surface the suffusion of the body blotches takes
place from the posterior to the anterior regions. Even on some newly
hatched specimens it is quite difficult to count the total number of
these markings due to a certain amount of darkening on the posterior
portions of the body while the snake was still in the egg. In most
newly hatched specimens the total number of these dorsal markings
can be counted easily.

The mean number of dorsal blotches in young racers varies geo-
graphically, as has been shown by Ortenburger (1928: 180, 213, and
216). Populations of this species which inhabit the western portions
of North America have a higher mean number of body blotches than
do populations from more eastern localities. Furthermore, this geo-
graphic variation is a gradual cline, except possibly in central Ohio
(Conant, 1938: 54). This character gradient is best described as a
geocline, and more specifically — an internal cline.

The subspecies, C. c. stejnegerianus, is thought to be closest to the
ancestral stock, and this form possesses the highest number of dorsal
blotches. However, the total number is seldom countable due to the
fact that they are rather indistinct posteriorly. They range from
42-53 on the anterior third of the body and are not saddle-shaped as
in the eastern forms, but crossbands (Auffenberg, 1949: 55). C. c.

No. 6 Auffenberg: Coluber constrictor 121

flaviventris has 43-80 blotches. Specimens of this race from south-
eastern Louisiana have fewer dorsal saddles than do those from more
northern parts of the range. In 23 newly hatched specimens from
Jefferson Parish, Louisiana (WA 3673-96), the number varies from
43-63 (mean, 55). This is considerably lower than in a series of 18
specimens of flaviventris from Oakland County, Michigan (WA
3026-44), in which they range from 68-76 (mean, 71.3). In Florida,
from which a large number of young are available, a definite trend
is indicated, whereby specimens from southern Florida seem to have
fewer blotches than do those from northern Florida. Of 56 newly
hatched specimens from northern Florida the total number of blotches
varies from 43-73, mean 52. In 32 specimens from central Florida
the number of blotches ranges from 45-76, mean 49.7. In 10 speci-
mens from extreme southern Florida in which the total number of
dorsal blotches could be counted, they vary from 40-57, mean 45.8.

Conant (1942: 195) has also shown that geographic variation
exists in the coloration of newly hatched racers. He mentioned that
specimens of priapus from five miles south of Brighton, Glades
County, Florida, had body blotches which were "rich chestnut brown",
while in a series from Philadelphia, Pennsylvania these markings
were "dark brown". The specimens from Florida were also more
reddish in other parts of the body and these colors were suggested as
additional means of distinguishing juvenile specimens of priapus and
constrictor.

An examination of the material has shown that not all specimens
of priapus, or even all specimens of priapus from Florida, have
such brilliant colors. Ditmars (1939: 192) mentioned dark brown
blotches on a specimen from Marion County, Florida. Although he
recognized a difference concerning the color of the chin and supra-
labials of adult specimens from the southeast as compared with those
of the northeast, still, in describing the typical young of black racers
he used a specimen from Florida. In a series of 18 specimens from
Columbia County, Florida, the dorsal blotches are a very dark reddish-
brown and, when placed next to a series of hatchlings from Scioto
County, Ohio, no appreciable difference can be discerned. However,
in eleven specimens from Alachua County, which is directly south
of Columbia County, the blotches are decidedly more reddish. An-
other juvenile described by Ditmars (ibid, p. 192) from Connecticut
has dorsal blotches which are "grayish-brown", being lighter than
the markings of the Columbia County, Florida specimens. A hatch-
ling from Pensacola, Escambia County, Florida, has pinkish-gray
blotches, while one from Tampa, Hillsborough County, Florida, has
dorsal blotches which are very dark brown.

On the whole, most juvenile specimens from northern Florida
northward, including the remainder of the range of priapus and all
of that of constrictor, have markings which are usually some shade
of dark brown, frequently dark reddish-brown. Conant's specimens
from Glades County, which is quite far southward in the state, shows

122 Tulane Studies in Zoology Vol. 2

a portion of the dine exhibited by this coloration in peninsular
Florida. In the Everglades of southern Florida practically all of the
juveniles have light chestnut, or even reddish or pinkish blotches.
Such light colors are never found in specimens from northern Florida,
but are frequently found in individuals from south-central Florida.
These colors are not diagnostic of the race as a whole, but the char-
acter exists as a geocline throughout peninsular Florida.

Specimens of flaviventris from southeastern Louisiana, which seem
to have fewer dorsal blotches than do specimens of the same race
from farther north, also have dorsal blotches which are decidedly
more reddish than do specimens from farther north.

Conant {op. cit.) also mentioned the general reddish cast to the
entire body as of possible significance in the identification of juvenile
specimens of priapus. This characteristic, as with the lighter dorsal
markings, is most prevalent in the Everglades race, pdudicolus.
Throughout the major portion of the range of priapus (excluding
south-central Florida) most juveniles have a gray, or grayish-brown
ground color as do specimens of constrictor. Of special significance
is the fact that specimens of flaviventris from southeastern Louisiana
show the reddish suffusion of the general body colors which is char-
acteristic of juvenile paludicolus* Specimens of flaviventris from
more northern parts of the range do not show this coloration, but are
similar to constrictor.

The ventral color of juveniles from south-central Florida is also
different from that of specimens in the Philadelphia area (Conant,
ibid). This difference involves both the ground color and the small
ventral spots. Conant's specimens from Glades County, Florida had
"spots on the belly orange brown; ground color of belly (posteriorly)
bright pinkish orange". Specimens from the Philadelphia area had
ventral spots which were "very dark brown or black", and the ground
color of the belly was "bluish-gray". Ditmars (1939: 192), regard-
ing a juvenile from Marion County, Florida, wrote of "brick-red"
ventral spots. In 18 juveniles from Columbia County, Florida, the
posterior ground color of the belly is not pinkish and the ventral
spots are dark reddish-brown. In USNM 17655 from Tampa, Hills-
borough County, Florida, the ventral spots are reddish-brown. In
USNM 4736 from Micanopy, Alachua County, Florida, these spots
are brown, while in eleven others from Alachua County (WA
3647-56, and 3680) the ventral markings are light reddish-brown.
In WA 3657 from 4 miles north of Kissimmee, Osceola County,
Florida, the ventral spots are dark pinkish-orange. These few examples
indicate that the color of the ventral spots is probably clinal in the
state of Florida. Juvenile specimens from northern areas of the
peninsula usually have reddish-brown ventral spots with no reddish
suffusion posteriorly, while those from southern Florida have reddish
ventral spots with a similar color suffusion posteriorly. Thus, for
the same reason that the color of the dorsal blotches cannot be used
as a diagnostic character of the race priapus, neither can the color-

No. 6 Auffenberg: Coluber constrictor \T^>

ation of the ventral spots be so used; namely, the character is an
internal cline in Florida (fig. 16). The ventral spots of both flavi-
ventris from southeastern Louisiana and paludicolus are decidedly
reddish or orange. The general ground color of the belly is slightly
pinkish in specimens from both localities. The ventral color of
juvenile flaviventris from the northern parts of its range resembles
that of juvenile constrictor.

Throughout the entire range of constrictor, and most of that of
priapus (except south-central Florida), the ventral markings are
rather large. They are about one-fourth to one-sixth as wide as the
width of the ventral plates themselves. In priapus from south-central
Florida, as well as in paludicolus, these markings are much smaller.
They are practically never larger than one-eighth the width of the
ventral plates and are frequently absent entirely. This reduction in
the size of the ventral spots in specimens from southern Florida is
again paralleled in flaviventris from southeastern Louisiana. The
small number of live specimens which I have examined from this
area have fewer and smaller ventral markings than do specimens from
Michigan and adjacent states.

Scutellation. — Certain scales of the head and body of racers have
been shown to vary geographically ( Ortenburger, 1928: 174 ff). This
is especially true of the number of ventral and subcaudal scales, both
of which form continuous clines.

Ventral scales. — It was first pointed out by Ortenburger {ibid.)
that the mean number of ventral scales in Coluber constrictor varies
geographically as a continuous gradient. The highest means are
found in Florida and the extreme northeastern portions of the United
States, while the lowest mean count is found in southern Texas.
This last area is inhabited by the subspecies stejnegerianus. During
the early part of the century only a few specimens of this race were
available and it was thought to represent nothing more than a
variant of C. c. flaviventris. During the past few years many speci-
mens have become available and studies of this form by Mulaik and
Mulaik (1942: 13-15) indicated that the form deserves subspecific
designation. Later studies (Auffenberg, 1948: 53 ff), based on a
much larger number of specimens than were available to Ortenburger,
showed that the mean number of ventral scales of this race is lower
than that of any other subspecies in North America.

It is significant that of the specimens from Florida, those from
the southern parts of the peninsular have a higher mean number of
ventral scales than do those from the more northern portions of the
peninsula. This portion of the state is believed to have been popu-
lated from more northern localities. This, in part, constitutes the
basis of Ortenburger's hypothesis which states that low ventral scale
counts in Coluber constrictor are ancestral to higher ones. On the
whole the present study has corroborated this hypothesis, though cer-
tain complications have become evident with additional material.

In Trapido's study of the genus Storeria (1944), as well as in

124

Tulane Studies in Zoology

Vol. 2

Ortenburger's monograph of the whipsnakes and racers, the cline in
the number of ventral scales is more rapid and exaggerated in Florida
than in any other portion of the ranges of the genera involved in
these studies. This is also true of the subcaudal scale counts.

According to Ortenburger the mean number of ventral plates in
specimens from the northern parts of Florida is 177.4. For the
southern portions, this average is stated as 181.6. Figure 8 shows

WESTERN HIGHLANDS

MARlANNA LOWLANDS CENTRAL HIGHLANDS

173.2 V TALLAHASSEE HILLS

..%'.0:,"*lT7.O

176.7

Figure 8. Mean number of ventral scales for various localities in
Florida, based on data from localities from which at least five
specimens were available. The data presented on the map represent
scale counts of 872 specimens.

the results of mapping the mean number of ventral scales in 872
specimens from the state of Florida. Means are shown for various
localities, but only if at least five specimens are available from that
area. A number of other specimens (136) were examined, but not
included due to the unreliability of the counts of only a few speci-
mens scattered over rather large areas. Also drawn on the map are
the physiographic provinces of the state as outlined by Cooke (1939:

No. 6 Auffenberg: Coluber constrictor 125

14). Because the specimens from the extreme southern parts of
the peninsula are so noticeably different from those on the Keys, the
mean number of ventral scales was tabulated separately for these two
areas. By separating these counts the mean number of ventral plates
for the populations of the Lower Keys is 176.7 ±.49, while in a large
sample of specimens from Alachua County, in northern Florida, the
mean is 178.3 ±0.38. On the other hand, specimens from Dade
County, at the southern extremity of the peninsula, have a mean of
184.0 ±.36. It is obvious that the number of ventral plates of the
specimens from the Keys is much closer to those from Alachua
County in northern Florida than to the geographically closer locality
of Dade County.

The difference between the means for northern and southern Flor-
ida, as found by Ortenburger, is 4.2 scales. By separating the Keys
populations from those of the southern mainland I find a difference
between the populations of northern and southern peninsular Florida
of 6.7 scales (P=>.001). Specimens from Cape Canaveral, Bre-
vard County, which are identical in color with the Everglades race,
paludicolus, also have a higher number of ventrals than do popula-
tions of typical priapus from the mainland (fig. 8). From the Cape
the mean is 184.6 ±0.39, while for the adjacent mainland population
of priapus it is 182.0 ±0.52. However, the difference between the
means of the two populations is not significant. It should be pointed
out, however, that the samples from both of these areas are rather
small.

Series of specimens from a number of localities over the state of
Florida suggest that the cline in the number of mean ventral scales
is not simply north-south in the peninsula. In figure 8 there is shown
a Central Highlands District. This area may be termed the backbone
of the state, and is composed of significantly older formations than
are the surrounding Coastal Areas (Cooke, 1939, 1945; MacNeill,
1951, et al). Large portions of it are the site of the Floridian Archi-
pelago, formed during former high sea levels and which may be
important in the distribution of the fauna of the extreme southeastern
states. Within each one of these districts there seems to be a slight
suggestion of a north-south trend in the number of ventral scales.
Specimens from the northern portions of each district have lower
counts (though not statistically significant) than do those from the
southern parts of the district. The mean for the entire northeastern
part of the Coastal Lowlands is 180.1 ±0.42, while for Dade County,
in extreme southern peninsular Florida, it is 184.0 ±0.36. In the
Central Highlands the mean in the entire northern portion is 178.2 ±
0.32, while in the extreme southern tip of this district it is 179-6±
0.65. Specimens from the western panhandle of the state have the
lowest mean number of ventral scales of any area in Florida. This
region includes the northwestern Coastal Lowlands, the Western
Highlands, Mariana Lowlands and the Tallahassee Hills. The mean
for this entire region is 174.1 ±0.72.

126 Tulane Studies in Zoology Vol. 2

Reference has been made to specimens from Sarasota County, on
the west coast of Florida which approach intergrade coloration
(paludicolus X priapus). It seems significant that this region also
has a higher mean number of ventral scales than the surrounding
counties, and that this count approaches that of the race paludicolus.
In the region of Lake and Sumpter Counties in central Florida, the
number of ventrals (181.3) is higher than in populations to the
north or south of them. This general region also seems to be the
one in which it is believed a seaway extended from the east to the
west coasts of Florida during a large portion of the early Pleistocene.
Whether or not there is any direct correlation is unknown at present.
Specimens from the geologically recent portions of Florida, compris-
ing the Pamlico Terrace of the extreme eastern coastal strip and
much of southern peninsular Florida, have the highest mean number
of ventrals of any area in the state. The mean number of ventral
scales for all of the western "panhandle" is 174.1 ±0.72; for the
Central Highlands, 178.1 ±0.35; and for the peninsular Coastal Low-
lands, 181.8±0.38. There is a difference of 4.0 scales (P= >.001)
between the means for the western districts and the Central High-
lands; and 3.7 scales difference (P=>.001) between the means
for the Central Highlands and the peninsular Coastal Lowlands. In
the specimens from Sumpter and Lake Counties in the middle por-
tion of the Central Highlands, the mean is 181.3, while to the north
the mean for Marion and Alachua Counties is 177.5. The difference
between the two areas is 3.5 scales, but is not statistically significant.
The mean for Orange, Polk and northwestern Osceola Counties
(south of Lake and Sumpter Counties) is 176.8, the difference be-
tween them being 4.2 scales. Again, the difference is not significant,
but only suggestive. The difference between the means of speci-
mens from the Everglades and the Keys is quite great, being 7.3
(P— >.001). This difference, though actually gradual, takes place
over approximately 150 miles. Although the difference between the
mean number of ventral scales within the State of Florida is quite
small, it seems significant that the differences among the mean num-
ber of ventrals of a sample of stejnegerianus from extreme southern
Texas, and another sample of flaviventris from Nebraska is 10 scales
(Pr= >.001). This involves a distance of approximately 1,500 miles.

An isophene is drawn for the various mean numbers of ventral
scales in Florida in figure 9. This is based entirely on figure 8. Figure
10 shows the results of drawing an isophene of the mean number of
ventral scales throughout the United States. Due to the proportion-
ately lesser number of specimens used in this evaluation as compared
with that of Florida, this representation is more generalized than
that for Florida. However, it is evident that there are two areas
with very high mean ventral scales: (1) the New York, Pennsyl-
vania, Maryland and eastern Ohio region; and (2) the eastern and
southern coastal regions of peninsular Florida. The area with the
lowest mean count is in southern Texas. The pattern is one of more

No. 6

Auffenberg: Coluber constrictor

111

170

*• * .MO**'"' x-|8°

Figure 9. An isophene based on the mean number of ventral scales
of specimens from Florida, based on data obtained from figure 8.

or less concentric circles away from this area having a low mean
number of ventral scales. It will also be noted that the counts are
lower along the northern parts of the entire range of the species.
This is especially evident in the New England States, where the counts
drop from a mean of 180.1 in New York state to practically 170 in
Maine. This has been thought to be due to the effects of minimum
vital conditions found in this northern region ( Ortenburger, 1928:
231), and not an integral part of the inherent tendencies along the
paths of dispersal.

SttbcMidal scales. — The geographic variation of the subcaudal scales
is very similar to those of the ventral scales. Analysis of the available
data by Ortenburger {op. cit.) indicated that a geocline exists in the
mean number of subcaudals. The present study has substantiated
these data, at least for the eastern United States.

Figure 11 is an isophene drawn from the available material of
eastern United States. Unfortunately, a large number of specimens
have broken tails, so that even large series from one locality have

128

Tulane Studies in Zoology

Vol. 2

many specimens that cannot be utilized for an analysis of this char-
acter. As a result, the isophene based on subcaudals is necessarily
more generalized than that for the ventral plates.

Similarities between the two isophenes (figs. 10, 11) are appar-
ent. This is especially true in Florida. The western portion of the
state has the lowest mean number of subcaudals of any area in the
state (97.5 ±1.02). The southern portion of the peninsula has a
higher mean number of caudals (107.9±0.82) than does the north-
ern part of the peninsula ( 103.2 ±0.93). On the Keys the mean
number is 102.5 ±1.31, being more similar to the mean of the
population from northern peninsular Florida. It is also significant
that the isophene representing a mean number of 105 caudals is very
similar to that representing a mean of 180 ventrals in figure 9- In
New York State, where specimens also have a high number of
ventral plates, the number of subcaudals is also relatively high, but
not significantly higher than those from many other portions of the
eastern United States.

Reduction of scale rows. — Ortenburger (1928: 205) discussed the
mean point of reduction of the dorsal scale rows, especially in terms
of geographic variation. He showed that the point of scale row
reduction is "farther posteriorly in specimens from the Northeast
and from the Southeast than from the southwestern part of the
range". This is to be expected if the point of decrease is correlated
with the total number of ventral scales in this species. The mean
number of ventral scales at which this reduction occurs in specimens
from Alachua, Marion and Columbia Counties, Florida, is 108.4±
1.32; for specimens from Volusia and Seminole Counties, Florida;
it is 109.0±1.02; for the Keys, 107.5 ±1.72; and from Collier and

Figure 10. An isophene based on the mean number of ventral scales,
of specimens of Coluber constrictor from the United States.

No. 6

Auffenberg: Coluber constrictor

129

90 85

80 75 70

1

i

^—j

-~Vg>

> sC

45

//( sly \ \ ^
/// / ■ i \
/It i
6(1 /,*. i /
J / (s\ ' (

j j ( y

45

40

V /

0 5^

f r

S \ /

-v/ [ \ I ^}

40

)/J

y 4st/ — 95

35

-so ff

1 f~

~^^^^~

V^rT^ 95

35

nA

) r^

/ 95^\

30

\ -4

!/ 1

^—100

30

Ay105

sy

7°) ^>

25

i

i

<Z.P6

25

90

85

80 75

Figure 11. An isophene based on the mean number of subcaudal
scales in specimens of Coluber constrictor from eastern United States.

Dade Counties in extreme southern Florida it is 115.7 ±1.01. As
in the distribution of the number of ventral and subcaudal scales,
the populations from the Keys are more similar to those from north-
ern peninsular Florida than to those from the geographically closer
locality of extreme southern mainland Florida. Populations from
localities between the northern part of the peninsula and the Ever-
glades of southern Florida have the scale reduction occurring between
the 108th and 115th ventral plates and exhibit a geocline in that the
scale reduction is found farther back on the body as populations are

130 Tulane Studies in Zoology Vol. 2

encountered which have progressively higher ventral scales. When
the mean number of ventral scales of any population is divided by
the mean ventral scale at which the scale reduction occurs, the result-
ant ratio varies from 1.6 to 1.8.

Osteology. — During the latter part of this study some skulls of
Coluber constrictor were examined. Certain variations became evi-
dent, but because of the small number of prepared crania most of
these cannot be discussed at present with any degree of certainty.
However, one character-gradient dealing with the bones of the skull
is rather obvious. This is the shape of the premaxillary.

Premaxillary shape. — Bogert (1947: 11) has mentioned the use of
this bone as a possible indicator of phylogenetic relationships among
various genera and species of snakes. In Coluber constrictor the
premaxillary is roughly "skate-shaped", having two "wings" extended
out in a horizontal plane, perpendicular to the axis. At the posterior
inner edge of these flat, lateral "wings", are two posterior projections
of the processes.

An examination of 32 specimens for this character has disclosed
that a geocline seems to exist from mormon and stejnegerianus
through flaviventris to priapus. This cline is concerned with the
shape and plane of the axis of the two posterior processes mentioned
above. In flaviventris, mormon and stejnegerianus these processes
are rather short, being shortest in the last two subspecies. In all of
these races they lie in a nearly horizontal plane, but decidedly less
so in flaviventris. In constrictor they are in a horizontal plane, but
longer than in flaviventris. They are longest of all in southeastern
specimens of priapus and paludicolus, and are, furthermore, twisted
in these two races so that the outer edges of the two posterior pro-
cesses have been rolled upwards, causing them to lie in a vertical,
rather than a more or less horizontal plane. The shape and nature
of the axis of these processes varies both geographically and in-
dividually.

The shape of the lateral "wings" of the premaxillary may also vary
individually and geographically. In the western races the tips of
these processes usually form less acute angles than do the tips on
specimens from the east, and especially the southeast. In some of
these specimens this tip may be formed into a sharp, slightly re-
curved hook.

In the description of a fossil Pleistocene snake, Coluber acuminatus,
Cope (1899: 197) described the shape of the premaxillary as "sub-
conic", and "narrower and more protuberant than in Z. constrictor
. . . with which I have compared it". The degree of protuberance
of the anterior tip of this bone is variable, individually, geographically
and ontogenetically. Adult specimens from the southeastern states
have a much more protuberant profile than specimens from the west-
ern states. Juvenile specimens from the southeastern states have a
less protuberant profile than do the adults from these areas and are

No. 6

Aujjenberg: Coluber constrictor

131

similar to adult specimens of more western races.4

Dentition. — That the number of teeth in this species varies geo-
graphically has already been shown by Ortenburger {op. cit.). More-
over, the number of teeth increases from the western races to those
of the eastern parts of North America. Figure 12 shows the range
of variation in the number of dentary, palatine, pterygoid and maxil-
lary teeth in the races of Coluber constrictor.

U

»-

UJ
00

2
z>

NUMBER OF
SPECIMENS

203

□

31

PALATINE

DENTARY

56

51

68

MAXILLARY

PTERYGOID

Figure 12. Range of variation in the number of teeth in some of
the subspecies of Coluber constrictor. The relationship of the number
of palatine and dentary teeth is especially interesting. In priapus
the numbers overlap slightly, while in the other races the numbers
of these teeth are discrete. There is an apparent increase in the
number of palatine teeth from the western to the eastern races.
There is also a general increase in the numbers of other teeth. The
decrease in number of dentary teeth from constrictor to priapus and
paludicolus is interesting in view of a general increase from mormon
to constrictor (*data from Ortenburger, 1928: 174).

4 A paper by the author (in print) compares the variation in the
skeletal elements of Coluber constrictor with Coluber acuminatum,
and shows that acuminatus should be referred to the synonymy of
Coluber constrictor.

132 Tulane Studies in Zoology Vol. 2

Onto dines 5

Color of the iris. — The color of the iris in specimens of Coluber
constrictor is important. Conant (1942: 195) has suggested that
the red eyes observed in both adult and juvenile racers from Florida
may be characteristic of the race priapus, and may be of value in
distinguishing both sexes of this subspecies from constrictor. The
latter usually have brown-colored eyes. The range of variation, either
individual or geographic, was practically unknown, although Perkins
(1949: 10) suggested that the degree of consistency in this char-
acter needed more confirmation, especially in adults.

In all living examples of constrictor, flaviventris and priapus, in
the more northern and western portions of its range, the iris is usually
brown. Sometimes it may be dark amber, tan or reddish-brown,
but never distinctly reddish or orange. Specimens with a red, or
orange iris occur in Florida. However, analysis of this character is
complicated by the simultaneous existence of a geocline and an onto-
genetic change in the color of the iris in specimens from peninsular
Florida.

Of the living adults that I have examined from the extreme north-
ern portions of Florida, none possesses red eyes. However, of eleven
newly hatched specimens from Columbia County, which is situated
in the northern part of the peninsula, four possessed red eyes, while
those of the remainder were brown. Both parents of these hatch-
lings (WA 3345-6) had brown-colored eyes. The female parent
of nine red-eyed hatchlings from Alachua County, which borders
Columbia County on the south, was also brown-eyed. Seven other
newly hatched specimens from Columbia County and six others from
Alachua County had brown eyes. In eight young from Kissimmee,
Osceola County, in central Florida, seven had red eyes and those of
the remaining snake were brown. Moreover, there were many adults
from Osceola County that had red eyes, while in Alachua County
there were only a few. Of 38 living specimens of paludicolus from
southern Florida, practically all of the juveniles and adults had red
eyes. Those that were not red were yellow, or rarely reddish-brown.
In the case of specimens from northern Florida, the number of young
with bright red eyes (25.3%), and the fact that only a very few
red-eyed adults were taken (5.2%) seem to indicate that this char-
acter is, at least in part, ontogenetic. That is, most red-eyed juveniles
simply change to brown-eyed adults. This is somewhat substantiated
by the brown-eyed parents of red-eyed juveniles. In living speci-
mens from the Kissimmee area, not only are red eyes more frequent
in newly hatched specimens than in those from northern Florida
(69-8%), but many retain this color into maturity (58.2%). In
extreme southern Florida, practically all of the juveniles possess red
eyes (78.4%). None seem to change this color, for a large number

5 An ontocline may be either a geo- or ecocline, and at the same
time an external or internal cline. It is a condition in which an onto-
genetic change varies geographically as a character-gradient.

No. 6

Auffenberg: Coluber constrictor

133

of the adults examined were also red-eyed (75.3%). The relation-
ship of the color of the eyes in juveniles and the geographic locality
of various populations is shown in figure 13.

That the eyes of juvenile snakes from northern peninsular Florida
change color is strongly substantiated by living specimens of all ages
in which all gradations from red to brown have been observed. In
specimens with red eyes the first indication of change is in the lower
half of the iris, next to the pupil (fig. 14). Here there is formed a
reddish-brown crescent that gradually gets larger. In specimens of
about 30-40 inches there is frequently noted a red crescent on the
upper portion of the iris, while the remainder is brown. This cres-
cent in the iris of adult specimens is probably an indication of its
previously red-colored iris as a juvenile.

NUMBER OF SPECIMENS

13

14

28

15

14

-J

O

I-

u.
o

o

<

CD

-I
O
O

<
3
I
O
<

OB <

CO
UJ 3
* -J

< o
_l >

<0 w

UJ

o

z
<

o

o

Z
ui

UJ

£uj

0>

Ui

UJ

o

<

UJ

o
cc

z

O

Figure 13. Geographic variation in the coloration of the iris in
juvenile specimens of Coluber constrictor from Florida. Note the
gradual increase in the number of red-eyed specimens from Columbia
County to Dade County, Florida, and the similarity between the
specimens from the Outer, or Lower Keys and northern peninsular
Florida.

134

Tulane Studies in Zoology

Vol. 2

©

I.

5.

6.

Figure 14. Ontogenetic changes in the coloration of the iris of speci-
mens of Coluber constrictor. Shaded areas represent regions of
the iris which are dark brown in life. Clear areas represent regions
which are bright red or yellow in life. The eyes, numbered 1-6,
show the change of red-eyed hatchlings through successive stages
in which the brown color increases, to the practically all brown eye
of the adult. Faint traces of red are usually found above the pupil
in even very large adults (6).

Another interesting parallel between the race paludicolus and speci-
mens of flaviventris from southeastern Louisiana is that the iris of
juveniles from both areas is usually reddish. All the live juvenile
specimens of flaviventris which I have seen from Michigan and Ohio
have brown eyes (fig. 16).

Dimorphism

Geographically variable dimorphism. — A character not mentioned
by any earlier workers, and of frequent occurrence in the southeastern
part of North America, is the contact of the first supralabial and the
loreal. This is made possible by an elongation of the upper posterior
corner of the first supralabial making contact with the lower, anterior
corner of the loreal. It may be present on one side, but not on the
other.

The presence of a contact between these two scales is found mainly
in specimens from the southern states, and then only in a fair degree
of regularity. However, it seems to be generally absent in specimens
from the northern portions of the ranges of C. c. flaviventris (except
in Michigan, and Ohio, where it occurs in 12.2 percent) and C. c.
constrictor. It is absent in the few specimens of flaviventris that I

No. 6 Auffenberg: Coluber constrictor 135

have examined from the central parts of its range, as well as in all
specimens of the race mormon* It occurs in at least 10% of all
specimens of priapus except those from extreme eastern North Caro-
lina (fig. 16). I have found this contact in 6.6% of the specimens
of constrictor from north of the Fall Line in Georgia, and in 7.1%
of the specimens from Tennessee. It does not seem to occur in any
other portions of the range of this form.

In the southern portions of the states of Mississippi, Alabama,
Georgia and South Carolina, this contact occurs in at least 15% of
the specimens, while in northern peninsular Florida it occurs in
10-35% of the specimens. In the central part of the peninsula it
occurs in 25-45%, while in the southern counties it is found in
40-50% of the individuals.

Although a cline is indicated by the frequency with which this
contact is found, this cline is not steady and gradual. Various in-
consistencies occur throughout the peninsula, such as, for example,
the counts obtained from specimens from northern Florida. Of one
hundred and seventy-eight specimens from Alachua County 45.9%
have a contact between these two scales. However, in specimens
from adjacent counties it occurs in only 20-30% of all of the in-
dividuals examined. In Volusia County, from which 134 specimens
are available, it occurs in only 10%, while in the surrounding counties
it occurs in closer to 30% of the individuals examined (102). In
Pinellas County it is found in 50% of the specimens, while in adjacent
counties it occurs in 26.9-34.3%.

The means for the degree of frequency in which such contact
occurs for northern, central and southern Florida are, respectively,
28.2%, 33.8% and 44.2%. Thus, there is a tendency for more
specimens to have a loreal-first supralabial contact in southern Florida
than in the northern parts of the state. Furthermore, such contact
seems confined to the southern half of the United States, and is more
frequent below the Fall Line (fig. 16). It also seems significant that
the population from the Florida Keys have this contact in a degree
of frequency (27.3%) which is more similar to that of northern
Florida (28.2%) than to that of southern Florida (44.2%).

Sexual dimorphism. — Aside from rather obvious sexual dimorphism
such as that concerning the reproductive organs, certain other char-
acters are evidently related to the sex of the individuals. However,
most of these differences are not obvious, and are based on mean
values of significantly large samples.

Scutellation. — The only scale character which I have found exhibit-
ing statistically significant sexual divergence is the number of caudals,
but only on the basis of local populations. A comparison of the
mean number of caudals for males and females of large samples
from restricted localities indicates that differences in the means are
significant (P= >0.001). On the other hand, when the means for
all of the males and females are analyzed, the observed differences
are not significant (P= <0.03).

136 Tulane Studies in Zoology Vol. 2

Other scale characters, such as the point of reduction of dorsal
scale rows and the total number of ventrals, have been mentioned by
Ortenburger (1928: 185) as showing sexual dimorphism, expressed
by mean values. However, on statistical analysis such differences as
might exist in these characters could be due to effects of random
sampling (P= <0.03).

A valuable statistic, the coefficient of divergence, has been given
for a few species of snakes. A comparison of the coefficients of
divergence that have been determined for the ventrals and caudals of
two species by Klauber (1940: 208), two races by Grobman (1941:
21) and one by Dowling (1950: 21), as well as those of the eastern
races of Coluber constrictor, are shown in Table 2. As in other species
for which such coefficients are available, Coluber constrictor exhibits
more divergence in the mean number of caudals than in the number
of ventrals. In fact, sexual differences in the mean number of
ventrals do not approach statistically significant levels, even for large
samples from restricted localities (P = 0.8-0.3).

TABLE 2.
Coefficients of Sexual Divergence

Number of
Form Specimens Ventrals Caudals

Lampropeltis g. californiae
Opheodrys v. vernalis
Opheodrys v. blanchardi
Phyllorhynchus d. perkinsi
Seminatrix p. pygaea
Coluber c. constrictor
Coluber c. priapus
Coluber c. flaviventris
Coluber c. paludicolus

Although the coefficients of divergence of the mean number of
ventral scales of the eastern races of Coluber constrictor seem to form
a group, the same is not true of the caudal scales. The coefficient
of sexual divergence in the mean number of caudals in Coluber c.
paludicolus is quite different from those of the other subspecies, and
the low value of this statistic may be of some diagnostic importance.

Size and proportion. — Very few specimens were measured during
the course of this study, partly because of the difficulty encountered
in obtaining accurate measurements of preserved specimens. How-
ever, measurements of 103 living specimens seem to indicate that
females attain a larger size than do males of the same subspecies.
The mean length of 51 male specimens of Coluber c. priapus is 943
mm, while the mean for 26 females of the same race is 955 mm.
The largest specimen of this race that I have examined is a male,
1473 mm in total length. Of 13 male specimens of paludicolus, the
mean is 951 mm, while for eight females it is 1043 mm. The largest
specimen is a female, 1689 mm in total length. The mean length
of 13 male specimens of constrictor is 949 mm, while for 12 females

271

.009

.073

391

.006

.147

56

.072

.163

225

.073

.256

203

1

.228

278

.008

.068

1008

.010

.049

102

.012

.071

58

.017

.008

No. 6 Auffenberg: Coluber constrictor 137

it is 999 mm. The largest example of this race which I have measured
is a female, 1588 mm in total length.

Ortenburger (1928: 186) has shown that there is considerable
sexual dimorphism in proportionate tail length; "the male has a much
longer tail than the female; the averages are 0.254 and 0.233 respec-
tively". He has also shown that there is a positive correlation between
the number of caudals and the proportionate tail length within each
sex.

Environmentally Produced Characters

For the most part practically all of the characters used in the
description of variations of snakes, either individual or geographic,
are assumed to be genetic in origin. However, Fox (1948) has
shown that the number of scales on various parts of at least certain
species of snakes may be influenced by environmental conditions.

Character-gradients, or geoclines, are frequently found to have as
their basis of expression, environmental factors which exist in a
similar gradient. This has been shown both experimentally and in
field studies for a number of species of fishes (Taning, 1952; Hubbs,
1922a, 1922b, 1940; Jordan, 1893; Schultz, 1926). In all of these
studies temperature was of utmost importance in the expression of
meristic characters. It is entirely within the realm of possibility that
the number of caudal and ventral scales in Coluber constrictor, as
well as other clinal characters, are environmentally influenced during
development of the embryo within the egg, either by temperature
alone, or in combination with other ecological factors. Aside from
these possibilities of environmental influence on the embryo of young
racers, it was found that such factors may affect the appearance of
racers after they are hatched. This change is concerned with the
color of the ventral surface.

Coloration of the ventral surface. — Cope (1900: 793) mentioned
black punctations on the ventral surface of a racer from Galveston,
Texas (USNM 10481). This coloration is a common phenomenon
in many individuals from southern United States, even being evident
on the black bellies of many specimens of priapus. These marks are
rarely encountered in specimens of flaviventris or constrictor from
the northern parts of their ranges. I have never noticed such marks
on newly hatched or juvenile specimens. They are present on adult
specimens of flaviventris from east central Texas, and are very evi-
dent on adult anthicus. The latter frequently has such marks on the
first few rows of dorsal scales as well as on the ventral surface. They
are also common in many adult specimens of priapus from southern
Mississippi, Alabama, Georgia, South Carolina and Florida. Many
specimens of paludicolus also have them (fig. 16).

Close examination of these minute black punctations shows that a
very large number of them contain a small mite (Ophioptes sp., near
tropicalis Ewing), which is buried beneath the noncellular outer
layer of the skin. Presumably they are not discarded when the skin
is cast. In preservation the mite is pale buff to white, and the black

138 Tulane Studies in Zoology Vol. 2

punctations are composed of very minute black specks or streaks.

The taxonomic implications of this phenomenon are intriguing.
The distribution of specimens with black punctations on the venter
may indicate the Coastal Plain range of the mite. Of more signifi-
cance is the fact that this distribution may reflect a physiologically
variable character in which the racers from southern United States
may be more susceptable as host animals to the parasitic mite, and
that the mite is actually found over a much greater range.

The influence of the environment on cold-blooded vertebrates, and
especially reptiles, is very little known. Of particular importance to
the taxonomist interested in these groups is its effect on meristic
characters. Jordan (1893) found a general correlation between water
temperature and the number of vertebrae in marine fishes, while
Schultz (1926) showed the existence of a similar phenomenon in
the fresh water fish Notemegonus crysoleucas. Both of these studies
were based on field data and indicated that most species of fishes
tend to have a higher number of vertebrae or fin rays with decreasing
temperatures. Hubbs (1922a and b, 1940) and Taning (1952)
worked with specimens in which the temperature during the de-
velopmental period was controlled. The results of these experiments
tended to substantiate theories based on field data. It was also found
that decreasing oxygen pressure and rapid changes in temperature
during the sensitive period of embryonic development produced a
difference in the mean number of vertebrae (Taning, 1952). Field
observations concerning the effect of salinity on meristic characters
have been rather inconclusive.

In reptiles the only studies of this nature have been on one
ovoviviparous species of snake, Tbamnophis e. atratus (Fox, 1948).
The subject is of vast importance in the study of herpetology in view
of our present practice of utilizing the mean number of scales in the
description of nominal populations. As in fishes, racial studies in-
volving geographical variation are more difficult than previously sup-
posed because of the possibility of dependence of some meristic
characters on environmental factors. The subject affords many op-
portunities for worthwhile contributions to our knowledge of variation.

UNCORRELATED VARIATION

In this section are discussed all of the external morphological
variations which do not seem correlated with age, sex, environment
or geographic locality.

Scutellation
Dorsal scales. — For the most part, practically all of the specimens
of Coluber constrictor have a dorsal scale formula of 17-17-15 rows.
However, occasionally specimens are examined that have a vertebral
reduction in addition to the normal lateral type. Ortenburger (1928:
204-205) stated that this type of reduction occurred only rarely in
Coluber constrictor, being found in only 4 out of over 400 examined.
However, of 86 specimens of priapus that I have examined for this

No. 6 Auffenberg: Coluber constrictor 139

character, I find that it occurs on the body in 5.6%. A vertebral
reduction occurs over the anal plate in 34.3%, and on the base of the
tail just behind the anal plate, in 60.1%. A typical dorsal scale
formula of a specimen possessing a vertebral reduction on the body
is given in the manner suggested by Dowling (1951: 131-133). The
specimen is from Columbia County, Florida.

3 4(2) 3 4(8) 3 4(100) 7 7

21 19 17 15 14 13(178)

3 4(3) 3 4(9) 3 4(102) 8(146) 7(148)

The high incidence of a vertebral reduction in Coluber and the
fact that the race ruthveni of Masticophis taeniatus is more or less
characterized by having no vertebral reduction (58% of the speci-
mens have a scale formula of 15-13 fide Ortenburger, 1928: 87)
lends support to the proposition that perhaps the two genera should
be united, and not given generic distinction as indicated by Inger
and Clark (1943: 492). No subspecies of Coluber constrictor is
characterized by having a vertebral reduction, although the condition
seems to occur sparingly throughout the range of the species.

Scales between the chin shields. — The presence or absence of small
scales between the posterior chin shields of specimens is also of
taxonomic importance (Inger and Clark, ibid.). A small proportion
of the specimens (24.5%) examined do not have these scales present
between the chin shields. When present, they vary from 1-8.

Supralabials. — The number of supralabials in this species may vary
from 3-12. Both extremes are rather rare and are produced by a
considerable amount of fusion or splitting of the scales. They usu-
ally range from 6-8. In the eastern races they are usually 7 (mean
7.2). Occasionally 8 supralabials (7.1%) are found in specimens
from eastern United States. Although this number is characteristic
of two races (stejnegerianus and mormon) from the western parts
of North America, it is not found in any definite area in eastern
United States in a higher frequency than in any other, and the number
seems uncorrelated with locality in this area. The mean number for
flaviventris is 7.11 ±0.11, for priapus it is 7. 10 ±0.09, for paludicolus
7.05±0.27, and for helvigularis 7.01 ±0.34. The possibility of sexual
dimorphism is discussed under infralabials.

Infralabials, — The number of infralabials in the races of Coluber
constrictor from the eastern portion of North America varies from
6-12, usually 8-9, mean 8.36±0.31. The occurrence of a number of
infralabials other than 8 or 9 does not seem correlated with any geo-
graphic region. The mean number for flaviventris is 8.75 ±0.57, for
constrictor 8.31 ±0.82, for priapus 8.32 ±0.48, for paludicolus 8.41 ±
0.78 and for helvigularis 8.51 ±0.92. The suggested geographic
variation is not statistically significant (P = 0.03).

Ortenburger (1928: 186 //) mentioned the possibility of a slight
sexual dimorphism in the number of supralabials and infralabials,
at least in the race flaviventris. However, statistical analysis shows
that the observed results could have been obtained from random

140 Tulane Studies in Zoology Vol. 2

sampling ( P = 0.3 ) • Table 3 shows the mean number of infrala-
bials and supralabials for a sample of flaviventris from Michigan,
Ohio and Indiana, and a sample of priapus from northern Florida.
The mean numbers of these scales in males and females are not sig-
nificantly different and the coefficients of divergences are low, as
would be expected ( P = 0.8-0.3 ) .

TABLE 3.
Number of Supralabials and Infralabials

Subspecies

Sex

Number of
Specimens

Supralabials
Mean C. D.

Infralabials
Mean C. D.

Coluber c.
flaviventris

9

$

59
53

7.12
7.08

.005

8.66 .017
8.51

Coluber c.
priapus

9
S

154
231

7.12
7.08

.005

8.36 .014
8.28

Postoculars. — The number of postoculars varies from 2-4 (mean
3.01 ±0.08). The occurrence of either 2 or 4 postocular scales does
not seem correlated with any geographic area, as it occurs rarely
throughout the entire range of the species. The small differences
among the mean number of these scales for specimens from various
parts of the United States do not reach statistical significance.

Preoculars. — The preoculars are usually 2, rarely 1, mean 1.97 ±
0.06. The occurrence of 1 preocular is not correlated with any geo-
graphic area of North America.

Loreal scale. — The loreal scale is usually single (97.01%) but may
be divided into two or three smaller scales. The shape is usually
square to slightly higher than long. The posterior edge may be
straight or slightly convex. Variations as might exist seem indi-
vidual rather than geographic.

Temporals. — The number of temporals is quite variable. They are
usually disposed in three pairs, one behind the other, indicated by
the formula 2-2-2 (80.5%). Four pairs are also common (16.9%),
shown by 2-2-2-2, or in two pairs (1.3%), indicated by 2-2, or even
five pairs (1.3%), 2-2-2-2-2. Frequently an additional scale may be
present in any vertical row, in which case the formula is expressed
as 2-2-3, 2-3-2, or 3-2-2. Three scales may occur in more than one
row, as 3-3-2, 3-3-3, or 2-3-3. Occasionally the pair of scales in
each vertical row may be reduced to form a condition expressed as
1-2-2, 2-1-1, etc. Frequently an additional scale is present between
two pairs of scales, so that the formula is expressed as 2-2-1-2, 3-1-2-2,
etc. Practically all combinations of numbers from 1-5, either in
pairs or not, are found. There is no apparent correlation with geog-
raphy. The lowest number of temporals which I have counted is
expressed as 2-2, the highest as 3-3-2-2 and 2-3-3-2, both totaling
ten scales in all.

Color

In general, the coloration of individuals, both juveniles and adults,
is correlated with geographic locality and has been discussed previ-

No. 6 Auffenberg: Coluber constrictor 141

ously. However, certain facets concerning the coloration of the
eastern races, such as the manner in which the juvenile pattern changes
to that of the adult, seem uncorrelated with geographic locality.
This is also true of aberrant colors of both young and adult specimens.

Ontogenetic color and pattern changes. — It has been mentioned that
the transformation of color and markings of juvenile specimens,
which are blotched, to the adults, which are uniform in color, follows
rather basic patterns. On the dorsal surface the markings are first
obliterated posteriorly. This suffusion slowly moves anteriorly. On
many newly hatched specimens, such posterior suffusion is already
present, so that accurate counts of the total number of blotches is
difficult. Numerous authors have reported that these markings dis-
appear completely during the second or third year. The total length
at which this suffusion is complete is about 500 mm. WA 3423
from Astor Park, Lake County, Florida, which is 458 mm in total
length, has a dorsal surface spotted with faint dark brown blotches.
In WA 3204 from the same locality and 519 mm in total length,
there is no indication of any dorsal markings. A specimen from
Winter Haven, Polk County, Florida (SRT 250), which is 623 mm,
does not have any visible dorsal markings. In USNM 4423 from
Georgetown, South Carolina, which is 526 mm in length, the ventral
spots are quite evident while the dorsal spots are completely suffused
by black. Practically all of the specimens which I have examined that
are under 480 mm in total length, had visible dorsal spots, while
most of those over 600 mm showed no indication of such markings.
McCauley (1945: 74) stated that in Maryland the largest specimen
that he has examined that had noticeable dorsal blotches was 584 mm.
The smallest one that was completely transformed was 640 mm.

In the subspecies with a dark, slate-colored belly (priapus, con-
strictor and helvigularis) there is frequently observed a light stripe
extending down the middle of the ventral surface. Loennberg ( 1894)
described such a specimen from Arcadia, DeSoto County, Florida, in
which there is a median stripe of small white, cloudy spots. This
is a consequence of the method by which the ventral suffusion takes
place. A specimen 521 mm in total length, from 4 miles north of
Kissimmee, Osceola County, Florida, was kept alive for a long period
and notes were taken on color changes that occurred over a period
of nine months. Ventrally this specimen was bluish-gray with
pinkish-orange ventral spots (Sept. 14, 1951). By November 16
there was a noticeable area of dark grayish-blue on the lateral edges
of the ventral plates in the mid-body region. This left a lighter
stripe of the original ground color down the center of the belly.
Anteriorly, and on the underside of the tail, the specimen was still
light-colored. By April, 1952, it was 617 mm in total length. The
darker pigment which was slowly suffusing the midbody outer sur-
faces of the ventral plates was slaty-black and the light mid-ventral
stripe became very evident.

This observed change is also borne out by many living and pre-

142

Tulane Stttdies in Zoology

Vol. 2

served specimens. In most adult individuals the belly is slaty-black
with only the chin and first few ventral plates being white. The
underside of the tail is also frequently white. This coloration seems
typical of most large specimens from peninsular Florida. Large in-
dividuals from New York, New Jersey and other northeastern states
usually have much more black on the chin, throat and underside of
the tail. In many specimens from such widely separated states as
New York and Florida, the anterior portion of the underside of the
body and throat may be white, blotched with black. A similar con-
dition is also frequently noted on the underside of the tail. McCauley
(1945: 74) reported a specimen from Allegheny County, Maryland,

m®-.

%&¥ •'*&£ Mm

■:;-.v~:*" ■■■;■ ..ri

&£:*y?y?&

v •,[" "~T;. m

l~s;v-*-~

^-.•■;tV»//.>.t

W^i*

5S*

S5.Vfi5VSi.

3

'•$&£5

B

Figure 15. Ontogenetic color and pattern changes in subadult speci-
mens of Coluber constrictor from eastern United States. Black
blotches on the anterior portion of the belly of young subadults (A)
slowly get larger. This causes light, faint reticulations to be formed
in older specimens (B). The final stage is complete fusion of the
blotches, causing a uniform coloration.

No. 6 Auffenberg: Coluber constrictor 143

1130 mm in length, which was mottled with black about 100 mm
down the belly. UMNZ 57059 from Enterprise, Volusia County,
Florida, has very large blotches on the anterior portion of the ventral
surface. The same is true of AMNH 6719, labeled "Arkansas".
WA 3561 from Astor Park, Lake County, Florida, is similar. None
of these specimens are full-grown, but are about 760 mm in total
length. They all come from areas in which full-grown specimens
have a black throat and belly. This would seem to indicate that
while the mid-body ventral suffusion takes place over a large area
on the lateral surfaces of the ventral plates, in the region of the
throat this suffusion takes place by a gradual growth, and finally, a
coalescence of slaty-black blotches. In many specimens from Florida
that are between 760-1020 mm in total length, the anterior portion
of the belly and throat is mostly black, but with faint reticulations
of bluish-white. These specimens represent the next to the last stage
in the ventral ontogenetic color change, in which the black blotches
are practically completely fused (fig. 15).

In the races that have a light-colored ventral surface (i. e., flavi-
ventris, anthicus, mormon, stejnegerianus, and paludicolus) the ven-
tral markings of the juvenile may be retained for a long time. This
is especially true near the anterior end. Pope (1944: 172) men-
tioned a male flaviventris 1067 mm long which possessed traces of
the juvenile spots. I have seen two specimens from Michigan that
were 914 and 1001 mm long on which traces of ventral markings
were quite noticeable anteriorly. Occasionally specimens of paludi-
colus also retain this juvenile ventral coloration anteriorly, though
always faint as in the other subspecies.

Abberrant head color. — Occasional individuals are examined that
have the supralabials, infralabials and sometimes the chin, variously
colored, usually some shade of yellow, or even orange in some cases.
Although occurring rarely, they are found throughout the range of
the species, at least in the eastern portions of North America.

Acknowledgements

This research could not have been carried out without assistance
from a great many sources, and it is indeed pleasant to acknowledge
my indebtedness to the persons and institutions mentioned below.

My thanks are especially due to those numerous individuals not
directly concerned with the study of herpetology, but who, when the
opportunity presented itself, collected and sent many live specimens.
In particular I wish to thank Mr. Wm. J. Epps, Sanford, Florida; Mr.
Kenneth Hansen, formerly of Stetson University; Mr. Wm. Lanigan,
Wauchula, Florida; and Mr. Harold J. Smith, Fort Pierce, Florida.

I am primarily grateful to Dr. Arnold B. Grobman, University of
Florida, who directed the later portion of this study with unfailing
interest. To Doctors Coleman J. Goin, W. C. Allee, H. K. Wallace
and R. A. Edwards, all of the University of Florida, I am indebted
for many constructive suggestions. To Dr. Archie F. Carr, Jr., of
the University of Florida, I wish to show appreciation for color notes

144

Tulane Studies in Zoology

Vol. 2

o c

m 5 c

fn 3 3

0> O '

No. 6 Auffenberg: Coluber constrictor 145

on this group which he made available to me. I wish to thank Messrs.
E. Ross Allen and Wilfred T. Neill, Silver Springs, Florida, who have
made every attempt to obtain specimens from critical areas in the
Southeast. I also wish to thank Dr. James A. Oliver, now at the
New York Zoological Park, and Mr. John W. Crenshaw, University
of Florida, for valuable criticisms.

For the loan of specimens, use of data, or for the gift of specimens,
I owe thanks to Dr. Doris Cochran, United States National Museum;
Mr. Arthur Loveridge, Museum of Comparative Zoology, Harvard
University; Mr. Charles M. Bogert, American Museum of Natural
History; Dr. Karl P. Schmidt and Mr. Clifford H. Pope, Chicago
Natural History Museum; Doctors Charles Walker and Norman Hart-
weg, Museum of Zoology, University of Michigan; Mr. N. Richmond,
Carnegie Museum; Dr. Fred R. Cagle, Tulane University; Dr. Ralph
L. Chermock, University of Alabama; Mr. Roger Conant, Philadelphia
Zoological Gardens; Dr. C. B. Segars, Mississippi Southern College;
Mr. W. E. Haast, Kendall, Florida; Mr. and Mrs. Lewis H. Babbitt,
Petersham, Mass.; Messrs. R. M. Johnson and R. E. Etheridge, for-
merly of Tulane University; Dr. W. F. Blair and Messrs. Wm. Mil-
stead and J. Mecham, University of Texas; Mr. J. Hanahan, Charleston
Museum; Mr. M. Hopkins, Fitzgerald, Georgia; and Messrs. J. W.
Crenshaw, R. Highton, S. Telford, K. Hansen, R. Leake and the many
others at the University of Florida who have been so helpful in many
diverse ways.

I would also like to thank Dr. Edward W. Baker, United States
National Museum, for identification of some mites.

Summary

As now recognized, the genus Coluber includes four species. These
are:

Coluber oaxaca (Jan) 1863: 63. A species known from a very
few specimens. It has been recorded from the Mexican states
of Oaxaca and Colima.

Coluber spinalis (Peters) 1866: 91. An Asiatic form of northern
China and Mongolia.

Coluber acuminatus (Cope) 1899: 197. A fossil species which is
reported from the Pleistocene of Montgomery County, Penn-
sylvania. Examination of the skulls of a few specimens of
Coluber constrictor seems to indicate that there is reason to
doubt the validity of the fossil species.

Coluber constrictor Linneaus 1758: 216. This is a polytypic
species dispersed throughout most of North America. It is
found as far north as Halifax, Canada, and as far south as the
Florida Keys on the east coast. On the Pacific coast it occurs
from British Columbia to extreme southern California. In
middle North America it is found from North Dakota to
southern Texas, and in the eastern Gulf Coastal Plain of
Mexico to the state of Vera Cruz, thence to Guatemala.

146 Tulane Studies in Zoology Vol. 2

Included in the species Coluber constrictor are eight races. These
are:

Coluber c. mormon Baird and Girard 1852: 351. No intergrada-
tion is known between this race and its closest relative geo-
graphically (flaviventris), and is included as a subspecies
because of its extreme similarity to other members of the
group. Although mormon and flaviventris have been found
within 130 miles of one another (Stuart, 1930: 44), large
series will have to become available from suspected areas of
intergradation before intermediate populations can be demon-
strated. Undoubtedly such series will show that intergrada-
tion exists, at least in Montana, and probably along the Front
Ranges of the Rocky Mountains in other states farther south.
It is distributed from southern British Columbia to extreme
southern California and from the Pacific Coast to the Rocky
Mountains. It seems absent from the Sonoran Physiographic
Province of southeastern California and western Arizona.

Coluber c. stejnegerianus (Cope) 1895: 797. A subspecies of ex-
treme southern Texas and Mexico, being found from Nueces
County south in the Gulf Coastal Plain to Vera Cruz. Coluber
c. ortenburgeri Stuart is referred to the synonymy list of this
form, as there seems to be no reliable character with which to
separate the two races, thus extending the range of stejnegeri-
anus to Guatemala.

Coluber c. flaviventris Say 1823: 185. A subspecies of the Great
Plains and Interior Lowlands as well as restricted portions of
the Gulf Coastal Plain in eastern Texas and southern Louisi-
ana. Its northern limits of distribution are from Montana to
North Dakota, southern Minnesota and Lower Michigan. It
extends westward as far as the Front Ranges of the Rocky
Mountains and the Edwards Plateau in southern Texas. In
the northern parts of its range its distribution has been
shown to be correlated with the prairie provinces in both
Ohio and Minnesota by Conant (1938: 50) and Breckenridge
(1944: 116). Further collecting will probably show that
this is true of the remainder of the northern parts of its range.
Specimens of this race, although so far reported only from
Ontario in Canada (Mills, 1948: 10), will probably be found
in the southern portions of Saskatchewan and Alberta. Inter-
gradation between this race and constrictor is to be sought
along the northern shore of Lake Erie in southern Ontario.

Coluber c. constrictor Linnaeus 1758: 216. This is a subspecies
of the eastern forests. It does not seem to range above 4,000
feet elevation in the Appalachian Mountains. It is found
from Halifax, Canada, south to the Fall Line; west to the Mis-
sissippi Lowlands and the unglaciated portions of Ohio. It
seems absent from the non-coastal areas of Maine, the greater
part of New Hampshire and Vermont and the Adirondack

No. 6 Aufjenberg: Coluber constrictor 147

Mountains. The distributional patterns of this race and jlavi-
ventris are nearly congruent with those of many other groups
of snakes (Schmidt, 1938; Ruthven, 1908; Grobman, 1941;
et. al.) .

Coluber c. priapus Dunn and Wood 1939: 4. A lowland form
of southern United States being found in the Atlantic and
Gulf Coastal Plain from North Carolina to Lake Pontchar-
train, Louisiana, and north in the Mississippi Lowlands to ex-
treme southwestern Ohio. Considerable interdigitation seems
to exist in Arkansas between this race and flaviventris. It
is a syngeograph of many southeastern snakes, including
Farancia abacura, Agkistrodon piscivorus, Crotalus h. atricau-
datus, etc., as well as many frogs, such as Rana grylio, Hyla
cinerea, etc.

Coluber c. anthicus (Cope) 1862: 238. A subspecies found in
western Louisiana and eastern Texas.

Coluber c. helvigularis subsp. nov. This is a form inhabiting an
area along the Appalachicola and Chipola Rivers in western
Florida. Lampropeltis g. goini seems to have a somewhat
equivalent range.

Coluber c, paludicolus Auffenberg and Babbitt 1953: 44. A race
mainly confined to the Everglades of southern Florida. An
outlying population occurs on Cape Canaveral, Brevard County,
Florida. Geographical equivalents are found in such forms
as Seminatrix p. cyclas, Liodytes a. lineapiatus and Pseudo-
branchus s. belli.

Three new subspecies have been proposed since Ortenburger's
monograph, and even since the 5th Edition of the Checklist (Stej-
neger and Barbour, 1943). All of these are from Florida and are:
C. c. paludicolus Auffenberg and Babbitt (1953), C. c. helvigularis
subsp. nov. and C. c. haasti Bell (1952). The last named race does
not seem worthy of subspecific designation. Of the five characters
stated as being diagnostic of haasti, one is always found in typical
specimens of priapus (black belly); one is occasionally found in both
priapus and haasti (brown snouts); another is found in all specimens
of priapus except the central and south-central Florida peninsular
populations (black supralabials ) ; and I consider the comparative
luster and size extremely difficult to measure (there being consider-
able overlap), and of too small an average difference, if they exist
at all, to warrant the erection of a new subspecies based on these
characters alone. It also seems highly significant that of all of the
characters which have been examined, the populations from the
Middle and Lower Keys seem to have closer affinities to those from
the northern parts of the peninsula than to those from any other
portion of the southeast. In fact, the two populations seem prac-
tically identical, and I find it impossible to tell them apart on the
basis of any character, or group of characters, which I have studied
so far.

148 Tulane Studies in Zoology Vol. 2

Three names have been revived from the synonymy of C. c. flavi-
ventris since Ortenburger's monograph. These are C. c. anthicus,
C. c. foxi and C. c stejnegerianus. Two races {anthicus and stejne-
gerianus) seem worthy of this distinction. Coluber c. foxi, con-
sidered a synonym of C. c. flaviventris in this paper, may deserve
subspecific designation if it is shown that the area of intergradation
between the two populations exhibits an external, rather than an
internal cline as suggested in this paper.

The variations found in the eastern races have been examined
critically. Some of these were found to be correlated with either
locality, sex, environment, or age.

Of the correlated variations, those concerned with geographic
variation are most obvious. When plotted, they appear to be internal,
external and ontogenetic clines or character gradients. Some of these
have been summarized in figure 16 for easier reference and to show
the varying limits of distribution from one character to the next.

External Clines:

Length of the hemipenial spines. — An examination of this char-
acter has indicated that the range of C. c. priapus, heretofore thought
to be restricted to peninsular Florida, includes the Atlantic and Gulf
Coastal Plains and the Mississippi Lowlands. Coluber c. constrictor
is the only subspecies which does not possess an enlarged basal
hemipenial spine which may be termed a hook.

Coloration, — The color of the temporals, gulars, supralabials, chin,
and ventral and dorsal surfaces are examples of this type of cline.
On the basis of the dorsal and ventral coloration the ranges of C. c.
constrictor and C. c. flaviventris have been changed from those
given by Ortenburger (1928). The eastern edge of the range of
flaviventris has been placed farther westward in Arkansas and north-
eastern Louisiana. The southern edge of the range of this form in
the states of Illinois and Indiana has been placed a little farther
northward.

Internal Clines:

Number of hemipenial spines. — The number of spines on the
hemipenis of male specimens, which may be termed hooks, is variable
and seems to form a continuous gradient. A lower number of hooks
are found in the eastern forms than in those from western United
States. Although specimens of Coluber c. flaviventris normally have
at least one hook, many specimens from the northern portions of the
range of this form have basal spines which are not large enough to
be termed hooks. Specimens which are intermediate in coloration
between the races flaviventris and constrictor from central Ohio do
not seem to be intermediate in this spinal character.

Coloration of the young. — Many colors suspected of being char-
acteristic of populations from the Southeast have been shown to be
character-gradients. These include the color of the dorsal and ven-
tral surfaces and the number and size of the dorsal and ventral
markings. The mean number of dorsal spots is highest in the West,

No. 6

Auffenberg: Coluber constrictor

149

Figure 17. Generalized directions for the major internal geoclines
which are evident in the morphological variations of Coluber con-
strictor. 1. A primary focal point, represented by the race C. c.
stejnegerianus, is the basis for many clines running in northerly
directions, such as the internal clines represented by the number
of ventral and caudal scales, the point of reduction of the dorsal
scale rows, and adult size. Most of these characters increase in
value or degree as one moves away from southern Texas. 2. A very
broad and generalized cline from western to the eastern portions
of North America, which seems to terminate in southeastern United
States. The characters involved in this cline are the general in-
crease in the number of some of the teeth, the number of hemipenial
spines, and the changing shape of the premaxillary bone. 3. An
obvious reversed cline is found in the extreme northern portions of
the range of this species, at least in central and eastern United
States. This internal cline deals with the number of ventrals and
caudals, and the point of reduction of dorsal scale rows. This is a
reversal of cline number 1. 4. Another reversed cline is found from
the Lower to the Upper Keys of Florida. It is represented by the
increase of ventrals, caudals, point of reduction of the dorsal scale
rows, larger size (?), and the gradual change from brown to red-
dish eyes in the adults, as well as an increase of red colors in the
juveniles on both dorsal and ventral surfaces. 5. A cline from the
continental portions of the southern Coastal Plain into peninsular
Florida, terminating in the extreme southern tip of the mainland.
It is represented by the increase in the frequency of a loreal-first
supralabial contact, increase in the number of ventrals and caudals,
more posterior reduction of the dorsal scales, a gradual change from
brownish to red eyes, larger size, and decrease of number of dorsal
and ventral spots in the juveniles, as well as a change of color in
these markings. 6. A cline from flaviventris from the Great Plains
and Coastal Prairies of Texas along the Coastal Prairies of Louisi-
ana into the delta region. It is represented by an increase in the
reddish colors on the ventral surface of juveniles, the decrease of
ventral markings in number and size, and the increase of darkly
colored temporals in adults. 7. An internal cline from the popu-
lations of flaviventris of the Great Plains northeastward into the
Steppe Peninsula, represented by the increase of the number and
size of juvenile belly markings, and the increase of the frequency ot
darkly-colored temporals.

150 Tulane Studies in Zoology Vol. 2

where the spots are smallest, and lowest in the East, where they are
largest. The number and size of the ventral markings are largest
and most numerous in the northern latitudes, and smallest and least
numerous in the southern latitudes.

Scutellation. — The mean number of both ventral and caudal scales
for various localities have been shown to be internal clines. The
lowest mean numbers of ventral and caudal scales are found in the
western races, while the highest are found in peninsular Florida and
the greater part of the glaciated regions of North America. The
highest mean number of ventrals and caudals is found in the south
Florida race, paludicolus. The lowest number occurs in the south-
eastern Texas and Mexican Gulf Coast race, stejnegerianus. The
mean number of ventral and caudal scales of populations from Florida
show fair correlation with the physiographic divisions of the state;
the geologically "older" portions of the state having significantly
lower mean counts than do those from the geologically "newer", and
lower coastal areas. The northern portions of these peninsular physio-
graphic areas tend to have lower mean counts than do the southern
portions, though the differences are not statistically significant.
However, significant differences occur between specimens from north-
ern, southern and the Keys areas of Florida. The coastal areas of
Georgia and South Carolina have higher mean counts than do popu-
lations from the Piedmont and Appalachian Mountain Provinces of
these two states.

A high percentage of the specimens from southern United States
possess a contact between the loreal and first supralabial. However,
these percentages vary with locality and the character is a good
example of geographically variable dimorphism. This contact occurs
infrequently in specimens from Michigan, Ohio and Illinois, but in
at least 10% of the specimens from south of the Fall Line. It occurs
in about 60% of the specimens from southern Florida.

The point of reduction of the dorsal scale formula has been found
to vary geographically and represents an internal cline. It varies in
direct proportion to the total number of mean ventral scales, and is
contained in that number from 1.6-1.8 times, regardless of locality.

Osteology. — The shape of the premaxillary bone has been found
to vary geographically as an internal cline. In the western subspecies
the two posterior processes are shorter and in a horizontal plane,
whereas in the eastern races, and especially in priapus and paludi-
colus, they are in a vertical plane, and longer.

Dentition. — The number of teeth also varies geographically, and
in general the means are lower in the western races than in the eastern
forms.

Dimorphism:

Scutellation, — The contact of the first supralabial and loreal has
already been mentioned as a dimorphic character, but it also varies
geographically as an internal cline.

Sexual dimorphism. — Dimorphism of this nature is indicated in

No. 6 Auffenberg: Coluber constrictor 151

the number of caudals, though not statistically significant. Very
slight differences in the mean numbers of ventrals and supralabials,
mentioned by Ortenburger as being sexually dimorphic characters,
are shown to be non-significant. Associated with the number of
caudals is the relative length of the tail, and as in the caudals, the
differences in length are probably not statistically significant.

Environmentally produced characters. — Small black punctations on
the ventral and lower lateral surfaces of adults have been shown to
be caused by a small burrowing mite (Ophioptes sp., near tropicalis
Ewing). Although seemingly restricted to the Mississippi Lowlands
and the Coastal Plain, the character does not have obvious taxonomic
significance.

Variations in the following characters have been found to be un-
corrected with environment, age, sex or locality: dorsal scales, num-
ber of scales between the chin shields, number of supralabials, infra-
labials, postoculars, preoculars, loreals and temporal scales. Variations
in the ontogenetic color change from a blotched pattern of the
juvenile to the uniform coloration of the adult do not seem correlated
with locality.

References Cited
Allen, E. R. and R. Slatten 1945. A herpetological collection from
the vicinity of Key West, Florida. Herpetologica, 3(1) : 25-26.

Anderson, P. 1942. Missouri amphibians and reptiles. Bull. Chicago
Acad. Sci., 6: 203-220.

Atkinson, D. A. 1901. The reptiles of Allegheny County, Pennsyl-
vania. Ann. Carnegie Mus., 1: 145-157.

Auffenberg, W. 1949. The racer, Coluber c. stejnegerianus (Cope),
in Texas. Herpetologica, 5(4) : 53-58.

1955. The status of the fossil snake Coluber

acuminatus (Cope): Copeia, 1955 (in press).

Auffenberg, W. and L. H. Babbitt 1953. A new subspecies of
Coluber constrictor from Florida. Copeia, 1953 (1) : 44.

Babbitt, L. H. 1932. Remarks on Connecticut herpetology. Bull.
Boston Soc. Nat. Hist., (63) : 23-28.

Baird, S. F. and C. Girard 1853. Catalogue of North American
Reptiles in the Museum of the Smithsonian Institution. Part I, —
Serpents. Smithsonian Inst., Wash., D. C, pp. 1-172.

Bell, L. N. 1952. A new subspecies of the racer Coluber constrictor.
Herpetologica, 8(2): 21.

Blair, W. F. and T. H. Hubbell 1938. The biotic districts of Okla-
homa. Amer. Midi. Nat., 20(2) : 425-454.

Blanchard, F. N. 1924. A collection of amphibians and reptiles from
southeastern Missouri and southern Illinois. Pap. Mich. Acad.
Sci. Arts and Letters, 4 : 533-541.

Bogert, C. M. 1947. The status of the genus Leptodrymus Amaral,
with comments on modification of Colubrid premaxillae. Amer.
Mus. Novitates, No. 1952 : 1-14.

Bogert, C. M. and R. B. Cowles 1947. Moisture loss in relation to
habitat selection in some Florida reptiles. Amer. Mus. Novitates,
No. 1858: 1-32.

Boulenger, G. A. 1893. Catalogue of the Snakes in the British
Museum, London. 1 : pp. 1-448.

152 Tulane Studies in Zoology Vol. 2

Branson, E. B. 1904. The snakes of Kansas. \Kans. Univ. Sci. Bull.,
2(12): 353-430.

Breckenridge, W. J. 1944. The Reptiles and Amphibians of Minne-
sota. Univ. Minn. Press, Minneapolis, pp. 1-202.

Brimley, C. S. 1942. Reptiles and amphibians of North Carolina.
Carolina Tips, No. 21: [No pagination].

Brown, B. C. 1950. An annotated check list of the reptiles and am-
phibians of Texas. Baylor Univ. Studies, pp. 1-257.

Burger, W. L., Jr. 1948. Farancia abacura reinwardtii in Oklahoma.
Copeia, 1948 (2) : 133.

Burt, C. E. 1935. Further records of the ecology and distribution of
amphibians and reptiles in the middlewest. Amer. Midi. Nat.f
311-336.

Burt, C. E. and M. D. Burt 1929. A collection of reptiles and am-
phibians from the Mississippi Valley with field observations.
Amer. Mus. Novitates, No. 381: 1-14.

Burt, C. E. and W. L. Hoyle 1934. Additional records of the reptiles
of the central prairie regions of the United States. Trans. Kans.
Acad. Sci., 37: 193-216.

Cagle, F. R. 1941. A key to the reptiles and amphibians of Illinois.

Mus. Nat. and Soc. Sci., Southern III. Normal Univ., Contrib*

No. 5: 1-32.
1942. Herpetological fauna of Jackson and Union

Counties, Illinois. Amer. Midi. Nat., 28(1): 164-200.

Carr, A. F., Jr. 1940. A contribution to the herpetology of Florida.
Univ. Fla. Publ, Biol. Sci. Series, 3(1) : 1-118.

Conant, R. 1930. Field notes on a collecting trip. Bull. Antivenin
Inst. Amer., 4(3) : 60-64.

1938. Reptiles of Ohio. Amer. Midi. Nat., 20(1):

1-200.

1942. Notes on the young of three recently de-
scribed snakes with comments on their relationship. Bull..
Chicago Acad. Sci., 6(10) : 194-198.

Cooke, C. W. 1939. Scenery of Florida interpreted by a geologist.

Fla. Geol. Bull, 17: 1-118.
1945. Geology of Florida. Fla. Geol. Bull, 29 r

1-339.

Cope, E. D. 1862. Notes upon some reptiles of the Old World. Proc.
Acad. Nat. Sci. Phila., 14: 337-344.

1899. Vertebrate remains from Port Kennedy bone

deposit. Jour. Acad. Nat. Sci. Phila., (2nd ser.), 11: 193-267.

1900. Crocodilians, lizards and snakes of North

America. U. S. Nat. Mus. Rept., 1898: 153-1270.

Dellinger, S. C. and J. D. Black 1938. Herpetology of Arkansas.
Part One, the Reptiles. Occ. Pap. Univ. Ark. Mus., No. 1 : 1-47.

Ditmars, R. L. 1907. The Reptile Book. Doubleday Page and Co.,
New York, pp. 1-472.

1936. The Reptiles of North America. Garden

City Publ. Co., N. Y., pp. 1-477.

1939. A Field Book of North American Snakes-

Doubleday Doran and Co., New York, pp. 1-305.

Dowling, H. G. 1950. Studies of the black swamp snake, Seminatrix
pygaea (Cope), with descriptions of two new subspecies. Misc.
Publ Mus. Zool Univ. Mich., No. 76 : 1-38.

No. 6 Auffenberg: Coluber constrictor 153

1951. A proposed method of expressing scale re-
duction in snakes. Copeia, 1951 (2) : 131-133.

Dumeril, A. M. C, G. Bibron and A. Dumeril 1854. Erpetologie
Generate ou Histoire Naturelle Complete des Reptiles, 7(1):
1-780. Librairie Encyclopedique de Roret, Paris.

Dumeril, A. and F. Bocourt 1870. Etudes sur les reptiles et les
batraciens. Mission scientifique au Mexique et dans VAmerique
Centrale [Paris], 3(3) : 1-1012.

Dunn, E. R. and C. Wood 1939. Notes on eastern snakes of the
genus Coluber. Notulae Naturae, No. 5: 1-4.

Eckel, E. C. and F. C. Paulmier 1902. Catalogue of New York rep-
tiles and batrachians. N. Y. State Mus. Bull., 51: 356-414.

Ellis, M. M. and J. Henderson 1913. The amphibia and reptilia of
Colorado. Univ. Col. Studies, 10(2) : 35-129.

Etheridge, R. E. 1952. The southern range of the racer Coluber
constrictor stejne g erianus (Cope), with remarks on the Guate-
malan species Coluber ortenburgeri Stuart. Copeia, 1952 (3) :
189-190.

Fox, W. 1948. Effect of temperature on development of scutellation
in the gartersnake, Thamnophis e. atratus. Copeia, 1948 (4) :
252-262.

Grobman, A. B. 1941. A contribution to the knowledge of variation
in Opheodrys vervalis (Harlan), with the description of a new
subspecies. Misc. Publ. Mus. Zool. Univ. Mich., No. 50: 1-37.

Gunther, A. C. L. G. 1850. Catalogue of the Colubrine Snakes in
the British Museum. London., pp. 1-281.

Guthrie, J. E. 1926. The snakes of Iowa. Bull. Iowa St. Agric.
Exper. Sta., No. 239 : 1-192.

Holbrook, J. E. 1842. North American Herpetology, (2nd ed), 3:
1-128, J. Dobson Co., Philadelphia.

Hubbs, C. L. 1922a. Variations in the number of vertebrae and other
meristic characters of fishes correlated with the temperature of
the water during development. Amer. Nat., 56: 360-372.

1922b. Seasonal variation in the number of verte-
brae of fishes. Papers Mich. Acad. Sci., Arts & Lttrs. 2: 207-214.

1940. Increased number and delayed development

of scales in abnormal suckers. Papers Mich. Acad. Sci., Arts &
Lttrs. 26: 229-237.

Huxley, J. 1943. Evolution, the Modern Synthesis. Harper Brothers,
New York, pp. 1-645.

Inger, R. F. and P. L. Clark 1943. Partition of the genus Coluber.
Copeia, 1943 (3) : 141-145.

Jan, G. 1863. Elenco Sistematico degli Ofidi, Descritti e Desegnati
per L'Iconografia Generale. Lombardi, Milano, pp. 1-143.

1867. Iconographie Generale des Ophidiens. liv.

22, pis. 1-3, Paris.

Jordan, D. S. 1893. Temperature and Vertebrae : A Study in Evolu-
tion, Being a Discussion of the Relations of the Number of
Vertebrae among Fishes to the Temperature of the Water and
to the Character of the Struggle for Existence. Wilder Quarter-
Century Book, Ithaca, pp. 13-36.

Kelly, H. A. 1936. Snakes of Maryland. Nat. Hist. Soc. Maryland,
pp. 1-103.

King, F. W. 1932. Herpetological records and notes from the vicinity
of Tuscon, Arizona, July and August. Copeia, 1932 (4) : 175-177.

154 Tulane Studies in Zoology Vol. 2

Klauber, L. M. 1940. Two new subspecies of Phyllorhynchus, the
leaf -nosed snakes, with notes on the genus. Trans. San Diego
Soc. Nat. Hist, 9: 192-214.

Linnaeus, C. 1758. Sy sterna Naturae per Regna Tria Naturae,
Secundum Classes, Ordines, Genera, Species, cum, Characteribus,
Differentis Synonymis, Locis, 10th. ed. Laurentii, Holiae, 1:
1-824.

Linsdale, J. M. 1940. Amphibians and reptiles in Nevada. Proc.
Amer. Acad. Arts and Sci., 73(8) : 197-257.

Loennberg, E. 1894. Notes on the reptiles and batrachians collected
in Florida in 1892 and 1893. Proc. U. S. Nat. Mus., 17: 317-339.

Lueth, F. X. 1941. Manual of Illinois Snakes. 111. Dept. Cons.,
Springfield, pp. 1-48.

MacNeill, F. S . 1951. Pleistocene shore lines in Florida and Georgia.
Geol. Surv. Prof. Paper 221-F., pp. 95-107.

McCauley, R. H., Jr. 1945. The Reptiles of Maryland and the Dis-
trict of Columbia., publ. by Author, Baltimore, pp. 1-194.

Mills, R. C. 1948. A checklist of the amphibians and reptiles of
Canada. Herpetologica, 4(2) : 1-15.

Mittleman, M. B. 1947. Miscellaneous notes on Indiana amphibians
and reptiles. Amer. Midi. Nat., 38(2): 466-484.

Mulaik, S. and D. Mulaik 1942. A neglected species of Coluber.
Copeia, 1942 (1) : 13-15.

Ortenburger, A. I. 1926. A report of the amphibians and reptiles
of Oklahoma. Proc. Okla. Acad. Sci., 6: 89-100.

1927. A key to the snakes of Oklahoma. Proc.

Okla. Acad. Sci., 6(1) : 197-218.

1928. The whipsnakes and racers, genera Masti-

cophis and Coluber. Mem. Univ. Mich. Mus., 1: 1-247.

1930a. Reptiles and amphibians from Pawnee

County, Oklahoma. Copeia, 1930 (3) : 94-95.

1930b. A key to the lizards and snakes of Okla-
homa. Publ Univ. Okla. Biol. Surv., 2(4): 209-239.

Perkins, C. B. 1949. A key to the snakes of the United States (2nd
ed.). Bull. Zool. Soc. San Diego, No. 24: 1-79.

Peters, W. 1866. Mitteilung iiber neue amphibien und fische.
Monatsber. Akad. Wiss. Berlin, 1866: 86-96.

Pope, C. H. 1944. Amphibians and Reptiles of the Chicago Area.
Chicago Nat. Hist. Mus., Chicago, pp. 1-275.

Roddy, H. J. 1923. Reptiles of Lancaster County, and the State of
Pennsylvania. Dept. Nat. Hist., Franklin and Marshall Coll.,
pp. 1-53.

Ruthven, A. G. 1908. The faunal affinities of the prairie region of
central North America. Amer. Nat., 42: 388-393.

Say, T. 1823. Account of An Expedition from Pittsburgh to the
Rocky Mountains Under the Command of Major S. H. Long.
H. C. Peary and I. Lea, Philadelphia, 2: 1-442.

Schmidt, K. P. 1938. Herpetological evidence for the postglacial
eastward extension of the steppe in North America. Ecology,
19(3): 396-407.

1953. A Check List of North American Amphibi-
ans and Reptiles (6th ed.) Amer. Soc. Ichthy. and Herp., Chi-
cago, pp. 1-280.

No. 6 Aufjenberg: Coluber constrictor 155

Schultz, L. P. 1926. Temperature-controlled variation in the golden
shiner, Notemegonus crysoleucas. Papers Mich. Acad. Sci. Arts
& Lttrs., 7 : 417-429.

Smith, H. M. and A. B. Leonard 1934. Distributional records of
reptiles and amphibians in Oklahoma. Amer. Midi. Nat., 15:
190-196.

Stejneger, L. and T. Barbour 1917. A Check List of North Ameri-
can Amphibians and Reptiles (1st. ed.), Harvard Univ. Press,
Cambridge, pp. 1-125.

1943. A Check list of North American amphibians

and reptiles (5th ed.), Bull. Mus. Comp. Zool, 93(1) : 1-260.

Strecker, J. K. 1908. Reptiles and batrachians of McLennan County,
Texas. Proc. Biol. Soc. Wash., 21: 69-85.

1915. Reptiles and amphibians of Texas. Baylor

Bull., 18(4): 1-82.

Strecker, J. K. and W. J. Williams 1927. Herpetological records
from the vicinity of San Marcos, Texas, with distributional data
on the amphibians and reptiles of the Edwards Plateau region
and central Texas. Contrib. Baylor Univ. Mus., No. 12: 1-16.

Stuart, L. C. 1930. An extension of the range of Coluber constrictor
mormon (Baird and Girard). Copeia, 1930 (2): 44.

1934. Studies on neotropical Colubrinae. IV. An

apparently new species of Coluber from Guatemala. Occ. Papers
Mus. Zool. Univ. Mich., No. 284: 1-4.

o

Taning, A. V. 1952. Experimental study of meristic characters in
fishes. Biol. Rev., 27: 169-193.

Taylor, E. H. 1935. Arkansas amphibians and reptiles in the Kansas
University Museum. Univ. Kans. Sci. Bull, 22(10): 207-218.

Trapido, H. 1944. The snakes of the genus Storeria. Amer. Midi.
Nat., 31(1) : 1-84.

Yarrow, H. C. and H. W. Henshaw 1878. Report upon the reptiles
and batrachians collected by the expedition of 1875, 1876, 1877
in California, Arizona and Nevada. Ann. Rept. Chief Eng.,
Surv. West 100th. Merid., App. L. of App. NN., Washington,
pp. 1-233.

\

Volume 2, Number 7

February 25, 1955

NOTROPIS HYPSILEPIS, A NEW CYPRINID FISH FROM THE

APALACHICOLA RIVER SYSTEM OF GEORGIA

AND ALABAMA

ROYAL D. SUTTKUS,

DEPARTMENT OF ZOOLOGY, TULANE UNIVERSITY,

NEW ORLEANS, LOUISIANA

and

EDWARD C. RANEY,

DEPARTMENT OF CONSERVATION, CORNELL UNIVERSITY,
ITHACA, NEW YORK

mi.

MAS 8 1955

urnrn
mnmw

TULANE UNIVERSITY
NEW ORLEANS

TULANE STUDIES IN ZOOLOGY is devoted primarily to the
zoology of the area bordering the Gulf of Mexico and the Caribbean
Sea. Each number is issued separately and deals with an individual
study. As volumes are completed, title pages and tables of contents
are distributed to institutions exchanging the entire series.

Manuscripts submitted for publication are evaluated by the editor
and by an editorial committee selected for each paper. Contributors
need not be members of the Tulane University faculty.

EDITORIAL COMMITTEE FOR THIS NUMBER

Carl L. Hubbs, Professor of Biology, Scripps Institution of
Oceanography, La Jolla, California.

Reeve M. Bailey, Curator of Fishes, Museum of Zoology, Uni-
versity of Michigan, Ann Arbor, Michigan.

Ernest A. Lachner, Associate Curator of Fishes, United States
National Museum, Washington, D. C.

Manuscripts should be submitted on good paper, as original type-
written copy, double-spaced, and carefully corrected.

Separate numbers may be purchased by individuals, but subscriptions
are not accepted. Authors may obtain copies for personal use at cost.
Address all communications concerning exchanges, manuscripts, edi-
torial matters, and orders for individual numbers to the editor. Re-
mittances should be made payable to Tulane University.

When citing this series authors are requested to use the following
abbreviations: Tulane Stud. Zool.

Price for this number: $0.30.

Assistants to the Editor:
Carol L. Freret
Donald W. Tinkle

George Henry Penn, Editor,

c/o Department of Zoology,
Tulane University,
New Orleans, U. S. A.

MA* 8 195

NOTROPIS HYPSILEPIS, A NEW CYPRINID FISH FROM THE

APALACHICOLA RIVER SYSTEM OF GEORGIA

AND ALABAMA

ROYAL D. SUTTKUS,

Department of Zoology, Tulane University,
New Orleans, Louisiana

and

EDWARD C. RANEY,i

Department of Conservation, Cornell University,

Ithaca, New York

The small species of Notropis described below appears to be endemic
to the Apalachicola River system. It has been overlooked, or con-
fused with Notropis xaenocephalus (Jordan), which is native in the
upper Alabama River system of Alabama and Georgia. Little is
known of its life history or habitat and judging from the few times
it has been taken it seems to be scarce, or, possibly, it prefers a habitat
not normally investigated.

The material is deposited in the Cornell University fish collection
and was collected by the authors with the aid of R. H. Backus, J.
Willard Colston, R. W. Crawford, J. S. Dendy, J. N. Layne, C. R.
Robins, and R. L. Wigley. Counts and a description of the lateral
line head pores were kindly made available by Helen J. Illick. Counts
and measurements were taken using the methods described by Hubbs
and Lagler (1947: 8-15).

NOTROPIS HYPSILEPIS, sp. nov.
Figs. 1 and 2, Map 1

The 50 available specimens, all designated as types, range from 19
to 51 mm in standard length and are from eight localities. In the
enumeration of the material below the number of specimens and
range of standard length of the series measured in millimeters are
given in parentheses, e. g. (6, 41-51). In addition to standard ab-
breviations for compass directions, with the following "of" deleted,
the following are used: Co. = County, Cr. = Creek, Hwy. = High-
way, mi. = mile or miles, junct. = junction, R. = River, trib. =
tributory (of), CU = Cornell University fish collection, UMMZ =
University of Michigan, Museum of Zoology.

Material. — Holotype, CU 17442, an adult male, 51 mm from trib.
Soque R., 0.4 mi. W. Soque R. and 1.5 mi. W. Clarksville on Ga.
Hwy. 115, Chattahoochee R. system, Habersham Co., Georgia, on
April 1, 1950, by Edward C. Raney, Royal D. Suttkus, Richard H.
Backus and C. Richard Robins. Designated as paratypes are CU
28237 (6, 41-51) which were taken with the holotype.

Other paratypes from the Apalachicola River system are listed

1 A grant from the Cornell University Faculty Research Grants
Committee made possible the ichthyological explorations which dis-
covered this species.

162 Tulane Studies in Zoology Vol. 2

below. Chattahoochee River system of Georgia: CU 18102 (8, 34-
50), Grant Cr., trib. Soque R. near junct. with Chattahoochee R.,
near Cornelia, Habersham Co., June, 1949; CU 19625 (9, 45-51),
Chestatee R. at mouth of Yahoola Cr., 2.1 mi. S.E. Dahlonega, Lump-
kin Co., March 26, 1951; CU 17138 (19, 34-51), Vickery Cr. at
junct. with Chattahoochee R. at Roswell, Fulton Co., March 30, 1950;
CU 17279 (3, 19-27), Hodchodkee Cr., 1.4 mi. S. Lumpkin, Stewart
Co, March 28, 1950; UMMZ 157957 (1, 48), Cane Cr, VA mi.
W.S.W. Dahlonega, Lumpkin Co, August 25, 1939- Chattahoochee
River system of Alabama: UMMZ 142910 (2, 41-43), Halawakee
Cr, 10 mi. N.E. Opelika, Lee Co, October 17, 1938. Flint River
system: CU 21138 (1, 40), trib. Lazier Cr, 2.9 mi. W. Talbotton,
Talbot Co, March 26, 1952. Also examined was a single adult
specimen, not designated as a paratype, taken from Little Uchee Cr,
trib. Chattahoochee R, at US Hwy. 80. Russel Co, Alabama, October
10, 1953, by Jack S. Dendy. This specimen is now in the collection
of the Alabama Polytechnic Institute.

Diagnosis — A small, pale-colored species of Notropis with 2, 4 — 4,
2 teeth and 7 anal rays. Other fin rays: dorsal 8, rarely 7; caudal
19; pectoral 13-16, usually 14; pelvic 8, or occasionally 9- Anterior
lateral line scales elevated; exposed portion of third scale 4 times as
high as wide. Scales: predorsal rows usually 15, rarely 16 or 17;
above lateral line to dorsal origin 5, rarely 6; below lateral line to
anal origin 3 or 4; in lateral line 35 or 36; around body before dorsal
fin 21 to 26, usually 22 or 23; around caudal peduncle 12. Snout
blunt, mouth inferior. Body subterete. Lateral line on body com-
plete. Middorsal dark stripe little developed before and scarcely if
at all behind the dorsal fin. Lateral band on body weak. Diagnostic
dark patches on body at base of the first 4 or 5 dorsal and anal rays.
Area about the anus light. Small wedge-shaped basicaudal spot
separated from the lateral band on body by a light area. Oblique
dark scapular bar lacking. Size small, to 51 mm.

Description. — Proportional measurements for the holotype and 11
paratypes are given in Table 1. In the following enumeration of fin,
teeth, and scale counts, the count for the holotype is in bold face. The
frequency is shown in parentheses. Dorsal rays, 8 (44), 7 (3); anal
rays 7 (47); caudal rays, 19 (20); pectoral rays, 13 (3), 14 (12),
15 (6), 16 (1); pelvic rays, 8 (71), 9 (15), holotype 9—8. In
upstream samples the pelvic count averages high: the two samples
(CU 17442, 18102) from tributaries of the Soque R. have 8 (18),
9 (12). Teeth 2, 4—4, 2 (9), 1, 4—4, 2(1). Predorsal scale rows,
15 (8), 16 (2), 17 (1); scales from dorsal origin to lateral line,
5 (10), 6 (1); scales from anal origin to lateral line, 3 (5), 4 (6);
lateral line scales, 35 (10), 36 (9); scales around body before dorsal
fin, 21 (1), 22 (7), 23 (5), 24 (3), 25 (2), 26 (1), with 10 to
12, usually 11, above the lateral line and 9 to 12, usually 9 or 10,

11 in holotype below the lateral line; caudal peduncle scale rows,

12 (11), the formula being 5—2—5.

No. 7

Suttkus and Raney: A New Cyprinid

163

TABLE 1.

Proportional Measurements of Eleven Adult Notropis hypsilepis

in Thousandths of the Standard Length. Seven Males

and Four Females Were Measured.

Holotype

Paratypes

Measurement CU 17442

CU 28237 and 18102

Male

Males

Females

Range

Avg.*

Range

Avg.

Standard length, mm

51.2

41.2-51.2

47.0

37.1-48.2

43.9

Dorsal' origin to snout tip

502

491-510

501

495-525

505

Dorsal origin to caudal

base

529

514-546

530

527-545

538

Dorsal origin to occiput

307

287-316

301

269-321

296

Pelvic insertion to snout

tip

488

488-499

494

480-484

482

Anal origin to caudal base

344

336-365

348

338-347

341

Body, greatest depth

193

177-210

198

184-244

212

greatest width

127

119-137

126

116-135

129

Dorsal origin to lateral

line

127

112-134

126

121-139

133

Pelvic insertion to lateral

line

092

082-102

092

084-112

100

Caudal peduncle, length

226

226-248

240

226-252

239

least depth

094

090-105

097

085-097

091

Head, length

244

239-259

247

242-258

250

depth

161

152-161

157

154-156

155

width

131

123-136

129

119-130

124

Interorbital, least fleshy

width

088

085-095

090

085-090

087

Snout length

080

080-087

084

079-084

082

Eye length

065

065-075

070

069-074

071

Upper jaw length

073

072-078

075

067-075

071

Suborbital, least width

041

032-041

036

031-136

034

Dorsal fin, depressed

length

227

226-246

235

225-231

228

Anal fin, depressed

length

184

174-193

185

169-179

174

Caudal fin, length

270

270-304

285

271-288

276

Pectoral fin length

191

191-213

202

188-194

190

Pelvic fin length

164

162-181

171

155-174

166

* Holotype included.

The body shape and other characters are illustrated in Figures 1
and 2. The body is subterete. The subtriangular head terminates in
a blunt snout, which extends slightly beyond the upper lip and far
in front of the lower lip. The contour of the dorsum posterior to
the occiput is straight and rises gradually to the dorsal origin; it dips
at the dorsal fin base and then extends in an almost straight line
posteriorly to the procurrent caudal rays. Ventrally the body curves
gently downward to a point two-thirds the distance between the
pectoral fin base and the pelvic fin insertion; posteriorly the contour
is almost a straight line to the anal origin and then curves in a gentle
arc to the procurrent caudal rays. The least depth of the caudal

164 Tulane Studies in Zoology Vol. 2

peduncle is just anterior to the procurrent caudal rays. The upper
jaw does not reach to below the anterior margin of the eye. The gape
of the mouth is little inclined. The upper lip is curved about mid-
way, its anterior half is almost horizontal and at its anterior tip it
barely reaches the lower level of the eye. Viewed from below the
gape is U-shaped. The eye is shorter than the snout and is rather
high. The interorbital region is moderately convex. The center
of the eye is slightly closer to the tip of snout than to the margin of
opercle.

The lateral line is complete. Anteriorly its path is marked by
small aggregations of melanophores above and below each pore.
Posteriorly it curves gently downward to a point opposite the pelvic
fin insertion and then curves upward to a point opposite the posterior
end of the dorsal fin base. The remaining part is straight and lies
within the upper part of the relatively narrow dark lateral band. In
the following description of the cephalic lateral line pores, the num-
ber of specimens is given in parentheses. In making the pore counts
on a canal the anteriormost is designated number one. Five speci-
mens (CU 17138) were studied. The supratemporal canal is in-
complete and may be broken irregularly. In three specimens, the
count is 2 on the left and 1 on the right; in two, there are 2 on the
left, 2 isolated in or near the center and 1 on the right. The supra-
orbital canal is complete in four of the five specimens studied, with
pore counts of 8 (2), 9 (1), 10 (l);in one specimen it is incom-
plete, with a break between pores four and five. The supraorbital
may end opposite a vertical line projected from the posterior margin
of the eye but usually this line falls on the penultimate pore. The
infraorbital canal is complete in three out of the five studied, with
pore counts of 14 (1), 15 (1), 17 (1). In two specimens, with
counts of 17 and 16 respectively, the canal is incomplete and irregu-
larly broken; in one the break comes between pores four and five;
in the other, in the temporal region, between pores fourteen and
fifteen. A vertical from the anterior margin of the eye falls between
pores four and five in four specimens and on pore seven in one.
The preoperculomandibular canal is incomplete in four, with a count
of 13, and complete in one, with a count of 12. When this canal
is incomplete, it is broken between pores five and six. In respect
to the high count of this canal, bypsilepis stands extreme among
species with 2, 4 — 4, 2 teeth and 7 or 8 anal rays.

The extended dorsal and anal fins are almost straight on the distal
border; when depressed the tip of the anterior rays of the dorsal
exceed the tip of the posterior ray while the reverse is true of the
anal fin. The dorsal fin is relatively high. The distance from the
midcaudal base to the midfork is less than half the distance from the
caudal base to the tip of the longest lobe. The posterior margin of
the pectoral and pelvic fins is gently rounded. The origin of the
dorsal fin is directly over or slightly posterior to the insertion of the
pelvic fin.

No. 7

Suttkus and Raney: A New Cyprinid

165

Figures 1-2. Notropis hypsilepis, sp. nov. 1. (top) Side view of
holctype, an adult male, 51 mm in standard length, from trib. Soque
River, 0.4 mi. W. Soque River and 1.5 mi. W. Clarksville, Chattahoo-
chee River system, Habersham Co., Georgia. 2. (bottom) Top view
of a paratype, an adult male, 51 mm in standard length, from the
same series as the holotype. (Photographs by Douglass M. Payne.)

166

Tulane Studies in Zoology

Vol. 2

Map 1. Distribution of Notropis hypsilepis. (Circle designates the

type locality)

No. 7 Suttkus and Raney: A New Cyprinid 167

The pharyngeal arch which bears the teeth is shaped like a fish
hook and the extreme tip of the short arm is curved abruptly out-
ward. The upper two-fifths of the arch is widened prominently.
The teeth in the main row are three to four times higher than wide
and the hooked tips partly overhang the well developed grinding
surface, which has serrulate edges. The small teeth of the lesser
row are similarly shaped and have a well developed grinding surface,
but the hook is not as pronounced. There are only three developed,
bluntish gill rakers on the first arch. One lies near the angle and the
other two on the upper half of the lower arch. The nape and breast
are scaled.

In separating the light-colored hypsilepis from the other species of
Notropis in the southeastern United States, many details of pigmen-
tation are diagnostic. The other forms with 2, 4 — 4, 2 teeth and 7
anal rays, are usually darkish, with a prominent dark lateral band
running the length of the body. In hypsilepis the lateral band is
weakly developed throughout. Anteriorly it consists of a scattering
of melanophores which lie above the lateral line and in addition there
is a small group of melanophores immediately above and below each
lateral line pore. Posteriorly the lateral band is darker, narrower,
and more continuous close to the lateral line. On the rear half of
the caudal peduncle the melanophores are arranged in oblique lines,
some of which form chevrons. The dark lateral band is paralleled
above by a prominent clear band which is wider. The lower side is
white. On the snout and opercle the lateral band is weak. The
adjacent areas behind the eye and the fleshy margin of the opercle
lack or have very few chromatophores. The dark appearance of the
opercle of the holotype (fig. 1) is due to the darkness of the under-
lying gills which changed upon preservation.

The white area from the eye to and including the nostrils con-
trasts with the internarial region which bears scattered melanophores.
On the top of the head between the eyes are two subquadrate blotches,
separated by a clear area. The heart-shaped mark on the rear of the
head is divided into three parts, each with a lighter center. The
temporal region has scattered melanophores. There is no oblique dark
bar posterior to the opercular slit. The anterior quarter of the upper
lip and the groove above bear scattered melanophores; the lower lip
is white, as is the lower half of the head and the inside of the mouth.

Each of the dorsal scales, except those on the midline, has a light
center and the posterior third bears small scattered chromatophores.
The middorsal streak anterior to the dorsal fin origin consists of a
thin line of one to three rows of melanophores and it lacks an ex-
panded concentration immediately before the dorsal fin. Diagnostic
are the dark marks on the body that extend along the base of the
first 4 to 5 developed rays of the dorsal and anal fins. Posterior to
the dorsal fin base the middorsal streak is absent or consists of a
weak line just behind the dorsal base. The basicaudal dark spot is a
wedge with its apex pointing forward; about half is on the caudal

168 Tulane Studies in Zoology Vol. 2

rays. It is separated from the dark lateral band on the caudal ped-
uncle, which is also expanded posteriorly, by a clear space which
sometimes has a few scattered melanophores. Just above the diffuse
dark area on the posterior caudal peduncle is a faint band of melano-
phores, which extends to the base of the caudal rays. The area im-
mediately before and behind the anus is white. On the lower edge
of the caudal peduncle two or three rows of melanophores form a
weak band that normally reaches to the procurrent caudal rays. The
peritoneum is silvery, with large scattered melanophores which be-
come more common dorsolaterally. The fins are white and trans-
parent; the dorsal and caudal have the ray margins faintly outlined
by a row of melanophores and the first two pectoral rays are similarly
marked. No bright colors were noted in the field.

Nuptial tubercles are well developed on the male. Small scattered
tubercles are present on the opercle and elsewhere on the head.
Anteriorly they are coarse and irregularly arranged on the snout, below
the eye, and on the dentary. Fine tubercles cover the lips, isthmus,
and branchiostegals. The scales from the occiput to the dorsal fin
have small tubercles, mostly in an irregular row on the posterior
border. A vertical row of tubercles is present on the front third of
each lateral line scale on the anterior half of the body. These are
absent on the scales in the row above the lateral line and none were
noted on the scales below the lateral line row, except for a few
scattered, coarse tubercles on two or three transverse scale rows im-
mediately behind the opercle. The small tubercles on scales and head
were red in specimens taken on March 26 (CU 19625). Tubercles
are absent on fins, except on the upper side of the first 8 pectoral
rays, where they are arranged in several rows; the maximum develop-
ment is on rays 3 to 6. The female has small tubercles and many
developed "sensory" pits on the head but no tubercles on body or fins.
Other than in the degree of development of nuptial tubercles, sexual
dimorphism is not marked but the pelvic fins are slightly longer and
the pectoral fins somewhat broader in the male. Judged from the
large ovarian eggs and the high development of tubercles, spawning
occurs sometime in June rather than in March or April.

Relationships. — Among the species of Notropis with 2, 4 — 4, 2
teeth hypsilepis seems to be most closely related to roseus, xaenocep-
halus and an undescribed species from the Mobile Bay drainage. It
differs from chalybaeus, cbrosomus and lutipinnis, which have 8 anal
rays, in numerous characters but especially in body shape, the type
of mouth, the shape of caudal spot, the development of the middorsal
stripe, the nature of the dark lateral band on body and head, and the
elevated anterior lateral line scales. A description of many of these
characters was given for the above-mentioned species by Suttkus and
Raney (1955: 73-84).

Besides the elevated anterior lateral line scales, the poorly developed
dark lateral band, the light colored body, inferior mouth and blunt
snout hypsilepis differs from baileyi, petersoni, xaenocephalus and

No. 7 Suttkus and Raney: A New Cyprinid 169

roseus, which have a black lateral stripe, 2, 4 — 4, 2 teeth and 7 anal
rays, in numerous details of pigmentation. The following characters
are helpful in separating hypsilepis from the forms listed immediately
above.

(1) Body light-colored, with contrasting dark patch on body at
the anterior base of the first 4 or 5 dorsal and anal fin rays, rather
than being darker, with the dark blotch extending along entire base.

(2) Melanophores lacking rather than present immediately be-
hind or beside anus.

(3) Dark lateral band on body very weak and mostly developed
posteriorly, rather than being well developed for its entire length.

(4) Basicaudal dark spot separated from the posterior extension
of the lateral band and forming a small wedge no wider than 3 caudal
rays at its posterior end, rather than being well developed and con-
tinuous with the dark lateral band, or only slightly separated (N.
peter soni) , and quadrate, or, if wedge-shaped, as wide as 5 or 6
caudal rays.

(5) Lacks oblique scapular bar, rather than being strong to weak.

(6) Upper lip white on its posterior two-thirds and lower lips
white, rather than each being dusky to black in whole or in part.

(7) Heart-shaped mark on top of head divided into three sections,
rather than two.

(8) Peritoneum silvery with, no or few scattered melanophores
midventrally, rather than being somewhat dusky and blotched ventrally.

Superficially and in the elevation of the anterior lateral line scales
hypsilepis resembles volucellus but the similarity is attributable to
convergence. These species differ trenchantly in tooth and anal ray
counts (4 — 4 and 8 in volucellus), as well as many details of pro-
portions and coloration.

Distribution and ecology. — Notropis hypsilepis is apparently en-
demic to the Apalachicola River drainage. Eight stations are known
from the Chattahoochee River system and one collection is available
from the upper Flint River system. It apparently prefers "Piedmont"
type streams with sand bottom. Although it has been collected
three times in small creeks from 10 to 20 feet wide, it has been taken
in greatest abundance in or near the mouths of streams at a point
where they enter large rivers. In color and body shape it seems to
be adapted for life on or near the bottom.

The name hypsilepis, derived from hypsi, high, and lepis, scale,
refers to the elevated anterior lateral line scales.

References Cited

Hubbs, Carl L. and Karl F. Lagler. 1947 (and 2nd printing, 1949).
Fishes of the Great Lakes Region. Bull. Cranbrook Inst. Sci.,
26:i-xi, 1-186, many figs.

Suttkus, Royal D. and Edward C. Raney. 1955. Notropis baileyi,
a new cyprinid fish from the Pascagoula and Mobile Bay drain-

170 Tulane Studies in Zoology Vol. 2

ages of Mississippi and Alabama. Tulane Stud. Zool., 2(5) :
69-86, 4 figs., 1 map.

Corrigendum
Material housed in the University of Michigan, Museum of Zoology-
should be acknowledged in the second paragraph of paper.

Addendum

A very fine series of 42 specimens of Notropis hypsilepis was
examined by the senior author while the paper was in press. These
specimens are hereby designated as paratypes.

Pigmentation, counts and measurements of these additional speci-
mens agree favorably (with one exception) with those given in the
diagnosis, description and elsewhere in the paper. The one exception
noticed was an anal-ray count of 8 rather than 7.

This additional series bears the following data: UMMZ 136086
(42, 34-48), trib. Chattahoochee R., just below Hwy. 43, just E.
Dahlonega, Georgia, collected by C. L. and L. C. Hubbs, May 15, 1937.

Volume 2, Number 8

May 4, 1955

THE FISHES OF TWO GULF COASTAL
MARSH AREAS OF FLORIDA

JOHN D. KILBY,

DEPARTMENT OF BIOLOGY, UNIVERSITY OF FLORIDA,
GAINESVILLE, FLORIDA

m. oo::p. zqol

btfAiti
MAY 1 9 1955

TULANE UNIVERSITY
NEW ORLEANS

TULANE STUDIES IN ZOOLOGY is devoted primarily to the
zoology of the area bordering the Gulf of Mexico and the Caribbean
Sea. Each number is issued separately and deals with an individual
study. As volumes are completed, title pages and tables of contents
are distributed to institutions exchanging the entire series.

Manuscripts submitted for publication are evaluated by the editor
and by an editorial committee selected for each paper. Contributors
need not be members of the Tulane University faculty.

EDITORIAL COMMITTEE FOR THIS NUMBER

Gordon Gunter, Director, Institute of Marine Science, Uni-
versity of Texas, Port Aransas, Texas.

G. Robert Lunz, Director, Bears Bluff Laboratories, Wadmalaw
Island, South Carolina.

Royal D. Suttkus, Assistant Professor of Zoology, Tulane Uni-
versity, New Orleans, Louisiana.

Manuscripts should be submitted on good paper, as original type-
written copy, double-spaced, and carefully corrected.

Separate numbers may be purchased by individuals, but subscriptions
are not accepted. Authors may obtain copies for personal use at cost.
Address all communications concerning exchanges, manuscripts, edi-
torial matters, and orders for individual numbers to the editor. Re-
mittances should be made payable to Tulane University.

When citing this series authors are requested to use the following
abbreviations: Tulane Stud. Zool.

Price for this number: $1.00.

Assistants to the Editor:
Carol L. Freret
Donald W. Tinkle

George Henry Penn, Editor,
Meade Natural History Library
Tulane University,
New Orleans, U. S. A.

THE FISHES OF TWO GULF COASTAL
MARSH AREAS OF FLORIDA

JOHN D. KILBY.i

Department of Biology, University of Florida,
Gainesville, Florida

The low shores of the currently exposed peninsula of Florida slowly
emerge from a much larger land mass, the almost flat Floridan
Plateau, barely covered by the waters of the Gulf of Mexico on the
west and the Atlantic on the east. This plateau is considered one of
the earth's most stable areas and is believed to be of great age, having
changed only slightly during its many millions of years of existence.
At times it has been exposed almost entirely and at others it has
been almost submerged. Throughout the Pleistocene the records of
these fluctuations in water levels have been preserved with sufficient
clarity for geologists to show at least seven old shorelines during
that period. The intermittent rises and falls of the water levels,
although not great, made conspicuous changes in the amount of land
exposed, reducing it at times to only a few small islands and increas-
ing it at others to an area at least double that of the present. In
addition to these events, there appears to have been a long-time but
gentle tilting of the plateau toward the west, which has gradually
and almost imperceptibly drowned a large segment of the Gulf coast
of the state and at the same time elevated, with equal gentleness, a
major fraction of the Atlantic coast, {cf. Campbell 1940; Cook 1945).

Thus, where the shallow seas which cover the submerged portions
of the Floridan Plateau meet the almost flat shoreline of the penin-
sula, a wide intertidal zone results along much of the coast. This
zone, between land and sea, is particularly conspicuous in estaurine
localities where it often extends many miles inland, at the southern
tip of the state where it is covered primarily by mangrove swamps,
and along the Gulf side of the northern half of the peninsula where
extensive coastal marshes have developed. If we can judge the past
by the present, it is not difficult to conclude that intertidal marshes
have been long-term features of the Florida coast; and certainly since
the Pleistocene they have been much as we see them today, an
ecologically distinct habitat fraught with the inhospitalities of un-
stable water levels, fluctuating temperatures, and rapidly varying
salinities.

Although this intertidal (or "coastal" or "salt" or "brackish") marsh
reaches its maximum extent on the Gulf coast of the upper peninsula
of Florida, it is found in greater or lesser degree all along the more
than twelve hundred mile shore of the state, and everywhere it harbors
a fauna of fishes conspicuous for its variety of species and its abun-
dance of individual representatives. Questions concerning these fishes
are: What is the composition of the brackish marsh fauna? Was it

1 Contribution from the Department of Biology, University of
Florida.

MAY 1 6 1955

"MTV

176 Tulane Studies in Zoology Vol. 2

derived from fresh waters or the sea? Or, perhaps, is it unique in
itself? What are some of the more conspicuous ecological factors of
the marshes that mold its fish fauna?

Apparently no intensive ecological or faunistic study has been made
of brackish marsh fishes, although there are several studies of the
fishes of bays and other inshore waters. In addition, the problem of
the relations of fish distribution to water salinities has received no
detailed attention in Florida. As Gunter (1938, 1942, 1945) has
pointed out, this relationship is known imperfectly for most species
of fish which are neither wholly marine nor strictly fresh water in-
habitants. Yet the members of this euryhaline group include some
of the most important commercial fishes of the world, and in Florida
much of the income from fisheries is derived from species which
make extensive use of the brackish waters of the state.

The present study was designed to gather data about the fishes of
the brackish water marshes as well as of those species which cross
that zone by moving back and forth between marine and fresh waters.
The principal objectives of the undertaking were to determine which
fishes were present and to what extent they inhabited the brackish
marshes. A secondary aim was to investigate the influence of salinity
on the distribution of marsh fishes. Incidental to the work, some
information was obtained relative to the life histories of the species
encountered. Certain of these findings are included in the present
paper and others have been presented previously by the writer (Kilby,
1949).

Two marshes were selected for the field work. One, at Bayport,
receives fresh water from river discharge; the other, at Cedar Key,
has a negligible dilution of its water from stream flow. The areas
proved to be quite different in both their physical and faunistic
features.

Monthly collections of fishes made during 1948 were augmented
by many preliminary collections during the preceding six months.
Collecting was done in and adjacent to the coastal or salt marshes
bordering the Gulf of Mexico. In addition, a number of exploratory
collections were made in the shallower waters of the Gulf, the bays,
and the fresh waters close to the mainland margins of the salt marshes.

Although no detailed faunal study of intertidal marsh fishes has
been attempted in North America, several workers have been con-
cerned with the effects of salinity on fish distribution and the fish
populations of bays and other open waters near the shores of the sea.
Foremost are Gunter, who studied the temperature and salinity ef-
fects and the fish faunas of Texas bay waters; Longley, whose find-
ings of a twenty-five year study of the fishes of the Dry Tortugas
were completed and edited by Hildebrand; Hildebrand and Schroeder,
who compiled a pioneering work on the fishes of Chesapeake Bay;
and, more recently, Reid, who worked on the eulittoral zone fishes
of Cedar Key. Without the findings of these and other workers,
this study would never have gained much of its substance and, with-

No. 8 Kilby: Gulf Coastal Marsh Fishes 111

out the challenge presented by Homer Smith's (1943) stimulating
lectures on the kidney, it is doubtful that it would have been
undertaken.

Methods
An attempt was made to collect random samples of the fish popu-
lations. This goal was difficult to maintain and when bias was
recognized appropriate note was taken. D-type dip nets, tea strainers,
and poison were employed but by far the most frequently used col-
lecting devices were seines of ten and twenty-five foot lengths having
a depth of four feet, mesh of a quarter inch, and a four-foot bag
located in the center. Whenever possible, the seine hauls were planned
so that the body of water being worked was covered by the net from
shore to shore and from top to bottom. Seining was continued until
it was believed that practically all species of fish present were repre-
sented in the collection. The details of these collecting procedures,
and those of the methods which follow are available in a manuscript
in the University of Florida Library (Kilby, 1950); consequently they
are given here with considerable abbreviation.

The fishes were preserved in ten percent formalin and each col-
lection was tagged by a field number identical with that on a field
catalog card bearing entries descriptive of the station at the time of
the collecting.

McBee Keysort Cards were used for the field catalog, and as the
fishes of each collection were identified and examined in the laboratory
many of the resulting data were recorded on the card for the proper
collection. Subsequent coded punching of the cards permitted rapid
recovery of the data.

Unless otherwise stated, all measurements are in millimeters.
Lengths are standard lengths and were made with the aid of a fish
measuring board.

Salinities were calculated in the laboratory from water temperature
and density data collected in the field at the time each collection was
made. The equipment and the method varied within permissible
limits from the technique employed by the U. S. Coast and Geodetic
Survey as described by Schureman (1941: 81-85).

The calculation of interspecific association coefficients, as im-
proved and explained by Cole (1949), was used to answer the fol-
lowing questions: "Are these two species of fish occurring together
more (or less) often than is to be expected by chance?" "To what
extent do the two species appear to be associated either positively
(occurring together), or negatively (occurring separately)?" "How
reliable (in terms of its standard error) is the obtained coefficient?"

Calculation of the coefficient produces two figures: the coefficient
itself (C) and its standard error (S E). For the species pair Fundulus
c. confluentus and Cyprinodon v. variegatus at Bayport, for example,
C±S E = 0.81 ±0.1 3. This means that the two species were found
together frequently, and the best measure of the association is 81

178 Tulane Studies in Zoology Vol. 2

percent. (If two species were always together, C^+i-CK); if their
association was random, C = 0.00; and if they were never together,
C := — 1.00). The standard error gives an estimate of the reliability
of the coefficient in the usual statistical sense and consequently it
decreases in magnitude as the number of collections increases.

Other statistical techniques and terms used in the treatment of data
are based on those recommended by Simpson and Roe (1939). All
errors shown are standard errors.

With the exception of a few specimens deposited elsewhere, all of
the fishes taken during this study are available in the University of
Florida Collections maintained jointly by the Florida State Museum
and the University's Department of Biology.

The Cedar Key Area

The extreme diversity of the coastal physiography at Cedar Key
can be appreciated fully only by an aerial view. Complexity of the
land and water features of the intertidal zone, which ranges up to
three miles in width, is a result of the irregularities in the shore line
of the mainland, the presence of numerous islands and oyster bars
in the tidal area, the fringing beaches of the Gulf, and the tortuous
passageways which facilitate the tidal ebbs and flows. To the eye
of a person on the ground the extensiveness of the open water in the
intertidal zone is minimized by the presence of emergent vegetation
which covers, and also fringes, the shallows around the islands and,
to a lesser extent, many of the larger or higher oyster bars. The
dominant emergent hydrophyte of the tidal area is oyster grass,
Spartina alterniflora Linnaeus, but occasional strips and patches of
black rush, J uncus Roemerianus Scheele, are conspicuous, especially
near the mainland and in the vicinities of the larger islands. On top
of the oyster bars and skirting the shallows adjacent to the islands,
black mangrove, Avicennia nitida Linnaeus, frequently forms dense
thickets; these, as much as the sweeps of grass, contribute to the
impression that the tidal area is much more heavily vegetated than
it is shown to be by an aerial observation.

Elevations in the area, although not great, have a marked influ-
ence on the vegetation. The mainland near the coast, for instance,
is a mere three to five feet above normal high tide level and, from
the evidence of the change in tide levels, the gradient across the tidal
area is approximately one foot per mile. Thus, an oyster bar which
is only twelve inches above the surrounding mud flats may support
a dense growth of Spartina, or even Avicennia, and the slightly more
elevated islands and the mainland produce pine woods.

At times of extreme high tide the entire tidal area is submerged
and takes the appearance of a vast sheet of water being occasionally
interrupted by the more prominent islands. Conversely, at extreme
low tide, all waters are withdrawn except for those in pools and chan-
nels. At such times one would not hesitate to walk through the
marshes from the mainland to the town of Cedar Key at the edge

No. 8 Kilby: Gulf Coastal Marsh Fishes 179

of the Gulf and could do so without getting in water over knee
deep. Such a trip, however, would impress upon the venturer that
mud is one of the most conspicuous features of the entire area. Mud
overlays all of the surfaces of the marshes, except for the outlying
strips of beach where white, wave-worked sand may be encountered.
The mud is a soft, black, highly organic ooze of considerable depth
into which a walker may sink as much as three feet, but usually
it is sufficiently compacted to permit fairly easy progress on foot.

In planning the collecting at Cedar Key it was considered logical
to follow a transect across the marshes in an attempt to encounter as
many different fish habitats as possible, and to work all such stations
on an intensive basis for at least a consecutive twelve-month period.
It was soon evident, however, that the number of stations had to be
limited, although a good means of access to the marshes exists by a
road which could be considered a transect line. As a result eleven
stations were picked for periodic examination along the road. Five
of these were pools situated inside the marshes near the mainland;
two were similarly constituted pools near the Gulf margin of the
marshes; three were bayous or creeks penetrating the marshes near
the mainland; and three were shallows along the shores of the larger
bays and of the Gulf. The characteristics of these stations, which
are designated by the terms Inner Pools, Outer Pools, Water-courses,
and Open Waters, respectively, are outlined in Table 1. The major
features of these four categories of habitats are summarized in order
to emphasize some of the more pertinent factors of each of them.

Inner Pools

For the purposes of this study a pool is defined as a small, shallow
body of water usually devoid of emergent vegetation, which either
may retain water during periods of low tide, or become entirely
drained at such times. The inner pools at Cedar Key were all sur-
rounded by expanses of oyster grass, Spartina alterniflora. They con-
sisted of slight depressions in the surface of the surrounding mud
which, in addition to supporting the emergent vegetation, were
flooded regularly at high tide. Thus, at extreme low tide some of
the pools were drained completely while others, further removed
from channels or open waters, might retain up to three and one-half
feet of water. At high tide, the influx of water not only raised the
levels of the pools, but overflowed into the adjacent marshes in all
directions. At such times, fishes could and did leave most of the
pools and either fed or took refuge in the surrounding marshes. This
flooding also permitted the fishes free access to all parts of the
marshes, and therefore facilitated their movements from area to area.
The receding waters of the ebb tide forced the fishes to concentrate
in the pools, water-courses, and margins of open waters. At such
times of concentration they could be netted with facility and con-
sequently nearly all the pool collecting was done at low tide.

The salinity of the pool waters varied through a greater range (1.2
to 35.6 ppt) than that for any other of the groups of habitats, and

180

Tulane Studies in Zoology

Vol. 2

<
<

<
O

H

o

H
X
H

h4

DO

c

PQ £
<!

g

H

o

H
pj

o
u

fc.
o

co

O

H
Hi

«

O
02

w
Q

OB

^*,

a

fl

V

Ej

EQ

c8

&

-4->

CI

j=

-c

-rr

-4—

efl

<

-4—

c

m

o

c

££

o a.

Z«2

Ph

«

ZO

3J;5

sa-B

S E

■a 9
2b

i

i-l

«

H

c

y

c

a

a

■a

7-

o

fc

O

Ph

00 cs-^*
e c-S
5 H s

.§■"3 s

is lit

CO

6

- °'s

<o-3

^c^-i

•2 °

Saw

"SCioo

CO

■

o

3

_S '-Of CO

"g -CO-J

8 -S3

Os so G P *

CO -fc.

§>u5c"»

.*> ^ ~H

•I S§-2

«

a

a

o

o

&

z

Z

T3

a

o 2

i -o

a -73

.* a

^ a

T3 3

3 o

§*

I'a

«"3

CO

"a

>, SP

i

i

53.9.

P-m

o_S

Ph

Ph

in 5

a;^

T s

<->a

°a

oa

0)

o
o

o

*m

PM

"C

H-d

<^< h

co-c

^<

c

a

£^ s

co"-SfS
c B •

. oo*"1

^^ s

c

a

o

o

£

Z

i"0

o 3

^a

-D

2 »

J2'5

co

pa ca

M

I

Pi xn

Ph

If

o

lO

(M

C^

lO

^

e3

a

3

K

-

a

"o

u

S

•*3

Ph

■^

-r

(-.

\n

c

o

3Hi

—
o
o

Ph

OS
W
H
!3

C

"3 h .

§" .d

"Co-"1
St ..»

coi, e

atr .

-g HtffN

115
U0

i -O

°a

J2 3

bD

Mm

s= a

■§«*

Sol

■a
•-S >,

M O

& a

-> «3

^ —

oPl,
P3

No. 8

Kilby: Gulf Coastal Marsh Fishes

181

c a

.■^co"
O-* g

-«* — 2

^s ft.

1 So

•po

rn

00

w

o

rn

CO

«

W

W

P

H

o

■<

u

£

t-
<

s

"5

q3

w

Oh

o

o

-7= 0) C

~ as

'J-ga

03

4

i— *•*

J ■!»

3 S-2

OH .a
■pi -** c»

£ ££ -

S -O-S

°o:= 5.

&a2 & °

_.5 -s
US-S

o >>

o us

■3

*c

-e n

,~ g ^ c

U . *— ' »

Boc-§*
to

x

a_

e >-

c t_

C3"D

rt H

CD

pj O m

•si

Im

•a g

a

3 >>-0

3 >>X1

CQ

s°

2°

a

..5 .

J? c

Flat;
des of
steep

CO

>>

OS

a

~-S

©.3

c-5e

■^

00

o - e
"-2 cc to

o
o

«

-a S~
a © —■

pq

182 lulane Studies in Zoology Vol- 2

the same is true for temperature, which showed a range of 12 to
37.5 °C. This observed variation is not unexpected when it is con-
sidered that the tide waters are spread over the muds of the marshes
in a thin sheet, usually to a depth of only a few inches, and that
under these conditions the effects of precipitation, evaporation,
radiation, and dilution by seepage waters are quickly discernible.

In these areas the characteristic combination of species is Mollienesia
latipinna, Cyprinodon v. variegatus and Adinia xenica ( Table 1 ) .

Outer Pools
The outer pools at Cedar Key are substantially like those nearer
the mainland except that they are closer to the high salinity waters
of the Gulf and their areas of surrounding marsh are less extensive.
These factors probably account for the higher mean salinity encoun-
tered, its smaller range (15.4 to 30.2 ppt), and possibly the more
moderate temperature range (16° to 36°C) in comparison with the
inner pools.

Species were more numerous, but not necessarily more abundant in
individuals, in the outer pools than in any other habitat except the
Gulf shore waters. The variety is due to the presence of the marsh
species plus additions from the nearby Gulf.

In addition to Mollienesia latipinna, Adinia xenica, and Cyprin-
odon v. variegatus, which were also found abundantly in the inner
pools, appreciable numbers of Fundulus similis always were present
in the outer pools.

Water-courses
The bayous, creeks, tide guts, and channels were grouped together
under this title, and all collections from such habitats were considered
together after a separate study showed that the similarities between
stations warranted such a procedure.

At Cedar Key the water-courses studied were in the mainland edge
of the marsh and proved particularly difficult as collecting stations.
Seining during high tide was unproductive and at low tide was not
much better. Apparently no species is well adapted to the swift
waters and the rapidly changing levels of the bayous.

Salinities in the water-courses near the mainland are low to medium
(0.8 to 17.5 ppt). This condition is due almost entirely to dilution
bv underground seepage water from the mainland, and this fresh
water is supplied in sufficient quantities to offset the rise in salinity
of the marsh waters due to evaporation. At dead low tide the bayou
water does not taste salty and can be seen seeping into the channel
from the sands along the banks.

The species which are characteristic of the water-courses include
Fundulus g. grandis and Fundulus similis when taken with Cyprinodon
v. variegatus, Adinia xenica, and Mollienesia latipinna. Also char-
acteristic, however, is the size of the specimens found. The pool
fishes usually are of all sizes, with an emphasis in numbers on the
smaller individuals, whereas the fishes of the water-courses are almost

No. 8 Kilby: Gulf Coastal Marsh Fishes 183

always adults or at least half-grown individuals.

Open Waters

This category was assigned to include those stations where the
collecting was accomplished in the shallows along the margins of the
bays and the Gulf. These habitats v/ere examined principally in
order to determine, in part, the availability of species for populating
the marshes. Collecting was done by seining the shallows both parallel
and perpendicular to the shore, and the exploring was kept within
two hundred feet of the land.

Salinities of these open waters showed a short range varying from
16.3 to 31-1 ppt for all locations. The range, however, was much
less for individual stations. Temperatures were neither quite so low
nor quite so high as at other stations, and it is likely that if the data
had been collected further from shore, the range of both salinity and
temperature would have been lessened.

Fundulus similis and Menidia beryllina were present in most of the
catches, but other species were caught with less regularity.

The pools, water-courses, and inshore waters at Cedar Key are
under the influence of the nearby offshore waters from both a physi-
cal and a faunistic standpoint. The offshore region presents a variety
of fish habitats and the fish fauna is rich in both species and numbers
of individuals. One of the more obvious causes for the abundance
of fish found in the offshore region lies in the luxuriant and extensive
bottom vegetation. Such is not the case in the intertidal zone where
submerged aquatic plants are so poorly represented that they could
not be of real importance in either the food or the cover relationships
of the marsh fishes.

The causes for this paucity of submerged vegetation may rest
partly in salinity. Although the salinity range is extensive at Cedar
Key (0.8 to 37.6 ppt), the mean salinity is high (21.1 ppt) (fig. 1).
Such a high mean salinity appears to prevent the development of the
dense stands of aquatic spermatophytes so often found in the fresh
and the mildly brackish waters of Florida.

Water temperature data are shown in figure 2. The range was
from 12 °C to 37.5 °C; the mean was 25.9 °C. The Bayport temper-
atures appear to be less variable than those for Cedar Key (fig. 2).
The Coefficient of Variation for temperatures taken at Cedar Key is
25.4% and the equivalent for Bayport is 17.4%. These coefficients
support the above statement that the Bayport temperatures are less
variable than those at Cedar Key. (The difference between the co-
efficients contains its error 4.2 times. This indicates that the differ-
ence is real insofar as these data give a true picture of the situation.)

The Bayport Area
The Bayport area has much in common with that at Cedar Key,
but an aerial view emphasizes the differences. From a physiographic
viewpoint the Bayport tidal area appears much less diverse than the

184

Tulane Studies in Zoology

Vol. 2

one at Cedar Key; although the mainland has an irregular outline,
the tidal area is dotted with islands, and the Gulf margin of the land
is tortuous. The most obvious simplifying factor is the absence of
expanses of open water within the tidal zone. At Bayport the mud
flats are vegetated principally with dense stands of black rush, Juncus
Roemerianus, which is intermixed with small patches of oyster grass,
Spartina alterniflora. Thus the oyster bars are eliminated, as well as
the mud-bottomed embayments of open water which are so con-

SALINITIES

PERCENTAGES OF OBSERVATIONS BY CLASSES

i-
z

o

a.

o

z

V)

<

Bayport Cedar Key

129 observations 117 observations

i 17.05 n

4 25.58 Wmmm
7 26.36 mmmm

I 0

13

i 6

I 9
22
25
28
3 I
34

37

13.95 m%m&

9.3 0 9M.
5.43

?

Mean- 7.1

^ 2.56

rf 2.56
1 256
! .86

I 5.98

iiiHii

5.13

.86

11.71

Mean = 21.1

2 0 10 0 10

PER CENT

2 0

Figure 1. Salinities. Salinities, in parts per thousand, are shown
by percentages of total' observations for Bayport and Cedar Key
respectively.

No. 8

Kilby: Gulf Coastal Marsh Fishes

185

spicuous at Cedar Key.

Another major difference lies in the width of the two intertidal
areas. At Bayport the zone is about one-third that at Cedar Key and
this circumstance is reflected in the distribution of the fishes. At
Cedar Key a fish must cross the mud-bottomed, oyster-bar-dotted
embayment to move from areas near the Gulf to areas near the main-
land, whereas at Bayport ready access by a much shorter route up
the bayous is available for the crossing.

A third difference of importance is the influx of fresh waters into
the Bayport area in sufficient quantities to affect materially the
salinity, and quite possibly the temperature, of the water. The two

WATER TEMPERATURES

PERCENTAGES OF OBSERVATIONS BY CLASSES

Bayport
126 observations

Cedar Key
I 7 observations

z o

10 0 10

PER CENT

2 0

Figure 2. Water Temperatures. Temperatures, in degrees Centi-
grade, are shown by percentages of observations for Bayport and
Cedar Key respectively.

186 Tulane Studies in Zoology Vol. 2

major streams emptying into the Bayport marshes are the Mud and
the Weekiwatchee rivers. Both are of spring origin and receive
some contribution from runoff along their courses. The water of
the two streams is faintly brackish due to the mild salinities of some
of their springs. This almost-fresh water of the rivers tends to
reduce salinities in the marshes and also, because of the relatively
constant temperatures of the springs, to partially stabilize the tem-
perature of the diluted marsh waters. Figure 2 presents the tem-
perature data. As has been discussed previously, the temperatures
at Bayport appear to be less variable ( Coefficient of Variation =
17.4%) than those at Cedar Key (CV. = 25.4%).

Despite these major differences, which have an effect on the fish
populations, the transect method of selecting collecting stations was
again employed, and the habitats worked were grouped into the same
categories as described for Cedar Key. These groups are discussed
in the following paragraphs, and the major features of the individual
stations are summarized in Table 2.

Inner Pools

Four salt marsh pools and a group of mainland fresh water pools
were selected. The inner marsh pools were near the mainland, and
the fresh water ones were directly connected to brackish water by
their drainages, but were elevated sufficiently to escape salt water
influx except during maximum storm tides. Ordinary high tides
rarely reached the inner pools of the marshes but, as a rule, the bi-
weekly spring tides were sufficiently high for the purpose.

The fresh water pools differed most conspicuously from those of
the marshes by being vegetated with cattail (Typha) , pondweed
(Nais), widgeon grass {Ruppia) , spike rush (Eleocharis) , and arrow
head (Sagittaria) . Together this vegetation gave a wholly fresh
water aspect to the habitats. Salinity in these pools, however, reached
7.5 ppt on one occasion, but usually it was much closer to zero. The
marsh pools maintained salinity means of 10.4 ppt or less, although
the salinity range was from 0.0 ppt to 24.7 ppt.

The temperatures recorded showed a range of 14.5 °C to 32 °C.

At Cedar Key none of the inner pools was vegetated, but at Bay-
port some of the inner marsh pools contained submerged aquatics
and these plants were present during all months.

The inner Bayport pools contained Mollienesia latipinna, Cyprino-
don v. variegaUts, Gambusia affinis holbrooki, Lucania parva. and
Fundulus confluentus as constant and conspicuous elements of the
fauna. Adinia xenica, which was also conspicuous in the inner pools
at Cedar Key, was found only rarely at Bayport.

Outer Pools

The three outer pools were close to the bay area at Bayport and
were, as at Cedar Key, very similar to the inner pools.

All of the outer pools were vegetated with Ruppia or a combina-
tion of Ruppia and Chara. The vegetation, however, was sparse, and

No. 8 Kilby: Gulf Coastal Marsh Fishes 187

tended to be in patches. Water levels consistently rose and fell
almost the full extent of the total change for the area in all of the
outer pools.

Salinities showed a maximum variation at a given station of about
11 ppt, and the range for all stations taken together was 3-3 to 17.1
ppt. The mean salinities for the outer pools were 8.3, 9.1, and 10.4
ppt, which are high for the Bayport area (Table 2). Temperatures
of from 14.5 °C to 33 °C were recorded.

The fishes which were both prominent and consistently present in
the outer pool collections were Menidia beryllina, Eucinostomus argen-
teus and Lucania parva.

Water-courses

In keeping with the practice at Cedar Key, the channels, bayous,
creeks, and tide guts were grouped into this category. Unlike the
Cedar Key habitats, however, these were heavily vegetated over much
of their bottoms with dense growths composed principally of Ruppia,
which formed mats from the bottom to the surface. Open water
areas existed, however, and where no vegetation grew the bottom
was of sand, rock, or shells. Elsewhere, an organic mud was present
and this overlaid the calcareous rock substratum for various depths
up to several feet.

The water-courses facilitated the movement of tidal water and some
of them assisted in the discharge of the influx from the Mud and
Weekiwatchee Rivers. Currents were never more than about four
miles per hour and were usually barely perceptible except in the main
channels during the latter half of the ebb.

Salinities in the water-courses ranged from 1.1 ppt to 11.8 ppt and
their means were between 4 and 7 ppt. Temperatures varied from
15°C to 35°C.

The water-courses contained more species of fish than any of the
other groups of habitats at Bayport. This was due to the presence of
both fresh water and Gulf species, which were added to the normal
marsh complement.

Prominent in the list from the standpoint of both numbers and
regular appearance were Lucania parva, Menidia beryllina, Eucinos-
tomus argenteus, Mollienesia latipinna, Gambusia affinis holbrooki
and Eundulus g. grandis*

Open Waters

The shallows adjacent to the Gulf beach were assigned to this cate-
gory. The floor of the Gulf at Bayport slopes very gently, an esti-
mated eight inches in the first quarter to half mile. The bottom
inshore is vegetated to the normal low tide mark, and between low
and high tide levels a dense stand of black rush, J uncus Roemerianus,
is present. During low tides, seining near shore in the shallows
proved to be productive and it was here that collecting was done.

Salinities ranged from 2.0 ppt to 16.2 ppt but averaged only 8.4
ppt, which indicated the diluting effects of the river discharges in

188

Tulane Studies in Zoology

Vol. 2

CO

55

O

t— i
H
<
H
02

O
55

H
U

w
►J
-
o
u

H
M

C

Ph

PQ W
H o

CO

o

H

Ph

«
u

CO
H

Q

«

s

s
p

02

co^, o.

■sis

-< J—

CO -^ fc

S Oj3

<3§°

.CO

o ^

• Cv
o a

55 <»

H
Ph
Ph

:s

PS

zo

-1

a . g

"o -P 'cc

pq1-

d

•?.

5 o>

=

K!

-.

<U

f

OJ

ti

a

&

a

O

5 f

55

fcto •»

.2 3

00 s u

sf<*s

s Sic t
. ■* -fe ^_-

CO

1-1

o

O
Pi

PS
W
55
55

o

c

d

V

ed

—

id

CO m o
. <i) § ^

§.S ■ "*«
= fe.^ — -

■£%«;© to

~« e o
*> — a; J?

■« C T

.00 S>

.©CO

00

Jfe.~

!»• -

;s.»l

£too2^

•SQO «

• o tco

^ -O

5«5 K
to^S

o

o

W;

...

-*-»

d

17)

-

0

Ph

ON «

31

a

c 'S r "

2 a .

•e'-J'-J .

o
o

>^a r

rt OT

Only

storm
tides

^3

J3

43

^^ CO

-^»

■73

CO

a
o
55

A few
sprigs o
Ruppia

Dense
beds of
Chara

a,
o.

3
3h

.2 -a

a.2J=

§

3-Q

-s

ce
. 3

£ a

o
-J2

ft, bla
rganic
mud

. o

^ o u

O rt o
CC O o

Soft,
Icareo
ud an

—
a

E9

CO

03

sf a

o o

"3

g a

o

cc

3

Uj

Flat;

undercut

banks

•^ t*

— .2
"Hi:

1

*C CO

'U

Ph

tn
3

.2

ca
>

QJ CO

fcPl

■5i b

•°* c -

• o cj .

■r

CO

>-i

O

3

Ph

«

W

H

U>

O

M

s

cti

* -

"3 3X5

■StfO

Pn-»M
O

' -^

g-ajna

?^ 3

S " a

S^T3

™ o S

^^ CO

>».S 1

S t* c

pq

« a

to g o

No. 8

Kilby: Gulf Coastal Marsh Fishes

189

1 r- os ■«
ojj o §
tad -H-5

»:»§

3 g w «
c --=. ■

i~ ~ 00

2^-od

c_; 2 <=>' -~

w — — -a

03

.•8

ig1
= 03C

«

iff

u-1

si

1 1"

Si

S?2 &oi

■>N

CO '
3 K =-.

s. ? =

J3T3

'* a

S e

0Q n

>-

3

a

C

^Z

> a

ra

"o
o

p.

CO

X

CM

Bay

margin of

marsh

Bay

margin of

marsh

m

~3
o

On

8

Sridge

ool

ti

^

On

S o -

CO

&) - ■ ,

-© o "^ co

rr,

•-J

W

fH

r/i

X

P

O

-H

o

03

W

kt

H

3

to

<

03

as:

e^"^

3 D t,

°-ga

03

~ajco
e us
'&- e

^ g-s

■CI B

•^S Set.

« ca ?

- »G

CQ

5P,

-1 CD —

TO

: 5 »

O

-^- Ci CO

O en

2 e s
J= §

Kj

a fc-5 s

R. g w Co
-J fe] Bl] fej

rn

03

e>i

Is)

00

=:

Z

W

09

—

3

o

to
a)

03

£ES

-a B'o.

iJ § o.

£1=03

•a

T3 I

C >v

■S^-o

"C3

is g a

c8=s

c >

0J"a3

o^

C3
2Q

190 Tulane Studies in Zoology Vol. 2

the area. Temperatures in the shallows varied widely from 14 °C to
33.5 °C.

The fishes most prominent in the collections were Lucania parva,
Eucinostomus argenteus, Floridichthys c. carpio, Menidia beryllina,
Fundulus similis, Microgobius gulosus and Eucinostomus gula.

The marshes of the Bayport area, like those at Cedar Key, are
bordered on the seaward side by the inshore shallows of the Gulf.
These shallows are much less diverse in habitats than those at Cedar
Key and support a fish fauna which is limited, if not in variety of
forms, then certainly in numbers of individuals. Toward the main-
land, on the other hand, the marshes come in contact with the mouths
of spring-fed streams, other marshes where the water is almost fresh,
coastal swamp waters, and the margins of mesophytic hammocks,
pine woods, and palm thickets.

Thus at Bayport the mainland waters and their fish faunas would
be expected to have a marked influence on the waters and fishes
of the intertidal zone. One of these influences is well illustrated by
a consideration of salinity. As previously mentioned, the Bayport
marshes receive a considerable amount of almost fresh water from
mainland streams. Salinities varied from 0.0 ppt to 24.7 ppt and
showed a mean of 7.1 ppt. The salinities in the marshes at Bayport
and at Cedar Key (Mean = 21.1 ppt) were conspicuously dissimilar
in distribution (fig. 1).

The more conspicuous similarities and differences between the
habitats of the two areas and the physiographic features underlying
those factors may be restated in tabular form as follows:

Cedar Key Bayport

1. The marsh area is about three 1. The marsh area is only about
miles in width a mile wide

2. Influx of fresh water is neg- 2. Influx of fresh water from
ligible, consequently: two moderately small rivers
Salinity relatively high; conspicuously reduces salin-
Range: 0.8-37.6 ppt ities, consequently:

Mean: 20.5 ppt Salinity relatively low;

Range: 0.0-26.1 ppt
Mean: 7.1 ppt

3. Water temperatures some- 3. Water temperatures less vari-
what variable able than those for Cedar Key
Range: 11.5-37.5 °C Range: 16-35 °C

Mean: 25.9°C Mean: 26.7°C

Coefficient of Variation = Coefficient of Variation =

25.4 17.4

4. The dominant emergent of the 4. The marsh dominant is Jun-
marshes is Spartina alterni- cus Roemerianus

flora

5. Submerged hydrophytes ab- 5. Submerged hydrophytes
sent from marsh pools and abundant in water-courses
water-courses and present in varying

amounts in pools

6. Offshore Gulf shallows have 6. Offshore Gulf shallows are
diverse bottoms, both vege- mostly smooth bottomed; only
tated and clean sparse patches of vegetation

in limited areas

No. 8 Kilby: Gulf Coastal Marsh Fishes 191

7. Eulittoral waters densely 7. Eulittoral waters much less

populated with fishes densely populated

All of the fishes taken from the marshes and their immediately
adjacent waters are presented in the following annotated list of species.
When the data seem to warrant it, conclusions relative to the in-
fluence of some of the ecological factors on the occurrence, abundance,
and distribution of certain of the fishes are advanced. Particular
attention has been directed to the influence of salinity on the fishes.

Annotated List of Species
DASYATIDAE: Stingrays

1. Dasyatis sabina (Le Sueur), Southern Stingray.

One immature specimen with a total length of 239 mm was taken
in July from the channel near the public dock at Bayport. The
salinity was 2.5 ppt and the water temperature 25 °C.

Lepisosteidae: Gars

2. Lepisosteus platyrhyncus DeKay, Florida Spotted Gar.

Gars were not encountered at Cedar Key, but at Bayport they were
numerous and reached sizes of about four feet in total length. Oc-
casional individuals fed in the open waters of the Gulf at least a
mile offshore, and even these few cause the bay fishermen consider-
able annoyance, especially during the summer. Field observations
indicated that the gars were most numerous in the Weekiwatchee
River, especially near its mouth, but only rarely do they enter the
strictly marsh waters.

Three gars were seen in the marshes which, when collected, proved
to be L, platyrhyncus. Two were taken in July at a salinity of 1.1
ppt and the other in November at a salinity of 2.8 ppt. All came
from dense mats of Ruppia near the channels. The November in-
dividual was 169 mm in length; the July specimens measured 445
mm and 495 mm.

Elopidae: Tenpounders

3. Elops saurus Linnaeus, Ladyfish (Tenpounder).

Ladyfish occur only in the open bay at Bayport. At Cedar Key,
on the other hand, twelve collections included 19 specimens ranging
in size from 31 to 153 mm, with the smallest taken in May and the
largest in June. These young fish were taken only during the period
April-July inclusive, and only from marsh pools. Salinities at the
times the fish were taken ranged from 9-0 ppt to 35.6 ppt (avg =
23.3 ppt).

In the open bay at Cedar Key ladyfish are abundant, at least sea-
sonally, and in order to maintain the observed numbers of adults the
young would necessarily have to be much more numerous than is
here indicated. Thus the Cedar Key marshes probably contain only
strays of the young concentrated elsewhere.

Although the number of specimens collected was small, and conse-

192 Tulane Studies in Zoology Vol. 2

quently may not represent a reliable sample of the population, from
the data available growth appears to be rapid. This trend is shown
by the following data.

Month of
Collection

Mean Length
in mm

No. of
Specimens

April

50

1

May

55

12

June

96

5

July

101

1

Clupeidae: Herrings

4. Harengula pemacolae Goode and Bean, Pilchard.

Four specimens were taken from the shallow waters of the Gulf
beach at Cedar Key during April (salinity 20.8 ppt) and October
(salinity 30.2 ppt). The October individuals measured 34 and 62
mm, respectively, and those of April were 115 mm and 125 mm. In
both instances the tide was low but rising, the water had a milky or
whitish color, the temperature of the water was the same (25°C),
and Anchoa m. diaphana was by far the most dominant fish in the
collections, constituting 81.8% of the April catch and 91.8% of the
October catch.

Engraulidae: Anchovies

5. Anchoa mitchiili diaphana Hildebrand, Southern Bay Anchovy.

Seventeen of the Cedar Key collections included 1160 specimens
of this anchovy. None was caught at Bayport. With the exception
of one fish, all were taken from either the shallow waters along the
shores of the Gulf or the larger bays (1152 specimens), or from
marsh pools very close to the beach (7 specimens) (Table 3), and
nearly all of Reid's (1954) specimens came from along the beaches.
Most were collected in October (962) but some were found in
February (1), April (109), June (16), July (29), August (1),
November (1), and December (41) (Table 4).

Sizes of 16 to 19 mm were present in June, July, and October;
specimens less than 25 mm were caught in those months, and also
in December. The two largest fish were 67 mm each. The presence
of the small fish in June and October indicates at least two breeding
seasons, or possibly a very extended one. Emphasis on the October
season is indicated by the preponderance of specimens caught at that
time, and by their small sizes. Gunter (1945: 34) found strong
indications of an extended breeding season, "probably extending from
spring to fall," and Hildebrand and Cable (1930: 288-9) found the
same season in North Carolina.

Salinities of the waters where catches included the anchovies ranged
from 6.2 ppt (1 specimen) to 37.6 ppt (41 specimens), and showed
a mean of 25.2 ppt. Six of the collections contained more specimens
of anchovies than of any other species. These showed a salinity

No. 8 Kilby: Gulf Coastal Marsh Fishes 193

range of 20.8 ppt to 37.6 ppt (mean = 28.2 ppt). The Cedar Key
area anchovies are more abundant in the areas of high salinities, and
the salinity index of 31.03 (Table 6) is the second highest for any
species. Gunter (1945: 34) took more than eleven thousand speci-
mens along the Texas coast; his data do not indicate that the fish
show a marked preference for any particular salinity. He caught
specimens in large numbers at salinities of less than 5 ppt, and in
those waters his average catch per haul was slightly greater than in
saltier places. His work, however, was done in open waters. Thus,
on the basis of present findings and those of Gunter and of Reid,
it is clear that Anchoa m. diaphana is a fish of the open bays, par-
ticularly their shallower parts, and that the degree of salinity is not
an important factor in their local distribution, provided the water is
at least mildly brackish (2.3 ppt or more).

Synodontidae: Lizard Fishes

6. Synodus foetens (Linnaeus), Lizard Fish.

Only five specimens were taken during the study and these were
from the shallow waters along the Gulf beach at Cedar Key. A 57
mm specimen was taken in August (salinity 24.4 ppt), and the other
four, ranging from 44 mm to 74 mm, were taken in November
(salinity 31.2 ppt). Obviously the fish is not a marsh inhabitant.

Cyprinidae: Carps

7. Notemigonus crysoleucas bosci (Valenciennes), Southeastern
Golden Shiner.

Seventy-four specimens were taken at Bayport in four collections
obtained from mainland waters. Salinities varied from freshwater,
in a tributary of the Weekiwatchee River, to 5.6 ppt in a pool near
the marshes subject to an influx of salt water at times of abnormally
high, storm tides. In this latter habitat, where both the aquatic
plants and the fish population indicated a fresh water condition, 35
specimens were taken in February at a salinity of 0.8 ppt, and 36
specimens were caught in June when the salinity was 5.6 ppt. Many
others, particularly adults, were observed in the Weekiwatchee River
and in the spring at its head. Thus Notemigonus crysoleucas bosci
had ready access to the coastal marshes and penetrated to their edges,
but was never found where waters were more than weakly saline,
although the fish is able to tolerate salinities up to at least 5.6 ppt.
It appears that the fish is confined to fresh water, but in areas very
near the coast it is found occasionally in slightly brackish situations.

8. Notropis petersoni Fowler, Coastal Shiner.

Only 37 specimens were caught and these were found in two col-
lections. Thirty-six were taken in the mainland edge of the marshes
at the mouth of the Weekiwatchee River when the salinity was 1.4
ppt. The other specimen was caught upstream at a salinity of 2.5 ppt.

Since no individuals were taken in the marshes proper, and the

194 Tulane Studies in Zoology Vol. 2

fish was found only at low salinities in mainland waters, it is shown
that this fresh water fish occasionally is found in weakly saline waters
near the coast.

Ameiuridae: Catfishes

9. Ameiurus natalis erebennus Jordan, Southeastern Yellow Bull-
head.

Ten specimens were taken in four Bayport collections but none was
found at Cedar Key. Five came from the Weekiwatchee River (two
collections) and the other five were caught from a pool deep within
the marshes where salinities were 2.5 ppt and 15.0 ppt on the two
occasions involved. These pool individuals were young, ranging
in size from 53 mm to 73 mm, and were taken in February and April.
The fish from the Weekiwatchee were less than 26 mm in length
and were taken in July. Adults were caught by hook and line fisher-
men as far bayward as the Channel at Bayport and were not difficult
to obtain when fished for at night in other water-courses of the
vicinity. (Except for several exploratory night collections and a
number of night observations at Bayport, all of the field work of this
study was done during daylight hours.)

The species is a well known freshwater form which, at least at
Bayport, is not uncommon in slightly saline waters and which oc-
casionally tolerates salinities up to at least 15.0 ppt.

Hemirhamphidae: Half beaks

10. Hyporbampbus unifasciatus (Ranzani), Ballyhoo.

Halfbeaks were caught at both Cedar Key (3 collections, 22 speci-
mens) and Bayport (3 collections, 3 specimens). All collections
were from open waters near the shores of bays or of the Gulf. Many
schools of halfbeaks, presumably of this species, were observed in the
shallow, open waters of the Gulf, and on several occasions much time
was spent in unsuccessful attempts to catch representatives. Un-
doubtedly these fish are among those forms most successful at avoid-
ing nets. They take flight at the slightest disturbance, and the term
"flight" is used advisedly since they almost can leave the water, some-
times skittering fifty feet or more over the surface with only their
tails submerged.

The Cedar Key specimens were taken in May, June, and July at
salinities of 21.3 ppt, 24.4 ppt, and 27.2 ppt, respectively. Sizes
were distributed fairly evenly between 60 mm and 95 mm. Of the
three Bayport individuals two were caught in April (180 mm, salinity
7.5 ppt; and 184 mm, salinity 16.2 ppt) and one in November (186
mm, salinity not taken).

Its total absence from the marshes and its presence in the bays
nearby excludes this species from consideration with the marsh fauna.

Belonidae: Needlefishes

11. Strongylura notata (Poey), Needlefish.

No. 8 Kilby: Gulf Coastal Marsh fishes 195

Seven Cedar Key collections included 14 specimens. Two were
taken in June (salinities 24.7 ppt and 26.1 ppt), two in July (salin-
ity 27.2 ppt), and ten in October (salinities 30.2, 27.3, 26.1, 27.7
ppt). Six came from tide pools adjacent to the Gulf and eight were
taken from shallow water along the Gulf beach. None was collected
deep within the marshes. Sizes ranged from 90 mm to 230 mm and
showed no observable pattern by months.

S. notata was present in twelve Bayport collections and was repre-
sented by 40 specimens taken in March, April, June, July, and No-
vember. Sizes varied from 50 mm (April and July) to 260 mm
(March and April). All were taken from either the shallow waters
of the Gulf or from pools and ditches having open water connections
with them. Salinities ranged from 0.8 ppt to 17.1 ppt (mean = 9-3
ppt). Temperatures varied from 25 °C to 32 °C.

This species and Strongylura timucu appear to be associated in the
collections; this observation is discussed under the latter species.

12. Strongylura timucu (Walbaum), Needlefish, Timucu.

Nine Cedar Key collections included 40 individuals of S. timucu.
These were taken in April, May, June, July, September, and October
at salinities ranging from 4.5 ppt to 30.2 ppt (mean = 22.9 ppt).
Temperatures ranged from 20°C (October) to 32.5 °C (September).
Four were taken near the mainland margin of the marshes. One of
these was from a marsh pool and the other three were from a bayou.
All others came from shallow waters along the Gulf shore except for
two individuals caught in a pool close to the open Gulf. Sizes
ranged from 80 mm to 190 mm. The small number of individuals
and their distribution through the months by sizes gave little clue to
the growth rates or to the breeding season.

Eighteen Bayport collections included 53 specimens caught in
January, February, March, April, May, June, July, October, and No-
vember at salinities from 3-3 ppt to 17.1 ppt (mean =10.0 ppt).
Temperatures varied between 21.5°C and 33 °C. All specimens were
obtained from either Gulf shore waters or the pools and bayous
directly connected to the Gulf by open water passageways. Sizes
ranged from 35 mm to 440 mm but gave no conclusive information
on growth rate. The 35 mm specimen was taken in June and was
the only individual sufficiently small to indicate a breeding date close
to that of the time of capture.

S. notata is rather weakly associated with S. timucu at both Cedar
Key (C = 0.38±0.10) and Bayport (C = 0.52±0.11) (cf. p. 177).

Needlefishes were seen in the Weekiwatchee River on many oc-
casions and they ascend that stream to its headwaters. It was thought
that both species here discussed were represented, but since speci-
mens were not collected the identifications are uncertain.

Anguillidae: Eels

13. Anguilla bostoniensis (Le Sueur), Common Eel.

A single 292 mm individual was taken at Bayport in November

196 Tulane Studies in Zoology Vol. 2

by net from the mouth of a large bayou. The fishermen at Bayport
occasionally catch a specimen from the brackish waters of the marshes
where I have observed them at night on several occasions.

Myridae: Worm Eels

14. My r op his punctatus Lutken, Worm Eel.

A single specimen 165 mm in length was captured among the
damp shells of an oyster bar exposed during an extremely low tide
at Cedar Key in August, 1948. That they are more abundant than
here indicated has been established by later collecting with rotenone
in shell bottomed pools near the mouths of bayous.

Cyprinodontidae: Killifishes

15. Fundulus chrysotus Holbrook, Golden Topminnow.

Nine Bayport collections included 29 specimens, but only two of
these were taken in marsh waters. The others were collected in
mainland swamps, streams, and springs. The presence of the two
specimens in the brackish marshes is attributed to accidental straying
from fresh waters nearby, but since the salinities were 15.0 ppt and
24.7 ppt at the times Fs chrysotus was taken, it is obvious that the
species can tolerate a rather high salinity, at least for short periods.
The species is a resident of Salt Spring at Bayport where salinities
usually are close to 1.8 ppt.

16. Fundulus similis (Baird and Girard), Longnose Killifish.

One thousand nine hundred fifty-four specimens were collected and
of these 1750 came from 87 Cedar Key collections and 204 were
from 34 Bayport collections. The species represented 5.23% of all
specimens taken at Cedar Key and only 0.96% of those from Bayport.

The fish from Cedar Key ranged in size from 13 mm to 111 mm;
larger specimens were observed but not caught. Bayport individuals
were from 12 mm to 120 mm. Very young fish were present in the
collections during all seasons at both areas. These occurred at Bay-
port about equally through the months, but they showed perceptible
increases in January and in May at Cedar Key. These observations
are in substantial agreement, but less marked, than those of Gunter
(1945: 44).

Growth charts on the species were not particularly illuminating,
but the one for Cedar Key suggested that the young of January prob-
ably reached a length of at least 45-55 mm by June, and that the
young of May probably reached 40 mm by late July and 55 mm in
late October. On this basis it might be assumed that specimens
exceeding 60 mm are at least six months old. These conclusions are
similar to those reached by Gunter (1945: 44).

The species is widely distributed throughout the marshes in spite
of its greater abundance in the shallow shore waters and pools where
salinities are constantly high (Table 3). Field observations repeated-
ly showed that the fish was more abundant in shore waters than
elsewhere, but the catches did not reflect that abundance in its true

No. 8 Kilby: Gulf Coastal Marsh Fishes 197

perspective because of the skill of the larger specimens in avoiding
capture. It is certain that the proportions of large fish are less in
the collections than those in the populations.

The percentage of Fundulus similis was greatest in catches made
during the summer and early fall (Tables 4, 5).

Table 8 indicates at Cedar Key a weak positive association (0.36±
0.17) between F. similis and Menidia beryllina, which are both "open
water" species; a negative association ( — 0.34±0.10) with Gambu-
sia affinis holbrooki, which is a "protected water" form; and a positive
association (0.50±0.11) with Fundulus g. grandis, which is a widely
distributed marsh form. Other correlations are not marked and
emphasize the fact that F. similis is widely distributed in the marshes.

Bangham (1942: 297) lists F. similis from the fresh waters of
Silver Springs, Florida, but this is almost certainly in error as has
been noted by Hubbs and Allen (1943: 22).

17. Fundulus grandis grandis Baird and Girard, Gulf Killifish.

One thousand nine hundred seventy-nine specimens were obtained
and of these 1476 were taken in 75 collections at Cedar Key, and
503 were from 56 collections made at Bayport.

The presence of fish of all size ranges (20 mm or less to 100 mm
or more) at all seasons of the year indicates a continuous breeding
season. This killifish exhibited increases in abundance during May,
November, and December at Cedar Key (Table 4), and during
January, June, and November at Bayport (Table 5). These are not
striking increases and are mentioned for comparison with Gunter's
(1945: 43) findings in Texas where "the fish was caught in greatest
abundance in summer and again in midwinter."

At Cedar Key F, g. grandis was taken at salinities from 0.8 to 35.6
ppt, and at Bayport it was taken at salinities from 0.8 to 16.2 ppt.
The fish was taken in all possible salinity classes at Cedar Key and
showed a slight increase above expectancy for salinities centering
around both 2.5 and 22.5 ppt (Table 6). Contrarily, the species
was more abundant at low salinities for Bayport (Table 7). Gunter
(1945: 126) showed that his Texas specimens were more abundant
at salinities less than 15 ppt. Thus it becomes clear that this species
can tolerate a wide range in salinity but reaches its population peaks
at salinities between about 5 and 25 ppt. The contradictory findings
of the present study are resolved, to a great extent, by a consideration
of habitat preferences.

F. g. grandis occurs most abundantly in outer pools, water-courses,
and, to a less extent, open waters at both Cedar Key and Bayport
(Table 3). And, as noted previously, the salinities of those habitats
average higher at Cedar Key than at Bayport. It appears, then, that
F. g. grandis is more responsive to the type of habitat than to the
salinity of its water, provided that the water is at least brackish and
not above about 25 ppt.

At Cedar Key F. g. grandis exhibits a moderate positive association
(C is 0.50±0.11) with F. similis and both of these species show a

198 Tulane Studies in Zoology Vol. 2

preference for the same habitats. A negative association is shown
with Gambusia ajjinis holbrooki by both F. similis and F. g. grandis
(C is — 0.62 ±0.12). Gambusia is a fish of the inner pools and
sheltered places, whereas both species of Fundulus under consider-
ation prefer the outer pools and shore waters. At Bayport a positive
association is indicated with Cyprinodon v. variegatus (0.50 ±0.12)
and with F. similis (0.38±0.13), and a possible negative but weak
association with Gambusia ( — 0.21 ±0.09).

18. Fundulus conjluentus conjluentus Goode and Bean, Eastern
Marsh Killifish.

One thousand six hundred seventy-nine specimens were collected
from the marshes and several hundred others were taken in the inland
fresh waters of Levy County many miles from the coast. The Cedar
Key specimens numbered 1,080 and were included in 75 collections.
Forty-eight Bayport catches accounted for the remaining 599 in-
dividuals.

Sizes ranged from 11 mm to 69 mm, but fish of less than 25 mm
were poorly represented in the collections. At Bayport, however,
small fish were taken with some regularity, being absent (or nearly
so) only in November and December. They were also absent for
these months at Cedar Key and, in addition, were not present in the
March, April, and May collections. The growth charts prepared for
this species are not particularly informative except to suggest that
the breeding season is prolonged, the mortality rate high, and that
the majority of the fish live only a single season during which they
attain a maximum size of a little over 55 mm. These interpretations
are tentative, and further observation on the species in these respects
is needed.

At Cedar Key F. c. conjluentus was taken in greater abundance
than expected on the basis of random distribution at salinities be-
tween 20 and 30 ppt, and less than expected both above and below
those limits (Table 6). The comparable optimum salinities at Bay-
port were between 10 and 30 ppt.

The species occurred most abundantly in the inner pools at Cedar
Key, but at Bayport it occurred abundantly in both the inner and
the outer pools (Table 3). These occurrences explain, for the most
part, the differences in salinity distribution, since the salinities of
these habitats vary in the two areas. These are correlated with the
habits of the species in freshwater where they are found most abundant
in sheltered pools.

A positive interspecific association is indicated at Cedar Key
(Table 8) between F. c. conjluentus and Mollienesia latipinna
(0.94±0.13), Adinia xenica (0.64±0.11), Cyprinodon vs variegatus
(0.89±0.13), and Gambusia afjinis holbrooki (0.81±0.21). Nega-
tive associations are with Menidia beryllina ( — 0.35 ±0.1 1 ) , Eucinos-
tomus argenteus ( — 0.39±0.13), Anchoa m. diaphana ( — 0.81 ±
0.17), and Floridichthys c. carpio ( — 0.64 ±0.23).

At Bayport (Table 9) the positive associations are with Mollienesia

No. 8 Kilby: Gulf Coastal Marsh Fishes 199

latipinna (0.70±0.19), Adinia xenica (0.64±0.15), Cyprinodon v.
variegatus (0.81±:0.13), Mugil cephalus (0.83±0.18), Jordanella
floridae (0.79 ±0.18), and Ga??ibusia aj finis holbrooki (0.52 ±0.14).
Negative association is indicated possibly for Eucinostomus argenteus
( — 0.34±0.12). These associations emphasize the occurrence in
the inner pools and the sparcity of numbers found in the water-
courses and the shallows of the bays and Gulf.

19- Adinia xenica (Jordan and Gilbert), Diamond Killifish.

Five thousand three hundred and three specimens were collected;
of these 4,939 were taken in 84 collections at Cedar Key, while 364
came from 22 Bayport catches.

Sizes ranged from 12 mm to 42 mm. Specimens of 15 mm or less
were present in nearly all months (fig. 3). Thus the breeding sea-
son at Cedar Key seems to be continuous, except for possible breaks
in the March-April period and again in December. The rate of
growth appeared to be rapid. Considering the period June-September,
it seems that certain of the size groups can be followed through the
several months. For instance, the 19 mm group of June is possibly
the 22 mm group of July, the 24 mm one of August, and the 25 mm
one of September. This same trend of size increase is developed by
following the maximum-size group for the same months, and is
substantiated to a limited degree by the record of a 9 mm specimen
of December which grew to 37 mm in twelve months in a large
aquarium.

The monthly catches for Cedar Key (Table 4) indicate that the
population reached a peak during the fall, maintained a high level
during the winter, declined in February and March, and increased
from that low through the spring and summer to the fall peak.

At Cedar Key considerably more specimens of Adinia were caught
where salinities ranged from 20 to 24.9 ppt than from any other range
of comparable size, and that the percentage (56.29%) at these salini-
ties was greater than expected (36.68%) when assuming a random
distribution (Table 6). The comparable Bayport figures (Table 7)
indicate a greater abundance in salinities between 15 and 20 ppt
than in any other class.

At Cedar Key this species is distributed by habitats almost the
same as are all other species taken together ( Table 3 ) . There existed,
however, a tendency to avoid open waters and to be more abundant
in inner pools. Table 3 indicates a strong tendency at Bayport to
avoid inner pools and open waters and to concentrate in the water-
courses and particularly in the outer pools. These data suggest that
the fish prefers medium salinities between 15 and 25 ppt, and that
outer pools and water-courses are chosen for habitats. The fish occurs
in much greater numbers at Cedar Key than at Bayport, possibly
because the salinities in the former area more nearly meet its require-
ments.

At Cedar Key Adinia xenica showed a positive association (Table

200

Tulane Studies in Zoology

Vol. 2

8) with Mollienesia latipinna (0.72±0.11), Cyprinodon v. variegatus
(0.70±0.11), Gambusia a, holbrooki (0.77±0.25), and Fundulus
c. confluentus (0.64±0.11). Negative association was indicated be-
tween it and Menidia beryllina ( — 0.35 ±0.11) and Anchoa m.
diaphana ( — 0.67 ±0.15). These associations add strength to the
indications that this species avoids open waters and is most abundant
in medium salinities. Positive Bayport associations are with M.
latipinna (0.67±0.32), C. variegatus (0.89±0.23), F. C, confluentus
(0.64±0.15), and possibly Floridichthys c. carpio (0.37 ±0.12). No
negative associations are shown. This emphasizes the scattered nature

Figure 3. Adinia xenica,
frequency curves.

LENGTH IN MM
Cedar Key specimens.

Monthly length-

No. 8

Kilby: Gulf Coastal Marsh Fishes

201

of the population at Bayport where the species is not abundant and
where it is probably finding difficulty in adapting itself to a habitat
in which salinities are bordering on the unsuitable.

20. Lucania parva (Baird and Girard), Rainwater Killifish.

Four thousand two hundred and five specimens were obtained.
Ninety-eight collections at Bayport included 4,192 of these and the
other 13 specimens were taken in nine collections at Cedar Key.

20 _

15 _

10 _

5 _

JAN.
532 sp

is
io[_

5
0

20
15
10

6

0

25
2 0
15
10

5

0
IS
10

s

0

. 15

= 10

FEB.
334 sp.

MAR.
49 sp.

APR.
Ill sp.

MAY
4 7 sp.

o

<£IS
W|0
°-8

o

25
20
IS
10

s
o

IS
10
5

0
20
15
10
5
0
10

s

0
IS
10

5

JUN.
46 8 sp.

OCT.
2 0 9 s p.

NOV.
5 8 8 sp

DEC.
7 6 sp.

LENGTH IN MM

Figure 4. Lucania parva, Bayport specimens,
frequency curves.

Monthly length-

202 Tulane Studies in Zoology Vol. 2

Reid (1954: 28) found only three individuals in the shallow off-
shore waters of the Gulf at Cedar Key.

Sizes ranged from 7 to 39 mm and young were present in all sea-
sons. The size distribution by months at Bayport (fig. 4) indicates
a more or less continuous breeding season but is not informative
relative to growth rates. The use of time intervals shorter than a
month for making collections in the study of this species might have
given more growth rate information. The absence of any determinable
trends in the growth of the population renders further speculation on
the matter inadvisable.

Lucania parva was second only to Mollienesia latipinna in abun-
dance at Bayport and it was found in more collections (98) than
any other species (Table 5). Its relative abundance, however, fluctu-
ated widely from month to month. Over the year it averaged 19.64%
of all specimens taken, but in January it constituted nearly 55% of
the total catch, and in May it made up only a little more than 2%
of all the fish taken. It showed a decrease from the January peak
through the months to June when it began an ascension lasting
through August. Another peak was present in November and for
December the population took its average position for the year.

Salinities at which the species was taken varied from 0.1 to 28.2
ppt; however, 81% of all specimens were taken from waters of 10
ppt or less ( Table 7 ) .

A positive association existed at Bayport between Lucania parva
and Cyprinodon v. variegatus (052 ±0.13), Menidia beryllina
(0.66 ±0.14), and Eucinostomus argenteus (0.49 ±0.16). No nega-
tive associations were evident ( Table 9 ) •

The positive association indicated with C. v. variegatus is due in
part to the wide distribution of the latter in the Bayport marshes and,
it is believed, in part to collecting errors. As is shown later, C. v.
variegatus is most abundant in unvegetated waters, whereas L. parva
appears to avoid such places. In seining at Bayport, however, it is
difficult to avoid encompassing both situations in a given seine haul.
The same error is present in the case of Menidia beryllina, and to a
lesser extent in the association with Eucinostomus argenteus. All
of these forms, however, occupied the water-courses and thus show
an association which would not be in evidence if the seining could
have been sufficiently selective.

21. Chriopeops goodei (Jordan), Redtail Killifish.

One hundred twenty-six specimens of Chriopeops goodei (Lucania
goodei of several recent authors) were obtained from fifteen Bayport
collections and none was taken at Cedar Key. The data on these fish
are summarized:

No. 8

Kilby: Gulf Coastal Marsh Fishes

203

No. of

No. of

Size

Sal.

Temp.

Month

Colls.

Spec.

(mm)

(ppt)

CO)

Feb.

1

1

27

0.8

25

June

3

27

12-29

1.4, 5.6, 10.3

27,31,25.5

July

8

87

9-30

0.0 - 7.5
Mean: 2.8

24.5-26
Mean: 25.6

Sept.

2

8

15-24

0.0-1.8

24,24

Nov.

1

3

20-27

—

—

All but four of the specimens were taken from mainland waters
where the salinities recorded were due to a mild or periodic influx
of tide water or were derived from slightly saline springs. The four
individuals which constituted exceptions were from the Gulf edge
of the marsh and possibly were stragglers. The occasional tolerance,
however, of salinities up to 10.3 ppt would definitely enable this
species to penetrate into brackish waters and thus have an added
chance for wider distribution by moving from one drainage system to
another by way of the coastal marshes which connect their mouths.

22. Cyprinodon variegatus variegatus Lacepede, Southern Sheeps-
head Killifish.

A total of 6,191 specimens was collected. Of these 4,747 were in
91 collections from Cedar Key and 1,444 were found in 75 Bayport
catches.

Sizes of individuals ranged from 8 mm to 63 mm. Young fish of
15 mm or less were present during all months except January,
February and March, when fewer specimens were taken than for any
other period of equal length (fig. 5). In a study conducted at Beau-
fort, N. C, Hildebrand (1917) found that C. variegatus had a long
breeding season, both individually and collectively. For example,
one of his females laid eggs over a period of about five consecutive
months, and other females were found in spawning condition from
mid-March to October. Thus in the warmer waters of the Gulf, it
would not be surprising to find the species spawning on a year-round
basis.

Little information on the rate of growth is obtainable from length-
frequency curves (fig. 5).

Cyprinodon v. variegatus was taken from waters of salinities vary-
ing between 0.0 and 35.6 ppt. A preference was shown by the fish
for salinities less than 20 ppt at both Bayport and Cedar Key (Tables
6,7).

At Bayport the species exhibited a decided preference for outer
pools and a lesser one for inner pools (Table 3). At Cedar Key the
preference appeared to be for inner pools and water-courses (Table
3). These choices placed the fish in waters of the two areas where
the submerged aquatics were either absent or not particularly abun-
dant, and at the same time permitted it to avoid the shallows of the
open bays and of the Gulf. This agrees with the findings of Reid
(1954: 29) who found no individuals in the open Gulf at Cedar

204

Tulane Studies in Zoology

Vol. 2

Key except in the shallows very near the marsh borders of islands.
Field observations were consistent with these conclusions in that this
species was encountered with practical certainty where areas of un-
vegetated bottom were present in pools and water-courses. During
extremely low tides the fish were found buried in the detritus and
mud of the bottoms where they also went when disturbed.

At Cedar Key C. v. variegatus showed a positive association with
Fundulus c. confluentus (0.89 ±0.13) and negative associations with
Menidia beryllina ( — 0.56±0.15), Eucinostomus argenteus ( — 0.3 5 ±
0.10), Anchoa m. diapbana ( — 0.77 ±0.13), and Floridichthys c. carpio
( — 0.61 ±0.17) (Table 8). These associations emphasize the prefer-
ence exhibited by C. v. variegatus for marsh waters rather than for
the shallows along the shores of the bays and Gulf.

At Bayport C. v. variegatus exhibits a positive association with all
but three of the twelve species with which it is compared in Table 9,

L ENGT H IN MM

Figure 5. Cyprinadon v. variegatus, Cedar Key specimens. Monthly
length-frequency curves.

No. 8 Kilby: Gulf Coastal Marsh Fishes 205

and which constitute the bulk of the marsh fishes in that area. In
short, at Bayport C. v. variegatus is distributed throughout the marshes
to such an extent that it is to be found almost everywhere other fish
occur, despite its evident habitat preferences.

23. Floridichthys carpio carpio (Giinther), Florida Goldspotted
Killifish.

Seven hundred eleven specimens were obtained. Thirty-six Bay-
port collections included 663 examples, and 48 were taken in ten
collections from Cedar Key. Reid (1954: 29) found four individuals
in trawl catches from open Gulf waters at Cedar Key and four others
on shallow flats in the vicinity of islands.

Sizes ranged from 12 to 58 mm. With the exception of March,
when very few (17) specimens were taken, and April and August
when none was caught, young specimens of 15 mm or less appeared
in all months. These young were particularly abundant from late May
until mid-July at Bayport, and scattered in the remaining portion of
the year. From February through July the fish were of two distinct
sizes: less than 29 mm and greater than 30 mm. The young fish
averaged approximately 19 mm, whereas the older group averaged
about 50 mm. Except for a few stragglers, the older group dis-
appeared during late July and the young group grew rapidly to
average about 34 mm by late October. In November and December
very few (35) specimens were taken at Bayport and only 39 speci-
mens were caught at Cedar Key. These indicated that growth was
slower during the colder months, but some individuals of the year
reached 45 mm in December. Growth is rapid during the first
several months and adulthood is reached in one year. Fish larger
than sizes reached the first year are very rare. This may be the result
of decreased growth rate or a failure to survive the breeding season
for any appreciable time. While at Bayport the fish is most abun-
dant during the summer, at Cedar Key only two specimens were
caught between January and October.

At both Cedar Key and Bayport, F. c. carpio was found exclusively
in two habitats (outer pools and open waters) and in both 70% of
the specimens were taken from open waters (Table 3).

At Cedar Key no specimens were caught at salinities below 20.0
ppt or above 37.6 ppt; at Bayport all specimens came from salinities
between 2.0 and 15.8 ppt. These data indicate that the fish is more
severely restricted by habitat than by salinity, and that the combined
favorable conditions, i.e. open waters plus moderate salinities, pos-
sibly are accountable for the greater density of population at Bayport
than at Cedar Key.

As would be expected, F. c. carpio is shown ( Table 9 ) at Bayport to
have a positive association with species which have some preference
for open waters, such as Menidia beryllina (0.47±0.15), Fundulus
similis (0.47 ±0.09), Eucinostomus argenteus (0.74±0.13), and
Adinia xenica (0.37 ±0.12), as well as with the ubiquitous Cyprino-

206 Tulane Studies in Zoology Vol. 2

don variegatus (0.68±0.16). At Cedar Key, however, a negative
association ( — 0.61 ±0.17) exists between F. carpio and Cyprinodon
variegatus, which illustrates very nicely the fact that Cyprinodon and
Floridichthys occupy different habitats when permitted to do so by
the availability of their choices. Cyprinodon is a fish of the marshes,
whereas Floridichthys, at least in the Bayport and Cedar Key areas,
mostly occurs in the shallow shore waters of the Gulf which lie at the
very edges of the marshes.

24. Jordanella floridae Goode and Bean, Flagfish.

Six hundred forty-four examples were taken in fifteen Bayport col-
lections and none has been recorded for Cedar Key.

Sizes ranged from 12 mm to 39 mm and specimens less than 20
mm were taken during all seasons. Nearly 500 were caught in May
and the numbers taken during the other months ranged from five to
eighty-four. An inspection of the growth charts prepared for the
species showed no recognizable growth rate trends.

This fish was found in waters where the salinity varied from 1.1
to 26.1 ppt, but since only 44 individuals were from salinities below
10 ppt and over 600 were from higher salinities, the data suggest
that this species prefers saltier waters. This is not in agreement
with the known distribution of the fish, however, for it is found in
entirely freshwater situations at many points in Florida. The anomaly
arises from the fact that those pools in the salt marshes where /.
floridae was found normally have much lower salinities than happened
to be present when the two collections, which contained the majority
of the specimens under consideration, were made. Their presence in
the waters of high salinity, however, demonstrates their tolerance of
salt for at least limited periods, and this would allow the species to
use the marshes as avenues in range extension. Since no specimens
were taken at Cedar Key, it is assumed that the salinities of that
area, which are higher than at Bayport, may act as a barrier.

All of the Bayport specimens were taken from marsh pools near
the mainland and, insofar as I am aware, these records are the first to
extend the ecological range of the species into brackish waters.

POECILIIDAE: Poeciliids

25. Gambusia affinis holbrooki Girard, Eastern Gambusia.

Eighty Bayport collections included 2026 specimens, and 29 col-
lections from Cedar Key yielded 1,132 to make a total of 3,158,
which is not a high number for G. a. holbrooki in Florida. In fresh
waters, where it invades every possible type of habitat, I have taken
over 4,000 specimens in a single short haul with a ten-foot net, and,
judging from field observations, it is not unlikely that it is both the
most abundant and the most frequently encountered fish in Florida
fresh waters.

In the brackish marshes fish were caught which ranged from 7 mm
to 47 mm. Gravid females were taken in all months of the year, as
were young fish of 18 mm or less, and at any season it would have

No. 8 Kilby: Gulf Coastal Marsh Fishes 207

been possible to take a very small fish by using fine nets in the
waters where vegetation or other cover was present.

At Cedar Key specimens were collected only from the inner pools
near the mainland. Salinities in these pools are often high, sometimes
quite low, and changes are rapid. At Bayport the fish was found
everywhere except in open waters, but was most abundant in the
inner pools and in the heavily vegetated waters. These places at
Bayport normally have rather low salinities. It thus appears that
the fish is most numerous in protected waters such as shallow pools
and vegetated areas where the salinities are lowest, at least periodi-
cally, but it can tolerate salinities up to at least 26 ppt (Table 6).

Table 8 indicates that G. a. holbrooki is positively associated at
Cedar Key with Mollienesia latipinna (1±0.31), Adinia xenica
(0.77 '±0.25), and Fundulus c. confluentus (0.81 ±0.21), and at
Bayport with M. latipinna (0.64±0.11) and F. c. confluentus
(0.52 ±0.14), all of which occur in the same protected waters.
Fishes not of the protected water group such as Menidia beryllina
(—0.61 ±0.18) and Fundulus similis (— 0.34±0.10) at Cedar Key,
and Eucinostomus argenteus ( — 0.44 ±0.10) at Bayport show nega-
tive associations (Tables 8, 9).

26. Mollienesia latipinna LeSueur, Sailfin Molly.

Cedar Key Area. — Ten thousand eight hundred and forty-six speci-
mens representing 32.43% of all the fish caught at Cedar Key were
of this species which occurred in 93 of the 119 collections studied.
On this basis, this species was more than twice as abundant and
slightly more frequently encountered than the next most abundant
species, Adinia xenica, which totaled 4,939 specimens in 84 collections.

Sizes ranged from 12 mm to 69 mm as shown by the length-
frequency curves (fig. 6). Individuals less than 20 mm were en-
countered during all months of the year except March and April.
Sizes of from 12 mm to 14 mm were obtained from mid-May to the
last of September and indicate a prolonged breeding season during
the warmer months of the year, at least.

In general, the size groups tend to segregate, and it was not un-
usual to find the entire sample of several hundred specimens from
a shallow pool deep within the marshes restricted to fish of less than
35 mm in length. On the same day, a collection from a deeper pool
might not contain a single specimen less than 37 mm, and yet a
third collection would contain fish from 14 mm to 64 mm.

The calculation of growth rates from length-frequency curves is
unreliable when it is impossible to follow the development of any
one age group for consecutive months. Such a situation exists with
M* latipinna because of the extended breeding season, the tendency
to segregate into size groups, and possibly other reasons.

M. latipinna tolerates a wide salinity range and occurs in Florida in
inland freshwater habitats, as well as along the coasts in protected
areas of brackish water. At Cedar Key, specimens were collected in
waters of salinities varying from 1.2 ppt to 37.6 ppt. Temperatures

208

Tulane Studies in Zoology

Vol. 2

ranged from 11.5°C to 37 °C. Open waters of the Gulf, bays, and
bayous were avoided (Table 3). Nearly every collection made in
the small marsh pools revealed the species, but it was absent in 14
of the 16 collections along the Gulf beaches and, except for one
instance, all the other collections in which it was not found were
from the shores of bays, from creeks or bayous, or from pools very
close to the open Gulf beaches. Gunter (1945: 46) obtained only
a single specimen during his work in Texas bays where the open
waters, rather than the sheltered places, were studied, and Reid ( 1954:
29) took none in trawl catches in the Gulf at Cedar Key.

With a single exception, M. latipinna was dominant in the fish
fauna of all pools located deep within the marshes and without direct
connections with the bays or bayous. The number of individuals
was as high as 61.1% of the total catch at one such pool, and as low
as 43-9% in another. In pools having direct connections with larger

LENGTH IN MM

Figure 6. Mollie?iesia latipinna, Cedar Key specimens,
length-frequency curves.

Monthly

No. 8 Kilby: Gulf Coastal Marsh Fishes 209

bodies of water, M. latipinna was the dominant species in numbers
in two instances, but was exceeded by another species in one instance.
In the bayous, along the bay shores, and at the open Gulf beach it
totaled between 4% and 5.6% of the catches for the year, and from
4 to 5 species exceeded it in abundance at those stations.

At Cedar Key, Ms latipinna appeared to be associated positively
(Table 8) with Adinia xenica (0.72±0.11), Cyprinodon v. varie-
gatus ( 0.60 ± 0.03), Gambusia affinis holbrooki (1±0.31), and
Fundulus c. confluentus (0.94±0.13). All of these species are con-
spicuous in the fauna of the sheltered waters. Negative associations
existed between M. latipinna and Menidia beryllina ( — 0.59±0.15),
and Anchoa m. diaphana ( — 0.40±0.12). Both of these species
occur most abundantly in open water.

Bayport Area. — From the Bayport area 4,889 specimens of Af,
latipinna were collected and these accounted for 22.91% of all the
fishes caught there. This species was taken from all the habitats
studied and occurred in 95 of the 130 collections made. The next
most abundant species was Lucania parva with 4,192 specimens
(19.64%) in 98 of the 130 collections.

Sizes varied from 9 mm to 60 mm, and specimens less than 20 mm
were found in all months except February, March, and December.
Sizes of 14 mm or less were taken in June, July, September, October
and November (fig. 7).

Segregation into size groups was apparent at times, but was not as
pronounced as at Cedar Key. As was the case at Cedar Key, an
inspection of the length-frequency curves gave no recognizable in-
formation on growth rates.

M. latipinna was taken at Bayport in salinities ranging from 0.0 ppt
to 26.1 ppt but the distribution of the species in the marshes is not
a matter of salinity, but rather one of habitat. In marsh pools isolated
from water-courses, i.e., the most sheltered places, 42 collections were
made and M. latipinna was present in all but two of them. On the
other hand, 42 collections were made in unsheltered places including
the Gulf beach and the marsh water -courses and AT, latipinna was
absent from 19 of those collections. In 26 collections made in par-
tially sheltered situations within the marshes M. latipinna was absent
in five instances.

At Bayport M. latipinna appeared to be associated positively (Table
9) with sheltered water species including Gambusia a. holbrooki
(0.64±0.11), and Fundulus c. confluentus (0.70 ±0.19). No nega-
tive associations were indicated.

Summary. — At both Cedar Key and Bayport M. latipinna was the
most abundant species in the collections. In both areas it was found
in all of the habitats studied, and showed greatest abundance in
sheltered pools deep within the marshes. Frequently such pools at
Bayport were vegetated, principally with Chara and Ruppia, whereas
those at Cedar Key were not. Salinities appeared to have no deter-
minable effect on distribution.

210

Tulcme Studies in Zoology

Vol. 2

At Bayport a higher proportion of small fish and a lower percent-
age of large specimens were taken than at Cedar Key. The very
small individuals were collected, almost invariably, from the heavily
vegetated areas and, as has been mentioned, submerged vegetation
was much more abundant at Bayport than at Cedar Key. It is as-
sumed that many of the very small fish at Cedar Key were not col-
lected because they were localized in the shallows in emergent vegeta-
tion where effective seining was not possible.

The smallest specimens and the greatest abundances of the small
sizes were found during the warmer months of the year at both areas,
which indicate an extended breeding season and eliminate the pos-
sibility of calculating growth rates by an examination of the collec-
tions. Size segregation was noticeable but not absolute. The smaller
sizes were found in the shallowest and most sheltered pools, and

10

5

0

10

5

0

25

20

15

Z15-
UJ 10 _
O 5 _

-A/M^

SEP.
886 sp.

DEC.
20 sp.

Ll)

15 _

0. 10
5
0

N^~

JAN.
3 0 sp.

FEB.
68 6 sp.

MAR.
115 sp.

APR.
853 sp.

MAY
796sp

LENGTH IN MM

Figure 7. Mollienesia latipinna, Bayport specimens. Monthly length-
frequency curves.

No. 8 Kilby: Gulf Coastal Marsh Fishes 211

larger fish in the deeper and less sheltered places.

The total Cedar Key fish population was composed of a greater
proportion (32.43%) of M. latipinna than the one at Bayport
(22.91%). Two very obvious differences existed between the two
areas: Cedar Key stations showed a higher mean salinity (20.9 ppt)
than Bayport (7.1 ppt) and a paucity of the submerged aquatic
plants which were fairly abundant in the Bayport habitats. Since the
species occurs abundantly in fresh water, and was more abundant
at Cedar Key than at Bayport, and since Cedar Key salinities were
higher than the Bayport ones, salinity was not considered a major
factor in the distribution or abundance of the species. On the other
hand, M. latipinna was almost always the dominant species in numbers
at those stations, both Bayport and Cedar Key, where the following
conditions obtained: soft mud bottom, no submerged aquatics,
sheltered pool not directly connected with a large body of water or
marsh water-course. At Bayport it was frequently dominant in
numbers, even when submerged aquatics were present, in the mud-
bottomed, sheltered pools, but the degree of dominance was less con-
spicuous in these places than in the non-vegetated pools described.

27. Heterandria jormosa (Agassiz), Least Killifish.

Four hundred seventy specimens were present in a total of 30 Bay-
port collections, but none was encountered at Cedar Key.

Sizes ranged from 5 mm to 23 mm. Most of the specimens (331)
were taken in July, but this was due to their presence in a half dozen
exploratory collections during which vegetated waters were searched
intensively. Permanent, although small, populations lived in the
marshes on a year-round basis. During months other than July, in-
dividuals appeared in the collections at a rate believed to reflect their
relative position in the populations of these marshes much more
accurately than did the large catches of July. Young fish of 15 mm
or less were found at all seasons and gravid females were present dur-
ing all months in which specimens were captured.

H. jormosa was confined in the marshes to waters which were
vegetated heavily with submerged aquatics. These places were, in
order of the abundance of the vegetation, water courses, inner pools,
and outer pools. The order of abundance for the fish is the same,
with emphasis on numbers from the water courses (Table 3)-

Although the species is a well known fresh water form in Florida
and, from my observations, occurs most abundantly where the sub-
merged vegetation is thickest, it is certain that the fish can also live
in brackish marshes where it can maintain a population provided
salinities are mild and cover is abundant. H. jormosa was taken at
salinities of 0.0 to 15.0 ppt, but most of the specimens were found
where and when salinities were between 0.0 and 4.9 ppt (Table 7).
On the basis of these data this fish may be characterized as a fresh
water species which sometimes invades heavily vegetated coastal
waters of low salinity.

212 Tulane Studies in Zoology Vol. 2

Monacanthidae: Filefishes

28. Stephanolepis hispidus (Linnaeus), Common Filefish.

A single 18 mm example was taken in July at Bayport from shallow,
sand-bottomed, shore-waters of the Gulf (salinity 12.0 ppt, temper-
ature 26.5 °C). Exploratory collecting offshore in the bay at Cedar
Key indicated that this fish was not uncommon in the areas where
the bottom was vegetated.

Tetradontidae: Swellfishes

29. Spheroides nephelus (Goode and Bean), Florida Swellfish.

Sixteen immature individuals were taken in seven collections at
Cedar Key. The data on the Cedar Key collections is summarized
as follows:

Month

No.

No.

Sizes

Sal.

Temp.

Taken

Colls.

Spec.

(mm)

(ppt)

(°C)

Apr.

1

5

10-25

21.2

24.5

May

1

2

16-17

—

—

Oct.

2

5

15-32

25.9 & 27.7

21.5 & 31

Nov.

2

3

20-25

21.2 & 31.1

22 &25

Dec.

1

1

18

37.6

19

All were taken from mud flats or mud-bottomed marsh tide pools
at the edge of the Gulf. They are frequently found on the offshore
bottoms in the Gulf at both Cedar Key, where Reid (1954: 71)
found it not uncommon, and at Bayport.

SPHYRAENIDAE: Barracudas

30. Sphyraena barracuda (Walbaum), Great Barracuda.

One immature specimen, 48 mm in length, provisionally assigned
to this species, was taken in October from a pool near the Gulf edge
of the marsh at Bayport. The salinity was 16.7 ppt, and the tem-
perature 28.5 °C.

Mugilidae: Mullets
During this study more than two thousand young mullet were col-
lected and their identification was difficult, principally because of
my inexperience with the fish. During one phase of the study I was
convinced that all my material was one species, Mugil cepbalus
(Kilby 1949). Fortunately, Dr. Carl L. Hubbs read that paper,
kindly directed my attention to the possibility of the presence of
other species in the material, and furnished me with his notes on the
characteristics of the several forms. Those notes, and examination
of specimens at the Museum of Zoology, University of Michigan,
enabled me to review the material and to identify three species in it.
Dr. Hubbs then verified the identifications of representative samples
of the three species, M. cepbalus, M. curema and M. tricbodon.

No. 8

Kilby: Gulf Coastal Marsh Fishes

213

31. Mugil cephalus Linnaeus, Striped Mullet.

Two thousand one hundred thirty specimens were taken: 1,924 in
37 collections from Cedar Key; 203 in 19 Bayport catches.

Sizes ranged from 16 mm to 103 mm. This latter specimen is not
shown on figure 8 which illustrates the size distribution of the fish
by months. The growth of the young has been discussed previously
(Kilby 1949: 19), and subtraction of the specimens of M. curema
and M, trichodon from the specimens used does not change the con-
clusions reached on either the length of the breeding season or the
rate of growth. In summary, the breeding season is long, extending
from October through May at Cedar Key and from December through
July at Bayport. The rate of growth, which agrees with that of
Texas fish studied by Gunter (1945: 51-2), permits an October
individual of approximately 18 mm to reach about 65 mm by mid-
April, at which time the fish begins an exodus from the marshes

LENGTH IN MM

Figure 8. Mugil cephalus, Cedar Key specimens,
frequency curves.

Monthly length-

214 Tulane Studies in Zoology Vol. 2

resulting at Cedar Key in a depletion of the marsh population by
June. At Bayport the trends are not as sharply defined, and this is
due, probably, to temperatures, which are more moderate than at
Cedar Key.

At Cedar Key the young appeared in October collections and became
more prominent in the fauna until January, when they constituted
39.13% of the specimens caught. Through April the young were
conspicuous in the catches, but declined in May, and by the end of
June disappeared entirely. At Bayport the young first appeared in
November, reached a peak of 10.57% of the catches in December,
and then continued irregularly in the collections through September.
None was found in October, although a thorough search was con-
ducted for them during that month.

Young M. cephalus were restricted almost entirely to the pools
(outer and inner) at both Cedar Key and Bayport (Table 3). The
open water specimens at Cedar Key were less than 20 mm in length,
and it was believed that these would have entered the marshes at the
next high tide, thus vacating the beach waters where they were taken.

Although Af, cephalus was taken at salinities ranging from 2.5 ppt
to 35.6 ppt at Cedar Key, most of the specimens there came from
waters with salinities between 15 and 30 ppt (Table 6). The Bay-
port specimens were restricted largely to waters with salinities be-
tween 5 and 20 ppt. These differences are understandable when it
is noted that the young fish are found most often in sheltered, mud-
bottomed pools which are devoid of submerged aquatics. Both the
inner and outer pools (low salinities) at Bayport, and the outer
pools (high salinity) at Cedar Key, met those habitat requirements.
Since during comparable collecting about ten times as many speci-
mens were taken at Cedar Key as at Bayport, it may be that the com-
bination of high salinity with other requirements accounts for the
greater density of the former population. However, the adult popu-
lation in the bay at Cedar Key is much denser than the one at Bay-
port, and that probably affects the number of young produced.

At Cedar Key M. cephalus exhibited no conspicuous association
(Table 8) with any of the species with which it was tested, but at
Bayport it appeared to be associated with Jordanella floridae (0.53rh
0.10), F. c. conjluentus (0.83±0.l6), and Cyprinodon variegatus
(0.88±0.25). These associations reflect its occurrence in the pools
of that area.

32. Mugil curema Cuvier and Valenciennes, White Mullet.

Fifty-three were taken in eight collections at Cedar Key and eight
were caught in six Bayport collections. The data on these are sum-
marized as follows:

No. 8 Kilby: Gulf Coastal Marsh Fishes 215

No. of

No. of

Size

Sal.

Temp.

Month

Colls.

Spec.

(mm)

(ppt)

(°C)

Cedar Key

Mar.

2

2

30,30

19.1-15.4

19.5 - 23

Apr,

2

3

29-31

21.2-20.8

24.5 - 25

May

3

40

24-65

20.6-25.2

32-33

June

1

8

24-36

24.7

35.4

Bay port

May

1

1

53

24.7

24.5

June

2

2

35,38

11.8-11.2

29.5-32

July

1

3

36-38

17.8

31

Aug-.

1

1

65

4.1

27

Sept.

1

1

49

4.8

29

All of the Cedar Key specimens were collected in mud-bottomed,
marsh pools which were very close to the open Gulf, thus accounting
for the rather high salinities observed. On the other hand, all of the
Bayport fish were taken from mud-bottomed, marsh pools which were
near the mainland side of the marshes and (except for two specimens)
at times when the salinities were relatively high for Bayport situations.

Their absence during the colder months is a reflection, perhaps, of
the greater abundance of the species in more tropical waters, although
it is recorded from Cape Cod on the Atlantic coast. Breeding is
accomplished at Cedar Key, as is attested by the presence of speci-
mens 24 mm in length. The season is lengthy, as fish of 30 mm or
less were found from March to June. The greatest number (40) of
these were caught in May.

33- Mugil trichodon Poey, Fan-tail Mullet.

Only 44 specimens of this interesting little mullet were caught.
All came from three collections made at Cedar Key. The data re-
corded are summarized as follows:

No. of Sizes Sal. Temp.

Month Spec. (mm) (ppt) (°C)

May 15 23-30 24.7 32.0

Jun. 4 47-58 24.7 34.5

Nov. 25 19-24 30.7 24.5

All were caught in mud-bottomed tide pools in the marshes ad-
jacent to the open Gulf where the salinities were relatively high.

Atherinidae: Silversides
34. Membras vagrans vagrans (Goode and Bean), Rough Silver-
sides.
One hundred and ten specimens were taken in five collections at

216 Tulane Studies in Zoology Vol. 2

Cedar Key. None was encountered at Bayport. The data are sum-
marized as follows:

No. of

Sizes

Sal.

Temp.

Month

Spec.

(mm)

(ppt)

(°C)

Apr.

9

53-62

21.3

30

May

3

30-45

20.8

25

Jun.

31

26-43

24.4

31

Jul.

58

40-59

27.2

31

Oct.

9

40-52

30.2

25

Each collection was made in the shallow waters along the Gulf
beach when the salinity was high; no specimens were encountered in
the marshes.

35. Menidia beryllina (Cope), Glassminnow.

Three thousand seven hundred sixty-eight glassminnows were col-
lected during this study. Sixty-nine of the Bayport collections in-
cluded 1,913 individuals; the remaining 1,855 were in fifty-two of
the catches at Cedar Key.

The fish were caught during all months of the year at both areas,
and sizes ranged from 10 to 99 mm. The Bayport fish seldom
reached 50 mm, whereas the ones from Cedar Key were about as
well represented in sizes larger than 50 mm as they were in the
smaller sizes. The largest fish at Cedar Key (99 mm) probably
constitutes a size record. Very young fish were taken in all seasons
at both localities.

An inspection of size charts prepared by months gave no definite
clue to the growth rate, although field experience suggested a very
rapid turnover in the population. Gunter (1945: 49) suggested a
short life cycle for Texas representatives.

At Cedar Key the fish was most abundant during March and
September, and at Bayport it remained a conspicuous element in the
population during all seasons. The relatively large numbers taken
in August were due, in part at least, to selective collecting, and thus
are not to be judged as comparable to the figures for the other months.

Field experience showed that M. beryllina was most abundant in
open waters at all sizes above about 35 mm. Below that size, par-
ticularly at Bayport, the fish was found frequently near, or more
often in, the open water just above the dense mats of submerged
aquatics. In a water-course or pool, however, where a central area
of open water existed, even if this open water was limited to a sur-
face stratum only a few inches in depth, M. beryllina could be ob-
served in schools which kept out of the vegetation and very near
the surface. Unfortunately it was not practicable to make the regular
collections in a manner that would demonstrate this obvious occu-
pancy of the open waters and, consequently, the data recorded in the
vegetated areas at Bayport are misleading on this point. The almost
total lack of aquatics at Cedar Key presented a different situation;

No. 8 Kilby: Gulf Coastal Marsh Fishes 217

there M. beryllina was most abundant in the larger water bodies and
tended to congregate in the centers of pools rather than in the
shallows near shore. In all cases, however, the fish remained near
the surface and, unless disturbed, assembled in the top twelve or
eighteen inches of water.

Most of the specimens from Cedar Key were taken in the outer
pools and the open waters along the Gulf shore ( Table 3 ) • At Bay-
port ( Table 3 ) the species was most abundant in the outer pools
and water courses, but occurred also in the open waters and the inner
pools. Of these latter situations it preferred those with submerged
aquatics, and few specimens were taken elsewhere.

The fish at Cedar Key were taken substantially from salinities be-
tween 15 and 30 ppt, but some were found where the salinity was
as low as 2.5 ppt and others where the salinity reached 35.6 ppt
(Table 6). More than half of the Bayport specimens (Table 7)
were taken at salinities between 5 and 10 ppt, and almost all of the
others came from within 5 ppt above or below those limits. Since
this fish was about as abundant at Cedar Key as at Bayport, it ap-
pears that the degree of salinity is not important to this species,
although at Cedar Key, where the salinity was higher than at Bayport,
the fish grew to larger size. Gunter (1945: 44) found the fish at
salinities of 0.0 ppt and on one occasion at 71.3 ppt, and in summary
stated "this species was more or less equally distributed in all sa-
linities."

At Cedar Key M. beryllina (Table 8) was associated positively with
E. argenteus (0.45±0.15), A. m. diaphana (0.90±0.20) and F. c.
carpio (0.82 ±0.27) which are all characterized as open water and/or
outer pool species. Negative association at Cedar Key is shown with
such protected water species as M. latipinna ( — 0.59±0.15), A* xenica
(— 0.48±0.13), C. variegatus (— 0.56±0.15), G. a. holbrooki
(—0.61 ±0.18) and F. c. conjluentus (—0.35 ±0.11).

Because of the non-selective character of the seining operations at
Bayport, as explained above, the associations at that area may not be
characteristic.

Centrarchidae: Sunfishes
36. Micropterus salmoides floridanus (LeSueur), Florida Large-
mouth Bass.

Eighteen specimens were taken by seining at Bayport, and hundreds
were observed in the channels where it is difficult to seine with
sufficient rapidity to catch so wary a fish. Hook and line techniques
give good results in the brackish waters if cut bait is used, and many
of the local fishermen depend on the fish as a mainstay of their
sport. The bass bite best during low, or nearly low, tides, and appear
to favor a late afternoon feeding period. Stomachs examined con-
tained blue crabs, needlefish, pinfish, snappers, mojarras, shrimp,
Gambusia and catfish.

No bass were taken from the marsh pools, although the fish was
common in the adjacent channels, and adults were observed on

218 Tulane Studies in Zoology Vol. 2

numerous occasions apparently feeding in the marshes at high tide.
Thus, by avoiding the pools, the fish also avoided high salinities, and
only one specimen was taken where the salinity reached 11.8 ppt.

Specimens as small as 16 mm were taken in the marshes in July
but since only 18 small fish were taken, no definite conclusions could
be reached relative to growth rates or the length of the breeding
season in the marshes. Marjorie Carr ( 1942 ) has discussed the
early development and growth of young bass from Florida waters.

37. Lepomis punctatus punctatus (Valenciennes), Eastern Spotted
Sunfish.

One hundred sixty-six specimens were present in 28 Bayport col-
lections, but none was found at Cedar Key.

Sizes ranged from 12 to 128 mm in the net catches, but larger
specimens were taken by local sport fishermen. Very young fish
were present in all seasons and nesting was observed during all months
of the year, although field observation indicated that the height of
the season was reached in mid-summer.

The adult fish were caught on hook and line as far out as the
open bay, but, in general, the fish tended to occur abundantly only
in the very mildly brackish waters near the mainland where the salt
was usually less than 5 ppt. The salinity range was from 0.0 to
11.8 ppt.

The species was found at Bayport almost exclusively in inner pools
and water-courses, and exhibited a decided preference for the latter
(Table 3). It was absent almost invariably from all situations where
the submerged aquatic vegetation was not abundant, and thus at
Bayport would be restricted to the two stated habitats.

38. Lepomis microlophus (Giinther), Redear Sunfish.
Thirty-eight specimens were obtained; these were found only in

the Bayport area at salinities of 12.3 ppt or less. Five of the nine
collections including this species were from the brackish channels
and pools of the marshes. The other four collections were from
mainland pools nearby, where occasional storm tides bring in salt
water which raises the pool salinities to at least 5.6 ppt. On the Gulf
side, the water yielding L. microlophus showed salinities up to 5.3
ppt, and in marsh pools up to 12.3 ppt, but nearly 90% of the in-
dividuals were caught where the salt was less than 5 ppt.

The smallest specimens taken were in June (12 mm) and in
November (19 mm). A graph of the distribution in time by sizes,
however, failed to indicate the growth rate. The breeding season is
obviously a long one, since individuals 30 mm or less were taken
from January through November.

Like L. p. punctatus, the species showed preference for the vege-
tated waters, particularly the inner pools, and occasionally it ventured
into heavily vegetated water-courses (Table 3).

39. Lepomis macrochirus purpurescens Cope, Southeastern Blue-

gill.

Thirty-eight bluegills were taken from mainland fresh waters at

No. 8 Kilby: Gulf Coastal Marsh Fishes 219

Bayport. However, one collection in June from a pool separated
from the marshes by the width of a road yielded nineteen specimens
at a salinity of 5.6 ppt. A month later the salinity of this pool was
down to 1.2 ppt and L. my purpurescens was still abundant. No
specimens were encountered in the brackish marshes, although the
species had ready access to those places in the Bayport area. Fish as
small as 13 mm were taken in July and 16 mm specimens were found
in June.

40. Chaenobryttus coronarius (Bartram), Warmouth.

Many specimens were collected from fresh water habitats near Bay-
port during the course of this study, but only three individuals were
found in brackish waters. It was never taken at salinities above 1.8
ppt, but it is a resident of Salt Creek at Bayport where the salinity is
usually around 1.5 ppt. Its ready access to the marshes, and its almost
total absence there, is a strong indication that the fish is practically
restricted to fresh water.

41. Elassoma evergladei Jordan, Everglades Pigmy Sunfish.

Specimens were locally abundant in the swamp waters of the main-
land at Bayport and pathways were available for its emigration to
the marshes nearby, but no individual was discovered in these marshes,
nor in any nearby mainland waters showing the least hint of tide
water influence. On this basis, the species appears to be entirely
restricted to fresh water.

LUTJANIDAE: Snappers

42. Lutjanus griseus (Linnaeus), Gray Snapper.

Three snappers were caught in as many collections at Cedar Key,
and seventeen specimens were taken in a total of eight catches at
Bayport. It is much more abundant at Bayport (and possibly at
Cedar Key) than the catches indicate. The bottoms of the channels
at Bayport literally are covered with schools of individuals during
the summer and winter months, and at low tide almost every nook
and crevice in the rocks and shells contains at least one individual,
even if the fish had to curl into a ball to squeeze itself into a niche.
Occasional specimens would bite a baited hook, but nearly all in-
dividuals were incredibly successful at avoiding nets of any type.
For instance, schools of snappers in less than five inches of water
would repeatedly escape a cast net thrown over scores of them.

Longley, in Longley and Hildebrand (1941: 115-116), made inter-
esting notes on this snapper at Tortugas, where it is "in many respects
the dominant fish in the local fauna." Herald and Strickland ( 1949:
105) recorded it as "fairly common" in the fresh waters of Homo-
sassa Springs, Florida.

The data for the fish caught in the present study are summarized
as follows:

220 Tulane Studies in Zoology Vol. 2

No. of

No. of

Size

Sal.

Temp.

Month

Colls.

Spec.

(mm)

(ppt)

(°C)

Cedar Key

Sep.

1

1

55

24.6

31.5

Oct.

1

1

49

4.5

24.0

Nov.

1

1

57

5.5

23.0

Bayport

Apr.

2

2

118-122

7.5, 17.1

25,26

Jun.

2

8

75 - 128

10.2, 11.8

25, 29.5

Jul.

1

1

113

7.7

33

Sep.

1

2

41-49

8.5

32

Nov.

2

4

41-115

— and 3.9

— and 22

The three Cedar Key specimens were taken from shallow bayous
and marsh pools near the mainland edge of the brackish marshes.
Three of the Bayport fish were from marsh pools, and the others
were caught in channels of the marshes. Salinity does not appear to
be a primary factor in their distribution at Bayport, as they ascend
both the Weekiwatchee River and Salt Creek where salinities are
usually less than 2.0 ppt.

Gerridae: Mojarras

43. Eucinostomus gula (Cuvier and Valenciennes), Common Mo-
jarra.

A total of 312 specimens was obtained. Of these, 211 were in 9
Cedar Key collections, and 101 came from 14 catches at Bayport.

Sizes ranged from 13 to 77 mm in the marshes, but larger speci-
mens were observed in the nets of fishermen operating offshore. At
both Bayport and Cedar Key the first specimens appeared in June
and by early December the species was no longer present. Gunter
(1945: 65) also found it absent from his early year catches, which
covered the deeper waters of the bay extending into the Gulf, as
well as the shallows. Since only immature fish were taken, it appears
almost certain that the adults spawn in the open Gulf and the young
fish make some use of the marshes. No clue to growth rate was dis-
cernible in the charts showing monthly distribution by size.

At Cedar Key nearly all specimens were taken from the shallow
waters along the Gulf beach, and at Bayport most (73%) came from
the same category of habitat while, except for a few specimens, the
remainder (26%) were in outer pools close by the Gulf (Table 3).

At Bayport (Table 7) nearly 90% of the specimens of E. gula
were from waters of 5 to 15 ppt salt, whereas at Cedar Key (Table
6) over 96% were taken where the salinity was between 20 and 30
ppt. In both instances, however, the fish occupied the waters further-
est from the mainland regardless of salinity, type of bottom, or
vegetation.

Concerning the habits of the fish during the early part of the

No. 8 Kilby: Gulf Coastal Marsh Fishes 221

calendar year, apparently little is known for the Gulf coast. A part
of the lack may be due to the confusion of this species with E.
argenteus, for several authors have mentioned experiencing that dif-
ficulty.

44. Eucinostomus argenteus Baird and Girard, Silver Mojarra.

Two thousand four hundred seventeen individuals were collected
and most of these were from Bayport (2,141 specimens in 62
catches). The remainder (276 in 26 collections) were from Cedar
Key.

Sizes ranged from 10 mm to 90 mm. At Bayport the species is a
year round resident of the marshes where young fish of 16 mm or
less were present during all months. At Cedar Key no specimens
were found in catches for the January through April period, and very
young fish were taken there only in May, October, December. Very
few of the individuals were larger than 50 mm in either area, and
consequently it appeared that only the young made any appreciable
use of the marshes or the inshore waters of the Gulf. The young
fish were most abundant at both Cedar Key and Bayport during the
period May through October, but at Cedar Key a drop in numbers
occurred in July, August and September.

E. argenteus, although found in all habitats, showed concentrations
at Cedar Key in water courses and open waters, whereas at Bayport
the concentrations were in the outer pools and the open waters (Table
3). The chart also points out the ecological difference between E.
gula and E. argenteus. At both localities E. gula reaches the marsh
area from the Gulf side but does not penetrate it to any appreciable
extent, while E. argenteus not only reaches the marsh from the Gulf,
but also conspicuously enters all its habitats except the inner pools,
where it occurs rarely.

As indicated by Tables 6 and 7, salinity preferences of the two
species are close, but at both Cedar Key and Bayport E. argenteus
was found distributed throughout the possible salinities and £, gula
was confined to a much narrower range. E. argenteus was recorded
by Herald and Strickland (1948: 106) from the fresh waters of
Homosassa Springs, Florida.

Table 8 indicates that E. argenteus is associated positively at Cedar
Key with such open water forms as M. beryllina (0.45 ±0.15) and
A. m. diaphana (0.62 ±0.12), and negatively with the protected water
species represented by Adinia xenica ( — 0.35 ±0.11), Cyprinodon
variegatus ( — 0.35 ±0.10), and F. c. confluentus ( — 0.39±0.13).
At Bayport its positive association with M. beryllina (0.49 ±0.10)
and Floridichthys c. carpio (0.74±0.13) indicates its preference for
the open waters, whereas its additional preference for the water-
courses is shown by the positive association with Lucania parva
(0.49±0.16). Negative associations at Bayport with Gambusia a.
holbrooki ( — 0.44 ±0.10), and Jordanella floridae ( — 0.72 ±0.26)
are reflections of the tendency of E. argenteus to avoid the inner pools.

222 Tulane Studies in Zoology Vol. 2

Carangidae: Pompanos, Jacks.

45. Chloroscombrus cbrysurus (Linnaeus), Bumper.

Six specimens were taken in a single collection from the shallows
of the Gulf beach at Cedar Key during October. Sizes ranged from
42 to 46 mm; salinity, 30.2 ppt; temperature, 25 °C. None was
caught at Bayport.

46. Oligoplites saurus (Bloch and Schneider), Leather- jacket.

Seventy-three specimens were taken in a total of twelve Cedar Key
collections, but none was caught at Bayport. A summary of the data
is presented:

No. of

No. of

Sizes

Sal.

Temp.

Month

Colls.

Spec.

(mm)

(ppt)

(°C)

Jun.

2

6

29-39

24.4, 24.7

31,34.5

Jul.

2

23

30-79

26.7, 27.2

31,33

Sep.

1

2

40,46

28.1

30

Oct.

7

42

25-68

27.7-30.2

21.5-34.5

All the fish were taken from shallows along the Gulf beach or
from similar shallows along the shores of the largest bays. No speci-
mens were found inside the marshes, although their presence in the
bays would probably permit them to wander into the larger channels
of the marshes on occasions.

The sizes of the smallest specimens from Cedar Key indicated an
extended breeding season of at least four months, but since the data
did not give a clue to growth rates, the time of the season is not
determinable.

47. Trachinotus falcatus (Linnaeus), Round Pompano.

Only a single specimen (43 mm) of this pompano was encountered.
It came from shallows of the bay east of Cedar Key where the bottom
was of soft mud liberally mixed with shells.

Haemulidae: Grunts

48. Orthopristis chrysopterus (Linnaeus), Pigfish.

Two specimens were taken. One individual (173 mm) was
caught in April (salinity 20.8 ppt, temperature 25 °C) in the shal-
lows along the Gulf beach at Cedar Key, and the other (143 mm)
was taken in November from a marsh channel at Bayport (salinity
and temperature not recorded).

Sparidae: Porgies

49. Lagodon rhomboides (Linnaeus), Pinfish.

Pinfish are very conspicuous in the offshore fauna of the Gulf at
both Cedar Key and Bayport wherever the shallows are vegetated.
At Cedar Key Reid (1954: 44) took over five thousand individuals
from the open waters of the Gulf. Inshore they also show a prefer-
ence for vegetated areas and at Bayport the channels contain both the

No. 8 Kilby: Gulf Coastal Marsh Fishes 223

plants and the pinfish. The catches at Bayport do not reflect the true
abundance of this form because pinfish are adept at escaping capture.
Seven Cedar Key catches netted 116 specimens, and 41 of the Bay-
port collections totalled 190 individuals. Sizes ranged from 13 to 146
mm, but more than half of the fish were less than 50 mm in length.

The smallest fish were taken between late December and late Febru-
ary, and the growth charts prepared (but not included here) indicated
that two size groups were present in January. The first was of small
fish between 13 and 20 mm. The second group was less defined
and included sizes from about 50 mm. By the end of the year the
small group had reached sizes of between 45 and 80 mm and were
approximately one year old. By July the larger group was repre-
sented by specimens of 100 mm or greater. The lack of large num-
bers of specimens does not permit precise analysis of these growth
trends, but they are in substantial agreement with the Texas findings
of Gunter (1945: 63) and the Cedar Key conclusions of Reid (loc.
cit. ) .

At Cedar Key 85% of the specimens were from outer pools and
nearly all of the remainder ( 14% ) were taken from the shallows
along the shore of the Gulf and of the bays. At Bayport the fish
seldom was taken from the inner pools but was found in all other
habitats (Table 3). Preference there was for the heavily vegetated
water-courses and this preference is reflected in the salinity figures.

At Bayport the species preferred low salinities and more than half
of the specimens were taken from waters of 5 to 10 ppt salt (Table 7).
The remainder were mostly from waters within 5 ppt above and below
that range. Substantially those salinities are the ones of the water-
courses. At Cedar Key the salinities at which L. rhomboides was
found are higher (15 to 30 ppt), and correspond to the salinities of
the outer waters there. Herald and Strickland (1949: 106) reported
the species in the fresh waters of Homosassa Springs, Florida.

In summary, as also concluded by Gunter (1945: 63), the fish does
not seek out a particular salinity, and the evidence of the present study
indicates that aquatic vegetation is of greater importance than the
degree of salinity in the choice of habitat.

50. Archosargus probatocephalus (Walbaum), Sheepshead.

Only one sheepshead, a specimen of 23 mm, was taken. It came
from a July collection (salinity 2.2 ppt, temperature 33°C) made in
the dense mats of Ruppia maritima in a marsh channel at Bayport.
This and other channels at Bayport contained hundreds of larger
sheepshead which were repeatedly observed at night by the aid of
lights. Most of these fish were 120 mm to 250 mm in length, but
fishermen occasionally take larger individuals.

At Bayport the sheepshead ascends the Weekiwatchee River for at
least a mile, where it was observed in waters of 1.2 ppt salinity in
July. Herald and Strickland (1948: 106) record the species as
common in the fresh waters of Homosassa Springs, Florida.

224 Tulane Studies in Zoology Vol. 2

Sciaenidae: Croakers and Drums

51. Bairdiella cbrysura (Lacepede), Silver Perch, Yellow Tail.

Eight fish were caught at Cedar Key and three at Bayport. The
data are listed:

No. of

Size

Sal.

Temp.

Month

Spec.

(mm)

(ppt)

(°C)

Cedar Key

Apr.

1

141

20.8

25.0

Jun.

4

66-76

26.1

34.5

Oct.

3

97-108

27.3

20.0

Bayport

Jul.

2

34 and 53

6.4

30

Jul.

1

47

5.6

31

The April specimen from Cedar Key was caught in the shallows
along the Gulf beach and all others were taken in a nearby marsh
pool. Both collections at Bayport were at dead low tide at the head
of the Channel in dense mats of Ruppia maritima.

Hildebrand and Cable (1930: 411) found B. cbrysura spawning
from late April to mid-July at Beaufort, N. C; Welsh and Breder
(1924: 171) found the height of the North Carolina season to be in
May; and Gunter (1945: 66) found fish in breeding condition in
April, May, and June in Texas. The sizes of specimens taken at Bay-
port and Cedar Key indicate a spring and early summer breeding
season. Welsh and Breder discussed the growth rates and stated that
the first breeding occurs at twenty-four months after hatching.

52. Cynoscion nebulosus (Cuvier), Spotted Sea Trout.

Six specimens were caught at Cedar Key and three at Bayport. The
data are as follows:

No. of

Size

Sal.

Temp.

Month

Spec.

(mm)

(ppt)

(°C)

Cedar Key

Jun.

1

35

24.7

34.5

Jul.

1

53

26.7

33

Oct.

3

30,38,45

25.9

31

Nov.

1

58

31.1

25

Bayport

Jul.

2

100, 109

12.3

26

Jul.

1

130

5.1

30

One of the Cedar Key specimens was from the beach -line waters of
the Gulf, another came from a nearby mud flat, and the others were
taken in the marsh tide pools near the Gulf.

Two of the Bayport fish were from the Gulf margin, and the other

No. 8

Kilby: Gulf Coastal Marsh Fishes

225

JAN.
72 sp.

FEB.
iea sp.

MaR
451 sp

aPR.
151 spi

A /I AA/UU

May

16 sp

JUN
20 sp'

iZL

LENGTH IN MM

'plus l sp 95mm and I cp 96 mrr

Figure 9. Leiostomus xanthurus, Cedar Key specimens. Monthly
length-frequency curves.

from a marsh pool.

Reid (1954: 53-54) found this species not uncommon at Cedar
Key and summarized Moody's (1950) study of the form in the Cedar
Key area. Moody's findings were in substantial agreement with those
of both Pearson (1929) and Gunter (1945: 76-78).

53. Cynoscion arenarius Ginsburg, Silver Sea Trout.

Four specimens were caught at Cedar Key and none at Bayport.
Two (19 mm and 66 mm) were taken in June (salinity 24.4 ppt,
temperature 31° C) from the Gulf beach, and two July specimens
(56 mm and 70 mm) were from a mud-bottomed, tide pool near the
Gulf beach (salinity 26.7 ppt, temperature 33 °C).

54. Leiostomus xanthurus Lacepede, Spot.

Eight hundred eighty-nine spots were collected at Cedar Key.
These were contained in twenty catches. Only 39 specimens were
caught at Bayport and these were in five collections.

Sizes at Cedar Key (fig. 9) ranged from 15 to 96 mm. Specimens
less than 20 mm were present during January, February, March and
May; these small individuals indicate a breeding season lasting about
the first three months of the year. Some spawning may occur for
the next two months, but it is not pronounced. The length-frequency
curves (fig. 9) indicate rather strongly that the January young reach
sizes between about 70 and 85 mm in June, when they were taken
for the last time during the year. Apparently the fish move out of
the marshes in June when they have reached lengths of about 80 mm.
These observations correspond with the conclusions of Reid (1954:
49-50) and of Gunter (1945: 70), who also summarized previously
reported accounts. According to Gunter, spawning occurs offshore

226 Tulane Studies in Zoology Vol. 2

during the winter, and the young spread into the inshore shallows.

The spot showed a rather high incidence of occurrence at salinities
between 15 and 20 ppt at Cedar Key (Table 6) and between 5 and
10 ppt at Bayport (fable 7). At both Cedar Key and Bayport,
however, the hsh was particularly partial to outer pools and was tound
rarely, or not at all, in the other habitats, with the exception of the
water-courses (Table 3). Thus it appears that habitat, rather than
salinity, is the deciding factor accounting for the young in the marshes.
On the basis of these findings and those of Gunter( op. cit.), it is
thought that the young which penetrated into the marshes at Cedar
Key and Bayport were part of a much larger population which con-
centrated in the nearby shallows of the bays and of the Gulf.

55. Pogonias cromis (Linnaeus), Black Drum.

Nine specimens were taken in six May and June collections at
Cedar Key. The seven May fish ranged from 17 to 43 mm, and the
two June ones were 49 and 83 mm respectively. All specimens came
from tide pools near the open Gulf. Salinities ranged from 20.6 to
26.1 ppt, and showed a mean of 24.3 ppt. Temperatures were all
high (32°C to 34.5 °C; mean 33.2 °C).

56. Sciaenops ocellata (Linnaeus), Redfish.

One hundred forty specimens were taken at Cedar Key in a total
of twenty-four collections. Oniy eleven redfish were caught at Bay-
port and these appeared in three November collections.

Sizes were from 12 to 146 mm, excluding specimens taken by hook
and line. The smallest fish appeared in the fall catches (September
and October at Cedar Key, and November at Bayport). The Cedar
Key fish grew rapidly, by late November their sizes centering around
46 mm, and by the following March most of these passing 100 mm
in length.

These conclusions are not at variance with the findings of Pearson
(1929) and Gunter (1945) who concluded that their Texas speci-
mens were a year old at about 300 mm.

At Cedar Key the young redfish were most abundant in open
waters, outer pools, and water-courses, in that order (Table 3). At
Bayport, the few fish caught came from open waters and water-
courses.

More than half of the specimens taken at Cedar Key were from
waters where the salinity ranged between 25 and 30 ppt, but the fish
was also caught at salinities which went as low as 0.8 ppt and as high
at 37.6 ppt (Table 6).

The adults are well known, marine littoral fish which feed both on
the bottom and pelagically, and my field experience supports the
findings of Gunter (1945) and Pearson (1929) in these respects.
Night observations in the clear waters at Bayport showed that the
adults were most abundant in the channels, and almost invariably
they were seen on, or very near, the bottom where they appeared to
be searching for food.

No. 8 Kilby: Gtdj Coastal Marsh Fishes 227

Ephippidae: Spadefishes

57. Chaetodipterus faber ( Broussonet ) , Spadefish.

One 20 mm specimen was taken in October at Cedar Key from
the inshore shallows of the bay. The salinity was 26.5 ppt and the
temperature 26 °C. The bottom was of mud and shell, soft and un-
vegetated.

Triglidae: Sea Robins

58. Prionotus tribulus (Cuvier), Bighead Sea Robin.

Seven of these sea robins were caught in a total of five Cedar Key
collections, but none was taken at Bayport. The catches were made
in February ( 1 specimen of 30 mm, salinity 27.2 ppt, temperature
24°C), October (2, 10 and 14 mm; 25.9 ppt; 31°C), October (2,
20 and 24 mm; 27.7 ppt; 21.5 °C), November (1, 13 mm; 23.8 ppt;
17.3°C), and November (1, 13 mm; 21.2 ppt; 22 °C). All of the
catches were made where the Gulf meets the marsh, and at times of
relatively low temperatures and high salinities.

Batrachoididae: Toadfishes
59- Opsanus beta (Goode and Bean), Toadfish.

Nineteen specimens were taken in three Cedar Key collections and
sixteen were included in nine Bayport catches. Sizes ranged from
19 to 175 mm at Bayport, and from 32 to 130 mm at Cedar Key.
The species is a secretive form and hides in any cover available. It
is abundant about oyster bars; 18 of the 19 Cedar Key specimens
were caught in the Gulf shallows by hand at low tide from tiny
basins of water under clumps of exposed oyster shells. The Bayport
specimens were obtained by seining in dense mats of Ruppia maritima.

The smallest Bayport individuals were caught in two July collec-
tions, and by November this age group may have grown to approxi-
mately 65 mm in length.

At Bayport, salinities ranged from 1.8 to 7.9 ppt (mean, 5.5 ppt)
and temperatures 23 °C to 32 °C (mean, 27.5 °C). The Cedar Key
salinities were 24.4, 27.7, and 37.6 ppt; the respective temperatures
were 31°C, 21.5°C, and 19 °C. These collections were in August,
October, and December, while the Bayport ones were made in Febru-
ary, June, July, September and November.

The species is undoubtedly more abundant than is indicated for
either area, but it is not a marsh form. It occasionally may be found
on the Gulf edge of the marshes, and may even penetrate into them
where cover in the form of shells or vegetation is available in the
channels.

Syngnathidae: Pipefishes and Seahorses

60. Syngnathus scovelli (Everman and Kendall), Scovell's Pipe-
fish.

Twenty specimens were taken: twelve from Cedar Key in two col-
lections, and eight from Bayport in eight collections. Sizes ranged

228 Tulane Studies in Zoology Vol. 2

from 33 to 78 mm at Bayport and from 35 to 78 mm at Cedar Key.
Specimens 35 mm or less were taken in November at Cedar Key and
in June at Bayport.

The Cedar Key individuals were all taken in November, where the
salinities were 31.1 and 23-8 ppt, and the respective temperatures
25°C and 17°C, from a shallow tide basin on the beach (11 speci-
mens) or in the adjacent shallows of the open Gulf (1 specimen).

All of the Bayport individuals were taken in seine hauls over
Ruppia covered bottoms in either the exposed Gulf beach situations,
or in waters directly connected with those places. Salinities ranged
from 4.8 to 11.8 ppt (mean = 7.6 ppt). Temperatures of 14.5°C
to 33.5 °C were recorded (mean = 29-2 °C). Specimens were taken
at Bayport in June ( 3 ) , July ( 3 ) , September ( 1 ) , and November
( 1 ) . No individuals were encountered in the interior of the marshes
or in the fresh waters of the nearby streams, although the fish has
been shown to be an abundant, year-round, continuously breeding,
resident of the inshore Gulf waters at Cedar Key (Reid 1954: 24-25).

61. Syngnathus floridae floridae (Jordan and Gilbert), Florida
Pipefish.

Twenty-one specimens were caught at Cedar Key. Nineteen of
these came from a shallow tide basin on the open sand beach of the
Gulf (salinity 31.1 ppt, temperature 25 °C) during November, and
the others were taken from the nearby shallow, shore-line waters of
the open Gulf in November (salinity 23.8 ppt, temperature 17°C)
and December (salinity and temperature not taken). Sizes ranged
from 58 to 102 mm in November, and the December specimen was
74 mm in length. The species is met with frequently in the grassy
areas of the open Gulf, as shown by Reid (1954: 26).

62. Syngnathus louisianae (Giinther), Louisiana Pipefish.

Two specimens were taken from a very small tide basin on the open
sand beach at Cedar Key in November. The salinity was 31.1 ppt,
the temperature 25 °C, and the sizes of the fish 89 mm and 164 mm.
The species is not a common resident in the nearby shallows of the
open Gulf where Reid (1954: 25) took only thirteen specimens
during his intensive survey.

63. Micrognathus crinigera (Bean and Dresel), Least Pipefish.

One diminutive specimen of 42 mm was found during November
in a tiny tide basin on the open sand beach of the Gulf at Cedar Key
where the water was only about two inches deep. At the same time,
thirty-two other pipefish, representing the previously mentioned three
species, were taken, but subsequent collecting in the same basin failed
to produce specimens of any of the forms. The pipefish were caught
in water which had a salinity of 31.1 ppt and a temperature of 25 °C.
Reid (1954: 27) found them regularly, but not abundantly, in his
collections in the offshore waters at Cedar Key.

No. 8 Kilby: Gulf Coastal Marsh Fishes 229

64. Hippocampus hudsonius punctulatus Guichenot, Seahorse.

One specimen of 26 mm (salinity 31.1 ppt, temperature 25 °C)
was taken in November and another of approximately 100 mm
(salinity and temperature not recorded) was caught in December.
Both came from vegetated bottoms in the shallow waters along the
Gulf margin at Cedar Key. No seahorses were taken at Bayport.

Gobiidae: Gobies

65. Bathygobius soporator catulus (Girard), Mapo.

Four Cedar Key collections included one specimen each and all
were taken from shallow, shore-line waters of the Gulf, or from a
nearby tide pool in the marsh. The data are: February (64 mm,
salinity 26.1 ppt, temperature 25 °C), November (30 mm, 23.8 ppt,
17 °C), December (26 mm, 37.6 ppt, 19 °C; 37 mm, 28.1 ppt, 23 °C).

66. Micro gobius gulosus (Girard), Large-mouthed Goby.

Two hundred fifty-six specimens of this fish were caught. The
majority (235) were taken in fourteen collections at Cedar Key;
twenty-one came from ten Bayport catches.

Sizes ranged from 12 to 62 mm. The smallest fish were taken at
Bayport in late September and at Cedar Key during May. This in-
dicates an extended breeding season, and the presence of individuals
of 25 mm or less during all seasons of the year indicates either a
continuous breeding season, or very slow growth of the young. Al-
though the data do not prove the point, the growth charts prepared
suggest a rather rapid development which produced 35 mm specimens
in approximately one year; these reached maximum sizes within the
next six to eight months.

The majority of the Cedar Key specimens were taken from outer
pools and the few Bayport individuals were scattered throughout the
possible habitats (Table 3). It was noted at Bayport, however, that
the fish were taken almost exclusively where the bottom was muddy
and the submerged aquatic vegetation was dense, and that at Cedar
Key the species was most abundant in the deepest part of a mud-
bottomed, outer pool.

At Bayport M. gulosus was never taken in water above salinity 10.5
ppt, and at Cedar Key it was never found below 15.4 ppt. This in-
dicates that salinity is not of prime consideration, although the greater
abundance of the species at Cedar Key may be a reflection of the
higher salinities there. On the other hand, its occurrence in the St.
Johns River as far south as Welaka, where the water is almost fresh,
strengthens the idea of wide salinity tolerance (Everman and Kendall
1900: 91).

67. Gobiosoma bosci Lacepede, Naked Goby.

Four collections at Cedar Key yielded twenty-nine specimens. The
collections were made in February (11 specimens; salinity 26.1 ppt;
temperature 25 °C), April (13, 20.8 ppt; 25 °C), May (3, 24.7 ppt;
32 CC), and July (2, 26.1 ppt; 32 °C). None was caught at Bayport,
and all of those from Cedar Key came from one marsh pool near the

230 Tulane Studies in Zoology Vol. 2

open Gulf. The smallest specimen ( 1 5 mm ) was caught in February
and the largest (46 mm) in May. The size distribution by months
does not give information regarding growth trends.

68. Gobionellus stigmaturus Goode and Bean, Spot-tailed Goby.

Two specimens were obtained at Cedar Key. One was caught in
October (salinity 22.9 ppt, temperature 27 °C) and the other in
November (salinity 21.2 ppt, temperature 22°C). Both were in Gulf
shore waters with mud bottoms. None was found at Bayport and,
since Reid (1954: 58) did not find the species at Cedar Key, the
form appears to be rare in the areas studied.

69- Gobiosoma robustum Ginsburg, Robust Goby.

Five specimens were collected at Cedar Key in two December
catches (salinity 37.6 ppt, temperature 19°C, and salinity 28.1 ppt,
temperature 23 °C). Two were taken at Bayport in one January
catch (salinity 7.0 ppt, temperature 18°C). All specimens were
from Gulf shore waters where the bottom was muddy. Sizes ranged
from 16 mm to 28 mm in December, and from 18 mm to 21 mm in
January.

Reid (1954: 59) found G. robustum abundant on the shallow
flats at Cedar Key, but much less so in deeper waters. These data
indicate that the species lives in the bays and approaches the shore
only in the coldest months of the year.

Blenniidae: Blennies

70. Chasmodes saburrae Jordan and Gilbert, Florida Blenny.

In July, 1947, at Bayport there was an extremely low tide which
lasted most of the night of the sixth. While walking on the exposed
bottom of the Gulf approximately half a mile from the normal shore-
line, I found one 27 mm specimen of C. saburrae in a half-cup of
water under a cluster of oyster shells. Many specimens of sizes rang-
ing up to 103 mm have been taken from the open Gulf at Cedar Key
by myself and by Reid (1954: 61-62), but at no time was the form
found in the marshes.

Bothidae: Left-eyed Flounders

71. Etropus crossotus atlanticus (Parr), Atlantic Fringed Flounder.
One 60 mm specimen was taken in August from the shallow water

near the Gulf beach at Cedar Key where the salinity was 24.4 ppt
and the temperature 31°C. Reid (1954: 64) regularly encountered
it in small numbers at Cedar Key, particularly in channels and on deep
flats of the open Gulf.

72. Paralichthys albigutta Jordan and Gilbert, Gulf Flounder.

A single, 65 mm specimen was caught at Cedar Key from a marsh
pool near the open Gulf in April (salinity 20.8 ppt, temperature
25 CC). Reid (1954: 65) found it not uncommon at Cedar Key
where he took many small individuals from the grassy flats near
shore and larger fish from the deeper waters worked by trawl.

No. 8 Kilby: Gulf Coastal Marsh Fishes 231

Achiridae: Broad Soles
73- Trinectes maculatus jasciatus (Lacepede), Southern Hog-
choker.

Five Cedar Key collections included nine specimens of T. m. jas-
ciatus {Achirus jasciatus Lacepede of many authors) and none was
taken at Bayport, although the form is locally abundant in Weeki-
watchee Spring at the head of the Weekiwatchee River. It has also
been recorded from a number of other fresh water situations in
Florida reaching from tributaries of the Apalachicola River in Liberty
County to Silver Springs in Marion County.

The Cedar Key specimens were taken in March (salinity 2.5 ppt,
temperature 15.5 °C), April (19.6 ppt, 29°C), and October (18.8,
19.4 and 21.8 ppt; 33.5 °C, 34.5 °C, and 28°C). Sizes of individuals
ranged from 10 to 30 mm in October to 24 to 36 mm in March and
April. All specimens were taken from shallows with mud bottoms,
either in marsh pools or imediately adjacent to the marshes in the
shore waters of the smaller bays. Reid (1954: 66) encountered the
species irregularly in trawl catches at Cedar Key.

Gunter (1945: 87-88) found the fish an inhabitant of both Gulf
and bay habitats, and he caught most of his specimens in bottom
trawls where the salinity was above 10.0 ppt. The species inhabits
fresh, brackish, and saline waters along the Gulf coast and thus may
be said to be euryhaline.

74. Achirus lineatus (Linnaeus), Striped Sole.

Twelve Cedar Key collections included a total of 46 specimens of
this sole and two were taken at Bayport in as many collections. The
Cedar Key fish were from three areas; fourteen were caught from a
small cove of the Gulf, 24 were in a marsh pool near the Gulf, and
eight were found in marsh pools close to high ground of the main-
land. All were taken from mud bottoms. Seventeen of the smallest
specimens were taken in late October from waters no more than an
inch deep in a Gulf cove where the mud was soft, deep, and mixed
with oyster shells. Three of these fish were 11 mm, and twelve were
less than 17 mm. Subsequent catches indicate a fairly regular growth
through the months until the following September, when the smallest
was 33 mm. Although the total number of specimens involved was
small, no exceptions to this growth rate were observed. It is possible,
then, that the breeding season is restricted to the early fall or late
summer and that growth is slow: from 11 mm in October to approxi-
mately 33 mm a year later. Reid (1954: 67), who took thirty-three
specimens at Cedar Key from a number of habitats, agrees with this
assumption of a slow growth rate.

One specimen was taken when the salinity was 2.5 ppt. The others
were all caught when it was from 19.6 to 30.2 ppt. Temperatures
varied from 15 °C to 31°C.

The two Bayport specimens were both taken from shallows at the
edge of the open Gulf in September (salinity 5.4 ppt, temperature

232 Tulane Studies in Zoology Vol. 2

30.5 °C) and October (salinity 2.0 ppt, temperature 25 °C), and were
possibly year old fish of 28 mm and 33 mm respectively.

Gunter (1945: 88) caught 125 of these soles from the bays and
Gulf. His data on temperature and salinity are in striking agree-
ment with those presented here.

Considering both the findings of Gunter {op. cit.) and those of
the present study, A. lineatus appears to be a slow-growing, bottom
species inhabiting the bays where salinities are moderate, but oc-
casionally wandering into bay and marsh areas of low salinity, and
also into the more saline Gulf waters.

Cynoglossidae: Tonguefishes

75. Symphurus plagiusa (Linnaeus), Tonguefish.

Two specimens, one in March (83 mm), one in December (30
mm), were taken at Cedar Key from waters of salinities 15.4 and
37.6 ppt respectively. Both came from muddy bottoms of waters
near the open Gulf: one from a shallow cove, the other from a marsh
pool. Water temperatures were relatively low, 23 °C in March, and
19°C in December.

Discussion

It will be recalled that compared with Bayport, the Cedar Key
marshes were described as being much wider, more saline, less inti-
mately interdigitating with freshwater habitats, almost lacking in sub-
merged aquatics, and having more variable water temperatures.

There can be little doubt that these and other differences are re-
flected in the fish populations. Figure 10 shows the percentage of
the total catch at Bayport represented by each of the most abundant
fifteen species and the comparable percentages for fourteen species
at Cedar Key. The figure makes it clear that, although there is con-
siderable similarity in population composition by species, there is a
marked difference in abundance of most of the species between the
two areas. A consideration of the entire species list, however, showed
other conspicuous differences. These are summarized as follows:

Cedar Key Bayport

Number of species collected 55 49

Number of species found in one

area but not in the other 26 20

Number of species known to be

permanent residents of fresh water

but also found in brackish water 5 19

Number of marine species also

found in brackish water 36 17

Twenty-nine species were found at both Cedar Key and Bayport.
Only five of these were restricted to the brackish marshes and the
immediately adjacent waters. These were: Fundulus similis, Fundu-
lus grandis grandis, Adinia xenica, Cyprinodon variegatus variegatus,

No. 8

Kilby: Gulf Coastal Marsh Fishes

233

POPULATION COMPOSITION

PERCENTAGES OF TOTAL CATCHES

Species

B oy port

Cedar Key

3.1 I

2.81 m

Mollienesio latipinna 22.91 ^f^±:'y^:::i>t:^mi

LuconiQ porvo

Eucinostomus orgenleus

Gombusio o holbrooki

Menidia beryl lino

Cypnnodon v.variegotus

Flor i di chthys corpio

Jo r d onel I o f Ion d oe

Fundulus c. eonf luentus

Fu ndulus g. gron d i s

Heterondno formoso

Ad i nio xen ico

Fundulus sim ilis

Mu gi I cephaius

Logodon rhomboides

Lepomis p. punctotus

Leiostomus xanthurus

An c h oo m. diaphana

3.02 m .0 0

X

totals 9 6.59%
J |

532 43

14.19

.14

3.23

2 36... .5lS 4.41

14.77

96.44%

_L

10 0 10

PER CENT

3 0

Figure 10. Population composition for Bayport and Cedar Key
fishes. Percentage composition is shown for the most abundant
species.

and Floridichthys carpio carpio*

These species are all members of the family Cyprinodontidae and
their numbers constituted 38.75% of all fish taken at Cedar Key and
14.8% of the total at Bayport. If other cyprinodontid fishes which
are adapted equally to brackish and to fresh waters were included,
the list would show the addition of Fundulus conjluentus confluentus,
Lucania parva, Gambusia affinis holbrooki, and Mollienesia latipinna.

The equivalent percentages of the total catch would be raised to
77.84 at Cedar Key and to 69.74 for Bayport, and the fact that the
two percentages are now subequal re-emphasizes that the primary
ecological difference between the Cedar Key and Bayport habitats is

234 Tulane S Unites in Zoology Vol. 2

one of salinity. A reflection of this same difference may be seen in
the composition of the lists of the species which occurred at one of
the two areas but not at the other. These species can be characterized
briefly as marine (M), fresh water (F), catadromous (C), and
fresh water and marine (FM). The lists are as follows:

Species found only at Bayport

Dasyatis sabina (M), Lepisosteus platyrbyncus (F), Notemigonus
chrysoleucas bosci (F), Notropis peter soni (F), Ameiurus natalis
erebennus (F), Anguilla bostoniensis (C), Fundulus cbrysotus (F),
Chriopeops goodei (F), Jordanella floridae (F), Heterandria formosa
(F), Stepbanolepis hispidus (M), Spbyraena barracuda (M), Microp-
terus salmoides floridanus (F), Lepomis punctatus punctatus (F),
Lepomis microlophus (F), Lepomis macrochims purpurescens (F),
Chaenobryttus coronarius (F), Elassoma evergladei (F), Archosargus
probatocephalus (M), and Cbasmodes saburrae (M). Totals: Fresh
water species, 14; Marine, 5; Catadromous, 1.

Species found only at Cedar Key

Elops saurus (M), Harengula pensacolae (M), Anchoa mit chilli
diaphana (M), Synodus foetens (M), Myropbis punctatus (M),
Spheroides nephelus (M), Mugil trichodon (M), Membras vagrans
vagrans (M), Cbloroscombrus chrysutus (M), Oligoplites saurus
(M), Trachinotus falcatus (M), Pogonias cromis (M), Chaetodip-
tenis faber (M), Prionotus tribulus (M), Syngnatbus floridae floridae
(M), Syngnatbus louisianae (M), Micrognathus crinigera (M),
Hippocampus hudsonius punctulatus (M), Batbygobius soporator
catulus (M), Gobiosoma bosci (M), Gobionellus stigmaturus (M),
Etropus crossotus atlanticus (M), Paralichtbys albigutta (M), Tri-
nectes maculatus fasciatus (FM), Symphurus plagiusa (M). Totals:
Fresh water species, 1 (?); Marine, 25.

Thus it is evident that those species peculiar to Cedar Key include,
at best, only one freshwater representative. On the other hand, more
than two-thirds of those found only at Bayport live principally in
fresh waters.

On the basis of these differences in population composition, there
appears little doubt that salinity is an important factor in shaping the
faunas of these two brackish marshes. Other possibly significant
factors have been suggested previously. Among these the availability
of species capable of adjusting to the total ecology of the marshes
must be considered.

Both of the areas studied border directly on the shallow, eulittoral
zone of the Gulf of Mexico and, although this zone has not been
studied at Bayport as intensively as it has at Cedar Key, it is obvious
to anyone familiar with both localities that the offshore fish faunas
are similar, if only their qualitative make up be considered. Equally
obvious, however, is the fact that the overall population density at
Cedar Key far exceeds that at Bayport. It seems very likely that this
greater population density would provide more opportunity for in-

No. 8 Kilby: Gulf Coastal Marsh Fishes 235

dividual marine fish to invade the brackish marshes there than in the
case of Bayport waters. Contrariwise, fresh waters at Bayport, oc-
cupied by a variety of fishes, intimately interdigitate with the brack-
ish marshes, while at Cedar Key such fresh waters are limited and
discontinuously distributed. Thus, to the factors of propinquity and
accessibility alone may be attributed part of the faunal divergence
(both quantitative and qualitative) of the two areas. Obviously,
however, an accurate appraisal of the contribution of these influences,
as compared to other ecological factors, would be difficult, if not
impossible, to make.

Regardless of the relative importance of the factors of salinity and
the availability of species, the fishes which occur in greatest abun-
dance in the brackish marshes at both Cedar Key and Bayport are
mostly species which belong to freshwater groups. Sixteen species
contributed more than 96% of the total catch of individuals at Bay-
port (fig. 10). Of these, only three (Eucinostomous argenteus,
Mugil cephalus and Lagodon rhomboides) can be considered euryha-
line representatives of marine groups of fishes. The other thirteen,
clearly belonging to freshwater families, contributed 87.72% of the
total catches of individuals. In like manner, fourteen species account
for more than 96% of the catches at Cedar Key. Of these, only five
{Eucinostomous argenteus, Mugil cephalus, Lagodon rhomboides,
Leiostomus xanthurus and Anchoa m. diaphana) are from marine
families, while the remaining nine, all of freshwater groups, con-
tributed 83.37% of the Cedar Key total. Expressing this another
way, approximately 85% of the specimens caught during this study
in the brackish marshes were of species belonging to freshwater, or
to freshwater-brackish water, groups.

It is also of interest that within this list of five marine species,
none is known to breed in the marshes. Contrarily, all of the non-
marine group of species not only breed in the brackish marshes but
maintain permanent populations there. On the basis of these data
rests the conclusion that the brackish waters of the intertidal marshes
at Cedar Key and at Bayport are populated principally by species
which have freshwater, rather than marine, affinities. Moreover,
although many marine species invade the brackish marshes tempor-
arily, it is the freshwater fishes which are contributing most conspicu-
ously to the permanent occupancy of the marshes by their establish-
ment of resident breeding populations.

In general, the above condition usually obtains in Florida waters,
but there are several striking exceptions of typically marine groups
with species which have established breeding populations in fresh
waters. Also an imposing list of typically marine species temporarily
go far inland in Florida fresh waters.

Summary
A study was made of the fish faunas of two coastal marshes on the
Gulf Coast of Florida to determine, primarily, the composition of

236

Tulane Studies in Zoology

Vol. 2

m

55

H

g^

WZ
b O
O co

« «

m &
«

iJ o

o

w % ■

33 Eh 5
H S ^

& « HH

"Is

. H
E* K J
Ph H ■<

fflzg

< < o fe

HjZ

o 2

S 3 «

Q -< (-«

W H «

O O H

H 2

fe a H

Eh fe Eh

5 o 3

S EH "<

o

«

X

w

a

a

BE

=

Q

— i

55

Ed

<!

O

en

55

H

O

_

►--

W

H

Ph

0

X

pq

(x

K

pq

H

7.

TT -SO

5 5
OOh

«-hPh

&■

.-

rf _o.

F]

o

a c

H

3

55

TO

(3 L

.-..

02

*•*

3

6"

3 O

CPh

c s

6*3.

0.

DO

I I

• Is 1

Tf o o r- g qo o
qqooioqq

r-» in cir* i-vi (—i ■ co

cog* n

I I

-OCO

I I ss M I

CO CO <M CO i— ' iO«5 *— « t— <

N --; (N ^ CO *# O C^ OS (

td«o«rHt>! I — ' od I ci -^>

CO-^COt-C^ ' »OW '<M

5 OflJ

0 GCO

* I O©

I I

d I n

aocoNeciocoto

p"* 'Hta^ V'» ' *- V* »W *.'V -■ ^— ' — u.; —-

- i - i . _ , r - — I G* -*j" c^i <m"

O i-t «D »-« C4 -^ ' OW5COtJ*

Hc<500ooTjiiN.ocooMi-iOoiTfcoa''»t,NO^n-*oo.Ort

G_o0 ^

I Ido'dd I |
1 Hoac '

•-it^r-J-- I I CO ai I

i>- *o

I I

11-111 • I I I I •

I la I 'I

I I

t^ CO

■<j;co

I O '-* G »C C! 1 -" G
1 -tj< «0 CO CO tO ' MN

I »

G OoOh-^h*

I ci I I ddr^dd

I I ' ! COMOiCCC

GiO
G CO

I I ©^ I

3 0ffl'1,0>0000|N'l,OOHOO©OOMONOOOiN«00©OS

^-»cO Cl'—^f IOI^-QOC01-^■^^CO^», i-*C^I

-— t>- -** O &> ->• »-t QO OS

"=; ^_2 2

^ a "5? -°

| ?■ S ^ -s §

■S g « a, g 3 -g
3i-2 a. ^Hi 0 fe

O *> — ' ° K 3>JP

d -; fej a; ^ oq 5?

oo 3

is

S O 3

all

C 5 3

o2«

3

■8.E

oo^ ^«S'

o R^ « dnj c

§5^

o 3 C
■S'-faJS-SXH - ~

~h=^:^;
S e. 3 3

Sfce S

3 3"

s "3 .2
-e"o 3

3 3^=

o to s =• j
a t, o cj v

3 -a. si— o a v. j~ -iz *-*.«-».

3 -O '**

"a I :
c S *j 2J

No. 8

Kilby: Gulf Coastal Marsh Fishes

237

s I I I I

I *o r-^ co in" | cc I -«£ r^l

2 I Ig I I

I I

= o
© | a

I I Isl

©so© ©
© <m © ©

00

o

O

wo
r- »n

©•o© © 1
o ■«■ © © '

1 «

1 CM

» 1

1 eo d

l» | | | | M | | |

s lis i

""t r"7 *"?

oi io ^j«

-- CO

I I

I I

■^iNHiNqQcqcic in

I I 1 i - I I I I I I I I I I I I I I I I g° I I I I I I I I I

Q0OO«00cfiC0Q0Nrtt^'-HOOO'-OHCCP*OCftOHOOc0Q0
TT} <-" ZO CO CO CM ^^ o -^ OS CO ^ *•+

ooooo^hocmo-hooono

Mh-OOJOCfll

in «o © aq © © r-^

«.. — — >. v_r — I"" ■ .

I oi •—" o ifi o o* cq I cicddoc

ot^oo^t^c;— *©©o©o2©^>

»OCOOOC'--inOt--OOOOOOOC^

i-!co©in©'©"©©©©"*0'*«< I ©' © I o

CO ©
OS ©
CO o

coin

I I I Ig is I I

o — •

-1 = 11

1 1 1 1 I I I I 1 1 I I 1 1 I 1 1

r- ^j« »—

I I

*-« CO

3 I I

I I

I I

I ©' — m' :
1 I © — •*■

I oo t-^

1 00— i

"■» (M tN

HH©OQO!0"J'0)0)0^'NOi«f^«Hn'f»OOiiflNOHH01(DM
^ CO ■* ,- ■ rH <m mtM «*

GO " -«*

ao —i

*- w -o ? ft.
S C * ^ » »

Q To O
•S 3 °^

si §1

§.§.§1

S 3

^ 3)

« " « ? 3-2.

Bg 1 § 22 3

"3 g

=s a
■g 5

K g E

-2 ^ S

Co go <J "

3 SO w;

C o o «5^'S
See? S-es

3 3 3 -e ;?? >^

1 1

o a «

. g^ k e

: c» o e e

1 S * S «j

l^Iii

? § £ 'c

.3 o"a -s

as 2 I

« J sw

g>-2 o-»

I 0, 05 o ft. O 0Q Co &3

-2 a a s g, 3 g

«S«v . 53 s

b fe=;^ » 2 « =•§

t«« « J 5.3 §-°

«a» jis.;.;8

ccceoggog

53 C» Cb CS K Q --

S 3
«c te £

S=3^3

1 <-, £. o

c -^

,<s"2

: ° §>s

3 3
9 ft

»■■§ ■§;

i »."3 2-fe ft
&Jft.t~^S5

238

Tulane Studies in Zoology

Vol. 2

-#

02

a

t-H

w

CM

to

m

03

H

a

03

w

si
<
O
w
O

fc
c

W

" CO

PQ a
< s

<
U

(*

j

w

o

*«— "ff* ^

° 6

eS

o

c

M
03

a
a>
o
<-.

Ah

o

(D

Q

>

o

bC

<

3

1-5

a

h

c

ci
its

02

CONOOiOMrHNCC'MCCOOCOMiOMNO'f^MlMONtDOtO

^■^•^idioiCT^cococooi

CO »— I H

OCNNLOOfflONOOffliOHOtOOWWMffl'fOOIMHOlO!
-*fCOtcHCMiOiOI^-cOeOOOOOr^CO-H-*i— ii— <t^.iO-*l»#-*tl''rlCMCMCMHH

O^ rfH'

iCCOHOi-iCOTtiClClO C5 ■— ii- 1 rf<

CM <M Ol 00° T}i tJH CO H ° ■■* ,
CM CM CM

CO lO OS t^ i-3 id id 'i— I

Tj<H©00COTj<t>.CDCO«D

^"kjh ' to ■* to ci id -I

CO i— l i—i i—i

lO^N tDOQ OC.

t> co os i ' cm' co i h •*

"<f CM

CO

co i— i co coocooiooo

tfOo i * O CM* ' •<*

CO i— < CO i— <

ittOLl c? C-1 ZD "P iO CQ

iOh© i ^H 1>J i-H i— I i— I i— I
LO ^H

co to
'co

co i—i co qo

CO o o

co

o

00

i O

00 H
CM H

CM i-H

o^r-

•* CM

o o

CM

rf — < Tti

Ocqi-O'^iocooooqt^cqt^ooco'-; co *q ^ ■*
oo'ooo ° id co ' id »d 'h ' ' I hh

■^ i-H

Ol o

CO

cqiqoooqocoo hco^cco h Tt< h rf hioohco
LOori'*m'ri'* i qn 'hh i I ■— 3 I

COr-H H

CfflilJHoo^MIC^O i> cm cm Lq cm CO
■O CO -* CO id CM h id i— i oo" I •* i "i-I '| 'i

CN l-H H tH

K0 CM CO 00 co CO """^ COSOo O CO t~- i— ii— ICM

id ■* [>l id jv! © to , (^i . . 'hi i

CM

"* CO CM CM co O W i— i Ol O o C5 CM CO

cc t*; oo co t^ cd c-i ■ ci <N co , >d , ' h

— h — i

N ■* r- H qq ^ O Tt< "* CO H

CO H CC OS ' uj H . ^

r-NN H

oo

<M

CM

'§'«

5i

to
8

ci

.<"
Q ~ .*>

^'S't' r

R -s: ^ 'o

-S -^"° *

S *> M R

R ^v ->^ ^

. "« •

c^ R

~» ci 2
S <S s
"e ^a -o

t»5

to

Cj ^ ^

la.

to ~

to

i-5

■S o o

"2 3'

O to
Cxo o

^^cS

k,KH

^

s s

v R

e s

. <o

» to

CO "»

C3-"3

*. "??

to

to

o

r :

R

S to o

, to to

5 to to ^

rf3

^^c^^^o^

■*<» bo to

a. I

^ s

* s

■o 13
R to

OK]

No. 8

Kilby: Gulf Coastal Marsh Fishes

239

ooooooooooooooooooooooooooo
ooooooooooo

so ■* co (M © oi co i> so so us io ■* Tf Tt> cc ci ci 01 01 i

CM

CM

co

r— t ^ O

CI

o

CO

o

O i— ' >— ' ©

<M

CO

CI

© ©

re

©

to

s

«o si

=o 13
S3 S

coco

S3 to

si

S3

■3 §-§

to

a
s

to

^3 §

** *5 ?*

« i 8

-^. to

5 »

SS8^2

S3 to

^cos-,

2 »-

S3 S3

S -S3 <;
^. *i <s

-o a v.

8.2 §

C5

ss

ss

to '

fe 2-2
O, cq a, 'o <o cj

u
to

o
j. 8

Co Co

feoS^

to .^ -^

~3

S "o

■^4 to

S3-S ^Ss* *?
■S2-S to-£j g

^co^gco^

S3
s^
tu

tij '^

«•&

to

8 5

K s

-©* S3

§ "S3

«-2

S3

la

5 "*■■*
-2 53

2"5

JSS S3

co
co"

CO

I-

00

r— I

©

o
©
10

GO

10
©

CO

SO

©
CM

CM

©

3!

©

X

0-1

o

CM

01

CM

00

co

o

240

Tulane Studies in Zoology

Vol. 2

W
.-J
PQ
<

co

w
o

H

Cm
CO

|H

pa

CO

to

«
O

Oh

<

PQ

fc
o

03

H
S3

H
<!

u

K
Eh

Z

o

o
o

H
P

«

M

H

co

*— "3 -^
fc-P ° «

° «

£co

CO

U

-^
o3

o

bfc

03

C

o>

ej
1-
oj
Ph

>
c

O

ft

o
CO

bC
3

C »-3

c

3
i-s

a
<

c

03
»-s

CO

'3

CP

ft

CO

<H'l,MO!tOr»'Hf)-<tOOi-i(DiCCSCOCBNiOiOOSOOCCCONTl<00
01!OO^iC51N'HO»MNN010CCNiO^,WN'-ii-ihi-ihiHihih

ci oi © © oo to co co cn ci ci r-5

©C^HCDCO-<*CO-'tlCiCOO-'t'H/icO©©©'--iT4HcO©©OOOOI>.©CMH
OOOl^C'l'H^cOTfoONtOOOCSONONin^MMMMflNN
CCHrtOO'*(OcOiOiO'*MN(M'-iiHrHrt

•<f ■* of of T* i-T

N O CO O !0 ■* iO COLO © O © CO

iO © ci -^ t^- 16 ' I i— h ' i hh'o

t— l CO *— t t— t

co

OXiOr- 1 CO CR CO iONhiOOhOON'H Mh t}h

■* © co ci ci co ci . i-h co ' co co ' rA ci ,

Ttl ,-H I-H

lOCOMCSNiOCOionHO'fl©

1(5 IC) •* Tj< ■* rt T)i ° 1(0 rH i-I C2

O) i— i i— i i— I

© H

OC0l--.O100C000cOC0C0i0r-.C0 CI t>. ,* ,-h

co © t^ © io oo ci " co i-i »-I i— 1 1— i i ' i-J

t^ iO CO O) OJ

tJh CD ci i Oh
•* CO

CO Ol

ci i-5

COO)

ci r-i

^©^©©COtHh^^lOtHI^-COOJ-h/lO-'^OOhOIOI

oi >o «-! t-J ic »o io ' co ic <-J 1-1 " tA

hC-Ih i-h

CO
CO«D(NON^C]cot>:1^WNcONcCr- i © -*T C) CO ©

rA ic co co oo © ■>* ■ ' ec th ' i-H ' ci h ' i-H

HHdH

OS *-H H 00 "* """J LO T)H '"J C"l I— I © 01

TjHCNCNOOOOCO^^eo ' . ' i-i
CO -h CN

C CD O CS N CO O <>• "* 00 iOhh

6<Ohh CM . u^ © ,-i i i-i -rp

LO i-H

O © h 't "3 oj ^) t>. O i— i CD

06'_I|,!CH1— '^COi-hMi-H. i , lOTl< .

NH I hH

CM'O-^CO^'^iOi— i^NfqcCMC^fflH

^ ri ^ •* rH ^ * i-i ■* i-S ' »h i-H -

^Ohcooco qcorto © © t~-

CO Jg rA 00 © . O^'cNCO ) hh '

>o

©CI

co

oo ■— i co

CI CM

© CM © i-H d

iO©(M

l> © uo

co'cm,_;

co ^ oi oo
ci I r-i

co

CO

iO

U3 '

HHIQ

s

^0 O

i'l

a
•=•3 |"S

"^ s <*- ^

^3 33

■J c

1-2 §

k a ».

cxc

5

r"^» Ci

ij ai.2

^. >. a

s a a

^ Vj ft

•o ^ S
S? O <o

^ s « a

O >o « «

^•S 2 §

to

a

to
to

*«■

*^ o

Si ^

K a
•Q"a

o 2

^ I

w, C3 ^ ^: "

^ k5 t*q to tq < co co ^

■? §§

i 2 =

tjj -

a t3 C3 .

** S: S; ^

^ § 5.2

K N tl J

ft ft^.^

_ Co ^ Co ■

(O "■•*

ft ft ^ a

to

a

2 ^

IS

^^

to Cxo

i§

O txo

a: o

No. 8

Kilby: Gulf Coastal Marsh Fishes

241

ooooooooooooooooooooo
ooooooooo

ooi^cc^©oooocococococo<n<M'

CO
CO

co

co co oo

CO

NON

CO

HCOO

CO

(M

co
o

ri

CO <N CO

ooo

"5

a

a

•S3
«? .-

a
s

^-

^>

*}
>~

O S
"3 to

■s^a

"3 » ^ 'K

Co o

2 e

S3 .^

III

to

a

8

"5

a

I

-Si

a o i

§ 2

■ft« to

** va

£5-g

sag
<! o S

© -ft, "2

£-2

©*-,£■>

6u0 <J>

a ©

*j s -~

© a -a,

r*\ © *»

8

<u
9

• ^ r M "

^ 2

00

oo

CO

o

CO

o
co
o

CO

CO
00

CO

CO
C5
CO

o
oo

CM

co

CO
CO

"3

X

— i

a
a>

a

242

Tulane Studies in Zoology

Vol. 2

w
w

X

CO
H

M

Q
H

o

U

►J
h4 fc,

CD

<!
CO

r-
W

o
EH

«

03

- CD

~^

CO fl

i— i

s~e

3

3 _!

g

s^

*3

X «3

• S

dec

%%

|S

£#

OS

©

CO

lO

co

©

CO

TfH

CD

co

C/J

1

</J

©

CO

u

©

!>>

©

—

CM

p

"3
CM

C3

m

©

CM

CO

O

43

CM

a

©

O

©

ft

r-I

ou

lO

o

t»

©

01

-1

Th

03

+j

o

CD

i-H

a

lH

CD

CI

HH

©

LO

©

"*

o

«-i

0)

40.

o

n

*-i CD

° a

•j

CO

^

CO

CD

O

CD

a

co

O^ffiClrtOOco^OOMOOOTOCOHH

r-I r-I CD rh CO CM oi oi NNHrtHoiddo'oi
CO CO CM CM CM Ol CM Ol Ol Ol CM CM CM CM CM CM CM r- i

O O CD r^ CD 01 CD CD CD CD CM CD CD ,— i CD •— i CD CD

t~- t— 1>- i> t- oo id t^ lo t-^ lo oI t~- cd t> • © id id

COCOCOOICOCMCOCOCOCOOICOCOOICOCMCOCO

III Il'll

lO O © Tt< iO tJH 00 L0 LO 00 Ol CM CO iO 00 Ol 00 00

t^ © CM- id CO id o ■"* CM* O r-i r-1 ©' -^ O '-' CD ©

i— I Ol Ol ■— I r-I

©00©.rO©©0©r-icO©iOr- i © © © © ©
OOOoO^OloOMcOO-hNOIOCOIOI

coooio'OOffl^n-ooccn'^oNO'i'w

i0©-*o©00©©'— !©-*'— ' o lO O lO Ol

co © r-I ' co ° oi

LO

cooor^ooocococDcooooOc-^otooi

^OC0OC1]OoNTt<HO'1<Oo|0OCSN

id cm oi ' co ' id ■■* r-I ' co ' co ' cm

t^r^oiiocDoO(M©iooiioocDro©r^t^'*t<

r- _ CD i-O 00 LO O CM © © iD CO CO '00 cD © ~] © "+1
CD CD r-I © r-i CO l>I CD ■* r-I ■* cd oi /-; co oi oo oi

i— I i-l 00 Tfl ■<* OI OI CO r-I OI -^f OI LO Jo CM Ol r-I OI

00©©©-rtH->*c»©T+,©©Ol"*cDt^t^t^l>

Nrt(OhOLOHOH^|nXHrtM«OrH

©' © ■* © oi r-I id cd cd co oi id r>- 06 r-I cd cd co

H W rH Tj< OI O LO CO O O CO CO r- 1 rt CO M ■* C»)

©©©CO©00t^©LOCOLO>rtl-*ro'tll^t^-*
©©©lOt>;CO©©^tl OI CD LO'— '©!>--* Oll^

' id © id r-i r-I ci ■<# "3 th i> r-i oi cd t^ oi

Ol r-I r-I CM 1— I OI CO OI LO i— I CO

©0©©01©C5©©OOOOlO©©01r-*©

qoqofiqooq^qoooootoio

t— I 1— I ' CD Ol' ^"h rH

©0©0-*©©©0©COci©©CO©i-i©

© ©_ © © Ol © LO OJ -rf ID t^ Ol © © "* O © ©

CM CO 1— 1 1— 1 O ©

CO

©©©©LO©cD01-*COt^©Oo©t^©LO

qqqoqqNCOrtooqoqoqotoM
r-I oi r-i oi t-~ oi © " 1*- ' © ©

rH Ol

©00r-llOLOCO©©t^©Ol©©©©©©t^
O'TflrHCOLOr-IOOt^OlCOCO-rtl'^r-HLOOOr^-^1
r^ CMO100 O'MDDrHXrHHNOO'^b-

r-l r-H HTf HO

r-I Tt<

•c: v>

8^1

a g

r^ r^* r^

"> >~ fe •

s tt S u

U !> ()

e o s;

^ r*

^S;^^^^

Q 5 "S

to -s» C3 "^C
>^ *^^ 10

i -I 's «

^ CJ K

2 .° s
.^ *-^ 'S -0

S ^-§ a

C3 O

§.3

^5 r^>

s

^

a

8 ci
.3 s

ci -s:
■a ^

S S

a s

r-

GO R

, o
o

■2 s -Ci,

CM

co
00
©"

CO

co
'o

OJ

ft

CO

O

cu
+->

c

C3
w

C

'S

&

73

CO

co

r2

CJ

C3
CO

«(H

o

co

03

rQ

cu

rC

C
o

T3
CD

rS

o

O

CO

cu

03

CO

CU .
-G CO
•+J co

a

<Mr^

o «

CU CO

5rC

.2 os

•■"■ cu

cu s
rD""

• SrC

CO

03 to
CO CU

be
2 «

G
CU
CJ

rH
CU

No. 8

Kilby: Gulf Coastal Marsh Fishes

243

m
H

n

CO

«

O
fin
!*
<l

w

Q
H

U
H

►J

W h

-t o

^9 CO

■<

C/2

PQ

O

H
P

«

«
CO

a *

a s

o aj

OS

oi

eo

i

CO
C2

-*'

co

■

o

CO

O o

03
CM

>>

m

a
cu

.5 C5

C5
CM
i
US
CM

05

rH
CN
i
O
CM

a

o

CO

u,

i

r/)

"3

i— I

o

m

OS

0>
03

i— 1

o

a>

o

tH

OS

iO

o a

CO

02

C000'*0>C0INOtD'*T)(O'H>OMe0OCSOM

o

rHThi^'c5oJoJoJodoscx5o6cx3i>t^t^i>«6«D'*

CM r-l i-H H

ci

T-irtWHi0NirjHiHi0N00MNCC00N(MO

T— 1

CN OI OJ OI >— ihhNNi- i(NH (Ni-iiHINHrH

lllllllllll 1

CO
Ol

OrtHOHOOOOOOSHOMHTfOiHOOO

1

o

driddHOciddMHoiiodrtNodd

o

OOOOOOa.Oi'OfflOOiONONOlii
OOOHOOWOOOiOOOOMOiOHO

CD

' 1— 1 ' CO ' I— " ^JH

ooooooooooooooooooo
ooooooooooooooooooo

o
o

ooooooooooooooooooo
ooooooooooooooooooo

o
o

OcO^-^OOOOOiOOOOOOOOOOO

Mcooowoqqi;ioqqqqqooooo

o
o

■-5 oi ""f US CO

Tfl

CO

NOOMWOOOOOMOMOONOO'I'OO

MM<OiOOOOOWONOOnoOH00

us

^'n«ocj ' ' oi co
co <-i

■*

OlcSMHTfiOMOOOlMNOOOMOOMH
OOCOOWtOOlHtsO'OHOL'JOOHNN

o

oo

dddd^NcinM ' i> co 'oi ' cd oi

00

!0O)NNCJ00MOO)NC0NOi0O!10nON
NOCON'OMtDCONMNCOOTtiOOOlCNOlCO

CO
00

r^ooi'Ldcdcdt^cdt-^cdodoi '©Nhl^'do

CN CN <— i i— i CO OI OI CO >-t OI i— i ^h ,-h ^h ,-h ,_(

T— 1

ON^O^OOoOM'J^iOONcONCON^
OhiOMMOl'3^N'-<NOO'Oc4C»MC3)M

' oi us' co >— i o ci oi ©' <>• oi co o us' us oo' r-i uj oi

CNCNiHCO'C'^COM'OsOffiOiOiOiOiO'^'H

MC)<*HiO00i-iON(NiOtOO'*OlCOe0OO
KJ^O-^NHfflooOiOtO^OCioOOONO

' CO t-J CO O CO 00
OI i-H CN CO CO CO

tOiOHiONCOOONCOOrn
CN CO OI i— I CN CO l-HH

^MO)ffli<*HT|lfflrHiHMHnOC5NnO

•*OCiX<000'*(NCN'J<(Ohho)C<:c»ON
CO CN US 00 CO OI i— 1 Tt< O r-iCO C5_r-( ^h US "*

■>*" i-Tof of i-T tjT

a

:a

8
8

a

VJ

a
8

■S,

3

to

a

3 s>"*

v> a s <o

g a g> x a

"a ? a "a

'So'^a-S a "a
a 5: o -« i

a 0

2 s "^
a ^ ^ ^i

^ <3 . 53

a S ^°

111

t-j a

g s a

■*^ ^3 -»i "^ ■*-» >^ >Cs

■§£•§

^ 8

u y >. <3
IS a.£-

a a

a

a <»

»«• . -?
a &4.a

a a «
a'a v.

S a ^

02

x

cu

c

.5

'o

w

CO

C3
CO

CU

co

C3

J3

O

-a
cu
*>

,5

CJ

5J

CO

CD

a

■j.

a)

-c

t«

4J

C/)

cS

«H

.— .

O

CJ

r

>»

c3

+J

cu

c

s

»— i

.-5

cu

ro

-c

+j

^3

u

7.

S3

a>

X

cu

C

'O

• rH

a

-c

B5

>.'«S

-i-»

c

8f-l

o

C3

CO

DO

i>

bun

4»

-i

nfl

-»^

H

c

CU

*

u

^4

CU

A

244

Tulane Studies in Zoology

Vol. 2

00

W

PQ
<

M

P3
«4

O
H

o

s

o

K

OS

H

to

a
w
eh
u

H

H
■ CO

o

x

-
O

O

o

t— I

H
<!

O

O

CO

02

o

l-l

3

H
Ch
cc
K
W
H
to

is
I 5 7

.e <m rt
•2 £ ±1 +i

^ ^ OO N

.S -^

"&

3 M M rf

^ CI CI y- 1

■S +[ +[ +1

K -^ © CM

3 © CO .— i

5 -S

"3

+i +i

cc os

CM CM

l"

C CM

CM ^-

*o CO

3 =»

§ "3

» §

1

I S3

•g +i

CJ CO

C2 O

> IS) '

^ J^ M ^

+j +? +i +j
O CO gj oo

g S 3 S £;

-i +f +f +f +; +j

S 2 o « g B

q * ic « ^, h

S

c-.

CI

3

CM

CI

+1

CO

+1

•
1C

+1

CM

cC

I I

'C "§ *"! ""! • " 1 *"! ^ *i

s ■? +' t1 ,m ±i -ti +i +j +1

ss. KiOt^iOCMcOOOCiOO
B. ^ • " I ••,•••

C

©

©ioococor~o.-iioco
'E +! +f +1 +1 +1 +1 +f +\ +i

aionccMcOHt,nr.

o i" r ■ ■ f * • • r

050DhMh«0»0^o
(N t- CO -^ N « H im w ^

+r +f +f +f +r +r +\ +\ +i +i

*H CO NHH^^
nr\ rr—i .^i en fxi *-^

it j ^ n sn "
oo to io <o CO

I

to CO © i— < o

s +f +i +i +i

•<S i-h O O i— i

1

+1
_ ^

t~- CM

5, '

a

+i +i +i

© co

s

.-t Tf -^

CO *-> f- O -^

±1 ±1 +1 +1 +1
M O ^ CO M
O *0 CM CO <M

'g.S

e -s

Q

c^

6B

OB

fc.

o

■^

>C.

»

Ca>

CI

£
;
^

Ct.

u

a
•<*

00

O

a

tt

°9

ac

1 k

■o ts -S "o

_ k e >- r

3 3 3^3"°

M

C5

w

PQ

«
O

pq
s

o

OS

H
»-t
O

t/J

O
W
H
O

►J
H
CO

o

H

t-
fe
H
O

o

£

Eh

<

o

CO

CO

"5

•S •<

cs

r—

H

+ 1

iC

=

^

W- 3 (M CO ^H

c5 3 ^1 *n "*

*» *« +1 +1 +'

. K CO »0 G>

<J 3 CO i-t ^JH

S,^ ' ''

v +i +i +i

fe N N CO

^ * n -fli

fc. " l'

•S 3
"e "3

Ci r-l IO CO

.. * (D T)< M o
3 CM CM CO 3D w

^ O

3 '~i

-O +1

S CO

3 CO

CM CO t^ CO *-

+1 +1 +f +1 +|

CO OO ~* CO |^

h m fi w o

cc
S

e
'8

=3 3

5 b

c •=>
& .g o
'n "a '~'-
e 'S +1

+1 +1 +1 +1 +1 +1 +1

m *f •* CO ^CM CO CM
1—t CO t~- CO t— i m x^-

+1 +1

*T5 CM
CM —

+1 +1 +1 +1

t>- O0 CO CO
"* CO CO IO

I- -

.3 C +i

i— I N M O N K io CC
H. ^ 1 ^ w w N

+i ±! +i ±i +i +•' +i +i

mOthmOIOCOOT
HiOMCDNCOJDCD

-<=

o

c^

«

s

~

CJ

C:

o*

C

•<**

CO

c^

-r

h-

c

K

O

•— 1

i-*

CO

-—

CM

CM

CO

S,

-^

+1

+1

+1

+1

+1

-"-1

+1

+1

+:

+1

cr

r-

■^t*

is

OO

OJ

to

3

c

^t*

iO

CM

"*«

*#

«

to

KhhwhHi-jonWWCO

^* +1 +1 +1 +1 +1 +1 +1 +i +1 +l +r

•;5M*'MC'i-iCMOiO-*b-Cl'-'

ftCO«rHlN(NNOHtC(0

^

O Ci CM CO ^

'g O *H 1-H ^

e +■ +! +i +i

y X OC N <D

=J o o *o CO

±\ +1 +1 +1 +1 +1 +1 +1

OJM^Clh.tNN'-'
T^rJHCO^-OCOCO'-'

00»0O»0OOIO00cOtJ<CM— 140

OOlOONOtCiO^COCOCNi-iH

^ « ^ ft.

fe S e -« g»3

iS "° K '2 "£ "^3 "O

"S I I s § § §

fc. ft, ^ ^ 4

o
a.

c?

m

No. 8 Kilby: Gulf Coastal Marsh Fishes 245

those faunas and to discover, if possible, the more obvious ecological
factors which were shaping them.

The two areas were different ecologically. The one at Bayport had
less saline waters and more stable temperatures than the one at Cedar
Key in addition to other differences which were described.

Seventy-five species were recorded during the study and species data
are presented in an annotated list. This treatment is supplemented
by tables, graphs, and other compilations designed to present certain
of the data which could be summarized.

Five species, Fundulus similis, Fundulus grandis grandis, Adinia xeni-
ca, Cyprinodon variegatus variegatus, and Floridichthys carpio carpio
were found only in the brackish waters of the marshes or immediately
adjacent brackish waters. These are considered to be brackish zone
fishes and are all members of the family Cyprinodontidae. They
comprized 14.89% of the total catch at Bayport and 38.75% of that
from Cedar Key. If four other cyprinodontids which reside in both
brackish and fresh waters are added to the list, the respective percent-
ages of the total catches are raised to 69.74 for Bayport and 77.84
for Cedar Key. Thus the cyprinodontids were the dominant family
of fishes in the brackish waters studied.

The availability of species capable of adjusting to the total ecology
of the brackish marshes was considered. At Cedar Key, where main-
land fresh water is at a minimum, dense populations of eulittoral
marine species were available to the marshes. At Bayport the marine
fishes, although substantially equal to those at Cedar Key in species
present, occurred in much fewer numbers of individuals, but the pro-
pinquity of fresh waters permitted easy access to the marshes for
many freshwater fishes.

Despite the relative importance of the factors of salinity and the
availability of species, it was noted that the fishes which occur in
greatest abundance in the brackish marshes at both Cedar Key and
Bayport belong to freshwater groups. In other words, approximately
85% of the specimens caught in the brackish marshes during this
study are of species representing either freshwater or freshwater-
brackish water groups of fishes. Furthermore, of the sixteen species
at Bayport contributing 96% of the total catch only three represented
marine species, and of the fourteen species at Cedar Key contributing
96% of the total catch only five were marine. None of these five
marine species (Eucinostomus argenteus, Mugil cephalus, Lagodon
rhomboides, Leiostomous xanthurus and Anchoa mitchilli diaphana)
is known to breed in the brackish marshes, whereas all of the fresh-
water affinates in reference not only breed in the marshes but also
maintain permanent populations there. Thus it is concluded that the
intertidal marshes at Cedar Key and at Bayport are populated prin-
cipally by fishes with freshwater, rather than marine affinities.

Acknowledgements
The present paper is a modification of a manuscript submitted in

246 Tulane Studies in Zoology Vol. 2

1952 as a doctorate dissertation to the University of Florida. The
original accumulation and presentation of data was facilitated and
made possible by the generous and willing assistance of the staff of
the University's Department of Biology and by the writer's Graduate
Committee in particular. I recall with pleasure and gratitude the
many favors received from various individuals and wish to thank in
particular Doctors Harley B. Sherman, W. C. Allee, Arnold Grobman,
Archie Carr, George Reid, J. C. Dickinson, Jr., E. Lowe Pierce, C. J.
Goin and Miss Esther Coogle.

My sincerest thanks and appreciation also go to Dr. Reeve M.
Bailey, who temporarily abandoned his own researches in order to
review selected samples of the Bayport and Cedar Key fishes and to
furnish identifications on the majority of the species involved in this
study; to Mr. Isaac Ginsburg, who kindly identified several of the
gobies and the pompano; to Dr. Carl Hubbs, who took an active in-
terest in the mullets collected; and to my wife, Evelyn, who not only
provided the expected encouragement but also attacked with determi-
nation and skill the laborious chores of seining, typing, and verifying
the statistical calculations.

References Cited

Bangham, Ralph V. 1942 (1940). Parasites of fresh-water fish of
southern Florida. Proc. Fla. Acad. Sci., 5: 289-307.

Campbell, R. B. 1940. Outline of the geological history of Florida.
Proc. Fla. Acad. Sci., 4: 87-105.

Carr, Marjorie Harris 1942. The breeding habits, embryology, and
larval' development of the large-mouth black bass in Florida.
Proc. Neiv England Zool. Club, 20: 43-77.

Cole, Lamont C. 1949. The measurement of interspecific association.
Ecology, 30(4) : 411-424.

Cooke, C. Wythe 1945. Geology of Florida, Bull. Fla. Geol. Surv.,
29: i-ix, 1-339.

Everman, Barton Warren and William Converse Kendall 1899
(1900). Check-list of the fishes of Florida. Rept. U. S. Comm.
Fish and Fisheries, 1899: 35-103.

Gunter, Gordon 1938. Notes on the invasion of freshwater by fishes
of the Gulf of Mexico, with special reference to the Mississippi-
Atchafalaya River System. Copeia, 1938 (2) : 69-72.

1942. A list of the fishes of the mainland of North

and Middle America recorded from both freshwater and sea
water. Amer. Midi. Nat., 28 (2) : 305-326.

1945. Studies on marine fishes of Texas. Publ.

Inst. Marine Science, 1 (1) : 1-190.
Herald, Earl S. and Roy R. Strickland 1948 (1949). Annotated

list of the fishes of Homosassa Springs, Florida. Quart. Jour.

Fla. Acad. Sci., 11 (4) : 99-109.
Hildebrand, Samuel F. 1917. Notes on the life history of the min-
nows Gambusia affinis and Cyprinodon variegatus. Rept. U. S.

Comm. of Fish, 1917, Appendix VI, 1-15. (Doc. No. 857).
Hildebrand, Samuel F. and Louella E. Cable 1930. Development

and life history of fourteen teleostean fishes at Beaufort, N. C.

Bull. U. S. Bur. Fish., 46 : 383-499.

No. 8 Kilby: Gulf Coastal Marsh Fishes 247

Hildebrand, Samuel F. and William C. Schroeder 1928. The fishes
of Chesapeake Bay. Bull. U. S. Bur. Fish., 43 (Part 1) : 1-366.

Hubbs, Carl L. and E. Ross Allen 1943. Fishes of Silver Springs,
Florida. Proc. Fla. Acad. Scl, 6: 110-130.

Kilby, John- D. 1948 (1949). A preliminary report on young striped
mullet (Mugil ceplalus Linnaeus) in two Gulf coastal areas of
Florida. Quart. Jour. Fla. Acad. Scl, 11 (1) : 7-23.

1950. The fishes of two Gulf coastal marsh areas

of Florida. (Unpublished Manuscript, University of Florida
Library)

Longley, William H. and Samuel F. Hildebrand 1941. Systemic
catalog of the fishes of Tortugas, Florida. Pap. Tortugas Lab.,
34: i-xiii, 1-331. [Carnegie Inst. Wash. Publ. No. 535]

Pearson, John C. 1928 (1929). Natural history and conservation of
the redfish and other commercial sciaenids on the Texas coast.
Bull, U. S. Bur. Fish, 44: 129-214.

Reid, George K., Jr. 1954. An ecological study of the Gulf of Mexico
fishes in the vicinity of Cedar Key, Florida. Bull. Marine Sci.
Gulf and Carib., 4: 1-94.

Schureman, Paul (Prepared by) 1941. Manual of tide observations.
U. S. Dept. Commerce, Coast and Geod. Surv., Special Publ. No.
196, (Rev. 1941), iv, 1-92.

Simpson, George G. and Ann Roe 1939. Quantitative Zoology.
McGraw-Hill, New York, pp. 1-414.

Smith, Homer W. 1943. Lectures on the Kidney, Univ. Kan., Law-
rence, pp. 1-134.

Welsh, William W. and Charles M. Breder, Jr. 1924. Contribu-
tions to the life histories of the Sciaenidae of the eastern United
States Coast. Bull. U. S. Bur. Fish, 39: 141-201.

>- /vn

/v^«w(yr/«4^sj

TULANE STUDIES IN ZOOLOGY

VOLUME 2

INDEX TO AUTHORS AND SCIENTIFIC NAMES
(New species and genera in boldface)

Abacion, 63

Achiridae, 231

Aehirus lineatus, 231-232

Acris, 33

Adinia xenica, 199-201

Agkistrodon piscivorous, 147

Ameiuridae, 194

Ameiurus natalis erebennus, 194

Amphibia, 13-46

Anchoa mitchilli diaphana, 192-193

Anderson, Paul K. (article), 13-46

Anguilla bostoniensis, 195-196

Anguillidae, 195-196

Apheloria reducta, 67

Archosargus probatocephalus, 223

Atherinidae, 215-217

Auffenbere, Walter (article), 87-

155
Bairdiella chrysura, 224
Batliygobins soporator catulus, 229
Batrachoididae, 227
Belonidae, 194-195
Blenniidae. 230
Bothidae, 230
Bufo, 36, 37
Calanoida, 49-60
Carangidae, 222

Causey, Nell B. (article), 61-68
Centrarchidae, 217-219
Chaenobryttus coronarius, 219
Chaetodipterus faber, 227
Chasmodes saburrae, 230
Chloroscombrus chrysurus, 222
Chriopeops goodei, 202-203
Cleidogona

arkansana, sp. nov., 66-67
aspera, 65, 66
fustis, 66
mississipp ia na, 66
sublettei, sp. nov., 66
unit a, 66
Clupeidae, 192
Coluber

acuminatus, 130, 145
constrictor, 87-155

anthicus, 90, 91, 92, 112, 114,
117, 119, 120, 137, 143, 147
constrictor, 90, 91, 94-97, 110,
112, 114,115, 117, 121, 122,
123, 130, 132, 134, 135,
136, 137, 139, 141, 146
flaviventris, 90, 91, 92, 110,
112, 114, 115, 117, 119,
120, 121, 122, 123, 126,
130, 132, 134, 137, 139,
140, 143, 146
foxi, 92, 114, 115

helvigularis, subsp. nov., 90,
91, 93, 105-108, 110, 112,
115, 117, 119, 139, 141,

147

moimion, 90, 91, 92, 94, 112,
115, 117, 120, 130, 135,
139, 143, 146

ortenburgei, 146

paludicohis, 90, 91, 92, 102-105,
110, 112, 113, 114, 115,
117, 118, 122, 123, 125,
126, 130, 132, 134, 136,
137, 139, 143, 147

priapus, 90, 91, 92, 93, 97-101,
109, 110, 111, 112, 113,

114, 115, 117, 120, 121,
122, 123, 125, 126, 130,
132, 135, 136, 137, 138,
139, 140, 141, 147

stejnegerianus, 90, 91, 94, 112,

115, 117, 120, 123, 126,
130, 139, 140, 143

oaxaca, 145

spinalis, 145
Conchostraca, 1-10
Copepoda, 49-60

Crotalus horridus atricaudatus, 147
Crustacea, 1-10, 49-60
Cynoglossidae, 232
Cynoscion

arenarius, 225

nebidosus, 224-225
Cyprinidae, 193-194
Cyprinodon variegatus variegatus,

203-205
Cyprinodontidae, 196-206
Dasyatidae, 191
Dasyatis sabina, 191
Delophon

carolinum, 63

georgianum, 63

serrulatum, sp. nov., 63, 64
Diadophis punctatns, 92
Diaptomus

angustilobus, 54

ashlandi, 54

clavipes, 51

connexus, 56, 57, 58

cuauhtemoci, 59

garciai, 58

jndayi, 55, 56, 57, 58

louisianensis, 51

moorei, sp. nov., 51-54, 55, 56,
57, 58, 59

yiovamexicanus, 55, 56, 57, 58, 59

nudus, 54, 55, 56, 57, 59

pallidus, 51

INDEX TO AUTHORS AND SCIENTIFIC NAMES— Continued

probilofensis, 54

sicilis, 54

siciloides, 56, 57, 58

signicauda, 51, 55, 56, 57, 58, 59

spinicornis, 56, 57

trybomi, 54, 55, 58

tyrelli, 54

virginiensis, 51

washingtoniensis, 58
Diplopoda, 61-68
Dorypetalidae, 63
Elassoma evevgladei, 219
Elopidae, 191
Elops saurus, 191-192
Engraulidae, 192-193
Ephippidae, 227

Etropus crossotas atlanticus, 230
Eucinostomus

argenteus, 221

gula, 220-221
Eulimnadia

agassizii, 8

alineata, 3, 6, 7

antillarum, 7

antlei, 8

diversa, 6, 8

francesae, 7

inflecta, 8

oryzae, sp. nov., 3-5, 9

stoning toniensis, 9

texana, 8

thompsoni, 9

ventricosa, 4, 5, 9
Eurymerodesmidae, 67
Eurymerodesmus

minimus, 67

varius, 67
Eutrichodiaptomus, 54
Farancia abacura, 147
Floridichthys carpio carpio, 205-206

chrysotus, 196

confluentus confluentus, 198-199

grcmdis grandis, 197-198

similis, 196-197
Gambusia af finis holbrooki, 206-207
Gerridae, 220-221
Gobiidae, 229-230
Gobionella stigmaturus, 230
Gobiosoma

bosci, 229-230

robustum, 230
Haemulidae, 222
Harengula pensacolae, 192
Hemirhamphidae, 194
Hippocampus hudsonius punctula-

tus, 229
Heterandria formosa, 211

cinerea, 23, 147
squirella, 23

Hybopsis harperi, 84
Hyporhamphus unifasciatus, 194
lridomyrmex

humilis, 16

pruinosus, 23
Jordanella floridae, 206
Julus ornatus, 67
Kilby, John D. (article), 173-247
Lagodon rhornboides, 222-223
Lampropeltis

doliata, 92

getulus, 92
goini, 147
Leiostomus xanthurus, 225-226
Lepisosteidae, 191
Lepisosteus platyrhinus, 191
Lepomis

macrochirus purpiirescens, 218-
219

microlophus, 218

punctatus punctatus, 218
Leptodiaptomus, 49-60
Limnadia, 6
Limnadiidae, 1-10
Liodytes alleni, 92
Lucania parva, 201-202
Lutjanidae, 219-220
Lutjanus griseus, 219-220
Lysiopetalida, 63
Mattox, N. T. (article), 1-10
Membras vcugrans vagrans, 215-216
Menidia beryllina, 216-217
Microhyla

carolinensis

carolinensis, 13-46
mazatlanensis, 15
olivacea, 15, 39
Microgobius gulosus, 229
Micrognathus crinigera, 228
Micropterus salmoides floridanus,

217-218
Mollienesia latipinna, 207-211
Monocanthidae, 212

cephalus, 213-214

curema, 214-215

trichodon, 215
Mugilidae, 212-215
Myridae, 196
Myrophis punctatus, 196
Nematophora, 63
Notemigonus crysoleucas bosci, 193-

194
A^ofropis

baileyi, sp. nov., 71-85, 168

chalybaeus, 168

chromosomus, 71, 73, 74, 75, 76,
77, 80-81, 82, 83, 84, 168

hypsilepis, sp. noV. 161-168,

169, 170

INDEX TO AUTHORS AND SCIENTIFIC NAMES— Continued

lutipennis, 71, 73, 74, 75, 76, 77,

80-81, 82, 83, 84

norrisi, 84

petersoni, 84, 168, 193-194

roseifs, 84, 168, 169

williami, 84

xaenocephalus, 84, 85, 161, 168

volucellus, 169
Oligoplites saurus, 222
Ophioptes sp. (near tropicalis) ,

137, 151
Opsanus beta, 227
Orthoporus ornatus, 67
Orthopristis chrysopterus, 222
Osphranticum labronectum, 51
Osteichthyes, 69-86, 159-170, 173-

247
Paralichthys albigutta, 230
Pelorodiaptomus, 54, 55
Poeciliidae, 206-211
Pogonias cromis, 226
Polydesmida, 67
Prionotus tribulus, 227
c seudobranchus striatus belli, 147
Psychrodiaptonius, 54

catesbeiana, 32

clamitans, 32

grylio, 147

pipiens, 32
Eanev, Edward
86, 159-170
Reptilia, 87-155

C. (articles), 69-

Sciaenidae, 224-226
Sciaenops ocellata, 226
Seminatrix pygea cyclas, 147
Sparidae, 222
Spheroides nephelus, 212
Sphyraena barracuda, 212
Sph.yraenidae, 212
Spirostreptida, 67
Spirostreptidae, 67
Stephanolepis hispidus, 212
Strongylura

uotata, 194-195

timucu, 195
Suttkus, Royal' D. (articles), 69-86,

159-170
Symphiirus plagiusa, 232
Syngnathidae, 227-229
Syngnathus

floridae floridae, 228

loiiisianae, 228

sco velli, 227-228
Synodontidae, 193
Synodus foetens, 193
Tetradontidae, 212
Texophon, 63
T hamnophis sauritus, 92
Trachinotus falcattis, 222
Trig-lidae, 227

Trinectes macidatus fasciatus, 231
Wilson Mildred Stratton (article),

49-60
Xj stodesmidae, 67

Q 1Q*#

»