HARVARD UNIVERSITY

Library of the

Museum of

Comparative Zoology

MUS. COMF». ZOOIJ,
L ' P5 R A e? V

OCCASIONAL PAPERS ^- ^'-^

r

of the

MUSEUM OF NATURAL HISTORY
The University of Kansas
Lawrence, Kansas

NUMBER 102, PAGES 1-20 JULY 27, 1982

DESCRIPTION OF A NEW SPECIES, FUNDULUS

JULISIA, WITH A REDESCRIPTION OF FUNDULUS

ALBOLINEATUS AND A DIAGNOSIS OF THE

SUBGENUS XENISMA (TELEOSTEI:

CYPRINODONTIDAE ) .

James D. Williams^ and David A. Etnier-

Introduction

This paper describes a new species of Funduhis from the Barrens
Plateau area of middle Tennessee. It also provides additional de-
scriptive data for the poorly known and presumably extinct white-
line topminnow, Funduhis oIl)oJineatus Gilbert, from Big Spring,
Hunts\illc, Madison County, Alabama. The Barrens topminnow,
described herein as Funduhis juUsia, was first collected during pre-
impoundment surveys conducted by the Tennessee Valley Authority
(TVA) in the late 1930s. Most of the specimens were small, but one
tuberculate male, 42 mm SL, was taken on April 1, 1937. The three
collections of Barrens topminnows taken in 1937 were recognized
as a distinct species by C. L. Hubbs in 1938. These collections were
deposited in the University of Michigan Museum of Zoology where
they were subsequently catalogued as Fundidus albohneotus. After
the original collections in the spring of 1937, it was not taken again
until the mid 1960s. In recent years, the new species and F. olho-
Uneatus had been considered conspecific until recent work by J. S.
Ramsey, R. W. Bouchard, and ourselves indicated it is more likely
that they represented two species. These species share many meris-
tic, morphological, and pigment characters with three other upland
species; Funduhis catenatus (Storer), the northern studfish; F.

' Office of Endangered Species, U. S. Fish and Wildlife Service, Washington,
D.C. 20240.

"Zoology Department, University of Tennessee, Knoxville, Tennessee 37916.

2 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

stellifer (Jordan), tlic southern studfish; and F. rathhuni (Jordan
and Meek), the speekl(>d kilhfisli. The conib'ned ^roup appears to
represent a distinet pliyletie hne in the genus Fundulus. \^arious
authors (Brown, 1957; Farris, 1968; Chen, 1971) have placed some
or all of the abo\'e species in a species group, for which the sub-
gener'c name Xeirsma Jordan has l)een used. Brown (1957) in-
cluded F. cafcnatus, F. steUifcr, F. ratJihimi, and F. alholincatus
in Xcwsmci (his concept of F. alholincatits was based on specimens
of the species from middle Tennessee, described herein as new), but
Farris (1968) and Chen (1971) suggested that the subgenus
Xovsina also includes addit'onal species of Fiinchihis such as F.
d'liphanus, F. uaccamemis, F. scminolis, and F. parvipinnis. Farris'
(1968) analysis was based primarily on skull morphology. Chen
( 1971) compared karyotypes of 20 species of Fundulus, but unfortu-
nately the only member of Xen'snici (sensu. Brown, 1957) examined
was F. rathhuni, which showed closest karyological affinities with
the species pair F. diaphanus and F. waccamensis. Miller (1955)
did not utilize subgeneric names within Fundulus but implied a
relationship among the Xcn'sma species by listing F. cafcnatus, F.
stellifer, F. rathlmni, and F. alholincatus together in his "tentative
phylogenetlc sequence." We contend that Xen'sma (scnsu. Brown,
1957), exhibits and/or shares sufficient character states either not
exhibited or not shared by other Fundtilus to be recognized as a
\'alid and mon jplnJrtic subgenus.

The diagn!)sis of the subgenus Xenisma presented herein should
facil tate future efforts t;) determine subgeneric limits in the genus
Fundulus. In cvxamining the gen\is Fundulus, Miller (1955) recog-
nized 26 recent species and pointed out that additional study was
needed before subgeneric divisions could be mad(\ Brown (1957)
provided a key to 33 species and subspecies of the genus Fundidus
which he plaec^d in five subgenera, non(> of which were diagnosed.
Subsecjuent workers have generally followed the subgeneric limits
of Brown (1957) but there ha\'e been no attempts to diagnose anv
subgenus of the genus Funduhis. Ilublxs and Burnside (1972), on
the basis of developmental and karNological data, concluded that
F. notatus and F. olivaccus were sufficiently distinct from other
Fundidus to warrant generic separation. The oldest axailable name
for these species is 7j/ii,()ncctcs. A tliird speces, Fundulus curyzona-
tus, recently described by Suttkus and Cashner (1981) was referred
to tlie subgenus Zyfionectes. Most recently Parenti (1981) diag-
nosed the genus Fundtilus and noted that the decision to recognize
subgenera was "a problem for a future rexisor."

I'hree species ol the subgenus Xcn'sina have been the subject of
detailed studes. jirown (1955) reported on \ariation and relation-
ships of F. rathhuni, and Thomerson (1969) published on the
systeinatics of F. cat mat us and F. stellifer. Fu}}dulus alholineatus

A NEW SPECIES OF FUNDULUS 3

was described by Gilbert (1S91) and is known only from 25
specinicMis taken from Big Spring Branch at nnnts\ill(\ Alabama.

Methods and Materials

Fin ra\' connts were made following methods described by
Ilnbbs and Lagler (195S). The method of counting dorsal and
anal fin rays used by Brown (1957) and Thomerson (1969) result
in one more fin ray per fin than the standard method. These fin ray
counts are adjusted herein to make them similar to the coimts ob-
tained using the methods of Ihibbs and Lagler (1958). Scale
counts were made according to the methods described by Ilubbs
and Lagler (1958) except as noted below. Transverse or vertical
scale rows were counted on the left side only, in a zig-zag fashion
from and including the scale on the dorsal midline at the anterior
origin of the dorsal fin ventrally to the last scale before the ventral
midline (Thomerson, 1969). Measurements were made following
methods of Thomerson ( 1969 ) . The terminology used for describ-
ing sensory canal system of the head follows that of Gosline ( 1949).
Head squamation pattern follows the description presented by
Hoedeman (1958). The genus Fundulus is implied for all species
epithets that appear without a generic name or initial. Specimens
examined have been kindly provided by the following museums:
University of Alabama Ichthyological Collection (L^AIC), Auburn
University (AU), California Academy of Sciences (CAS) and the
Stanford University (SU) collection stored at the California Acad-
emy of Sciences, Cornell University (CU), Universit\' of Michigan
Museum of Zoology (UMMZ), University of North Carolina at
Charlotte (UNCC), Tulane University' (TU), University of Ten-
nessee (UT), and United States National Museum (USNM). We
thank the curators in charge of these collections for making this
material available.

Information on the karyotype of Fundulu'i julisia was graciously
provided for inclusion in this paper by \\\ M. Howell and Ann
Black, Samford University, Birmingham, Alabama. The data were
taken from one adult male and three adult females collected at the
type locality on 14 June 1979.

Chromosome MethodoIog.y. — The fishes were injected intraperi-
toneally with 0.5 ml of 0.1 percent colchicine and were returned to
the minnow bucket. After 8-12 hours, they were sacrified by decapi-
tati(;n. Tissues from gills, scales and testes were removed and fixed
in two changes of 3:1 methanol glacial acetic acid for 15 minutes
each. Prepared tissue was dabbed on to the surface of a micro-
scope slide. This action released a slurry of cells which was air-
dried and stained in 4 percent Giemsa stain.

4 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

Subgenus Xenisma Jordan

Xenisma Jordan, in Jordan and Copeland 1877:142. Type spe-
cies, Xenisma stellifera Jordan, by original designation.

Included species. — Fundidus catenattis, F. stellifer, F. ratldnini,
F. alhoJineatus, and F. jidisia (described herein).

Di(ig,nosis. — In all species of the subgenus Xenisma, all stages of
both sexes lack a broad, darkly pigmented lateral band, a large,
darkly pigmented subocular "teardrop" blotch, and vertical bars
on sides of body. Nuptial males have well developed red, blue,
and yellow breeding colors (nuptial colors unknown for aJ])olinea-
tus). Nuptial males with terminal and/or subterminal dark and/or
chromatic band on caudal fin. Nuptial males with tubercles con-
sistently absent from nape and top of head, but present on dorsal
and anal fins, cheeks and opercles (alholineatiis lacks tubercles on the
cheeks and opercles) and in some species developing on paired fins
(catenatus, rathhuni, stellifer) and breast scales (rathhuni). Mid-
lateral body scales of nuptial males with 3-5 tubercles per scale.
Dorsal fin origin posterior to anal fin origin. Vertebrae 33-38. Cau-
dal rays 14-19. Branchiostegal rays 6-6, rarely 5 in rathhuni. Total
gill rakers 4-7, very blunt, most slender and elongate ones with their
basal width as great as their length. Lacking a conspicuous ridge
of connective tissue on the dorsal portion of the pharynx, extending
anteriad from the level of the uppermost gill raker.

Additional Characters. — Dorsal fin rays 9-15. Anal fin rays 10-17.
Pectoral fin rays 14-19. Lateral line scales 34-52. Transverse scale
rows 11-18. Caudal peduncle scales 16-24. Gill openings vary from
restricted in julis-ia to only moderately restricted in the other species.
Head scales consistently in the "A pattern" of Hoedcman (1958).
Preopercular canal pores typically 7 in ratlihuni, alholineatiis, and
jidisia and 8 in catenatus and stellifer. Mandibular canal typically
with 4 pores, with anterior pores (Z and Z') more widely spaced
than pores Z and Y (these distances about equal in rathhuni and
occasional specimens of catenatus and stellifer). Preoperacular and
mandibular canals widely separated. Supraorbital canals with pore
2b behind or rarely (catenatus) even with middle of posterior
nostril, pores 2a and 2b more widely spaced than pores 3 and 4a.
Postorbital canal extremely variable, but typically with pore 6 com-
pletely or partly divided; and with the canal between pores 4b and
5 persistent. Cosline (1949) indicated that this canal was typically
absent in .stellifer, but it was consistently present in the 20 speci-
mens we examined. Pit organs near tip of snout not connected by
canal.

Di.scu.ssion. — In the subgenus Xenisma, the absence of a broad,
darkly pigmented lateral band, a darkly pigmented subocular "tear-
drop" blotch, and vertical bars on the sides distinguish it fi-om all
species of Fundulns with one exception, sciadicus. Three species of

A NEW SPECIES OE FUNDULUS 5

Fundulus. ciinjzonafiis. iiotatus, and oUvaceiis have a broad, darkly
pigmented lateral band and are usually assigned to the subgenus
Zi/fionectes (see Suttkus and Cashner, 1981; Ilubbs and Burnside,
1972). The nottii speeies group is eharaeterized by ha\ing a darkly
pigmented subocular "teardrop" bloteh and numerous stripes be-
tween the flank scales of females (Wiley, 1977). All other species
of the genus Fundulus, with the exception of sciadicus, have vertical
bars on the sides of one or both sexes. Fundtdiis sciadicus can be
distinguished from species of the subgenus Xenisma in ha\'ing one
or two breeding tubercles per scale on mid-lateral body scales (3-5
tubercles per scale in Xenisnui)\ 31-34 vertebrae (33-38 in Xenis-
ma); 19-22 caudal rays (14-19 in Xenisma) and 5-5 branchiostegal
rays (6-6 in Xenisma, rarely 5 in ratJd)tini).

Distribution. — Species of the subgenus Xenisma are virtually
restricted to fresh waters above the Fall Line in southeastern and
southcentral United States. The exceptions are stellifer with popu-
lations on the coastal plain in the Alabama Rixer drainage and
catenatus with populations in tributaries to the Mississippi River
in southwest Mississippi.

Fundulus albolineatus Gilbert

Whiteline Topminnow

Fig. 1

Fundulus albolineatus Gilbert, 1891:149 (original description).

Types'.— Lectotype: We select USNM 125055 (male, 70 mm
SL), collected by P. H. Kirsch, U.S. Bureau of Fisheries, in Spring
Creek, Huntsville, Madison County, Alabama, 27-28 May 1889
from the svnt)qDic series as the lectotype. The remaining syntypes
(USNM 225996(1), UMMZ 157629(1), CAS 44176(9) and SU
1029(8)) became paralectotypes.

Additional material. — The only other known collection of F.
albolineatus (USNM 63133) was originally catalogued as F. catena-
tus. These five specimens were also collected by P. H. Kirsch from
Huntsville, Madison County, Alabama, 27-28 May 1889.

Fig. 1.— Fundulus albolineatus. Gilbert, 189L Lectotype (USNM 125055),
male 70 mm SL, collected in Spring Creek, Huntsville, Madison County, Ala-
bama, May 27-28, 1889. This figure is reprinted from the original plate XLIII,
Fig. 1, published by Gilbert, 1891. Drawing by S. F. Denton.

6 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

Diap,nosis. — A member of the subgenus Xcn'stna as diagnosed
in tliis paper. Most similar to F. julisia from wliieli it differs in
"ha\'ing the rows of seales with interrupted whitisli streaks, most
conspicuous on hinder lialf of body" in adult males (Gilbert, 1891).
Whitish streaks are still discernable even though specimens are
badly faded. Additional color and pigmentation differences be-
tween alhoUneatus and julisia can be inferred, but are relegated
to the description due to the unclear and sketch)- nature of Gilbert's
( 1891 ) comments and our difficult)- in interpreting pigment pat-
terns in these extremely faded specimens. Fiimhihts alhoUneatus
lacks breeding tubercles on the cheeks and opercles (present in
other species of Xenisma). Meristically most similar to rathhuni
and julisia, but differing from both (Table 1) in having more lat-
eral-line scales (usually 40 vs. usually 39 or fewer), and fewer verte-
brae (usually 34 vs. 35 or more). Further differs from rathhuni in
lacking a darkly pigmented line extending from under eye obliquely
forward to angle of the jaw (Fig. 2), and in having the supraorbital
canal system of head very abbreviated, with pores 2b, 3, and 4a hav-
ing their diameters approximately equal to the intervening segments
of the canal. Differs from both catcmiius and stellifcr in having 7
rather than 8 preopercular canal pores, fewer dorsal rays (9-11, usu-
ally 10 vs. 11-15, usually 12-14), fewer anal fin rays (10-12, usually
11 vs. 12-17, usually 13-15), fewer vertebrae (33-35, usually 34 vs.
35-38, usuallv 36-37), and fewer lateral-line scales (37-43, usually
40 vs. 38-52,' usually 41-48) (Table 1).

Description. — Males reaching lengths of 78 mm SL, females
reaching 75 mm SL. Due to the condition of the 25 extant speci-
mens it was not possible to obtain precise body proportions. The
general body shape is illustrated in Fig. 1. This is the drawing by
S. F. Denton used by Gilbert ( 1891 ) in the original description of
F. alhoUneatus. The specimen illustrated appears to be the large
(70 mm SL) male we selected as the lectot\'pe. Measurements
and body proportions reported by Gilbert (1891) are as follows:
head 3'A to 3% into length; depth 4 to 4^2 into length; least depth
of caudal peduncle equals snout plus % eye; interorbital distance
2)4 to 2)3 into head length; snout Yi length of head; dorsal and anal
fin bases equal in length and short, equaling length of snout plus
half of eye; in males both dorsal and anal fins become elevated,
the longest anal ray equaling % of head length; depressed pectoral
fin reaches pelvic insertion; depressed pelvic fin reaches to or nearly
to the anus; pectoral and pelvic fins shorter in females. Gilbert
(1891) reported that dorsal and anal fin origins were at the same
level or the dorsal origin was slightly anterior to anal origin. This
appears to be in error since the illustration of the species and our
observations indicate the dorsal fin origin to be slightly posterior to
the anal origin. Examination of radiographs shows the first inter-
neural spine extending anteriad and not interdigitating with neural

A NEW SPECIES OF FUNDULUS

S
o

n3

.S2

t3 C

•So

C3

C
c

«
>

C

a:

ic

a

CO
V.

O

CO

CO CD in o^ i^
as C30 t^ t> T— I
CO CO t— I

O lO CO O 00
O OC t— CO ^H
CO CO -— I

t- CO

Tr in

CO 00

.-H CO

C-l O
CO CI

CD

m

oq -r ^ c CO

1—1 CO (M (M

(M CO Tf T}<

00 (M

>— I C35 ■— I
t- ^ i-H

cn

CO

M

05 00

-rtf 1— (

O CD

o 00

lO t^ C5 O Ol
00 00 O l^ CM
CM CO I— I

C35

00

Cl)

00 O 00

00 '—I

O 00

a> m

CD

lO

CO Ol

CD -t CD

+

2 ■^ 5 « ;-
t ::= S -H g

ct] t; t fc^ t,'

+

s; .3 ,:c:

S ^ 2 ■= -^

tC ti^ ki ti,' ft]

(M -t O

■-H CM >— I

UO O CM

CO 00 IC

05 r^ CO in 00

■—I CD -n* I— I

CD m CO CO CO
■— I •— I -^ I— I

CM —( r-(

*

;j '^ i-

CO

3

■=: a

[t. k, k. t, t^

OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

-s3

C

o

g

CT>

^ 00

<u

O

c
ca

a

CD

in

a

CO

in m 00 c 05

— I (M 05 t-
1—1 (M

CO

cq

05 CO

OC Cj^ CO CM CO

O O I- lO C^
in (M (M CO

Tf 00 -r Di

CM TT

CM CO
(M

oc

CO

~2 l~

^ CO

CO
CO

in

CO

CO

CO
CO

00 in m 00 m
CT> in '— I — I (M

tM oq

-H 00

CD (N

in CO

CM O]

^ in

C-l 05 CI C£

in

so

z

in CD

1—1

CD -^
CO (M

(M

CO

00

05
1— (

r-l Cq

O}

C/2

CI

u

00

t-

,-( CO

•— I C? -H 1— I 00

'-^ l/> r^ 3

' --H CO ^

—I CO m CO
in CO (N

i> TT ,— I

in

T cq

CI

*

= 1 •= ' §

~ ^ ^ .2 ~

S = S .2 g

St? g .= "a

k.' tt^ fi] ti,' ti,"

^ -^ -c .^^ >e
*j t, ■«- "C: -c

^ ^ I, •-_ a

ti< U, ti. k, ti.

+

a

a 1^

(i, U. ti. k, ti<

A NEW SPECIES OF FUNDULUS

c

o

in

xo

o

05

00

a
0:1

^ IX)

I

"a in

O

05

• oo
n

Xr-

CD

n

in

CO

CO

CO 03 t^ CD in

00 t^ O CO (M

Cq CO --H

^ ^

in ^

CO CO

CD

-1* CD
CO C<1

CO
CO

CO

05

CO

CO

00

CO o

CO in

csi in

in -t CM oi
■-I CO

^ CO CD in
--1 CO

CSI CO r-l CD o

CM 05 Tf in in

^

Oi

<—{

0

CM

05

1— 1

CI

CO

^

0

CM

»— (

00

CO

in

Co

3

o

CO

C

c

H i-H

CM l^ CM CO O

00 CO I- ^ <M
(M CO

CO

<M

(M

2 * —

5j

c ^ ^ .2 ~

-S ^ -S ;S -§

2 ^ 2 .= "3

tt; [i^ ti] lii til

Gi ■— I

CO

CO Tf

CO CD C3> t- CM
-TT CD CM >— (

00 CM ^ CO l>

CM CO in ^

in m ^ --I
00 ^

00

ci

*

Co
3

1 "s

CO S

b, [i, t, t, tc

10 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

spines; the interneuial spine of the second dorsal fin ray Hes be-
tween the neural spines of the 16th and 17th vertebrae. The in-
terhemal spines associated with the first two anal rays are between
the hemal spines of the 15th and 16th vertebrae. Sensory canal
system of head basically as described for subgenus Xenismo.

Gilbert (1891) apparently saw the original types only after
they had been in preservative for some time, since he reported no
chromatic colors which surely would ha\'e been present on these
nuptial specimens. He reported the color of males as "blackish
brown, the sides plumbeous, the rows of scales with interrupted
whitish streaks, most conspicuous on hinder half of body. A black
streak along middle line of back. Vertical fins dusky, the caudal
becoming translucent on distal half, its margin abruptly and nar-
rowly black-edged." The translucent band adjacent to the black
margin of the caudal fin was probably brightly colored, yellow to
orange in life, as is characteristic of other species of the subgenus
Xenisma. Females were described by Gilbert (1891) as being
"olivaceous, dusky on back, silvery below, the back and side with
narrow black lines following the rows of scales. Fins translucent,
the dorsal sometimes with fine specks at base, the caudal black
edged." Very little of the pigment pattern remains on the 25 extant
specimens which have faded to an almost uniform brown color. The
discernable pigment pattern on the body is as follows: males with
a prominent white predorsal streak partially overlain with melano-
phores, and with 6-7 broken white lines laterally on the posterior
half of body, each line follows the central Yi to /'s of the exposed
field of the scale; females with dark spots forming thin lines along
the center of the exposed field of scales on 3-5 rows of scales on
each side of the dorsal midline; larger spots present on sides of
females but do not appear to form rows. The dorsal, anal, and
pectoral fins of males appears dusky with pigment present on rays
and membranes; on the dorsal and anal fins the concentration of
dusky pigment appears to be heavier on the posterior 2-3 rays than
on the anterior ravs. The caudal fin in males has a narrow marginal
dark band, a light submarginal Ixmd, and the basal % of the fin is
dusky. Females with a narrow dark marginal band on caudal fin that
is less prominent than that of males. Nuptial tubercles, present only
on males, occur on the dorsal and anal fins and mid-lateral l)ody
scales. Some males have a few tubercles on the anterior caudal pe-
duncle scales. On the dorsal fin, they are present on the first eight
rays. Tubercles are present on the distal ]'i to % of the anal fin, with
the largest tubercles on the branched rays. There are about 3-5 tu-
bercles on the periphery of the mid-lateral body scales. The 15
nuptial males examined lack tubercles on the paired fins, cheeks,
and opercles.

Habitat. — The only known habitat of the whiteline topminnow
was Big Spring and its outflow. Spring Creek, in Huntsville, Madi-

A NEW SPECIES OF FUNDULUS 11

son County, Alabama. Tlic habitat was described by Gilbert (1891)
as follows: "Spring Creek, IIuntsN'ille, May 27 [1(S<S9]. Tempera-
ture 65° F. This small stream about IS feet wide is formed b\' a
single spring in the town of Iluntsville. It is about }i of a mile
long and flows into Pin-hook Creek. Its bottom is similar to that
of the former [Pin-hook Creek, Iluntsville, bottom of blue lime-
stone]. It is full of fishes, darters being very numerous." Armstrong
and Williams ( 1971 ) reported on two collections from Big Spring
in IIuntSN'ille but no F. albolincatus were taken. Big Spring has
been drastically altered since the turn of the century. According
to local residents, the spring has been pumped dry on several oc-
casions. The banks of the spring and the spring run have been
lined with cement. The spring outflow flows through a city park
and has been impounded and stocked with carp and goldfish.
Native fishes reported from Big Spring by Armstrong and Williams
(1971) include Semot'lus atromactilatus, RJiinicJitJiys atrattthis,
Campostoma anoinahun, Catostomus commersoni, and Etheostoma
squom'ccps.

Status. — Spring type habitats near the lower bend of the Ten-
nessee River and in the adjacent upper portion of the Mobile Basin
drainage contain at least five additional endemic species. In the
upper Mobile Basin springs, Cottus pypnacus is known only from
Coldwater Spring, Calhoun County, Alabama (Williams, 1968);
Etheostoma nucJialc (;ccurs in several springs near Birmingham,
Jefferson County, Alabama (Ilowell and Black, 1976); Etheostoma
ditrema is somewhat more widespread and occurs in at least se\'en
localities in Alabama, Georgia, and Tennessee ( Ramsey, 1976; Stiles
and Etnier, 1971, and an unpublished record, UT 91.782). In the
lower bend of the Tennessee River Elassoma sp., formerly known
from Cave Spring, Lauderdale County, Alabama, and Pryor Springs,
Limestone County, Alabama has been extirpated from these areas
but was recently rediscovered in Moss Spring, Limestone County,
Alabama; Etheostoma tuscuml)ia persists in about 10 springs in
northern Alabama (Ramsey, 1976). Rediscovery of Elassoma sp.
and a thriving E. tuscumhia population in Moss Spring, Limestone
County, Alabama, in 1973 raised hopes that F. alholineatus might
occur there. After considerable collecting effort in Moss Spring,
this no longer seems likely. Although disovery of an extant popu-
lation of F. alhoVneatus in springs near the lower bend of the Ten-
nessee River remains a possibility, it seems more likely that the
species is extinct.

Eunduhis jiilisUi, sp. nov.

Barrens Topminnow

Fig. 2

Funduhis alholineatus Gilbert. — Miller, 1955:9 (distribution in Ten-

12 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

ncssee). — Brown, 1957:71 and 74 ( distribntion in Tennessee). —
Farris, 1968:4 (used in osteological comparisons). — Armstrong
and Williams, 1971:110 (reported from Coffey County, Tennes-
see).— Ramsey, 1976:63 (reported related form from eastern
Highland Rim of central Tennessee).

Fundutiis sp. — Miller, 1972:214 (reference to Barrens topminnow
in Tennessee). — Bouchard, 1973:40 (reference to undescribed
Fiindulus). —Srhastava and Griffith, 1974:137-139 (reported
from springs in Tennessee ) .—Griffith, 1974:320, 321, 324, 326
( unnamed form from Tennessee, Barrens topminnow ) . — Deacon
et al., 1979:41 (Barrens topminnow, color photograph).

Holotype.— Adult male, USNM 225997, 53.8 mm SL, spring
tributary to West Fork Hickory Creek, Caney Fork River system,
Cumberland River drainage, at Tennessee Highway 55, 0.3 rd. mile
northeast of junction with County Road 4292, Summitville, Coffee
County, Tennessee (spring originates from cave on Joseph R. Banks
property, on north side of Tennessee Highway 55), 1 April 1973,
collected by David Etnier, Ray Bouchard, and Frank V. Oakberg.

Paratopotypes. — UT 60. 103 ( 1 ), same date as for holotype; UAIC
2417(1), 18 December 1966; UAIC 2536(7), 13 April 1967; UT
60.28(13), 20 October 1968; AU 15728 (11), 13 September 1969;
UT 60.53(1), and UT 60.55(5), 3 May 1972; UT 60.61(2), 30
September 1972; UT 60.218(5), 9 July 1977; University of Tennes-
see Zooarchaeology Collection 4405(1), disarticulated skeleton, 5
May 1978; USNM 225999(1), 11 March 1979; Florida State Mu-
seum (UF) 31452(2), Illinois Natural Historv Survey (INHS)
87282(2), Academy of Natural Sciences of Philadelphia (ANSP)
146524(2), University of Kansas (KU) 18845(2), UMMZ 207690
(2), USNM 225998(2), all from 14 June 1979, and originally prop-
erty of Samford Universitv, Birmingham, Alabama; UT 60.225(2),
23 May 1981.

Additional Parotypes. — Caney Fork River system, Cumberland
River drainage. Coffee County, Tennessee: TU 33495(1), tributary
to Hickory Creek, 13.1 mi southwest of McMinnville, on Tennessee
Highway 55, 17 July 1964; AU 11059(5), West Fork Hickory Creek
0.8 mi. south of Summitville, 0.2 mi. south of Tennessee Highway
55, on County Road 4292, 28 August 1975; AU 12715(1), same lo-
cality as AU 11059, 18 March 1976. Duck River system, Tennessee
River drainage, Coffee County, Tennessee: UMMZ 120861(3),

Fig. 2. — Breeding males of four species of the subgenus Xenisma. The
four species from top to liottom are Fttndulus catcnatiis { i 115 mm SL col-
lected 12 May 197.3, Mulherrin Creek, Snuth County, Tennessee), F. stellifer
( S 97 mm SL collected 29 April 1973, Conasauga River, Bradley County,
Tennessee), F. rathhuni ( i 75 mm SL collected 28 April 1979, Back Creek,
Mecklenburg County, North Carolina), and F. julLsia ( c^ 68 mm SL collected
23 May 1981, spring tributary to West Fork Hickory Creek, Coffee County,
Tennessee).

A NEW SPECIES OF FUNDULUS

13

;,?y"i-'rt«a»»^

14 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

spring branch above Ovaca Lake, north of Tullahonia, 26 March
1937;" UMMZ 120914(8), Little Duck River at Tennessee Highway
55, northeast of Manchester, 30 March 1937; UMMZ 121014(7),
Doak Springs on Hunt Creek, southeast of Manchester, 1 April
1937; TU 33511(3), same locality as UMMZ 120914, 17 July 1964.

Dia<i.nos's. — A member of the subgenus Xenisma as diagnosed
herein. Most similar to alhoVincatus, but differs most notably in the
absence of interrupted whitish streaks on posterior half of the body
in breeding males. The body of breeding males of julisia is irides-
cent yellowish green with scattered orange to reddish orange spots.
It also differs from alhoUneatus in having a lower modal lateral-line
scale count (3S or 39 vs. 40), and a higher modal vertebral count (35
vs. 34). Differs from cafenatus and stclUfer in having fewer dorsal
fin rays (9-12 vs. 11-15), fewer anal fin rays ( 10-12 vs. 12-17), fewer
vertebrae (34-36 vs. 35-38), and fewer lateral-line scales (36-42 vs.
38-52) (Table 1). Difl:ers from rafhJnini, catenatus, and stcUifer
in lacking nuptial tubercles on paired fins, and caudal peduncle
scales; differs from alhoUneatus in having tubercles on the cheeks
and opercles; and in having an extremely abbreviated supraorbital
canal, in which pore diameters of pores 2b, 3, and 4a are essentially
equal to lengths of canal segments between these pores ( segments of
this canal are nuich longer than pore diameters in those three
Xenisma). Further differs from rathhuni in lacking tubercles on
breast scales and in lacking a darkly pigmented line extending from
under eye obli(^uely forward to angle of jaw (Fig. 1).

Description. — Measurements expressed as thousands of SL, for
10 adult males and 10 adult females (mean and range for males,
followed bv datum for holotvpe, followed by mean and range for
females) are as follows: SL 47.4, 35.5-58.8, 53.8, 50.5, 42.8-78.2;
head length 287.1, 276-299, 285, 274.8, 263-285; head dc^pth 170.3,
155-189, 162, 159.4, 148-177; head width 170.7, 160-180, 169, 162.4,
147-179; snout length 97.1, 85-107, 99, 87.1, 78-102; orbital diameter
81.7, 74-89, 74, 75.2, 68-82; predorsal length 660.0, 645-675, 668,
677.6, 653-694; caudal peduncle length 234.5, 223-250, 240, 232.5,
206-246; caudal peduncle depth 130.0, 124-133, 132, 123.5, 117-133;
body depth 218.3, 202-227, 222, 221.0, 199-236; length of anal fin
base 138.9, 130-154, 136, 124.6, 106-139; anal fin length 294.6, 282-
306, 294, 222.5, 202-236; caudal fin length 256.5, 236-280, 246, 234.5,
211-247; dorsal fin length 272.6, 248-308, 275, 215.2, 203-234; pec-
toral fin length 177.4, 167-190, 179, 168.1, 156-186; pelvic fin length
123.0, 107-132, 132, 99.4, 91-107.

Nuptial males with blackground color of pale blue-green to
yellow-green on sides of head and b:)dy, fading to white on ventral
surface; diffuse orange to dark red spots, each approximately occu-
pying the exposed field of a scale, liberally scattered on sides, and
on head Ijclow cheek, on operculum, and occasionally present on
ventral surface of head and body; dorsum olivaceous, with a promi-

A NEW SPECIES OF FUNDULVS 15

nent iridescent, wliite to pale gold mid-dorsal stripe extending from
dorsal-fin origin about % distance to occiput; anal fin orange-yellow,
with scattered orange spots; paired fins )'ellow, occasionally with a
few orange spots near base; bt)th caudal and dorsal fins with a few
orange spots near base; caudal fin and occasionally posterior portion
of dorsal fin witli a broad pale yellow submarginal band and narrow
dark margin. All fins with many small dark chromatophores along
rays and on membranes; eye with iris bright yellow darkened by
scattered dark chromatophores. Freshly captined females and juve-
niles much more subdued, brownish gray on sides with widely
scattered dark brown spots about size of orange spots of male (not
with "narrow black lines following the rows of scales" as reported
for alholineatus by Gilbert); dorsal and caudal fins with small
dark chromatophores margining rays and abundant on membranes;
dark chromatophores absent from paired fin membranes and few
on anal fin; dark predorsal streak extends from dorsal-fin origin to
occiput. Pigmentation of females and juveniles changes little in
preservative. Males in preservative lose orange and red spots, leav-
ing sides uniformly pigmented (not with "the rows of scales with
interrupted streaks" as reported for alholineatus by Gilbert); pale
yellow band on caudal fin detectable only as a pale area near fin
margin (black marginal band on the caudal fin noticed by Gilbert
on alholineatus is often not evident in preserved specimens of
julisia ) ; iridescent predorsal streak becomes dark brown.

Somatic chromosome counts from both gill and scale tissues were
2n = 4S in both male and female (Table 2). No chromosomal
sexual dimorphism is apparent. The chromosomes are all acrocen-
tric except for a single pair of small submetacentrics (Fig. 3).
Meiotic chromosome counts from testes of the male were n := 24.
Although this karyotype is similar to that reported for F. rathhuni
(Chen, 1970), it differs very little from that of several other species
of Funclulus reported in that paper. It is also similar to the karyo-
type of F. stellifer reported by Denton & Howell (1969). Our lack
of expertise in this area forces us to relegate the critical comparison
of Fumlulus karyotypes to serious students of this technique.

Hahitat. — The Barrens topminnow is known from springs, spring
runs, and small streams strongly influenced by groundwater. In
both the type locality and Little Duck River, watercress (Nas-
turtium officinale) occins in lush growths. At the type locality
the spring emanates from a cave about 100 m north of Tennessee
Highway 55 on the Joseph R. Banks property, Summitville, Coffee
County. Until recently, the spring run was about 2-3 m wide
and 10-30 cm deep, with moderate ciurent. Chert-gravel substrate
predominates in the current-swept areas, while the depositional
areas associated with the watercress mat and the few im vege-
tated pools contain mud and silt substrates. Barrens topmin-
nows typically occur in the calm-water habitats. W^ithin the past

16

OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

o
a
a

a
c
H

c

l-l

m

a

.3_

C

c
<-t-l

lU
C
"rt

o

c

3
o

U

e

o

O

s

o

;-!
J2
U

n^

'o

ID
_>

"S
o

s

CS
CO

^'

o
a

s

c

G
<

C
03

O

Q '^

O O

C PL,

4->

Q

U

I-

O P c

u

00

c

o
O

"a
Q

CD

10

O T)
S^ N

-Q^ >^

^ 3
z ^

o c^J
I, qj a;

qj ;:^ N

C O a!
§ n !=

00 00

05 00

CI

X

en

CO

ca

CO

*o ot

c

hO

"C

3

c

a

o
•x:
a

S

o

•t-t

QJ

3

T3

O
O

42

o
a

o

3

s
o
o

G G

CO

Crt

u -0

A NEW SPECIES OF FUNDULUS 17

iew N-ears, Mr. Banks placed a low dam across the spring run just
abo\e Hi,u;hwav 55 in an attempt to provide a pool area for rainbow
trout. Although the trout rapidly disappeared downstream, the
pool, fortunately, provides a fax'orablc habitat for the topminnows.
One is often able to see several schools of them, about 10-20 per
school, while standing on the edge of this M-acre pond. The spring
run extends only about 400 m before entering West Fork Hickory
Creek, which is currently receiving the effluent from a trailer court.
Barrens top minnows have been occasionally taken in West Fork
Hickory Creek near the mouth of the spring run. Noteworthy as-
sociated species at both the type locality and the Little Duck River
include Hemitremia flammea, Chologastcr ap^assizi, and Etheostoma
lufeovinctum. A male of the phryganeid caddisfly OJi^ostomis ocel-
ligcra (Walker), otherwise known only from as far south as Massa-
chusetts and Michigan, was reported from the spring run by Etnier
(1973).

Biology. — Spawning apparently takes place in early spring, based
on presence of tuberculate males in available collections. W. M.
Howell and S. K. Osborn, Samford University, have successfully
reared this species in aquaria. They found (pers. comm.) that
sexual activity could easily be initiated at other times of the year
by providing the appropriate photoperiod. In aquaria adult males
were extremely aggressive, often killing conspecific males and fe-
males, and on one occasion quickly killing a Fundulus oUvaceiis
of comparable size.

Aquaria were maintained at room temperature, and food pro-
vided included frozen brine shrimp and Tetramin. Spawning oc-
curred at water temperatures of 20-22° C. Spawning activity was
apparently initiated by the male's chasing the female until she
retreated into the spawning cover provided. Actual spawning was
not observed because of the density of the cover, which consisted
of floating mats of vegetation from the type locality, and a nylon
mop suspended from a float on the surface of the aquarium. Fe-

I MIA« • » ^

Fig. 3. — Mitotic karyotype of a female Barrens topminnow, Fundulus
juJisia. This cell was derived from scale epithelium and shows a 2n = 48. The
karyotype contains 46 acrocentric chromosomes of graded sizes and 2 small
submetacentrics. Scale bar = 5 microns.

18 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

cundity is unknown, but a single spawning pair produced about 25
eggs daily which were hiter found adhering to the spawning sub-
strate.

Status. — Fundulus jtdisia at present appears to be virtually re-
stricted to the spring-run habitat at the type locality. This popula-
tion, based on visual observation of the easily seen topminnows,
probably does not exceed 200-.500 individuals. In spite of repeated
efforts, the species has not been collected in the Little Duck River
locality since 1964, and has apparently disappeared from both Doak
Springs and the Lake Ovaca locality, where only the 1938 collec-
tions are known. Dr. J. S. Ramsey, Auburn Universit)' (in litt.),
surveyed about 40 small stream localities in Coffee County in 1975,
but failed to locate additional populations. At present, this species
must be considered as one of the most critically Endangered fishes
in eastern North America. We applaud the recent efforts of Ten-
nessee Wildlife Resources Agency, the Nature Conservancy, and
Mr. Joseph R. Banks, who, on August 13, 1980, joined in a coopera-
tive agreement designed to maintain the population of Barrens
topminnows at the type locality, and to monitor this habitat care-
fully in order to detect any changes that might be detrimental to
this population (Rogers, 1980). One of us (DAE) has received
support from the U.S. Fish and Wildlife Service, Office of Endan-
gered Species, to continue to search for additional populations of
this species.

Subsequent to the preparation of this manuscript an additional
population was found in the West Fork Hickory Creek System
(Caney Fork drainage) in Coffee County, Tennessee. The popula-
tion was discovered in October of 1981 in Meadow Creek about 150
yards above the L&N Railroad bridge, north of Tennessee Highway
55. This locality is approximately 2.6 air miles north-northeast of
the type locality.

Etymology. — The species epithet, jiiUsia, is derived from the
Cherokee "amjulisi" = watercress, and "atsat" = fish (King, 1975).
The common name. Barrens topminnow, was suggested by J. S.
Ramsey, and refers to the Barrens Plateau area of the Highland
Rim physiographic province to which the species is apparently
restricted.

LiTERATUHE CiTED

Armstrong, J. G. and J. D. Williams. 1971. Cave and spring fishes of the
southern liend of the Tennessee River. T- Tenn. Acad. Sci. 46(3):
107-115.

Bouchard, R. W. 1973. A preliminary survey of the distribution of fislies in
the Caney Fork Ri\er svsteni, Tennessee. Bull. Assoc. Southeast. Biol.
20(2):40.

Browx, J. L. 1955. Local \ariation and relationships of the cyprinodont fish,
Fundulus rathbtini Jordan and Meek. T- Elisha Mitchell Soc. 71(2):
207-213.

A NEW SPECIES OF FUNDULUS 19

Brown, J. L. 1957. A key to the species and subspecies of the cyprinodont
genus Fimdulus in the United States and Canada east of the Conti-
nental Divide. J. Wash. Acad. Sci. 47(3):69-77.

Chex, T. R. 1971. A comparative chromosome study of tlie genus Fundidus
(Teleo.stei, Cyi^rinodontidae). Chromosoma. 32:436-453.

Deacox, J. E., G. KoBETiCH, J. D. Williams and S. Coxtheras. 1979. Fishes
of North America endangered, threatened, or of special concern: 1979.
Fisheries Bull. Am. Fish Soc. 4(2):29-44.

Dextox, T. E. and W. M. Howell. 1969. A technique for obtaining chromo-
somes from tlie scale epithelium of Teleost fishes. Copeia. 1969(2):
392-393.

Etnier, D. a. 1973. Extensions of the known ranges of northern Trichoptera
into the southern Appalachians. J. Ga. Ent. Soc. 8:272-274.

Farris, J. S. 1968. The e\'olutionary relationship between the species of the
killfish genera Fmidulus and Profunduhis (Teleostoi, Cyprinodontidae).
Ph.D. Dissert. Univ. Mich. 73 p.

Gilbert, C. H. 1891. Report of explorations made in Alabama during
1889, with notes of the fishes of the Tennessee, Alabama, and Escam-
bia rivers. Bull. U. S. Fish Comm. 9:14.3-159.

GosLixE, W. A. 1949. The sensory canals of the head in some cyprinodont
fishes, with particular reference to the genus Fundulus. Occ. Fdp. Mus.
Zool. Univ. Mich. 519:1-21.

Griffith, R. W. 1974. Enxironmental and salinity tolerance in the genus
Fundulus. Copeia. 1974(2) :319-331.

HoEDEMAx, J. J. 1958. The frontal scalation pattern in some groups of tooth-
carps. Bull. Aquatic Biol. 1:2.3-28.

Howell, W. M. and A. Bl.\ck. 1976. Status of the watercress darter. Proc.

Southeast. Fish. Council. l(3):l-3.
HuBBS, C. and D. F. Burxside. 1972. De\elopmental sequences of Zygoncctes

notatus at Severn] temperatures. Copeia. 1972(4) :862-865.

HuBBS, C. L. and K. F. Lagler. 1958. Fishes of the Great Lakes region. Univ.
Mich. Press, Ann Arbor, Mich. 213 p.

JoRDAx, D. S. and H. E. Copelaxd. 1877. Check list of the fishes of the
fresh waters of North America. Bull. Bufl'alo Soc. Natur. Sci. 3:133-164.

KixG, D. H. 1975. A grammar and dictionary of the Cherokee language.
Ph.D. Dissert. Univ. Ga. 294 p.

Miller, R. R. 1955. An annotated list of the American cyprinodontid fi.shes
of the genus Fundulus, with the description of Fundulus pcrsimilis from
Yucatan. Occ. Pap. Mus. Zool. Univ. Mich. 568:1-25.

Miller, R. R. 1972. Threatened freshwater fishes of the United States. Trans.

Am. Fish Soc. 101(2 ) :2.39-252.
Parexti, L. R. 1981. A phylogenetic and biogeographic analysis of Cyprino-

dontiforni fishes (Teleostoi, Atherinomorpha ) . Bull. Am. Mus. Nat. Hist.

168(4) :.335-557.
Ramsey, J. S. 1976. Freshwater fishes, in Endangered and threatened plants

and animals of Alabama. Bull. Ala. Mus. Nat. Hist. No. 2:53-65.
Rogers, T. 1980. Tiny fish finds refuge in family's prixate pond. The Ten-

nessean. 2 Nov. 1980. Nash\'ille, Tennessee.

Srivstava, a. K. and R. W. Griffith. 1974. Erythrocyte morphology and ecol-
ogy of species of Func/i/Zi/ v. Copeia. 1974( 1 ) :136-141.

Stiles, R. A., and D. A. Etxier. 1971. Fishes of the Conasauga River drain-
age of Polk and Bradley counties, Tennessee. J. Tenn. Acad. Sci. 46:
12-16.

20 OCCASIONAL PAPERS MUSEUM OF NATURAL HISTORY

SuTTKus, R. D. and R. C. Cashxer. 1981. A new species of cyprionndontid
fish, genus Fundulus (Zygonectes), from Lake Pontchartrain tiilnitaries
in Louisiana and Mississippi. Bull. Ala. Mus. Nat. Hist. fi:l-lS.

Thomersox, T. E. 1969. Variation and relationship of the stiidfishes Fundulus
catcnatus and Fundulus stcUifer ( Cvprinodontidae, Pisces). Tulane
Stud. Zool. Bot. 16(1):1-21.

Williams, J. D. 1968. A new species of sculpin, Cottus pi/fimocus, from a
spring in the Alabama Ri\er basin. Copeia. 1968(2) :334-342.

University of Kansas Publications

MUSEUM OF NATURAL HISTORY

The University of Kansas Publications, Museum of Natural
History, beginning with volume 1 in 1946, was discontinued with
volume 20 in 1971 . Shorter research papers formerly published in the
above series are now published as Occasional Papers, Museum of
NaUiral History. The Miscellaneous Publications, Museum of Natural
History, began with number 1 in 1946. Longer research papers are
published in that series. Monographs of the Museum of Natural
History were initiated in 1970. All manuscripts are subject to critical
review by intra- and extramural specialists; final acceptance is at the
discretion of the publications committee.

Institutional libraries interested in exchanging publications may
obtain the Occasional Papers and Miscellaneous Publications by
addressing the Exchange Librarian, The University of Kansas Li-
brary, Lawrence, Kansas 66045. Individuals may purchase separate
numbers of all series. Prices may be obtained upon request addressed
to Publications Secretary, Museum of Natural History, The Univer-
sity of Kansas, Lawrence, Kansas 66045.

Editor: E. O. Wiley
Managing Editor: Joseph T. Collins

PRINTED BY

UNIVERSITY OF KANSAS PRINTING SERVICE

LAWRENCE, KANSAS

7 07 0 \\k

3 2044 093 361 699

DATE DUE

ARfl:i-a-fe04

DEMCO, INC. 38-2931