ie —_——. — 2. ..-_3 Rare ae ee ee 2 a Leer ee ge eed = Kata ba sibs ans 2 ae

a 2 VS = ‘S ar NESS ie ie = se

) W. SX > oO ‘ th My , O's 2 | 5 \

ME 2 2 $G4G 2 a

- 2 = 2 Uy = 2 ‘e.¥

: > = 7 > @G = > Pie.

z= ” = 79) ee ” Mf

Ai PMITHSONIAN _INSTITUTION NOILOLILSNI_ NVINOSHLIWS $3

7 : = ese

=! = “mt Be 4 WNW 24

a / Oe a \ 4 SS So :

a Oe a per S WS oc

Zz ii ‘i 2 a S a

NOILALILSNI NVINOSHLINS S3IYVYUSIT LIBRARIES SMITHSONIAN INS

6 Si S 0 _ 2 ee

ca ow — oo — wm

= A 5 5 ee 2 [Sf

ae > Ee > = 2 > let

= = - 2 ral — oe

= o — pia an. F

Saluyvudit s LIBRARI id ibis eoieltael pNOMLOLILSNI _ NVINOSHLINS, $3 1%

| = = < = < =

a = = a 4 |

oa e) 2 yy QS ZB 2 oY

= z = \S SS edi _ ae

A ae Gian see 5 :

INSTITUTION 1p NOILNLILSNI_ NVINOSHLINS Sa fyYVvug 11 LIBRARIES SMITHSONIAN _INS q

We) rr rT) Ff a 2 -

c 2 c = z 4

< = a S Y; J 4 =

oc + &€ & [oa ; fre ve

= fs) ee 3S ay 5 a

J Pa ‘ad z % .

saiIvvyaly _t! BRARIES _MITHSONIAN _INSTITUTION NOILNLILSNI ~NVINOSHLIWS “sar

= S) tte S 2 z

7 a ve je —- fa

> ae = = D E LA

te : rm > ; =<

a = = eZ 2 = We

m = m ” - a ‘ee

” = wo = ye = iF

INSTITUTION NOLLALILSNI_ pS tINWYGIT_ LIBRARIES, SMITHSONIAN INSTI

ei 2 ny DB nh a SS

= Be S - g = XN s\

> Z = > = i & =

ARG ” = ” Wie. ae a

ss LIBRARIES SMITHSONIAN _ INSTITUTION NOILNLILSNI NVINOSHLINS $3

ir Pa) 8 Wa Ww

= oe! na oc 4 Ag ow Ke

Fa <x 4, = < | ie SS <x (5&

sy a 4 ac ¢ \ cc

2 : 5 = 5 yo

he 1 af Zz ~! aye .

INSTITUTION | NOILNLILSNI _NVINOSHLINS 4 IYVUdIT LIBRARI ES ~~ SMITHSONIAN~ INS T

ro) apg ) = 5 BY - | ‘

a oe) . — = Le

5 7 Wa = ms ie Wy, » ~

E “0 A = oa = GL = Ke

NN Si : iW AP 2 X&

= w sh = o 7 S on a

SSIUVUSIT_ULIBRARIES SMITHSONIAN INSTITUTION NOILNIILSN!_NVINOSHLIWS $3 uy

Wi = = Sues Z i z ey = SONG z vy

; . 4+ /s EN | ye cr tf, > y < a Yyy

Vly, 5 Sr 2M, z KR 5 G2 Kam 1 G

q — ball : _ Y) Se eS ce

NOLLNLILSNI_NVINOSHLINS, Sa IUVUGIT_ LIBRARIES, SMITHSONIAN INSTITUTION NOILNI

re Py ee as ZG. §

6 x Bh S = BNO Z|

2 SG4e 2 z 2NX 8

2 2G = 2 = Z

| / = > ad = > = >

. oo. Zz 7) z | an 2

LIBRARIES SMITHSONIAN NOILNLILSNI NVINOSHLINS

& = w WW

oe . = oc a, XS we a

ee | = o = i a a —4

ee ro) 3 o = o

aul = ot = apne =

NVINOSHLIWS LIBRARIES. SMITHSONIAN INSTITUTION NOILN!

os z a Sua z

oO a w —_ My ee) a

ae - if

Aa wa ify Fa

> = > = YA Ee

2 = 2 a OG fr 2 =

SMITHSONIAN INSTITUTION NOILNLILSNE NVINOSHLINS sasiuvud I1 LIBRA

” Ze w z 27) : =

X = =< = Wy = = : nee

IW 2 2 7 bi lr * Z 2

‘\ z ie 2,74 fe" E é =

ie > = ~ > = >" =

1 * z 77) Se 0h za (9) = MO

K OILNLILSNI_NVINOSHLINS S3 bYvVddit LIBRARI ES SMITHSONIAN INSTITUTION -

G ws Fe ts OD ra

ra ft =. we ea oe

. ee = < en. <

2 oc c oc ar cr

2 0 a. oO = 0

: Wash Oo SS iss A a re oO ae

; : a ed : zZ ae 2 3

LIBRARIES SMITHSONIAN INSTITUTION Saiuvudiia LIBRA

2 “00 2 m0 = w

4 a E : = 2

ie = . " E =

. m poly ee. — rm) m

= Ca ee w . £ o

NOILALILSNI NVINOSHLINS SSIYVYSIT LIBRARIES, SMITHSONIAN NOILN

t = Uy = weet op) — w

= = Pe oe eS ie =

S 2 i sp ENS = S a

ss SGee FW 8S 2 S

= 2 “yy = 2 = 2

i Z s > ae = > = a .

* 7) z 7) ee an eae

t IBRARI ES SMITHSONIAN INSTITUTION NOINLIISNI_ Sauvage _ LIBR,

ge 5 o = ae =

a = S & S

te eee o = rs) oa o

a al = onl 4 wal 9

NOILALILSN!I NVINOSHLINS S3IYVUIT INSTITUTION NOILA

°; ee rs = S ~ S

“i aa > 2) = ) —

i, > - > ri > =

, = r ai e i =

{ IBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLIWS S3IMVUSIT LIBR,

a. ” Be ie w z n 2" &

i oN STS = ee | eae ee) lee): SS

ANNALS

OF THE

SOUTH AFRICAN MUSEUM

VOLUME 55

PRINTED FOR THE

TRUSTEES OF THE SOUTH AFRICAN MUSEUM

1969-1970

LIST OF CONTENTS

Page

Hu ey, P. A.

An investigation of the Rajidae of the west and south coasts of southern Africa

(published October 1970) .. mile aie 3g ue hs re Be ee ih

PEnrITH, M.-L.

Apletodon pellegrint (Chabanaud) and other clingfishes ood Gobiesocidae) from

South West Africa (published October 1969) _—«... - she 123

PenritH, M.-L.

The distribution of the fishes of the family Clinidae in southern Africa ee

Wuly 1970) —.. oe ae a =f a ; 135

PEnrRITH, M.-L.

The systematics of the fishes of the family Clinidae in South Africa ea

September 1969) .... ute a rs : I

NEW NAMES PROPOSED IN THIS VOLUME

SUB-GENERA

Blennophis Penrith, 1969 (Clinus, Clinidae), 20

Cirrhibarbis Penrith, 1969 (Clinus, Clinidae), 25

Climacoporus Penrith, 1969 (Clinus, Clinidae), 30

Clinus Penrith, 1969 (Clinus, Clinidae), 32

Fucomimus Penrith, 1969 (Pavoclinus, Clinidae), 72

Labroclinus Penrith, 1969 (Pavoclinus, Clinidae), 74

Muraenoclinus Penrith, 1969 (Clinus, Clinidae), 28

Pavoclinus Penrith, 1969 (Pavoclinus, Clinidae), 76

Smithichthys Penrith, 1969 (Pavoclinus, Clinidae), 85

SPECIES

confundens (Raja) Hulley, 1970, 203

dissimilis (Raja) Hulley, 1970, 199

ravidula (Raja) Hulley, 1970, 196

robertst (Raja) Hulley, 1970, 190

wallace: (Raja) Hulley, 1970, 210

, Ve: hi Ps OH cet as

baa ee Te) eet

)

2 -

why

j , \ ives

a i

‘ ey

/ '

‘|

‘o

‘

¥ ;

y 4

fi.

ve x ty ;

Ty oy a)

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 55 °+#£Band

September 1969 September

Part 1 Deel

THE SYSTEMATICS OF THE FISHES

OF THE FAMILY CLINIDAE

IN SOUTH AFRICA

MARY-LOUISE PENRITH

Cape Town Kaapstad

EATHSONT

DEC 9 1969

CiBRARIE2 ‘

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na beskikbaarheid

van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad

OUT OF PRINT/UIT DRUK

I, 2(1, 3, 5, 7-8), 3(1-2, 5, t—p.i.), 5(2, 5) 7-9);

6(1, t-—pi.), 7(1, 3), 8, g(1-2), 10(1-3),

11(1-2, 7, t.—p.i.), 21, 24(2), 27, 31(1-3), 38,

44(4).

Price of this part/Prys van hierdie deel

R5.00

Trustees of the South African Museum © _ Trustees van die Suid-Afrikaanse Museum

1969

Printed in South Africa by In Suid-Afrika gedruk deur

The Rustica Press, Pty., Ltd. Die Rustica-pers, Edms., Bpk.

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

THE SYSTEMATICS OF THE FISHES OF THE

FAMILY CLINIDAE IN SOUTH AFRICA

Mary-Louise PENRITH

South African Museum, Cape Town

(With 48 figures)

[MS. received 24 September 1967]

CONTENTS

PAGE

Introduction : , : I

Relationships and eeahention of the fermnily @inidae : : 2

Systematics of the South African Clinidae : ; ; 5 al

Origin of the South African Clinidae ; : ; : LOS

Summary . i 3 ’ : ; : Lys

ec eements ; : 2 ; : : : 5 LTR

References . ; : . s : , : ; 5 150g)

INTRODUCTION

The family Clinidae includes the klipfishes, which are characteristic of the

intertidal pools of the South African coast. Most of the common shore fish

species of the Cape Province belong to this family, which is widely distributed,

with endemic species in each region where it occurs. The relationships between

the clinids of different parts of the world, their zoogeography and possible

origins are important in indicating wider faunal relationships and paths along

which faunal dispersal may have taken place.

‘The present work is a taxonomic revision of the South African Clinidae.

The species were previously split into two groups. The species contained in

one of the groups did appear to be more closely related to each other than to

any of the species in the other group, but the second group seemed to be a

heterogeneous assemblage of species with little in common. Furthermore, the

characters used in defining numerous genera were such as sometimes to create

difficulties in the placing of undescribed species. The generic arrangement

proposed in the present work is based largely on external characters and may

change, pending more detailed osteological studies. It was felt, however, that

at this stage some rearrangement of previous classifications was necessary, and

that a publication containing figures and uniform descriptions of all the species

presently recognized, and all information known of the distribution of those

species, would form a useful basis for future work.

Ann. S. Afr. Mus. 55(1), 1969: 1-121, 48 figs.

24 ANNALS OF THE SOUTH AFRICAN MUSEUM

Since 1962 the South African Museum has carried out intensive collecting

of intertidal fishes from Mocamedes in southern Angola to Durban. This has

supplied large quantities of material of the common species so that the extent

of variability of characters in those species could be determined, and additional

material of the rarer species has been collected. This has enabled me to examine

fresh (as well as preserved) material of 28 of the 33 species of Clinidae described

below.

RELATIONSHIPS AND CLASSIFICATION OF THE FAMILY CLINIDAE

The Clinidae are a family of blennioid fishes. The blennioid fishes have

been defined by Gosline (1968) as perciform teleosts with the pelvic fins anterior

to the pectorals, and the number of dorsal and posterior soft anal rays cor-

responding exactly with the number of vertebrae between them. The caudal fin

is usually rounded.

The limits of the suborder Blennioidei have been subject to much change

by systematists during the last century. The definitive recent work of Gosline

(1968) has established the position of the Blennioidei among the perciform

fishes, and the limits of the suborder. Other recent reviews of the classification

of blennioid fishes have been published by Hubbs (1952), who revised the classi-

fication of tropical blennioids, and Makushok (1958), who revised the northern

blennioids. The families at present recognized as constituting the suborder

Blennioidei are listed in the comprehensive classification of fishes by Greenwood

et al. (1966), and also by Gosline (1968).

Family CLINIDAE

Diagnosis. Small, weakly-swimming blennioid fishes. Body usually covered with

cycloid scales. Suborbital bony ring weak and flexible. Jaw teeth conical and

fixed, usually in more than one series; vomerine teeth usually present, and

sometimes palatine teeth as well. Lateral line canals on head covered; lateral

line running in upper half of body in front to behind pectoral fin, then curving

down fairly sharply to mid-lateral position. Dorsal fin long, more or less con-

tinuous; first three spines may be separated to varying degree from rest of fin;

more spinous than soft elements. Anal fin long and continuous, with two spines

anteriorly. Pelvic fins with minute spine and two to four rays, jugular in posi-

tion. Dorsal, anal, pectoral, and pelvic rays unbranched. Caudal rays branched

or simple. Pseudobranchiae present. Gill membranes united, forming fold

across throat. Branchiostegal rays six to seven on either side. An upturned, hook-

like process on anterior border of cleithrum present or not.

Distribution. The Clinidae are distributed along both coasts of North, Central

and South America, the West Indies, West Africa, South and South West

Africa, Australia, New Zealand, the East Indies, Japan, the Philippines, and

the Mediterranean. Their greatest representation is in America, the West

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 3

Indies and the temperate waters of South Africa and Australia. They generally

occur in the intertidal zone of the shore or on coral reefs.

Relationships. The work of Hubbs (1952), Makushok (1958), Ford (1959), and

Gosline (1968) confirms the close relationship of the Clinidae, Tripterygidae,

and Blenniidae. Their relationship with other perciform fishes is as yet undeter-

mined (Gosline, 1968; Springer, 1968). The only clinid and blenniid fossils

known are from the Mediterranean Miocene, during which period the genera

known today were probably differentiated. Hubbs (1952) considered that the

Blenniidae diverged early from the other families, since the absence of scales

and the position of the eyes, well forward and high up in the head, indicate that

they have specialized along different lines from the Clinidae and the Triptery-

giidae, and presumably must have required a considerable amount of time to

perfect these specializations. The fossil family Pterygocephalidae, known only

from Eocene deposits, shows a mixture of percoid and blennioid features, and

evidently represents a very early line, but Hubbs (1952) stated that the Ptery-

gocephalidae show certain specializations that would remove them from the

direct line of descent of any of the modern blennioids.

Both the Clinidae and the Tripterygiidae have retained more generalized

features than the Blenntidae, but neither can be said with certainty to be more

primitive than the other. The usually compressed body of the Clinidae is more

typically percoid than the usually depressed body of the Tripterygiidae, but the

‘Tripterygiidae retain more branched fin rays than do the Clinidae, in which

branched caudal rays appear in only a few of the apparently most primitive

members of the family. The Clinidae and the Tripterygiidae have more in

common with each other than either family has with the Blenniidae, and in the

vast majority of earlier works have been treated as a single family. However,

Hubbs (1952), in his key to the families he included in a superfamily Blen-

nioidae, gave some of the features on which the families can be separated, and

they have been treated as separate by later workers, in particular Rosenblatt

(e.g. 1963).

A fairly strong link between the Clinidae and the Blenniidae appears to be

the family Chaenopsidae, which has long been of uncertain position in blen-

nioid classification. It has usually been either partially or completely submerged

in the Clinidae or the Blenniidae. Stephens (1963), revising the Chaenopsidae,

gave them family status. Springer (1964), reviewing Stephens’s work, was not

entirely satisfied that the characters on which Stephens raised the Chaenopsidae

to family status justified such an action. They have features in some ways

intermediate between those of the Clinidae and those of the Blenniidae; they

lack scales, but their affinities appear on the whole to lie more with the Clinidae

than with the Blenniidae (Bohlke, 1957). Springer (1966) continued to treat

them as a separate family in his discussion of Medusablennius chani, and they are

listed as a separate family in the classifications of Greenwood et al. (1966) and

Gosline (1968).

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE

i *6G61 ‘9gS61 ‘Sh61 “6261 ‘Aap ‘Po6r ‘tynzng f1961 ‘susydajg {E161 ‘sya

£16r ‘yywurg {661 ‘zi_nyog £Eg6r “9yTeMA 2 NE[Quesoy ‘o661 “eg ‘gS61 ‘uas[O ‘gggI ane (Sei Bi eee ae : ace} an

€sz61 ‘S161 ‘yoorngoy {4661 ‘qpeysreyy £gS6r ‘zitmasnepy fgh61 ‘puvsgapyipy fg&6r ‘orrapy fo061 “szaqyty $66g1 ‘uUeutey £661 “oS61 ‘Lb61

‘hh6r ‘zbGr ‘1€61 ‘szpmMog {0061 ‘ysiepy 2 UULUTIEAq fog61 ‘uosmeq ‘PS61 ‘JeMpyeD ‘oS61 ‘euapey ‘sg61 ‘ueng ‘g6gr1 ‘siogq ‘PE61

‘9c61 ‘ueA-20y, 2 eqoog £1961 ‘uemdeyD 2 WoONvag :syIOM SuIMOT[OF DY} WOIZ spJOIII [LUOTIIPpe YIIAA ‘Sp1ooat winasnyy ysutag pur

winasnyy uvorpy ynog pure £gS61 ‘GS6r “PS61 ‘sasuridg f6P6r ‘yMUg $SS61 “6E61 “SE61 43095 foggr ‘keorqoy ‘6061 ‘go061 ‘yOO]NDOW £9061

‘syivig 2 uepsof fq ‘vSS6x ‘2S61 ‘sqqnyy $6861 ‘ousapy Sgz61 ‘uyyiiy ‘ 1061-g6gr “uyjed ‘0961 ‘uewIIA-[Ty JO syIOM ay} UO ATUTeU paseg

Turua1307d Arr) WV

Sipas apjsaqut + Turdsa1uyyy WV

tprurpodoux §=© QVUIWOSTIQe'T

rurury) TuTUIpeIe Vv

Se) POLS TES jes

way O ipmpyeD =

TUIPOXATA] © . TurWOSLIqe’y]

pe ooS Ni x iprmosuqeyT =]

Gruvfang agit, aqunqny quunfqns aque I, _ agungns

SLL

Os SEL OG Sv ; O Gv O06 SEL

Fic. 1. World distribution of the tribes and subtribes of the family Clinidae.

ANNALS OF THE SOUTH AFRICAN MUSEUM 5

The relationships within the family Clinidae are not by any means fully

determined. Hubbs (1952) discussed the relationships of the South African

and American Clinidae, but did not include the Australasian, Japanese, and

Mediterranean clinids in his scheme of classification. He divided the Clinidae

into two subfamilies, the Labrisominae and the Clininae. The Clininae are

distinguished from the Labrisominae by the presence of an upturned hooklike

process on the anterior border of the cleithrum. This process is absent in the vast

majority of the Labrisominae, although in two labrisominid species, Mala-

coctenus erdmant Smith and M. aurolineatus Smith, the hook is usually present in

adults (Springer, 1958). The Clininae are further distinguished by the presence

of radii on all scale margins. The scales are usually smaller than those of the

Labrisominae (although exceptional genera without scales occur in both

subfamilies), and the scales are often embedded in the skin.

The Labrisominae occur mainly in the tropical waters of both the Pacific

and the Atlantic American coasts and the West Indies, one species occurs in

Japan, and two species occur in tropical West Africa. The Clininae occur

mainly in the temperate waters of the west coast of America, South Africa, and

Australia, with a few representatives in New Zealand, two in Argentina, and

one in Japan and the Philippines, and the East Indies.

Hubbs (1952) divided the Clininae into two tribes, the Clinidi and the

Myxodidi. It should be noted here that Hubbs used the suffix -idi for tribes

and the suffix -ini for subtribes, but in order to comply with the recommenda-

tion of the International Code for Zoological Nomenclature (1961), the process

is reversed here and henceforth in the present work -ini will denote tribes and

-idi will denote subtribes. Hubbs considered the Myxodini to be the more

primitive group, since they are all oviparous, and the male lacks a penis for

the transmission of sperm to the female. All the American Clininae belong to

this tribe. The Clinini are distinguished by the presence in the male of a fleshy

penis, and the fact that the young are born alive in at least all the species whose

breeding habits are definitely known. All the South African clinids belong to

the Clinini. Hubbs (1952) did not mention the Australian clinids and the

Japanese clinine in his system of classification. Milward (1967) recognized

three genera of Clinidae from Australia, of which two, Cristiceps and Petraites,

are represented in western Australia. Milward found all the western Australian

species, at least, to be live-bearing and to have an intromittent organ in the

male. The Australian Clinidae of the genera Clinus, Petraites, and Cristiceps

are here included in the tribe Clinini. The Japanese species, Petraites flammeus

(Jordan & Starks, 1906), is referable to the Clinini, as it has a well-marked hook

on the cleithrum (Jordan & Starks, 1906), and a delicate intromittent organ

in the male (Dr. V. G. Springer, pers. comm.). The classification of the family

Clinidae is shown in table 1. Their world distribution is shown in figure 1.

Two of the known species of Clinidae do not fit into Hubbs’s scheme of

classification. Clinitrachus argentatus (Risso), the Mediterranean clinid, must for

the time being be regarded as incertae sedis. It lacks a hooklike projection on the

anterior border of the cleithrum, and is oviparous, the male lacking a penis.

6 ANNALS OF THE SOUTH AFRICAN MUSEUM

TABLE 1. Classification of the family Clinidae (main arrangement after Hubbs,

1952, 1953a; naming of certain genera modified according to Springer, 1955,

and Bohlke & Springer, 1961; classification of the tribe Clinini new).

Family: CLINIDAE Regan, 1912

Subfamily: LABRISOMINAE Hubbs, 1952

Tribe: NEOCLININI Hubbs, 1953

NEOCLINus Girard, 1858

Tribe: MNIERPINI Hubbs, 1952

MNIERPES Jordan & Evermann, 1896;

DIALOMMUS Gilbert, 1891

Tribe: PARACLININI Hubbs, 1952

STATHMONOTUS Bean, 1885; PARACLINUS Mocquard, 1889;

EXERPES Jordan & Evermann, 1806.

Tribe: STARKSIINI Hubbs, 1952

STARKSIA Jordan & Evermann, 1896

Tribe: CRYPTOTREMINI Hubbs, 1952

ALLoCLINuUsS C.L. Hubbs, 1927; cRYPTOTREMA Gilbert, 1890.

Tribe: LABRISOMINI Hubbs, 1952

Subtribe: CALLICLINIDI Hubbs, 1952

AUCHENIONCHUS Gill, 1860; CALLICLINUS Gill, 1860;

MYERSICHTHYS Hubbs, 1952

Subtribe: LABRISOMIDI Hubbs, 1952

LABRISOMUS Swainson, 1839; MALACOCTENUS Gill, 1860

Subfamily: CLININAE Gill, 1885

Tribe: MYXODINI Hubbs, 1952

MYXODES Cuvier, 1829; HETEROSTICHUS Girard, 1854; GIBBONSIA

Cooper, 1864.

Tribe: CLININI Hubbs, 1952

Subtribe: CLINIDI Hubbs, 1952

cLinus Cuvier, 1817; PAVOCLINUS Smith, 1945; CLINOPORUS

Barnard, 1927; GYNUTOCLINUS Smith, 1945; BLENNIO-

cLinus Gill, 1860; ?PPETRAITEs Ogilby, 1886; cRISTICEPS

Valenciennes in Cuvier & Valenciennes, 1836.

Subtribe: XENOPOCLINIDI Hubbs, 1952

XENOPOCLINUS Smith, 1947; CANCELLOxus Smith, 1961.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 7

Guitel (1893) stated that although Valenciennes (zn Cuvier & Valenciennes,

1836) believed that further examination of this species would result in the male

being found to have a penis, his detailed examination of many specimens

revealed no such structure, the anal papilla always being small. He described

the breeding of this species in some detail; the female lays her eggs among the

fronds of algae before they are fertilized by the male. By the present definition

of the group, the lack of a hook on the cleithrum excludes it from either tribe

of the Clininae, yet it more closely resembles the Clininae than the Labri-

sominae, particularly in the nature of the small embedded scales. It is possible

that, during the long isolation of this species from other members of the Clininae,

the hook on the cleithrum may have been secondarily lost; or that this species

separated from the line leading to the Clininae before the hook had appeared.

Clinus nematopterus Giinther was described from the Sea of China in 1861.

Mr. A. Wheeler of the British Museum of Natural History has kindly examined

the type for me. It fits into the Clininae, having small, embedded scales, a hook

on the anterior border of the pectoral girdle, and a penis. It is known from a

single specimen. Dr. V. G. Springer (pers. comm.) suggests that it is probably

the same as the Japanese and Philippine clinids.

Of the three genera of Clinidae recognized in Australia (Milward, 1967),

Cristiceps is a well-defined genus occurring mainly or only in the Australian

region. Two Argentinian species of Clinidae, ezgenmanni Jordan, 1887 and

argentinus Berg, 1898, were placed in the genus Cristiceps by their original authors.

The descriptions are both very brief, but various points in them suggest that

these two South American species may not be referable to the same genus as the

Australian Cristiceps. However, they are provisionally treated as Clinini in

figure I. |

The Australian species of Clinus and Petraites that I have examined are

externally at least very similar to one another and to the South African species

of the genus Clinus. Both genera have been used for South African and Australian

Clinidae. The validity of the genus Petraites is doubtful. Ogilby (1886) created

the genus Petraites to contain certain Australian species. McCulloch (1908) said

that Petraites cannot be clearly separated from Clinus, as there are several

intermediate forms, but surprisingly went on to say that ‘no good purpose can

be served by uniting the two genera’. It seems unlikely that any good purpose

is served by retaining as separate two genera that are difficult to distinguish

from one another. However, a wider range of Australian material than has at

present been available to me would be necessary to determine whether the

genus Clinus should include both Australian and South African species, and

whether there is any justification for retaining the genus Petraztes. The Japanese

flammeus may belong to an Australian genus or may warrant a genus of its own.

SYSTEMATICS OF THE SouTH AFRICAN CLINIDAE

Maierial and methods. Most of this study is based on the clinid material in the

South African Museum collection. The collection was considerably augmented

8 ANNALS OF THE SOUTH AFRICAN MUSEUM

during my study by additions of fresh material collected on the coast from

Swakopmund in South West Africa to Durban. Rotenone poison and hand-nets

were used in the collection of samples. Details of material examined are given

under the description of each species.

Where a syntype series is present in the South African Museum collection,

a specimen in good condition and agreeing with the original description has

been selected and designated as a lectotype.

Measurements were made with needle-point dividers as follows: Standard

length (tip of snout to mid-base of caudal fin); head length (tip of snout to

upper extremity of opercle); depth at anal origin; greatest diameter of orbit;

length of upper jaw (inner edge of premaxilla to upper end of maxilla) ; length

of caudal peduncle (base of last anal ray to mid-base of caudal fin) ; least depth

of caudal peduncle; height of first dorsal spine; height of fourth dorsal spine;

length of portion of fourth dorsal spine free from fin membrane. Head length

was expressed as the number of times the head was contained in the standard

length, as was body depth, and orbit diameter as the number of times it was

contained in the head length, since these measurements were so expressed by

Fic. 2. Disposition of head pore series.

a = mandibular; b = preopercular; c = nasal; d = interorbital; e = suborbital;

f = postorbital; g = supraorbital; h = occipital.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 9

previous authors; other measurements were expressed as percentage of standard

length or of head length. Head length, body depth, orbit diameter, and caudal

peduncle measurements are given under each description; other measurements

are given where relevant. Counts were made of the number of spines and rays

in the dorsal, anal, pectoral and pelvic fins, by the usual methods; all the fin

rays in all South African Clinidae are undivided. Counts given in parentheses

are modal counts. Vertebral counts of small samples of each species were made

from X-ray photographs. Caudal vertebrae were taken to be those which had a

haemal spine. Counts of the gill-rakers of the outer series on the first arch are

given where possible. The disposition of the head pore series referred to is

shown in figure 2.

Abbreviations used in the lists of material given at the end of each species

description are $S.A.M. (South African Museum) and R.U.C. (Department of

Ichthyology, Rhodes University, Grahamstown).

The localities referred to in the distribution notes for each species are

shown in a map, figure 48, placed at the end of the paper for easy reference.

Subfamily CLININAE Gill, 1885

Diagnosis. Clinidae with an upturned, hook-like process on the anterior edge

of the cleithrum. Scales usually small and embedded, with radii on all margins.

Caudal rays always unbranched. No nuchal cirri.

Tribe CLININI Hubbs, 1952

Diagnosis. Clininae which bear their young alive. Males with a fleshy penis.

Always a cirrus above the anterior nostril. Body scales usually small and

embedded, exceptionally absent. Hook on cleithrum usually well developed,

rarely reduced or absent.

Subtribe CLINIDI Hubbs, 1952

Diagnosis. Clinini with head compressed, eyes lateral. Pelvic rays free from

membrane for greater part of length. Hook on anterior border of cleithrum

well developed.

ENstorical discussion of the classification of the South African Clinini

Various attempts have been made to classify the South African Clinini.

Valenciennes (zn Cuvier & Valenciennes, 1836), who described many of the

species, placed all the South African species that he described, as well as

Blennius superciliosus Linnaeus, in the genus Clinus Cuvier, with the exception

of capensis, for which Cuvier (1817) had erected the genus Cirrhibarbis, dis-

tinguished from Clinus by the presence of barbels on the chin and snout.

Swainson (1839) erected various genera of Clinidae, besides recognizing

Cuvier’s genera Clinus and Cirrhibarbis. Besides those two genera, three of

Swainson’s genera (Clinitrachus, Blennophis, and Labrisomus) contained South

IO ANNALS OF THE SOUTH AFRICAN MUSEUM

African species. He used body form and form of the dorsal fin (whether raised

anteriorly to form a crest or not), as well as dentition, in the definition of his

genera. He included two South African species, latipennis Valenciennes and

brachycephalus Valenciennes (= linearis Swainson) in the genus Labrisomus,

whose representatives are now included in a different subfamily from the South

African forms (table 1).

Gill (1860) discussed Swainson’s division of Cuvier’s genus Clinus. He

regarded Clinus and Clinitrachus as congeneric, as he considered the presence or

absence of a dorsal crest to be insufficient grounds for generic separation, but

Blennophis he considered valid. He removed from the genus Labrisomus most of

the species placed in it by Swainson (1839), including Jatipennis and brachy-

cephalus, erecting the genera Ophthalmolophus for the former and Blennioclinus for

the latter. He stated that, while latipennis and brachycephalus were related, the

former could be distinguished by the presence of supraorbital tentacles.

Swain (1882), in reviewing Swainson’s (1839) work, followed Gill (1860)

in synonymising Clinitrachus with Clinus, and considered Blennophis and Labrisomus

to be subgenera of Clinus. He made little attempt to sort out exactly which

species should be placed in each genus or subgenus, being chiefly concerned

with the validity or otherwise of Swainson’s proposed generic names.

The earlier South African workers, Gilchrist & Thompson (1908) and

Barnard (1927), disregarded most of the clinid genera then available, and

placed the majority of the South African species in the genus Clinus as defined

by Cuvier (1817). Gilchrist & Thompson (1908) placed one species, mentalis

Gilchrist & Thompson, in the genus Cristiceps Valenciennes on account of its

completely separate dorsal crest. Several of the other species, notably mus

Gilchrist & Thompson and laurentii Gilchrist & Thompson, also with completely

separate dorsal crests and otherwise rather similar to mentalis, were retained in

the genus Clinus. Cristiceps is an Australian genus with distinctive characters not

found in any of the South African species. In their introductory discussion,

Gilchrist & Thompson (1908: 98) also suggested that the generic name Clinoides

be used for the non-crested species, but did not use the name in their systematic

discussions.

Barnard (1927) transferred mentalis to another Australian genus, Petraites

Ogilby, and with it laurentsz, but Petraites is characterized by the presence of a

supraorbital tentacle, lacking in both mentalis and laurentii. Barnard (1927) also

described a new genus, Clinoporus, for Clinus biporosus Gilchrist & Thompson,

which lacks scales and has a peculiar lateral line very different from that of the

other clinids. In 1935 he described the species navalis and placed it in a new

genus Climacoporus, based on lateral line differences.

Smith (1945) revised the South African Clinidae and divided the group

into two subfamilies (Clininae and Myxodinae), distinguished only by the

presence or absence of a supraorbital tentacle. Among the 15 genera, seven

of them monospecific, which he recognized, nine were old genera: Clinus,

Cirrhibarbis, Blennophis, Ophthalmolophus, FPetraites, Blennioclinus, Clinoporus,

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE II

Climacoporus, and Myxodes Cuvier. Of these Myxodes had been used previously

only for South American species; the inclusion in it by Smith of the South

African live-bearing species fucorum Gilchrist & Thompson was later pointed

out to be an error (Hubbs, 1952). In addition to the above genera, Smith

described six new genera: Muraenoclinus, Blenniomimus, Fucomimus, Pavoclinus,

Labroclinus, and Gynutoclinus, four of which are monospecific.

Hubbs (1952), in a revision of the higher taxonomic categories of the

Clinidae, accepted Smith’s (1945) division of the South African species, but

reduced Smith’s subfamilies to the level of subtribes of the Clinini, which

includes all the South African species (see table 1). Hubbs named them the

‘Clinini’ and the ‘Blennioclinini’, any form of the name Myxodes being unaccept-

able for South African species. He pointed out that the South African species

fucorum, which Smith placed in the genus Myxodes on dental characters, could

not be a Myxodes on account of the penis in the male and the fact that the young

are born alive. Hubbs created a new genus, Smithichthys, for fucorum, revived

Swainson’s (1839) genus Clinitrachus for the species superciliosus Linnaeus and

robustus Gilchrist & Thompson, and used the name Clinus for the species placed

by Smith (1945) in the genus Ophthalmolophus. Apart from the changes in the

nomenclature of the genera and the lowering of the two subfamilies to the

status of subtribes, he did nothing to alter Smith’s system of classification.

Smith (1966) pointed out that the genus Clinitrachus is applicable to the

oviparous Mediterranean clinid argentatus Risso by type designation by Swainson

(1839), and created a new genus, Caboclinus, for superciliosus and robustus.

From the above it can be seen that the only definitive attempt to group the

South African Clinidae above the species level was Smith’s (1945) revision.

The major objection to this classification is his union of the non-tentacled

species into one group. The genera Blennioclinus, Gynutoclinus, and Clinoporus

have little in common with the other ‘Blennioclinini’’ or with each other apart

from the lack of a supraorbital tentacle in all except Gynutoclinus, which has a

supraorbital papilla.

A potentially useful contribution to South African clinid taxonomy was

made by Jackson (1950), but this unfortunately has remained unpublished, in

thesis form. Jackson examined the South African Clinidae and concluded that

to separate groups of species at a higher level than the generic one was incorrect.

He placed the species with a supraorbital tentacle in one genus, Clinus, having

the subgenera Clinus, Blennophis, Petraites, Ophthalmolophus, Cirrhibarbis and

Climacoporus. Most of the species lacking a supraorbital tentacle he placed in

another genus, Myxodes (this work was, of course, done prior to Hubbs’ 1952

study), having subgenera Myxodes, Blennioclinus and Labroclinus. He left Clino-

porus and Gynutoclinus, both monospecific and rare, with full generic value,

mainly (he stated) owing to lack of material for study.

Jackson’s scheme is one which, with several important modifications,

appears to me to give a more valid interpretation of the evolutionary history

of the group than any of the others. He distinguished his genera Clinus and

I2 ANNALS OF THE SOUTH AFRICAN MUSEUM

Myxodes on the presence or absence of a supraorbital tentacle and the relative

size of the mouth. Clinus was defined as having a tentacle over the eye and a

large mouth with powerful jaws, while Myxodes lacked a supraorbital tentacle

and had a small mouth with weak, narrow jaws.

I have made measurements of the upper jaw of samples of all the South

African species of Clinini, and the results are shown in a diagram (figure 3).

The statistical method used in preparing the diagram was that described by

Hubbs (1952); details of the method are given in the legend to the figure.

OR tse 1 Clinus anguillaris

2 Clinus striatus

3 Clinus capensis

4 Clinus dorsalis

Ra ee se 5 Clinus navalis

6 Clinus acuminatus

__ cme linus. agilis

el. 8 Clinus berrisfordi

Millis carey 9 Clinus brevicristatus

___ Clint cottoes aaa

Sue Lo ete 11 Clinus helenae

aieka Sieh, SETH ene 12 Clinus latipennis

BOOM NES] (ome sr Trade es,

13. Clinus heterodon

14 Clinus robustus

oO 2 15 Clinus superciliosus

16 Clinus taurus

Le come rs ocaieiis

‘Leaps 18 Clinus Woodi

AS ie MD IARG e cey ante tras

20 Pavoclinus graminis

21 Pavoclinus laurentii

2 ee 22 Pavoclinus litorafontis

| 23 Pavoclinus mentalis

ERA 24 Pavoclinus pavo

| 25 Pavoclinus profundus

26 Pavoclinus fucorum

_ oes 27 Blennioclinus brachycephalus

28 Blennioclinus stella

hay On | dee: 29 Clinoporus biporosus

sae es 30 Gynutoclinus rotundifrons

Se | Ee ee eee

20 25 30 35 40 45 50 55 60

°lo of head length

Fic. 3. Comparison of upper jaw lengths of the South African species of Clinini.

The method of representation is that used by Hubbs (1952) for comparative morphological

studies. In each diagram the base line represents the range of measurements recorded for the

sample, the upright line represents the mean, the open area represents the standard deviation

on either side of the mean, and the solid area represents two standard errors on either side of

the mean.

Samples of 25 specimens covering the full size range available were used where possible.

Where fewer than 25 specimens were available, all specimens were used. For species of which

fewer than 10 specimens were available, only the range and the mean are given. A full set of

data is given in Appendix A.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 13

All the species that have a fringed supraorbital tentacle (numbers 1-18

in figure 3) have an upper jaw averaging more than 40% of the head length

in the specimens examined (which usually covered a reasonably wide size

range). Most of the species included in Jackson’s Myxodes (numbers 19-28 in

figure 3) have the upper jaw averaging less than 40% of the head length. How-

ever, the difference between the lower limits of the ranges of one group and the

upper limits of the ranges of the other group is to my mind too small to be of

much use as a definitive generic character; the overlap would render deter-

mination difficult in the case of individual specimens. There is clearly a ten-

dency for the species with fringed supraorbital tentacles (genus Clinus) to have

larger mouths than most of the other species, and this is certainly of importance

in uniting the group, although it would apparently be of little use in placing

individuals in one group or the other.

Jackson suggested that ecological considerations should be taken into

account in attempting to classify the South African Clinidae. He stated that,

while his Clinus species frequented rocky crevices and stony pools, his Myxodes

species lived in seaweed. All the species that he included in Myxodes other than

those belonging to his subgenus Blennioclinus do live in seaweed, and are modified

in body form and colour for this habitat. However, Clinus brevicristatus Gilchrist

& Thompson, a rare tentacled clinid not different from the other species of

Clinus in body form or type of colour pattern, is usually taken in seaweed

together with the seaweed dwellers of Jackson’s genus Myxodes. The occurrence

of C’. brevicristatus in seaweed, as well as other tentacled clinids on occasions,

indicates that the habitat in which any individual specimen may have been

collected is not necessarily an indication of its relationships. However, the sug-

gestion by Inger (1958) that species united by resemblance and phylogenetic

considerations will usually be adapted for the same type of habitat is in general

true of the South African Clinidae.

There are relatively few characters that can be used to differentiate

between the South African members of the Clinini above the species level,

since this group is fairly homogeneous in most respects, and many features are

common to the group as a whole. I have divided the group into five genera,

somewhat unevenly, on external characters. Two are large genera, containing

between them 26 of the 30 recognized species of Clinidi. Subgenera have been

used in these two genera to denote pairs or groups of species which appear to be

more Closely related to each other than to the other species in the genus. Of the

other three genera, two are monospecific and rare, and the third contains two

species. The five genera are compared in table 2.

Key to the genera of the South African Clinidi

1. Body covered with small scales

Body without scales ..

_ ee.

we ae Clinoporus

2. A tentacle or minute papilla over the eye Be oe a ae a Se ea:

No tentacle or papilla over the eye un e e on a ue liens

14. ANNALS OF THE SOUTH AFRICAN MUSEUM

3g. A minute, simple papilla over the eye We Br ie uw a Gynutoclinus

A fringed or branched tentacle over the eye a ae Clinus

4. Posteriormost dorsal spine much shorter than first aoa soft eve a dees notch in the

margin of the dorsal fin before the soft rays at ie Blennioclinus

Posteriormost dorsal spine almost equal to first dorsal soft Sere ; no deep notch in the

margin of the dorsal fin before the soft rays ye Be Ae oe .. Pavoclinus

Genus CLINUS Cuvier, 1817

Clinus Cuvier, 1817: 173 (type-species Blennius acuminatus Bloch & Schneider, by subsequent

designation by Swainson, 1839: 75).

Cirrhibarbis Cuvier, 1817: 174 (type-species Cirrhibarbis capensis Valenciennes in Cuvier & Valen-

ciennes, by monotypy).

Blennophis Swainson, 1839: 75 (type-species Clinus anguillaris Valenciennes in Cuvier & Valen-

ciennes, by monotypy).

Ophthalmolophus Gill, 1860: 104. (type-species Clinus latipennis Valenciennes in Cuvier & Valen-

ciennes, by monotypy).

Climacoporus Barnard, 1935: 646 (type-species Climacoporus navalis Barnard, by monotypy).

Muraenoclinus Smith, 1945: 538 (type-species Clinus dorsalis Castelnau, by monotypy).

Blenniomimus Smith, 1945: 538 (type-species Clinus taurus Gilchrist & Thompson, by original

designation).

Caboclinus Smith, 1966: 73 (type-species Clinus robustus Gilchrist & Thompson, by original

designation).

Diagnosis. A distally flattened, at least bifid tentacle over eye. Lateral line

usually of double pores and/or single pores opening above and below line ante-

riorly; posterior to post-pectoral curve usually of short separate horizontal

tubes with a pore at either end. Intromittent organ of male usually consists of

fairly long basal portion with small tip retractile between one or two pairs of

fleshy lips. Body covered with small embedded cycloid scales. 'Two bands of

villiform teeth in each jaw, outer row larger. Vomer always toothed. Upper

jaw averages more than 40% head length. Body moderately robust. Caudal

peduncle short.

Note on the type-species of Clinus

The problem of the type-species of Clinus is discussed by Hubbs (1952) and

Smith (1966). Blennius superciliosus Linnaeus was generally regarded as the type-

species of the genus, and was so designated by Swain (1882), because it is

evident from the expanded description of the genus by Valenciennes (in Cuvier

& Valenciennes, 1836) that the description was based chiefly on specimens of

superciliosus. However, Swainson’s (1839) arbitrary designation of acuminatus

Bloch & Schneider as the type-species of Clinus must stand according to the

rules of the International Code of Zoological Nomenclature (1961), Article

~ 69(a) (iii).

Discussion

Most of the 18 species of the genus Clinus are small and occur almost

exclusively in the intertidal zone; a few extend into deeper water, and a few

reach a fairly large size (150-350 mm). Most species live in rock-pools among

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE

oejided uo uedO

poms

quosolg

jenbs ynoqy

yuosqyy

possorduroos ATy stp]

yisus] prsy

ov uvy} doy

poyivoysus ‘]yeus

dy {suo j1ed jeseg

‘pus yove 3e

orod YIM saqny

[eiuozi10y 1410Ys

jo ased s0119;s0d

¢Ayyetpour surusdo

g]8uIs sarod 1o112;0y

suiddry]

-IZAO OU ‘OINUITL

eyided

a] dunts ‘aynuryy

snurjs0jnudy)

ysny usdO

OUTS

yuosqy

jenbo ynoqy

yuosqy

QYI[-[99 “IsNqorYy,

Yysue]

pesy %oP uvy} ssavy

poyleoysua

you ‘{[eus

dij ‘suoy j1ed jeseg

ynoysno1y}

Ajse[NSIIIT MOTIQ

JO 2Aoqe sutusdo

osrey AIOA sarod

yuosqy

quosqy

snsogoury’)

ysny uedO

¥0U$

yuosqy

Ja}10Ys YONI,

juasolg

deIIpOp[

Yisu9s]

pesy %Sv—o&

peyieoysua

‘lopusys ‘suoT

dy fo;e19pour

0} JA0ys j1ed [eseg

‘puo

yous ye o10d YIM

soqn} [e,u0ZzII0Y

yro0ys jo yAed

Jo1193sod fajqnop

saiod 1o11ajuy

suiddey]

-I19A0 JOU ‘oINUTYT

yuosqy

snurjsouualg

ysny usdO

Suo]/91219poyy

quosoig

jenbo jnoqy

yuasqVy

possezdui09

ATYSIy 0} 91e19pojy

ysuey peoy %ovr

ULY} SSI] SASVIBAV

sutpnsjoid ‘osiey]

di) {9}v19pout

0} [[ews Javed [eseg

pus yore jz

910d yiIM soqny

jeuozts0y 110ys

jo yaed s0119}sod

‘MOTOq pue

aAoqe 10 AT[erpaut

suruodo ‘apduts

saiod 10119}0y

surddey]

-I9AO U91JO

‘QINUIU IO [[eVWS

yuosqy

snuns00vg

seyided

uo Jo ysny uedOQ

quosqe Ayyensp)

jenbs jnoqy

juosoid us1IO

DI]-[99 sous

*jsNnqOl 0} 9}e19pPOP[

yisus] peoy %ov

UCY} IIOUL SISVIZAY

jou JO poy}esysua

‘pews ATTensn

dy {suo 0}

oye1spow jied [eseg

puo yous 7e

o10d YIM saqn}y

[e}uoz110y 1210ys

jo Aypensn y1ed

Jotsaisod £ Mojaq

Jo aAoqez Sutusdo

‘Q]SUIS JI 9,]qQnop

uayjo sarod 10119juVy

suiddey|

-I2AO SOUIT]IULOS

‘QINUIUL IO [[BUIS

poyouviq pue

peusepy ATTeIsid

EEO)

‘IpIUl[D jo vous uvoUpy YINOG oY Jo UOsIIeduIO- *% ATAV,],

uo sozod snonyy

gpunped [epneyn

soueIquioul

ul] UI SMOpUIM,

yusoNysuvL I,

Awl [ess1op 1sIy

Joutds [esiop 4sv'T

sourds [esiop

UO TIIID JO s19}sNI[D

w10} Apog

Mef sodd¢

uvsd1O

7U9}} TWO TUT

oui [e191e'T

saTZoG

9joR1ua1

ye1qiovsdng

s1ojoeIeYyy)

16 ANNALS OF THE SOUTH AFRICAN MUSEUM

stones and in crevices, although many species frequently hunt in sea-weeds

as well, and one species habitually occurs in beds of the green alga Caulerpa;

none, however, are particularly modified or coloured for weed-dwelling (see

Pavoclinus). They are well camouflaged and able to change colour and pattern

almost instantaneously, although there is generally a basic pattern of pigmenta-

tion that may last for a considerable time after preservation. This pattern may

be of blotches or cross-bars, and there is almost invariably an ocellate spot on

the shoulder, anterior fin spines, or opercle. The species are entirely carni-

vorous, but the diet is very varied, both between and within species. There is a

tendency in many species to develop a notch in the membrane joining the third

and fourth dorsal spines. The first three dorsal spines are often elevated to form

a crest; this is considered to be a primitive feature in the family Clinidae (Hubbs,

19530). In C. superciliosus, which has a very well developed crest, the height of

the crest shows sexual dimorphism. The other species show little or no sexual

dimorphism other than the development of a penis in the male.

An examination of the gonads of many individuals of several species of

this genus suggests that, while parasitism of the ovaries by trematodes occurs

to a varying degree, the testes of the males are normally not affected.

Jackson (1950) divided Clinus into subgenera which, with one exception,

corresponded with Smith’s (1945) genera. In attempting to group the 18 species

contained in the genus Clinus as defined above, one is confronted with the

difficulty of deciding on a method that will reflect the natural relationships of

the species within the genus. This is a problem, because there is an indication

of development along at least two major and several minor lines within the

genus. Most of the species form a more or less generalized group in which each

minor line of development is represented by a range rather than a cluster of

species.

It is proposed to use five subgenera, one of which contains a large number

of generalized species. The other four are believed to represent particular lines

of development and are easily defined; they appear to be more closely related to

each other than to the more generalized species.

Clinus navalis, C’. dorsalis, C’. anguillaris, C. striatus, and C. capensis, previously

distributed among the genera Climacoporus, Muraenoclinus, Blennophis, and

Cirrhibarbis are all rather elongate, eel-like species. Smith (1945) separated them

from each other generically for the following reasons: (1) Clinus navalis has only

one dorsal soft ray, and the lateral line pores are double for most of the length of

the line, sometimes the whole length; it was placed in the genus Climacoporus

by Barnard (1935) and retained in that genus by Smith (1945). (2) C. dorsalis

has only one dorsal soft ray, and the lateral line is of the usual type posterior to

the post-pectoral curve (i.e. of short horizontal tubes with a pore at either end) ;

Smith (1945) placed it alone in the genus Muraenoclinus. (3) C. anguillans and

C. striatus have 2-4 dorsal soft rays and scaly cheeks, and were placed in the

genus Blennophis. (4) C’. capensis has barbels on the chin and snout, and scaly

cheeks, and was placed in the genus Cirrhibarbis.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 17

C. anguillaris and C’.. striatus are the two most eel-like species. They have

small scales on the cheeks as well as on the bases of the dorsal, anal and caudal

fins. They are very alike and form a distinct pair of species; it is proposed to

show their relationship by placing them together in a subgenus Blennophis.

C. capensis cannot be confused with any other species, on account of the

barbels on the chin and snout. In some families, e.g. Brotulidae, this feature is

considered sufficient grounds for generic distinction. Springer (1954) included

one species with a chin barbel in the clinid genus Paraclinus, whose members are

otherwise without facial barbels. In view of the fact that C. capensis resembles

Fic. 4. Distribution of C. anguillaris (open circles), C. striatus (closed circles), C. capensis (closed

triangles), C. dorsalis (open squares), C. navalis (closed squares), C. acuminatus (open triangles).

the other species of Clinus in all other respects, having the same type of supra-

orbital tentacle, lateral line, and intromittent organ, it seems preferable to

treat this species as a member of the genus Clinus. Since, however, it is set apart |

from the others by the facial barbels, it may be placed in a subgenus Cirrhi-

barbis.

The lateral line of C. navalis appears to be of the same type as that found in

the other species of Clinus, but the pores continue double for a varying distance

behind the post-pectoral curve, often along the whole length of the lateral line.

18 ANNALS OF THE SOUTH AFRICAN MUSEUM

As in C. dorsalis, there is one soft ray in the dorsal fin, and the cheeks are naked,

as in the majority of species. The supraorbital tentacle is similar in C. navalis and

C. dorsalis. It is very difficult to distinguish between juvenile specimens of these

two species except on fin counts. However, since C. dorsalis differs from C. navalis

in being a far more elongate species, with 19 precaudal vertebrae as opposed to

15 in C. navalis, and has virtually all the sensory pores of the head and lateral line

system single, while in C. navalis most of the sensory pores are double, and finally,

in view of the unusual form of the intromittent organ in C. navalis, Muraeno-

clinus and Climacoporus are maintained as subgenera.

The remaining 13 species constitute the subgenus Clinus.

C. taurus and C. cottoides were placed in a genus Blenniomimus by Smith

(1945) because both species have a bony ridge over the eye, which makes the

interorbital appear concave instead of flat. The ridge is by no means striking

in small specimens of C. cottoides; and it is slightly developed in C’. latipennis and

C. helenae, which Smith placed in the genus Ophthalmolophus. C. taurus and C. cot-

toides should therefore not be separated generically or even subgenerically from

the other species, since they, particularly C. taurus, represent the extreme of the

range of development of a particular feature. The bony ridge is furthermore a

feature which in some species becomes more pronounced with age; in C. heterodon

there is no sign of a meee) in small specimens, but it may be very pronounced in

large ones.

C. superciliosus, C.. Pinan C. brevicristatus and C. woodi were previously

separated from the other tentacled species because the first three dorsal spines

were higher than the succeeding ones, forming a crest. C. superciliosus and C.

robustus were placed in the genus Clinus in a restricted sense, because they have

a notch in the membrane between the third and fourth dorsal spines less than

50% of the height of the fourth spine. The two species were subsequently

transferred first to Clinitrachus, by Hubbs (1952), and then to Caboclinus, by

Smith (1966). |

C. brevicristatus and C. wood: were placed in the genus Petraites. Species of

this genus have a notch in the fin membrane between the third and fourth

dorsal spines more than 50% of the length of the fourth spine (McCulloch,

1908), although Smith (1945) stated that the notch is complete in Petraites,

and the membrane from the third spine barely reaches the base of the fourth

spine. McCulloch (1908) stated that the two genera Clinus and Petraites are

barely distinguishable from one another, and this is certainly true of the South

African species involved, since C. superciliosus resembles C’. woodi more closely

than it does C. robustus, and C. brevicristatus is as close to forms such as C. agilis,

C. venustris, and C. cottoides as it is to C. woodt. In C. superciliosus the depth of the

notch varies from o—100% of the fourth spine, but is generally 40-60% (see

table 3). C. robustus has an extremely shallow notch. In C. woodi the membrane

from the third spine barely reaches the base of the fourth spine. In C. brevi-

cristatus the notch varies from 25-100% of the fourth spine but is usually 50-

75%, and only rarely 100%. The crest is high and triangular in C. woodi and

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 19

particularly in mature male specimens of C’. superciliosus, and is very low, scarcely

elevated, in C. robustus and C. brevicristatus. Among the other species, C. agilis

and C’.. taurus have a notch in the membrane between the third and fourth

spines but the first three spines are not elevated, and C. venustris often has the

second and third dorsal spines somewhat elevated. There is a range of crest

and notch development in this genus, and it is thus impossible to base generic

distinctions on these features. It is often difficult to decide whether individual

specimens of C. robustus and C. brevicristatus have a crest at all, while in other

individuals it is quite marked. It is also possible that the crest has been developed

or reduced separately in the various species, and that its presence does not

indicate close affinity. There are no other features to suggest a closer affinity

amongst all the crested species than amongst crested and non-crested species.

The five species previously contained in the genus Ophthalmolophus were

apparently placed there because they lacked any of the structures that distin-

guished the other genera, although in fact indications of those structures appear

in the various members of this group of species. As noted above, C. latipennis

and C. helenae show signs of a bony ridge over the eye, and may in that feature

represent stages approaching the condition seen in C. taurus and C’. cottoides, so

that the classification should in some way relate these species. C. agilis, with a

notch between the third and fourth dorsal spines, is too similar to C. brevi-

cristatus on the one hand, with a deeper notch and the first three dorsal spines

slightly elevated, and to species such as C’. acuminatus on the other hand, with no

notch or crest, to be separated even subgenerically from either. C. venustris shows

rudiments of a crest (the second and third dorsal spines are sometimes higher

than the others), but this is not sufficiently marked in every individual for the

species to be defined as having a crest. Apart from these variations of the dorsal

fin, the 13 species placed in the subgenus Clinus are united by so many common

features and are often so hard to distinguish from one another that any attempt

at separation into subgenera would be unwarranted.

I believe that while the subgenera used do represent different lines of

evolution within the genus Clinus, the species contained in the genus as a whole

form a well-defined group that are more closely related to each other than to

any other clinid species. However, should another worker wish to separate the

eel-like subgenera Blennophis, Cirrhibarbis, Muraenoclinus and Climacoporus from

the subgenus Clinus and place them in a separate genus or separate genera, there

would be little reason for serious objection to such an arrangement.

Distribution of the genus Clinus

In South Africa Clinus occurs in temperate waters. Three species (C. navalis,

C’. helenae, C’. woodi) are confined to the waters north-east of Cape Agulhas; most

of the species are restricted to the coast south of the Kei River (figs 4, 11, 19).

Only one species (C. woodi) is known to extend north of Durban to Inhambane,

where it is rare, and it does not reach the tropical waters north of that point.

On the north-west coast, two species (C. superciliosus and C. heterodon) extend

20 ANNALS OF THE SOUTH AFRICAN MUSEUM

into the tropics north of Walvis Bay, but as the coast in that region is under the

influence of the cold Benguela current, conditions of water temperature and

fauna are not markedly tropical. The genus Clinus includes all the common species

of the Cape coast west of the Cape Peninsula. Twelve of the 18 species have been

recorded from the waters west of Cape Point, only one of these (C. cottoides)

being common north-east of Algoa Bay. The other five species are fairly rare

throughout the South African coastal region. It seems that this genus has its

greatest distribution in the colder regions of the South African coast.

Range. Rocky Point (18°59’S., 12°29’E) (South West Africa) to Inhambane

(23°51'S., 35°29 E) (Mozambique).

Key to the subgenera of Clinus

1. Scales on the cheeks .. op be be oa i a in i ve 22

No scales on the cheeks a weal, 23

2. Barbels on chin and snout Be 3 ae ia MM re Cirrhibarbis

(1 species, capensis)

No barbels on chin or snout = 5 3 He a ae Blennophis

(2 species)

3. One dorsal soft ray a

More than one dorsal soft ra ad a x5 sae ae + ye Clinus

(13 species)

4. More than 40 dorsal spines; most sensory pores single .. Be se: Muraenoclinus

(1 species, dorsalis)

Fewer than 40 dorsal spines; most sensory pores double I ia Climacoporus

(1 species, navalis)

Subgenus Blennophis Swainson, 1839

Blennophis Swainson, 1839: 75 (type-species Clinus anguillaris Valenciennes in Cuvier & Valen-

ciennes)

Diagnosis. Elongate eel-like clinids with small scales on cheek and dorsal, anal

and caudal fin bases, as well as over general body surface. Body robust. Dorsal

fin low, even. No cirri at dorsal spine tips. No barbels on chin or snout. Dorsal

soft rays 2-4.

Remarks. ‘Two species, one common only west of Cape Point, the other rare

throughout its range, but encountered more frequently east of Cape Point.

Both reach a large size and occur infratidally as well as in intertidal rock

pools.

Key to the species of Blennophis

1. Inner pelvic ray stout, equal in length to other two; dorsal spines 46 or more, anal rays

33 ormore .. me oe ae ve oe aM ou ae Be Clinus

(Blennophis) anguillaris

Inner (third) pelvic ray minute or absent; dorsal spines 45 or fewer, anal rays 31 or

fewer .. les », Be bd be, ait vy! Clinus (Blennophis) striatus

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE

(0) I v 8

0) 6 L vs

Oo L v lf

O O Co) O

oO O oO 0)

re gI 61 LS

%0o1 °% 66-06 % 69-08 %6L—oL 69-09

‘yojou jo yjdop aejnonsed v Suravy o[dures yore wio.1y suoutoads Jo Jaquinu juesoidedr ssoquinu 9y, J,

°/, 6G—0S

°/, 6b-ob

°/, 66-08

(0)

% 6-02

% 61-01

% 6-0

snanpy snui’)

sys snuy)

$njojs14910a1q

HO)

sngsngos snuary

¢poom snurj’y

snsoyiosaqns

STD)

sotvedg

‘(sueiquisut wos; daay ouTds [esIop YyIINO}J Jo % se possoidxa) sourds [esIop YANO} puv pAry} UsIMJaq sUBIqUISUT UT YI}0U JO yidoq °& a1aVv,],

22 ANNALS OF THE SOUTH AFRICAN MUSEUM

Clinus (Blennophis) anguillaris Valenciennes in

Cuvier & Valenciennes, 1836

(Fig. 5)

Blennius rubescens Lichtenstein, 1823: 117 (nomen oblitum).

Clinus anguillaris Valenciennes in Cuvier & Valenciennes, 1836: 390. Gilchrist & Thompson,

1908: 133. Barnard, 1927: 862.

Blennophis anguillaris : Swainson, 1839: 75. Smith, 1945: 539, 1949: 352, pl. 78 fig. 980.

Blennophis rubescens : Smith, 1962: 40.

Description. D. XLVI-L (XLVIII-L) 2-4 (3-4); A. IT 33-37; P. 13-14 (13);

V. 13; C. 13. Gill-rakers in outer series on first arch 2-3 + 5-7. Vertebrae

18 + 38-40. Dorsal fin long, low, even. No clusters of cirri on dorsal spine tips.

Pectoral fin rounded. Inner pelvic ray well developed, stout, equal to others.

Caudal peduncle very short, length 15-5-25°% head length, depth 28-5-35:5%

head length. Caudal fin short, sub-truncate.

Body elongate, robust, eel-like, with small embedded scales extending on to

caudal, dorsal, and anal fin bases, upper edge of opercle, and cheeks. Depth

6—7-75 in standard length. Head bluntly wedge-shaped, 4:75-5:75 in standard

length. Eye 3°25-5 in head, relative size of eye decreasing with increase in

(ii) lateral

(i) ventral

GS)

(iii) anterior

(c)

Steeda phil rg eww WU, OT CNY Oy

aT Onan Wang, : ~

(b) (d) en

Fic. 5. Clinus (Blennophis) anguillaris: (a) Lateral view, male, 99 mm, S.A.M. 24206; (b) Head

pore system; (c) Intromittent organ of male; (d) Lateral line.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 23

standard length of fish. Orbital tentacle prominent, with flat stalk and fine

fringe of cirri at tip. Cirrus on anterior nostril elongate, simple, with serrated

edges. Upper jaw 38-42°5% head length. Lips thick. Vomer toothed. Sensory

pores of head single in nasal, interorbital, occipital and supraorbital series,

mainly single in mandibular and preopercular series, first pore of mandibular

and last pore of preopercular series double; postorbital and suborbital series

mostly of paired pores (fig. 5 (b)).

Lateral line mostly of vertical pairs of pores in front to post-pectoral curve,

then of short separate horizontal tubes with pore at either end (fig. 5 (d)).

Intromittent organ of male with moderate basal portion and single pair of oval

dorso-lateral lips (fig. 5 (c)).

Colouring. Variable, reddish to dark maroon, or pink with reddish blotches.

Fins red- or white-tipped. Sometimes light-edged, dark ocellus on shoulder,

always present in juveniles. Often two dark radiating bars from eye across

cheek. Snout white in large, dark-red specimens. Plain buff or pinkish preserved

in alcohol. :

Location of type material. Muséum National d’Histoire Naturelle, Paris.

Material examined. 36 specimens, 65-278 mm in standard length. 2 from ‘Rade

du Cap’, Paris Museum Cat. Nos. A. 2079 & 6341 (syntypes); 11 from Liide-

ritzbucht, S.A.M. 24208; 1 from Gert du Toit Bay, southern Namaqualand,

S.A.M. 24019; 4 from Doring Bay, southern Namaqualand, S$.A.M. 24020;

2 from Lambert’s Bay, S.A.M. 23887, S.A.M. 24010; 1 from Kommetjie, Cape

Peninsula, S.A.M. 23874; 9 from St. James, False Bay, S.A.M. 10526, S.A.M.

13721; 2 from Dalebrook, False Bay, S.A.M. 21538, S.A.M. 21539; 1 from

Still Bay, S.A.M. 25059; 1 from East London, 8.A.M. 19745; 2, no locality,

S.A.M. 21708, S.A.M. 21709.

Remarks. The scales on the head in this and the related species, Clinus (Blen-

nophis) striatus, might possibly be primitive characters, as also the degree of

development of the inner pelvic ray of C. anguillaris.

Nomenclature. Smith (1962) pointed out that Lichtenstein’s (1823) description

of Blennius rubescens from the Cape of Good Hope fits Clinus anguillaris. However,

the name anguillaris has been used for the present species since 1836, while

Lichtenstein’s name had not been used for well over a hundred years until

Smith revived it in 1962. According to the International Code of Zoological

Nomenclature (1961), Article 23(b), unless the International Commission

should decide otherwise, ‘a name that has remained unused as a senior synonym

in the primary zoological literature for more than fifty years is considered to be

a forgotten name (nomen oblitum)’. Forgotten names may not be used unless

the Commission so directs. Smith (1962) did not place the matter before the

Commission. The name anguillaris Valenciennes in Cuvier & Valenciennes,

1836, is thus retained for this species.

Distribution (fig. 4). The known range is Liideritzbucht (26°38’S., 15°10’E)

(South West Africa) to East London (33°00’S., 27°54’E). Fairly common on

24 ANNALS OF THE SOUTH AFRICAN MUSEUM

the west coast as far south as Lambert’s Bay, becoming rather rare east of Cape

Point. Small specimens occur under stones in shallow pools near the top of the

shore, but larger specimens occur only at the bottom of the shore, and extend

into the sublittoral fringe. The species blends perfectly with the stipes of kelp

among which it occurs.

Clinus (Blennophis) striatus Gilchrist & ‘Thompson, 1908

(Fig. 6)

Clinus striatus Gilchrist & Thompson, 1908: 134. Barnard, 1927: 862.

Blennophis striatus: Smith, 1945: 539, 1949: 353, fig. 981.

Description. D. XL-XLV 2-4 (2-3); A. II 28-31 (30-31); P. 12-13 (13);

V. I 2-3; C. 13. Gillrakers in outer series on first arch 2-3 + 7-8. Vertebrae

17 + 34-35. Dorsal fin long, low, even. No clusters of cirri at tips of dorsal

spines. Pectoral fin rounded. Inner pelvic ray reduced or absent. Caudal ped-

uncle very short, length 10-20% head length, depth 23-27% head length.

Caudal fin short, subtruncate.

Body elongate, robust, eel-like, with small embedded scales extending on

to caudal, dorsal and anal fin bases, upper edge of opercle, and cheeks. Depth

6-8-25 in standard length. Head bluntly wedge-shaped, 4:5-—5:25 in standard

length. Eye 3:5—5 in head. Supraorbital tentacle prominent, with flat stalk and

fringe of fine cirri at tip. Cirrus on anterior nostril elongate, spatulate, and

slightly lobed ventrally. Upper jaw 37-44°5% head length. Lips thick. Vomer

toothed. Sensory pores of head single in nasal and interorbital series; mainly

single in mandibular and preopercular series, first mandibular and last pre-

opercular pore paired. Pores of other series double. Multiple pore in postorbital

series (fig. 6(b)).

Lateral line as in Clinus anguillaris, mainly of vertical pairs of pores opening

above and below line in front to post-pectoral curve, then of short, separate

horizontal tubes with pore at either end (fig. 6(d)). Intromittent organ of male

with moderate basal portion and single pair of narrow, crescentic dorso-lateral

lips (fig. 6(c)).

Colouring. Pink with brownish-red to dark-red mottling, fins usually reddish.

Adults occasionally have faint white longitudinal stripes. Light-edged dark

ocellus on shoulder. Two dark radiating lines from eye across cheek. Juveniles

white with longitudinal black stripes. Uniform buff to pinkish preserved in

alcohol.

Location of type material. South African Museum, Cape Town.

Lectotype. Female, 142 mm standard length, Simon’s Bay, False Bay, S.A.M.

25230.

Material examined. 11 specimens, 42-145°5 mm in standard length. 1 from

Saldanha Bay, S.A.M. 23878; 3 from Simon’s Bay, False Bay, S.A.M. 25239

(lectotype), S.A.M. 10527 (paralectotypes), 2 from Kalk Bay, False Bay,

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 25

S.A.M. 17844, S.A.M. 18086; 3 from Strandfontein, False Bay, S.A.M. 24240;

2, no locality, S.A.M. 21710, S.A.M. 21706.

Remarks. This species is very similar to Clinus anguillaris, differing from it exter-

nally only in the fin counts, in the degree of reduction of the inner pelvic ray

and in the greater number of double sensory pores on the head.

Distribution (fig. 4). The known range is from Saldanha Bay (33°00’S., 17°56’E)

to East London (33°00’S., 27°54’E). The juveniles occur high up on the shore,

but larger specimens are taken at the bottom of the shore and infratidally,

among kelp stipes. Rare.

(ili) anterior

(i)ventral (ii) lateral

(c)

SDH AT ALAA AR LR

(b) SUI LEAVE UT CNS

7 8... ay Ca *

(d)

Fic. 6. Clinus (Blennophis) striatus: (a) Lateral view, female, 142 mm, S.A.M. 25239, lectotype;

(b) Head pore system; (c) Intromittent organ of male; (d) Lateral line.

Subgenus Cirrhibarbis Cuvier, 1817

Cirrhibarbis Cuvier, 1817: 174 (type-species Cirrhibarbis capensis Valenciennes in Cuvier & Valen-

ciennes).

Diagnosis: Elongate clinids. Snout pointed, with three stout, simple cirri at tip;

chin with eight stout, simple cirri. Small scales on cheek, dorsal, caudal and

anal fin bases, as well as over general body surface. Dorsal fin low, even, with

clusters of fine cirri at tips of spines.

One species, Clinus (Cirribarbis) capensis (Valenciennes in Cuvier & Valen-

ciennes).

26 ANNALS OF THE SOUTH AFRICAN MUSEUM

Clinus (Cirrhibarbis) capensis (Valenciennes in Cuvier & Valenciennes, 1836)

(Fig. 7)

Cirrhibarbis capensis Valenciennes in Cuvier & Valenciennes, 1836: 405. Swainson, 1839: 277

fig. 86. Smith, 1945: 539, 1949: 352, pl. 77 fig. 977, and fig. 977.

Clinus capensis: Giinther, 1861: 269. Gilchrist & Thompson, 1908: 131. Barnard, 1927: 853.

Description. D. XXXVII-XLIV (XXXVIII-XLI) 5-9 (6); A. II 26-34

(26-29); P. 12-14 (13); V. I 2-3; C. 13. Gill-rakers in outer series on first arch

2-3 + 5-7. Vertebrae 18-21 + 32-37. Dorsal fin low, even. Clusters of 3-5

fine cirri at tips of dorsal spines. Pectoral fin rounded. Inner pelvic ray reduced,

with free tip very short, or absent. Caudal peduncle short, length 20-30%

head length, depth 18-26% head length. Caudal fin subtruncate.

Body elongate, slightly compressed, covered with small embedded scales

extending on to dorsal, caudal and anal fin bases, and cheeks. Depth 5-6 in

standard length, occasionally greater. Head narrow, pointed, 3:5-4°5 in

standard length. Snout with three stout, simple cirri at tip; chin with eight

stout, simple cirri clustered about jaw symphysis. Supraorbital tentacle pro-

minent, with broad flat stalk and fringe of fine cirri at tip. Eye 3°5-5:5 in head.

Cirrus on anterior nostril large, spatulate, with about four shallow, flat lobes

at tip. Upper jaw 37°5-47°5% head length. Lips thick. Vomer toothed. Most of

sensory pores of head double; multiple pores occur mainly in postorbital and

occipital series (fig. 7(b)).

Lateral line of vertical pairs of pores in front to post-pectoral curve, then

of short separate horizontal tubes with pore at each end (fig. 7(d)). Intromittent

organ of male with long basal portion; pair of large confluent dorsal lips and

pair of rounded ventral lips ensheathing tip (fig. 7(c)).

Colouring. Variable, pinkish, grey, or greenish mottled. Light-edged dark

ocellus on shoulder. Fins and facial cirri usually red. A juvenile specimen

from Lambert’s Bay and two specimens from the west coast of the Cape Penin-

sula, one juvenile and one adult, were milky white with longitudinal black

stripes and red fins and facial cirri. Buff to pinkish preserved in alcohol, shoulder

spot may remain distinct.

Location of type material. Muséum National d’Histoire Naturelle, Paris.

Material examined. 97 specimens, 28-246 mm in standard length. 2 from Cape

of Good Hope, Paris Museum Cat. Nos. 710, 6334 (syntypes) ; 1 from Lambert’s

Bay, S.A.M. 24237; 1 from Kommetjie, Cape Peninsula, S.A.M. 23926; 1 from

Froggy Pond, False Bay, S.A.M. 23930; 2 from Kalk Bay, False Bay, S.A.M.

18224, S.A.M. 18276; 11 from Dalebrook, False Bay, 19/2/1965 and 18/4/1965,

S.A.M. not catalogued; 14 from St. James, False Bay, S.A.M. 10530, S.A.M.

12018, S.A.M. 23583, and 1962, S.A.M. not catalogued; 29 from Strandfontein,

False Bay, S.A.M. 23975, S.A.M. 24236; 1 from Gordon’s Bay, False Bay,

S.A.M. 23290; 4 from Onrust River Mouth, S.A.M. 24238: to from Still Bay,

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 27

S.A.M. 25098, S.A.M. 24665; 4 from Mossel Bay, S.A.M. 23929; 4 from Port

Elizabeth, S.A.M. 23928; 2 from Igoda Mouth, East London, $.A.M. 25099;

8 from East London, S.A.M. 23927; 3 from Gonubie River Mouth, East

London, S.A.M. 25100.

Remarks. This species can be distinguished at once from all other African clinid

species by the facial cirri, but it is considered similar enough to the other

tentacled species to be included in the genus Clinus. It is most similar to the

eel-like forms of the subgenus Blennophis in body form, type of habitat, and the

very short caudal peduncle, as well as the scaly cheeks; the clusters of cirri at

the tips of the dorsal spines, the intromittent organ with its two pairs of lips and

the high dorsal soft ray count are more similar to the conditions found in some

of the species of the subgenus Clinus.

Distribution (fig. 4). The known range is Lambert’s Bay (32°04’S., 18°20’E) to

East London (33°00’S., 17°56’E). The source of Smith’s (1949) Port Nolloth

record for this species is not known, and in spite of extensive collecting there

its occurrence has not been confirmed. It is rare west of Cape Point, but is

fairly common from False Bay eastwards. It lives in shallow pools when

small, but large adults appear to be more or less confined to the sublittoral

fringe.

(i)ventral (ji) jateral (ii anterior

(c)

(d) x

(b) aon

Fic. 7. Clinus (Cirrhibarbis) capensis: (a) Lateral view, male, 76 mm, S.A.M. 24236; (b) Head

pore system; (c) Intromittent organ of male; (d) Lateral line.

28 ANNALS OF THE SOUTH AFRICAN MUSEUM

Subgenus Muraenoclinus Smith, 1945

Muraenoclinus Smith, 1945: 538 (type-species Clinus dorsalis Castelnau in Bleeker).

Diagnosis. Elongate clinids with one dorsal soft ray. No scales on cheeks. Dorsal

fin low, even. No clusters of cirri at tips of dorsal spines. No barbels on chin or

snout. Most sensory pores single. Posterior part of lateral line of short separate

horizontal tubes with pore at either end.

One species, small and intertidal; fairly common on the west coast, less so

eastwards to the south coast of Natal.

Clinus (Muraenoclinus) dorsalis Castelnau in Bleeker, 1860

(Fig. 8)

Clinus dorsalis Castelnau in Bleeker, 1860: 54. Castelnau, 1861: 54. Gilchrist & Thompson,

1908: 132. Barnard, 1927: 86.

Muraenoclinus dorsalis: Smith, 1945: 538, 1949: 352, pl. 78 fig. 979.

Description. D. XLI-XLVIII (XLV-—XLVI) 1; A. IT 25-31 (26-29); P. 11-13

(12); VI 2-9; C. 13. Gill-rakers in outer series on first arch 2 + 0-3. Vertebrae

19 + 32-35. Dorsal fin low, even. No clusters of cirri at dorsal spine tips.

Pectoral fin rounded. Inner pelvic ray reduced, often absent, if present minute.

Caudal peduncle very short, length 16-5-20% head length, depth 26-33-5%

head length. Caudal fin short, subtruncate.

Body elongate, eel-like, slightly compressed, covered with small embedded

scales not extending on to fin bases or head. Depth 5-5-7. Head narrowly

wedge-shaped, snout acutely pointed, head 3-5-5 in standard length. Eye

3-4°5 in head. Supraorbital tentacle prominent, with flat stalk and terminal

fringe of fine cirri. Cirrus on anterior nostril narrow at base, tip broadly spatu-

late, with deeply serrated edge. Upper jaw 34°5-46% head length. Lips thick.

Vomer toothed. Sensory pores of head single in most series, last pore of pre-

opercular series and most pores of suborbital series double (fig. 8(b)).

Lateral line of single pores opening above and below line in front to post-

pectoral curve, then of short separate horizontal tubes with pore at either end

(fig. 8(d)). Intromittent organ of male with long basal portion; pair of dorso-

lateral lips and pair of rounded ventral lips ensheathing tip (fig. 8(c)).

Colouring. Plain dark green or red, or light greenish or brownish mottled with

darker greens, browns, and mauves. Mottled specimens have ocellate spot on

shoulder. Broad white stripe from base of first dorsal spine to tip of snout. Fins

same as general body colour. Juveniles usually uniform light brown or black

with white stripe from dorsal origin to tip of snout. Plain yellowish buff pre-

served in alcohol.

Location of type material. Unknown.

Material examined. 138 specimens, 23—71°5 mm in standard length. 26 from

Liideritzbucht, S.A.M. 10535, S.A.M. 24207; 17 from Port Nolloth, S.A.M.

24222; 22 from Lambert’s Bay, S.A.M. 23931; 3 from Saldanha Bay, S.A.M.

a —— —

— a

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 29

17913; 6 from Sea Point, S.A.M. 23932 and 17/2/1965, S.A.M. not catalogued;

14 from Kalk Bay, S.A.M. 10536; 2 from Dalebrook, False Bay, S.A.M. 24239

and 18/4/1965, S.A.M. not catalogued; 20 from St. James, False Bay, 1962,

S.A.M. 23935; 1 from False Bay, S.A.M. 23934; 5 from Onrust River Mouth,

S.A.M. 24258; 1 from Hermanus, 8.A.M. 23933; 9 from Die Dam, Bredasdorp

district, S.A.M. 24505; 3 from Still Bay, S.A.M. 18077 and April, 1965, S.A.M.

not catalogued; 4 from Great Fish Point and East London, $.A.M. 18091; 4

from East London, $.A.M. 23936.

Remarks. Clinus dorsalis most closely resembles Clinus navalis. Jackson (1950)

placed it in his subgenus Blennophis with Clinus anguillaris and Clinus striatus, but

in view of the naked cheeks and the single soft dorsal ray it seems more appro-

priate to place it in a separate subgenus.

Note on the author of Clinus dorsalis

Gilchrist & Thompson (1908), Thompson (1918), Barnard (1927) and

Smith (1945, 1949) gave Bleeker (1860) as the author of this species, but this

appears to be incorrect. The first record of the species was published by Bleeker

(1860), but it was listed as Clinus dorsalis Castelnau although Castelnau’s

description of the species was published a year later, in 1861. In a brief note

(i) ventral (ii) lateral (ili) anterior

(C)

(d) ‘.

(b) re

Fic. 8. Clinus (Muraenoclinus) dorsalis: (a) Lateral view, female, 64 mm, S.A.M. 24207; (b) Head

pore system; (c) Intromittent organ of male; (d) Lateral line.

30 ANNALS OF THE SOUTH AFRICAN MUSEUM

published in 1860 Castelnau discussed the fish fauna of South Africa and

stated that he would be producing a memoir on the ichthyological fauna of

southern Africa, in which 69 new species would be described. It seems probable

that Bleeker saw and quoted from Castelnau’s list of species prior to its publica-

tion in 1861. He should not be regarded as the author of the species, since he

acknowledged Castelnau as the author, and the full description of the species

was written and published by Castelnau.

Distribution (fig. 4). The known range of this species is Liideritzbucht (26°38'S.,

15°10’E) (South West Africa) to the Natal South Coast (Smith, 1947a). Smith

(1937) gave the north-western limit of this species as Walvis Bay, stating that he

took this record from Barnard (1927). However, Barnard (1927) gave the

north-western limit as Angra Pequena (Liideritzbucht), so that the 1937

Walvis Bay record was due to an error in transcription. The occurrence of the

species north of Liideritzbucht has not been confirmed.

Usually found among stones in shallow pools high on the shore, but occurs

at all levels of the intertidal zone. This species appears to be more common west

of Cape Agulhas.

Subgenus Climacoporus Barnard, 1935

Climacoporus Barnard, 1935: 646 (type-species Climacoporus navalis Barnard).

Nemacoclinus Smith, 1937: 195 (referring to Climacoporus navalis Barnard).

Diagnosis. Fairly elongate clinids with one dorsal soft ray. No scales on cheeks.

Dorsal fin low, even. No clusters of cirri at tips of dorsal spines. No barbels on

chin or snout. Most sensory pores double; vertical pairs of pores in posterior

part of lateral line.

One species, small and intertidal, occurring only east of Cape Agulhas and

rare.

Clinus (Climacoporus) navalis (Barnard, 1935)

(Fig. 9)

Climacoporus navalis Barnard, 1935: 646. Smith, 1945: 538, 1949: 352, pl. 78 fig. 978.

Nemacoclinus navalis: Smith, 1937: 195.

Description. D. XXXV-XXXVIII 1; A. II 23; P. 12; V. I 2-3; C. 13. Gill-

rakers in outer series on first arch 1-2 + 5. Vertebrae 15 + 27-29. Dorsal fin

low, even. No clusters of cirri at tips of dorsal spines. Pectoral fin rounded.

Inner (third) ray of pelvic fin reduced or absent. Caudal peduncle short, length

25-33°5% head length, depth 25~-33-5°% head length. Caudal fin subtruncate.

- Body elongate, slightly compressed, covered with small, embedded, non-

imbricating scales extending on to dorsal, caudal, and anal fin bases but not

head. Depth 5-6:25. Head 4-5-25 in standard length, snout rounded. Eye

3-4 in head. Supraorbital tentacle prominent, with flat stalk and fringe of fine

cirri terminally. Cirrus on anterior nostril small, flap-like, slightly emarginate

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 31

below apex. Upper jaw 38-44°5% head length. Lips thick. Vomer toothed.

Sensory pores of head single in nasal, interorbital, and mandibular series,

mainly double in all other series (fig. 9(b)).

Lateral line of vertical pairs of pores throughout; narrows and may

become obscure towards tail (fig. 9(d)). Intromittent organ of male long, slender

with apparently no definition into tip, basal part and lips (fig. 9(c)).

Colouring. Yellow-brown to greenish, mottled and barred irregularly with

darker green or brown; head green or reddish brown; belly cream. Light-edged

dark ocellus on shoulder. Plain buff preserved in alcohol.

Location of type material. South African Museum, Cape Town.

Material examined. 10 specimens, 18-57 mm in standard length. 1 from ship’s

bottom, Simonstown, False Bay, $.A.M. 18287 (holotype); 2 from Still Bay,

S.A.M. 25101; 1 from Port Elizabeth, S.A.M. 23889; 1 from Boknes Point,

R.U.C.; 2 from Great Fish Point, R.U.C.; 1 from Igoda Mouth, East London,

S.A.M. 25102; 2 from Xora Mouth, R.U.C.

Remarks. The arrangement of the double pores of the lateral line of this species

is exactly the same as the arrangement seen in the anterior part of the lateral

line of other species that have double pores, such as Clinus capensis. In view of

the many features which Clinus navalis has in common with other species of the

(b) ) SEH

Fic. 9. Clinus (Climacoporus) navalis: (a) Lateral view, male, 55°5 mm, S.A.M. 18287, holotype;

(b) Head pore system; (c) Intromittent organ of male; (d) Lateral line.

32 ANNALS OF THE SOUTH AFRICAN MUSEUM

genus Clinus, particularly Clinus dorsalis, I believe that to set this species apart

on account of the lateral line would obscure its relationships with the rest of the

group. The intromittent organ, however, is very peculiar. In the few male

specimens available it was very strongly erect; the finer structure may have been

obscured to some extent. There is some resemblance to the erect organ of

C.. brevicristatus, which shows but little definition at the tip.

Distribution (fig. 4). The known range is Still Bay (34°23’S., 21°24’) to Port

St. Johns (31°37’S., 29°32’E). The first specimen described was found among

barnacles on a ship’s bottom at Simonstown, but the species does not normally

occur in the region of the Cape Peninsula.

Subgenus Clinus Cuvier, 1817

Clinus Cuvier, 1817: 173 (type-species Blennius acuminatus Bloch & Schneider).

Clinitrachus Swainson, 1839: 75 (in part) ; (type-species Blennius variabilis Rafinesque).

Ophthalmolophus Gill, 1860: 104 (type-species Clinus latipennis Valenciennes).

Blenniomimus Smith, 1945: 538 (type-species Clinus taurus Gilchrist & Thompson).

Caboclinus Smith, 1966: 73 (type-species Clinus robustus Gilchrist & 'Thompson).

Diagnosis. Body not elongate. Head naked. No barbels on chin or snout. More

than one dorsal soft ray.

Thirteen species, occurring mainly among rocks in intertidal pools. ‘These

species are difficult to identify from a key, as they are extremely variable. Two

keys are given, but species should always be compared with the description as

well.

Key to the species of the subgenus Clinus

1. Anterior three dorsal spines, or at least the second and third dorsal gaia elevated, higher

than fourth dorsal spine .. be Ae SR

Anterior three dorsal spines saat to or Sno oa fourth dora oe aye 6

2. Dorsal soft rays three or less. Second spine highest; no notch in membrane peareee Gard

and fourth dorsal spines .. : Clinus (Clinus) venustris

Dorsal soft rays four or more; a said of ee epin in membrane between third and

fourth dorsal spines

3. Pectoral rays 15 or more .. if ae or ye Gir (Clinus) eee

Pectoral rays 14 or fewer a ae ate

4. Crest high, triangular; dorsal fin anecnaves. over gerd sian e Se Clinus (Clinus) woodi

Crest low, rectangular to rounded; dorsal fin originates over hind margin of preopercle 5

5. Lips with vertical corrugations; diosall soft rays nine or more .. Clinus (Clinus) robustus

Lips smooth; dorsal soft rays seven or fewer .. : Clinus (Clinus) brevicristatus

6. Anotch of varying depth in membrane between third ariel fourth dorsal spines area y 17,

No notch in membrane between third and fourth dorsal spines a ae pe

7. Interorbital concave, strong bony ridges over eyes; occiput with deep transverse groove

Clinus (Clinus) taurus

Interorbital flat, no ridges over eyes; no deep occipital groove Clinus (Clinus) agilis

8. Interorbital strongly concave, a nee over eye; ee groove deep; dorsal soft

rays six orfewer .. Pe: Clinus (Clinus) cottoides

Interorbital flat, or only ce nrly concave; Mecca anaes if present shallow ST OL0)

g. Dorsal soft rays 8-9. . a ae ae LA 0) Clinus (Clinus) latipennis

Dorsal soft rays seven or fewer .. ae ae he ee ae a, on |) LO

10. Dorsal soft rays 2-3.. b 6 oy! tb: i.e i ae Clinus (Clinus) venustris

Dorsal soft rays 4-7. . ite wh: ah un br a Be At Se

If.

13.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 33

Dorsal spines 34-37... Sy = #5 a ai a a a's My 12

Dorsal spines 33 or fewer .. a HoEMe aiey BILD

Mucous pores on head open on aepiles: : enon Say Poumcied Clinus (Clinus) helenae

Mucous pores on head open flush with surface; snout more or less wedge-shaped

Clinus (Clinus) berrisfordi

First dorsal spine shorter than fourth spine; pectoral rays usually 12; intromittent organ

of male broad and flattened towards tip a a Cling (Clinus) acuminatus

First dorsal spine equal to or very slightly longer ea fourth; pectoral rays 13; intromittent

organ of male slender, cylindrical ay oe if .. Clinus (Clinus) heterodon

Alternative key to the species of the subgenus Clinus

No. of

soft dorsal

rays

2-3 1. Dorsal spines 39-41.. se wf ne 33 Sf Clinus (Clinus) venustris

Dorsal spines 34-37. . ae a: ve a fe Clinus (Clinus) agilis

4. 1. Interorbital concave 2s sig sits ae ae Clinus (C..) cottoides

Interorbital flat Sy r, 2 SNL 2

2. First three dorsal spines Meo en Powis: tuna a ee Mt <#f Et ie}

First three dorsal spines lower than fourth, not forming a crest

3. Pectoral rays 15 or more. .. eae ae 52 Clinus (C. ) auton oars

Pectoral rays 14 or fewer Clinus (C..) brevicristatus

4. A notch in membrane between fhird aid eee dlvsell spines; dorsal spines 34-37

Clinus (C.) agilis

No notch in membrane between third and fourth dorsal spines; dorsal spines 33 or

fewer ; ‘ she Cli us ee acuminatus

BR) ial) Pirst three dessa spines iene es fourth, eae acrest .. ei ae

First three dorsal spines lower than or equal to fourth, not forming a ies r

2. Pectoral rays 15 or more .. B. oe an ms Clinus (C.) superciliosus

Pectoral rays 14 or fewer 3 ase toe

3. Crest high, triangular, pHieiaedes 3 over dl eee of ae .. Clinus (C.) wood

Crest low, originates over hind margin of preopercle .. Clinus (C.) brevicristatus

4. A notch in membrane between third and fourth dorsal spines Clinus (C.) taurus

No notch in membrane between third and fourth dorsal spines ae Ses

5. A large round black spot on the opercle . Be ete Clinus (C.) cottoides

No large black spot on opercle ; an alg Sesh nO

6. First dorsal spine lower than fourth : an oe eis

First dorsal spine equal to or a little longer than fourth Clinus (C.) berrisfordi

7. Dorsal spines 33 or fewer; pectoral rays 12 aye a Clinus (C.) acuminatus

Dorsal spines 34 or more; pectoral rays 13 HM ae Clinus (C..) helenae

6 1. First three dorsal spines high, forming a crest .. i si ae ate) he

First three dorsal spines not forming a crest Ye te sis sph, eee

2. Pectoral rays 15 or more .. te Me ie Clinus (C.) superciliosus

Pectoral rays 14 or fewer : Clinus (C.) brevicristatus

3. A notch in membrane between ined and foun Ansell spines Clinus (C.) taurus

No notch in membrane between third and fourth dorsal spines od BAe Nae

4. A large round black spot on the opercle .. As Le Clinus (C.) cottoides

No large black spot on opercle.. a i - si ae aia

5. First dorsal spine lower than fourth nF fe ee +wli/36

First dorsal spine equal to or a little longer fae Dae a be

6. Pectoral rays 12; dorsal spines 33 or fewer =i ad Clinus (C.) acuminatus

Pectoral rays 13; dorsal spines 34. or more 22 ¥- Clinus (C.) helenae

7s» Rectoralraysi12, . .. ae Na on ei ey: Clinus (C.) berrisfordi

Pectoral rays13... - ae Clinus (C.) heterodon

7 1. First three dorsal spines Sen, faerie a ee ¥ a ie = stan Peo

First three dorsal spines not forming a crest ah fe a aes Aah" WS

34. ANNALS OF THE SOUTH AFRICAN MUSEUM

No. of

soft dorsal

rays

7 2. Pectoral rays 15 or more .. ar a Ne ie Clinus (C.) superciliosus

Pectoral rays 14 or fewer : .. Clinus (C.) brevicristatus

3. First dorsal spine lower than fate eel see 12; intromittent organ of male

broad and flattened toward tip... nee Clinus (C.) acuminatus

First dorsal spine equal to or a little ene Gram fourth; pectoral rays 13; intro-

mittent organ of male slender, cylindrical a be Clinus (C.) heterodon

8 1. First three dorsal spines high, forming acrest .. we Clinus (C.) superciliosus

First three dorsal spines not forming a crest ae of Clinus (C.) latipennis

g 1. Lips with vertical mee ae Ne is A Clinus (C.) robustus

Lips smooth : he a8 Ns si sive. 2

2. First three dorsal spines fee fugarie a eres a ay Clinus (C.) superciliosus

First three dorsal spines not forming a crest ae Si Clinus (C.) latipennis

10 1. Lips with vertical corrugations .. re ae a Clinus (C.) robustus

Lips smooth ve na x sat Se ve Clinus (C.) superciliosus

11-14 ne a she iA se Ae = a3 Clinus (C.) robustus

Clinus (Clinus) acuminatus (Bloch & Schneider, 1801)

(Fig. 10)

Blennius acuminatus Bloch & Schneider, 1801: 169.

Clinus acuminatus: Cuvier, 1817: 175. Valenciennes in Cuvier & Valenciennes, 1836: 370. Swain-

son, 1839: 75. Gilchrist & Thompson, 1908: 124. Barnard, 1927: 859. Hubbs, 1952: 106.

Smith, 1966: 73. Penrith 1967: 43, figs 1, 4(a).

Ophthalmolophus acuminatus: Smith, 1945: 542, 1949: 355, pl. 79 fig. gor.

Description. D. XX XI-XXXIII (XXXII-XXXITI) 5-7; A. II 20-24 (21-22);

P. 12-13 (12); V. I 2-3; C. 13. Gill-rakers in outer series on first arch 1-2 +

4-6. Vertebrae 16 + 27-29. Dorsal fin low, even. Clusters of 2-3 cirri at tips of

dorsal spines for about half length of dorsal fin. Pectoral fin rounded. Inner

pelvic ray reduced or absent. Caudal peduncle short, length 20°5-33°5% head

length, depth 20-35% head length. Caudal fin subtruncate.

Body slightly compressed, covered with small scales extending on to

dorsal fin base; caudal and anal fin bases and head naked. Depth 4-5-6 in

standard length. Head large, 3-2-4 in standard length, snout wedge-shaped,

angle of profile acute. Eye 2-75-5 in head. Supraorbital tentacle prominent,

with short, flattened stalk terminating in several short, simple branches. Cirrus

on anterior nostril short, flattened, trilobed. Upper jaw 34°5-50% head length,

increasing with size of fish. Lips moderate. Vomer toothed. Sensory pores of

all head series single excepting last pore of preopercular series and one pore of

suborbital series, which are double (fig. 1o0(b)).

Lateral line usually of 2-4 vertical pairs of pores in front above opercle,

then of mainly single pores opening above and below the line to post-pectoral

curve; after that of short separate horizontal tubes with pore at either end

(fig. 10(d)). Intromittent organ of male large, flattened, and spade-shaped,

with moderately long, thick basal portion, pair of small dorsal lips, and large,

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 35

broad pair of ventro-lateral lips ensheathing tip (fig. 10(c)).

Colouring. Light green with broad, well-defined dark brown cross-bars, or with

black speckling in sparse large spots with mosaic-like pattern. Tips of all fins

and of supraorbital tentacles reddish orange. Belly silvery white. Dark ocellate

spot on shoulder, and two dark radiating bands from eye across cheek. Small

juveniles white with well-defined black cross-bars. Pattern retained for several

years after preservation.

Location of type material. Unknown.

Material examined. 182 specimens, 19-113°5 mm in standard length. 17 from

Liideritzbucht, S.A.M. 10544, S.A.M. 24206; 38 from McDougall’s Bay, Port

(ii) anterior

(i) lateral

Gii) ventral

(d) a. & (b)

Fic. 10. Clinus (Clinus) acuminatus: (a) Lateral view, female, 73 mm, S.A.M. 24255; (b) Head

pore system; (c) Intromittent organ of male; (d) Lateral line.

Nolloth, S.A.M. 24220; 12 from Port Nolloth, 12/7/1960, R.U.C.; 15 from

Lambert’s Bay, S.A.M. 21477, S.A.M. 23922 and 17/1/1964, S.A.M. not cata-

logued; 18 from Lambert’s Bay, R.U.C.; 2 from Lambert’s Bay, University of

Cape Town; 2 from Schaapen Island, Langebaan, $.A.M. 23925; 8 from Sea

Point, S.A.M. 22840, S.A.M. 23211; 3 from Kommetijie, S.A.M. 10541; 3 from

Froggy Pond, False Bay, S.A.M. 23924; 28 from St. James, False Bay, S.A.M.

12023 and 1962, S.A.M. not catalogued; 2 from St. James, False Bay, R.U.C.;

21 from False Bay, S.A.M. 10542; 2 from Cape Peninsula, University of Cape

36 ANNALS OF THE SOUTH AFRICAN MUSEUM

Town; 4 from Onrust River Mouth, 6/11/1964, S.A.M. not catalogued;

2 from Die Dam, Bredasdorp District, S.A.M. 24508; 5, no locality,

1. OIG

Remarks. Clinus acuminatus is similar on the one hand to Clinus agilis, having a

flat interorbital, and on the other hand to species such as Clinus cottoides, Clinus

heterodon, Clinus latipennis, and Clinus helenae, which lack a notch in the membrane

between the third and fourth dorsal spines. It is also very similar to Clinus

berrisfordt. Small specimens are rather similar in appearance to Clinus dorsalis,

and occupy the same type of habitat. Smith’s (1931) suggestion that Clinus

agilis may be found to be a subspecies of Clinus acuminatus is not supported by

a comparison of the two species, which are not sufficiently alike in appearance

to be confused even in the field. Furthermore, they occur over the same geo-

graphical range and often in the same pools.

: "Ny

ted: ‘ ° ‘ y , rhe. \ ’ . - / oe

on ey

ah eg

‘: ay : , ry B ~. ee

Oh, ae F / id ; i BETES: @ A

FN ae a Beh

CCE eng ‘

Soler... Be oletee LigiezeieR SO

OSs 0-8 &

pA @

Fic. 11. Distribution of C. agilis (open circles), C. berrisfordi (closed squares), C. brevicristatus

(open squares), C. cottoides (closed circles), C. helenae (closed triangles), C. heterodon (open triangles).

Distribution (fig. 4). The known range is Liideritzbucht (26°28’S., 15°10’E)

(South West Africa) to Cape Agulhas (34°50’S., 20°00’E). It is more common

west of Cape Point. This species occurs in shallow, barren pools under stones

and ledges at the top of the intertidal region of the shore.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 37

Clinus (Clinus) agilis Smith, 1931

(Fig. 12)

Clinus agilis Smith, 1931: 154, pl. 16, 1966: 73.

Ophthalmolophus agilis: Smith, 1945: 54, 1949: 355, pl. 77 fig. 988.

Ophthalmolophus anne Smith, 194.7): 733, fig. 2, 1949: 355, pl. 77 fig. 980.

Clinus anne: Smith, 1966: 73.

Description. D. XX XII-XXXVITI (XXXV-XXXVIT) 2-4 (3); A. IT 20-25

(22-23); P. 13-15; V. I 2-3; C. 13. Gillrakers in outer series on first arch 1-2 +

3-5. Vertebrae 16 (rarely 17) + 27-30. Dorsal fin low, first three spines not

elevated to form crest, but notch in membrane between third and fourth dorsal

spines, varying in depth, rarely absent (see table 3). Clusters of 3-4 cirri at tips

of dorsal spines for about half length of fin. Pectoral fin rounded. Inner pelvic

ray, if present, minute; usually absent. Caudal peduncle short, length 23-31 %

head length, depth 25:5-29% head length. Caudal fin subtruncate.

Body slightly compressed, covered with small scales not extending on to

fin bases or head. Depth 4°5—5-25 in standard length. Head 3-5—4:5 in standard

length, snout rounded. Eye 3—4:25 in head. Supraorbital tentacle prominent,

with flat stalk dividing into several short simple branches at tip. Cirrus on

anterior nostril long, narrow, pointed. Upper jaw 40-50% head length. Lips

fairly thick. Vomer toothed. Sensory pores of head single in nasal and

interorbital series, mainly single in mandibular and preopercular series; pores

in remaining series double or multiple (fig. 12(b)).

Lateral line of vertical pairs of pores and single pores opening above and

below line in front to post-pectoral curve, then of short separate horizontal

tubes with pore at either end (fig. 12(d)). Intromittent organ of male with long

basal portion; tip completely ensheathed by 2 pairs of confluent lips. Pair of

rounded fleshy lobes within fold, between which is minute tip (fig. 12(c)).

Colouring. Ground colour usually dark greyish or greenish, with about 7 dis-

tinct dark red and green cross-bars. Occasional specimens have pinkish or

milky ground colour with darker pink cross-bars, or vermilion ground colour

with dark brown cross-bars. Fins dark, irregularly barred; characteristic

translucent area in membrane joining dorsal soft rays. Belly silvery white to

greyish. Sprinkling of white spots over body sometimes. Head with irregular

lacy patterns in body colours, or plain dark grey. Light-edged dark ocellate

spot on shoulder usually. Juveniles white with well-defined red cross-bars.

Pattern remains for several years after preservation.

Location of type material. Department of Ichthyology, Rhodes University,

Grahamstown. One paratype in South African Museum.

Material examined. 672 specimens, 17-113 mm in standard length. 191 from

Liideritzbucht, S.A.M. 24205; 1 from Sinclair’s Island, R.U.C.; 3 from Orange

River Mouth, 5 fms, 8.A.M. 23964; 246 from Port Nolloth, S.A.M. 24214;

1 from Hondeklip Bay, March 1965, S.A.M. not catalogued; 6 from Gert du

38 ANNALS OF THE SOUTH AFRICAN MUSEUM

Toit Bay and 4 from Doring Bay, southern Namaqualand, S.A.M. 24232;

5 from Lambert’s Bay, R.U.C. (syntypes of anne); 108 from Lambert’s Bay,

S.A.M. 23919; 1 from Lambert’s Bay, R.U.C.; 13 from Steenberg’s Cove,

St. Helena Bay, 1962, S.A.M. not catalogued; 3 from Saldanha Bay, S.A.M.

22072, S.A.M. 23921; 5 from Langebaan, S.A.M. 21475, S.A.M. 21479;

38 from Sea Point, S.A.M. 23212, S.A.M. 23920, and 17/2/1965, S.A.M. not

catalogued; 6 from Froggy Pond, False Bay, S.A.M. 23906; 2 from Dalebrook,

False Bay, 19/2/1965, S.A.M. not catalogued; 1 from Knysna, S.A.M. 18278

(paratype); 38 from Knysna, $.A.M. 24233.

(iii) anterior

tn.

Nee

(i) ventral (ii) lateral

(c)

eae meee

in eee

(d) in

‘ee

(b)

Fic. 12. Clinus (Clinus) agilis: (a) Lateral view, female, 113 mm, Hondeklip Bay, S.A.M.; (b) Head

pore system; (c) Intromittent organ of male; (d) Lateral line.

Remarks. It has been believed for some time at the South African Museum that

this species and Ophthalmolophus anne Smith, 1947, are probably conspecific.

In attempting to identify large samples from the west coast, one might arrive

at either species in Smith’s (1949) key, or at both species for different specimens

of the same sample. Smith (1931) created agilis for seven specimens taken from

Knysna estuary, ranging in size from 55-65 mm. In 1947 he created anne for

five specimens, 24-50 mm, from the west coast, mainly Lambert’s Bay. He

differentiated anne from agilis on the following points: (1) number of dorsal

spines (34-35 for agilis, 36-37 for anne); (2) eye size (3-4 in head in agilis,

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 39

4°2-4'5 in head in anne); (3) snout length (greater in anne, no figures given) ;

(4) pectoral length (greater in anne, no figures given); (5) pelvic length (less in

anne, no figures given). Since so few specimens were examined, and the size

range of the samples did not overlap, the four points relating to body propor-

tions may probably be ignored. In the Clinidae body proportions have been

found to vary considerably with size, so that comparable size ranges are essential

for the comparison of body proportions of different species; furthermore, the

ranges found for body proportions tend to be very wide in the Clinidae, so that

other features are usually more important for delimiting species. In any case,

the orbit diameter of 19 specimens of Clinus agilis, all from the west coast and

therefore falling in Smith’s species anne, was measured and was found to -be

2-6—-4°25 in head (size range of specimens 31-93 mm in standard length), very

close to the range found by Smith (1931) for his Knysna specimens. 15 speci-

mens from Knysna, ranging from 31-59 mm in standard length, showed an

orbit diameter range 2°5-3°5 (see table 4). As far as the dorsal spine count is

TABLE 4. Comparison of orbit diameter of Clinus agilis from

south-east and south-west coasts of the Cape.

East Coast (agilis Smith) West Coast (anne Smith)

Std. length Eye in head Std. length Eye in head

(mm) (mm)

31 3-3 31 3-3

31 a. 7 31 2.8

32 3-3 31 3+3

33 3-3 33 3-3

35) 3-0 35 2.9

3 3-0 37 2.7

39 2.5 38 2.6

39 3-0 39 3-1

42 259 42 3.0

48 2.9 47 3.0

49 3-4 49-5 3-3

5! 3-5 5! 3-4

52 | 3-5 53 3-5

55 2.9 55 3-1

59 3-3 60 3.4

67 3-5

70 3.6

82 4.25

93 3.8

concerned, samples from the west coast showed a range of 33-38 dorsal spines,

which covers the range of 34-35 of agilis. Furthermore, samples from different

areas along the coast showed variation in the average number of dorsal spines

for each population. This is shown in a histogram (fig. 13). The variation from

one population to another is an interesting indication of the lack of genetic

exchange between populations along the coast, a lack not surprising in view

of the sedentary mode of life and the methods of reproduction of these small

fishes. That such variation between localities occurs makes one hesitate to

recognize even a racial difference between the east and west coast forms. The

40 ANNALS OF THE SOUTH AFRICAN MUSEUM

identical intromittent organ of east and west coast specimens is the strongest

argument in favour of joining these species, since the shape of the intromittent

organ is strongly species-specific in the genus Clznus.

Smith (19475) stated that if the seven Knysna specimens and the five west

coast (Lambert’s Bay) specimens had not been so widely separated geographi-

cally he would have placed them in the same species. A study of clinid distri-

bution in South Africa has shown that there are few species from the west coast

of South Africa that do not occur at least as far east as Port Elizabeth, so that the

occurrence of a species at both Lambert’s Bay and Knysna is not unexpected,

40 60

“= 20 3 3

g g

39 38 37 36 35 34 33 32 39 38 37 36 35 34 33 32

dorsal spines dorsal spines

(a) Luderitzbucht (d) Lambert's Bay

40 40

o rh

wv =

5 20 ‘6 20

2 g

39 38 37 36 35 34 33 32 39 38 37 36 35 34 33 32

_ dorsal spines dorsal spines

‘(b) Port Nolloth (e) Sea Point

40 40

£ a

= . 20

3 z

: o

o c

39 38 37 36 35 34 33 32 39 38 37 36 35 34 33 32

dorsal spines dorsal spines:

Fic. 13. Histogram showing dorsal spine count variation in different populations of Clinus agilis.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 41

particularly when the species has been recorded from False Bay on several

occasions as well.

It is proposed therefore that Ophthalmolophus anne Smith, 1947, should

become a synonym of Clinus agilis Smith, 1931, since the two species are indis-

tinguishable. Clinus agilis is common on the west coast of South Africa but

becomes rare east of Cape Point.

Barnard (1948) suggested that Clinus agilis might be the young of Clinus

taurus Gilchrist & ‘Thompson, but the two species are easily distinguishable in

all stages by the shape of the interorbital, which is strongly concave in Clinus

taurus and flat in Clinus agilis, and by the number of dorsal soft rays (2-4 in

Clinus agilis and 5-6 in Clinus taurus).

Clinus agilis is in some ways intermediate between forms with a low, even

dorsal fin and no notch in the dorsal fin membrane between the third and

fourth spines, such as Clinus acuminatus, and forms with a low dorsal crest and a

notch in the dorsal fin membrane such as Clinus brevicristatus. There is a marked

similarity in appearance between Clinus agilis and Clinus brevicristatus both in

body form and colour pattern, but in the latter the first three dorsal spines are

elevated, forming a crest, and there are more dorsal! soft rays.

Distribution (fig. 11). The known range of this species is Liideritzbucht (26°28’S.,

15°10 E) (South West Africa) to Port Alfred (33°36’S., 26°54’E), common

on the west coast in pools at all levels of the shore and also taken infratidally,

but on the whole rare on the east coast, although a fairly large number of

specimens was taken in Knysna lagoon. The biota of the Knysna estuary has

been discussed by Day et al. (1952) and Day (1967). It is an exceptionally rich

faunal area, and has also received a great deal more attention than any other

area on the south coast of South Africa. A similar case of apparently discon-

tinuous distribution involving a lagoon is observed in Blennioclinus brachyce-

phalus on the west coast (Penrith, 19650), without racial differentiation of the

separated populations. |

Clinus (Clinus) berrisford: Penrith, 1967

(Fig. 14)

Clinus berrisfordi Penrith, 1967: 48, figs 3, 4(c).

Description. D. XX XITI-XXXVI (XXXIV-XXXV) 5-6; A. II 23-25 (23-

24); P. 11-12 (12); V. 1 3; C. 13. Gill-rakers in outer series on first arch 3-4 ++

6—7. Vertebrae 16 (rarely 17) + 28-31. First three dorsal spines not elevated

to form crest, but equal to or little longer than fourth dorsal spine. No notch

in membrane between third and fourth dorsal spines. Dorsal spines with clusters

of three fine cirri at tips for about half length of fin. Pectoral fin rounded. Inner

pelvic ray minute but present in all specimens examined. Caudal peduncle

short, length 20-35% head length, depth 20-35% head length. Caudal fin

subtruncate.

Body slightly compressed, covered with small scales extending on to dorsal

and caudal fin bases but not anal fin base or head. Depth 4-5-5 in standard

42 ANNALS OF THE SOUTH AFRICAN MUSEUM

length. Head 3-4-4 in standard length, snout wedge-shaped, profile acute. Eye

3-4:2 in head. Supraorbital tentacle prominent, with flattened stalk ending in

several long slender filaments. Cirrus on anterior nostril with narrow stalk and.

flattened, bilobed tip. Upper jaw 36:4-50% of head length. Lips moderate.

Vomer toothed. Sensory pores of nasal, interorbital and mandibular series

single with exception of double first pore in mandibular series; pores of remain-

ing series double or multiple (fig. 14(b)).

(ii)anterior

i) lateral

ii) ventral

Sf ARAN A A AN... Si ASA

oN Set care) MW WN

ie SS

(d) ba east

Fic. 14. Clinus (Clinus) berrisfordi: (a) Lateral view, female, 95 mm, S.A.M. 24221, holotype;

(b) Head pore system; (c) Intromittent organ of male; (d) Lateral line.

Lateral line of usually vertical pairs of pores in front to post-pectoral curve,

few single pores opening above or below line, then of short separate horizontal

tubes with pore at either end (fig. 14(d)). Intromittent organ of male with a

long basal portion and a slender upturned tip ensheathed by a pair of bilobed

ventro-lateral lips and a small pair of rounded dorsal lips (fig. 14(c)).

Colouring. Reddish orange with about seven faint broad darker cross-bars. Two

dark radiating lines from eye across cheek. Dark ocellate spot on shoulder. Fins

reddish orange, faintly mottled and barred. Plain yellow preserved in alcohol.

Location of type material. South African Museum, Cape Town.

Material examined. 16 specimens, 37:-5-106 mm in standard length. 16 from

Onrust River Mouth, S.A.M. 24601 (holotype), S.A.M. 24221 (paratypes).

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 43

Remarks. This species is similar in appearance to Clinus acuminatus and Clinus

heterodon but differs from the former in the dorsal and anal fin counts, the form

of the intromittent organ, the height of the first dorsal spine, the form of the

supraorbital tentacle, the anterior part of the lateral line, and the habitat,

occurring at a much lower level on the shore. It differs from the latter in the

dorsal, anal, and pectoral fin counts, the intromittent organ (although of all the

species of Clinus, Clinus berrisfordi and Clinus heterodon have the most similar form

of intromittent organ), the anterior part of the lateral line, the clusters of cirri

on the dorsal spines, and in having a narrower interorbital.

Distribution (fig. 11). So far taken only at Onrust River Mouth (34°25’S.,

19°10’). The 16 specimens were taken from a weedy pool at the bottom of the

intertidal region.

Clinus (Clinus) brevicristatus Gilchrist & Thompson, 1908

(Fig. 15)

Clinus brevicristatus Gilchrist & Thompson, 1908: 118. Barnard, 1927: 856.

Petraites brevicristatus: Smith, 1945: 540, 1949: 353, pl. 78 fig. 985.

Description. D. XXXIII-XXXVI (XXXV-XXXVI) 4-7 (5); A. IT 21-24;

P. 12-14 (12-13); V. I 3; C. 13. Gill-rakers in outer series on first arch 1-2 +

3-4. Vertebrae 17-18 + 27-31. First three dorsal spines elevated to form low

crest, fourth dorsal spine 70% to over 90% first dorsal spine. Notch of varying

depth in membrane between third and fourth dorsal spines (see table 3).

Clusters of 4-6 cirri at tips of all dorsal spines. Pectoral fin rounded. Inner

pelvic ray minute. Caudal peduncle short, length 21-34:5% head length,

depth 21-28% head length. Caudal fin subtruncate.

Body slightly compressed, covered with small scales extending on to

dorsal fin base but not caudal or anal fin bases or head. Depth 4-5-5 in standard

length. Head 3-°75-4:75 in standard length, snout bluntly rounded. Eye 2-75-—

3°25 in head. Supraorbital tentacle prominent, with short subcylindrical stalk

and round, flattened tip ending in several simple branches. Cirrus on anterior

nostril prominent, elongate, flat, and narrow, with slightly indented margin.

Upper jaw 40-48-5°% head length. Lips fairly thin. Vomer toothed. Sensory

pores of head single in nasal and interorbital series and in mandibular series

with exception of first, double pore; multiple as well as double pores in supra-

orbital, postorbital, occipital and preopercular series (fig. 15(b)).

Lateral line of usually about 25 vertical pairs of pores in front (few may be

single opening above or below line), to post-pectoral curve, then of short

separate horizontal tubes with pore at either end (fig. 15(d)). Intromittent

organ of male with fairly long basal portion and single pair of dorso-lateral

lips ensheathing base of tip, which is slightly swollen, narrowing terminally

(fig. 15(c)).

Colouring. Ground colour usually light grey, with about seven vermilion cross-

bars, having shorter brown bars between them. The cross-bars may, however,

44 ANNALS OF THE SOUTH AFRICAN MUSEUM

be various shades of green, brown, or deep red. The cross-bars extend on to

dorsal and anal fins, but membrane joining dorsal soft rays has irregular

translucent patches. Caudal and pectoral fins finely barred with dark brown.

A 3-shaped dark mark at base of pectoral fin, and a dark stripe from edge of

branchiostegal membrane to central edge of pectoral fin base. Operculum with

irregular dark mark; two irregular dark bands radiate from eye across cheek.

Pelvic fins barred with brown. Belly silvery white to greyish. Head grey or

pinkish. Pattern retained for several years after preservation.

Location of type material. South African Museum, Cape Town.

Material examined. 36 specimens, 37:5-106 mm in standard length. 2 from

Lambert’s Bay, R.U.C.; 1 from Simon’s Bay, False Bay, S.A.M. 24243; 1 from

Kalk Bay, False Bay, S.A.M. 9988 (holotype); 5 from Dalebrook, False Bay,

S.A.M. 23870 and 18/12/1964, S.A.M. not catalogued; 1 from St. James, False

Bay, 16/5/1965, S.A.M. not catalogued; 26 from Strandfontein, False Bay’

S.A.M. 23871, S.A.M. 23872, S.A.M. 23954, S.A.M. 23971, S.A.M. 23976,

S.A.M. 24242.

Remarks. Clinus brevicristatus is very similar to the species with a low dorsal fin,

such as Clinus cottoides, Clinus acuminatus, Clinus venustris, and particularly Clinus

agilis. ‘The species was described from a single female specimen from Kalk Bay,

é a, TIT VI Me Se ag

é iro i} Lit qog

FEL GHEE

(i) ventral Gi) lateral (iii) anterior

(c)

Cae SAA

A “9

—F..

ea de Pe eS oe ow aa ey ey 5

_ Sa ID a

(d) eet

(b) ae

Fic. 15. Clinus (Clinus) brevicristatus: (a) Lateral view, female, 70 mm, S.A.M. 23870; (b) Head

pore system; (c) Intromittent organ of male; (d) Lateral line.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 45

in False Bay. I have collected additional specimens of both sexes from weedy

areas at Strandfontein and Dalebrook in False Bay. The new material covers a

fairly wide size range. This species is no more similar to Clinus woodi, with which

it was placed in the genus Petraiies by Smith (1945), than it is to Clinus venustris

or Clinus agilis ; in fact, it resembles the latter two species far more closely. The

development of the crest is poor in both male and female specimens of C. brevi-

cristatus, unlike that in Clinus woodi, in which the crest is high and triangular

in both sexes, and almost entirely separate from the rest of the fin. In Clinus

brevicristatus the crest is low and rounded, and resembles the crest of Clinus

agilis (with a notch in the membrane between the third and fourth dorsal spines

and the first three dorsal spines not elevated) more closely than it does the crest

of Clinus woodi. Clinus brevicristatus does not appear to reach a large size. This

species is unusual in the genus Clinus in that it lives almost exclusively in weed-

beds. The peculiar intromittent organ of C. navalis is nearest to that of C. brevi-

cristatus, which shows little definition of the thin dorso-lateral lips when fully

erect.

Distribution (fig. 11). The known range is from Lambert’s Bay (32°04’S.,

18°20'E) to False Bay (+34°10’S., 18°50’E). The species is not common.

Smith (1937) gave the range as False Bay to Great Fish Point, but the latter

limit was presumably an error, as it was not repeated in Smith’s 1945 and 1949

works. Taken at low tide in False Bay from dense beds of the green alga Caulerpa

filiformis, intertidal. _

Clinus (Clinus) cottordes Valenciennes in Cuvier & Valenciennes, 1836

(Fig. 16)

-Clinus cottoides Valenciennes in Cuvier & Valenciennes, 1836: 367. Swainson, 1839: 276. Gilchrist

& Thompson, 1908: 125. Barnard, 1927: 858.

Blenniomimus cottoides: Smith, 1945: 539, 1949: 353, pl. 78 fig. 983.

Description. D. XXXI-XXXVI (XXXITI-XXXIV) 4-6 (5-6); A. IT 21-25;

P. 12-14 (12-13); V. I 2-3; C. 13. Gill-rakers in outer series on first arch

2 + 5-6. Vertebrae 15 + 29-31. Dorsal fin low, even, anterior spines shortest;

no notch in membrane between third and fourth dorsal spines. Pectoral

fin rounded. Inner (third) pelvic ray minute or absent. Caudal peduncle

short, length 30°5-34°5% head length, depth 21-26% head length. Caudal fin

subtruncate.

Body slightly compressed, tapering markedly towards tail, covered with

small embedded scales extending on to dorsal and caudal fin bases but not anal

fin base or head. Depth 4:5-5:5 in standard length. Head very large, heavy,

3°25-4°25 in standard length. Snout bluntly rounded. Head becomes heavier

in relation to body with age. Eye very large, 2°-25-3°5 in head. Interorbital

concave, bony ridge above each eye. Deep groove across occiput. Supraorbital

tentacle prominent, on bony ridge, with flattened stalk and numerous long

filamentous cirri terminally. Cirrus on anterior nostril short, flattened, with

46 ANNALS OF THE SOUTH AFRICAN MUSEUM

about four short, simple branches terminally. Mouth large, increasing with

size of fish, upper jaw 42-60% head length. Lips thin. Vomer toothed. Sensory

pores of head single in nasal and interorbital series and on mandible except

for first, double pore; mainly double in other series, and some multiple in

occipital region (fig. 16(b)). Some pores in supraorbital and occipital regions

Open on minute papillae.

Lateral line usually of about 18-20 vertical pairs of pores, interspersed

with or followed by few single pores opening above or below line, in front to

post-pectoral curve, then of short separate horizontal tubes with pore at either

end (fig. 16(d)). Intromittent organ of male with fairly long basal portion and

single pair of slightly papillose dorso-lateral lips below which tip, which is

slightly swollen at base with narrow termination, protrudes (fig. 16(c)).

Colouring. Ground colour slaty grey or dull greenish with irregular mottling,

often in lacy pattern with vague crossbars extending on to dorsal and anal fins,

chiefly dark grey, dark red, and whitish. Fins dusky, faintly mottled and barred.

Three small dark lines on pectoral base. Prominent round black spot on opercle.

Dark mark behind centre of eye, below which two narrow dark lines radiate

from eye across cheek. Head and lips dusky to slate-grey; terminal filaments of

supraorbital tentacles white. Belly white to greyish. Juveniles milky with fine

lacy dark markings particularly in upper half of body; opercular spot well

defined at all stages. Pattern retained for years after preservation.

Location of type material. Muséum National d’Histoire Naturelle, Paris.

Material examined. 951 specimens, 18-120 mm in standard length. 1 specimen,

63 mm in standard length, Cape, Paris Mus. Cat. No. A2078 (lectotype) ;

26 from Doring Bay, southern Namaqualand, S.A.M. 24229; 4 from southern

Namaqualand, R.U.C.; 101 from Lambert’s Bay, S.A.M. 23940; 12 from

Lambert’s Bay, R.U.C.; 3 from Saldanha Bay, S.A.M. 17914; 1 from Lange-

baan, S.A.M. 21880; 6 from Mouille Point, Cape Peninsula, 5/5/1965, S.A.M.

not catalogued; 42 from Sea Point, S.A.M. 23941, S.A.M. 23942 and 17/2/1965,

S.A.M. not catalogued; 1 from Hout Bay, S.A.M. 12013; 2 from Miller’s

Point, False Bay, S.A.M. 22906, S.A.M. 23937; 164 from Froggy Pond and

Miller’s Point, False Bay, S.A.M. 23939; 113 from Dalebrook, False Bay,

19/2/1965, 3/3/1965 and 18/4/1965 S.A.M. not catalogued; 25 from St. James,

False Bay, S.A.M. 10540; 80 from Strandfontein, False Bay, S.A.M. 23978

and 12/1/1964 and 13/8/1964, S.A.M. not catalogued; 33 from Onrust River

Mouth, S.A.M. 23938, S.A.M. 24230; 2 from Hermanus, S.A.M. 18095; 4

from Die Dam, Bredasdorp district, S.A.M. 24510; 59 from Still Bay, S.A.M.

24666; 35 from Mossel Bay, S.A.M. 23947; 4 from Knysna, S.A.M. 24230;

53 from Plettenberg Bay, S.A.M. 23946; 72 from Port Elizabeth, S.A.M.

23945, S.A.M. 23985; 1 from Kidd’s Beach, East London, S.A.M. 20564;

14 from Igoda Mouth, East London, $.A.M. 25060; 104 from East London,

S.A.M. 23943, S.A.M. 23944, S.A.M. 24228, and S.A.M. 20562; 2 from

Gonubie River Mouth, East London, S.A.M. 25061; 2 from Kei Mouth,

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 47

S.A.M. 20563; 1 juvenile from Mozambique Island, R.U.C.

Lectotype of Clinus cottoides

I have examined three of the four syntypes of Clinus cottoides. Of these

three syntypes, one small specimen is identifiable as the species regarded as

Clinus cotioides in South Africa since the time of Gilchrist & Thompson (1908).

The two larger syntypes are examples of a species known as Clinus taurus Gil-

christ & Thompson. Dr. M. L. Bauchot kindly examined the remaining, dried

syntype in the Paris Museum and concluded that it too represented Clinus

taurus.

aS 2

CARS

” (ii) lateral (iii) anterior

(c)

(i) ventra}

(b) AS

Tae

Fic. 16. Clinus (Clinus) cottoides: (a) Lateral view, female, 85 mm, S.A.M. 23945; (b) Head

pore system; (c) Intromittent organ of male; (d) Lateral line.

Although Valenciennes (1836 in Cuvier & Valenciennes) almost certainly

based his description of C’.. cottoides on the larger specimens, the description

applies more or less to both species; and the species was not figured. In order

not to complicate the nomenclature of the two species, I here designate the

small specimen, 63 mm, Paris Museum Catalogue No. A2078, of the four

syntypes of Clinus cotioides the lectotype.

48 ANNALS OF THE SOUTH AFRICAN MUSEUM

Remarks. Smith (1945) considered the concave interorbital of Clinus cottoides

and of Clinus taurus to be of generic significance. However, Clinus latipennis,

large specimens of Clinus heterodon, and to a slight degree Clinus helenae, also

have a concave interorbital with low bony ridges over the eyes. Instead of a

clear-cut division between species with and without a ridge over the eyes, there

is a series, from Clinus heterodon, in which the supraorbital ridges develop only

in large specimens, to Clinus taurus, in which the ridges are very heavy at all

stages. These species all appear to be closely related to each other and to species

lacking the ridge, such as Clinus acuminatus and Clinus berrisfordt.

Specimens of Clinus cottoides from East London were compared with speci-

mens from Lambert’s Bay, and no east-west clinal or individual population

differences were found in the fin counts of this species. Clinus cottoides does, how-

ever, attain a considerably greater size on the west coast than it does east of

Cape Point.

Distribution (fig. 11). The known range is from southern Namaqualand

(+31°40’S., 18°15’E) (Olifants River Mouth) to the Kei River (32°41’S.,

28°23’E). It is one of the most abundant species from Lambert’s Bay to the

Kei River, particularly from False Bay eastwards. A single tiny juvenile from

Mozambique Island was seen in the Rhodes University Ichthyology Depart-

ment, but this seems to be well outside the normal range for this species, which

has not been recorded otherwise much north of the Kei River. Clinus cottoides

was recorded by Kner (1865-67) from Java, but De Beaufort & Chapman

(1955) stated that this record is certainly erroneous; its occurrence outside

South African waters is most improbable.

Clinus (Clinus) helenae (Smith, 1945)

(Fig. 17)

Ophthalmolophus helenae Smith, 1945: 542, fig. 33 1949: 355, pl. 79 fig. go2.

Clinus helenae: Smith, 1966: 73.

Description. D. XXXIV-XXXVII 5-6; A II 24-25; P. 13; V. I 3; C. 13.

Vertebrae 16 + 30. Dorsal fin low, even; no notch in membrane between

third and fourth dorsal spines. No clusters of cirri at tips of dorsal spines.

Pectoral fin rounded. Inner pelvic ray minute. Caudal peduncle short, length,

30-35°5% head length, depth 20-24% head length, noticeably longer than

deep. Caudal fin subtruncate.

Body slightly compressed, covered with small embedded scales not extend-

ing on to dorsal, anal, or caudal fin bases or head. Depth 4:5—-5 in standard

length. Head 3-75-4-75 in standard length, snout bluntly rounded. Inter-

orbital slightly concave, low bony ridges over eyes; occipital grooves moderately

deep. Eye 3-5 in head. Supraorbital tentacle prominent, with flattened stalk

and terminal fringe of long fine filamentous cirri. Cirrus on anterior nostril

small, simple, flaplike. Upper jaw 39°5-52:5% head length. Lips moderately

thick. Vomer toothed. Sensory pores of head single in nasal, interorbital, and

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 49

mandibular series except for first double mandibular pore; double or multiple

in other series (fig. 17(b)). Many pores open on raised papillae, giving head a

rough appearance.

Lateral line of vertical pairs of pores in front to post-pectoral curve, then

of short separate horizontal tubes with pore at either end (fig. 17(d)). Intro-

mittent organ of male with moderately long basal portion; small pair of circular

dorsal lips and large pair of ventro-lateral lips with finely serrated ventral

margins ensheathing tip. Small pair of lateral lobes on basal portion (fig. 17(c)).

Colouring. A single fresh specimen from Kei Mouth was light grey, with darker

mottling in the form of irregular cross-bars; body and head speckled with white.

Smith (1945) described the colouring of his specimens as ‘light brown, with

seven darker broken crossbands. Darker spots and speckles on head and body.’

(i) ventral

(iii) anterior

(c)

Gi) lateral

ae a a Eee

“2.

b S

(b) (a) \

Fic. 17. Clinus (Clinus) helenae: (a) Lateral view, female, 71 mm, S.A.M. 25065; (b) Head

pore system; (c) Intromittent organ of male; (d) Lateral line.

The fins are finely spotted and barred. Pattern retained for years after pre-

servation.

Location of type material. Department of Ichthyology, Rhodes University,

Grahamstown.

Material examined. 8 specimens, 67:-5-78:5 mm in standard length. 4 from

50 ANNALS OF THE SOUTH AFRICAN MUSEUM

Boknes Point, R.U.C. (syntypes); 3 from Cape Morgan, R.U.C. (syntypes) ;

1 from Kei mouth, $.A.M. 25065.

Remarks. Clinus helenae resembles Clinus cottoides, Clinus taurus and particularly

Clinus latipennis in the form of the head, although the concave interorbital is

least marked in Clinus helenae. It is also similar to Clinus heterodon (in which the

interorbital becomes concave only in large specimens), Clinus berrisford: and

Clinus acuminatus. The intromittent organ is similar to that of Clinus latipennis,

the only other species which has a pair of subsidiary lobes on the basal portion.

Distribution (fig. 11). The known range is Port Alfred (33°36’S., 26°54’E) to

the Bashee River (31°55'S., 28°27’E) ; rare.

Clinus (Clinus) heterodon Valenciennes in Cuvier & Valenciennes, 1836

(Fig. 18)

Clinus heterodon Valenciennes in Cuvier & Valenciennes, 1836: 394.

Clinus obtusifrons Penrith, 1967: 46, figs 2, 4(b).

Description. D. XXX-XXXII (XXX-XXXI) 6-7; A. II 20-22 (20-21);

P. 13; V. I 2-3; C. 13. Gill-rakers in outer series on first arch 1-2 + 5-6.

Vertebrae 16 + 27-30. First three dorsal spines not forming crest, but equal

to or a little longer than fourth spine, second spine longest. No notch in mem-

brane between third and fourth dorsal spines. No clusters of cirri at tips of

dorsal spines. Pectoral fin rounded. Inner (third) pelvic ray minute or absent.

Caudal peduncle short, length 20-35°% head length, depth 25-37:5% head

length. Caudal fin subtruncate.

Body slightly compressed, covered with small embedded scales extending

on to dorsal and caudal fin bases; anal fin base and head naked. Depth 4-5

in standard length. Head heavy, 3-2—4°5 in standard length. Interorbital con-

cave with low supraorbital ridges in large specimens. Snout bluntly rounded,

angle of profile obtuse. Eye 3-4-5 in head. Supraorbital tentacle prominent,

with flattened stalk, terminating in several short, simple branches. Cirrus on

anterior nostril flattened, expanded, and roughly triangular at tip. Upper jaw

36-4-50% head length, increasing with size of fish. Lips moderate. Vomer

toothed. Sensory pores of head single in nasal and interorbital series, mainly

single in mandibular, suborbital, and preopercular series. Mainly double and

multiple in remaining series (fig. 18(b)). A few pores in occipital region open

on raised papillae.

Lateral line of single pores, opening above or below line, in front to post-

pectoral curve, then of short separate horizontal tubes with pore at either end

(fig. 18(d)). Intromittent organ of male with moderately long basal portion

and slender, upwardly hooked tip between pair of lateral lips and small rounded

pair of dorsal lips (fig. 18(c)).

Colouring. Dusky or slaty with darker grey, irregular, lacy cross-bars, speckled

with white and sometimes dark blue when fresh. Tips of anal and pelvic rays

red. Prominent blue-edged spot on shoulder. Two dark radiating bars across

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE aie!

cheek. Tips of dorsal fin and orbital tentacles white. Pattern retained for years

after preservation.

Location of type material. Muséum National d’ Histoire Naturelle, Paris.

Material examined. 113 specimens, 20-116 mm in standard length. 1 specimen,

Cape, Paris Museum Cat. No. A 1885 (holotype); 1 from Milestone 26, north

of Swakopmund, S.A.M. 24201; 2 from Liideritzbucht, S.A.M. 24211; 1 from

Sinclair’s Island, August 1947, R.U.C.; 12 from Port Nolloth, $.A.M. 24216, 5

from Port Nolloth, 27/4/1950 R.U.C.; 4 from Hondeklip Bay, S.A.M. 24671,

S.A.M. 24670 (holotype and paratypes of obtusifrons); 1 from Hondeklip Bay,

28/4/1960, R.U.C.; 3 from Gert du Toit Bay, southern Namaqualand, $.A.M.

24253; 1 from Doring Bay, southern Namaqualand, 26/4/1960, R.U.C.; 58

from Lambert’s Bay, R.U.C.; 3 from Ysterfontein, 20/2/1948, R.U.C.; 7 from

Sea Point, S.A.M. 25068; 6 from False Bay, March 1947, R.U.C.; 8, no locality,

R.U.C.

(i) lateral

(ii) anterior

a) —=---

Fic. 18. Clinus (Clinus) heterodon: (a) Lateral view, female, 124 mm, S.A.M. 24671; (b) Head

pore system; (c) Intromittent organ of male; (d) Lateral line.

Remarks. The re-examination of Valenciennes’ types indicated that the type-

specimen of the species that he named heterodon was a tentacled clinid referable

to the genus Clinus. It is in no way identifiable with the species for which the

name heterodon has been used in South African literature since Barnard (1927)

52 ANNALS OF THE SOUTH AFRICAN MUSEUM

identified it as such from Valenciennes’ very brief description. The description

is very inadequate and, translated, reads ‘I did not see a tentacle over the

eye ...’. It has naturally been assumed, as it happens, incorrectly, that the

specimen he examined did not have a tentacle over the eye. Gilchrist & Thomp-

son (1908) did not identify any of the South African species they collected and

described with Valenciennes’ C. heterodon. However, Barnard (1927), assuming

the lack of a supraorbital tentacle, and using the dorsal and anal fin counts

given by Valenciennes for C. heterodon, identified a species of weed-dwelling

clinid described by Gilchrist & Thompson under the name Clinus graminis as

Valenciennes’ heterodon. This identification has been followed by subsequent

workers, so that the name heterodon has come to be applied to a species of the

genus Pavoclinus (Smith, 1945, 1949).

The species to which Valenciennes’ type and description of C. heterodon

refer is fairly common on the west coast of South Africa.

Owing to its superficial similarity in fin counts to C. acuminatus it was not

rediscovered until recently, and was described as a new species Clinus obtusifrons

(Penrith, 1967). However, since my specimens of obtusifrons agree in every

Keuse ee

Carer ne aoe ne : oe aS PES ze ee

box Ae OA i

A@ oO f

O ot A 4 4

Fic. 19. Distribution of C. latipennis (open squares), C. robustus (open triangles), C. superciliosus

(open circles), C. taurus (closed triangles), C. venustris (closed circles), C. woodi (closed squares).

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 53

way with Valenciennes’ holotype of C. heterodon, the name heterodon must

replace obtusifrons. The name graminis Gilchrist & Thompson is revived for the

species that has been masquerading under the name of heterodon in the South

African literature.

Clinus heterodon has very similar dorsal counts to Clinus acuminatus, but can

be distinguished from that species by the shape of the intromittent organ, the shape

of the snout and profile, the height of the first three dorsal spines, the greater

width of the interorbital, and the number of pectoral rays. Large specimens of

Clinus obtusifrons can be distinguished from Clinus acuminatus by the concave

interorbital and the supraorbital ridges. The intromittent organ is similar to

the type found in Clinus berrisfordi, Clinus cottoides, and Clinus brevicristatus..

Disiribution (fig. 11). The known range is from north of Swakopmund (22°40’S.,

14°34’E) (South West Africa) to False Bay (+34°10’S., 18°50’E). The

species is fairly common in pools in the middle and lower regions of the inter-

tidal zone on the coast west of Cape Point, particularly from the Olifants River

to the Orange River, but is extremely rare in False Bay.

Clinus (Clinus) latipennis Valenciennes in

Cuvier & Valenciennes, 1836

(Fig. 20)

Clinus latipennis Valenciennes in Cuvier & Valenciennes, 1836: 394. Barnard, 1927: 860. Hubbs,

1952: 106. Smith, 1966: 73.

Clinus latipinnis: Gilchrist & Thompson, 1908, 127.

Labrisomus latipennis: Swainson, 1839: 277.

Ophthalmolophus latipinnis: Gill, 1860: 104. Smith, 1945: 542.

Ophthalmolophus latipennis: Smith, 1949: 355, fig. 993.

Description. D. XX XIII-XXXVI 8-9; A. IT 23-26; P. 13-14; V. I 2-3; C. 13.

Vertebrae 17 + 30-33. Dorsal fin low, even; no notch in membrane between

third and fourth dorsal spines. No clusters of cirri at tips of dorsal spines.

Pectoral fin rounded. Inner pelvic ray minute or absent. Caudal peduncle

short, length 27-5-33°3% head length, depth 20-33% head length. Caudal fin

subtruncate.

Body slightly compressed, covered with small embedded scales extending

on to dorsal fin base; caudal and anal fin bases and head naked. Depth

4°85-5°2 in standard length. Head 3-5-5-5 in standard length, snout bluntly

rounded. Interorbital grooves moderately deep. Eye 3:2—4:6 in head. Supra-

orbital tentacle prominent, with flattened stalk, expanded tip, and terminal

fringe of fine filamentous cirri. Cirrus on anterior nostril flattened, narrow at

base with expanded trilobate tip. Upper jaw 35-43°5% standard length. Lips

moderately thin. Vomer toothed. Sensory pores of head single in nasal and

interorbital series, single except for first pore in mandibular series, and double

in preopercular series; double or multiple in remaining series (fig. 20(b)).

Many pores open on papillae, giving head a rough appearance.

Lateral line of vertical pairs of pores in front to post-pectoral curve, then of

54 ANNALS OF THE SOUTH AFRICAN MUSEUM

short separate horizontal tubes with pore at either end (fig. 20(d)). Intro-

mittent organ of male with moderately long basal portion and swollen tip

ensheathed by pair of large, flattened ventral lobes and small flattened dorsal

lobe. Three lobes at base of organ, two lateral rounded ones and one dorsal

flattened lobe (fig. 20(c)).

Colouring. Pinkish brown with faint irregular darker mottling and speckling

Fins whitish, dorsal with about seven dark cross bars not reaching upper edge

of fin. Pectoral and caudal fins with clusters of small black dots forming irregu-

lar spots. Head grey-brown, mottled. A few white spots on pectoral base and

along sides. Pattern remains for a long time after preservation.

Location of type material. Muséum National d’Histoire Naturelle, Paris.

Material examined. 7 specimens, 47-102 mm in standard length. 3 specimens,

Cape, Paris Museum Cat. Nos. Azo10, Ago11, A2o12 (syntypes); 1 from St.

James, S.A.M. 10533; 2 from Die Dam, Bredasdorp district, S.A.M. 25066;

1 from Fish Hoek, December, 1966, S.A.M. 25067.

(i) ve (ii) lateral (iii) anterior (iv) dorsal

(c)

SPER '

SSE WEEE jg

AY,

arn

Pea

Pk a .

(d) “8 @= > eme- -=

(b)

Fic. 20. Clinus (Clinus) latipennis: (a) Lateral view, female, 82 mm, S.A.M. 25066; (b) Head

pore system; (c) Intromittent organ of male; (d) Lateral line.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 55

Remarks. This species is poorly known, although it was one of the earliest

described South African Clinidae. There are four specimens in the South

African Museum, three of which were acquired recently, so that fresh material

was available for study. This species is extremely similar to Clinus helenae,

particularly in the occurrence of basal lobes around the intromittent organ.

It is distinguished from Clinus helenae in having more dorsal soft rays. It is very

similar in appearance to Clinus cottoides.

Distribution (fig. 19). The known range is from Table Bay (+33°45'S., 18°20’E)

to Cape Agulhas (34°50’S., 20°00’E); rare.

Clinus (Clinus) robustus Gilchrist & Thompson, 1908

(Fig. 21)

Clinus robustus Gilchrist & Thompson, 1908: 128. Barnard, 1927: 850. Smith, 1945: 541, 1949:

354, pl. 81 fig. 987 and fig. 987.

Clinetrachus robustus: Hubbs, 1952: 107.

Caboclinus robustus: Smith, 1966: 73.

Description. D. XX XII-XX XIV (XXXII) 9-14 (10-11); A. IT 26-28 (26-27) ;

P. 12; V.13; C. 13. Gill-rakers in outer series on first arch 1-2 + 4-5. Vertebrae

17 + 33. First three dorsal spines slightly elevated, forming low crest, second

spine longest (first dorsal spine 0-5-4 mm longer than fourth dorsal spine).

Shallow notch in membrane between third and fourth dorsal spines, depth

variable (see table 3). Anterior dorsal spines with clusters of 3-4 cirri at tips.

Inner pelvic ray well developed, as least half length and thickness of other

rays. Caudal peduncle short, length 25-33:3°% head length, depth 24-30%

head length. Caudal fin subtruncate.

Body slightly compressed, covered with small embedded scales extending

on to bases of dorsal and caudal fins but not anal fin base or head. Depth

4°5-5°5 in standard length. Head large, heavy, 3:25-4 in standard length,

snout bluntly rounded in large specimens, somewhat subconical in smaller

ones. Eye 5—7-5 in head. Supraorbital tentacle prominent, with flattened stalk

and expanded flat tip, terminating in several short, flat, simple branches.

Cirrus on anterior nostril somewhat elongate, narrower at base, margin irregu-

larly indented. Mouth large, upper jaw 45°5-53°5% head length. Lips thick,

with distinct vertical corrugations. Vomer toothed. Sensory pores of head

double or multiple in suborbital, supraorbital, and occipital regions, otherwise

mostly single (fig. 21(b)).

Lateral line of vertical pairs of pores and single pores opening above and

below the line in front to post-pectoral curve, then of short separate horizontal

tubes with pore at either end (fig. 21(d)). Intromittent organ of male with

moderately long basal portion and two pairs of lips ensheathing tip, small

rounded dorsal pair and large ventro-lateral pair, latter with finely serrate

ventral margins (fig. 21(c)).

Colouring. Variable, olive yellow to dusky, with speckling in dark olive, red,

56 ANNALS OF THE SOUTH AFRICAN MUSEUM

black, green, and white, forming obscure crossbars continuing on to dorsal and

anal fins, or dark greyish brown with vague mottling. Sometimes a few bright

orange speckles along anterior part of lateral line. Pectoral, pelvic, and caudal

fins barred. Dorsal, anal, and pelvic fins red- or orange-tipped. Usually two

dark radiating lines from eye across cheek. Uniform yellowish preserved in

alcohol.

Location of type material. South African Museum, Cape ‘Town.

Lectotype. Female, 280 mm standard Jength, Seal Island, False Bay, S.A.M.

10539.

(ii) fateral

® ventral

(iii) anterior (¢)

.(b)

Fic. 21. Clinus (Clinus) robustus: (a) Lateral view, female, 121 mm, S.A.M. 23878; (b) Head

pore system; (c) Intromittent organ of male; (d) Lateral line.

Material examined. 12 specimens, 123-314 mm in standard length. 1 from Melk-

bosch, S.A.M. 24072; 8 from Kalk Bay, False Bay, S.A.M. 10537, S.A.M.

10538 (paralectotypes), S.A.M. 10539 (lectotype), S.A.M. 18087; 2 from

Dalebrook, False Bay, S.A.M. 23873; 1 from St. James, False Bay, S.A.M.

12019.

Remarks. Although Smith (1945), and following him Hubbs (1952), treated

Clinus robustus as if it were very close to Clinus superciliosus, the similarity in size

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 57

and habit between these two species probably accounted for this as much as

actual resemblance, as the two species are not easily confused. Clinus robustus

appears to be a rather primitive and generalized species; it is at least as closely

related to forms such as Clinus taurus and Clinus agilis as to Clinus superciliosus,

and at smaller sizes bears a superficial resemblance to large specimens of Clinus

acuminatus.

Distribution (fig. 19). The known range is west coast of the Cape Peninsula

(34°21'S., 18°29’E) to East London (33°00’S., 27°54’E). Fowler (1934)

recorded this species from Natal, but its occurrence there is most improbable.

The validity of all Fowler’s (1934) clinid records is doubtful. Infratidal except

in the young stages. Appears to be rare.

Clinus (Clinus) superciliosus (Linnaeus, 1758)

(Fig. 22)

Blennius superciliosus Linnaeus, 1758: 257.

Blennius mustelaris Linnaeus, 1758: 257. Gronovius, ed. Gray, 1854: 98 (mustellaris).

Blennius punctulatus Lacépéde, 1800: 460.

Blennius mustela Lacépéde, 1800: 459.

Blennius spadiceus Bloch & Schneider, 1801: 172.

Blennius capensis Forster, in Bloch & Schneider, 1801: 175.

Clinus superciliosus: Cuvier, 1817: 173. Valenciennes in Cuvier & Valenciennes, 1836: 360.

Gilchrist & Thompson, 1908: 113. Thompson, 1918: 149. Barnard, 1927: 855. Smith, 1945;

541, 1949: 354, pl. 77 fig. 986 and fig. 986.

Clinitrachus superciliosus: Swainson, 1839: 276. Hubbs, 1952: 106 (Clinetrachus).

Blennius versicolor Pappe, 1853: 27.

Blennius mycterizans Gronovius, ed. Gray, 1854: 97.

Blennius ignobilis Gronovius, ed. Gray, 1854: 96.

Clinus dubius Castelnau, 1861: 51.

Clinus pantherinus Castelnau, 1861: 52.

Clinus marmoratus Castelnau, 1861: 52.

Clinus ornatus Gilchrist & Thompson, 1908: 116.

Clinus superciliosus var. arborescens Gilchrist & Thompson, 1908: 115.

Caboclinus superciliosus: Smith, 1966: 73.

Doubtful synonym: Blennius varius Seba, 1758: 90, 93.

Description. D. XXXI-XLII (XXXIV-XXXVI) 5-10 (7-8); A. II 21-30

(24-27); P. 15-18 (15-16); V. I 2; C. 13. Gill-rakers in outer series on first

arch 2-3 + 7-8. Vertebrae 18 + 30-32. Dorsal fin with first three spines con-

siderably elevated to form crest, higher in mature males than in females and

juveniles (table 5). Notch of varying depth in membrane between third and

fourth dorsal spines (table 3). Pectoral fin rounded. Third pelvic ray invariably

absent; present in two specimens of whole sample examined. Caudal peduncle

short, length 24-26-5% head length, depth 29-33:5% head length. Caudal fin

subtruncate. Clusters of cirri usually present at tips of first three dorsal spines.

Body slightly compressed, covered with small embedded scales, more or

less overlapping at least on front half of body, extending on to dorsal and caudal

fin bases; anal fin base and head naked. Depth 4-5 in standard length. Head

3°25-4 in standard length, snout conical to rounded in large specimens. Eye

58 ANNALS OF THE SOUTH AFRICAN MUSEUM

2°5-3'75 in head. Supraorbital tentacle variable, usually small, with narrow

subcylindrical stalk and flattened spatulate tip with a few short, simple branches,

but occasionally more prominent, with fairly long subcylindrical stalk giving

off many fine filamentous branches towards tip. Cirrus on anterior nostril small,

flaplike. Upper jaw 33-55°5% head length, increasing with increase in stan-

dard length. Lips thick. Vomer toothed. Sensory pores of head mostly single,

double in occipital and supraorbital series (fig. 22(b)).

Lateral line of single and few vertical pairs of pores in front to post-pectoral

curve, then of short separate horizontal tubes with pores at either end (fig.

22(d)). Intromittent organ of male with moderately long basal portion and tip

ensheathed by complicated skinny fold, presumably derived from confluent

ventral lips, with two flat dorsal lips above; appears square when retracted

(fig. 22(c)).

Colouring. Very variable. Usually mottled and blotched pattern with con-

spicuous dark, roughly diamond-shaped blotches, lighter in centre, along base

of, and continued on to, dorsal fin. Ground colour usually buff or grey, with

red and darker mottling, but plain scarlet, crimson, bronze and green speci-

mens with fine black speckling, and red, bright green, or olive specimens with

the usual pattern in a darker shade occur. A round dark spot, which may

appear metallic blue when fresh, on dorsal crest. Head dark above, barred and

reticulated below with longitudinal stripes of lighter and darker shades, rising

from lips to eye and curving down again across cheek. Dark, comma-shaped

mark on opercle. Belly mainly white; may have reticulate markings. Branchio-

stegal membranes with reticulate markings. Fins irregularly barred; often with

bright red and orange markings in mature males. Juveniles milky with faint

dark marks along dorsal base. Pattern remains for years after preservation.

Location of type material. Unknown.

Material examined. 1,118 specimens, 29-251 mm in standard length. 57 from

Swakopmund and Milestone 26, S.A.M. 24200; 56 from Walvis Bay, S.A.M.

1381, S.A.M. 9866, S.A.M. 24203; 127 from Liideritzbucht, S.A.M. 24204;

28 from Port Nolloth, S.A.M. 24215; 1 from Kleinsee, Namaqualand, S.A.M.

18225; 17 from Hondeklip Bay, March 1965, S.A.M. not catalogued; 53 from

Gert du Toit Bay and Doring Bay, southern Namaqualand, S.A.M. 24225;

417 from Lambert’s Bay, S.A.M. 23905, S.A.M. 23914; 4 from Hoetjies Bay,

Saldanha Bay, S.A.M. 9868, S.A.M. 9869; 9 from Saldanha Bay, S.A.M. 23970,

S.A.M. 23911; 4 from Robben Island, Table Bay, S.A.M. 9859; 3 from Green

Point, Cape Peninsula, S.A.M. 22779; 88 from Sea Point, S.A.M. 23912,

S.A.M. 23915, and 17/2/1965, S.A.M. not catalogued; 1 from Hout Bay, Cape

Peninsula, S.A.M. 23908; 2 from Kommetjie, Cape Peninsula, S.A.M. 9865;

32 from Froggy Pond and Miller’s Point, False Bay, S.A.M. 23910; 8 from Kalk

Bay, False Bay, S.A.M. 9875 (syntype of ornatus) ; 11 from Dalebrook, False Bay,

S.A.M. 23913, S.A.M. 23916, S.A.M. 23917 and 18/4/1965 and 16/5/1965,

S.A.M. not catalogued; 41 from Strandfontein, False Bay, S.A.M. 23977,

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 59

S.A.M. 24223; 10 from Strandfontein and Dalebrook, 19/2/1965, S.A.M. not

catalogued; g from False Bay, S.A.M. 9871, S.A.M. 9872, S.A.M. 9873

(syntypes of var. arborescens), S.A.M. 9874 (syntype of ornatus); 51 from

Onrust River Mouth, $.A.M. 23904, S.A.M. 24226; 13 from Die Dam, Bredas-

dorp district, S.A.M. 24507, S.A.M. 24547; 10 from Still Bay, S.A.M. 25070;

2 from Mossel Bay, S.A.M. 23909; 5 from Knysna (Leisure Isle and the Heads)

S.A.M. 24227; 44 from Port Elizabeth, $.A.M. 23918, S.A.M. 23986, S.A.M.

24224; 2 from Kidd’s Beach, East London, $.A.M. 25069; 3 from Kei Mouth,

9S.A.M. 25071.

. t =o } .

(i) ventral i) lateral (ii) anteriop

(c)

(b)

Fic. 22. Clinus (Clinus) superciliosus: (a) Lateral view, male, 82-5 mm, S.A.M. 23907; (b) Head

pore system; (c) Intromittent organ of male; (d) Lateral line.

Remarks. Clinus superciliosus is the most abundant South African clinid, and also,

as suggested by the extensive synonymy, the most variable. Clinus ornatus was

described by Gilchrist & Thompson (1908) for nine mature male specimens from

Table Bay; these specimens had rather different markings from the typical

60 ANNALS OF THE SOUTH AFRICAN MUSEUM

form, a better developed supraorbital tentacle, and the dorsal fin originating

further forward than in the usual variety (see table 6). It seems that, as stated

by Smith (1945), these specimens are simply an unusual variety of male

Clinus superciliosus, and unless females and juveniles can be found to correspond

with the ornatus males, they must be included in the species supercaliosus. This

species shows a stronger tendency to sexual dimorphism than the other South

African clinid species, since, apart from the intromittent organ, the crest is

4 2 %

ro Bi S

g g g

0 0 ' ' }

39 38 37 36 35 34 33 32 31 39 38 37 36 35 34 33 32 31 39 38 37 36 35 34 33 3231

dorsal spines dorsal spines dorsal spines

(a) Swakopmund (t) Lambert's Bay (k) Strandfontein

no. of fish

° 8

no. of fish

0 39 38 37 36 35 34 33 32 31 0 39 38 37 36 35 34 33 32 31° c 39 38 37 36 35 34 33 32 31

dorsal spines dorsal spines dorsal spines

(b) Walvis Bay (g) Saldanha Bay (1) Onrust River Mouth

§ i E

& S10 = 10

° b. °o

. 9° ;

° c 2

2 c

Y 39 38 37 36 35 34 33 32 31 P 39 38 37 36 5 i

; } 35 34 33 32 31 39 38 37 36 35 3

dorsal spines dorsal spines dorsal ieee sd

(e} Lideritzbucht (h) Melkbos (m) Bredasdorp dist.

no. of fish

* n

no. of fish

3S 8

“i : - , 0

39 38 37 36 35 34 33 32 31 39 38 37 36 35 34 33 32 31 39 38 37 36 35 34 33 32 31

dorsal spines dorsal spines dorsal spines

(d) Orange River Mouth (i) Sea Point (n) Still Bay

no. of fish

_ nm

oO o

no. of fish

= nN

o o Oo

no, of fish

=) 8

0 0 -

39 38 37 36.35 34 33 32 31 39 38 37 36 35 34 33 32 31 39 38 37 36 35 34 33 32 31

dorsal spines dorsal spines dorsal spines

(e) Port Nolloth (j) Froggy Pond (False Bay) (0) Port Elizabeth

Fic. 23. Histogram showing dorsal spine count variation in different populations of C. superciliosus.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 61

much higher in mature males than in females or juveniles, a condition not

found in any of the other crested South African species (see table 5).

A variety of Clinus superciliosus, C.s. arborescens, was also described by Gil-

christ & Thompson (1908) for specimens which have the larger, more fila-

mentous type of tentacle over the eye, but this also appears to develop only in

some mature males. A few specimens with tentacles of the arborescens type were

taken in a large sample obtained from Onrust River mouth (January, 1963),

and there were no other differences between them and the other specimens of

the sample.

Clinus superciliosus resembles Clinus wood: most closely of all the species of

Clinus, but has higher fin counts, and the dorsal fin origin further back (table 6).

The crest of Clinus woodt, which is also high and well developed, does not appear

to show sexual dimorphism. Clinus superciliosus resembles Clinus robustus mainly

in the size attained and the habitat in which it occurs.

In a study of large samples of Clinus superciliosus, two variations that may

be related to distribution have become apparent. The height of the crest of

mature males appears to be greater east of Cape Point and the number of

dorsal spines tends to be lower. The latter is shown in fig. 23. The crest requires

further study, as no large mature males were obtained east of Hermanus, but

there was a marked difference between Hermanus district specimens and

specimens from Lambert’s Bay, St. Helena Bay, Saldanha Bay and Sea Point

on the coast west of the Cape Peninsula. Jackson (1950) also noted the occur-

rence of such variation in Clinus superciliosus, but did not give measurements.

The fin counts, which do not vary with the size of the fish, show distinct varia-

tion from one population to another (fig. 23), and the east coast populations

show on average a lower number of dorsal spines than the west coast popula-

tions.

Distribution (fig. 19). The known range is Rocky Point (18°59’S., 12°29’)

(South West Africa) to the Kei River (32°41’S., 28°23’E). The species is

exceedingly common from Swakopmund to Port Elizabeth and less common

northwards from Swakopmund and eastwards from Port Elizabeth. Intertidal

and infratidal. Linnaeus (1758) gave the type locality as India, but as his descrip-

tion seems to pertain to the South African species described above, the locality

he gave is certainly erroneous.

TABLE 5. Height of dorsal crest in Clinus superciliosus

Height of crest (% of standard length)

10-12°9% 13-15°9% 16-18:9% 19%

Females

(no. of 63 6 O 0)

Juvenile males

(no. of 25 21 O )

fish)

Mature males

(no. of 2 6 17 32

fish)

62 ANNALS OF THE SOUTH AFRICAN MUSEUM

TaBLE 6. Comparison of Clinus superciliosus (mature males) ‘ornatus’ males, and Clinus woodt

(males and females) (Abbreviations: S.L. = standard length; H.L. = head length; Snt. =

snout; D.O. = dorsal fin origin).

S.L H.L. as Snt.-D.O

(mm) % of S.L as % H.L

superciliosus 163 25°2 78-0

165 26-6 79°5

103 26-7 80:0

112 24°0 96-5

131 26-0 76°5

‘ornatus’ 70 24.°2 67°5

88-5 24°8 68-0

72 25°7 19°9

I51 PB 70:0

166°5 22°8 70°0

14! 24°0 66-0

077) 23°8 62-0

167 22°8 68-5

woodt 81 28-4 61-0

96 29°2 59°0

95 27°4 54°09

112 27°6 61-0

146 27°0 58:0

Clinus (Clinus) taurus Gilchrist & Thompson, 1908

(Fig. 24)

Clinus taurus Gilchrist & Thompson, 1908: 126. Barnard, 1927: 858.

Blenniomimus taurus: Smith, 1945: 539, 1949: 353, pl. 78 fig. 982.

Description. D. XXX-XXXVI (XXXII-XXXIV) 5-6; A. II 20-24 (21-23);

P. 12; V. I 2-3; C. 13. Gill-rakers in outer series on first arch 3-4 + 7-8.

Vertebrae 16 + 28-30. Dorsal fin low, even. Notch in membrane between third

and fourth dorsal spines (table 3). No clusters of cirri at tips of dorsal spines.

Pectoral fin rounded. Inner pelvic ray minute or absent. Caudal peduncle

short, length 20-32% head length, depth 21-23:5°% head length. Caudal fin

subtruncate.

Body slightly compressed, tapering towards tail, covered with small

embedded scales extending on to dorsal and caudal fin bases but not on to anal

fin base or head. Depth 4-5 in standard length. Head very heavy, 3°5-4°25 in

standard length. Snout bluntly rounded. Interorbital markedly concave, heavy

bony ridges over eyes. Occipital groves deep. Eye 3—-4°5 in head. Supraorbital

tentacle prominent, with flattened stalk and terminal fringe of fine cirri,

filamentous. Cirrus on anterior nostril small, flattened, flaplike. Mouth large,

upper jaw 42-53% head length. Lips moderately thin. Vomer toothed. Sensory

pores of head single in nasal and mandibular series except first, double mandi-

bular pore; multiple in all other series (fig. 24(b)). Most pores open on raised

papillae; head very rough.

Lateral line of vertical pairs of pores, with multiple pores in front to post-

pectoral curve, then of short separate horizontal tubes with pore at each end

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 63

(fig. 24(b)). Intromittent organ of male with fairly long basal portion; pair of

crescentic dorsal lips and pair of rounded, confluent ventral lips ensheathing

slender tip (fig. 24(c)).

Colouring. Ground colour in the fresh specimens examined pale green or yellow,

with about seven heavy dark brown cross-bars; head mottled with lilac. Fins

orange-tipped. Plain yellow preserved in alcohol.

Location of type material. South African Museum, Cape Town.

Lectotype. Male, 154. mm standard length, St. James, no other data, S.A.M.

10451.

(a)

ii lateral

Se eee 2

iii anterior

(c)

, 4,

fp SAFER Ha Aut cn re

SIAR Het

(a) ‘%

(b) \,

Fic. 24. Clinus (Clinus) taurus: (a) Lateral view, male, 116 mm, Dalebrook, 12/1/1966, S.A.M.;

(b) Head pore system; (c) Intromittent organ of male; (d) Lateral line.

Material examined. 17 specimens, 74-1885 mm in standard length. 1 from

Lambert’s Bay, S.A.M. 24008; 1 from Kommetjie, Cape Peninsula, $.A.M.

10454 (paralectotype) ; 4 from Strandfontein, False Bay, S.A.M. 24234; 11 from

False Bay, S.A.M. 10451 (lectotype), S.A.M. 10450, S.A.M. 10452, S.A.M.

10453 (paralectotypes), 2 specimens, Cape, Paris Museum Cat. No. A. 2077

(syntypes of C’. cottozdes).

64 ANNALS OF THE SOUTH AFRICAN MUSEUM

Remarks. Clinus taurus resembles Clinus cottoides, Clinus latipennis, Clinus helenae, and

Clinus heterodon in having a concave interorbital, and represents the greatest

development of the bony supraorbital ridges. It attains a larger size than the

other species with a concave interorbital, and differs from them in having

notches in the membrane between the four or five anterior dorsal spines,

deepest between the third and fourth spines. In these respects it resembles

Clinus robustus more closely, although Clinus robustus has a crest. The head

pore system and the anterior part of the lateral line have more numerous pores

than any of the other species of the genus Clinus.

Distribution (fig. 19). The known range of this species is Lambert’s Bay

(32°04'S., 18°20’E) to Port Alfred (33°36’S., 26°54’E). It is rare, and occurs at

the bottom of the intertidal zone and infratidally.

Clinus (Clinus) venustris Gilchrist & Thompson, 1908

(Fig. 25)

Clinus venustris Gilchrist & Thompson, 1908: 130. Barnard, 1927: 861. Smith, 1966: 73.

Ophthalmolophus venustris: Smith, 1945: 542. Smith, 1949: 355, fig. ggo.

Description. D. XXXVII-XLI (XXXIX-XLI) 2-3; A. II 23-28 (24-27);

P. 14; V. I 2; C. 13. Gill-rakers in outer series on first arch 3 + 6-8. Vertebrae

18 (rarely 17) + 28-31. First dorsal spine low, about equal to fourth; second,

and to lesser extent third, dorsal spines elevated, second spine 0-5-3 mm

higher than first or fourth spines. No notch in membrane between third and

fourth dorsal spines, but second and third spines project above membrane.

Clusters of 2-3 cirri at tips of dorsal spines for about half length of fin. Pectoral

fin rounded. Third pelvic ray absent. Caudal peduncle short, length 26-5-33%

head length, depth 23°5-29% head length. Caudal fin subtruncate.

Body slightly compressed, covered with small embedded scales not extend-

ing on to fin bases or head. Depth 4-5 in standard length. Head 3°75-4.75 in

standard length, forehead sloping rather steeply to eyes. Snout bluntly rounded.

Eye 2°25-3°75 in head, noticeably large. Supraorbital tentacle with short,

flattish stalk and flattened tip ending in several short, simple branches. Cirrus

on anterior nostril flattened and spoon-shaped, margin shallowly indented.

Upper jaw 39-48% head length. Lips moderately thick. Vomer toothed.

Sensory pores of head single in nasal and interorbital series, mainly single in

mandibular, preopercular, and postorbital series, double in remaining series

(fig. 25(b)).

Lateral line of vertical pairs of pores in front to post-pectoral curve, then

of short separate horizontal tubes with pore at either end (fig. 25(d)). Intro-

mittent organ of male with moderately long basal portion; confluent pair

of dorsal lips and large pair of ventro-lateral lips, both with serrated inner

margins, ensheathing slender tip (fig. 25(c)).

Colouring. Very variable. Specimens from the coast north of Lambert’s Bay

pale buff with brown streaks, speckles, and reticulations. Bright blue red-

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 65

edged spot over first three dorsal spines. Fins red-tipped, anal fin often entirely

red. Branchiostegal membranes pale with fine black dots. Characteristic dark

line along edge of united gill membranes. Four specimens taken at Sea Point

were plain crimson. Uniform yellow preserved in alcohol.

Location of type material. South African Museum, Cape Town.

Lectotype. Male, 103 mm standard length, Fish Hoek, False Bay, S.A.M. 25238.

Material examined. 43 specimens, 40-106 mm in standard length. 13 from Liide-

ritzbucht, S.A.M 24212; 2 from Port Nolloth, $.A.M. 24217; 1 from Honde-

klip Bay, S.A.M. 25072; 1 from southern Namaqualand, S.A.M. 24017;

7 from Saldanha Bay, S.A.M. 18462; 4 from Sea Point, $.A.M. 23948 and

17/2/1965, S.A.M. not catalogued; 15 from Fish Hoek, False Bay, S.A.M.

= g Se

= SS

oS Swe eee eee

(ii)lateral (iii) anterior

(c)

(i) ventral

ap PRARRARA RAR AAAS LAA A en 4

RR a A 20,

(d) X

(b)

Fic. 25. Clinus (Clinus) venustris: (a) Lateral view, male, 91 mm, S.A.M. 25072; (b) Head pore

system; (c) Intromittent organ of male; (d) Lateral line.

25238 (lectotype), S.A.M. 10543 (paralectotypes).

Remarks. This species is not strikingly similar to any of the other species of

Clinus, and appears to be almost as close to forms such as Clinus superciliosus and

Clinus woodi as to forms such as Clinus acuminatus and Clinus agilis. The form of the

dorsal fin is different from that of any of the other species.

66 ANNALS OF THE SOUTH AFRICAN MUSEUM

Distribution (fig. 19). The known range is from Liideritzbucht (26°28'S.,

15°10’E) (South West Africa) to Port Alfred (33°36’S., 26°54’E). This rare

species occurs infratidally and in pools at the bottom of the intertidal region.

It is taken intertidally more frequently west of Cape Point than from False

Bay eastwards.

Clinus (Clinus) wood: (Smith, 1945)

(Fig. 26)

Petraites woodi Smith, 1945: 540, fig. 2, 1949: 353, pl. 77 fig. 984.

Description. D. XXVII-XXXI 5; A. II 21-24; P. 12-13; V. I 2; C. 13. Verte-

brae 15 + 27. First three dorsal spines elevated, forming high, triangular crest;

third and fourth dorsal spines widely separated, membrane from third spine

barely reaches base of fourth (table 3). First dorsal spine originates far forward,

over hind margin of eye (table 6). Clusters of cirri at tips of at least first three

dorsal spines. Pectoral fin rounded. Third pelvic ray absent. Caudal peduncle

short, length 21-5-33°5% head length, depth 21-5-23°5°% head length. Caudal

fin subtruncate.

Body slightly compressed, covered with small, embedded, more or less

overlapping scales extending on to dorsal and caudal fin bases but not anal fin

base or head, apart from three scales at upper anterior edge of opercle. Depth

3°5-4 in standard length. Head 3-5-4 in standard length, snout rounded, rather

more conical in smaller specimens. Eye 3-4 in head. Supraorbital tentacle with

cylindrical stalk giving off fine filamentous branches towards tip. Cirrus on

anterior nostril short, flattened, bilobed. Mouth large, upper jaw 50-53°5%

head length. Lips moderately thick. Vomer toothed. Sensory pores of head very

small, single in nasal, interorbital and supraorbital series, mainly single in

mandibular series; mainly double, occasionally multiple in remaining series

(fig. 26(b)). |

Lateral line narrow in front, difficult to see in detail, of vertical pairs of

pores or single pores opening above or below line to post-pectoral curve, then

of short separate horizontal tubes with pore at each end (fig. 26(d)). Pores very

minute. Intromittent organ of male with fairly long basal portion; tip

ensheathed by small pair of round dorsal lips and large pair of ventro-lateral

lips, almost confluent, with constriction in centre (fig. 26(c)).

Colouring. No fresh specimens seen. Smith (1945) described the colouring as

‘vivid in marbled olive, brown, and red, with obscure irregular cross-bars. One

or two red oblique bars across cheek.’ Uniform yellow preserved in alcohol.

Location of type material. Department of Ichthyology, Rhodes University,

Grahamstown.

Material examined: 6 specimens, 81-146 mm in standard length. 1 from Xora

Mouth, S.A.M. 24241; 5 from Xora Mouth, R.U.C. (paratypes).

Remarks. Clinus woodt appears to be very close to Clinus superciliosus. The range

of distribution of these two species does not overlap, but they are clearly speci-

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 67

fically distinct. The male Clinus superciliosus described by Gilchrist & Thompson

(1908) as Clinus ornatus approaches Clinus woodi in the forward displacement of

the dorsal origin and the form of the supraorbital tentacle. The intromittent

organ of Clinus woodi approaches the type found in Clinus agilis, and a!so is

rather similar to that of Clinus superciliosus, in which the tip is surrounded by an

almost continuous fold of skin.

Distribution (fig. 19). The known range is Kei Mouth (32°41’S., 28°23’E) to

Inhambane (23°51’S., 35°29’E) (Mozambique). The species is said to be

fairly common (Mrs. M. M. Smith, personal communication).

(i)ventral Gi) lateral

(iii) anterior

(c)

5 wy) s

= ASS

(d) seer

Fic. 26. Clinus (Clinus) woodi: (a) Lateral view, female, 112 mm, Xora Mouth, July 1962, R.U.C.;

(b) Head pore system; (c) Intromittent organ of male; (d) Lateral line.

Genus PAVOCLINUS Smith, 1945

Cristiceps non Valenciennes, Gilchrist & Thompson, 1908: 138 (type-species Cristiceps australis

Valenciennes in Cuvier & Valenciennes, by monotypy).

Pavoclinus Smith, 1945: 545 (type-species Clinus pavo Gilchrist & Thompson, by original designa-

tion).

Labroclinus Smith, 1945: 544 (type-species Cristiceps mentalis Gilchrist & Thompson, by original

designation).

68 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fucomimus Smith, 1945: 544. (type-species Clinus mus Gilchrist & Thompson, by monotypy).

Myxodes non Cuvier, Smith, 1945: 544 (type-species Myxodes viridis Valenciennes in Cuvier &

Valenciennes, by monotypy).

Smithichthys Hubbs, 1952: 107 (type-species Clinus fucorum Gilchrist & ‘Thompson, by

monotypy).

Diagnosis. No tentacle over eye. Lateral line narrow in front, mainly of single

pores opening medially or above and below line to post-pectoral curve, then

of short separate horizontal tubes with pore at either end. Body covered with

small cycloid scales, imbricating or not. Intromittent organ of male with short

basal portion and very large conical tip surrounded at base by single pair of

dorso-lateral lips. Body compressed, sometimes deep. Caudal peduncle usually

elongate. Vomer toothed or not. Usually anterior crest of elevated first three

dorsal spines, separated from rest of fin or not. Exclusively weed-dwelling

species. Mouth fairly small, upper jaw averages less than 40% head length. No

clusters of cirri at tips of dorsal spines.

Discussion

Smith (1945) arranged six species in four genera as follows:

1. Pavoclinus: forms with teeth on the vomer, and the first three dorsal spines

elevated to form a crest, which is not separated from the rest of the fin by a

notch in the membrane. Two species, pavo Gilchrist & Thompson and graminis

Gilchrist & Thompson. Two additional species have subsequently been added

to this group, profundus Smith, 1960 and lztorafontis Penrith, 1965.

2. Labroclinus: forms with teeth on the vomer, and the first three dorsal spines

elevated to form a crest, which is separated from the rest of the fin by a deep

notch, the membrane from the third spine barely reaching the base of the fourth.

Two species, mentalis Gilchrist & ‘Thompson and Jaurenti Gilchrist & Thompson.

3. Fucomimus: forms lacking teeth on the vomer, with two bands of teeth

in each jaw. One species, mus Gilchrist & Thompson.

4. Myxodes: forms lacking teeth on the vomer, with one row of teeth in the

jaw. One species, fucorum Gilchrist & Thompson. Hubbs (1952) pointed out

that, as Myxodes is a South American genus of oviparous Clinidae, lacking a

fleshy penis in the male, and having external fertilization, fucorum, with a penis

in the male and internal fertilization, cannot be a member of the genus Myxodes,

and he proposed a new generic name, Smithichthys, for that species.

I believe that the features uniting the eight species mentioned above are

too numerous and important to allow their separation into more than one

genus, and they are therefore placed together in the genus Pavoclinus. I have

used the name Pavoclinus, which contains all the most typical members of the

genus, because this will necessitate the least number of alterations of name

and additions to the synonymy. The names Fucomimus and Labroclinus have page

preference over Pavoclinus. 'To use either Fucomimus or Labroclinus would result

in extensive additions to the synonymy of the species. In the interest of maintain-

ing the stability of the nomenclature, I therefore claim the privilege of the first

reviser of the group to use Pavoclinus (International Commission for Zoological

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 69

Nomenclature, 1964, article 24(a), recommendation 24A).

The elongation of the caudal peduncle, compression of the body, and the

coloration, including the frequently occurring translucent patches in the fin

membranes, are presumably all modifications for a weed-dwelling existence.

However, these similarities seem to be too many to be the result of convergence,

and appear to indicate close relationship. Clinus brevicristatus and Gynutoclinus

rotundifrons are also habitually weed-dwellers, but these two species do not share

the features common to the Pavoclinus species. There is also considerable uni-

formity in the form of the intromittent organ of the Pavoclinus species. I believe

that division of this group into more than one genus would obscure the relation-

ships of the species.

Four subgenera are used to indicate the relationships of the species within

the genus. Three of these are monospecific, and the fourth contains a closely

related group of species.

The lack of vomerine teeth in the two species mus and fucorum appears to

warrant their separation from the remaining species at a subgeneric level.

Although it is possible that they are more closely related to one another than

to any of the other species, mus and fucorum differ sufficiently from one another

to justify the retention of Smithichthys and Fucomimus as separate subgenera.

Pavoclinus mus differs from P. fucorum in having two rows of teeth in each jaw,

the snout not upturned, the form and position of the dorsal fin origin, and in

having 11 as opposed to 13 primary caudal rays (unique among the South

African Clinidae). —

The species favo, graminis, and laurentii are so similar that even subgeneric

distinction on account of the separation in laurentii of the dorsal crest seems

unwarranted. However, the close association of laurentii with mentalis (Barnard,

1927; Smith, 1945), does not appear to be correct. Pavoclinus graminis, P. pavo,

and P. laurentii all have fewer than 35 dorsal spines and fewer than 25 anal rays,

while P. mentalis is markedly elongate, with a high number of dorsal and anal

elements (35-39 dorsal spines and 29-30 anal rays). P. mentalis has a long skinny

projection at the lower jaw symphysis, and reaches a very large size. On account

of its peculiar features, and to emphasise the particularly close relationship

between the species favo, graminis, laurentii, litorafontis, and probably profundus,

mentalis is placed in a separate subgenus Labroclinus.

Pavoclinus graminis and P. laurentii are the two least strongly modified species,

having a fairly normal body shape rather like that of most species of the sub-

genus Clinus (Clinus). Pavoclinus pavo and P. litorafontis are somewhat more

compressed, with rather longer caudal peduncles. It is proposed to include

these four species in a subgenus Pavoclinus, together with, provisionally, Pavo-

clinus profundus, which is at present known only from the unique type.

The eight species of Pavoclinus are compared in table 7.

Distribution. West (26°38’S., 15°10’E) to east (23°51’S., 35°29’E) coasts of

South Africa (fig. 27), mainly east of Cape Point, only one species, Pavoclinus

pavo, occurring west of Cape Point. Mainly intertidal, in seaweed.

2A9 Jo UISIeUI

Joel EN @)

quasqy

o&-S .z

G.zo-Lb

G.12-GL.€1

G-€

ob—c&

G.>-G.¢

GL.v_-F

a1eINPOP[

(G1) &1-s1

12-61

(9-S) 9-?

(0€-62) 18-gz

wnsomnf

‘o19do01d

JIAO

yuosqy

quosoIg

ale

V.9&

Q-S

€

v.9&

g-¥

6.¢

onUIyy

€1

‘o1adoo1d “o1adoo1d

JIAO JIAO

quosqy yuosqy

yuasolg quosolg

1E—G .gz G.€6-G.cz

C.9b-of GL-G.9G

G.o1-G.G G.91-6

b-& G-C/.§

G.9&—az G.1b-G.€€

G-GL.§ GL.G-¢.7

GL.b-S.§¢ G-G/.§

aInuIyy pedopaasp-T][9

G1 G1

(G1) GI-11 SI-11

(3-02) Gz-oz (Ez) Ss-0z

(&) bz gL

(E-1£) g&-o§ (z&-18) zS-6z

ee ee eee eee

snpunfosg

oand syuofo10j1)

‘d19doa1d

IIAO

daeap ‘yuaso1g

yuasoIg

Gz—ze

Lv-€&

61-9

GL.€-GL.z

1b—9&

¢.¢_p

GL.0-Gz.P

a} 2IIPOp

(SZ—-13)

c-Y

(z&-0&) 8-6

quand]

SS—OG

‘d19do01d

I2AO

quasqy

yuasolg

1$—gz

G.96-S.9z

11-G.9g

GL.€-G.z

9&-0&

C_f

GL.P-G.§

a1 LIIPOP[

€1

(%-zz) b2-16

(S) 9?

(8-z&) SE-o8

SUMUDAS

‘o1adoo1d

JIAO

doap ‘jusso1g

quIsoIg

Gs—0%

ob-S&

G/L.o1-6

GL.G-G/.P

ob—SS

¢.¢-¢.P

CL.v-¥

pedojpeasp-Tj2

ZI-I1

(0-62) o&—Lz

(4-9) 8-9

(ZE-96) 6E-SE

SIDIUIUL

“otodoo1d

JIAO

daap ‘jusso1g

quosqy

G.Lo—bz

GL—09

¢.a1-9

GL.€-G.z

G.€&-Le

¢.v-G.§

c_V

2712 1Opoyy

(41-S1) g1-F1

(&) DE

(La-gz) gz—Gz

SNUL

‘O'd

jo uontsog

‘SP = ION

4199} DUIIOUIO A.

CTH %) ‘G'd'D

CTH %) “T'dO

CTS %) ‘sp 181

"TH ur ohg

CTH %) Mel

yideq Apog

"TS UlcTH

Aes orajod pis

sXKez jepnery

shed [210}99g

shel [euy

shei [esioq

souids [esioq

ee ee

srojoeIeyyy

‘gporedooid = -doaaid Jo -oradoaad {urS110 wy [esiop

‘O'd ‘wdep spunped jepnes = -q:g'y : yisuey aounped yepnes = “q-g°5 ‘auids jesiop

‘sp ‘mel toddn = mef -q ‘y)8u97 peoy = "TH ‘yisusy_ paepurys = -q-9 : SUOTIVIADIQGY

“snutjs0ang JO satsads ay} Jo uostredurory *Z aIavy,

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 71

Fic. 27. Distribution of P. mus (open triangles), P. fucorum (closed triangles), P. mentalis (half-

closed circles), P. pavo (open circles), P. graminis (closed circles), P. litorafontis (closed squares),

P. laurentii (+), P. profundus (open square).

Key to the species of Pavoclinus

Vomer toothed ai ae ae as Ae af: a aye a8 Be |e

Vomer edentate ee a we ie ee itis)

A prominent projecting dips of att on the een jaw at the symphysis

Pavoclinus (Labroclinus) mentalis

No flap of skin at lower jaw symphysis .. xe ae a: BY ae sash es

First three dorsal spines raised to form a crest a Aus si Sis ais

First three dorsal spines not forming a crest ae Pavoclinus (Pavoclinus) profundus

Membrane from third dorsal spine barely reaches base of fourth

Pavoclinus (Pavoclinus) laurentii

Membrane from third dorsal spine reaches more than halfway up fourth Le ha

Inner pelvic ray stout, equal to others; 7-8 dorsal soft rays

Pavoclinus (Pavoclinus) litorafontis

Inner pelvic ray reduced; 6 or fewer dorsal soft rays

Dorsal soft rays 4-6; caudal peduncle less than 40% of ee Tenet snout subconical

; Pavoclinus (Pavoclinus) graminis

Dorsal soft rays 2-4; caudal peduncle more than 40% of head length; snout acute

Pavoclinus (Pavoclinus) pavo

Snout upturned, pug-like; one row of teeth in each jaw; caudal rays 13

Pavoclinus (Smithichthys) fucorum

Snout normal; 2 rows of teeth in each jaw; caudal rays 11 Pavoclinus (Fucomimus) mus

72 ANNALS OF THE SOUTH AFRICAN MUSEUM

Subgenus Fucomimus Smith, 1945

Fucomimus Smith, 1945: 544 (type-species Clinus mus Gilchrist & Thompson).

Diagnosis: Vomer edentate; two rows of teeth in each jaw. First three dorsal

spines elevated to form crest. Membrane from third spine barely reaches base

of fourth. Snout not puglike. Body highly compressed. Scales minute, embedded,

non-imbricating. Lateral line scales not distinct. Caudal peduncle long. Caudal

rays Il.

One species, Pavoclinus (Fucomimus) mus (Gilchrist & Thompson).

Pavoclinus (Fucomimus) mus (Gilchrist & Thompson, 1908)

(Fig. 28)

Clinus mus Gilchrist & Thompson, 1908: 119. Barnard, 1927: 864.

Fucomimus mus: Smith, 1945: 544, 1949: 356, pl. 79 fig. 997 and fig. 997.

Description. D. XXV-XXVIII (XXVI-XXVII) 3-4 (3); A. IT 14-18 (15-17);

P. 10; V. I 3; C. 11. Gill-rakers in outer series on first arch 3-4 + 5-7. Verte-

brae 14 + 27-28. First three dorsal spines elevated to form low crest, well

separated from rest of fin by wide gap between third and fourth dorsal spines;

membrane from third dorsal spine barely reaches base of fourth. Pectoral fin

rather narrow, upper edge straight. Inner pelvic ray always present, about half

length and diameter of other two rays. Dorsal fin with translucent membranes

between groups of 2-4 spines. Caudal peduncle long, length 60-75% head

length, depth 24—27-5°% head length. Caudal fin subtruncate.

Body highly compressed, covered with minute, embedded, isolated scales

not extending on to dorsal, caudal, or anal fin bases, or head. Depth 3°5-4°5

in standard length, deepening with increase in standard length. Head 3°75—5

in standard length, snout rounded. Eye 2-5-4 in head. No supraorbital tentacle.

Cirrus on anterior nostril slightly elongate, curved forward over nostril. Upper

jaw 27-33°5% head length, mouth small. Lips moderately thick. Vomer eden-

tate. Sensory pores of head mostly single in all series (fig. 28(b)).

Lateral line narrow in front, of single, more or less medially opening pores

to post-pectoral curve, then of short separate horizontal tubes with pore at

either end (fig. 28(c)). Intromittent organ of male with short thick basal por-

tion, long, fairly slender conical tip, and pair of lobed dorso-lateral lips (fig.

28(b)).

Colouring. Very variable. Ground colour various shades of green or brown,

mottled and streaked in intricate patterns with yellow, mauve, olive, black,

white, dark green, dark brown, and silver. Dorsal and caudal fins with translu-

cent patches. Other fins green or brown. Belly not lighter than general ground

colour. Juveniles uniform dark brown. Uniform yellow preserved in alcohol.

Location of type material. South African Museum, Cape Town.

Lectotype. Male, 75 mm standard length, St. James, Kalk Bay, no other data,

S.A.M. 25233.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 73

Maiterial examined. 310 specimens, 22-86 mm in standard length. 2 from Froggy

Pond, False Bay, S.A.M. 23893; 1 from Glencairn, False Bay, S.A.M. 23883;

2 from Kalk Bay, False Bay, S.A.M. 17936, S.A.M. 18075; 82 from Dalebrook,

False Bay, S.A.M. 23880, S.A.M. 23895, S.A.M. 23897, and 18/12/1964,

19/2/1965 and 18/4/1965, S.A.M. not catalogued; 9 from St. James, Kalk

Bay, False Bay, S.A.M. 25233 (lectotype), S.A.M. 10531 (paralectotypes) ;

11 from St. James, False Bay, S.A.M. 12021, and 16/5/1965, S.A.M. not cata-

logued; 173 from Strandfontein, False Bay, S.A.M. 23879, S.A.M. 23884,

S.A.M. 23961, S.A.M. 23973, S.A.M. 24246, and 19/2/1965 S.A.M. not

catalogued; 4 from Strandfontein and Dalebrook, False Bay, S.A.M. 23882;

1 from Cape Peninsula, $.A.M. 23890; 1 from Gordon’s Bay, False Bay, $.A.M.

23292; 9 from Onrust River Mouth, $.A.M. 24247; 6 from Still Bay, S.A.M.

25073; 1 from Kidd’s Beach, East London, $.A.M. 25074; 1 from Igoda Mouth,

East London, $.A.M. 25075; 4 from Gonubie Mouth, East London, $.A.M.

25076; 3 from Kei Mouth, S.A.M. 25077.

Remarks. Pavoclinus mus is a highly modified weed-dwelling species. The deep,

strongly compressed body and the elongate caudal peduncle are like those of

Pavoclinus fucorum. I consider the difference in dentition important enough in

conjunction with other features to separate mus and fucorum at the subgeneric

(iii) anterlor

(i) ventral (ii) lateral

(c)

Fic. 28. Pavoclinus (Fucomimus) mus: (a) Lateral view, female, 61 mm, Strandfontein, 19/2/1965,

S.A.M.; (b) Head pore system; (c) Intromittent organ of male; (d) Lateral line.

74 ANNALS OF THE SOUTH AFRICAN MUSEUM

level, especially since Pavoclinus mus has on the whole as many or more features

in common with the species placed in the subgenus Pavoclinus than with Pavo-

clinus fucorum. Pavoclinus mus is rather similar to Pavoclinus pavo in having an

extremely small mouth, a similar, although more exaggerated, body shape, and

a rather delicate build compared with the other species.

Distribution (fig. 27). The known range is False Bay (--34°10’S., 18°50’E) to

the Kei River (32°41’S., 28°23’E). Quite abundant in dense growths of sea-

weed, particularly in the green alga Caulerpa filiformis, intertidal.

Subgenus Labroclinus Smith, 1945

Labroclinus Smith, 1945: 544 (type-species Cristiceps mentalis Gilchrist & Thompson).

Diagnosis: Vomer toothed; two rows of teeth in each jaw. First three dorsal

spines form high crest. Membrane from third dorsal spine barely reaches base

of fourth. Anal rays more than 25. Scales overlapping on anterior half of body.

Caudal rays 13. Prominent skinny flap on lower jaw symphysis.

One species Pavoclinus (Labroclinus) mentalis (Gilchrist & Thompson).

Pavoclinus (Labroclinus) mentalis (Gilchrist & Thompson, 1908)

(Fig. 29)

Cristiceps mentalis Gilchrist & Thompson, 1908: 139.

Petraites mentalis: Barnard, 1927: 866.

Labroclinus mentalis: Smith, 1945: 544, 1949: 357, pl. 81 fig. 998.

Description. D. XXXV-XXXIX (XXXVI-XXXVII) 6-8 (6-7) A. II 27-32

(29-30); P. 11-12; V. 1 3; C. 13. Vertebrae 16 + 37. First three dorsal spines

elevated to form crest, separated from rest of fin by wide gap between third and

fourth dorsal spines. Membrane from third dorsal spine barely reaches base of

fourth. Pectoral fin with upper edge more or less straight. Inner pelvic ray

stout, equal to others. Caudal peduncle fairly long, length 35-40% head length,

depth 20-25% head length. Caudal fin subtruncate.

Body highly compressed, covered with small embedded scales extending

on to dorsal and caudal fin bases but not anal base or head, more or less over-

lapping on front half of body. Depth 4:5—5-5 in standard length. Head 4-4°75

in standard length, snout narrowly pointed. Eye 4:25-5:75 in head. No supra-

orbital tentacle. Cirrus on anterior nostril small, flaplike over nostril. Upper

jaw 35-40% head length. Lips thick. Prominent flap of skin on lower jaw at

symphysis. Vomer toothed. Sensory pores of head mostly single, occipital series

double (fig. 29(b)).

Lateral line of mainly single pores opening above or below line or medially

in front to post-pectoral curve, then of short separate horizontal tubes with pore

at either end (fig. 29(d)). Intromittent organ of male with short basal portion

and large conical tip with pair of crescentic dorso-lateral lips ensheathing it at

base (fig. 29(c)).

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 75

Colouring. No fresh specimens seen. According to Smith (1949), ‘most vivid and

brilliant colouring, rather varied, aptly named ‘Rainbow-fish’. The young with

brilliant silvery iridescent spots along body’. Uniform yellowish preserved in

alcohol.

Location of type material. South African Museum, Cape Town.

Material examined. 8 specimens, 131-224 mm in standard length. 1 from East

London, 8.A.M. 9889 (holotype) ; 4 from East London, R.U.C.; 2 from Bizana

Coast, R.U.C.; 1 from Xora Mouth, R.U.C.

_ Gii) anterior

(i) ventral (ii) lateral

3 (c)

(b) (d) poe

Fic. 29. Pavoclinus (Labroclinus) mentalis: (a) Lateral view, female, 165 mm, East London, R.U.C.;

(b) Head pore system; (c) Intromittent organ of male; (d) Lateral line.

Remarks. Pavoclinus mentalis was placed by Gilchrist & Thompson (1908) in the

genus Cristiceps Valenciennes on account of its fully separated dorsal

crest, although the same authors placed all the other South African Clinidae

then described in the genus Clinus. The genus Cristiceps is Australasian and the

species included in it are characterised by a high, sickle-shaped separate dorsal

crest originating well forward over the eye, an elongate, very slender caudal

peduncle, and a long, simple tentacle over the eye. Pavoclinus mentalis has no

supraorbital tentacle, the crest originates in the usual position over the hind

76 ANNALS OF THE SOUTH AFRICAN MUSEUM

margin of the preopercle, and the caudal peduncle, although somewhat

elongate, is not very slender. The species therefore can not be included in the

genus Cristiceps.

Barnard (1927) commented on the similarity between mentalis and

laurentit and placed them in the genus Peirattes, also Australian. Petraites has a

fringed supraorbital tentacle and is probably not distinct from Clinus, so that the

inclusion of mentalis and laurentw in Petraites was unsuitable. In 1945 Smith

erected the genus Labroclinus for mentalis and laurentu. The association of mentalis

and laurenia by these authors appears to have been based entirely on the separa-

tion of the dorsal crest, and may also have been influenced by the very similar

distribution of the two species. As has been pointed out, neither of these species

should be separated from the species of the genus Pavoclinus, but mentalis is

considered to differ sufficiently from laurentii and the other species to warrant

subgeneric separation.

When Smith (1945) designated mentalis as type-species of Labroclinus he

cited it as ‘Clinus’ mentalis Gilchrist & Thompson, but this is incorrect, as the

name Clinus had never been used for mentalis.

Distribution (fig. 27). The known range is Port Alfred (33°36’S., 26°54’E) to

St. Lucia Bay (28°23’'S., 32°25’E). The species is rare, and occurs infratidally

as well.

Subgenus Pavoclinus Smith, 1945

Pavoclinus Smith, 1945: 545 (type-species Clinus pavo Gilchrist & Thompson).

Diagnosis. Vomer toothed; two rows of teeth in each jaw. First three dorsal

spines forming crest or not. With or without notch in membrane between

third and fourth dorsal spines. Scales usually overlapping on at least front half

of body. Body compressed. Caudal rays 13. No skinny flap on lower jaw sym-

physis. Anal rays less than 25.

Five species. Pavoclinus graminis is the most generalized species, having the

shortest caudal peduncle and the least compression of the body. Pavoclinus

laurentii has a rather longer caudal peduncle but is also not greatly compressed.

Pavoclinus pavo and Pavoclinus litorafontis are highly compressed forms with an

elongate caudal peduncle. Pavoclinus profundus is a peculiar infratidal species

known only from the holotype; it resembles the other species of this subgenus

more closely than it does any of the other clinids, but the dorsal fin is without

any crest and the caudal peduncle is quite short.

Pavoclinus (Pavoclinus) graminis (Gilchrist & Thompson, 1908)

(Fig. 30)

Clinus graminis Gilchrist & Thompson, 1908: 136.

Clinus heterodon non Valenciennes in Cuvier & Valenciennes, Barnard, 1927: 863. Smith, 1935:

217.

Pavoclinus heterodon: Smith, 1945: 545, 1949: 357, pl. 80 figs roor.

Description. D. XXX-XXXV (XXXII-XXXIID) 4-6 (5); A. IL 21-24 (22-29);

P. 12; V. 13; C. 13. Gill-rakers in outer series on first arch 2-3 + 4-5. Verte-

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE TA

brae 15 + 29-31. First three dorsal spines elevated to form crest, fairly low,

second spine highest. No notch in membrane between third and fourth dorsal

spines. Pectoral fin with upper edge more or less straight. Inner pelvic ray

slender, about half length and diameter of others. Caudal peduncle moderate,

26-5-38°5°% head length, depth 26-31% head length. Caudal fin subtruncate.

Body moderately compressed, covered with small embedded scales extend-

ing on to dorsal and caudal fin bases but not anal fin base or head, overlapping

over whole body surface. Depth 4-5 in standard length. Head 3:5-4:75 in

standard length, snout rounded to bluntly conical. Eye 2:5-3:75 in head. No

supraorbital tentacle. Cirrus on anterior nostril very small, flaplike, covering

nostril. Upper jaw 30-36% head length, mouth small. Lips moderately thick.

Vomer toothed. Sensory pores of head mainly single in nasal, interorbital,

mandibular and preopercular series, double or multiple in other series (fig.

30(b)).

Lateral line of single pores opening above or below line in front to post-

pectoral curve, then of short separate horizontal tubes with pore at either

end (fig. 30(d)). Intromittent organ of male with short basal portion and large

conical tip emerging between pair of curved more or less confluent dorso-lateral

lips (fig. 30(c)).

Colouring. Very variable, green, brown, red, often mottled, streaked or with

cross-bars of yellow, silver, or deeper shades of the ground colour; pearly spots

on body in young specimens, which are usually otherwise plain red. Fins plain

or barred, with small translucent areas particularly on dorsal and caudal fins.

Belly usually not lighter than ground colour. Colours tend to be brighter and

more variable further east along the coast. Uniform yellow preserved in

alcohol.

Location of type material. South African Museum, Cape Town.

Lectotype. Female, 126 mm, St. James, donated by Mr. Langschmidt, $.A.M.

25240.

Material examined. 100 specimens 24-141 mm in standard length. 1 from Miller’s

Point, False Bay, S.A.M. 23899; 8 from Dalebrook, False Bay, S.A.M. 23806,

S.A.M. 23900 and 18/12/1964 and 18/4/1965 S.A.M. not catalogued; 12 from

St. James, S.A.M. 25240 (lectotype), S.A.M. 10523 (paralectotypes) and

16/5/1965, S.A.M. not catalogued; 12 from Strandfontein, False Bay, S.A.M.

23885, S.A.M. 23901, S.A.M. 24249; 4 from Still Bay, S.A.M. 25078, 4 from

Kidd’s Beach, East London, $.A.M. 25079; 4 from Igoda Mouth, East London,

.S.A.M. 25080; 22 from East London, $.A.M. 23903, S.A.M. 25081; 4 from

Gonubie Mouth, $.A.M. 25083; 17 from Port St. John’s, S.A.M. 23902.

Remarks. Clinus heterodon was described from the Cape by Valenciennes (1836),

and listed by Gilchrist & Thompson (1908), who stated that they had not

rediscovered the species here. Barnard (1927) identified the specimens from

which they described graminis as C. heterodon of Valenciennes. However, the

original description of C. heterodon was so inadequate that one could not base

78 ANNALS OF THE SOUTH AFRICAN MUSEUM

an identification on it without examining the holotype. I found that the holo-

type is a tentacled clinid of the genus Clinus. Gilchrist & Thompson’s name

graminis therefore stands for the species of the genus Pavoclinus described above.

Distribution (fig. 27). The known range is False Bay (+34°06'S., 18°50’E) to

Inhambane (23°51'S., 35°29’E) (Mozambique), becoming more common

eastwards. Usually intertidal, in weed. Smith (1945 and 1949) gave Port

Nolloth as the north-western limit of this species, but its occurrence anywhere

west of Cape Point has not been confirmed, and it seems certain that the record

Was an error.

(ii) lateral (iii) anterior

(c)

Fic. 30. Pavoclinus (Pavoclinus) graminis: (a) Lateral view, female, 115 mm, S.A.M. 25083

(b) Head pore system; (c) Intromittent organ of male; (d) Lateral line.

Pavoclinus (Pavoclinus) laurentii (Gilchrist & Thompson, 1908)

(Fig. 31)

Clinus laurentii Gilchrist & Thompson, 1908: 120. Smith, 1935: 218, fig. 5.

Petraites laurentii: Barnard, 1927: 866.

Labroclinus laurentii: Smith, 1945: 544, 1949: 357, pl. 80 fig. 999 and fig. 999.

Description. D. XXIX-XXXITI (XXX-XXXII) 4-5; A. II 20-22 (21-22);

P. 12; V. 1 3; C. 13. Gill-rakers in outer series on first arch 2-3 + 4-6. Verte-

brae 15 + 27-29. First three dorsal spines elevated to form crest. Membrane

from third to fourth dorsal spine barely reaches base of fourth. Crest moderate,

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 79

second spine highest. Pectoral fin somewhat rounded, but upper edge fairly

straight. Inner pelvic ray about half length of other two, very slender. Caudal

peduncle moderate, length 33-47% head length, depth 22-33% head length.

Caudal fin subtruncate.

Body moderately compressed, covered with small embedded scales extend-

ing on to dorsal and caudal fin bases but not anal fin base or head, overlapping

on front half of body. Depth 4-5-5 in standard length. Head 4:25-4:75 in

standard length, snout rounded to bluntly conical. Eye 2°75-3-75 in head.

No supraorbital tentacle. Cirrus on anterior nostril small, flaplike. Upper jaw

36-41% head length. Lips moderately thick. Vomer toothed. Sensory pores

of head mostly double, many multiple pores in occipital region (fig. 31(b)).

(ii) lateral (iii) anterior

(c)

(i) ventral

(d) Ne

(b)

Fic. 31. Pavoclinus (Pavoclinus) laurentii: (a) Lateral view, male, 110 mm, Xora, May 1948,

R.U.C.; (b) Head pore system; (c) Intromittent organ of male; (d) Lateral line.

Lateral line of vertical pairs of pores or single pores opening above or

below line in front to post-pectoral curve, then of short separate horizontal

tubes with pore at either end (fig. 31(d)). Intromittent organ of male with

short basal portion and large, conical tip, ensheathed at base by pair of con-

fluent dorso-lateral lips (fig. 31(c)).

Colouring. No fresh specimens seen. Smith (1949) described the colouring as

‘most variable, but beautiful and delicate colours and markings, brown, red,

yellow, green, mottled and marbled.’ Uniform buff to yellow preserved in

alcohol.

80 ANNALS OF THE SOUTH AFRICAN MUSEUM

Location of type material. South African Museum, Cape Town.

Lectotype. Female, 108 mm standard length, Winkelspruit, Natal, collected by

L. Robinson, donated by R. Robinson, $.A.M. 25234.

Material examined. 15 specimens, 66-117 mm in standard length, 3 from Xora

Mouth, R.U.C.; 1 from Bizana Coast, R.U.C.; 1 from Port St. John’s, R.U.C.;

1 from Isipingo, R.U.C.; 3 from Winkelspruit, Natal, $.A.M. 25234 (lectotype),

S.A.M. 9888 (paralectotypes); 4 from Natal, S.A.M. 10927, S.A.M. 16157,

S.A.M. 18515; 1 from Inhaca Island, Mozambique, R.U.C., 1 from Inhambane,

Mozambique, R.U.C.

Remarks. Pavoclinus laurentii is similar to Pavoclinus mentalis in the separation of

the dorsal crest and in its distribution, but resembles Pavoclinus graminis in body

shape, head shape, fin counts, and relative size of eye, and appears to be closer

to the more generalized species than to Pavoclinus mentalis.

Distribution (fig. 27). Port Alfred (33°36’S., 26°54’E) to Inhambane (23°51’S.,

35°29’E) is the known range. The species is said to be not uncommon (Smith,

1949) in weedy pools.

Pavoclinus (Pavoclinus) litorafontis Penrith, 1965

(Fig. 32)

Pavoclinus litorafontis Penrith, 1965a: 212, pl. 5, and figs 1 and 2.

Description. D. XXIX-XXXII (XXXI-XXXII) 7-8; A. II 20-23 (23); P.

11-12; V. 13; C. 13. Gill-rakers in outer series on first arch 3-4 + 5-6. Verte-

brae 17 + 32-33. First three dorsal spines elevated to form crest, decreasing

in relative height with increase in size of fish. No notch in membrane between

third and fourth dorsal spines. Upper edge of pectoral fin more or less straight.

Inner pelvic ray stout, equal to others. Caudal peduncle long, length 58-5-75%

head length, depth 22-5~33-5°% head length. Caudal fin subtruncate.

Body compressed, covered with small embedded scales extending on to

dorsal and caudal fin bases but not anal fin base or head, overlapping over most

of body surface. Depth 4°5-5:75 in standard length. Head 3:75—5 in standard

length, snout subconical. Eye 3°5—4:25 in head. No supraorbital tentacle. Cirrus

on anterior nostril very small. Upper jaw 33°5-41°% head length. Lips mod-

erately thick. Vomer toothed. Sensory pores of head mostly single (fig. 32(b)).

Lateral line of single pores in front to post-pectoral curve, then of short

separate horizontal tubes with pore at either end (fig. 32(d)). Intromittent

organ of male with short basal portion and large, conical tip; single pair of

crescentic dorso-lateral lips ensheathing base of tip (fig. 32(c)).

Colouring. Green, striped in shades of darker green, yellow, and silver longitu-

dinally. Usually a yellow or silver stripe across cheek. Fins green, usually with

translucent patches; always translucent patches in the caudal fin. Belly no

lighter than ground colour, underparts green. Juveniles green or brown with a

broad silver longitudinal stripe mid-laterally. Uniform yellow preserved in

alcohol.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 81

Location of type material. South African Museum, Cape Town.

Material examined. 22 specimens, 27—190 mm in standard length. 14 from Strand-

fontein, False Bay, S.A.M. 23876, S.A.M. 23877, S.A.M. 23952, S.A.M. 23962,

S.A.M. 23972, S.A.M. 24052 (holotype), S.A.M. 24248, and March, 1965,

S.A.M. not catalogued; 8 from Onrust River Mouth, S.A.M. 24081, S.A.M.

24.257.

Remarks. Pavoclinus litorafontis appears to be the most strongly modified member

of the subgenus Pavoclinus, and has the most elongate caudal peduncle in that

subgenus. It is closest to Pavoclinus graminis in the form of the dorsal fin and the

(ii) lateral

G) ventral

@®)

(iii) anterior

(c)

“a,

(b) ‘

(d) Se

sae

Fic. 32. Pavoclinus (Pavoclinus) litorafontis: (a) Lateral view, male, 89 mm, S.A.M. 23876;

(b) Head pore system; (c) Intromittent organ of male; (d) Lateral line.

head. It resembles Pavoclinus mentalis in the body proportions but not the fin

counts other than the number of soft dorsal rays and the strong development

of the inner pelvic ray. It attains a larger size than the other species of the

subgenus Pavoclinus. Juveniles have a very poorly developed intromittent organ

and it is usually impossible to sex small specimens externally.

Distribution (fig. 27). The known range is False Bay (-34°10’S., 18°50’E) to

Cape Agulhas (34°50’S., 20°00’E); not common, in sea-weed low down on

the shore, particularly in beds of the green alga Caulerpa filiformis.

82 ANNALS OF THE SOUTH AFRICAN MUSEUM

Pavoclinus (Pavoclinus) pavo (Gilchrist & Thompson, 1908)

(Fig. 33)

Clinus pavo Gilchrist & Thompson, 1908: 123. Barnard, 1927: 857.

Pavoclinus pavo: Smith, 1945: 545, 1949: 357, pl. 81 figs 1000.

Description. D. XXX-XXXVI (XXXI-XXXIII) 2-4 (3); A. IT 20-23 (20-22);

P. 11-12 (12); V. I 2-3; C. 13. Gill-rakers in outer series on first arch 2-4 + 4-6.

Vertebrae 15 + 28-31. First three dorsal spines elevated to form crest, second

spine longest. No notch in membrane be'ween third and fourth dorsal spines.

Pectoral fin with upper edge more or less straight. Inner pelvic ray minute or

absent. Caudal peduncle long, 40-47 % head length, depth 20-30% head length.

Caudal fin subtruncate.

Body compressed, covered with small embedded scales extending on to

base of dorsal fin but not caudal and anal fin bases or head, overlapping on at

least front half of body. Depth 3-75-5 in standard length, body deepening with

increase in standard length. Head 3:5-4:75 in standard length, snout narrow,

acutely pointed. Eye 3-4 in head. No supraorbital tentacle. Cirrus on anterior

nostril flattened, flaplike, curving forward over nostril. Upper jaw 22-36:5%

head length, mouth small. Lips thick. Vomer toothed. Sensory pores of head

mostly single, occipital series and a few of preopercular and suborbital series

double (fig. 33(b)).

Lateral line of single pores opening above and below line in front to post-

pectoral curve, then of short separate horizontal tubes with pore at either end

(fig. 33(d)). Intromittent organ of male with moderate basal portion and large,

conical tip, swollen at base, and surrounded basally by large fleshy pair of

ovoid dorso-lateral lips (fig. 33(c)).

Colouring. Variable, green, brown or red, sometimes mottled and striped longi-

tudinally with yellow, lighter shades of the main colour and silvery white.

Belly of same shade as ground colour. Often pearly spots on body. Dorsal and

caudal fin with translucent patches. A specimen taken from kelp at Hondeklip

Bay was olive yellow, with three turquoise blue ocellate spots on side and a

turquoise blue line across cheek. Uniform yellow preserved in alcohol.

Location of type matertal. South African Museum, Cape Town.

Lectotype. Male, 88 mm standard length, St. James/Kalk Bay area, intertidal

pools, no other data, S.A.M. 25236.

Material examined. 59 specimens, 15-111 mm in standard length. 1 from Shear-

water Bay, Liideritzbucht, $.A.M. 24213; 2 from Port Nolloth, S.A.M. 24218;

3 from Hondeklip Bay, S.A.M. 25084; 3 from Lambert’s Bay, S.A.M. 23250;

1 from Langebaan, S.A.M. 21476; 1 from Melkbosch, 20/10/1964, S.A.M. not

catalogued; 1 from Table Bay, S.A.M. 4728; 1 from Miller’s Point, False Bay,

S.A.M. 23894; 1 from Glencairn, False Bay, S.A.M. 23891; 1 from Kalk Bay,

False Bay, S.A.M. 17935; 9 from Dalebrook, False Bay, S.A.M. 23888, S.A.M.

23898, and 18/4/1965, S.A.M. not catalogued; 2 from St. James/Kalk Bay,

False Bay, S.A.M. 25236 (lectotype), S.A.M. 10532 (paralectotype); 4 from

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 83

St. James, 16/5/1965, S.A.M. not catalogued; 3 from Onrust River Mouth,

S.A.M. 24251; 15 from Kidd’s Beach, East London, S.A.M. 25085; 4 from

Igoda Mouth, East London, $.A.M. 25086; 1 from East London, S.A.M.

23892; 6 from Kei Mouth, $.A.M. 25087.

Remarks. Pavoclinus pavo is apparently the only species of the genus Pavoclinus

occurring on the coast west of Cape Point, although it is much more common

east of Cape Agulhas. It is smaller and more delicately built than most of the

other species which occur intértidally.

Distribution (fig. 27). The known range is Liideritzbucht (26°38’S., 15°10’E)

(South West Africa) to the Kei River (32°41’S., 28°23’E). It is nowhere very

(ii) lateral (ii) anterior

(c)

(i) ventral

(b) \

Fic. 33. Pavoclinus (Pavoclinus) pavo: (a) Lateral view, male, 60 mm, S.A.M. 23888; (b) Head

pore system; (c) Intromittent organ of male; (d) Lateral line.

common, but is not uncommon towards the eastern limit of its range. It lives

intertidally in sea-weed. All the west coast specimens were taken in kelp, but

in False Bay and localities eastward most of the specimens were found curled

round the bases of short bushy growths of algae such as Hypnea spicifera; also in

Caulerpa beds. | |

Pavoclinus (Pavoclinus) profundus Smith, 1960

| (Fig. 34)

Pavoclinus profundus Smith, 1960: 689, fig. 1.

Description. D. XXX 4; A. IT 21; P. 12; V. 1 3; CG. 13. Vertebrae 14 + 27.

First three dorsal spines not elevated to form crest, but fourth spine slightly

84 ANNALS OF THE SOUTH AFRICAN MUSEUM

shorter than third or fifth. No notch in membrane between third and fourth

dorsal spines. Upper edge of pectoral almost straight. Inner pelvic ray minute,

bound to outer rays by membrane, tip not free. Caudal peduncle moderate,

length 36-4°% head length, depth 27-2°% head length. Caudal fin subtruncate.

Body highly compressed, sparsely covered with small embedded scales, not

overlapping. Head naked. Depth at anal origin 4:8 in standard length. Head

39 in standard length, snout acutely pointed. Eye 3 in head. No supraorbital

tentacle. Cirrus on anterior nostril a small flap. Upper jaw 36-4°% head length,

mouth small. Lips moderately thin. Vomer toothed. Sensory pores of head

mostly single (fig. 34(b)).

Lateral line of single pores opening above or below line or medially in

front to post-pectoral curve, then of short separate horizontal tubes with pore

at either end, becoming obscure on caudal peduncle (fig. 34(d)). Intromittent

organ of male with short basal portion and large conical tip; pair of elongate

fleshy dorso-lateral lips partially ensheathing tip (fig. 34(c)).

Colouring. Only the preserved type seen. Detailed description of fresh colouring

given by Smith (1960). Uniform buff preserved in alcohol.

(i) ventral (ii) lateral (iii) anterior

(c)

ee ee ee ee ne ee =!

Ae,

(Bee (d)

Fic. 34. Pavoclinus (Pavoclinus) profundus: (a) Lateral view, male, 43 mm, R.U.C., holotype;

(b) Head pore system; (c) Intromittent organ of male; (d) Lateral line.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 85

Location of type material. Department of Ichthyology, Rhodes University,

Grahamstown.

Material examined. 1 specimen, 43 mm in standard length, Knysna, 20 fms,

R.U.C. (holotype).

Remarks, This species is known only from the holotype, and in spite of the lack

of a marked dorsal crest it has so many features in common with the other

members of the subgenus Pavoclinus that it is placed in that subgenus. Not least

of these similarities is the form of the intromittent organ. The holotype was

stated by Smith (1960) to be a female, but the specimen is clearly a male with

an intromittent organ of a type very similar to that of the other species of

Pavoclinus.

Disiribution (fig. 27). Known from one locality, Knysna (34°03’S., 23°03’E);

dredged in 20 fathoms from a rocky bank.

Subgenus Smithichthys Hubbs, 1952

Myxodes non Cuvier, Smith, 1945: 544 (type-species Myxodes viridis Valenciennes in Cuvier &

Valenciennes).

Smithichthys Hubbs, 1952: 107 (type-species Clinus fucorum Gilchrist & Thompson).

Diagnosis. Vomer edentate; single row of teeth in each jaw. Snout upturned,

puglike. First four or five dorsal spines elevated to form high, rounded dorsal

crest. Body highly compressed. Scales minute, non-imbricating. Caudal

peduncle elongate; caudal rays 13.

One species, Pavoclinus (Smithichthys) fucorum (Gilchrist & Thompson).

Pavoclinus (Smithichthys) fucorum (Gilchrist & Thompson, 1908)

(Fig. 35)

Clinus fucorum Gilchrist & Thompson, 1908: 121. Barnard, 1927: 853.

Myxodes fucorum: Smith, 1945: 544, 194.9: 356, pl. 81 fig. go6.

Smithichthys fucorum: Hubbs, 1952: 107. Smith, 1953: 356, pl. 81 fig. 996.

Description D. XXVIII-XXXI (XXIX-XXX) 4-6 (5-6); A. II 19-21; P.

12-13 (12); V. 13; GC. 13. Vertebrae 15 + 28-29. First dorsal spine originates

well forward, over hind margin of eye. First four or five dorsal spines elevated

to form high rounded crest. No notch in dorsal fin membrane between any of

the spines; all spines evenly placed. Dorsal spines of varying lengths, giving

outline of fin markedly undulating appearance. Pectoral fin more or less

rounded, upper edge somewhat straight. Inner pelvic ray stout, almost equal

to others Caudal peduncle long, length 47—62-5% head length, depth 23°5-30%

head length. Caudal fin subtruncate, very large.

Body highly compressed, deep, covered with small embedded scales not

extending on to fin bases or head, not overlapping. Depth 3-5-4 in standard

length. Head 44:75 in standard length, snout rounded, upturned, puglike.

Eye 3-5 in head. No supraorbital tentacle. Cirrus on anterior nostril short,

with expanded trilobate tip. Upper jaw 32-42% head length. Lips fairly

86 ANNALS OF THE SOUTH AFRICAN MUSEUM

thick. Vomer edentate; single row of teeth in each jaw. Sensory pores of head

mostly single, a few double in supraorbital and suborbital series (fig. 35(b)).

Lateral line of single pores opening more or less medially in front to post-

pectoral curve, then of short separate horizontal tubes with pore at either end

(fig. 35(d)). Intromittent organ of male with moderate basal portion and long,

conical tip, ensheathed basally by wing-like pair of dorso-lateral lips (fig. 35(c)).

Colouring. Uniform dark brown or olive, edges of fins lighter or transparent; two

irregular translucent patches in caudal fin. A bright silvery iridescent spot

behind pectoral fin. Uniform yellow preserved in alcohol.

Location of type material. South African Museum, Cape Town.

Lectotype. Male, 160 mm standard length, St. James, no other data, S.A.M.

29239-

Material examined. 13 specimens, 42-196 mm in standard length. g from St.

James, False Bay, S.A.M. 25235 (lectotype), S.A.M. 10524 (paralectotypes),

(i) ventral

(ii) lateral

(c)

(d) ~

poe

(b)

Fic. 35. Pavoclinus (Smithichthys) fucorum: (a) Lateral view, male, 101 mm, St. James, 16/5/1965,

S.A.M.; (b) Head pore system; (c) Intromittent organ of male; (d) Lateral line.

- ae

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 87

S.A.M. 12020, and 16/5/1965, S.A.M. not catalogued; 2 from Muizenberg,

False Bay, S.A.M. 10525; 2 from Strandfontein, False Bay, S.A.M. 23886 and

March, 1965, S.A.M. not catalogued.

Remarks. Pavoclinus fucorum is a peculiar, rare, highly specialised weed-dwelling

species. It resembles Pavoclinus mus in lacking vomerine teeth, but otherwise

differs from it in as many ways as it does from the members of the other sub-

genera of Pavoclinus.

Distribution (fig. 27). The known range is False Bay (+34°10’S., 18°50’E) to

the Bashee River (31°55’S., 28°27’E); rare, in sea-weed. At the bottom of the

intertidal region and infratidal.

Genus BLENNIOCLINUS Gill, 1860

Blennioclinus Gill, 1860: 103 (type-species Clinus brachycephalus Valenciennes in Cuvier & Valen-

ciennes, by monotypy).

Diagnosis: No supraorbital tentacle. Last dorsal spine about half length of first

dorsal ray, deep notch in profile of dorsal fin before dorsal soft rays. Lateral line

of double pores anteriorly to post-pectoral curve, then of short separate hori-

oss F ee.%e -

eet : .

ee 8 oe ote 0 ree ete et eee. ws

lo ot 0 oo ote teen tS CAE

wy +6

¢ at [e)

Fic. 36. Distribution of B. brachycephalus (open circles), B. stella (closed circles), Clinoporus

biporosus (closed squares), G. rotundifrons (half-closed circles), Cancelloxus burrelli (open triangles),

X. leprosus (closed triangles), X. kochi (open squares).

88 ANNALS OF THE SOUTH AFRICAN MUSEUM

zontal tubes with pore at either end. Intromittent organ of male with short or

moderate basal portion and long slender tip concealed between pair of very

large lateral lips. Body covered with small embedded cycloid scales, not extend-

ing on to fin bases or head. First three dorsal spines elevated to form low crest.

Notch in membrane between third and fourth dorsal spines. Clusters of cirri

at tips of at least anterior dorsal spines. Vomer toothed. Head short, profile

rounded.

Two species.

Discussion. The genus Blennioclinus was included by Smith (1945) in his Myxo-

dinae on account of the lack of a supraorbital tentacle. However, the two

species brachycephalus and stella are not weed-dwellers and are further set apart

from the genus Pavoclinus by the notch in the profile of the dorsal fin before the

dorsal soft rays, the clusters of cirri at the tips of the dorsal spines, the body form,

the type of intromittent organ, and the high number of double pores in the

anterior part of the lateral line. They differ from the genus Clinus mainly in

having a notch before the dorsal soft rays, and in lacking a supraorbital tentacle.

I therefore regard the genus Blennioclinus as distinct from the other clinid genera.

Distribution. West and east coasts of South Africa (fig. 36).

Key to the species of Blennioclinus

1. 8-11 dorsal soft rays ey) aN vf a? ae Blennioclinus brachycephalus

6-7 dorsal soft rays - ie a We oe ad Blennioclinus stella

Blennioclinus brachycephalus (Valenciennes in Cuvier & Valenciennes, 1836)

(Fig. 37)

Clinus brachycephalus Valenciennes in Cuvier & Valenciennes, 1836: 371. Gilchrist & Thompson,

1908: 135. Barnard, 1927: 852.

Labrisomus linearis Swainson, 1839: 277.

Blennioclinus brachycephalus: Gill, 1860: 103. Smith, 1945: 543, 1949: 356; pl. 79 fig. 995.

Description. D. XXVI-XXX (XXVITI-XXIX) 8-11 (9-11); A. II 19-25

(21-24); P. 12-15 (13-14); V. I 3; C. 13. Gill-rakers in outer series on first

arch 2-3 + 3-5. Vertebrae 17 + 28-30. First three dorsal spines slightly

elevated, first spine o-2 mm longer than fourth, crest very low. Third and fourth

dorsal spines more widely spaced than others, very shallow notch in membrane

between them. Dorsal spines decrease in height posteriorly, last spine very

short, dorsal soft rays much higher; deep notch in profile of dorsal fin before

dorsal soft rays. Clusters of 3-4 cirri at tips of most dorsal spines. Pectoral fin

rounded. Inner pelvic ray short, slender. Caudal peduncle moderate, 35-40%

head length, depth 23-31% head length. Caudal fin subtruncate.

Body slightly compressed, covered with minute embedded scales not

extending on to dorsal, caudal, or anal fin bases or head. Depth 5-5-75 in

standard length. Head short, bluntly rounded, 4-5 in standard length. Eye

2°5-3°5 in head. No supraorbital tentacle. Cirrus on anterior nostril flattened,

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 89

irregularly trilobed. Upper jaw 28-5-34% head length, mouth small. Lips

thin. Vomer toothed. Sensory pores of head mostly double, with multiple pores

mainly in occipital series (fig. 37(b)).

Lateral line of about 30-35 multiple pores in front to post-pectoral curve,

then of short separate horizontal tubes with pore at either end (fig. 37(d)).

Intromittent organ of male with moderate basal portion and slender tip

ensheathed by large pair of lobed dorso-lateral lips (fig. 37(c)).

Colouring. Ground colour creamy fawn, much overlain with dark brown blotches

and reticulations, denser dorsally. Belly creamy or bright yellow, with dark

band running from isthmus towards vent, having broad dark symmetrical

vertical branches running up to the densely mottled area above. Sometimes

(iii) arterior

Lp (ii) lateral

(i) ventral

(c)

el ew

ANNs Uy

Sr a

& 3 (d) SL

Fic. 37. Blennioclinus brachycephalus: (a) Lateral view, male, 73 mm, S.A.M. 23974; (b) Head

pore system; (c) Intromittent organ of male; (d) Lateral line.

fine orange stripes alternate with dark bands. A red-edged bright blue semi-

circular spot behind pectoral fin. Head dark above to lighter brown below.

Two mauve wavy lines on preopercle and opercle. Dark irregular band from

below eye to angle of jaw. Spinous dorsal dark brown, with red tips to spines.

Dorsal soft rays dark brown, membrane joining them cream barred with brown

between first five rays, then plain cream. Anal fin with broad brown cross-bars,

ground colour creamy, tips of rays red. Pectoral fin cream with dark brown

cross-bars. Pelvic fins dark at base, then cream with two narrow red bars

distally. Caudal fin cream with irregular dark brown markings. Branchiostegal

gO ANNALS OF THE SOUTH AFRICAN MUSEUM

membranes mauve, with dark brown cross-bars. Juveniles plain dark red with

longitudinal row of large silvery spots along side. Pattern remains defined for

some time after preservation, slowly fading to yellow-buff.

Location of type material. Muséum National d’Histoire Naturelle, Paris.

Material examined. 223 specimens, 17-100 mm in standard length. 1 from Cape,

Paris Museum Cat. No. A2013 (holotype); 12 from Liideritzbucht, S.A.M.

24210; 2 from Simon’s Bay, False Bay, S.A.M. 10534; 13 from Dalebrook,

False Bay, 18/12/1964, 19/2/1965, 18/4/1965 and 16/5/1965, S.A.M. not cata-

logued; 2 from St. James, False Bay, S.A.M. 12014; 146 from Strandfontein,

False Bay, S.A.M. 23875, S.A.M. 23953, S.A.M. 23970, S.A.M. 23974 and

20/11/1964, and 13/8/1964, S.A.M. not catalogued; 1 from Gordon’s Bay,

False Bay, S.A.M. 23293; 1 from Die Dam, Bredasdorp district, S.A.M. 25088;

25 from Still Bay, S.A.M 18265, S.A.M. 25089; 1 from Port Elizabeth, S.A.M.

24244; 2 from East London, $.A.M. 18092; 2 from Kei Mouth, 8.A.M. 25090.

Remarks. Blennioclinus brachycephalus is the more widely distributed and common

of the two species of Blennioclinus. It is also the larger. ‘The notched profile of

the dorsal fin is interesting, since the same condition is found in some of the

American Labrisominae, e.g. Labrisomus and Malacoctenus, but not in any of the

other Clininae.

Distribution (fig. 36). Liideritzbucht (26°38’S., 15°10’E) (South West Africa)

to the Kei River (32°41’'S., 28°23’E); locally common east of Cape Point, in

pools at the low tide mark, usually among pebbles and sea-urchins. The species

has not been found at any localities on the west coast between Melkbosch (west

coast of Cape Peninsula) and Liideritzbucht (Penrith, 19655). The Lideritz-

bucht specimens were all collected in sheltered areas, not on the open coast.

They are in no way different from the south coast populations.

Blennioclinus stella Smith, 1945

(Fig. 38)

Blennioclinus stella Smith, 1945: 543, fig. 4, 1949: 386, pl. 79 fig. 994.

Description. D. XXIV-—XXVIII 6-7; A II 19-21; P. 12; V. I 3; C. 13. Gill-

rakers in outer series on first arch 1-2 + 4-6. Vertebrae 13 + 25-27. First

three dorsal spines elevated to form low crest. Gap between third and fourth

spines wider than between other spines; deep notch in membrane between

third and fourth spines. Dorsal spines decrease in height posteriorly, last spine

very short, dorsal soft rays much longer, resulting in deep notch in profile of

dorsal fin before dorsal soft rays. Clusters of cirri at tips of first three dorsal

spines. Pectoral fin rounded. Inner pelvic ray minute. Caudal peduncle short,

length 27°5-33% head length, depth 27-5-33:5°% head length. Caudal fin

subtruncate.

Body slightly compressed, covered with small embedded scales not extend-

ing on to dorsal, caudal, or anal fin bases or head. Depth 5:25-5:75 in standard

length. Head 3°75-5:25 in standard length, snout rounded. Eye 2-5-3:25 in

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE gi

head. No supraorbital tentacle. Cirrus on anterior nostril elongate, narrow

proximally, with tip expanded and deeply serrate. Upper jaw 37°5-47:5%

head length. Lips moderately thick. Vomer toothed. Sensory pores of head

mostly double, single in nasal, interorbital and in most of preopercular series

(fig. 38(b)).

Lateral line of 20-25 vertical pairs of pores in front to post-pectoral curve,

then of short separate horizontal tubes with pore at either end (fig. 38(d)).

Intromittent organ of male with slender tip ensheathed by pair of large, flat

ventro-lateral lips, with slightly serrate margins (fig. 38(c)).

Colouring. No live or unpreserved specimens seen. Smith (1945) described the

colouring as ‘variably brown, red and orange mottled. Three narrow bars

(b) NS

T= > exp

Fic. 38. Blennioclinus stella: (a) Lateral view, male, 39:5 mm, Xora, August 1947, R.U.C.;

(b) Head pore system; (c) Intromittent organ of male; (d) Lateral line.

across head, one through eye to cheek. Eight to ten cross-bars on body, spread-

ing to dorsal fin. Sometimes iridescent silvery patches along flanks.’ Two very

small freshly preserved juvenile specimens examined were yellowish red,

mottled and barred. Uniform yellow preserved in alcohol.

Location of type material. Department of Ichthyology, Rhodes University,

Grahamstown.

Material examined. 7 specimens, 26-38 mm in standard length. 2 from Port

Elizabeth, S.A.M. 24245; 3 from Cape Morgan, R.U.C.; 2 from Xora Mouth,

August 1947, R.U.C.

92 ANNALS OF THE SOUTH AFRICAN MUSEUM

Remarks. Blennioclinus stella is the smallest of the South African Clinidae. It is

very similar to Blennioclinus brachycephalus, but can be distinguished from it by

the lower number of dorsal soft rays.

Distribution (fig. 36). Port Elizabeth (33°58’S., 25°36’E) to Durban (29°53’S.,

31°oo’E); not common in southern part of range at least; usually in shallow

pools at the bottom of the shore, among sand and pebbles (Professor J. L. B.

Smith, personal communication).

Genus CLINOPoRUs Barnard, 1927

Clinoporus Barnard, 1927: 864 (type-species Clinus biporosus Gilchrist & Thompson, by monotypy).

Diagnosis: Body and head without scales, elongate, robust. No supraorbital

tentacle. Dorsal fin low, even, anterior spines not raised to form crest. Lateral

line of large pores opening above and below line throughout, very broad and

conspicuous. Pores on head large. Vomer toothed.

One species, Clinoporus biporosus (Gilchrist & Thompson).

Discussion. Clinoporus is a peculiar genus whose only known representative is

infra- rather than intertidal. The lateral line is very distinctive and unlike that

of any other clinid. The lack of scales is unique in the Clininae. C. L. Smith

(1957) investigated the possibility of the degree of scaling of clinids being related

to the amount of water movement in the environment of the adult, but conclu-

ded that the two factors are not correlated.

The inclusion of Clinoporus biporosus in the ‘Myxodinae’ by Smith (1945),

together with Pavoclinus, Blennioclinus and Gynutoclinus, solely on the basis of the

lack of a supraorbital tentacle is unjustifiable, since this species bears no resem-

blance to either the weed-dwellers of the genera Pavoclinus and Gynutoclinus or

to the species of Blennioclinus.

Clinoporus biporosus (Gilchrist & Thompson, 1908)

(Fig. 39)

Clinus biporosus Gilchrist & Thompson, 1908: 137.

Clinoporus biporosus: Barnard, 1927: 864. Smith, 1945: 545, 1949: 358, pl. 82 fig. 1003.

Description. D. XX XVIII-XLI 3; A. II 27-28; P. 12-13; V. 13; C. 13. Verte-

brae 17 + 32-35. Dorsal fin low, even, no notch in membrane between third

and fourth dorsal spines. No clusters of cirri at tips of dorsal spines. Pectoral fin

rounded. Inner pelvic ray minute. Caudal peduncle short, length 23-26% head

length, depth 28-5-31 °% head length. Caudal fin subtruncate.

Body robust, elongate, naked. Depth 6-25-6-85 in standard length. Head

4°5-6-25 in standard length, snout rounded. Sensory pores large and conspi-

cuous, mostly single (fig. 39(a)). Eye 2-75-4:25 in head. No supraorbital

tentacle. Cirrus on anterior nostril small, flaplike. Upper jaw 34-43% head

length. Lips moderately thick. Vomer toothed.

I , —, —_s

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 93

Lateral line throughout of large pores opening above and below line, close

together and numerous; line broad; ends in single pore on caudal peduncle

(fig. 39(a)). Intromittent organ of male with moderate basal portion and small

tip protruding between two rounded ventral lips (fig. 39(b)).

Colouring. No fresh specimens seen; uniform red, brown or orange (Smith,

1949). Uniform buff preserved in alcohol.

Location of type material. South African Museum, Cape Town.

Lectotype. Female, 117 mm, St. James, S.A.M. 25241.

Material examined. 7 specimens, 46-116 mm in standard length. 3 from Saldanha

Bay, S.A.M. 18463; 2 from St. James, False Bay, S.A.M. 25241 (lectotype),

(ii) lateral (iii) anterior

(i) ventral

(b)

Fic. 39. Clinoporus biporosus: (a) Lateral view, female, 117 mm, S.A.M. 25241, lectotype;

(b) Intromittent organ of male.

S.A.M. 10529 (paralectotype), 1 from False Bay, 14 fathoms, S.A.M. 10528

(paralectotype); 1 from False Bay, dredged, S.A.M. 23949.

Distribution. Saldanha Bay (33°00’S., 17°56’E) to False Bay (+34°10'S.,

18°50’E) is the known range. Rare, mainly infratidal.

Genus GYNUTOCLINUs Smith, 1945

Gynutoclinus Smith, 1945: 545 (type-species Clinus rotundifrons Barnard, by monotypy).

Diagnosis: A minute simple papilla over eye. Body highly compressed. Head

spherical, inflated, with pores opening on prominent papillae. Body covered

with minute, embedded, cycloid scales, which do not extend on to median fin

bases or head. Lateral line of single, more or less medially opening pores in

front to post-pectoral curve, then of short separate horizontal tubes with pore

at either end, becoming obscure caudally. Intromittent organ of male with long

basal portion; tip ensheathed by two pairs of lips. Vomer toothed.

A single, rare species, Gynutoclinus rotundifrons Barnard.

04. ANNALS OF THE SOUTH AFRICAN MUSEUM

Gynutoclinus rotundifrons (Barnard, 1937)

(Fig. 40)

Clinus rotundifrons Barnard, 1937: 63, fig. 4.

Gynutoclinus rotundifrons: Smith, 1945: 545, 1949: 358, fig. 1002. Penrith, 19654: 215, figs 3, 4.

Description. D. XXX-XXXII 8; A II 22-23; P. 12-14; V.13; C. 13. Vertebrae

16 + 28-29. First four dorsal spines slightly elevated to form low, rounded

crest, second and third spines highest. All spines evenly placed. No clusters of

cirri at tips of dorsal spines. No notches in membrane between any of the dorsal

spines. Profile of dorsal fin undulating. Pectoral fin rounded. Inner pelvic ray

reduced, not more than half other two rays. Caudal peduncle short, length

28-6-36:4% head length, depth 27-5-32°% head length. Caudal fin subtruncate.

Body highly compressed, deepening with age, covered with small scales

not extending on to bases of dorsal, caudal, or anal fins, or head. Depth 3°5—5-2

in standard length. Head 3:7—-4:15 in standard length, spherical, inflated,

broad, with pores opening on raised papillae. Eye 3:5-5 in head. Minute

simple papilla over eye. Anterior nostril tubular, cirrus large, flat, deeply

bilobed. Posterior nostril conspicuous, surrounded by short skinny lobes. Upper

jaw 42°5-54°5% of head length. Lips very thin. Vomer toothed. Sensory pores

of head mostly double (fig. 40(b)), opening on raised papillae.

Lateral line of single, more or less medially opening pores in front to post-

pectoral curve, then of short, separate horizontal tubes with pore at either end

(fig. 40(d)), obsolete on caudal peduncle on largest specimen, the holotype.

Intromittent organ of male with long basal portion and a clubshaped tip,

ensheathed by pair of thin, crescentic lateral lips and minute pair of more or

less confluent dorsal lips (fig. 40(c)).

Colouring. Male from Lambert’s Bay: ground colour pale brown with about

seven darker brown irregular cross-bars, edged with iridescent blue fine broken

lines. A very dark brown, narrow, vertical line behind and above the pectoral

axil, and another at end of caudal peduncle. Cross-bars continued on to dorsal

fin, with translucent patches between; small black dots on translucent areas.

Dorsal fin dark brown posteriorly, with single small translucent patch near

beginning of dorsal soft rays at base. Caudal fin translucent, with very faint

brown cross-bars, darkening at margins so that margins appear to be spotted.

Pectoral fin translucent with four very fine dark cross-bars, proximal one

curved. Anal fin mainly dark brown, with two translucent patches. Head mainly

light brown below. Dark stripe from below eye forwards in front of inflated

cheek to angle of jaw. Head above and opercular region deep pink. Snout pink

with darker pink bar between eyes and another above upper lip; fine darker

pink line down middle. Eye silvery with brown radii. Chin and lips light brown

mottled heavily with darker brown. Branchiostegal membranes and jugular

region silvery grey with fine black speckling. Pelvic fins light brown with dark

brown cross-bars. Belly silvery with golden brown sheen. Intromittent organ

greyish with fine black speckling. Juvenile from Lambert’s Bay yellow, under-

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 95

parts lighter. Dorsal fin with alternating pink and translucent patches. Anal

fin yellow. Caudal and pectoral fins translucent. Pelvic fins yellow with brown

cross-bars. Uniform pinkish to yellow preserved in alcohol.

Location of type material. South African Museum, Cape Town.

Material examined. 8 specimens, 26-95 mm in standard length. 1 from Liideritz-

bucht, S.A.M. 25091; 2 from Lambert’s Bay, $.A.M. 24007, S.A.M. 24082;

4 from Lambert’s Bay, R.U.C.; 1 from Oudekraal, Cape Peninsula, $.A.M.

18587 (holotype).

(iii) anterior

(i) ventral

(c)

Le a a arr ee

Sea ma —

(d) \

(b) Se

Fic. 40. Gynutoclinus rotundifrons: (a) Lateral view, female, 85 mm, S.A.M. 18587, holotype;

(b) Head pore system; (c) Intromittent organ of male; (d) Lateral line.

Remarks. This very rare species does not appear to be very close to any of the

other species, although it has certain modifications for a weed-dwelling habit

such as the strong compression of the body. It is known from eight specimens,

four in the South African Museum, and four in the Rhodes University Ichthy-

ology Department, the latter taken from the stomachs of sea-birds at Lambert’s

Bay and therefore in rather poor condition. The species was included by Smith

96 ANNALS OF THE SOUTH AFRICAN MUSEUM

(1945) in the ‘Myxodinae’, but in fact it has a small supraorbital tentacle,

although this is not very distinct in the long-preserved holotype. The tentacle

is quite prominent in fresh specimens. The intromittent organ of this species is

very similar to the type found in the genus Clinus.

Distribution (fig. 27). The known range is Liideritzbucht (26°38’S., 15°10’S) to

Cape Point (34°21'S., 18°29’E); in kelp at the bottom of the shore. Apparently

very rare.

Subtribe XENOPOCLINIDI Hubbs, 1952

Diagnosis: Clinini with hook on pectoral girdle sometimes reduced or absent.

Head depressed; eyes dorsal. Pelvic fins of three slender, equal rays, united

from base to tips by membrane. Dorsal fin low, even.

Discussion. Smith (19475) described a new genus and species of clinid, Xeno-

poclinus kocht, from a rock-pool at Lambert’s Bay on the south-western Cape

coast approximately 200 miles north of Cape Town. He placed this species in

the family Clinidae, but distinguished it by creating for it a separate ‘subfamily’

(equivalent to a subtribe in Hubbs’ 1952 classification). Subsequently another

genus and two more species, Xenopoclinus leprosus and Cancelloxus burrelli, obviously

closely related to Xenopoclinus kochi, were described from the Lambert’s Bay

area by Smith (1961). Smith then separated the three species from the Clinidae

and created for them a new family, Xenopoclinidae.

The three species are all adapted to a sand-burrowing existence. The body

is elongate, and the head is depressed, with the eyes in a dorsal, anterior posi-

tion. The pectoral fins are pointed, the middle rays being longer, and are used

in sweeping back the sand when burrowing. The jugular pelvic fins consist of

a minute spine and three rays of about equal length joined by a membrane that

extends well up between them, having the appearance of an amphibian webbed

foot. In Xenopoclinus kochi, but not in the other species, the opercular membranes

are expanded and overlap below. The opercular membranes of Cancelloxus

burrellt are expanded slightly above to form a fold over the pectoral axil.

In other respects they are much like the South African Clinidae. The

scales are minute, embedded, and cycloid. The jaw teeth are conical and fixed,

with usually a row of villiform teeth behind, at least at the jaw symphyses. The

two species of Xenopoclinus have teeth on the vomer, but in Cancelloxus the vomer

is edentate. There are no palatine teeth. The lateral line canals on the head are

covered; the lateral line itself is variable. In Xenopoclinus kochi the lateral line

follows a similar course to that of the Clinidae but is interrupted behind the

level of the pectoral fin, so that the anterior, dorsal portion is separate from the

posterior, mid-lateral portion. In Xenopoclinus leprosus the lateral line is similar

to that of the Clinidae, with no interruptions. In Cancelloxus burrelli the lateral

line is continuous, but curves down very much more gently than in the other

species to a mid-lateral position.

The dorsal fin is low and continuous, the anterior spines being the shortest,

and there are many more spinous than soft elements. There are two anal spines.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE Q7

All the soft fin rays are unbranched. The gill membranes are united, forming a

fold across the throat. There are six branchiostegal rays on either side. Xenopo-

clinus kochi has an upturned, hook-like process on the anterior border of the

pectoral girdle, but the other species show reduction or absence of this feature.

The young are born alive, and the male has a fleshy penis for the transmission

of sperm.

The question arises as to whether these three species, which have so much

in common with the South African Clinidae, should be placed in a separate

family.

Smith (1961) differentiated the Xenopoclinidae from the Clinidae by the

following characters: (i) the peculiar head, (11) the expanded opercular mem-

brane, (iii) the form of the pelvic fins, and (iv) the greater relative length of the

anal fin.

The form of the head and pelvic fins are associated with the sand-burrow-

ing habit of these fishes. The expanded opercular membranes occur only in one

of the three species, so cannot be considered as a strong differentiating character.

The species which shows this feature burrows in finer sand than the other two

species, and may have developed the expanded opercular membranes as a

result of this.

The meaning of the greater ‘relative’ length of the anal fin is not entirely

clear. The length of the anal fin base, about 60% standard length, is somewhat

greater in the Xenopoclinidi than in the South African Clinidi, about 40-50%.

This 10% difference seems hardly great enough to warrant separation at the

family level, in view of the considerable similarities between the two groups.

However, it is probable that by ‘relative’ length of the anal fin Smith (1961)

meant the number of anal elements relative to the degree of elongation of the

body as reflected by body depth in standard length. If this is the case, then the

anal fin is not relatively longer in the Xenopoclinidi than in the elongate species

of South African Clinidi. Xenopoctinus kochi and X. leprosus, with a body depth of

6-8, have 29-30 and 28-34 anal rays respectively; Clinus capensis, with a body

depth of 5-6, has 26-34 anal rays, C. dorsalis, body depth 5-5-7, has 25-31; and

C. anguillaris, the most elongate of the South African Clinidi, with a body depth

of 6-8, has 33-37 anal rays. Cancelloxus burrelli has a greater number of anal

rays (38-43) than any of the South African Clinidi, but the body depth is

8—11, greatly elongated in comparison with the other species. If only the number

of anal rays is considered, with no regard for the elongation of the body, then

Cancelloxus burrelli certainly has a longer anal fin than any of the South African

Clinidi, but the anal counts of Xenopoclinus kochi and X. leprosus fall well within

the ranges found among the South African Clinidi.

It seems, then, that the differences between the South African Clinidi

and the Xenopoclinidi are those concerned with the adaptation of the Xeno-

poclinidi to a burrowing existence; the depressed head with dorsally placed

eyes, the modified pelvic and pectoral fins, and the expansion of the opercular

membranes of Xenopoclinus kochi.

98 ANNALS OF THE SOUTH AFRICAN MUSEUM

Owing to the large number of features which the Xenopoclinidi share with

the Clinidi, particularly with the South African species, it is felt that separation

at the family level is unjustified, since it serves only to obscure the obvious

similarity of the two groups of species. That the Xenopoclinidi were derived

from the South African Clinidi, or at least from a common ancestor which had

already developed internal fertilization and a penis in the male, seems very

probable in view of the similarities of the two groups, particularly with regard

to the reproductive organs. Although the hook on the pectoral girdle, so charac-

teristic of the Clininae, is apparently in the process of being lost in the Xeno-

poclinidi, it is invariably present in Xenopoclinus kocht, and traces of it are present

in the other species. Sixteen specimens of Cancelloxus burrelli, including the

(b)

Fic. 41. Anterior border of cleithrum of: (a) Clinus superciliosus; (b) Xenopoclinus kochi (c) Xenopo-

clinus leprosus; (d) Cancelloxus burrellt.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 99

holotype, were examined, all of which had a projecting bony lamina on the

anterior border of the pectoral girdle. Of a sample of 89 specimens of Xeno-

poclinus leprosus examined, 32 had a small knob-like process on the anterior

border of the pectoral girdle. This feature is illustrated in figure 41.

Hubbs (1952) gave Smith’s (1949) subfamily Xenopoclininae the status

of a subtribe of the Clinini. He named the subtribe “Xenopoclinini’. It is pro-

posed here that this arrangement should be adopted, and that the three species

Aenopoclinus kochi, X. leprosus and Cancelloxus burrelli be placed in the subtribe

Xenopociinidi.

The resemblance the Xenopoclinidi bear to the American sand-burrowing

family known as the Dactyloscopidae is at first sight startling. The head, pec-

toral fins, and particularly the pelvic fins are strikingly similar in the two groups.

Smith (1961) noted this resemblance, but pointed out that they differ in that the

Xenopoclinidi have vomerine teeth (two of the three species), many more

dorsal spines than dorsal soft rays, small, embedded, non-imbricate scales,

united gill membranes forming a fold across the throat, and at least some

vestige of a hook on the pectoral girdle; on the other hand, they lack skinny

fringes on the opercle and lips, which are characteristic of the Dactyloscopidae.

The Xenopoclinidi differ further from the Dactyloscopidae in having a gene-

rally more elongate body, unbranched caudal rays, the uppermost pectoral

ray articulating with a pectoral radial (it articulates with the scapula in the

Dactyloscopidae, Dr. V. G. Springer, personal communication), the ecto-

pterygoid and mesopterygoid distinct (these two bones are fused in the Dacty-

loscopidae, Dr. V. G. Springer, personal communication), and a fleshy penis

in the male.

Key to the genera of the Xenopoclinidi —

1. WVomer toothed. Lateral line curves down sharply behind pectoral fin, may be discontinuous

over curve. A simple papilla or flap above the eye Ap Xenopoclinus

Vomer edentate. Lateral line curves down very gently from How to sasdhicen position. No

papilla or flap above the eye sm he aM er. ae ie .. Cancelloxus

Genus XENOPOCLINUS Smith, 1947

XAenopoclinus Smith, 194.7): 732 (type-species Xenopoclinus kochi Smith, by monotypy).

Xenopoclinops Smith, 1961: 354.

Diagnosis. A disc-like, fleshy papilla or simple flap over eye. Body elongate,

compressed, covered with minute embedded cycloid scales not extending on to

median fin bases or head. Vomer toothed. Lateral line curves down sharply to

medial position behind pectoral fin, discontinuous over curve.

Discussion. Smith (1961) divided this genus into two subgenera on the basis of

the presence or absence of a hook on the pectoral girdle, and whether or not

the opercular membranes were expanded and overlapped below. However, I

believe that, as suggested by Huxley (1940), the primary purpose of subgenera

is to indicate relationships between pairs or groups of species in genera, and

I0O ANNALS OF THE SOUTH AFRICAN MUSEUM

therefore that the use of the subgeneric rank in a genus containing only two

species is not warranted. The two species included in the genus Xenopoclinus

are obviously more closely related to one another than to the species for which

the other genus, Cancelloxus, was erected, and therefore should be included in

one genus.

Key to the species of Xenopoclinus

1. Hook on pectoral girdle well developed. Opercular membranes expanded, overlapping

below oF be Bis ale rt ae - x Xenopoclinus kochi

Hook on pectoral girdle reduced to a small knob, or absent. Opercular membranes not

expanded and overlapping below a: Ae 2 3 Xenopoclinus leprosus

Aenopoclinus kochi Smith, 1947

(Fig. 42)

Xenopoclinus kochi Smith, 19475: 732, fig. 1, 1949: 358, fig. 1004.

Xenopoclinus (Xenopoclinus) kochi: Smith, 1961: 352, fig. 2.

Description. D. XXXITI-XXXVII (XXXIV-XXXV) 8-11 (9); A. IT 29-30;

P. 12-13 (12); V. 1 3; C. 13. Gill-rakers in outer series on first arch 2 + 0-2.

Vertebrae 15 + 34-35. Dorsal fin low, even, anterior spines shortest. Third

spine a little shorter than first or fourth. No notch in membrane between third

and fourth dorsal spines; spines evenly placed. Soft dorsal rays a little longer

than spines. Spines fairly soft. No clusters of cirri at tips of dorsal spines. Pec-

toral fin with central rays elongate, upswept, fin pointed; lower rays short,

curling forward. Pelvic fin of three rays, equal in length, united by membrane

from base to tips, resembling a frog’s foot. Caudal peduncle short, length

30-40% head length, depth 25-35% head length. Caudal fin subtruncate.

Body elongate, compressed, covered with minute embedded scales not

extending on to bases of dorsal, caudal, or anal fins, or head. Depth 6-7-5

in standard length. Head depressed, 3°75—5 in standard length. Snout subacute.

Lower jaw projects. Eyes dorsally situated, almost adjacent, 4-6-5 in head.

Flat, fleshy papilla over eye. Anterior nostril tubular, cirrus short simple flap.

Upper jaw 30-40% head length. Lips moderate. Vomer toothed. Sensory pores

of head mostly double, multiple pores in postorbital region (fig. 42(b)). Oper-

cular membranes expanded, overlapping below.

Lateral line of mostly vertical pairs of pores in front to behind pectoral fin;

ends abruptly there and is continued below in the mid-lateral line, consisting

of short separate horizontal tubes with pore at either end (fig. 42(d)). Intro-

mittent organ of male with long basal portion and rounded tip ensheathed

below by two rounded ventro-lateral lips and above by hood-like fold (fig. 42(c))

Hook on anterior edge of pectoral girdle well developed.

Colouring. Ground colour stony pinkish mauve with about six irregular olive

green cross-bars, running into vague olive mottling posteriorly. Cross-bars

edged with darker olive. On head behind eyes are two adjacent olive rings.

Fins hyaline except dorsal, where olive cross-bars may continue on to fin.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE IOI

Pectoral base olive. Belly whitish. Dark olive bar at caudal base. Two dark

olive blotches on preopercle, sometimes lighter one between; olive stripe from

front of eye to upper lip on either side. Speckled and mottled all over body

and head with silvery white. Cross-bars more distinct in juveniles. Pupil black,

iris gold; papilla over eye white. Pattern retained for some time after preserva-

tion.

Location of type material. Department of Ichthyology, Rhodes University,

Grahamstown.

Material examined. 13 specimens, 26-73 mm in standard length. 8 from Lam-

bert’s Bay, S.A.M. 24667; 5 from Lambert’s Bay and Doring Bay, R.U.C.

(including holotype).

Distribution (fig. 36). Known only from Lambert’s Bay and Doring Bay in

southern Namaqualand (-(32°00’S., 18°20’E). Not common, in pools and

gullies at the bottom of the intertidal zone with much kelp. Burrows in fairly

coarse shingly sand.

eS)

i) ne Gi) [ateral Gii) anterior

(c)

(d) au

Fic. 42. Xenopoclinus kochi: (a) Lateral view, female, 72 mm, S.A.M. 24667; (b) Head pore

_ system; (c) Intromittent organ of male; (d) Lateral line.

Xenopoclinus leprosus Smith, 1961

(Fig. 43)

Xenopoclinus (Xenopoclinops) leprosus Smith, 1961: 354, fig. 3.

Description. D. XXXITI-XXXVI (XXXIV-XXXV) 7-12 (9-11); A. II

28-34 (31-32); P. 11-13 (11-12); V. I 3; C. 13. Gill-rakers in outer series on

first arch 1-2 + 2-5. Vertebrae 15 + 35-36. Dorsal fin low, even, spines

102 ANNALS OF THE SOUTH AFRICAN MUSEUM

increasing in height posteriorly. Third spine a little lower than fourth or second

first may be slightly higher, especially in juveniles. Soft dorsal rays a little

longer than spines. Spines soft. Pectoral fin not elongate, slightly upswept,

lower rays thickened. Pelvic fin of three equal rays joined throughout length

by membrane, resembling a webbed foot. Caudal peduncle short, length

20-33% head length, depth 20-33% head length. Caudal fin subtruncate.

Body elongate, compressed, covered with minute embedded scales not

extending on to dorsal, caudal, or anal fin bases or head. Depth 6-25-8-75 in

standard length. Head depressed, 4:75-5 in standard length in adults, 4-4°5

in standard length in juveniles. Snout subacute. Eyes dorsal, adjacent, 4-6°5

in head. Flattened, flaplike tentacle over eye, edge irregularly and shallowly

notched. Anterior nostril tubular, cirrus flaplike, sometimes terminally notched.

Upper jaw 33-43°5 %head length. Lips moderate. Vomer toothed. Opercular

membranes normal, not expanded or overlapping below. Sensory pores of head

mostly double (fig. 43(b)).

Lateral line of vertical pairs of pores and single pores in front to post-

pectoral curve, where it ends abruptly; continued in mid-line as short separate

horizontal tubes with pore at either end, very obscure in most specimens (fig.

43(d)). Intromittent organ of male with fairly long basal portion and short tip

ensheathed by pair of united ventral lips (fig. 43(c)).

‘ RS

SARK AG

QO A

WS,

4 SS. SSN

CHK SD

(iii) anterior

(ji) lateral

(c)

(i) ventral

ee fae Nee pete ee ee i

A LT WEES TES reseed SP

(b) (d)

Fic, 43. Xenopoclinus leprosus: (a) Lateral view, male, 62 mm, S.A.M. 24668; (b) Head pore

system; (c) Intromittent organ of male; (d) Lateral line.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 103

Hook on anterior border of pectoral girdle usually absent; if present,

reduced to small knob.

Colouring. Ground colour light brown below grading to chocolate above. Three

large white saddles in upper half of body. Head white, except for an area,

enclosed by a line on each side from dorsal origin to eye and from eye vertically

down cheek, which is sand-coloured in life, becoming dark brown or jet black on

preservation in spirit. Pectoral base yellow. Dorsal fin brown above dark areas

on body, hyaline above white saddles. Other fins hyaline. Iris golden, pupil jet

black. Belly light brown to cream. Pattern retained on preservation although

colours fade.

Location of type material. Department of Ichthyology, Rhodes University, Grahams-

town.

Material examined. 95 specimens, 22—68-5 mm in standard length. 1 off Orange

River Mouth, 10 fathoms, U.C.T.; 89 from Lambert’s Bay, S.A.M. 24668;

5 from Lambert’s Bay, R.U.C. (holotype and paratypes).

Distribution (fig. 36). The known range is Orange River Mouth (28°38’S.,

16°24’E) (infratidal, dredged) to Lambert’s Bay (32°04’S., 18°20’E) (in

intertidal pools and gullies at lowest levels of shore). Burrows in very coarse

shingle and broken mussel shells. Fairly common but patchily distributed.

Cancelloxus burrelli Smith, 1961

(Fig. 44)

Cancelloxus burrelli Smith, 1961: 355, figs 1, 4.

Description. D. XXXIV-XXXIX (XXXVI-XXXVII) 10-14 (10-13);

A. II 38-43 (41); P. 13-15 (13-14); V. 13; C. 13. Gill-rakers in outer series on

first arch 0-2 + 5-8. Vertebrae: females 20-21 + 44-45, males 19 + 43-46.

Dorsal fin low, even, anterior spines lowest, soft rays a little higher than spines.

Spines soft. Pectoral fin with middle rays elongate, pointed and upswept, lower

rays thickened. Pelvic fin of three equal rays, joined from bases to tips by

membrane, resembling a webbed foot. Dorsal fin originates well behind head.

Caudal peduncle short, length 23:5-31% head length, depth 20-28-5°% head

length. Caudal fin subtruncate.

Body elongate, compressed, covered sparsely with minute scales not

extending on to dorsal, caudal, or anal fin bases, or head. Depth 8-11 in standard

length. Head depressed, 4:25—5°5 in standard length, snout pointed. Eyes oval,

adjacent, 3°75-4°75 in head. No tentacle or papilla over eye. Anterior nostril

tubular, cirrus small simple flap. Upper jaw 26-29% head length. Lips fairly

thick. Lower lip produced at symphysis into acute process. Vomer edentate.

Few or no teeth at sides of jaw, patch at each symphysis. Opercular mem-

branes normal, not expanded or overlapping below. Sensory pores of head

mostly single, opening flush with skin surface (fig. 44(b)).

Lateral line of single pores opening more or less medially in front, then of

short separate horizontal tubes with pore at either end, sloping gradually to

104 ANNALS OF THE SOUTH AFRICAN MUSEUM

mid-line behind pectoral fin (fig. 44(d)). Intromittent organ of male extremely

large. Basal portion long and stout, terminating anteriorly in pair of horn-like

lobes between which tip protrudes. Tip conical, curving up towards belly.

Flat fleshy lobe dorsally at base of tip. Belly of male deeply grooved. Intro-

mittent organ illustrated in fig. 44(c). Vent of female large and crescent-

shaped, surrounded by fleshy fold produced posteriorly into two lobes, one on

either side of first anal spine. Each lobe with thick fleshy dorsal portion over-

laid by thin skinny lobe ventrally.

Colouring. Ground colour cream, semi-translucent. Bright silver broken stripe

mid-laterally. Irregular olive patches above. Eleven olive saddles above, edged

darker. Olive stripe from eye to posterior edge of opercle; below this stripe a

broad silver stripe. Lips and underparts cream. Pectoral base silver. Fins

hyaline. Pupils jet black, iris narrow and golden. In alcohol the ground colour

becomes opaque and the olive turns golden brown, later fades completely.

ta x, SN i

KN - ee a © S BRE RY, Re PTR

AERA TA TAI otal ate MCU LLCS Na he ANC RG RSS ERS ECC INC

( ) ARR, OX CAE SRS

Gi) lateral

(I) ventral (c)

(d)

Fic. 44. Cancelloxus burrelli: (a) Lateral view, female, 105 mm, S.A.M. 24669; (b) Head pore

system; (c) Intromittent organ of male; (d) Lateral line.

Location of type material. Department of Ichthyology, Rhodes University,

Grahamstown.

Material examined. 19 specimens, 75:5-107 mm in standard length. 1 from off

Orange River Mouth, 5 fathoms, $.A.M. 23272; 1 from Lambert’s Bay, R.U.C.

(holotype); 3 from Lambert’s Bay, R.U.C. (paratypes); 14 from Lambert’s

Bay, S.A.M. 246609.

Distribution (fig. 36). The known range is Orange River Mouth (28°38’S.,

16°24’E) (infratidal, 5 fathoms) to Lambert’s Bay (32°04’S., 18°20’E),

mainly in gullies and inlets open to the sea at the base of the intertidal zone, in

coarse shingle. Not common.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 105

ORIGIN OF THE SOUTH AFRICAN CLINIDAE

The presence of a hooklike projection on the anterior border of the clei-

thrum, small scales with radii on all margins, and the very similar general

appearance of the less specialized members of both groups, point to a common

origin of the Myxodini and the Clinini. Hubbs (1952) considered the Myxodini

to be ancestral to the Clinini because the Myxodini are oviparous, fertilization

is external, and there is no intromittent organ in the male.

The derivation of the South African species from the most generalized of

the American myxodids, the species of the genus Gibbonsia, would not be

difficult. If one of the American species such as Gibbonsia metzi C. L. Hubbs

were to develop viviparity, the result would be a clinid very similar to some of

the more generalized South African clinids such as Clinus robustus. I examined

specimens of Gzbbonsia, Myxodes, and Heterostichus, and, apart from the lack of an

intromittent organ in the male, they are strikingly similar to the South African

clinids.

Clinoporus

Blennioclinus

Pavoclinus

Clinus

Gynutoclinus

Fic. 45. Hypothetical intergeneric relationships of South African Clinidi.

106 ANNALS OF THE SOUTH AFRICAN MUSEUM

In a consideration of the origins of a group it is often difficult to decide

which characters are primitive and which are advanced. Table 8 gives what I

consider to be the primitive condition for twenty characters which have been

used in my study. In some cases the definition of the primitive condition is

based on the statements of Hubbs (19530) regarding typically clinid characters,

and in others on the conditions prevailing in the genus Gibbonsia.

Of the two subtribes of Clinini represented in South Africa, the Clinidi

have by far the greatest representation, with five genera and 30 species in

South African waters. The genera apparently represent two major lines of

evolution and three minor ones. The hypothetical intergeneric relationships are

shown in figure 45. The method used for devising the figure is similar to that

used by Stephens (1963). The lateral position is based purely on morphological

resemblance and the horizontal distances are arbitrary, but the vertical position

indicates the ‘primitiveness’ of the genera. The “primitiveness’ is determined by

the number of points awarded for each of the twenty characters listed in table 8,

from no points for the ‘primitive’ condition of a character up to ten points for an

‘advanced’ condition. The thickness of the lines indicates the relative number

of species.

100

dorsalis

capensis

venustris

striatus

brevicristatus

acuminatus

cottoides

berrisfordi

heterodon

latipennis

superciliosus

Fic. 46. Hypothetical interspecific relationships of the South African genus Clinus.

The two major lines, leading to the genera Clinus and Pavoclinus, retain a

fairly high number of ‘primitive’ characters, but these are not the same in the

two groups; for instance, the supraorbital tentacle is retained in Clinus but lost

in Pavoclinus, whereas the ‘primitive’ arrangement of the anterior dorsal spines

for this group is retained in several of the species of Pavoclinus, but in none of the

species of Clinus. The species of Pavoclinus are more specialized for a particular

habitat than are the species of Clinus.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 107

In the genus Clinus there appears to have been a tendency to reduce the

height of the anterior dorsal spines. (This tendency is apparently paralleled in

the Labrisomini (Hubbs, 1953)).) In those species of Clinus in which the

anterior dorsal spines are high, the membrane, especially that between the

third and fourth dorsal spines, is incised to a greater or lesser degree; this is also

Vv)

110 E

fucorum

graminis

laurentii

‘Davo

profundus

~“mentalis

41]

[as

Fic. 47. Hypothetical interspecific relationships of the South African genus Pavoclinus.

108 ANNALS OF THE SOUTH AFRICAN MUSEUM

the case in two species in which the anterior spines are low. In three of the four

subgenera (comprising, however, less than one third of the species in the genus)

there is a tendency towards elongation of the body and an eel-like habit. These

forms are thought to have diverged early from the main branch (see fig. 46,

Clinus anguillaris, C’. striatus, C. capensis, C. dorsalis and C. navalis). All the species

involved show considerable specialization, which would have required a rela-

tively long time, yet retain certain primitive characters such as the facial and

supraopercular scales in the subgenus Blennophis and the facial scales in the

subgenus Cirrhibarbis. All the less specialized species of Clinus have the head

region completely naked, except for C. woodi, which retains three supraopercular

scales on either side. ‘The tendency to reduction of scaling in the head region

has occurred in all the genera of South African Clinini, most of which lack

scales on the head, and the scaling of the fin bases has become reduced as well.

Figure 46 shows the hypothetical inter-relationships of the species of Clinus.

The figure has been drawn up in the same way as figure 45, using the characters

listed in table 8.

All the species of the genus Pavoclinus are modified for a weed-dwelling

habit. The supraorbital tentacle has been lost, and in this feature they are more

‘advanced’ than the species of Clinus, but the arrangement of the dorsal fin in

Pavoclinus graminis, P. litorafontis and P. pavo is very similar to that of the species

of Gibbonsia, and is therefore considered to be primitive. As in the genus Clinus

there has been a reduction in the extent of the scaling, and in all the species of

Pavoclinus the head region is entirely naked. The hypothetical relationships of

the species are shown in figure 47, which has been drawn up in the same way

as figures 45 and 46.

The subgenus Labroclinus, containing one species (Pavoclinus mentalis), may

have diverged fairly early. It retains several ‘primitive’ features; the fin counts

are higher than those in the other subgenera of Pavoclinus, being similar to the

intermediate fin counts of the Clinus (Clinus) group of species; the dorsal and

caudal fin bases are scaled; the vomer is toothed, and the caudal peduncle is

only moderately elongate. On the other hand, it has apparently specialized

features such as the complete separation of the dorsal crest and the skinny flap

on the lower jaw symphysis. It presumably arose by specialization from a form

such as Pavoclinus laurent, which has a separate dorsal crest.

The other two monospecific subgenera, Fucomimus and Smithicthys, are

highly specialized weed-dwellers and retain very few ‘primitive’ characters.

The fin counts are reduced, the scales are minute and do not extend to the

median fin bases, the vomer is edentate, and the caudal peduncle is very long.

Pavoclinus (Fucomimus) mus has a separate dorsal crest, and a low number of

dorsal soft rays; the number of caudal rays is reduced to eleven. Pavoclinus

(Smithichthys) fucorum has a peculiar form of crest with at least the first four

dorsal spines involved, an upturned snout, and the teeth reduced to one row in

each jaw. The lack of vomerine teeth in the subgenera Fucomimus and Smith-

ichthys is paralleled in the South American genus Myxodes, to which the species

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 109

Pavoclinus (Smithichthys) fucorum was referred by Smith (1945).

The remaining species constituting the subgenus Pavoclinus form a rela-

tively compact group; in one of the five species the fin membrane is incised

between the third and fourth dorsal spines, and in one of the species, Pavo-

clinus profundus, the anterior dorsal spines are low, and the scales are minute.

The three remaining genera are minor groups and have undergone little

or no speciation. |

Blennioclinus is apparently not close to any of the other genera, but in the

character of the anterior part of the lateral line and in the body form and habit

resembles Clinus rather than Pavoclinus. The notch in the profile of the dorsal fin

before the dorsal soft rays in this genus is unique in the Clininae, although it

commonly occurs in the labrisominid genera Labrisomus and Malacoctenus

(Springer, 1958, pls. I-VII). However, as Blennioclinus clearly belongs to the

Clininae, it is difficult to see how this character may have developed. In this

connexion it is interesting to note that there are various features which appear

scattered throughout the clinids in apparently not very closely related forms.

Examples are the flap of the skin on the lower jaw symphysis developed in some

of the American tribe Paraclinidi, e.g. Paraclinus barbatus Springer (Springer,

1954, fig. 1), in Pavoclinus mentalis (fig. 29), in Cancelloxus burrelli (fig. 44) and

in members of the related family Chaenopsidae, e.g. Chaenopsis alepidota alepidota

(Gilbert) (Stephens, 1963 pl. 12); the reduction of the number of dorsal soft

rays to one in the Paraclinidi (Hubbs, 1952) and the subgenera Climacoporus

and Muraenoclinus of the genus Clinus; and the development of the hook on the

pectoral girdle, so characteristic of the Clininae, in the adults of two Labri-

sominae, Malacoctenus erdmani Smith and Malacoctenus aurolineatus Smith (Sprin-

ger, 1958). I believe that these shared characters are strong evidence for a

monophyletic origin of the Clinidae as a whole.

Clinoporus, the most specialized of the five South African genera of Clinini,

is apparently most closely related to the genus Clinus and may have been derived

from the line leading to the eel-like species of Clinus by loss of the scales and

mod.fications of the lateral line. Since the single species, Clinoporus biporosus,

is the only species of the South African Clinini (other than the little-known

Pavoclinus profundus) that is normally taken in relatively deep water beyond the

limits of the infratidal fringe, it might be expected to differ markedly from the

predominantly intertidal Clinidae.

Gynutoclinus retains the greatest number of ‘primitive’ characters, although

the peculiar form of the head excludes it from the direct line of descent of any

of the other genera. The rarity of its single species suggests that it has not

proved as successful in those habitats where it has been taken as at least three

of the other four genera; Clinoporus is also represented by a single, apparently

rare species, but the inaccessibility of the usual habitat of Clinoporus may con-

tribute considerably to the impression of rarity.

Gynutoclinus has features in common with both Pavoclinus and Clinus, but

it appears to have followed an independent pattern of evolution. The intro-

I1O ANNALS OF THE SOUTH AFRICAN MUSEUM

mittent organ of the male is similar to the type found in the genus Clinus, and

there is a vestigial supraorbital tentacle; the characters shared with Pavoclinus

are chiefly those associated with adaptation to a weed-dwelling existence.

The other subtribe, the Xenopoclinidi, apparently have a common deriva-

tion with the Clinidi from a myxodidlike ancestor, and the high degree of

Table 8. Characters used in the consideration of specialization in South

African Clinini.

Characters

Presence or absence of scales on body

Degree of scaling on head

Degree of scaling on median fin

bases

Arrangement of jaw teeth

Presence or absence of teeth on

vomer

Position of dorsal fin origin

Height of anterior dorsal spines

Presence or absence of a notch in

membrane between third and

fourth dorsal spines

Number of dorsal spines

Number of dorsal rays

Number of anal rays

Development of third pelvic ray

Arrangement of posterior part of

lateral line

Arrangement of anterior part of

lateral line

Presence or absence of supraorbital

tentacle

Shape of head

Snout shape

Presence or absence of facial barbels

Condition of jaw symphysis

Length of caudal peduncle

‘Primitive’ condition

Scales present

Scale patches on cheek and above

opercle.

Scales on dorsal, caudal, and anal

fin bases

A row of small teeth behind front

row of jaw teeth

Vomer toothed

Over hind margin of preopercle

First three dorsal spines higher than

succeeding spines

Notch in membrane absent

30-40

5-10

20-30

Stout, equal to other two

Short separate horizontal tubes with

a pore at either end

Single pores opening more or less

medially

Supraorbital tentacle present

Compressed, not inflated

Sloping to mouth, not upturned

Facial barbels absent

Smooth, without projecting skinny

flap

Short to moderate, 20-40% of head

length

ANNALS OF THE SOUTH AFRICAN MUSEUM III

specialization of this group suggests that the divergence probably occurred

early. It is possible to derive the Xenopoclinidi from a form tending towards

elongation of the body such as might have given rise to Clinoporus and the eel-

like species of Clinus.

Note on the geographical origin of the Clininae

As can be seen from figure 1, the Clinidae are a widespread family, occurring

in the tropics and the temperate regions of the southern hemisphere. Because

the greatest occurrence of the family as a whole is in the tropics of Central

America, this region has been considered to be their centre of origin (Hubbs,

1952; Springer, 1958). The South African and Indo-Australian forms are all

viviparous and have been considered to have spread to those regions from

America some time after the origin of the group; closely related oviparous forms

occur in South America. Most of the clinids of the Pacific coast of America and

all the clinids except two species of the Atlantic coast of America belong to the

apparently more primitive subfamily the Labrisominae. The subtribe Calli-

clinidi, which Hubbs (1952) considered to be the most primitive of the Labri-

sominae, is apparently confined to the Pacific coast of South America. The

tribe Neoclinini, also primitive but of rather uncertain systematic position, is

also confined to the eastern Pacific, with one, or at most a few, representatives

in Japan, the Ryukyus and Korea.

It has generally been assumed that the Clininae and the Labrisominae have

had a common origin, and this view is supported by the shared features listed

in the previous section. However, it is as well to remember that insufficient

comparison -has been made between the Labrisominae and the Clininae to

determine whether either could actually be ancestral to the other; on the whole

a common origin from a more primitive group is most probable.

- Briggs (1955), following Matthew (1915), considered’ the centre of origin

or ‘centre of dispersal’ of a group to be that area in which the most progressive

species of the group are found, rather than the area where the greatest number

of species, regardless of their state of advancement, occur. The most primitive

species are then found to have a peripheral distribution about this area.

If the Labrisominae and the Clininae are considered separately, the follow-

ing emerges: the most progessive Labrisominae undoubtedly occur in the

tropical waters of Central America. The Calliclinidi occur on the periphery of

this region, in temperate Pacific South American waters. But the most progres-

sive Clininae do not occur in the American region; they occur in the temperate

waters of South Africa and Australia. The more primitive Clininae, the Myxo-

dini, the Japanese flammeus with its small intromittent organ, and probably the

Mediterranean clinid, occur mainly on the periphery of this region. This

suggests, then, that the Labrisominae had an Atlanto-East-Pacific origin,

whereas the Clininae had an Indo-West-Pacific origin.

Briggs (1955) pointed out that it has been proved for several plant and

land animal groups, where there is an extensive fossil record, that their centre

M4 THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE

of origin lies far from the area in which their greatest modern occurrence lies.

Unfortunately the inadequacy of the Tertiary fish fossil record has made it

necessary to rely to a large extent on modern distribution alone. However, there

are some indications that the Labrisominae may have originated and undergone

at least some radiation in the Tethys Sea before conditions there became

unsuitable for a tropical group, and that they then became established and

underwent extensive radiation in the Atlanto-East-Pacific while the Pacific and

Atlantic coasts of Central America were still in free communication. Labrisomus

nuchipinnis (Quoy & Gaimard) occurs on the tropical Atlantic coast of America,

in the West Indies, the Madeira Islands, the Canary Islands, the Cape Verde

Islands and the coast of tropical West Africa as far south as Mocamedes in

Angola. Malacoctenus africanus (Cadenat) has been recorded from Dakar,

Senegal, in West Africa. Arambourg (1921, 1927, 1943), described a fossil

species which he called Labrisomus pronuchipinnis from the Miocene ‘Tethys

deposits of Algeria. He referred it to the relatively small category of Mediter-

ranean fossil species having their main affinities with the fauna of the Atlantic

(Arambourg, 1943). Steinitz (1950, q. Hubbs, 19535) believed the occurrence

of Labrisomus nuchipinnis in West Africa to date from the time of the Tethys Sea.

Hubbs (19536) and Springer (1958) disagreed with this view, because they

considered that a species could not remain undifferentiated on both sides of

the Atlantic for so long a time, and Hubbs (19530) suggested that the fact that

Labrisomus nuchipinnis apparently has a rather longer pelagic larval stage than is

usual for clinids might account for its transport across the Atlantic. However,

the species of Malacoctenus apparently have a particularly short larval stage

(Springer, 1958), so that this would not account for the occurrence of a species

of Malacoctenus in West Africa, if it is assumed to be derived from the American

species of Malacoctenus. To explain the occurrence of Labrisomus pronuchipinnis

in Mediterranean deposits and the occurrence of Labrisomus nuchipinnis and

Malacoctenus africanus in West Africa, we must, if the group is considered to be

American in origin, postulate a Tethys crossing for Labrisomus pronuchipinnis

and two recent crossings for the two modern species. The two recent crossings,

at least, are not favoured by the prevailing current systems, although this

appears often to be the case in shore fish distribution (Briggs, 1967). Perhaps as

far as the Labrisominae are concerned it might be preferable to agree with

Steimitz (1950) and to suggest that, although on the whole one would expect

genetically isolated populations such as the American and West African

Labrisomus nuchipinnis to become specifically distinct, Labrisomus nuchipinnis is

clearly a robust and successful species where it occurs, and it may have had no

tendency to undergo any change. This latter is not surprising, since Ekman

(1953) pointed out that the West African region is in fact a faunally poor area

that few groups have found suitable for extensive radiation. Labrisomus nuch-

pinnis may be the exception that proves the rule regarding isolated populations;

whether we regard it as exceptional as regards speciation or exceptional as

regards crossing the Atlantic from America to Africa, it remains an exception;

pre

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE

‘ssorr) oder JO YRAOU SopIUI O1S Sor] 1 INq UMOYs jOU st ‘pa}d9][OD seM PTUT[D B YITYM 7B Jsvood ULITIFY 1S9AA YING IY}

uo yurod jsourussyj}s0U 94} “UIOg AxDOY 3X9} 9Y} UT O} PIIIOJII SOTII[VIOT OY) SUTMOYS VoLIPY UJOYINOSs Jo dep ‘gv ‘O17

Aeg

asie4 quiog 2ded

Sa]! OOL adoH poos jo adeD

ee — ee

P2aasV y4oY,

WIOd Ysi4 yea? auosped aded

uopuo7 4se3

YINOW 42aAIy 1ay

WY e40j0D

YINOW 42AIY 2ayseg ae ees

Aeg sjdeaquie7

5,uYyo,

Yor IS Wo ‘YY sqUEsO \Aeg Buls0og

P4EMPA 340d JUIOd J4ND

BONIAONd Adv

puejenbewen

Keg dij}2puoH

NO|YPLuy)

Aeg s,pueusiy as UJOIION WOd

Asenjs3 e19n7 4s

tats,

aN punWwafueso

? ny off

puejninz , ten

1zeBueduin

2ngZzp4epn7

sonbsuew O5U247O-}6 ; etedlsbe tes euanbag e4buy

Keg eobej2q

SSeS Sewen eens ence cnsnas sease (anseease

auequieyu

VolusyY 1SSM HLNOS

Keg sIAIeM

punwidoxemMs

SANOIGaWVZOW QZ 2U0jSaIIN

ssouD aded

Mees 99 ees ae es =

92 ; ve

I1i4 ANNALS OF THE SOUTH AFRICAN MUSEUM

at least by adopting the former view Labrisomus pronuchipinnis and Malacoctenus

africanus do not have to be regarded as similarly exceptional.

It seems possible that the common stock which gave rise to these two sub-

families was a tropical Tethys Sea group which, in common with most of the

early Tertiary Thethys fauna, died out when climatic conditions changed.

Ekman (1953) noted that both the Indo-West-Pacific and the western Atlantic

(Atlantic-East-Pacific region) received forms from the Tethys Sea which are

now either no longer predominant or are no longer found at all in the Mediter-

ranean or the eastern Atlantic. The Atlanto-East-Pacific would then have

received the Labrisominae, already to some extent differentiated, from the

Tethys Sea, and these forms would have found this area highly suitable for

adaptive radiation. The Clinidi, on the other hand would have become estab-

lished in the temperate Indo-West Pacific and there undergone adaptive

radiation and developed viviparity. The Myxodini may either have reached

South America by crossing the eastern Pacific from the Indo-West-Pacific, or

may have arisen separately from the viviparous Indo-West-Pacific Clinini. The

latter seems more likely; the hook on the pectoral girdle may not indicate as

close a relationship as it has been thought to do, since it does appear in two

Labrisominae. In any case, if the forerunners of the Clininae were present in

the Tethys Sea, it is possible that some migrated to the Indo-West-Pacific and

gave rise to the viviparous Clinini, while others, very similar, migrated with

the Labrisominae to the Atlanto-East-Pacific and there gave rise to the Myxo-

dini. However, such considerations of the zoogeographical origin of a group are.

always of a highly speculatory character.

| The mechanism of such dispersal of virtually sedentary forms, whether

. now or in Tertiary times, is of course unknown. Ekman (1953) stated that the |

. only conceivable method of dispersal over long distances for forms with a short ©‘

- or no pelagic larval period is by attachment to or entanglement in sea-weed.

He cited two cases of wide distribution of viviparous echinoderms which he

considered could have occurred in no way other than by transport in sea-weed.

One is the brittle-star Amphipholis squamata (Della Chiaje), which occurs from

the Red Sea to South Africa, Hawaii, the Society Islands and the west coast of

America; it also occurs at the southern tip of Patagonia, indicating that it was

transported round Cape Horn. It is obviously far more eurythermic than any

of the Clinidae. The other is the starfish Asterina exigua (Lamarck), endemic to

South Africa, which also occurs on the island of St. Helena, where it was

apparently carried in drifting material from South Africa by a branch of the

Benguela current.

The transport in sea-weed of the ancestor of, for instance, the Australian

Clinini from South Africa is quite likely. It would be quite possible for a weed-

dwelling clinid to live for a fairly long period of time in a floating mass of

weed, provided that the food supply lasted until land was reached, and to

become established in the new area if conditions were suitable. Viviparity

would favour this type of dispersal, since only one gravid female would be

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE 115

needed to carry quite a large number of fishes to the new area.

Finally it may be said that there is a modern record of a clinid surviving a

long journey when being transported by a ship, although the specimen met

with a fate on arrival that precluded its becoming established in the new area.

Clinus navalis (Barnard) .was first described from a specimen recovered

from amongst barnacles on a ship’s bottom at Simonstown in False Bay, Cape

Peninsula, some two hundred miles west of the western limit of the range of the

species, and some five hundred miles west of the last port where it could have

been picked up. It was apparently found in good health.

I am not suggesting that this example can be used to explain the distribu-

tion of, for instance, the West African clinids, since their existence in Africa

obviously antedates regular ship crossings; the example is merely given to show

that clinids can survive passive transport at least over a short period, by ships,

and the underside of ships, discharging oil, etc., might be considered to be a

less hospitable environment than a free-floating raft of weed.

SUMMARY

This work consists of a systematic revision of the South African Clinidae.

The relationships of the South African Clinidae with the Clinidae of other parts

of the world are briefly considered. Classification of the South African Clinidae

above the generic level is revised, and the several South African clinid species

are regrouped in seven genera, which arrangement seems to reflect the natural

relationships within the group better than the. previous assemblage of small

genera. The family Xenopoclinidae Smith, .1961 is included in the family

Glinidae. Some comment js':made on the evolution and zoogeography of the

Clinidae. lg en oe : iA

— ~

ACKNOWLEDGEMENTS

I am grateful to the council for Scientific and Industrial Research for a

grant which supported me for part of the time during which this research was

undertaken, and to the Council for Scientific and Industrial Research and the

South African National Committee for Oceanographic Research for grants

towards collecting expeditions.

I am indebted to Dr. N. A. H. Millard of the Zoology Department,

University of Cape Town, for helpful criticism and guidance throughout this

work; to Dr. V. G. Springer of the Division of Fishes, Smithsonian Institution,

United States National Museum, for many helpful suggestions and valuable

criticisms of the manuscript, as well as much useful comparative material; to

Dr. F. H. Talbot of the Australian Museum, Sydney, and formerly of the South

African Museum, for much help in the initial stages of the project; to the late

Professor J. L. B. Smith and to Mrs. M. M. Smith of the Ichthyology Depart-

ment, Rhodes University, and Professor J. H. Day of the Zoology Department,

University of Cape Town, for helpful discussions.

116 ANNALS OF THE SOUTH AFRICAN MUSEUM

I am grateful to the South West African Administration and in particular

Mr. C. G. Coetzee, Director of the State Museum, Windhoek, for making it

possible for me to collect on the South West African coast between Rocky Point

and the Kunene River Mouth, and for giving me every assistance in the

collecting.

I am indebted to the following individuals for placing specimens in existing

collections at my disposal, or where this was not possible, examining specimens

for me: the late Professor J. L. B. Smith (Ichthyology Department, Rhodes

University), Dr. M. L. Bauchot (Muséum National d’ Histoire Naturelle, Paris),

Mr. A. Wheeler (British Museum of Natural History), Dr. V. G. Springer

(Smithsonian Institution, United States National Museum), Professor J. H.

Day (University of Cape Town), Dr. F. H. Talbot (Australian Museum,

Sydney), Dr. Y. Tominaga (Misaki Marine Biological Station, Japan), and

Dr. J. E. Bohlke (Academy of Natural Sciences, Philadelphia).

Many people assisted by collecting and donating specimens for this pro-

ject; I am indebted particularly to the late Prof. J. L. B. Smith, Misses G. G.

Jones, S. C. Penrith, R. M. Tietz and Messrs. C. D. Berrisford, D. H. Eccles,

N. R. Fuller, S. X. Kannemeyer, B. F. Kensley, P. Pretorius for help with

collections and donation of material.

I am especially indebted to my husband, Mr. M. J. Penrith, of the Oceano-

graphy Department, University of Cape Town, who collected most of the

material for this project, and gave much valuable advice and help in the pre-

paration of the manuscript.

The Trustees of the South African Museum thank the South African

Council for Scientific and Industrial Research for a grant in aid of publication.

This work formed part of a thesis submitted for the degree of Ph.D. at

the University of Cape Town in October 1965.

REFERENCES

Au-Utuman, H. S. 1960. Revision of the Pacific forms of the tribe Starksiidi. Tex. 7. Scz. 12:

163-175.

ARAMBOURG, C. 1921. Sur la faune ichtyologique d’Oran. C. r. hebd. Séanc. Acad. Sci., Paris 172:

1232-1245.

ARAMBOURG, C. 1927. Les poissons fossiles d’Oran. Matér. Carie géol. Algér. (Paléont.) 6: 1-298.

ARAMBOURG, C. 1943. Sur la distribution mésogéenne de quelques poissons actuels et fossiles.

— C.4r. hebd. Séanc. Acad. Sci., Paris 217: 462-464.

BARNARD, K. H. 1927. A monograph of the marine fishes of South Africa. Ann. S. Afr. Mus. 21:

419-1065.

BARNARD, K. H. 1935. Notes on South African marine fishes. Ann. S. Afr. Mus. 30: 645-658.

BARNARD, K. H. 1937. Further notes on South African marine fishes. Ann. S. Afr. Mus. 32: 41-67.

BARNARD, K. H. 1948. Further notes on South African marine fishes. Ann S. Afr. Mus. 36:

341-404.

BeaurortT, L. F. pe & CHapmMan, W. M. 1951. The fishes of the Indo-Australian archipelago. IX.

Percomorphi (concluded), Blennoidea. Leiden: Brill.

BEEBE, W. & TEE-VAN, J. 1928. The fishes of Port-au-Prince Bay, Haiti. Zoologica, N.Y. 10:

1-279.

BrrBe, W. & TEE-VAN, J. 1934. A new genus and species of scaleless blenny, Somersia furcata,

from Bermuda. Am. Mus. Novit. 730: 1-3.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE BY]

Bere, C. 1898. Comunicaciones ictiolégicas. Comun. Mus. nac. Buenos Aires 12 9-13.

BLEEKER, P. 1860. Over eenige vischsoorten van de Kaap de Goede Hoop. Natuurk. Tijdschr.

Ned.-Indié 21: 49-80.

Biocu, M. E. & ScHNEIDER, J. G. 1801. Systema ichthyologia iconibus ex illustratum. Post obitum

auctoris onus inchoatum absolvit, correxit, interpolavit Johann Gottlob Schneider. Berolini.

BOuLKE, J. E. 1957. A review of the blenny genus Chaenopsis and a description of a related genus

from the Bahamas. Proc Acad. ‘nat. Sct. Philad. 109: 25-57.

BOutuxe, J. E. & SPRINGER, V. G. 1961. A review of the Atlantic species of the clinid fish genus

Starksia. Proc. Acad. nat. Sci. Philad. 113: 29-60.

Brices, J. C. 1955. A monograph of the clingfishes (order Xenopterygii). Stanford ichthyol. Bull.

6: 1-224.

Briccs, J. C. 1967. Dispersal of tropical marine shore animals: Coriolis parameters or competi-

tion ? Nature, Lond. 216: 350.

Buen, F. pg. 1962. Fauna chilena. Peces de la familia Clinidae. Montemar 2: 53-96.

CADENAT, J. 1950. Poissons de mer du Sénégal. Init. afr. 3: 1-345.

CALDWELL, D. K. 1954. Additions to the known fish fauna in the vicinity of Cedar Key, Florida.

Q. 7. Fla. Acad. Sct. 17: 182-184.

CASTELNAU, F. L. 1860. Note sur les poissons de |’Afrique australe. C. r. hebd. Séanc. Acad. Sci.,

Paris 50: 788-789.

CASTELNAU, F. L. 1861. Mémoire sur les poissons de l’ Afrique australe. Paris: Bailliére.

Cuvier, G. L. C. F. D. 1817. Le régne animal distribué d’aprés son organisation. 2. Poissons. Paris:

Fortin, Masson.

Cuvier, G. L. C. F. D. & VALENCIENNES, A. 1836. Histoire naturelle des poissons. 11. Paris: Levrault.

Dawson, C. 1960. Starksia ocellata, a new sponge inquiline from South Carolina. Copeia 1960: 75.

Day, J. H. 1967. The biology of Knysna estuary, South Africa. Estuaries 1967: 397-407.

Day, J. H., Mittarp, N. A. H. & Harrison, A. D. 1952. The ecology of South African estuaries.

Part III: Knysna, a clear open estuary. Trans. R. Soc. S. Afr. 33: 367-413.

De tFin, F. T. 1898-1901. Catalogo de los peces de Chile. Ruta chil. Hist. nat. 3: 95-99.

EKMAN, S. 1953. Coogeography of the sea. London: Sidgwick & Jackson.

EVERMANN, B. W. & Marsu, M. C. 1900. Descriptions of new genera and species of fishes from

Puerto Rico. Rep. U.S. Commar Fish. 25: 351-362.

Forp, R. F. 1959. A study of the major classification of blennioid fishes based upon cranial nerve evidence.

Thesis submittedfor degree of M.A. (Department of Biological Sciences, Stanford University.)

Fow er, H. W. 1931. Fishes obtained by the Barber Asphalt Company in Trinidad and Vene-

zuela in 1930. Proc. Acad. nat. Sci. Philad. 83: 391-410.

Fow er, H. W. 1934. Fishes obtained by Mr. H. W. Bell-Marley chiefly in Natal and Zululand

in 1929-1932. Proc. Acad. nat. Sci. Philad. 86: 485-514.

Fow ter, H. W. 1942. A list of fishes known from the coast of Brazil. Archos Zool. Est. S. Paulo 3:

115-184.

Fow ter, H. W. 1944. Results of the Fifth George Vanderbilt Expedition (1941) . . . The fishes.

Monogr. Acad. nat. Sci. Philad. 6: 57-529.

Fow.er, H. W. 194.7. Notes on Bahama fishes obtained by Mr. Charles G. Chaplin in 1947,

with descriptions of two new species. Notul. Nat. 190: 1-14.

Fow er, H. W. 1950. Results of the Catherwood-Chaplin West Indies Expedition, 1948. Part

III. The fishes. Proc. Acad. nat. Sci. Philad. 182: 69-93.

Fow er, H. W. 1953. On a collection of fishes made by Dr. Marshall Laird at Norfolk Island.

Trans. R. Soc. N.Z. 81: 257-267.

GarMAN, S. 1899. Reports on an exploration off the west coasts of Mexico, Central and South

America, and off the Galapagos Islands, in charge of Alexander Agassiz, by the U.S. Fish

Commission steamer ‘Albatross’, during 1891. XXVI. The fishes. Mem. Mus. comp. ool.

Harv. 24: 1-431.

GiILBERT, C. M. 1goo. Results of the Branner-Agassiz Expedition to Brazil. Proc. Wash. Acad.

Sci. 2: 161-184.

Gitcurist, J. D. & THompson, W. W. 1908. The Blenniidae of South Africa. Ann. S. Afr. Mus.

6: 97-142.

Git, T. 1860. Monograph of the genus Labrosomus Swainson. Proc. Acad. nat. Sci. Philad. 12:

102-108.

Gostine, W. A. 1968. The suborders of perciform fishes. Proc. U.S. nain. Mus. 124: 1-78.

118 ANNALS OF THE SOUTH AFRICAN MUSEUM

Gray, J. E. 1854. Catalogue of the fish collected and described by Laurence Theodore Gronov, now in the

British Museum. London: Woodfall & Kinder.

GREENWOOD, P. H., Rosen, D. E., Weitzman, S. H. & Myers, G. S. 1966. Phyletic studies of

teleostean fishes, with a provisional classification of living forms. Bull. Am. Mus. nat. Hist.

131: 341-456.

GrirFin, L. T. 1926. Description of New Zealand fishes. Trans. Proc. N.Z. Inst. 56: 538-546.

GuITEL, F. 1893. Observations sur les moeurs de trois Blenniidae, Clinus argentatus, Blennius

montagui et Blennius sphynx. Archs. Zool. exp. gén. (3) 13 325-384.

GUNTHER, A. 1861. Catalogue of the acanthopterygian fishes in the British Museum. 3. London: British

Museum.

Herre, A. W. 1936. Notes on the fishes in the zoological museum of Stanford University. V. New

or rare Philippine fishes from the Herre 1933 Philippine Expedition. Philipp. F. Sct. 59:

357-382.

Herre, A. W. 1939. The Philippine blennies. Philipp. 7. Sci. 70: 315-372.

HILDEBRAND, S. F. 1946. A descriptive catalog of the shore fishes of Peru. Bull. U.S. natn. Mus.

189: 1-530.

Husss, C. 1952. A contribution to the classification of the blennioid fishes of the family Clinidae,

with a partial revision of the eastern Pacific forms. Stanford ichthyol. Bull. 4: 41-165.

Husss, C. 1953a. Revision and systematic position of the blenniid fishes of the genus Weoclinus.

Copeia 1953: 11-23.

Husss, C. 1953). Revision of the eastern Pacific fishes of the clinid genus Labrisomus. Zoologica,

N.Y. 38: 113-136.

Hux.ey, J. S. (ed.) 1940. The new systematics. Oxford: Oxford University Press.

IncER, R. F. 1958. Comments on the definition of genera. Evolution, N.Y. 12: 370-384.

INTERNATIONAL TRUST FOR ZOOLOGICAL NOMENCLATURE, 1961. International code of zoological

nomenclature adopted by the XV International Congress of Zoology, London, July 1958. London:

The Trust for the International Commission on Zoological Nomenclature.

INTERNATIONAL TRUST FOR ZOOLOGICAL NOMENCLATURE, 1964. International code of zoological

nomenclature adopted by the XV International Congress of zoology, London, July 1958. London:

‘The Trust for the International Commission on Zoological Nomenclature.

Jackson, P. B. N. 1950. The fishes of the intertidal zone of the Cape Peninsula. Thess for the degree of

M.Sc. (Zoology Department, University of Cape Town).

JORDAN, D. S.-1887. Description of two new species of fishes from South America. Proc. Acad.

nat. ‘Sci. Philad. 39: 387-388.

Jorpan, D. S. & Starks, E. C. 1906. List of fishes collected on dicacen and Yalan, offshore

' ‘islands of southern Japan, by Robert Van Vleck Anderson, with descriptions of seven new

species. Proc. U.S. natn. Mus. 303-695-706. | ;

Kuausewitz, W. 1958. Fische aus dem Atlantik und Pasifik. Senckenberg. bah 39: “784 Lie

Kner, R. 1865-7. Reise dsterreichischen Fregatte ‘Novara’ um die Erde in den Jahren 1857-1859, unter

a Befehlen des Commodore B. von Wiillerstorf-Urbair. Zoologischer Theil, Fische. Wien: K.K.

Hof-u. Staatsdruckerei.

LackpEpE, B. G. E. DE. 1800. Histoire naturelle des poissons. 2. Paris: Plasson.

LicHTENSTEIN, M. H. C. 1823. Verzeichniss der Doubletten des zoologischen Museums der Konigliche

Universitit zu Berlin, nebst Beschreibung vieler bisher unbekannter Arten von Sdugethieren, Vogeln,

Amphibien und Fischen. Berlin: Trautwein.

Linnaeus, C. 1758. Systema naturae . . . Editio decima, reformata. 1. Holmiae.

MakusHok, V. M. 1958. The morphology and classification of the northern blennioid fishes

(Stichaeoidae, Blennioidei, Pisces). Trudy. zool. Inst. Akad. Nauk. S.S.S.R. 25: 3-129. (Trans-

lation by A. Gosline, U.S. National Museum, April 1959).

McCuttocn, A. R. 1908. Studies on Australian fishes. Rec. Aust. Mus. 7: 36-43.

McCuttocu, A. R. 1915. Notes on, and descriptions of, Australian fishes. Proc. Linn. Soc. N.S.W.

_ 402 259-277.

McCuttocn, A. R. 1922. The fishes and fish-like animals of New South Wales. Aust. Zool. 2:

86-130.

McCuttocn, A. R. 1929. A check-list of the fishes recorded from Australia. Mem. Aust. Mus. 5:

1-534:

McLeay, W. 1882. Descriptive catalogue of the fishes of Australia, part 3. Proc. Linn. Soc. N.S.W.

6: 1-138.

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE I19

MarsHAa.t, T. C. 1957. Ichthyolog ical notes. Ichthyol. Notes Qd 1: 117-137.

MatruHew, W. D. 1915. Climate and evolution. Ann. N.Y. Acad. Sci. 24: 171-318.

Meap, G. W. 1958. A catalog of the type specimens of fishes formerly in the collection of the

Department of Tropical Research, New York Zoological Society. <oologica, N.Y. 43:

131-134.

METZELAAR, J. 1922. On a collection of marine fishes from the Lesser Antilles. Bijdr. Dierk.

Feestnummer: 133-141. -

Mitwarop, N. E. 1967. The Clinidae of Western Australia (Teleostei, Blennioidae). 7. Proc. R.

Soc. West. Aust. 50: 1-9.

Morrow, J. E. 1957. Shore and pelagic fishes from Peru, with new records and the description

of a new species of Sphaeroides. Bull. Bingham oceanogr. Coll. 16: 5-55.

Ocitsy, J. D. 1886. Descriptions of new fishes from Port Jackson. Proc. Linn. Soc. N.S.W. 10:

225-230.

OtsEen, A. M. 1958. New fish records and notes on some uncommon Tasmanian species. Pap.

Proc. R. Soc. Tasm. g2? 155-159.

Pappe, L. 1853. Synopsis of the edible fishes at the Cape of Good Hope. Cape Town: Van de Sandt de

Villiers & Tier.

Parr, A. E. 1930. Teleostean shore and shallow-water fishes from the Bahamas and Turks

Island. Bull. Bingham oceanogr. Coll. 3: 1-148.

PenritH, M.-L. 1965a. Studies on the South African Clinidae. I. Description of a new species of

Pavoclinus, and redescription of Gynutoclinus rotundifrons (Barnard). Ann. S. Afr. Mus. 48:

QII-217.

PenritH, M.-L. 1965). Note on an extension of the known ranges of distribution of some species

of Clinidae (Pisces). S. Afr. F. Sci. 613 423-424.

PEnrRITH, M.-L. 1967. Studies on the South African Clinidae. II. Two new species of the genus

Clinus from the western Cape. Ann. S. Afr. Mus. 50: 43-59.

ROSENBLATT, R. H. 1963. Speciation in marine shore fishes. In Harding, J. P. & Tebble, N., eds.

Speciation in the sea: 171-180. London: Systematics Association.

ROSENBLATT, R. H. & WALKER, B. W. 1963. The marine fishes of the Galapagos Islands. Occ.

Pap. Calif. Acad. Sci. 44%-97-104.

ScHuttz, L. P. 1949. A further contribution to the ichthyology of Venezuela. Proc. U.S. nain.

Mus. 99: 1-211.

Scott, E. O. G. 1935. Observations on some Tasmanian fishes. Part II. Pap. Proc. R. Soc. Tasm.

1934: 63-73.

Scott, E. O. G. 1939. Observations on some Tasmanian fishes. Part IV. Pap. Proc. R. Soc. Tasm.

1938: 139-159. |

Scott, E. O. G. 1955. Observations on some Tasmanian fishes. Part VII. Pap. Proc. R. Soc.

Tasm. 89: 131-146.

SmiTH, C. L. 1957. Two new clinid blennies (Malacoctenus) from Puerto Rico Occ. Pap. Mus.

Kool. Univ. Mich. 585: 1-15.

Smit, J. L. B. 1931. New fishes from South Africa. Rec. Albany Mus. 4: 145-160.

SmiTH, J. L. B. 1935. New and little-known fishes from South Africa. Rec. Albany Mus. 4: 169-235.

SmiTH, J. L. B. 1937. New records of South African fishes. Ann. Natal Mus. 8: 167-197.

Situ, J. L. B. 1945. The fishes of the family Clinidae in South Africa. Ann. Mag. nat. Hist. (11)

T2? 535-546.

SmitTH, J. L. B. 1947a. Brief revisions and new records of South African marine fishes. Ann. Mag.

nat. Hist. (11) 14: 335-346.

SmitTH, J. L. B. 194.7. New clinid fishes from South Africa. Ann. Mag. nat. Hist. (11) 14: 732-736.

SMITH, J. L. B. 1949. The sea fishes of southern Africa. Cape Town: Central News Agency.

SmiTH, J. L. B. 1953. The sea fishes of southern Africa. Rev. ed. Cape Town: Central News Agency.

SmiTH, J. L. B. 1960. A new species of South African clinid fish. Ann. Mag. nat. Hist. (13) 3:

689-691.

SmitH, J. L. B. 1961. Fishes of the family Xenopoclinidae. Ichthyol. Bull. Rhodes Univ. 20: 351-356.

Situ, J. L. B. 1962. Fishes from the Cape described by Lichtenstein in 1823. S. Afr. F. Sct.

58: 390-40.

SMITH, J. L. B. 1966. Certain rare fishes from South Africa with other notes. Occ. Pap. Dept.

Ichthyol. Rhodes Univ. 7: 65-80.

SPRINGER, V. G. 1954. Western Atlantic fishes of the genus Paraclinus. Tex. 7. Sci. 6: 422-441.

120 ANNALS OF THE SOUTH AFRICAN MUSEUM

SPRINGER, V. G. 1955. The taxonomic status of the fishes of the genus Stathmonotus, including a

review of the Atlantic species. Bull. mar. Sci. Gulf Caribb. 5: 66-80.

SPRINGER, V. G. 1958. Systematics and zoogeography of the clinid fishes of the subtribe Labri-

somini Hubbs. Publs Inst. mar. Sci. Univ. Tex. 5: 418-491.

SPRINGER, V. G. 1964. A review of: A revised classification of the blennioid fishes of the American

family Chaenopsidae by J. S. Stephens, Jr. 1963. Copeza 1964: 591-593.

SPRINGER, V. G. 1966. Medusablennius chani, a new genus and species of blennioid fish from the

Tuamotu Archipelago: its implication on blennioid classification. Copeia 1966: 56-60.

SPRINGER, V. G. 1968. Osteology and classification of the fishes of the family Blenniidae. Bull.

U.S. natn. Mus. 284: 1-85.

Starks, E. C. 1913. The fishes of the Stanford Expedition to Brazil. Leland Stanford jr Univ.

Publs Univ. Ser. 12% 1-77.

StTEInITz, H. 1950. On the zoogeography of the teleostean genera Salarias, Ophioblennius and

Labrisomus. Archo zool. ital. 35: 325-348. (Not seen).

STEPHENS, J. S. Jr. 1961. Range extension of the temperate blennioid fish, Neoclinus bryope, into

the tropical western Pacific. Copeia 1961: 484-485.

STEPHENS, J. S. Jr. 1963. A revised classification of the blennioid fishes of the American family

Chaenopsidae. Univ. Calif. Publs Zool. 68: 1-133.

SuzukI, K. 1964. Results of the Amami Expedition. II. Fishes. Rep. Fac. Fish. Univ. Mie 5: 153-

188.

Swain, J. 1882. A review of Swainson’s genera of fishes. Proc. Acad. nat. Sci. Philad. 34: 272-284.

Swainson, W. 1839. The natural history and classification of fishes, amphibians and reptiles,

or monocardian animals. In The Cabinet Encyclopaedia. 8. London: Longman, Orme, Brown,

Green & Longmans.

TuHompson, W. W. 1918. Catalogue of the fishes of the Cape Province. Mar. biol. Rep. Cape Tn 4:

75-177:

WHITLEy, G. P. 1929. Johnston’s memoranda relating to the fishes of Tasmania. Pap. Proc. R.

Soc. Tasm. 1928: 44-69.

WHITLEY, G. P. 1945. New sharks and fishes from Western Australia. Aust. Zool. 11: 1-42.

Wuit ey, G. P. 1956. Ichthyolegical notes. Aust. Zool. 12: 251-261.

Wait ey, G. P. 1959. Ichthyological snippets. Aust. Zool. 12: 310-323.

APPENDIX A. Data for fig. 3 (comparison of upper jaw lengths of South African Clinidi).

(Data as % of head length.)

Std. Std. Std. No. of

dev. err. err. fish

Range Mean o om 20m N

Clinus anguillaris a -- 36°3-43:0 40°0 0°35 0:07 0-14 25

Clinus striatus .. cS fe 37 OAdar 40°2 — oe = 9

Clinus capensis .. aes eal SOAs Ng 2°48 0°51 1-02 25

Clinus dorsalis .. We -. 34°8-50°0 40°1 3°63 0:74 1°48 25

Clinus navalis .. at -- 38°0-44°5 40°7 — — — 5

Clinus acuminatus a 2. 35°9-45°5 40°2 2°53 0°52 1-04 25

Clinus agilis we, 8 .. 40°0-50°0 44°2 2°46 0°52 1°04 25

Clinus berrisfordi . . Me .- 36°4-50°0 43°3 3°94 1°02 2°04 16

Ciinus brevicristatus se -. = 38°5-46°5 Ane 1°70 036 0-72 23

Clinus cottoides .. if .. 43:°7-60°0 49°1 4°17 0°85 1-70 25

Clinus helenae... ae 22 89°5=52°5 44.°3 — — = 3

Clinus latipennis .. a 2s 34°5-43°5 39°3 — — — 3

Clinus heterodon .. ihe -. 36°4-50°0 44°2 4:26 0°87 Lan 25

Clinus robustus .. a 2) 45°5-53°5 49°4 — — —_ 8

Clinus superciliosus we fol Biase Ris V0) 41°2 2°85 0°58 1:16 25

Clinus taurus ae a SO BXSO ets 90) 44°4 4°36 0°97 1°94 19

Clinus venustris .. AY -. 35°3-48-0 41°2 3°92 0:82 ~=1°64 25

Clinus woodi oe a -» 48°0-51°5 49°8 — — = 5

Pavoclinus mus .. a se eOM Oa 29°8 B28 0-46 0°92 25

Pavoclinus graminis. oe .. 27°6-38°6 330 2°45 0°51 L02 25

THE SYSTEMATICS OF THE FISHES OF THE FAMILY CLINIDAE

Pavoclinus laurentii

Pavoclinus litorafontis

Pavoclinus mentalis

Pavoclinus pavo

Pavoclinus profundus

Pavoclinus fucorum a

Blennioclinus brachycephal.

Blenniclinus stella

Clinoporus biporosus ;

Gynutoclinus rotundifrons . .

36-0-41°

33°3-42°

ga) 0-40"

22°0-43°

36°4

32°4-42°

28° 6-36:

31° 547°

34° 1-43°

40°0-43°

ono Oo

oonwn ©

“I “I

COW NHB HO

1°39

Deay

4°36

2°06

0-40

0°45

0:89

0:42

0:80

0°86

—-

1-78

0:84

}

ove

INSTRUCTIONS TO AUTHORS

Based on

CONFERENCE OF BIOLOGICAL EDITORS, COMMITTEE ON FORM AND STYLE. 1960.

Style manual for biological journals. Washington: American Institute of Biological Sciences.

MANUSCRIPT

To be typewritten, double spaced, with good margins, arranged in the following order:

(1) Heading, consisting of informative but brief title, name(s) of author(s), address(es) of

author(s), number of illustrations (plates, figures, enumerated maps and tables) in the article.

(2) Contents. (3) The main text, divided into principal divisions with major headings; sub-

headings to be used sparingly and enumeration of headings to be avoided. (4) Summary.

(5) Acknowledgements. (6) References, as below. (7) Key to lettering of figures. (8) Explana-

tion to plates.

ILLUSTRATIONS

To be reducible to 4 in. X 7 in. (7 in. including caption). A metric scale to appear with

all photographs.

REFERENCES

Harvard system (name and year) to be used: author’s name and year of publication given

in text; full references at the end of the article, arranged alphabetically by names, chronologi-

cally within each name, with suffixes a, b, etc. to the year for more than one paper by the

same author in that year.

For books give title in italics, edition, volume number, place of publication, publisher.

For journal articles give title of article, title of journal in italics (abbreviated according to

the World list of scientific periodicals. 4th ed. London: Butterworths, 1963), series in

parentheses, volume number, part number (only if independently paged) in parentheses,

pagination.

Examples (note capitalization and punctuation)

BuLLoucuH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FiscHER, P.-H. 1948. Données sur la résistance et de le vitalité des mollusques. 7. Conch., Paris

88: 100-140. |

FiscHER, P.-H., DuvAt, M. & Rarry, A. 1933. Etudes sur les échanges respiratoires des littorines.

Archs Zool. exp. gén. 74: 627-634.

Koun, A. J. 1960a. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee

region of Ceylon. Ann. Mag. nat. Hist. (13) 2: 309-320.

Koun, A. J. 1960). Spawning behaviour, egg masses and larval development in Conus from

the Indian Ocean. Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyphacophora, Gastropoda marina, Bivalvia. In Schultze, L.

Koologische und anthropologische Ergebnisse einer Forschungsreise im westlichen und zentralen Siid-

Afrika. 4: 269-270. Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16: 269-270.

ZOOLOGICAL NOMENCLATURE

To be governed by the rulings of the latest International code of zoological nomenclature issued

by the International Trust for Zoological Nomenclature (particularly articles 22 and 51).

The Harvard system of reference to be used in the synonymy lists, with the full references

incorporated in the list at the end of the article, and not given in contracted form in the

synonymy list.

Example

Scalaria coronata Lamarck, 1816: pl. 451, figs 5 a, 6; Liste: 11. Turton, 1932: 80.

f ie |

Mie

& ye i ne) My

RM i YY ATA

AE Teoh done bul ‘taht Nie 4

i Sbe) andi re fe ee eae Pos ee A

San sea weel TT) i i ee era Sa a bong Obie stein:

ih ia if } ae ‘ :- D me} ‘ an i + %

Lis" nevis (astute Boke) aeRO,

‘ i Ni ie he ih 1 A Pepe ty LSC oa hebtane)) ce kta 3 th

saa hiss PUNE sivids Rea we Ua

we Ant giagk Weg can eaectnge bats ohygices ange “|

; ae AT OY AE UE wong i ol YS iy 1 (a) Gah

rr i ey Niko cha anu yn (0k eras

% bt if oO) (aRAy

“y te f ' “pit

i i LS tae + Un

j se » wohil¢

| '

es, ' P v's

| i it

h y

’ ‘ ae

;

, se? PAL

Li, “i che 8 is fee ik

: ts z \y | L 48 wit Ee Pe ith

| , , Wt ora) ‘Winey "

2 J 7 P 5 +3 t , 1 J ny ite a eh

| ;

O/-66

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 55 +#Band

October 1969 Oktober

Part. 2 Deel

_ APLETODON PELLEGRINI (CHABANAUD) AND

OTHER CLINGFISHES (PISCES: GOBIESOCIDAE)

FROM SOUTH WEST AFRICA

} MARY-LOUISE PENRITH

alt HSUny, ,

o™ 4,

Cape Town Kaapstad

DEC SYS Sos

f/8RARIED

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material -

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na beskikbaarheid

van stof

Verkrygbaar van die Suid-Afrikaanse iiceennt Posbus 61, Kaapstad

OUT OF PRINT/UIT DRUK

1, 2(1, 3, 5, 7-8), 3(1-2, 5, t-p-l.), 5(2, 5, 7-9),

6(1, t.—p.i.), TO By) 8, g(1-2), FO(1—2)55

(L250 fpr Pale) an 2 lene (2) 29 om Ta uae

44(4).

Price of this part/Prys van hierdie deel

R1.20

Trustees of the South African Museum © Trustees van die Suid-Afrikaanse Museum

1969

Printed in South Africa by ia In Suid-Afrika gedruk deur

The Rustica Press, Pty., Ltd. Die Rustica-pers, Edms., Bpk.

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

APLETODON PELLEGRINI (CHABANAUD) AND OTHER

CLINGFISHES (PISCES: GOBIESOCIDAE) FROM

SOUTH WEST AFRICA

Mary-LovuisE PENRITH

South African Museum, Cape Town

(With 5 figures)

[MS. received 14. May 1969]

CONTENTS

PAGE

Introduction . } : : Hh) neg

Systematic discussion . ; ey yp zs

Remarks on distribution. mae

Summary . OF calwent ait a 5, Ge

eerer iedee stent ‘ . Stee

INCIERCHCES RE a oe hes wy 2h B84

INTRODUCTION

It is not unusual that the presence of a group such as the clingfishes, most

of which are small and inconspicuous, should remain undetected in areas in

which very little collecting has been done. Smith (1964), discussing the cling-

fishes of the western Indian ocean and the Red Sea, pointed out that no cling-

fishes had up to that time been recorded from tropical West Africa south of

Ghana. His distribution maps showed no records of Gobiesocidae on the

western African coast between Ghana in the north and Port Nolloth in the

south.

Smith (1964) suggested that the lack of records of Gobiesocidae from the

west coast of Africa was probably due rather to lack of sampling of the area

than to a genuine deficit of clingfishes there, and this appears to be true.

Collecting on the South West African coast by the South African Museum

and the State Museum, Windhoek, has indicated that three species of Gobie-

socidae occur intertidally in South West Africa, and one of these has also been

found in southern Angola.

Smith (1964) listed five species of Gobiesocidae ee knysnaensis Srnith,

Chorisochismus dentex (Pallas), Eckloniaichthys scylliorhiniceps Smith, Lepadichthys

coccinotaenia Regan, Pherallodus smithi Briggs) as occurring intertidally in the

South African region. Of these, Pherallodus smithi is known from the single type

specimen collected at Durban in 1934, and Lepadichthys coccinotaenia, a species

of the tropical western Indian ocean, occurs rarely on the Natal coast. The

other three species are found further westwards. Briggs (1955) recorded four

species from north-west African shores and islands (Afletodon pellegrini (Chaba-

123

Ann. S. Afr. Mus. 55(2), 1969, 123-134, 5 figs.

124 ANNALS OF THE SOUTH AFRICAN MUSEUM

rly ae

ANGOLA

‘Kunene R,

A\Honolulu

A Liideritzbucht

O -Apletodon peliegrini

4 4 —Chorisochismus dentex

| @ 4 - Eckloniaichthys scylliorhinicleps

Fig. 1. Records of clingfishes from the South West African region.

APLETODON PELLEGRINI AND OTHER CLINGFISHES FROM S.W.A. 125

naud), Lepadogaster candolle: Risso, Lepadogaster lepadogaster lepadogaster (Bonna-

terre), Lepadogaster zebrina Lowe), while the genus Diplecogaster is represented

in deeper water in this area, as it is off southern Africa (Briggs, 1955; Smith,

1964). Two specimens of Apletodon pellegrini in the Muséum d’ Histoire Naturelle,

Paris, collected at Annobon island just south of the equator by A. Stauch,

represent the southernmost record for any clingfish in the West African

tropical region.

The three species here recorded from South West Africa are Chorisochismus

dentex, Eckloniaichthys scylliorhiniceps, and Apletodon pellegrim. The localities

from which these species were recorded are shown in figure 1. The South West

African records of Chorisochismus dentex and Eckloniaichthys scylliorhiniceps represent

considerable north-westward extensions of their distributional ranges.

The identification of the South West African specimens of Apfletodon

necessitated comparison with both northern and southern (Apletodon knysnaensis)

specimens, and this comparison led me to conclude that Apletodon pellegrint

and A. knysnaensis are not specifically distinct in spite of their apparent geo-

graphical separation.

Abbreviations used

P.M. — Muséum d’Histoire Naturelle, Paris.

R.U.C.—J. L. B. Smith Institute of Ichthyology, Rhodes University,

Grahamstown.

S.A.M.—South African Museum, Cape Town.

S.M.—State Museum, Windhoek, South West Africa.

U.Z.M. —Universitets Zoologiska Museum, Copenhagen, Denmark.

SYSTEMATIC Discussion

Apletodon pellegrint (Chabanaud, 1925)

(Fig. 2)

Lepadogasier (Mirbelia) Pellegrini Chabanaud, 1925: 283, 286.

Apletodon pellegrini: Briggs, 1955: 27 (synonymy and references). Blanc, Cadenat & Stauch,

1968: 252.

Apletodon knysnaensis Smith, 1964: 590, pl. 93 B, C, pl. 96 A-F, pl. 97 B-F.

Material on which description 1s based: 1 specimen, 29 mm standard length,

S.A.M. 24942, intertidal pool, Mogamedes, southern Angola; 1 specimen, 26

mm standard length, S.M. 5216, intertidal pool, Rocky Point, S.W.A.; 2

specimens, 25 & 22 mm standard length, S.A.M. 25247, intertidal pools,

Rocky Point, S.W.A.; 1 specimen, 34 mm standard length, S.A.M. 25263,

intertidal pool, Rocky Point, S.W.A.; 2 specimens, 42 & 32 mm standard

length, S.A.M. 25257, intertidal pools, Méwe Bay, S.W.A.; 1 specimen, 34 mm

standard length, $.M. 5268, intertidal pool, Méwe Bay, S.W.A.

Other material examined: 3 specimens, P.M. 25-239, 25-240, 25-241, Cap

126 ANNALS OF THE SOUTH AFRICAN MUSEUM

Blanc (syntypes); 1 specimen, P.M. 07-257, Senegal (syntype); 2 specimens,

P.M. 59-65, 59-66, Senegal; 2 specimens, P.M. 1965-623, Annobon Island;

3 specimens, U.Z.M., Senegal (Dakar); 1 specimen, R.U.C., Knysna (allotype

of A. knysnaensis); 1 specimen, R.U.C., Knysna (holotype of A. knysnaensis) ;

5 specimens, R.U.C., Knysna; 1 specimen, R.U.C., Lambert’s Bay; 5 speci-

mens, R.U.C., Bird Island, Port Elizabeth; 4 specimens, S.A.M. 25250,

Froggy Pond, False Bay; 9 specimens (10-20 mm standard length), S.A.M.

25257 & S.M. 5269-5271, Mowe Bay, S.W.A.

Description: D. 5-6; A. 5-6; P. 21-22; C. 10-12. Gills three and one half.

Third gill arch with six minute gill-rakers. Maximum depth of body 4:4—5-7

in standard length. |

Head broad, depressed, 2:3-2-4 in standard length, maximum breadth of

head 2:75-3:25 in standard length. Eye round, orbit diameter 4-5-6 in head.

Interorbital width greater than orbit diameter, 3-3-7 in head. Snout a little

longer than eye, rounded, 3-3-7 in head. Anterior nostril tubular, posterior

nostril a simple pore. Mouth terminal, upper jaw 2:25-3-7 in head, breadth

of mouth at posterior angles of jaw 2-2-8 in head. Teeth conical, caniniform;

four distinctly compressed, enlarged, incisiform teeth in front of upper jaw of

25, 34 (both), 32, and 42 mm specimens.

Dorsal and anal fins in posterior fourth of body. Anus anterior to dorsal

origin, slightly nearer to pectoral tip than to caudal base. Pelvic disc double,

fringed with small branched papillae, disc length 2-9—-3-8 in standard length.

Caudal peduncle very short, 4:5-5°5 in head length, deeper than long, depth

3-3-7 in head length. Caudal fin rounded.

Colouring: Upper parts dark red to reddish brown in Rocky Point specimens

and 34 mm Mowe Bay specimen, light green in 32 and 42 mm Mowe Bay

specimens. Angolan specimen red with greenish mottling. A light bar across

interorbital in the three largest specimens. Underparts of all specimens lighter,

creamy buff. Throat with fine red spots in all except the 32 and 42 mm speci-

mens. Pectoral fins hyaline; pelvic disc hyaline in most specimens, orange in

32 and 42 mm specimens. Dorsal, anal, and caudal fins barred with main

colour of upper parts.

Habitat: Intertidal; the Rocky Point specimens were collected amongst mussels

(Perna perna (Linnaeus)) and coralline algae in pools at the middle levels of

the intertidal zone of the shore. The Méwe Bay specimens were collected under

large stones in pools at the middle levels of the shore amongst brachiopods

(Discinisca tenuis (Sowerby)).

Distribution: Dakar, Senegal, to the Cape Verde, Canary, and Madeira Islands

(Briggs, 1955); Annobén Island (Blanc, Cadenat & Stauch, 1968); Lambert’s

Bay to Port Elizabeth (Smith, 1964).

Remarks

Three species have been included in the genus Apletodon Briggs: A. pellegrini;

A. microcephalus (Brook), from Sicily to Scotland (Briggs, 1955), with a sub-

APLETODON PELLEGRINI AND OTHER CLINGFISHES FROM S.W.A. 127

5mm

Fig. 2. Apletodon pellegrint (Chabanaud).

species A. m. bacescut Antoniu Murgoci from the Black Sea (Murgoci, 1964);

and A. knysnaensis Smith, from the southern coast of South Africa from Lambert’s

Bay to Port Elizabeth (Smith, 1964).

Comparison of specimens of A. pellegrint and the typical form of A. micro-

cephaius indicates that they are two clearly distinct species which differ in a

number of features, the most marked difference being the considerably narrower

head of A. microcephalus. A. pellegrini and A. knysnaensis are by no means as

readily separable.

Smith (1964) distinguished A. knysnaensis from A. pellegrini chiefly on the

length of the head in profile and the position of the anus with regard to the

pectoral fin and the caudal base. He stated that A. knysnaensis was extremely

similar to A. pellegrini, and that, but for the great distance separating the two

forms geographically, he might be tempted to accord the southern form sub-

specific status only.

128 ANNALS OF THE SOUTH AFRICAN MUSEUM

I can find no difference between the South West African and Angolan

specimens and specimens of A. pellegrin from the northern hemisphere and

Annoboén Island. Furthermore, comparison of the northern hemisphere, Angolan,

and South West African specimens of A. pellegrint with specimens, including the

types, of A. knysnaensis, indicates that these two species are in fact indistinguish-

able. The position of the anus with regard to the pectoral tip and the caudal

base appears to vary, but in the majority of specimens (all the northern

hemisphere, Annobon, and Angolan/South West African specimens and

several of the southern ones) the anus was found to be slightly nearer the

pectoral tip. As far as the profile head length is concerned, there is much

overlap: measurements are given in table 1. Judging from the proportions

given by Smith (1964), he may have made his measurements slightly differently

from mine, but the difference between southern and northern forms is in any

event so slight as not to warrant even subspecific separation.

TABLE I

Comparison of head lengths (profile) of Apletodon pellegrinit from the northern hemisphere,

Angola/South West Africa, and the southern coast of South Africa.

Locality Standard length Head length in

(mm) standard length

Cap Blanc . , : : 25 2a

Cap Blanc . : ‘ ; 29 B)A-

Senegal. ! : 3 é 28 22

Senegal. ; : : : 20 2°2

Dakar . : j : : 28 F437)

Dakar . : 5 : : 20 2°5

Dakar . : : : : 18 Bag

Mocamedes . ? ; , 29 2°4

Rocky Point . ; : : 34 2°4

Rocky Point . , ; : 25 P40:

Rocky Point . : : ‘ 26 2°*4

Rocky Point . : P : 22 2A

Mowe Bay . : : ; 42 2a

Mowe Bay . ; ; : 34 2°4

Mowe Bay . ; ; : 32 Qe

Lambert’s Bay : 5 ; 32 POG |

False Bay : : . 25 24g:

False Bay : : : 21 2723:

False Bay : : . 20 2°5

False Bay 99): , ‘ A 21 2°3

Knysna ; ; : , 33 Qf

Knysna , : : ; 24. 2°4

Knysna : ‘ : E 32 2°5

Knysna : : : : 29 29

Knysna 5 : é : 34. 2°4

Knysna ; ; 5 , 24 2°4

Knysna é : : : 23 PING |

Port Elizabeth : : q 25 2°5

Port Elizabeth : : : 23 2°6

Port Elizabeth ; : ‘ 26 2:6

Port Elizabeth i : : 25 2°8

Port Elizabeth ; ; ; 23 Bia

APLETODON PELLEGRINI AND OTHER CLINGFISHES FROM S.W.A. 129

It seems then that there are two species of the genus Apletodon, one con-

fined to the North Atlantic and Mediterranean region, and one widely distri-

buted along the African coast, including offshore islands.

Chorisochismus dentex (Pallas, 1769)

(Fig. 3)

Cyclopierus dentex Pallas, 1769: 6, pl. 1.

Chorisochismus dentex: Giinther, 1861: 490. Barnard, 1927: 423. Smith, 1949: 384, pl. 87 fig.

1086; 1964: 587, pl. 94 A, B, C. Briggs, 1955: 40 (synonymy and references).

Material : 7 specimens, 92—167 mm standard length, $.M. 4929-4934, intertidal

pools, Rocky Point, S.W.A.; 6 specimens, 99-190 mm standard length, $.A.M.

25221, intertidal pools, Rocky Point, S.W.A.; 1 specimen, 115 mm standard

length, $.M. 4405, intertidal pool, 42 miles north of Unjab River mouth,

S.W.A.; I specimen, 133 mm standard length, $S.A.M. 25228 intertidal pool,

42 miles north of Unjab River mouth; 1 specimen, 128 mm standard length,

S.A.M. 25231, Honolulu, $.W.A.; 13 specimens, 11-115 mm standard length,

50mm

Fig. 3. Chorisochismus dentex (Pallas).

I30 Ao ~ \'" ANNALS OF THE SOUTH. AFRICAN MUSEUM —

S.A.M. 24209, Liideritzbucht, S.W.A.; 22 specimens, 11-84 mm standard

length, S.A.M. 25248, Liideritzbucht, S. W.A.

Description: D. 7-8; A. 6; P. 21-23; C. 9-10. Gills three and one half. Reel

gill arch with four minute gill-rakers. Maximum depth of body 3-2—4-1 in

standard length in 84-190 mm specimens, 4:5-5°3 in standard length in

smaller specimens. _ .

Head broad, not strongly depressed, 1-9-2°5 in standard length, breadth

of head about equal to, usually very slightly less than, length; 1-9—2-9 in standard

length. Eye round, orbit diameter 3-2—4-9 in head. Interorbital width about

equal to eye in large specimens, less in smaller specimens, 2°7—5:2 in head

length. Snout a little longer than eye, bluntly rounded, '2:6—3-7 in head.

Anterior and posterior nostrils tubular, anterior nostril with a simple expanded

dermal flap on posterior rim. Mouth terminal, upper jaw 2:4-3:3 in head,

breadth of mouth at posterior angles of jaws 1-4~3-2 in head, broadening with

increase in standard length. Teeth mainly conical, enlarged, prominent, and

slightly compressed in front.

- Dorsal and. anal fins in posterior third of body. Anus well anterior to

dorsal origin. Urogenital papilla of male very much larger than that of female.

Pelvic disc single, length 2-6—3-2 in standard length. Caudal fin subtruncate.

Caudal peduncle 2:3-4°3 (3-4 in most specimens) in head length, longer than

deep, depth 4:3—5:8 in head length.

Colouring: Very variable; most specimens pinkish or greenish mottled to match

surroundings. Underparts creamy.

Habitat: Intertidal and infratidal; clings to rocks.

Distribution: Alexander Bay (South African Museum) ; Port Nolloth to northern

Natal (Zululand) coast (Smith, 1964). /

Remarks

This is the only species of clingfish that is common on South West African

shores.

Eckloniaichthys scylliorhiniceps Smith, 1943

(Fig. 4)

Eckloniaichthys scylliorhiniceps Smith, 1943: 67, fig. 1; 1949: 384, pl. 87 fig. 1087; 1964: 584,

ge pl. 92 D, pl. 93 D,; E. Briggs, 1955: 72.

Material: 1 specimen, 15 mm standard length, S.A.M. 25240, intertidal pool,

Agate Beach, Liideritzbucht. )

Description: D. 5; A. 5; P. 18; C. 10. Gills three. Each gill arch with four

minute gill-rakers. Maximum depth of body 7:5 in standard length.

Head depressed, 3-0 in standard length, maximum breadth of head 5:0

in standard length. Eye round, orbit diameter 5-0 in head. Interorbital width

equal to orbit diameter. Snout twice length of eye, narrowly rounded, 2-5 in

head. Mouth subterminal, inferior, upper jaw 5:0 in head, width of mouth at

APLETODON PELLEGRINI AND OTHER CLINGFISHES FROM S8.W.A. I31I

——_———— ee |

5mm

Fig. 4. Eckloniaichthys scylliorhiniceps Smith.

posterior angles of jaws 5:0 in head. Teeth minute, bluntly conical in upper

jaw, lower jaw with six blunt incisiform teeth at front of jaw, followed by three |

caniniform teeth on either side.

_ Dorsal and anal fins in posterior fourth of body. Anus in front of dorsal

origin. A flat fleshy pad on lower part of pectoral base. Pelvic disc single, length

50 in standard length. Caudal peduncle 1-7 in head length, much longer

than deep, depth 5-0 in head length. Caudal fin rounded.

Colouring: Upper parts plain dark olive brown, lower parts creamy.

Habitat: Clinging to kelp stipe in intertidal gully.

Distribution: Sea Point (west coast of Cape Peninsula) (South African Museum) ;

False Bay to East London (Smith, 1964); Kei River mouth (South African

Museum).

Remarks

This small, rather rare clingfish is always found attached to algae. On the

west coast, where it has been found only on kelp (Ecklonia and Laminaria), it

132 ANNALS OF THE SOUTH AFRICAN MUSEUM

is brown; from False Bay eastwards it usually occurs on Caulerpa, and is then

bright green.

The Liideritzbucht specimen is a female; males of this species are easily

distinguished by the very large urogenital papilla.

REMARKS ON DISTRIBUTION

The known distribution of the three species of clingfishes recorded from

South West Africa is shown in figure 5. Two of the species, Chorisochismus

dentex and Eckloniaichthys scylliorhiniceps, appear to be endemic to the southern

African region, while the third is very widely distributed from northern West

Africa to southern Africa.

Chorisochismus dentex is the only really common intertidally occurring

clingfish on the southern African coast. It is one of the few endemic southern

African intertidal fish species that extend to the northern coast of South West

Africa.

Eckloniaichthys scylliorhiniceps is known from relatively few specimens,

probably mainly because it is a minute and very cryptic species and its habitat

is difficult to sample. Both it and Apletodon pellegrini are likely to be found at

further localities.

At the present state of knowledge, there are several considerable gaps in

the distribution of Apletodon pellegrini, the greatest of which are the stretches of

coast between Dakar and Annobén Island, Annoboén Island and Mocgamedes,

and Rocky Point and Lambert’s Bay. Considering that these fishes are small

and inconspicuous, and the intertidal zone of the shore in tropical West Africa

is poorly known, at least some of these gaps may be apparent rather than real.

However, Rocky Point and Lambert’s Bay both lie in transitional areas where,

owing to changing hydrographic conditions, overlap of different faunal provinces

occurs. Between them is the stretch of coast where very low temperatures

prevail owing to the Benguela upwelling system; if Apletodon pellegrini really is

absent from this region, its absence may be related to hydrographic conditions.

SUMMARY

Three species of clingfishes are recorded from South West Africa, a

region from which no clingfishes were previously known. In the light of the

South West African material, a southern African species is synonymised with

a species from northern West Africa. The South West African material is

described, and the distribution of the three species is briefly discussed.

ACKNOWLEDGEMENTS

I am grateful to the South African National Committee for Oceanographic

Research for a grant towards the expenses of one of the expeditions to South

APLETODON PELLEGRINI AND OTHER CLINGFISHES FROM S.W.A. 1393

A fe

AOKao4

} —Apletodon pellegrini A

| -Chorisochismus dentex

-Eckloniaichthys scylliorhiniceps

Fig. 5. Distribution of Apletodon pellegrini, Chorisochismus dentex and Eckloniaichthys scylliorhiniceps.

194 ANNALS OF THE SOUTH AFRICAN MUSEUM

West Africa and Angola. I am indebted to the South West African Administra-

tion, and in particular Mr. C. G. Coetzee, Director of the State Museum,

Windhoek, for making it possible for me to visit the northern coast of South

West Africa.

My grateful thanks are due to those who assisted in the collection of

material: Mr. M. J. Penrith, of the Oceanography Department, University of

Cape Town; Mr. C. G. Coetzee, Dr. P. G. Olivier, and J. Batista of the State

Museum, Windhoek; Mr. B. F. Kensley, of the South African Museum; Mr.

C. D. Berrisford, of the Council for Scientific and Industrial Research, Durban;

Mr. N. A. S. Reay.

I am indebted to the following individuals for the loan of material for

comparison: Mrs. M. M. Smith, of the J. L. B. Smith Institute of Ichthyology,

Rhodes University, Grahamstown; Dr. M. L. Bauchot, of the Muséum National

d’Histoire Naturelle, Paris; Dr. J. Nielsen, of the Universitets' Zoologiska

Museum, Copenhagen; Dr. P. J. Whitehead, of the British Museum (Natural

History).

My very grateful thanks are due to Mr. B. F. Kensley for the illustrations

used in figures 2, 3, and 4, and to Mr. M. J. Penrith for commenting on the

manuscript. |

The Trustees of the South African Museum are grateful to the Council

for Scientific and Industrial Research for a grant to publish this paper.

REFERENCES

BARNARD, K. H. 1925-27. A monograph of the marine fishes of South Africa. Ann. S. Afr. Mus.

21: 1-1065.

Buianc, M., Capenat, J. & Staucu, A. 1968. Contribution a l’étude de l’ichthyofaune de Vile

onae uate Bull. Inst. fond. Afr. noire (A) 30: 238-256.

Briccs, J. C. 1955. A monograph of the clingfishes (order Xenopterygii). Stanf. ichthyol. Bull.

6: 1-224.

(CHaABANAUD, P. 1925. Lepadogaster (Mirbelia) bimaculatus Penn., microcephalus Brook et Pellegrini,

nov. sp. (Pisces Gobiesocidae). Bull. Mus. Hist. nat., Paris 1925: 283-287.

Guntuer, A. C. L. G. 1861. Catalogue of the acanthopterygian fishes in the collection of the British

Museum. 3. London: Trustees of the British Museum.

Mourcocr, A. 1964. Contribution a la connaissance des gobiesocides (ordre des Xenopterygii)

de la mer Noire. Rev. roum. Biol. (Zool.) 9: 297-306.

Patras, P. S. 1769. Spicilegia zoologica .. . 7. Berolini..(Not seen.)

SmirH, J. L. B. 1943. Interesting fishes Be three genera new to South Apter with a note on

Mobula diabolus (Shaw). Trans. R. Soc. S. Afr. 30: 67-77.

SmiTH, J. L. B. 1949. The sea fishes of southern Africa. Cape Town: Central News Agency.

SmiTH, J. L. B. 1964. The clingfishes of the western Indian ocean and the Red Sea. ehayok

Bull. Rhodes Univ. 30: 581-596.

INSTRUCTIONS TO AUTHORS

Based on

CONFERENCE OF BIOLOGICAL EDITORS, COMMITTEE ON FORM AND STYLE. 1960.

Style manual for biological journals. Washington: American Institute of Biological Sciences.

MANUSCRIPT

To be typewritten, double spaced, with good margins, arranged in the following order:

(1) Heading, consisting of informative but brief title, name(s) of author(s), address(es) of

author(s), number of illustrations (plates, figures, enumerated maps and tables) in the article.

(2) Contents. (3) The main text, divided into principal divisions with major headings; sub-

headings to be used sparingly and enumeration of headings to be avoided. (4) Summary.

(5) Acknowledgements. (6) References, as below. (7) Key to lettering of figures. (8) Explana-

tion to plates.

ILLUSTRATIONS

To be reducible to 42 in. x 7 in. (7 in. including caption). A metric scale to appear with

all photographs.

REFERENCES

Harvard system (name and year) to be used: author’s name and year of publication given

in text; full references at the end of the article, arranged alphabetically by names, chronologi-

cally within each name, with suffixes a, 5, etc. to the year for more than one paper by the

same author in that year.

For books give title in italics, edition, volume number, place of publication, publisher.

For journal articles give title of article, title of journal in italics (abbreviated according to

the World list of scientific periodicals. 4th ed. London: Butterworths, 1963), series in

parentheses, volume number, part number (only if independently paged) in parentheses,

pagination.

Examples (note capitalization and punctuation)

BuLLoucH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FIscHER, P.-H. 1948. Données sur la résistance et de le vitalité des mollusques. 7. Conch., Paris

88: 100-140.

FiscHER, P.-H., Duvat, M. & Rarry, A. 1933. Etudes sur les échanges respiratoires des littorines.

Archs Zool. exp. gén. 74: 627-634.

Koun, A. J. 1960a. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee

region of Ceylon. Ann. Mag. nat. Hist. (13) 2: 309-320.

Koun, A. J. 19605. Spawning behaviour, egg masses and larval development in Conus from

the Indian Ocean. Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. Jn Schultze, L.

Koologische und anthropologische Ergebnisse einer Forschungsreise im westlichen und zentralen Siid-

Afrika. 4: 269-270. Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16: 269-270.

ZOOLOGICAL NOMENCLATURE

To be governed by the rulings of the latest International code of zoological nomenclature issued

by the International Trust for Zoological Nomenclature (particularly articles 22 and 51).

The Harvard system of reference to be used in the synonymy lists, with the full references

incorporated in the list at the end of the article, and not given in contracted form in the

synonymy list.

Example

Scalaria coronata Lamarck, 1816: pl. 451, figs 5 a, 6; Liste: 11. Turton, 1932: 80.

eee

pre

, ; I - pr EX Re Ki i¥), 3 (

ae | f hie weet i ae Ab

j MPI j Pht i, . ‘* iy Lian te at Wee Aris ae Py ea j a Ass,

Vii Dead ony Dy

ae TRG? BRS

eM,

4 Veit ng!

bay eae Ry, ‘ n

p F © ad 7 i Wile, UG -

rie: 5k ieee ey le -

MRM ci. |) 3) D> i RAR |

ery ;

op) Seo yale Te - aay Wingo ty art ithe

‘

ys rT oe

; aye WA ae +

S } it

y 2 4 ape have ] Es An

‘*~

‘

z 4

- f

4 7s

‘

;

O/,66

PEN ALS “OF THE SOUTH AFRICAN,MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 55 #£Band

July 1970 Julie

Pare forge Deck

THE DISTRIBUTION OF THE FISHES OF THE

PeMILY CLINIDAE IN. SOUTHERN AFRICA

MARY-LOUISE, PENRITH

Cape Town Kaapstad

a HSOW—~

SEP 11 1970

“i8RARIL2

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na beskikbaarheid

van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad

OUT OF PRINT/UIT DRUK

I, Ain 3> D> 7-8), Be, 5> te ls 5(2, 5> 7-9)>

6(1, tea Dele) Fa S)e 8, O(1-2)), 10(1),

11(1-2, 5, 7, tpl), 21, 24(2), 27, 31(1-3),, 38,

44(4).

Price of this part/Prys van hierdie deel

R1.35

Trustees of the South African Museum (© ‘Trustees van die Suid-Afrikaanse Museum

1970

Printed in South Africa by In Suid-Afrika gedruk deur

The Rustica Press, Pty., Ltd. Die Rustica-pers, Edms., Bpk.

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

THE DISTRIBUTION OF THE FISHES OF THE FAMILY

CLINIDAE. IN SOUTHERN AFRICA

Mary-Louise PENRITH

South African Museum, Cape Town

(With 5 figures)

[MS. received 20 October 1969]

CONTENTS

PAGE

Introduction . : : F ; : : : é to

Methods . : , ; ‘ ‘ : ; ’ ergo

Note on the stability of South African clinid populations 136

South African intertidal faunal provinces . : : sp BO

Distribution of the South African Clinidae. : ; EO)

Discussion fs : ; : : : s : . 145

Summary. : : ; : . 148

Acknowledgements é : : : : : . 149

References ; : : , : : . 149

INTRODUCTION

Thirty-five species of the family Clinidae are known from the coast of

Africa. Two of them belong to the subfamily Labrisominae and these are

restricted to tropical West Africa. The remaining 33 all belong to the tribe

Clinini of the subfamily Clininae (Hubbs, 1952; Penrith, 1969), and all are

endemic to southern Africa between the Kunene River on the west coast and

Inhambane in Mocambique. In spite of intensive collecting by the late Pro-

fessor J. L. B. Smith in tropical East Africa, no clinids have been recorded from

the east coast north of Inhambane.

The distribution of the dominant elements of the intertidal fauna and

flora of South Africa was described by the late Professor T. A. Stephenson and

his colleagues before the Second World War, and the main conclusions are

summarized in three papers (Stephenson, 1939, 1944, 1948). These conclusions

were not based on shore fishes, which are difficult to catch, but Stephenson

(1948) noted that the distribution of the Clinidae outlined by Smith (1945)

agreed closely with the distribution of other shore animals. Stephenson further

suggested that, since the clinids dealt with by Smith are all endemic, the

details of their distribution would be particularly interesting. This stimulated

135

Ann. S. Afr. Mus. 55 (3), 1970, 135-150, 5 figs.

guasTHSOWAN §=gep 03 WI"

INSTITUTION S

136 ANNALS OF THE SOUTH AFRICAN MUSEUM

the present study, and the data obtained have provided new evidence to test

Stephenson’s conclusions regarding the limits of the faunistic provinces around

southern Africa.

In broad outline it may be said that the clinids are the dominant group

of fishes that live permanently on the shores of the Cape Province and southern

South West Africa. Further north along the shores of South West Africa the

clinids are largely replaced by other families of shore fishes, and this is also

true on the coasts of Natal and Mocambique. Around the Cape, from Liideritz-

bucht in southern South West Africa, to the Transkei, where clinids predominate,

there are changes in species composition and abundance that are similar to the

changes in other intertidal animals.

METHODS

Many collecting trips to various parts of the coast between Mocamedes

in southern Angola and Durban have been carried out by the South African

Museum since 1963. An outline of the southern African coast with the localities

where collections were made is shown in figure 1. Where possible, localities

were visited more than once. The records obtained from these field trips, both

with regard to occurrence and abundance of species, form the basis of the

present work. In addition, records supported by identified specimens in the

collections of the South African Museum, the J. L. B. Smith Institute of Ich-

thyology (formerly Department of Ichthyology), Rhodes University, Grahams-

town, and the Zoology Department, University of Cape Town, have been used.

Finally, certain individuals have made collections in particular areas on behalf

of the South African Museum; these are acknowledged below.

NOTE ON THE STABILITY OF SOUTH AFRICAN CLINID POPULATIONS

Many fishes, even ones which are permanent inhabitants of the intertidal

region of the shore, are unsuitable subjects on which to base conclusions regard-

ing faunal distribution boundaries, owing to their mobility. Many intertidal

fishes show seasonal variation in distribution (Dr. V. G. Springer, personal com-

munication). However, the South African Clinidae appear to be an extremely

sedentary group of fishes, forming a highly stable intertidal population.

Collecting in all months of the year over a period of more than four years on

both east and west coasts of the Cape Peninsula produced no evidence for any

kind of seasonal variation in either the occurrence or the proportions of the

different species making up the clinid population.

_ There seems to be little reason for the Clinidae to show seasonal variation

in their occurrence on South African shores. Conditions of temperature, at

least, normally one of the strongest factors influencing seasonal variation, tend

to vary relatively little at any given locality on the South African coast. A

series of maximal and minimal monthly mean surface temperatures for localities

137

FISHES OF THE FAMILY CLINIDAE IN SOUTHERN AFRICA

‘QPCUL JIM SUOTIIIT[OO YOIYM 3Y SOTI[VIO], SUIMOYS VOTIFY UIIYINOS Jo suITIsvOrD *1 “BI

anbut WI

OSuasioy

UMOL adpe

Avg odyuopipg

‘y biag Dg DUI)H "Is

Adg §42qWo7

‘YY SIUOTIO

"WY U2QD

hog diyapuoy

YIOHON 140g

WYONG ZytsapN7

Apg s32uads

WH DYAIAS

hog S!AjDM

punwdoyoms

$8015 adD>

VM'S

Apg bsJoy'

——S,02

Wiad 44204)

Ols4 ado’

"4 auauny

sapauirso/y

VIOONV

138 ANNALS OF THE SOUTH AFRICAN MUSEUM

from Saldanha Bay to Durban given by Isaac (1937) show differences of rarely

more than 6:5°C between maxima and minima, and frequently less. From data

given by Marchand (1932), Buys (1957) and Division of Sea Fisheries annual

reports (1963, 1964) the same appears to be true of localities north-west of

Saldanha Bay. These fluctuations are smaller than would occur in the intertidal

zone from day to night, and would not be expected to affect intertidal animals.

Furthermore, the habits of the South African Clinidae make it unlikely

that populations would move about freely. Seasonal mobility is often related

to breeding. All the South African Clinidae bear their young alive, fertilization

being internal. Mature females of several common species (Clinus acuminatus,

C’. cottoides, C’. superciliosus, Blennioclinus brachycephalus and Pavoclinus mus) from

one or two localities in False Bay, showed eggs and embryos in several stages

of development in all or almost all months of the year. From this it appears that

there is no fixed breeding season. Preliminary observations in an aquarium

tank indicated that raised temperature produces increased sexual activity in

Clinus acuminatus at least, but the degree of solar warming of intertidal pools

during periods of low tide is probably the determining factor here. Clinidae

have been observed mating in a tidal pool in False Bay on warm sunny days

in winter, when general sea temperatures in the area were low.

Another factor restricting the mobility of Clinidae may be territoriality.

Gilchrist & Thompson (1911) observed territorial behaviour in Clinidae both

in the field and in an aquarium. A further indication that territoriality exists

is the fact that a pool which has been more or less cleared of fishes by rotenone

poisoning will remain unoccupied for periods of six to eight months, although

the effects of the poison disappear with the first high tide, and then occupation

will be by small, maturing fishes. It seems, therefore, that the South African

Clinidae may be considered as reliable as any animal in indicating boundaries

to faunal distribution.

SoutTH AFRICAN INTERTIDAL FAUNAL PROVINCES

On the basis of a series of detailed surveys of rocky intertidal shores at a

number of localities from Port Nolloth to Durban, Stephenson (1944, 1948)

recognized three faunal provinces in the South African intertidal zone: the west

coast, from the Orange River to Cape Point; the south coast, from Cape Point

to the region of Qolora/Port St. John’s; the east coast, from Port St. John’s to

northern Natal. The areas between Kommetijie, on the west coast of the Cape

Peninsula, and Cape Agulhas, and between Port Elizabeth and Port Edward,

showed a marked overlap between faunal provinces. Ekman (1953) followed

closely Stephenson’s conclusions regarding the South African faunal provinces,

placing the boundary between the south and south-west (‘Namaqua’) faunas

in the region of the Cape of Good Hope, and the southern boundary of the

tropical south-east African fauna south-west of Durban. The position of this

latter boundary is not clear, but Ekman suggested that Algoa Bay might

FISHES OF THE FAMILY CLINIDAE IN SOUTHERN AFRICA 139

constitute the southern boundary of distribution of a subtropical fauna extending

from there to the area of Durban. Ekman recognized the area from Algoa Bay

to Cape Agulhas/Cape Point as an independent zoogeographical province,

and followed Stephenson (1944, 1948) in treating it as warm-temperate,

although he pointed out that according to the temperature divisions he used it

was subtropical rather than temperate. On the west coast, Ekman placed the

northern boundary of the ‘Namaqua’ faunal province in the region of Cape

ityetoy

Knox (1960), discussing the biogeography of the southern oceans, gave

a division of South Africa into a cold-temperate “West African Province’,

extending from the Cape Peninsula to about 18°S (i.e. the region of Cape Frio),

and a warm-temperate “Cape Province’. Surveys made on the coast of South

West Africa, north of the Orange River, show that the west coast province

extends at least as far north as Liideritzbucht, but that further north the inter-

tidal fauna of rocky shores is of a different nature (Penrith & Kensley, in press).

The northern boundary of the west coast province lies somewhere between

Liideritzbucht and Walvis Bay.

The three faunal provinces proposed by Stephenson are closely related to

temperature conditions around the South African coast. Owing to the Benguela

upwelling system, the water of the west coast is cold, inshore surface tempera-

tures on the open coast seldom rising above 15°C between Cape Point and

Liideritzbucht (Marchand, 1932; Isaac, 1937). From Walvis Bay northwards

warmer inshore surface temperatures (16-18°C in summer) are encountered

(Stander, 1964). The south coast east of Cape Point is warmer than the west

coast, with monthly mean inshore temperatures of 15-20°C (Marchand, 1932;

Isaac, 1937; Stephenson, Stephenson & Du Toit, 1936). The east coast is

subtropical in character; Marchand (1932) found monthly mean temperatures

inshore varying from 18 to 24°C.

Day (1967), discussing the distribution of polychaete worms around

southern Africa, considered that the west and south coast faunal provinces are

not distinct if the regions below the intertidal zone are considered as well. This

is probably because offshore bottom temperatures are more uniform east and

west of Cape Point (Day, 1967). Day noted that many of the differences

between the west and east coast faunal provinces are of relative abundance

rather than of occurrence. Apart from such differences in relative abundance,

there are many similarities between the west and south coast faunas; each

shows a high degree of endemism, about 45% of the species occurring there

being endemic (Day, 1967). Numerous species are common to both coasts.

DISTRIBUTION OF THE SOUTH AFRICAN CLINIDAE

Of the 33 species of South African Clinidae, 30 belong to the subtribe

Clinidi, and the following discussion will be concerned with these species.

The other three species, belonging to the subtribe Xenopoclinidi, are specia-

ANNALS OF THE SOUTH AFRICAN MUSEUM

140

auDqwupyy|

“WOLUTUTOD 0} JUaseId AyIePNSa1 aie satdads YOTIYM JOAO SoSULI 9}LOIPUT SIU] Pedy], “tpruryy Jo saroods uvowyy YIMOG oy} Jo uOTNGINSIC *% “SI

@ snpunjoud y

a OAD J

eS ees

SijDJuawW Yy

—SIJUOJ DIOP}

wnsoony g

SUOIJIPUNJOU +5

I1Poom *>

rr slJysnuaa *>

SN4NDdy *>

snsoljiouadns - PE TE tt

SNYDIIYS “>

snysnqol ‘>

SI|DADU ‘>

—— siuuadi3}d| ‘>

<0 P01 3324 5

aduajay ‘>

SJS—— SIIDSJOP “>

S2Pi0}0) ‘5

a ——— OSE) SS)

SNYDISIJIIABIG “>

® ipsojsissaqg ‘D tapj{in6ub “>

a LO EE RE SOI :

a se

SEE eee | SNDUIWINID “>

snsosodiq ‘ds

o—_—_-@———. oj/218 ‘9

EEE SEES LS, ene = = = He ~~ -@ SnidydashyoDIG ‘g

i w vu QO wn 0 mm w =

re) Pad 9 ce} © AEA o 2 =< zo a ie re) ao < bod ples Sy

: ° ° = aan

& a) = n nas ~< a 3 0 3 7 rs Q. =a a wo

a ig c re > om fe uv an 3° = = r* Lian x Pa) Za

p= ] aR : eR =e

ees 2g ee 8 Seem 8 Sy 2 8 z <8 Se ge

o oO . om — e g, a oo, Qo fom 3 Oo o 2,

Cc 5 3 N oO ns y = = le = Re = 5

z a ® a Qa cr fe) fe) =]

° a°¢ s< wo — xr a. ~

a Zz 9 eR a @ > fe.

= —— td co << a

c am > S

rt)

-FISHES OF THE FAMILY CLINIDAE IN SOUTHERN AFRICA 141

lized sand-burrowers which have been found intertidally only in the Lambert’s

Bay area, and I consider their distribution too imperfectly known to be dis-

cussed further here.

Nine of the 30 species of Clinidi occur both inter- and infratidally, larger

specimens in general being taken in deeper water; 14 have been taken more or

less exclusively intertidally, and seven occur mainly infratidally, although

juvenile and even adult specimens may occur intertidally from time to time in

most cases; Clinus woodi occurs intertidally in the northern part of its range, but

is infratidal further south (Mrs. M. M. Smith, personal communication). The

species in these three categories are listed for reference in table 1.

The known ranges of distribution of the 30 South African species of Clinidi

are shown in figure 2. Most of the species occur between Liideritzbucht and

the Kei River mouth; this is shown in a graph illustrating the number of

species recorded from different localities along the coast (fig. 3). Two of the

species, Clinus berrisfordi and Pavoclinus profundus, are known only from their

single type localities. Of the remaining 28, 14 are common, fairly common, or

at least regularly present over part or the whole of their known ranges of distri-

NO. OF SPECIES

anwhuaOrtoos

Cape Frio

Torra Bay

Swakopmund

Spencer Bay

Port Nolloth

Lamberts Bay

Hermanus

East London:

Durban

Inharmbane

Lourenco Marques

Fig. 3. Number of species of Clinidi present along each stretch of coast of approximately 100

miles, from Cape Frio to Inhambane.

142 ANNALS OF THE SOUTH AFRICAN MUSEUM

bution (indicated by thickened lines in figure 2); three are, at most, regularly

present in particular and limited habitats, and 11 are apparently rare (table 2).

It may be noted that all the mainly infratidal species fall in the categories of

less frequent occurrence, and this picture might change if sampling of the

infratidal region down to about 30 metres depth were more effective.

TABLE I

Inter- & Infratidal Mainly intertidal Mainly infratidal

Clinus acuminatus

Clinus agilis

Clinus berrisfordi

Clinus brevicristatus

Clinus cottoides

Clinus dorsalis

Clinus helenae

Clinus heterodon

Clinus latipennis

Clinus navalis

Blennioclinus brachycephalus

Blennioclinus stella

Pavoclinus mus

Pawoclinus pavo

Clinus anguillaris

Clinus capensis

Clinus striatus

Clinus superciliosus

Clinus venustris

Pavoclinus fucorum

Pavoclinus graminis

Pavoclinus laurentii

Pavwoclinus litorafontis

Clinoporus biporosus

Clinus robustus

Clinus taurus

Clinus woodi

Gynutoclinus rotundifrons

Pavoclinus mentalis

Pavoclinus profundus

TABLE 2

Frequency of occurrence of South African Clinidi

Single record only

Clinus berrisfordi

Pavoclinus profundus

Regularly present

in particular

habitats of rare

occurrence only

Rare/Little known

Clinoporus biporosus Blennioclinus stella

Clinus brevicristatus

Pavoclinus litorafontis

Clinus helenae

Clinus latipennis

Clinus navalis

Clinus robustus

Clinus striatus

Clinus taurus

* Clinus wood

Gynutoclinus rotundifrons

Pavoclinus fucorum

Pavoclinus mentalis

Common, fairly

common or regularly

present over part

or whole range

Blennioclinus brachy-

cephalus

Clinus acuminatus

Clinus agilis

Clinus anguillaris

Clinus capensis

Clinus cottoides

Clinus dorsalis

Clinus heterodon

Clinus superciliosus

Clinus venustris

Pavoclinus graminis

Pavoclinus laurentii

Pavoclinus mus

Pawoclinus pavo

*This species is said to occur in fair numbers (Mrs. M. M. Smith, personal communi-

cation) but is not easily accessible for collecting over some of the range and the range

over which it is common is consequently not known.

FISHES OF THE FAMILY CLINIDAE IN SOUTHERN AFRICA 14%

The extreme limits of distributional ranges of species are seldom fixed and

reliable; most species dwindle in numbers towards the ends of their ranges,

and as it is impossible to sample an area completely, the most easterly and

westerly members of a species may never be caught. It is not easy, however,

to be mistaken about whether a species is common in a given area or not. In

attempting to pinpoint areas where faunal changes occur, I have in the first

instance considered the ranges over which 14 species were regularly present

(i.e. an invariable constituent of the fauna) to common intertidally. Figure 4

shows the number of eastern and western limits of these ranges occurring on

each approximately 1oo-mile stretch of coastline between Cape Frio and

Inhambane.

From figure 4 it can be seen that 24 of the 28 range limits occur on the

following stretches of coast: (1) 100 miles south from Spencer Bay; (2) 100 miles

10)

Bee a

Or 25

jag

a ——ty

2 I Vie is > - 2 5

agg | ) Mm 2 ‘e) S = s

Eads ike = = 419) te ie

Oe or IS O 7 @ a c

ee ee = Soe 5 5 2 E

4 es eS ~ oS ; ‘a is

Os ht alae ie a 5 2 ~ °

coe & ee 2 o

ts 8

ie}

—J

Fig. 4. Number of limits of ranges over which 14 species of Clinidi occur as regularly

present to common along stretches of coast of approximately 100 miles between Cape

Frio and Inhambane.

west from Hermanus; (3) 100-mile stretch in the vicinity of East London/Kei

River mouth/Bashee River mouth. One of the remaining limits occurs immedi-

ately adjacent to (2), and the other three lie west and east of Spencer Bay or

East London/Bashee River respectively.

It is evident that major changes in the nature of the clinid fauna occur in

two of the areas (Cape Point/Cape Agulhas and East London/Bashee River)

suggested as faunal province boundaries by Stephenson, and that a third major

change occurs in the vicinity of Liideritzbucht. The clinid faunas of the different

intertidal provinces can now be considered.

1. The coast north of Liideritzbucht

Only one species of clinid, Clinus superciliosus, occurs at all commonly north

of Liideritzbucht. It was fairly common at Sylvia Hill, Walvis Bay and Swakop-

144 ANNALS OF THE SOUTH AFRICAN MUSEUM

mund. Small numbers were collected at Torra Bay, Mowe Point and Rocky

Point, where it was rare, two species of blennies (Blennius cornutus and Blennius

vandervekeni), and the clingfish Chorisochismus dentex being the common fishes

of the intertidal zone of northern South West Africa. Single specimens of Clinus

heterodon have been collected at Swakopmund and Sylvia Hill.

2. The west coast (Liideritzbucht to Cape Point)

Apart from a goby, Coryphopterus nudiceps, and a clingfish, Chorisochismus

dentex, both of which are fairly common, all the common intertidal fishes of

the west coast are Clinidae. The most abundant clinid of the intertidal

zone in this area is Clinus agilis. Clinus superciliosus is very common at

all levels of the intertidal zone and below. Clinus acuminatus is common and

Clinus dorsalis fairly common in pools at the top of the intertidal zone, juveniles

of the latter species occurring at lower levels as well. Clinus anguillaris and Clinus

venustris are regularly present in small numbers in pools at the bottom of the

intertidal zone, the former species occurring infratidally as well. Occasional

specimens of Pavoclinus pavo and Gynutoclinus rotundifrons have been taken

amongst algae.

From the region of Lambert’s Bay southwards to Cape Point, the typical

west coast clinid fauna described above is modified by the addition of typically

south coast forms, notably Clinus cottoides, which is common from Lambert’s

Bay southwards but has not been taken north of the Olifants River mouth.

3. The south coast (Cape Point to Kec River mouth)

The south coast has a larger number of common intertidal fish species than

the west coast, and again most of these are Clinidae. The gobies Coryphopterus

caffer and Ctenogobius saldanha and the clingfish Chortsochismus dentex are also

common, and from Knysna eastwards two blennies, Blennius cornutus and

Blennius steindachnert, became fairly common.

Clinus cottoides is the most abundant intertidal species from Cape Point

eastwards to East London. Clinus superciliosus is very common at all levels of the

shore at least as far east as Port Elizabeth, decreasing in numbers in the East

London area. Clinus dorsalis is fairly common at the top of the shore. Clinus

anguillaris is regularly present in small numbers, and the closely related species

Clinus striatus is occasionally found from Saldanha Bay eastwards. Clinus capensis

is fairly common in pools at all levels of the shore, the adults being mainly

infratidal. Blennioclinus brachycephalus is common in pools at the lowest levels of

the shore. In weed-beds, particularly of the green alga Caulerpa filiformis, and

in algal clumps in rock pools, the species Pavoclinus mus, Pavoclinus graminis, and

Pavoclinus pavo are common, and Pavoclinus fucorum occurs more rarely.

Clinus agilis and Clinus heterodon decrease sharply in numbers east of Cape

Point. The former occurs very rarely on the south coast as far as Port Alfred,

and fair numbers have been found in the Knysna lagoon (Penrith, 1969).

Clinus heterodon has not been recorded east of False Bay, where it is extremely

FISHES OF THE FAMILY CLINIDAE IN SOUTHERN AFRICA 145

rare. Clinus acuminatus is fairly common in False Bay, but has not been recorded

east of Hermanus, where it is rare. Clinus venustris is rare east of Cape Point, and

is usually taken infratidally. Clinus robustus and Clinus taurus, both large, mainly

infratidal species, occur on the south coast.

A. The east coast, north-east of the Kei River mouth, to Inhambane

As on the coast north of Walvis Bay, Clinidae do not form a major con-

stituent of the intertidal fish fauna. Pavoclinus laurenti is fairly common as far

north as Inhaca Island, Mocambique (Mrs. M. M. Smith, personal com-

munication) and Pavoclinus graminis as far as Port St. John’s. Blennioclinus stella

was taken in fair numbers from a single locality in the vicinity of Durban,

although it is not generally common. Most of the other species apparently

disappear before Durban is reached. The common intertidal fishes of the east

coast are blennies (mainly Blennius steindachneri, Omobranchus banditus, Hal-

mablennius striatomaculatus and Croaltus bifilum) and gobies (particularly Bathy-

gobius fuscus).

5. Lhe coast from Inhambane northwards

No Clinidae have been recorded north of Inhambane.

Discussion

As all the South African clinid species are endemic, it is not surprising

that the majority of them occur in the west and south coast intertidal faunal

provinces. Day (1967) pointed out that, as far as polychaete worms were

concerned, the Cape/South West African province (i.e. the west and south

coasts) is dominated by endemics. The Natal (east) coast has some endemics,

but also a considerable number of species from other sources; to the north-west

and north-east of these provinces are faunal provinces consisting mainly of West

African and Indo-Pacific tropical species respectively. Almost all the intertidal

fishes of the west and south coasts are endemic to the South African region,

while from the Bashee River north-eastwards and from Swakopmund north-

westwards the percentage of endemics is lower.

Within the South African region, the clinid faunas of the west, south, and

east coasts show differences in species composition and in the abundance of

species which occur in more than one intertidal faunal province.

From figure 2 it can be seen that the general pattern of distribution of

South African Clinidi, as far as it is known, corresponds with the intertidal

faunal provinces mentioned above. The 28 species of Clinidi known from more

than a single locality fall, at the present state of knowledge of their distribution,

into the following categories: (1) species characteristic of the south coast;

(2) species characteristic of the west coast; (3) species characteristic of the

east coast; (4) species characteristic of both west and south coasts; (5) species

characteristic of both south and east coasts; (6) species apparently restricted to

the south-west coast overlap; (7) species apparently restricted to the south-east

146

ANNALS OF THE SOUTH AFRICAN MUSEUM

coast overlap. The 28 species are arranged in these categories in table 3 and

figure 5.

TABLE 3

Distributional categories of South African Clinidi.

W. & S. S. & E. W/S E/S

West coast South coast East coast coasts coasts overlap overlap

I 2 3 4 5 6 7

Clinus Clinus Blennioclinus Clinus Pavoclinus Clinoporus Clinus

acuminatus capensis stella anguillaris graminis biporosus helenae

Clinus Clinus Pavoclinus Clinus Clinus Clinus

agilis cottoides laurentir dorsalis navalis brevicristatus

Clinus Blennio- Pavoclinus Clinus Clinus

heterodon clinus mentalis superciliosus latipennis

brachyce-

phalus

Clinus Pavoclinus Clinus Pavoclinus

venustris Sucorum woodi litorafontis

Gynuto- Pavoclinus

clinus mus

rotundifrons Pavoclinus

pavo

Clinus

robustus

Clinus

striatus

Clinus

taurus

| ——_

| £ oO x E 2

| z x0} £ (0), ey Se 6 sey EE ee 5

| 3 2 i) es £ we ens 2 |

E is 2 aT Se Ts : 2 9

a 2 z te 2 mt <a ior = 3 lee

=x i) 2 9 » w ~ x B =! ¥

: 3 5 §o ¢ = Sone Bie E wig

o a & a ty © x er ence oie A hes

Le es eaters

Fig. 5. Graphic representation of the various components of the South African Clinidi population.

I—west coast; 2—south coast; 3—east coast; 4—west and south coasts; 5—south and east

coasts; 6—west/south coast overlap; 7—east/south coast overlap. (See table 3.)

FISHES OF THE FAMILY CLINIDAE IN SOUTHERN AFRICA 147

All the species belonging to categories 6 and 7 are rare species, and these

categories may be apparent rather than real.

Four species, Clinus agilis, Clinus venustris, Blennioclinus brachycephalus and

Pavoclinus pavo, have been placed among the species characteristic of a single

intertidal faunal province only, although the extremes of their recorded ranges

of distribution encompass both west and south coast provinces. However, the

change in incidence of these species in the region of Cape Point is so marked

that they are here taken to be characteristic of only one intertidal faunal pro-

vince. Clinus agilis and Clinus venustris were collected at all localities between

Liideritzbucht and Cape Point on the west coast but are extremely rare east of

Cape Point. Blennioclinus brachycephalus has been recorded reliably from the

west coast only at Liideritzbucht, where small numbers were found in sheltered

embayed areas. Pavoclinus pavo was collected at a few west coast localities, but

is common from False Bay eastwards.

The fact that the distribution of the South African Clinidi corresponds

closely with the South African intertidal faunal provinces, which are largely

determined by temperature factors (Stephenson, 1948), suggests that South

African clinid distribution is dependent on temperature. The effect of tem-

perature on the distribution of a species may be twofold. Each species has a

wider or narrower temperature range at which it can successfully exist, and its

spread to areas outside this temperature range will be prevented by the direct

effect of temperature. Secondly, other organisms which make a habitat suitable

for a species may have a more limited temperature range, and their absence

may prevent that species from spreading to areas not unsuitable from a point

of view of temperature itself. Alternatively, a species may be prevented from

spreading towards the ends of its possible temperature range by the intrusion

of competitive species with overlapping temperature ranges.

Too little is known of the ecology of the South African Clinidi to estimate

the importance of direct and indirect influence of temperature, or of other

factors, in restricting their distribution. There are certain indications that at

least both temperature effects probably operate.

Most of the clinid species do not appear to be strongly specific to a parti-

cular type of habitat. However, in spite of the fact that intertidal pools are not

subject to the same conditions as the intertidal slopes that are exposed for

varying lengths of time during periods of low tide, the Clinidi show vertical

zonation comparable with that shown by other intertidal organisms (Stephen-

son, 1944, 1948). It has been pointed out by Day (1967) that the west and south

coast faunal provinces are more sharply divided intertidally than infratidally,

and that the faunas at least below 100 m are similar, as might be expected

since at this depth temperatures east and west of Cape Point are uniform. In

general, of the better known species, the more strictly intertidal ones (Clinus

acuminatus, Clinus agilis, Clinus cottoides, Clinus heterodon, Blennioclinus brachyceph-

alus, Pavoclinus mus, Pavoclinus pavo) are more restricted in at least their ranges

of optimal occurrence than most of the species which occur infratidally as well

148 ANNALS OF THE SOUTH AFRICAN MUSEUM

(Clinus anguillaris, Clinus superciliosus, Pavoclinus graminis). Clinus dorsalis is widely

distributed and occupies a similar niche at the top of the intertidal zone to

Clinus acuminatus, but, unlike Clinus acuminatus, occurs at all levels of the shore in

the juvenile stages. This may protect the young from the extremes of solar

heating to which the adults are subjected during periods of low tide. Clinus

venustris shows a change in zonation east of Cape Point which is presumably

the direct result of temperature, as it becomes infratidal as higher temperatures

are encountered. Clinus capensis, with a similar habitat to Clinus anguillaris, is less

widely distributed; other largely infratidal species, Clinus robustus, Clinus striatus,

Clinus taurus, and Clinus woodi, also appear to have restricted distributions.

Unfortunately, the greatest depths at which these species occur are not known.

Clinus superciliosus has been taken from depths of 60 metres.

A possible instance of dependence on another organism is shown by

Clinus cottoides. This species is abundant in midtidal pools, in the zone charac-

terized on the south coast by a dense belt of barnacles (Stephenson, 1948).

This balanoid zone is scarcely developed on the west coast, especially north

of Lambert’s Bay (Bright, 1938; Stephenson, Stephenson & Day, 1940).

Barnacle legs form an important constituent of the diet of Clinus cottoides, and

it is possible that this species avoids areas lacking barnacles.

Gynutoclinus rotundifrons and the species of the genus Pavoclinus are specia-

lized for living in algae. None of them are restricted to a particular alga, with

the possible exception of Gynutoclinus rotundifrons, which has so far been taken

only in kelp. However, of the species occurring on the south coast, Pavoclinus

pavo and Pavoclinus graminis show a wider preference for beds of Caulerpa filuformis,

which does not occur west of Cape Point (Simons, zn Day, 1969), and is fairly

characteristic of the south coast. P. mus has not been observed in infratidal

Caulerpa beds at a depth of approximately 10 metres (M. J. Penrith, personal

communication) and is probably intertidal.

Too little is known at present about the biology of South African intertidal

fishes to estimate the effects of interspecific competition. Towards the western

and eastern limits of the South African region the clinid fauna is replaced by

other mainly tropical groups, and in these regions competition with other

groups may occur.

SUMMARY

The distribution of the South African fishes belonging to the subtribe

Clinidi (Pisces: Clinidae) is discussed. All the species are endemic to the South

African region (Kunene River to Inhambane), but within this region the dis-

tribution of the species is not uniform. It is, as far as can be determined at

present, related to water temperatures, and agrees with the distribution patterns

of other groups of intertidal animals. Factors such as habitat, food preferences,

and competition might influence the distribution of individual species, the

influence of temperature in these cases being indirect.

FISHES OF THE FAMILY CLINIDAE IN SOUTHERN AFRICA 149

ACKNOWLEDGEMENTS

I should like to acknowledge the financial assistance received from the

Council for Scientific and Industrial Research and the South African National

Committee for Oceanographic Research towards field trips undertaken in

connection with this work. I am grateful to the South West African Administra-

tion and in particular Mr. C. G. Coetzee, Director of the State Museum, Wind-

hoek, for making expeditions to the northern coast of South West Africa possible.

I am grateful to Professor J. H. Day, of the Zoology Department, Univer-

sity of Cape Town, and the late Professor J. L. B. Smith and Mrs. M. M. Smith

of the J. L. B. Smith Institute of Ichthyology, Rhodes University, Grahamstown,

for making their records available to me and for many helpful discussions; in

particular to Professor J. H. Day for critical reading of the manuscript. I am

also grateful to Dr. N. A. H. Millard of the Zoology Department, University

of Cape Town, for guidance during the preparation of this work for a thesis.

I am grateful to the following individuals who made collections on my behalf:

Mr. A. de Freitas, of the Marine Biological Institute, Lourenco Marques;

Mr. C. D. Berrisford, of the National Institute for Water Research, Durban;

Miss R. M. Tietz, of the Albany Museum, Grahamstown; Mr. C. G. Coetzee,

of the State Museum, Windhoek. I should like to record my gratitude to the

numerous individuals who have assisted in collecting material and donated

interesting specimens.

I am indebted to my husband, Mr. M. J. Penrith, who collected most of

the material and gave much helpful advice and information.

Much of this work formed part of a thesis for the degree of Doctor of Philo-

sophy at the University of Cape Town (December 1965).

The Trustees of the South African Museum are grateful to the Council

for Scientific and Industrial Research for a grant to publish this paper.

REFERENCES

Bricut, K. M. F. 1938. The South African intertidal zone and its relation to ocean currents.

III. An area on the northern part of the west coast. Trans. R. Soc. S. Afr. 26: 67-88.

Buys, M. E. L. 1957. Temperature variations in the upper 50 metres in the St. Helena Bay area,

September 1950—August 1954. Investl Rep. Div. Sea Fish. Un. S. Afr. 2°73 1-114.

Day, J. H. 1967. A monograph on the Polychaeta of southern Africa. London: British Museum (Natural

History).

Day, J. H. 1969. A guide to marine life on South African shores. Cape Town: A. A. Balkema, for the

University of Cape Town.

EKMAN, 8S. 1953. <oogeography of the sea. London: Sidgwick & Jackson.

Gitcueist, J. D. F. & THompson, W. W. 1911. The Cape klipfishes of South Africa. S. Afr. 7.

Sct. 73 214-224.

Husss, C. 1952. A contribution to the classification of the blennioid fishes of the family Clinidae,

with a partial revision of the eastern Pacific forms. Stanf. ichthyol. Bull. 4: 42-165.

Isaac, W. E. 1937. South African coastal waters in relation to ocean currents. Geogr. Rev. 27:

651-664.

Knox, G. A. 1960. Littoral ecology and biogeography of the southern oceans. Proc. R. Soc.

(B) 152: 577-624.

150 ANNALS OF THE SOUTH AFRICAN MUSEUM

MarcHAND, J. M. 1932. Hydrographic investigations during 1930. Rep. Fish. mar. biol. Surv. Un.

S. Afr. 8: 1-30.

PenritH, M.-L. 1969. The systematics of the fishes of the family Clinidae in South Africa. Ann.

S. Afr. Mus. 55: 1-121.

PenritH, M.-L. & Kensey, B. F. (In press.) The constitution of the fauna of rocky intertidal

shores of South West Africa. Part 1. Liideritzbucht. Cimbebasia (A).

PenritH, M.-L. & Kenstey, B. F. (In press.) The constitution of the fauna of rocky intertidal

shores of South West Africa. Part 2. Rocky Point. Cimbebasia (A).

Situ, J. L. B. 1945. The fishes of the family Clinidae in South Africa. Ann. Mag. nat. Hist.

(11) 12: 535-546.

Soutu Arrica. Division of Sea Fisheries. 1963. Annual report for the period 1st April 1959 to

31st March 1960. Rep. Div. Sea Fish. 8. Afr. 312 1-242.

SoutH AFrica. Division of Sea Fisheries. 1964. Annual report for the period 1st April 1960 to

31st March 1961. Rep. Div. Sea Fish S. Afr. 32: 1-267.

STANDER, G. H. 1964. The pilchard of South West Africa (Sardinops ocellata). The Benguela

current off South West Africa. Investl Rep. mar. Res. Lab. S.W. Afr. 12: 1-43.

STEPHENSON, T. A. 1939. The constitution of the intertidal fauna and flora of South Africa.

Part I. 7. Linn. Soc. (Zool.) 40: 487-536.

STEPHENSON, T. A. 1944. The constitution of the intertidal fauna and flora of South Africa.

Part II. Ann. Natal Mus. 10: 261-357.

STEPHENSON, T. A. 1948. The constitution of the intertidal fauna and flora of South Africa.

Part III. Ann. Natal Mus. 11: 207-324.

STEPHENSON, T. A., STEPHENSON, A. & Day, J. H. 1940. The South African intertidal zone and

its relation to ocean currents. VIII. Lambert’s Bay and the west coast. Ann. Natal Mus. 9:

345-380.

STEPHENSON, T. A., STEPHENSON, A., & Du Tort, C. A. 1936. The South African intertidal zone

and its relation to ocean currents. I. A temperate Indian ocean shore. Trans. R. Soc. S.

Afr. 24% 34.1-382.

INST RUCEIONS FO AUTHORS

Based on

CONFERENCE OF BIOLOGICAL EDITORS, COMMITTEE ON FORM AND STYLE. 1960.

Style manual for, biological journals. Washington: American Institute of Biological Sciences.

MANUSCRIPT

To be typewritten, double spaced, with good margins, arranged in the following order:

(1) Heading, consisting of informative but brief title, name(s) of author(s), address(es) of

author(s), number of illustrations (plates, figures, enumerated maps and tables) in the article.

(2) Contents. (3) The main text, divided into principal divisions with major headings; sub-

headings to be used sparingly and enumeration of headings to be avoided. (4) Summary.

(5) Acknowledgements. (6) References, as below. (7) Key to lettering of figure. (8) Explana-

tion to plates.

ILLUSTRATIONS

To be reducible to 4? in. X 7 in. (74 in. including caption). A metric scale to appear with

all photographs.

REFERENCES

Harvard system (name and year) to be used: author’s name and year of publication given

in text; full references at the end of the article, arranged alphabetically by names, chronologi-

cally within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year.

For books give title in italics, edition, volume number, place of publication, publisher.

For journal articles give title of article, title of journal in italics (abbreviated according to the

World list of scientific periodicals. 4th ed. London: Butterworths, 1963), series in parentheses,

volume number, part number (only if independently paged) in parentheses, pagination.

Examples (note capitalization and punctuation)

BULLOUGH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

Fiscuer, P.-H. 1948. Données sur la résistance et de le vitalité des mollusques. 7. Conch., Parsi

88: 100-140.

FiscHer, P.-H., DuvaL, M. & Rarry, A. 1933. Etudes sur les échanges respiratoires des littorines.

Archs Zool. exp. gén. 74: 627-634.

Konn, A. J. 19604. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee region

of Ceylon. Ann. Mag. nat. Hist. (13) 2: 309-320.

Konan, A. J. 19606. Spawning behaviour, egg masses and larval development in Conus from the

Indian Ocean. Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. Jn Schultze, L.

Koologische und anthropologische Ergebnisse einer Forschungsreise im westlichen und zentralen Stid-

Afrika. 4: 269-270. Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16: 269-270.

ZOOLOGICAL NOMENCLATURE

To be governed by the rulings of the latest International code of zoological nomenclature issued

by the International Trust for Zoological Nomenclature (particularly articles 22 and 51).

The Harvard system of reference to be used in the synonymy lists, with the full references

incorporated in the list at the end of the article, and not given in contracted form in the synonymy

list.

Example

Scalaria coronata Lamarck, 1816: pl. 451, figs 5 a, b; Liste: 11. Turton, 1932: 80.

sa Tete i

if i

- |

“ip | seh ft

Bh | i

| | '

H.|

polly , en i Rear 4 ‘ hey "

‘ iPithy juin cas twa) L A AM,

‘i “Et "

j - nf eh ‘ ; aly q On a . ry cei radi

; eae 4 AH? i Ore k may a Mi

| ! eee eS ee ke caihgapaneet hehe a: aay): _ Siobwaiias

{' | We oe #1, favi i : ith y DADL ote Pavive Pek’? f 54 ri ryokan ro

i | il.) Bi Rai Bay Ty rah a teal ede aes itu) a aa,

1 ae . Chey! TSP! Aes i Bi 2) Lt Lae hel, blaeat) hh ig Yi riheeyy a:

| ¢ oe { PAA Ad ms (Hed sani Ty i AY ha a fi

f ty v jue

wr rey Ah TO! 7 eee ae eave aie a? vy yy

; \ V's

| ;

} i Hi }

| \ ' y t bi ay ' ’ a9 : The

1 I . % ¢ i ' f ‘

: : ea i rant Sp ig HK,

AAS. one

| | inp : !

: ' ; Lis ws MY oerrisy Vw ah

| ‘ t i ‘ if ; yiz

}

| ° i ‘ PPS ak rial

| A | [ J

ge) ~ ‘ea +? ht! AW

~*y

| ¥ ‘ (

tr) )

| -

} | 4 eye)

Bet}

1, | n

- A PPG Vey

\\| ;

vi" \’ (Wea

| cqehtigeh) | 1 si RE

it ‘ee sus

ia - Ph AL ’ i erg Cr ePey

mengaltk, Ai Nd

t a

_ ba ar nie eee ight a

‘ Yee ra Pecvaeit

: a een’ Nal Wen Bae ee Pith

Ki ¥

’

on 1s ae:

rc a oe > ti

: a1?

y : ee : fat jit TF

| ‘ oun) hs eae Mba) We

' ‘ ' ea : ee oars ie ae |

1 ‘ Re ¢ 4

oh 2 hosed Wiles sn giant) mati | By

Lame Se 3 Mss Ae, r

F 5

hay Wane na AOE

} 4 ,

| 4

% Pe ff

Be? / GS

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 55 °#Band

October 1970 Oktober

Part) «i -2- Deel

“ie

q li ay e

i Ww

AN INVESTIGATION OF THE RAJIDAE —

OF THE WEST AND SOUTH COASTS

OF SOUTHERN AFRICA

P. Ay HULLEY

HSON

a Ni4y

FEB 292 1971

{iBRARIE?

Cape ‘Town Kaapstad

The ANNALS OF THE SOUTH AFRICAN MUSEUM

are issued in parts at irregular intervals as material

becomes available

Obtainable from the South African Museum, P.O. Box 61, Cape Town

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na beskikbaarheid

van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad

OUT OF PRINT/UIT DRUK

I, 2(1, 3> 5» 7-8), 3(1-2, 5» t—pu.), 5(2, 5» 7-9);

6(1, tt.) HOS) hs 8, 9(1-2), 10(1),

I1(I-2, 5, 7, t.-p.i.), 21, 24(2), 27, 31(1-3), 33, 38,

44(4).

Price of this part/Prys van hierdie deel

R5.00

Trustees of the South African Museum © _ Trustees van die Suid-Afrikaanse Museum

1970

Printed in South Africa by In Suid-Afrika gedruk deur

The Rustica Press, Pty., Ltd. Die Rustica-pers, Edms., Bpk.

Court Road, Wynberg, Cape Courtweg, Wynberg, Kaap

AN INVESTIGATION OF THE RAJIDAE OF THE

WEST AND SOUTH COASTS OF SOUTHERN AFRICA

PAS PULLEY

South African Museum, Cape Town

(With 13 plates, 21 figures and 1g tables)

[MS. received 20 November 1969]

CONTENTS

PAGE

Introduction . : : é 5 : BSR

Method . : ‘ : : : 3 Bibb see

syseemmatic discussion’) fh. So. 153

Summary . : OA Ce Oe a eR? © Ge

Acknowledgements. Sh ee 2 De

References. 5 : : : : | (OES

INTRODUCTION

This paper represents the results of a three-year survey of the Rajidae of

the west and south coasts of southern Africa, and is intended as a paper

accompanying that of Wallace (1967) on the batoid fishes of the east coast of

southern Africa. All known species of southern African Rajidae have been

included in the given key, but, for descriptions of species which are exclusively

east coast in distribution, Wallace’s (1967) publication should be consulted.

The area originally covered by the survey extends from about Swakop-

mund on the west coast of South West Africa to Port Elizabeth on the south-

east coast of the Republic of South Africa. The majority of the specimens

have been obtained by commercial trawlers fishing on the continental shelf

between 250 and goo metres with otter trawls, while the rest have been taken

at shallower depths by research vessels. The area of survey has been extended

northwards along the west coast to 09°40’S by the examination of material

taken by R. V. Walther Herwig in 1967. The results of this cruise are given in

table 1.

The classification of South African skates is more or less tentative, for

members of this difficult family show great variability in their morphological

characters, which vary with age and/or sex in the individuals within a species.

Furthermore, some of the type specimens of South African rajids, obtained by

the Government Marine Survey at the beginning of this century, have been

lost, so that it is difficult to verify original descriptions, especially where these

are based on juvenile specimens. It is not surprising, therefore, that skate

taxonomy has received casual treatment from South African systematists.

Recently, however, Wallace (1967) has revised the east coast species, while

151

Ann. S. Afr. Mus. 55(4), 1970: 151-220, 13 pls, 21 figs, 19 tables.

152 ANNALS OF THE SOUTH AFRICAN MUSEUM

Hulley (1966, 1969) has shown that several species are identical with European

species from the corresponding latitudinal belt in the northern hemisphere,

and represent cases of bipolar distribution (Ekman, 1953).

Leigh-Sharpe (1920-6) has suggested the formation of pseudo-genera

within the Rajidae based on clasper structure, but detailed descriptions of the

anatomy of these organs in South African skates and their taxonomic significance

above the species level will be published in the near future.

Although precaudal vertebral counts appear to be of greater value for

separation of the family above the species level (Ishiyama, 1952), the external

and internal structures of the clasper glans are species specific without exception

(Ishiyama, 1958, 1967; Hubbs & Ishiyama, 1968) and may be used for com-

parison of geographically separated species (Hulley, 1966).

Zoogeographical and depth distributional patterns for the species are to

be dealt with in a later paper; in the interim period, however, some idea of

this may be obtained from the lists of material.

TABLE I. Rajidae taken by R.V. Walther Herwig off the west coast of southern Africa in 1967.

Station No. Position Depth Species No. of

WH |67 m Specimens

6 09°40’S, 12°58’E 100 . miraletus Linnaeus

10 10°28’S, 13°02’E 700 . straeleni Poll

12 10°41’S, 13°29’E 100 miraletus Linnaeus

14 11°04’S, 13°30’E 440 . miraletus Linnaeus

15 11°30’S, 13°25’E 110 miraletus Linnaeus

17 16°14’S, 11°33’E 100-120 miraletus Linnaeus

18 18°30’S, 11°27’E 310 . Straeleni Poll

miraletus Linnaeus

19 18°35/S, 11°25’E 400 . straeleni Poll

20 18°45/S, 11°20’E 500 . leopardus Von Bonde & Swart

confundens n.sp.

doutret Cadenat

32 20°25’S, 12°02’E 500 leopardus Von Bonde & Swart

confundens n.sp.

doutrer Cadenat

33 22°03’S, 13°12’E 200 leopardus Von Bonde & Swart

confundens n.sp.

straelent Poll

leopardus Von Bonde & Swart

confundens n.sp.

. confundens n.sp.

. caudaspinosa Von Bonde & Swart

clavata Linnaeus

. smith (Miiller & Henle)

confundens n.sp.

clavata Linnaeus

parcomaculata (Von Bonde & Swart)

dissimilis n.sp.

ravidula n.sp.

37 22°15’S, 12°46’E 500

43 23°00'S, 13°02’E 400

is 30°02’S, 14°39’E 510

89 32°06’S, 16°22’E 400

100 34°12’S, 17°34’E 630

126 27°13'S, 14°31’E 320

161 26°25'S, 14°18’E 300

178 32°39'S, 17°25’E 245

194. 33°47'S, 17°14’E 1 000

wee eNO NHN SB ee eee ND eH D eee eee ODD DH HIB DYNA BO}

195 33°49'5, 17°13’E 1 000 Sspinacidermis Barnard

. ravidula n.sp.

196 33°51'S, 17°14’E ~—s- 1: 350 . Spinacidermis Barnard

PO PO PRY PPO SD Py Fo Py Oe Pe Po Pu Pe By By Py Fe Dy 2 Do Pe Dy Fe Py De Pe De De Pe

. robertsi n.sp.

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 153

METHOD

Standard procedures in morphometry have been followed in this paper,

and to interpret the variations in the proportional dimensions which take

place during growth and which exist between the sexes, as many specimens as

possible from each of the species have been examined and measured. As a

check against the high intraspecific variability shown by rajids, counts of the

number of precaudal vertebrae have been made and, where possible, the morpho-

logical structures of the claspers of adult males have been examined. For each

species, figures of the external morphology of the clasper glans have been given.

In order to conform with previous work, the scheme of measurement

which has been adopted in this paper is that of Bigelow & Schroeder (1953),

and is represented diagrammatically in figure 1. Measurements of each specimen

have been recorded to the nearest millimetre and have been calculated as

permillage (thousandths) of the total length of the specimen. For a given

species, the mean of each particular measurement has been calculated and is

expressed in tabular form, together with the range of variation of that measure-

ment. In some cases, where few specimens were available, the proportional

dimensions of each specimen are given.

It should be noted that Hubbs & Ishiyama (1968) have suggested that

the disc width should be used as a basis for computing the proportional sizes

of the body parts, since the growth of the tail is negatively allometric (hetero-

gonic) and the tail is frequently damaged. However, this work has shown that

a much wider variation for each measurement (except the disc length) is

obtained when disc width is used, and the total length has therefore been

employed as the basis of calculation in all cases.

There would appear to be differences used in the terminology of the

structures of the clasper glans by Leigh-Sharpe (1920-6), by Ishiyama (1958)

and by Ishiyama & Hubbs (1968). Stehmann (1969) has evaluated these, and

the terminology used in this paper is mainly in accordance with his findings.

Vertebral counts were facilitated by the use of X-ray photography. The

number of precaudal vertebrae (Vprd) has been taken as the number of caudal

vertebrae up to the origin of the first dorsal fin (Ishiyama, 1952; Krefft, 1968q) ;

the number of trunk vertebrae (Vtr) and total count (V2’) are given according

to Krefft (1968q).

Although Bigelow & Schroeder (1953) have employed X-ray photography

in examinations of the snout, this method has not been used in the present

study, because of the small extent of calcification of the rostral cartilages

(Ishiyama & Hubbs, 1968). The rostral cartilages and rostral appendices were

examined by dissection.

SYSTEMATIC DiIscussION

Family Rajidae

Flat, depressed head and body, forming a rhomboidal disc. Eyes prominent,

rising above level of head; spiracles close behind eyes. Tail moderately slender,

ANNALS OF THE SOUTH AFRICAN MUSEUM

woe

“UJ [eSIOP ISI JO UISTIO 0} JUDA JO a[PprlU—Y ‘yp ynour—@&

‘goueysIp [eseusayuI—q {y}Sus] [eseusid—O SyjSuay peroord—NJ {3UaA Jo a[pPpruu 0} nous—JW SUy [esIop PUoeS Jo ySus] aseq—'T

‘gouds [es1opisqUI—y {Uy [esIop IsIy Jo ySusT aseq — [ {e0urjstp sepnovstdsssyuI—]J foouysip [eiqro193Ul—H Sapesds pue s49—*)

‘ako Jo Ja}aUIeIp [eUIPNSUO]— J SYIPIM osIp—W SYYPIM OsIp 389}8218 0} Jnous—] ‘yISuUsT Nous —1) SyyBus] osIp—g {YISsu2] [e101-—V

*suauutoods a}yeys 1OF JUIUIINSVIUL JO BULBS *I “OL

(=)

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 155

but not whip-like, with caudal fin reduced to a membranous fold. Two dorsal

fins. Pelvic fins with outer margins either weakly concave, or so deeply concave

that anterior lobe forms separate, three-jointed, limb-like structure. Skin on

dorsal surface of disc and tail with small spinules and spines, or larger thorn-

like denticles, or both, but without serrate tail spines. Ventral surface smooth,

or with pointed spines, or with small, flattened asperities.

In this family eight genera are distinguished, but only three of these occur

in the eastern South Atlantic.

Key to Genera

1 (a) Anterior lobe of pelvic fin forming slender, three-jointed limb-like structure, separate

externally from posterior fin-like region (fig.2 A) .. : Ps Cruriraja

(6) Anterior lobe of pelvic fin continuous externally with eeeenae anon along outer margin

of fin, not forming a separate limb (fig. 2 B) .. ie : ie : Se Gn

2 (a) Anterior radials of pectoral fins falling distinctly short of tip of snout; lees appendices

fused to stout, rod-like rostral bar throughout their aican Shield usually present in

clasper glans (figs 5-11, 15, 17-19) : Raja

(6) Anterior radials of pectoral fins Beers aig 5 tip oF en: eesel appendices

broadly united basally and hanging posteriorly free from soft, delicare rostral bar, without

anterior notch; each appendix separated posteriorly from axial bar by a notch about

4 as long as appendix. Shield absent in clasper glans (fig.21 B) ... .. Bathyraja

r I

Fic. 2. Ventral surface of pelvic fin: A. Cruriraja; B. Raja.

Genus cRURIRAJA Bigelow & Schroeder

Cruriraja Bigelow & Schroeder, 1948: 549; 1953: 313. Smith, 1964: 286. Wallace, 1967: 7, fig. 3.

Type-species: Cruriraja atlantis Bigelow & Schroeder, 1948.

Pectorals with radials of ordinary form, without lateral processes. Outer

margins of pelvics deeply notched, to form an anterior, limb-like structure,

consisting of three articulated segments, externally distinct from posterior,

fin-like lobe of pelvic. Tips of anterior rays of pectorals falling short of tip of

rostral cartilage.

I 56 ANNALS OF THE SOUTH AFRICAN MUSEUM

Three species of Cruriraja have been recorded in the southern African

region (Smith, 1964); one east coast species, one west coast species and one

species common to both coasts.

Key to species

1 (a) No thorns on tip of snout, along rostral ridge or in interspace between dorsal fins

C.. durbanensis

(b) A group of spines on tip of snout and along rostral ridge; thorns in dorsal interspace 2

2 (a) No thorns on mid-line of back above anterior half of abdominal region. Interdorsal

space usually less than half base length of first dorsal. Single enlarged thorn on anterior

dorsal border of clasper glans; ventral border with dermal denticles. . C.. triangularis

(b) Thorns present on mid-line of back above anterior half of abdominal region. Interdorsal

space usually greater than half base length of first dorsal. A single, enlarged thorn on

anterior dorsal border of clasper glans and an eperon on anterior ventral border (fig. 4)

C. parcomaculata

Cruriraja durbanensis (Von Bonde & Swart, 1923)

(Fig. 3)

Raia durbanensis Von Bonde & Swart, 1923: 11, pl. 22, fig. 1. Barnard, 1925: 69.

Cruriraja durbanensis: Bigelow & Schroeder, 1948: 550; 1953: 3153 1962: 199. Smith, 1964: 287.

Wallace, 1967: 7.

Types

The holotype, a juvenile male (232 mm total length) taken in 859 metres

at 30°10:00’S, 14°33:00’E, formerly in the collection of the Government

Marine Survey; now missing. The paratype, a female (311 mm total length)

also missing. The locality and depth for the paratype are noted as unknown.

Material

No specimens were available.

Unfortunately, the name C. durbanensis is misleading, as the type locality

given for this species, Station 343 (Von Bonde & Swart, 1923), is at a point

in the Atlantic Ocean about 640 km north-west of Cape Town. Bigelow &

Schroeder (1948, 1953) erroneously give the locality as off the Natal coast in

420 fathoms (769 m).

C. durbanensis is most easily distinguished from all other species of this

genus by the fact that there are no thorns on the snout and rostral cartilage,

and no thorns in the interdorsal space.

Description (Barnard, 1925: 69)

‘Width equal to distance from snout almost to middle of tail. In male

snout pointed but not produced, about 90°, anterior margin almost straight.

In female snout rounded, without point, anterior margins undulate. Outer

pectoral angle broadly rounded, hind margin moderately convex. Eye a little

less than interorbital width, 4 in preocular length of snout. Whole upper

surface of disc and upper and lateral surfaces of tail covered with spinelets;

in male 1 large spine in front of, 2 behind orbit; in female 5 and three respe-

tively; 2 (male) or 1 (female) suprascapular spines; a median row from occiput

————S””—”—”—”—”—”.TC—C PN rrr

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 157

to 1st dorsal in male; in female only extending about half-way along tail;

lower surface quite smooth.’

Colour

‘Reddish brown, lighter beneath.’

Fic. 3. Cruriraja durbanensis: the type, immature male. Dorsal view. (After

Von Bonde & Swart, 1923.)

Cruriraja parcomaculata (Von Bonde & Swart, 1923)

(Pl. 1A; Figs 4 A, 4 B)

Raia parcomaculata Von Bonde & Swart, 1923: 9, pl. 21, fig. 2.

Raja parcomaculata: Norman, 1935: 46.

Raia miraleius (non Linnaeus) Barnard, 1925: 68 (partim).

Raia smithi (non Miller & Henle) Smith, 1961: 66, fig. 68.

Raja caudaspinosa (non Von Bonde & Swart) Norman, 1935: 43 (partim).

Cruriraja parcomaculata: Bigelow & Schroeder, 1948: 5503; 1953: 315; 1962: 199. Smith, 1964:

288, pl. 26, fig. 27. Wallace, 1967: 11.

Types

The holotype, a juvenile (181 mm total length), trawled off Durban

(29°57°30'S, 31°34°15’E) in 545 metres, formerly in the collection of the

I 58 ANNALS OF THE SOUTH AFRICAN MUSEUM

Government Marine Survey; now missing. A single paratype, juvenile female

(108 mm total length) taken at the same locality and formerly in the same

collection, now in the collection of the British Museum (Natural History).

Material

II specimens of both sexes (160-550 mm total length) trawled off the

west coast from WNW Lideritzbucht to Cape Columbine in 267-622 metres

and from the east coast in Algoa Bay at 193 metres. 8 specimens preserved in the

collection of the South African Museum (SAM 24352-4, 24412, 24662).

C. parcomaculata was thought to be an east coast species (Smith, 1964),

but it is now known to occur along the south and west coasts of southern

Africa, where it is taken fairly regularly in commercial trawls. While the type

specimens were taken off Durban (Von Bonde & Swart, 1923), Wallace (1967)

has not recorded the species during a three-year survey, so that C. parcomaculata

probably does not extend further north than about Port Alfred on the east

coast, and is replaced by C’ triangularis in the region Durban to Barra da Falsa

(Wallace, 1967).

C. parcomaculata closely resembles the east coast species C. triangularis, but

differs from it in having a compatatively broader disc, a greater interdorsal

space and a continuous median series of thorns along the back, which in

adults, where there is development of lateral rows, gives rise to the distinctive

‘cross’ pattern of spines. It should be noted that the definitive character given

2-0cm

th . sn

Fic. 4. Cruriraja parcomaculata.

A: external view of right clasper from the dorsal side.

B: lateral view of right clasper, opened to show structural features of the glans.

ap—apopyle; bs—boss; hp—hypopyle; kn—knob; ‘ps’—‘pseudosiphon’; rh—rhipidion;

sh—shield; ep—eperon; sp—spike; st—sentinel; th—thorn; se—sentina.

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 159

by Smith (1964) for the separation of the two species—the position of the line

across the widest part of the disc—is not so marked in adults, where there is

an overlap of critical values. Comparison of the claspers of the two species has

shown that there are major differences in both external and internal structures.

This confirms that C. parcomaculata and C’. triangularis are distinct species.

A large specimen (BM. 1935. 5.2.64) taken by the Discovery and identified

as R. caudaspinosa by Norman (1935) is in fact C. parcomaculata.

TABLE 2. C. parcomaculata. Measurements expressed as permillage of the total length. Number |

of specimens II.

Character Mean Range

Total length : : : : : H I 000

Disc width : . ‘ : : : 591 568-627

Disc length ; : : ‘ : 451 427-521

Snout to greatest disc cee : é : 278 2490-314

Snout to middle of vent : ‘ ; 390 360-437

Middle of vent to 1st dorsal se, é : 479 459-519

Snout length 2 : : : : 3 95 84-108

Preoral length . : : , ; ; 110 99-118

Prenasal length . ; : 2 : 5 83 76-88

Eye, longitudinal diameter . ; é : 41 20-55

Eye and spiracle : i 3 : : 52 44-60

Spiracle . : : ‘ ‘ ; 23 16-31

Interorbital Dyeae : 4 : ; : 32 28-38

Interspiracular distance ; ; : i isi 66-82

Internasal distance. : : : : 50 44-59

Mouth width. . : é : s 61 52-70

Gill slit lengths: ist. : ; ‘ 4 16 14-21

ard’. : : é ; 17 15-20

5th : : ; : 14 10-17

Distance between inner ends of gill slits:

ESE vie : : : : 126 116-140

5th: : : : 2 53 46-60

Pelvic fin (anterior lobe) :

anterior margin . ; : i I14 101-133

length. : ‘ ‘ . : ‘ 97 79-116

base width . : : ‘ 3 ; 27 21-34,

1st dorsal fin: height : : : : 28 20-34

base length : ‘ : 42 32-59

and dorsal fin: height : : : 4 20) 20-35

base length ! : : 38 32-48

Interdorsal space : : : : 39 28-49

Description

Disc 1-:2-1°4 times as broad as long, its width 1-6—1-8 in total length;

angular in front, with maximum angle in front of spiracles 120° in juveniles

and females and 112° in adult males; anterior margins weakly concave close

behind tip of snout and again at level of spiracles; outer angles narrowly

rounded, posterior angles broadly so; posterior and inner margins convex.

Axis of greatest breadth 1-4—1-6 times as far from tip of snout as from posterior

edge of disc. Tail with lateral folds, well developed and expanded on posterior

third, so that tail is narrower at middle of length than distally; its length from

160 ANNALS OF THE SOUTH AFRICAN MUSEUM

middle of vent to origin of first dorsal fin 1-1-1-2 times the distance from middle

of vent to tip of snout; its length from middle of vent to tip of tail 1-4-1-7

times the distance from middle of vent to tip of snout.

Juvenile specimens with 5 thorns around inner edge of each orbit and 1

pair of widely separated thorns between spiracles; 1 median nuchal thorn and

2 scapular thorns on each side; a row of 23-27 thorns along mid-line of back

and tail, extending from immediately behind scapular region to origin of first

dorsal; median row flanked on each side by an irregular series of smaller

spines; 2—3 thorns in dorsal interspace. Whole upper surface of disc with fine

spinules, except on outer posterior edges; larger spines on tip of snout

and along anterior margins. Lower surface of disc and tail completely

smooth.

Adult specimens with 11-13 thorns around inner margin of each orbit

and spiracle; 2 pairs of thorns between spiracles; 2 (sometimes 4) median

nuchal thorns, usually arranged in a double series; 3 scapular thorns on each

side; 4-5 irregular rows of stellate-based thorns along mid-dorsal region from

nuchal thorns to origin of first dorsal, diminishing to a double series of slender,

recurved thorns about half-way along tail; a single lateral row of smaller

spines on each side of tail, from about posterior edge of pelvics to about level

of dorsal interspace; 4-8 thorns in interspace between dorsals. A group of

thorns on tip of snout and along rostral cartilage, and numerous, irregularly

arranged, stellate-based spines along anterior margins of disc to outer angles.

Otherwise smooth on upper surface, except in some cases a few small spines

at inner margin of pectoral. Lower surface of disc and tail without spines.

Snout slightly pointed but not produced, its length in front of orbits

2‘6-3'5 times as long as distance between orbits; its length in front of mouth

I'g—-2'4 times as long as spiracles; distance between orbits 1-0-1-8 in length

of orbit. Rostral cartilage projecting from cranium as hard bar; anterior rays

of pectorals extending about half the distance from front of orbits to tip of snout.

Mouth almost straight; nasal curtain not fringed; expanded posterior

margin of nostril conspicuously fringed. Teeth arranged in 39-44 rows in

upper jaw, juveniles with 30-31 rows; teeth blunt and flat and arranged

in quincunx in juveniles, but sharp pointed and regularly arranged in

adults.

Pelvic fins divided into slender, limb-like anterior lobe, arising separately

from ventral surface of disc, and posterior fin-like lobe.

Dorsal fins similar in shape with rounded apices; first dorsal usually

larger than second; dorsal interspace varying considerably from o-6-1-5 times

as long as base of first dorsal.

Number of precaudal vertebrae (Vprd) 66-69.

Colour

Juveniles uniformly brown, sometimes with scattered darker patches on

disc and tail. Adults typically uniformly brown, with irregular lighter and

darker areas. Lower surface of disc and tail white.

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 161

Genus RAJA Linnaeus, 1758

Raja Linnaeus, 1758: 231.

Type-species: Raja clavata Linnaeus, 1758 (as designated by Jordan & Gilbert,

1883).

Pectorals with radials of ordinary form, without lateral processes. Outer

margins of pelvics more or less concave, but not deeply notched and not

forming separate, limb-like structure. Tips of anterior rays of pectoral fins

falling short of rostral appendices. Rostral cartilage stout and bar-like, without

segment; rostral appendices fused to lateral edges of rostral cartilage throughout

length. Characters otherwise the same as for the family.

Genus BATHYRAJA Ishiyama, 1968

Bathyraja Ishiyama & Hubbs, 1968: 407, figs 1, 2.

Type-species: Bathyraja tsotrachys (Giinther, 1877)

Pectorals with radials of ordinary form, without lateral processes. Outer

margins of pelvics more or less concave, but not forming a separate, anterior,

limb-like structure. Rostral appendices broadly united basally with rostral bar,

and extending posteriorly as unnotched processes; posterior wings of rostral

appendices separate from rostral bar. Shield absent in clasper glans. Characters

otherwise the same as for the family.

Note: Ishiyama & Hubbs (1968) also define the genus Bathyraja by the presence

of a pseudosiphon in the clasper glans. Investigation of the claspers of the

radiata-complex has revealed that a true pseudosiphon, as defined by Ishiyama

(1958) and subsequently modified by Ishiyama & Hubbs (1968), is definitely

present in the species R. radiata, R. doellojuradoi, R. hyperborea and R. robertsi

n.sp. Furthermore, Ishiyama & Hubbs (1968) restrict Bathyraja to the Indo-

Pacific and Breviraja to the Atlantic, but the distributional pattern of Bathyraja

smithit shows this to be incorrect. Krefft (personal communication) suggests

that Breviraja is confined to slope areas in the tropics, whereas Bathyraja is a

bipolar, antitropical genus.

KEY TO SPECIES OF Raja AND Bathyraja

t (a) Mucus pores on ventral surface of disc pigmented, appearing as small black spots and

Sieeaks ~ *... uae ee. Ld Be abe

(6) Mucus pores on conta eee af ise ae piememed a i een aS

2 (a) Tail long and slender; snout markedly produced and very maimed R. lanceorostrata

(6) Tail not long and Specie: snout somewhat pointed but not very produced pe BG

3 (a) Ventral surface of disc completely covered by small, close-set asperities R. springert

(6) Ventral surface of disc never completely covered by small asperities, although spines

and spinules may be present on tip of snout, internasal region and anterior margins

of disc

4 (a) Distance from parila of vent to > tip of snout greater ‘than distance ees middle of vent

to tip of tail, by an amount about equal to or greater than the preorallength = R. doutrei

(5) Distance from middle of vent to tip of snout equal to or greater than the distance from

middle of vent to tip of tail, by an amount less than half the preoral length site LAG,

5 (a) Snout produced; interorbital distance about 5:5 in snout length; internasal distance

about 3 in preoral length + Se ~ est au v6 R. stenorhynchus

162

ANNALS OF THE SOUTH AFRICAN MUSEUM

KEY TO SPECIES OF Raja AND Bathyraja (continued)

(0)

6 (2)

)

20 (a)

(5)

Snout not produced; interorbital distance 3—4 in snout length; internasal distance about

2 in preoral length FS ; : #g R. pullopunctata

Distance from middle of vent to origin Mor frat doral fin eclates than distance from

middle of vent to tip of snout ..

Distance from middle of vent to ee a ee: dona fig be toan dig fun auniale

of vent to tip of snout .. : si ver

Upper surface of disc and tail couplet) sopeted sie fing eR a single row of

spines along mid-dorsal region of back and tail at ay ar R. spinacidermis

Upper surface of disc and tail without close-set spinules; 3 rows of thorns along mid-

dorsal region of back and 5 rows on tail ey ; : a: R. caudaspinosa

Upper surface of disc and tail liaiairiay, covered witla Gane close-set spinules; no

enlarged thorns on disc or tail . Le, : ae R. spinacidermis

Upper surface of disc and a never donne peed “en spinules (except some

juveniles) ; larger thorns always present on disc or tail xs ae 3 9

Less than 30 rows of teeth in upper jaw ae > ae AN ois) le aes

More than 30 rows of teeth in upper jaw__.. ae us oi dn Ppt ih fe)

Snout produced and sharply pointed .. Ag. re ad, bis cs R. alba

Snout not produced and not sharply pointed. . Bis at oe it rE

Ocelli absent on upper surface of disc. . a he oe ae oe ene md:

Ocelli black and yellow marbled ie oh af Ne ae 23 SP

Ocelli not marbled Ms : 13

Ocelli circular or only very settles pee: cite caussane of ches depaae rings of

colour... ae R. miraletus

Ocelli strongly horizontally we. consisting of a single Aas ring enclosing one or two

dark spots By aie .. R.clavaia

Distance from middle af vent to fap We snout about 1-5 times vereaees than distance from

middle of vent to tip of tail... : 15

Distance from middle of vent to tip a snout iss a eae Ga mate of oe to

tip of tail .. BA Ab. ve 33 et) a6

Upper surface of disc wath peleae nated fidras on anterior margins, rostral cartilage

and posterior angles. Ventral surface of disc white with some scattered darker blotches

R. radiata

Upper surface of disc without stellate based thorns on anterior margins, rostral cartilage

and posterior angles. Ventral surface of disc dark grey with white triangular patches

between nostrils, around mouth and at pectoral girdle al oF .. R. robertsi

Less than 50 rows of teeth in upper jaw a ey a ic By barn 7)

More than 50 rows of teeth in upper jaw... so oNgenet

Median row of thorns along back and tail absent. No apne on ronda canes

R. ravidula

Median row of thorns on back and tail se present. hors present on rostral

cartilage .. ‘ Bs

A single row of heme Plane paid’ daeel nomen oe back aad etl avealicen rows of

thorns (when present) always lateral. Number of precaudal vertebrae 46-54 .. 19

3 or more rows of thorns along mid-dorsal region of back and tail; auxiliary rows of

thorns semi-lateral in position; number of precaudal vertebrae 55-60 ; 20

Shield in clasper glans well developed and me (fig. 11 C); dorsal oaeee of disc

grey with numerous darker regular spots. . R. straeleni.

Shield in clasper glans poorly developed (fig. 10 C); dona es of ates ae or without

numerous irregular darker blotches. .. Be .. . clavata

Reduction in size to eventual loss of mid- durant rows ao Horas from disc to tail; lower

surface of tail uniformly grey with white distal tip; See vertebral count 65-69; ;

clasper with a slit and a cleft (fig. 15 C) Ae ne . R. dissimilis

Mid-dorsal rows of thorns large and continuous to aes of ba! dorsal fin; ie surface

of tail white or with scattered darker blotches; precaudal vertebral count 55-63;

clasper with two slits (fig. 17 C) ha} xt ne be ie R. confundens

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 163

KEY TO SPECIES OF Raja AND Bathyraja (continued)

21 (a) Distance from middle of vent to origin of first dorsal fin longer than distance from tip

of snout to greatest disc width, by an amount greater than the preoral length; precaudal

vertebral count 70; spike in clasper glans bilobed (fig. 19 C) by .. R. wallacei

(6) Distance from middle of vent to origin of first dorsal fin longer than distance from tip

of snout to greatest disc width, by an amount less than the preoral length; precaudal

vertebral count 55-58; spike in clasper glans bulbous but not bilobed (fig. 18 C)

R. leopardus

Note: R. ackleyi Garman has been recorded from the south Atlantic (Poll, 1951) at 3°10’S,

9°36’E and 4°58’S, 11°20’30’E, but this is outside the area of this survey.

Fic. 5. Raja doutrei.

A: rostral bar and rostral appendices.

B: external view of left clasper from the dorsal side.

C: lateral view of right clasper, opened to show structural features of the glans.

ap—apopyle; cf—cleft; hp—hypopyle; rh—rhipidion; sh—shield; sp—spike; st—sentinel.

164 ANNALS OF THE SOUTH AFRICAN MUSEUM

Raja doutrer Cadenat, 1960

(Pl 1B3, Figs 5 A,B, :G)

Raia doutrei Cadenat, 1960: 294, figs I-11, 13, 15.

Raja batis (non Linnaeus) Hulley, 1966: 512.

T ype

The holotype, a juvenile male (591 mm total length) taken at Sud Fosse

Kayar, Senegal, in 450-500 metres in the collection of the Muséum National

d’Histoire Naturelle, Paris.

Material

Two adult males (845, 931 mm total length) taken SW by W Liideritz-

bucht in 494 metres. Both specimens in the collection of the South African

Museum (SAM 24699, 24700). A single female (SAM 34564) from the same

locality not suitable for measurement.

This is a new record for this species in the South Atlantic. The specimens

were originally thought to be R. batis Linnaeus (Hulley, 1966), but comparison

of the claspers with those of R. batis from the North Atlantic (Hulley, 1966:

figs 6, 7) has revealed that these specimens do not belong to the latter species.

However, the specimens are identical in shape, coloration, teeth count and

spination pattern, particularly on the tail, to R. doutre: from the coast of

Senegal (Cadenat, 1960). Counts of the number of precaudal vertebrae

(Vprd 46 in the type) confirm this identification.

R. doutre: closely resembles the east coast species R. stenorhynchus, but may

be distinguished from this species by its comparatively shorter, narrower snout,

longer and wider disc, and by its longer tail. Furthermore, it appears that the

tooth count in the upper jaw is slightly lower than that of R. stenorhynchus,

although the validity of this small difference as a taxonomic character is open

to question.

Description

Disc about 1-2 times as broad as long, its width 1-3-1-4 in total length;

maximum angle in front of spiracles 72°; anterior margins sinuous from snout

to outer angles, convex at level of nostrils and concave a little behind level of

mouth; outer angles sharply rounded, posterior margins gently convex. Axis

of greatest breadth 1-5-1-9 times as far from tip of snout as from posterior

edge of disc. Tail expanded distally with well-developed lateral folds, especially

on posterior third; its length from middle of vent to origin of first dorsal fin

2°0-2°2 in length from middle of vent to tip of snout; its complete length from

middle of vent to tip of tail 1-5 in length from middle of vent to tip of snout.

Adults with 6-7 thorns around inner margins of each orbit and above

each spiracle; 0-1 very small, median nuchal thorn. A series of 13-26 thorns

along midline of tail from above vent to origin of first dorsal; 1-2 thorns in

dorsal interspace. No lateral rows of thorns on tail in adult males, but females

typically with thorns in a single series on each side. Spines on dorsal surface

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 165

TABLE 3. R. doutrei. Measurements expressed as permillage of the total length.

3 Number of specimens 2.

SAM SAM

Character 24699 24.700

Total length 3 ‘ t ‘ ‘ é I 000 I 000

Disc width : : ‘ P , : 723 751

Disc length : : : ‘ : ; 605 626

Snout to greatest disc width . : : 2 308 374

Snout to middle of vent ; : z : 505 600

Middle of vent to 1st dorsal origin. : : 300 269

Snout length : : A : é : 197 199

Preoral length . . . : . : 1Q2 192

Prenasal length . : : : ; ; 175 173

Eye: longitudinal diameter . : ‘ : 38 46

Eye and spiracle k : P : ; 53 56

Spiracle . : ; : : ! : 17 23

Interorbital distance . ‘ ; : : 51 50

Interspiracular distance : ; t : i 68

Internasal distance. : 4 : QI 95

Mouth width. 3 ‘ : : : QI 99

Gill slit lengths Ist. : : : ‘ 37) 19

ord). : : : : 17 20

5th «. : H 5 : 13 17

Distance between inner ends of gill slits:

Sty. 3 ; : F ‘ 153 164

5th : ‘ ‘ : 94 95

1st dorsal fin: height : : : 3 33 44.

base length ‘ j : 49 45

and dorsal fin: height ‘ ‘ : : 49 45

base length ; : ‘ 48 45

Interdorsal space ; : : : ; 18 17

on tip of snout and along anterior margins of disc from about opposite eyes

to outer angles. Ventral surface with spines on snout and along anterior

margins to level of mouth. Otherwise smooth on both dorsal and ventral

surfaces.

Snout produced and slightly pointed, its length in front of orbits 3-8—4-0

times as long as distance between orbits; its length in front of mouth 2-0-2°1

times as great as distance between nostrils. Orbits 2-0—2-2 times as long as

spiracles; distance between orbits 1-1-1-3 times as great as length of orbit.

Rostral cartilage projecting from cranium as hard bar, with rostral appendices

fused to bar throughout their length; anterior tips of pectoral fins falling short

of appendices.

Mouth slightly arched; nasal curtain fringed; expanded posterior margin

of nostril heavily fringed. Teeth arranged in 32 regular rows in upper jaw,

each with a single, large, posterior cusp.

Anterior lobes of pelvics fin-like and continuously connected with posterior

lobes along outer margin of fin.

Dorsal fins similar in size and shape, with broadly rounded apices. Small

but definite interspace between dorsals, its length 37°% as long as base of first

dorsal.

166 ANNALS OF THE SOUTH AFRICAN MUSEUM

Number of precaudal vertebrae (Vprd) 47-49 (male), 43 (female).

Colour

Upper surface of disc and tail uniformly brown, with some scattered,

irregular, darker blotches. Lower surface brown; terminal ends of mucus

pores pigmented, appearing as widely-spaced black spots on the snout and

anterior margins of the disc.

Raja pullopunctata Smith, 1964

(PI. 2 A, B. Figs 6 A, B,C)

Raia pullopunctata Smith, 1964: 285, pl. 25.

Raja pullopunctata: Hulley, 1966: 505, figs 4, 5. Wallace, 1967: 13, fig. 7.

Raia batis (non Linnaeus) Thompson, 1914: 156. Von Bonde & Swart, 1923: 3. Barnard, 1925:

70, non pl. 4, fig. 3. Smith, 1961: 66, pl. 3, fig. 65.

Raja batis (non Linnaeus) Norman, 1935: 39. Fowler, 1941: 385. Bigelow & Schroeder, 1953: 146.

Raja campbelli Wallace, 1967: 24, fig. 12.

Raia stabuliforis: Von Bonde & Swart, 1923: 12.

Types

The holotype of R. pullopunctata, a juvenile male (205 mm total length)

and paratype, a female (175 mm total length), both trawled in 183 metres in

Algoa Bay, in the collection of the J. L. B. Smith Institute of Ichthyology,

Grahamstown.

The holotype of R. campbelli, a female (663 mm total length) and paratype,

a male (386 mm total length), trawled east of Durban in 320 metres and

north-east of Durban in 137 metres respectively, in the collection of the Oceano-

graphic Research Institute, Durban.

Material

Seven female and four male specimens (221-1250 mm total length)

trawled from west of Cape Town to Port Alfred in 91-457 metres, including

the types and paratypes of R. pullopunctata and R. campbelli. Eight specimens

preserved in the collection of the South African Museum (SAM 22635, 22652,

24349-51, 24456, 24497).

On the basis of external morphology, Smith (1964) distinguished between

R. pullopunctata and the European species R. batis, to which the South African

material had previously been referred. This distinction has been confirmed by

Hulley (1966) after detailed examination of the clasper structures of the two

species.

Recently Wallace (1967) has described a new species of ‘black-bellied’

skate, R. campbelli, from the Natal coast. This species closely resembles R.

pullopunctata but was distinguished from it on the basis of the shape of the snout

and the spination of the rostral cartilage. It must be pointed out, however,

that Wallace only examined juvenile specimens of R. pullopunctata (340 mm

maximum total length). Specimens of R. pullopunctata in the collection of the

South African Museum, show a range of variation in the above characters.

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 167

In more mature specimens (387-1250 mm total length) spinules appear first

on the ventral tip and then on the dorsal tip of the rostral cartilage, while the

shape of the snout varies with age: width of disc at level of orbits 2-3—2-6 times

the preorbital length of the snout in juveniles (221-344 mm total length);

2°0-2°3 times in more mature specimens (387-540 mm total length); and

; 5-0cm \

Fic. 6. Raja pullopunctata.

A: rostral bar and rostral appendices.

B: external view of right clasper from the dorsal side.

C: lateral view of right clasper, opened to show structural features of the glans.

ap—apopyle; cf—cleft; hb —hypopyle; rh—rhipidion; sh—shield; sp—spike.

168 ANNALS OF THE SOUTH AFRICAN MUSEUM

1-8—1-9 times in adult specimens (660-1250 mm total length). It is therefore

held that specimens described as R. campbelli are specimens of R. pullopunctata.

This is further supported by the fact that there is no difference in the precaudal

vertebral count between the two species.

R. pullopunctata most closely resembles R. flavirostris Philippi from the

Falkland Islands, but may be distinguished from this species by the length of

the tail and by the number of teeth in the upper jaw. Examination of the

claspers of R. flavirostris has shown the two species to be distinct. Furthermore,

the precaudal vertebral count in R&. flavirostris appears to be slightly less than

that for R. pullopunctata.

R. pullopunctata is most easily distinguished from other South African

skates, except R. lanceorostrata and R. stenorhynchus, by the black-spotted ventral

surface of the disc and by the presence of a single large, median nuchal thorn

(sometimes 2). It may be separated from R. lanceorostrata by the shape of the

tail and snout, and from R. stenorhynchus by the shape of the disc and the nature

of the rostral cartilage.

TABLE 4. R. pullopunctata. Measurements expressel as permillage of the total length.

Number of specimens 10.

Character Mean Range

Total length ; ‘ : : : : I 000

Disc width : ; ‘ , . : 734 678-784

Disc length : : : ; : : 573 550-598

Snout to greatest disc width . : ‘ ; 320 294-366

Snout to middle of vent : : : : 526 493-567

Middle of vent to 1st dorsal origin. , ; 304. 272 eee

Snout length ‘ , : 5 ‘ ; 165 147-189

Preoral length . ‘ ; , ; : 167 156-183

Prenasal length . : : , . ; 136 120-152

Eye: longitudinal diameter . ; : ; 40 29-46

Eye and spiracle y : : : ; 54 50-53

Spiraclela : : : » , ; 22 19-26

Interorbital distance . : : ; ; 48 43-46

Interspiracular distance : : : : 72 70-75

Internasal distance . : ; ‘ : 88 79-95

Mouth width . : } : ; , 92 88-96

Gill slit lengths: 1st. . L ; ' 15 12-20

and . , : : A 16 12-23

5th : 4 4 i 13 I1-19

Distance between inner ends of gill slits:

TSt |e : : : , 162 155-171

Ae ata : ; ; : 93 82-98

ist dorsal fin: height : : : 34 24-39

base length . : A f 52 42-62

and dorsal fin: height : ‘ : : 30 18-35

base length . A 4 ; 49 40-58

Interdorsal space , : ; ¢ : 28 21-36

Description

Disc about 1-2-1-4 times as broad as long, obtuse in front, with anterior

angle in front of spiracles 92°-108°; anterior margins weakly concave posterior

to snout in juveniles (width of disc at level of orbits 2:3—-2-6 times in preorbital

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 169

length of snout), but more sinuous and particularly concave abreast of spiracles

in older specimens (width of disc at level of orbits 1-8—2-3 times in preorbital

length of snout); posterior and outer angles broadly rounded. Axis of greatest

breadth 1-1-1-4 times as far from tip of snout as from posterior edge of disc.

Tail with moderately wide lateral folds (wide in adults), its length from middle

of vent to origin of first dorsal fin 1-6-1-9 in length of middle vent to tip of

snout.

Juvenile specimens with 2 thorns in front of orbit and 1 behind; a single,

large, median nuchal thorn; a row of 8-12 thorns along mid-line of tail, from

above vent to origin of first dorsal fin; and 1 thorn in interspace between first

and second dorsals. No lateral rows of spines on tail. Otherwise entire dorsal

and ventral surfaces of disc and tail naked.

Larger specimens with 3-4 thorns around inner margin of orbit; a single,

median nuchal thorn (sometimes worn); a row of 10-12 thorns along mid-line

of tail, from above vent to origin of first dorsal; and 1-2 thorns in dorsal inter-

space. No lateral rows of thorns on tail. Upper surface of disc smooth, but

lower surface with spinules on tip of snout and anterior edges of disc to about

level of nostrils, and on internasal area.

Adult specimens with 5-8 thorns around inner margin of orbit; 1 median

nuchal thorn (sometimes 2), usually worn; a row of 26-27 thorns along mid-

line of tail, from above vent to origin of first dorsal; and 3—4 thorns in dorsal

interspace. A single, irregular row of lateral caudal thorns (15-20) on each

side of tail, from about posterior edge of pelvics to interspace between dorsals.

Dorsal surface of disc and tail with widely spaced, blunt asperities, except on

suprascapular region. Spinules on ventral surface of snout, anterior margins

to level of nostrils and on internasal area.

Snout produced, more markedly so in adults, its length in front of orbits

3°4-4'1 times as long as distance between orbits; its length in front of mouth

I-Q-2°0 times as great as distance between nostrils. Orbits 1-5-1-9 times as

long as spiracles; distance between orbits 1-5—1-9 times as great as length of

orbit. Rostral cartilage projecting from cranium as hard bar, with rostral

appendices fused to the bar throughout their length; tips of anterior rays of

pectoral fins falling short of appendices.

Mouth slightly arched; nasal curtain not fringed; expanded posterior

margin of nostril heavily fringed. Teeth with single, large cusp in males, but

more or less flat in females and juveniles, arranged in 53-58 rows in upper jaw.

Anterior lobes of pelvics fin-like and continuously connected with posterior

lobes along outer margin of fin.

First dorsal fin usually only slightly larger than second, but similar in

shape, with convex anterior margin and broadly rounded apex; interspace

between dorsals 47—73°% as long as base of first dorsal.

Number of precaudal vertebrae (Vprd) 50-58.

Colour

Dorsal surface light brown to biscuit, sometimes darker, with numerous

170 ANNALS OF THE SOUTH AFRICAN MUSEUM

small darker spots and blotches. Ventral surface greyish, with terminal ends

of mucus pores darkly pigmented, appearing as black spots and streaks.

Raja caudaspinosa (Von Bonde & Swart, 1923)

(PIS 9 A, Bs Pies 7 a5.)

Raia caudaspinosa Von Bonde & Swart, 1923: 8, pl. 21, fig. 1. Barnard, 1925: 66.

Raja caudaspinosa: Norman, 1935: 43 (partum). Fowler, 1941: 376.

non Raia caudaspinosa: Smith, 1961: 67, non fig. 72 (= R. leopardus).

Raia albalinea Von Bonde & Swart, 1923: 6, pl. 20, fig. 1.

Types

The holotype of R. caudaspinosa, a female (346 mm total length), trawled

off the Natal coast in 512 metres, formerly in the collection of the Government

Marine Survey; now missing.

The paratype of R. albalinea, a juvenile female (108 mm total length),

taken at 32°3:00’S, 16°2-00’E in 512 metres, formerly in the collection of the

Government Marine Survey; now in the British Museum (Natural History).

The type of R. albalinea taken by the Government Marine Survey, now missing.

Material

31 specimens of both sexes (304-580 mm total length) trawled between

292 and 914 metres from Port Nolloth to west of Cape Town. 29 specimens

preserved in the collection of the South African Museum (SAM 23187, 24420,

24449).

Although the number of precaudal vertebrae in the tail is indicative of

the genus Bathyraja (Ishiyama & Hubbs, 1968), this species has been retained

in the genus Raja because of the structure of the rostral bar and rostral

appendices, and because both a rhipidion and shield are present in the clasper

glans.

While resembling R. wallacei and R. confundens in the heavy spination of

the tail, R. caudaspinosa may be separated from these and from all other known

South African species except juvenile specimens of R. spinacidermis by virtue

of the fact that the distance from the middle of the vent to the origin of the

first dorsal fin is greater than the distance from the middle of the vent to the

tip of the snout. In this and other characters, R. caudaspinosa most closely

resembles R. fyllae Liitken from the northern Atlantic, but differs in being

comparatively narrower across the disc, having a shorter snout and a narrower

tail than the latter species. Unlike R. fyllae, juvenile specimens of R. caudaspinosa

possess dark cross-barring on the tail, which is more typical of R. senta Garman,

but they differ from this species in spination. Furthermore, Bigelow & Schroeder

(1953: 199) report that ‘the range of R. fyllae appears to be defined by its

preference for a narrow temperature range of a few degrees above the freezing

point of salt water’, so that while this condition may be satisfied on the South

African west coast, it seems unlikely that it applies to the waters off Natal

(type taken in 512 metres).

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA Lf

Examination of the type of R. albalinea led both Barnard (1925) and

Norman (1935) to include this species as a synonym of R. caudaspinosa. Unfor-

tunately the type is now missing, but a comparison of the position of the vent

in the paratype with its position in R. caudaspinosa has led the present author

to accept this synonymy.

2.0 cm

£3 BAN 8 2.0cm

cn BE Waal 3

e ‘ E 3

Fic. 7. Raja caudaspinosa.

A: rostral bar and rostral appendices.

B. external view of right clasper from the dorsal side.

C: lateral view of right clasper, opened to show structural features of the glans.

ap—apopyle; cf—cleft; hp—hypopyle; ae a se—sentina; sh—shield; sp—spike;

st—sentinel.

172 ANNALS OF THE SOUTH AFRICAN MUSEUM

TABLE 5. R. caudaspinosa. Measurements expressed as permillage of the total length.

Number of specimens 31.

Character Mean Range

Total length f , : : : : I 000

Disc width : 3 ; j , f 547 520-573

Disc length } . ‘ : 405 379-446

Snout to greatest disc = falas : ; ; 250 214-273

Snout to middle of vent 4 ; 4 393 372-409

Middle of vent to 1st dorsal ie ; : 468 450-494.

Snout length ; : : : ; ; 78 66-87

Preoral length . : ; : ; , 85 72-96

Prenasal length . : ; : : ; 63 55-74

Eye: longitudinal diameter . : : , 40 36-43

Eye and spiracle : : ; ; : 49 46-52

Spiracle . : : : : ; 19 16-23

Interorbital disauee : : : : , 33 30-36

Interspiracular distance : : : 66 59-74.

Internasal distance. f : : ; 60 55-64

Mouth width. ; : : : : 68 57-78

Gill slit lengths: 1st. : 3 : 4 15 11-18

ard)! : : ; : 17 13-18

5th : : . 13 10-16

Distance between inner ends ae oil slits:

AS OH, 00 , ; : ; 137 126-144

bb ne y : ‘ : 7h 58-83

1st dorsal fin: height ; : : : 26 20-36

base length . ; ; ; 51 44-57

ond dorsal fin: height ; : 5 ; 26 21-34

base length . ; ; : 51 43-59

Interdorsal space : ; : : , )

Description

Disc about 1:2—1-4 times as broad as long, its width 1-7—1-9 in total length;

conspicuously obtuse in front, with maximum angle in front of spiracles

125°-130°; anterior margins evenly convex in juveniles, but considerably

concave at level of spiracles in adults, and particularly indented in adult males;

outer angles broadly rounded, posterior margins evenly convex. Axis of

greatest breadth 1-1—-1-9 times as far from tip of snout as from posterior edge

of disc. Tail with moderately wide lateral folds along posterior third, its length

from middle of vent to origin of first dorsal fin 1-1-1-3 times as great as distance

from middle of vent to tip of snout.

Specimens typically with dorsal surface of disc smooth; large stellate-

based thorns on anterior margins arranged in 4-6 irregular rows and on snout

and rostral cartilage; 5-9 large thorns around inner margins of each orbit;

o-1 pair thorns between orbits and 1-2 pairs between spiracles; 4-5 median

nuchal thorns with 3-5 thorns in a group posteriorly and 3-4 scapular thorns;

a median series of 22-33 thorns along mid-line of back and tail to first dorsal

origin, decreasing in size posteriorly, but never absent. Median series flanked

by 1 row of equally large thorns on back, and 2 rows on each side on tail,

making 5 rows of thorns in this region; lateral rows extending to origin of

second dorsal. No interdorsal thorns. Small spines and spinules on anterior

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 173

edges of disc and on tail. Dorsal fins spinulose. Ventral surface smooth.

Snout obtuse and not produced, its length in front of orbits 2-0—2-6

times as long as distance between orbits; its length in front of mouth 1-2-1°6

times as great as distance between nostrils. Orbits 1-4-2-7 times as long as

spiracles; distance between orbits 1-1-1-4 in length of orbit. Rostral cartilage

projecting from cranium as hard bar with rostral appendices fused to bar

throughout their length; anterior rays of pectorals extending about two-thirds

the distance forward from level of front of orbits towards tip of snout.

Mouth arched medially; nasal curtain deeply fringed; expanded posterior

margin of nostril heavily fringed. ‘Teeth in 32-36 rows in upper jaw, arranged

in quincunx and closely crowded medially; flattened and blunt in females but

with sharp posterior cusp in males.

Anterior lobes of pelvics fin-like and continuously connected with posterior

lobes along outer margin of fin.

First and second dorsals confluent; second dorsal a little longer and lower

than first, and continuous with caudal.

Number of precaudal vertebrae (Vprd) 66—73.

Colour

Upper surface of disc dusky-grey to brown, with or without some darker

spots. Lower surface of disc and tail pale.

Raja spinacidermis Barnard, 1923

| (Pl. 4 A, B)

Raia spinacidermis Barnard, 1923: 440; 1925: 73, pl. 4, fig. 6. Smith, 1961: 66.

Raja spinacidermis: Norman, 1935: 46. Fowler, 1941: 392.

Raja mollis Bigelow & Schroeder, 1950: 388, pl. 2; 1953: 237, fig. 51. Templeman, 1965: 268,

figs 10-13. Krefft & Liibben, 1966: 380, figs 1, 2.

? Raia plutonia: Barnard, 1925: 68.

Type

The holotype of R. spinacidermis, a female (600 mm total length), probably

trawled off Cape Point in deep water (locality label lost), formerly in the collec-

tion of the South African Museum; now in the collection of the British Museum

(Natural History).

The holotype of R. mollis, a juvenile male (262 mm total length), from

41°53'N, 65°35’W in 1568 metres, in the collection of the United States National

Museum.

Material

The holotype and two specimens, a juvenile male (377 mm total length)

and a juvenile female (638-5 mm total length), taken by R.V. Walther Herwig

at Station WH 195/67 (33°49’S, 17°13’E) in 1 000 metres, and at Station WH

196/67 (33°51'S, 17°41'E) in about 1 350 metres. These two specimens in the

collection of the Institut fiir Seefischerei, Hamburg.

This species is retained in the genus Raa because the rostral bar extends

174 ANNALS OF THE SOUTH AFRICAN MUSEUM

forwards beyond the anterior extremities of the pectoral rays, and because the

precaudal vertebral count is within the range for this genus (Ishiyama, 1967).

R. spinacidermis was previously known only from a single adolescent

specimen, the holotype, but two specimens were taken by R.V. Walther Herwig

in deep water off Cape Town in 1967. It may be distinguished from all other

known South African species by the spinulose nature of the whole upper surface

of the disc and tail and by the complete absence of larger thorns.

R. mollis, which has been taken in deep water in the eastern and western

North Atlantic (Bigelow & Schroeder, 1950, 1953; ‘Templeman, 1965; Krefft

& Liibben, 1966), so closely resembles R. spinacidermis in proportional dimen-

sions, tooth count and in its peculiar spination pattern, that it is considered

that these two species are synonymous. However, the vertebral count in R.

mollis (Vtr 25-29; Vprd 67-72) is slightly higher than that of R. spinacidermis

(Vtr 25; Vprd 60-65), the difference probably being due to environmental

factors. Although the two are widely separated geographically, this fact seems

to have little systematic bearing so far as deep-water species are concerned, cf.

R. richardson: (Garrick, 1961; Forster, 1965).

Although R. spinacidermis resembles R. ravidula, also from deep water off

the Cape, in the spinulose nature of the upper surface of the disc, it differs

markedly in the absence of thorns around the orbits, absence of thorns along

the mid-dorsal region of the back and tail, and in tooth count. The snout is

not as produced as in R. ramdula.

Norman (1935) held that the two spinulose juveniles, reported as R.

plutonia Garman by Barnard (1925), are specimens of R. spinacidermis, for

despite the presence of enlarged thorns above the orbits and along the mid-

line of the back and tail, the median series of thorns in the larger specimen

showed signs of disappearance, i.e. a similar relationship between juveniles

and adults as in R. fullonica (Clark, 1926). Although these specimens were not

available to the present author, two almost identical specimens (SAM 22911,

24450), answering exactly to Barnard’s (1925) description of R. plutonia, were

examined. The tooth count in these specimens is 36-38 rows in the upper jaw,

slightly higher than that of R. plutonia (32-36; Barnard, 1925), but well below

the range for R. spinacidermis. Furthermore, the shape of the disc and length of

the tail suggest a closer similarity with R. caudaspinosa, i.e. a similar relationship

between juveniles and adults as in R. fyllae (Clark, 1926; Bigelow & Schroeder,

1953). However, until further material becomes available, Norman’s (1935)

synonymy has been tentatively accepted.

Description

Disc about 1-2—1-4 times as broad as long, its width 1-4—1°5 in total length;

obtuse in front with anterior angle in front of spiracles 107°; anterior margins

almost straight; posterior margins evenly convex; outer and posterior angles

broadly rounded. Axis of greatest breadth 1-4-1-7 times as far from tip of

snout as from posterior edge of disc. Tail with well-developed lateral folds on

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 175

TABLE 6. R. spinacidermis. Measurements expressed as permillage of the total length.

Character Type 3 Q

(377 mm) (6385 mm)

Total length : : : f 4 1 O00 I 000 I 000

Disc width : : : : «OA 660 701

Disc length : i : : SUP Reply 538 542

Snout to greatest disc width . : =| )307 335 315

Snout to middle of vent ; : A AO6 482 487

Middle of vent to 1st dorsal origin. se SARICT: 391 379

Snout length . : f i Hin pi (ills tg 166 I5I

Preoral length . f ; : a) (126) 166 148

Prenasal length . ; : : a xOG) = 136 114

Eye: longitudinal diameter . : : 33 a7 36

Eye and spiracle R : ; p 42 46 47

Spiracle . 5 : 2 ¢ ; 20 — =

Interorbital distance . : : : 45 40 36

Interspiracular distance : : : 65 70 64

Internasal distance. : : ‘ 83 89 79

Mouth width =. : ; ; : 83 go 87

Gill slit lengths: 1st. 4 : : 13 15 14

ord |. i é ‘ 17 17 16

pidclnan ; - d iit 13 10

Distance between inner ends of gill slits:

Ist) #5 ; s 4), HBO 140 147

5th). ; : : 89 84 gI

ist dorsal fin: height : : : 29 22 22

base length . 3 A 47 50 48

2nd dorsal fin: height ; t é 27 26 19

base length . : : 42 45 55

Interdorsal space ‘ : . : 0) O O

* Measurements made on damaged snout of type specimen.

posterior third; its length from middle of vent to origin of first dorsal fin 1-1-1-3

in length from middle of vent to tip of snout.

Entire upper surface of disc, and upper and lateral surfaces of tail with

closely-set, fine, setiform spinules, larger and coarser on tail than elsewhere;

enlarged spines or thorns absent on disc and tail in adults, but juveniles with

1-4 enlarged spines before orbit, 0-3 spines behind, and a median row of

spines along back and tail, becoming reduced and finally absent in front of

first dorsal origin. Enlarged spines are lost with growth (?). Ventral surface

with spines on tip of snout and on tail, except on median line of distal two-thirds.

Snout pointed but not produced; its length in front of orbits 3-8—4-2 times

as long as distance between orbits; its length in front of mouth 1-5—-1-9 times

as great as distance between nostrils. Orbits 1-6 times as long as spiracles;

distance between orbits 1-0-1°3 times as great as length of orbit. Rostral

cartilage projecting from cranium as hard bar, extending beyond anterior

extremities of pectoral rays.

Mouth slightly arched; nasal curtain not fringed; expanded posterior

margin of nostril fringed. Teeth arranged in 54-60 regular rows in upper jaw,

blunt with posterior cusp laterally, slightly pointed medially.

I 76 ANNALS OF THE SOUTH AFRICAN MUSEUM

Anterior lobes of pelvics fin-like and continuously connected with posterior

lobes along outer margin of fin.

Dorsal fins about similar in size and shape, with broadly rounded apices;

dorsal fins completely spinulose. First and second dorsals confluent.

Number of vertebrae Vtr 28; Vprd 65; V2' 93.

Colour

Pale slate grey, becoming slightly darker towards posterior margins of

pectorals, and distinctly darker on pelvics. Lower surface similarly and as

deeply coloured as upper surface, with white mottling on distal region of tail.

Raja alba Lacépéde, 1803

(Pl. 5 A, B; Figs 8 A, B, C)

Raja alba Lacépéde, 1803: 663, pl. 20, fig. 1. Norman, 1935: 40. Fowler, 1936: 115; 1941: 365:

Van Bruggen, 1965: 190. Hulley, 1966: 497, fig. 8. Wallace, 1967: 27, figs 13-15.

Raia alba: Von Bonde & Swart, 1923: 5. Smith, 1961: 66, fig. 67; 1964: 285. Thorpe, 1964: 27.

Raja marginata Lacépdée, 1803: 663, pl. 20, fig. 2. Thompson, 1914: 158.

Raia marginata: Regan, 1908: 242. Gilchrist & Thompson, 1916: 285. Barnard, 1925: 65, pl. 4,

fig. 1. Clark, 1926: 409, pls 28, 29, 30, 31a. Von Bonde, 1932: 33. Barnard, 1947: 26, pl. 3,

figs 11, Ila.

Type

The locality of the types of R. alba and R. marginata not given in the original

description.

Material

21 specimens of both sexes (286-816 mm total length), trawled from False

Bay to Port Alfred in 110-183 metres. 9 specimens preserved in the collection

of the South African Museum (SAM 3235, 13026, 22578, 22633, 23590, 24320,

24336).

Clark (1926) has shown that R. marginata is synonymous with R. alba, the

common spearnose skate of the north-eastern Atlantic and Mediterranean.

This species extends southwards from this region, along the western coasts of

Morocco (Pietschmann, 1906), West Africa (Fowler, 1936) and Angola

(Franca, 1959) and is reported in South Africa under the same name. Norman

(1935) pointed out that the snout appears to be slightly longer in European

specimens than in South African specimens, but he was unable to detect any

other important difference.

The claspers of a mature male of R. alba (1 700 mm total length) in the

collection of the Oceanographic Research Institute, Durban, have been

examined (fig. 8). The present author is unable to detect a single difference

between the structure of the clasper of this specimen and the description of

the clasper of European specimens of R. alba (Leigh-Sharpe, 1924, fig. 15),

and therefore holds that R. alba from southern Africa is identical to R. alba

from the northern hemisphere.

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 177

10-0cm

Pe a ea

ap a)

[ 5.0cm

5:0em

Fic. 8. Raja alba.

A: rostral bar and rostral appendices.

B: external view of right clasper from the dorsal side.

C: lateral view of right clasper, opened to show structural features of the glans.

ap—apopyle; cf—cleft; hp—hypopyle; pc—pecten; ‘pr’—‘pseudorhipidion’; sh—shield;

st—sentinel.

178 ANNALS OF THE SOUTH AFRICAN MUSEUM

TABLE 7. R. alba. Measurements expressed as permillage of the total length.

Number of specimens 21.

Character Mean Range

Total length ; ; : . , 3 I 000

Disc width : : ‘ ? : : ota fi 754-804

Disc length : . : ; : 537 511-566

Snout to greatest disc galin : : : 350 300-410

Snout to middle of vent , : ; 403 465-516

Middle of vent to 1st dorsal eee ; : 329 314-359

Snout length ; : : : ; : 155 141-164

Preoral length . k : : ; é 149 129-162

Prenasal length . : ; : ; : 119 103-132

Eye: longitudinal diameter . : 3 : 41 33-47

Eye and spiracle B ; ; . . 57 52-61

Spiracle . : ‘ : 5 : 25 19-35

Interorbital eens: ; : : ; : 54 49-59

Interspiracular distance : } ; ; 70 63-76

Internasal distance . i s , : gl 84-97

Mouth width. : : ; : : go 81-101

Gill slit lengths: 1st. : : : : 18 16-21

ord ; : : : 19 16-23

5th ? : : re 10-14

Distance between inner ends ioe eill slits:

Stole : ; ? : 175 166-186

Bia ; : ; ; 104 96—109

1st dorsal fin: height ; ; : : 33 Pej |

base length . ; : : 60 50-66

2nd dorsal fin: height ; : : : 29 20-35

base length . ; : : 62 56-69

Interdorsal space : : ; : : 7 0-34

Description

Disc about 1:4-1°5 times as broad as long, its width 1-2—1°3 in total length,

with maximum angle in front of spiracles about 105°; anterior margins concave

just behind tip of snout, convex opposite eyes and again concave opposite

spiracles; outer angles pointed, posterior angles broadly rounded. Axis of

greatest breadth 1-3-2-2 times as far from tip of snout as from posterior edge

of disc. Tail short and broad, tapering posteriorly, with lateral folds well

developed on posterior third; its length from middle of vent to first dorsal

origin 1°4-1-°6 in length from middle of vent to tip of snout.

Juvenile specimens with 1 thorn in front of orbit and o-1 thorn behind.

Nuchal and scapular regions and mid-line of back without thorns. 10-16 thorns

along mid-line of tail, from about posterior edge of pelvic to origin of first

dorsal, and o-2 thorns in dorsal interspace. 7-17 lateral caudal thorns in a

single row on each side. Otherwise dorsal surface of disc and tail smooth.

Ventral surface with spinules on tip of snout, internasal region and anterior

margins of disc to about level of mouth; spinules arranged in 3—4 irregular rows.

Adult specimens with a row of thorns about inner margins of orbits. Nuchal

and suprascapular thorns absent. 16-30 thorns along mid-line of tail, from

about posterior edge of pelvics to origin of first dorsal fin, and o—2 thorns in

dorsal interspace. A single row of 17-29 lateral caudal thorns on each side,

irregularly arranged and generally uneven on the two sides. Spinules on dorsal

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 179

surface of disc on snout, rostral cartilage, anterior margins and mid-line of

back. Ventral surface with spinules on snout, internasal region, anterior

margins, abdomen and gill slits. Spinules on both dorsal and ventral surfaces

of tail.

Snout abruptly narrowed and produced into a long, sharp point; its

length in front of orbits 2-5—-3-2 times as long as distance between orbits; its

length in front of mouth 1-4-2-4 times as great as distance between nostrils.

Orbits 1-4—2-4 times as long as spiracles; distance between orbits 1-2—1-6 times

as great as length of orbit. Rostral cartilage projecting from cranium as hard

bar, with rostral appendices fused to bar throughout their length; anterior

rays of pectorals extending half the distance forward from level of front of

orbits towards tip of snout.

Mouth slightly arched; nasal curtain not fringed; expanded posterior

margin of nostril heavily fringed. Teeth arranged in 40-45 regular rows in

upper jaw, with long conical points in middle of jaw, more obtuse with shorter

points laterally.

Anterior lobes of pelvics fin-like and continuously connected with posterior

lobes along outer margin of fin.

Dorsal fins large, similar in shape with broadly rounded apices; height of

first dorsal generally slightly greater than height of second, and base of first

shorter than second; interspace between dorsals distinct, 0-58% as long a

base of first dorsal.

Number of precaudal vertebrae (Vprd) 62-67.

Colour

Juveniles uniformly brownish, darker towards extremities of pectoral,

pelvic and caudal fins. Adults typically grey to brown, with or without scattered

white spots. Ventral surface of disc white, tail and margins of pectorals and

pelvics brownish to black, especially in juveniles.

Raja miraletus Linnaeus, 1758

(Pi iB; Biss 9. .A, B.C)

Raja miraleitus Linnaeus, 1758: 231. Thompson, 1914: 158. Fowler, 1936: 114; 1941: 375. Poll,

1949: 190, fig. 4; 1951: 107, fig. 53. Wallace, 1967: 31, figs 16, 17. Hulley, 1969: 137,

figs 1-3.

Raia miraletus : Von Bonde & Swart, 1923: 5. Barnard, 1925:68 (partim) ; 194.7: 26. Clark, 1926: 9.

Raja ocellifera: Thompson, 1914: 158. Fowler, 1925: 193. Norman, 1935: 42. Fowler, 1941: 375.

Van Bruggen, 1965: 190.

Raia ocellifera Regan, 1906: 2, pl. 2; 1908: 242. Garman, 1913: 365. Gilchrist & Thompson

1916: 286. Von Bonde & Swart, 1923: 5. Barnard, 1925: 67; 1947: 26, pl. 4, fig. 2. Von

Bonde, 1933: 32. Smith, 1961: 66, pl. 3, fig. 69; 1964: 285.

? Raza ocellifera: Samuel, 1963: 99.

Type

The holotype of R. ocellifera, a male (460 mm total length) from Algoa Bay,

and paratype, a female (480 mm total length) from Natal in 73 metres, in the

collection of the British Museum (Natural History).

180 ANNALS OF THE SOUTH AFRICAN MUSEUM

3-0cm j

Fic. 9. Raja miraletus.

A: rostral bar and rostral appendices.

B: external view of right clasper from the dorsal side.

C: lateral view of right clasper, opened to show structural features of the glans. ;

ap—apopyle; cf—cleft; ‘ps’—‘pseudosiphon’ (pocket); pr—pseudorhipidion; se—sentina;

sh—shield; si—signal; st—sentinel.

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 181

Maierial

24 specimens of both sexes (125-510 mm total length), trawled between

Cape Agulhas and Natal in 48-78 metres, and 23 specimens (272-504 mm

total length) taken by R.V. Walther Herwig from 18°30’S to 09°40’S in 100-440

metres. II specimens preserved in the collection of the South African Museum

(SAM 13022, 13032, 15480, 16228, 16347, 21274, 22789). Three specimens

from Angola (Belgian South Atlantic Expedition, Station 15: WNW Banana)

and two specimens from Sierra Leone (Walther Herwig Station 125/1964) also

examined.

Norman (1935) pointed out that R. ocellifera Regan from South Africa is

closely related to R. miraletus Linnaeus from the Mediterranean and the west

coasts of Africa, but considered the two to be distinct because of differences in

tail length, interorbital distance, tooth count and nature of the ocellus. After

comparative morphometric studies on the two species, Wallace (1967: 33)

considers ‘the variation recorded to be within the limits expected of two

geographically separated populations of the same species’, and holds that the

South African specimens should be referred to R. miraletus.

Hulley (1969) agrees with Norman (1935) in that the snout is longer and

the interorbital distance narrower in Mediterranean specimens of R. miraletus

than in South African specimens, but has pointed out that these differences

are less noticeable in West African and Angolan specimens when compared

with South African specimens. It would appear that these slight differences

are of secondary importance in the systematics of this species, as is the small

difference in the structure of the ocellus. Furthermore Hulley (1969) has shown

that while the tail length is comparatively shorter in West African specimens,

the range in number of precaudal vertebrae is the same in the two populations;

also there is no difference in tooth count. Comparative anatomical studies of

the claspers (Hulley, 1969) have confirmed that R. ocellifera is synonymous with

R. miraletus.

The presence of a bluish-black ocellus at the base of each pectoral fin

distinguishes R. miraletus from all other South African skates, except some

juveniles of R. clavaia. However, in R. miraletus the ocellus is circular (or only

very slightly ovate) and consists of three distinct rings of colour, while in R.

clavata the ocellus is markedly horizontally ovate and is usually composed of a

single dark ring enclosing one or two small dark spots.

Description

Disc about 1-3-1-4 times as broad as long, its width 1-4-1-6 in total

length; obtuse in front, with anterior angle in front of spiracles 110°-116°;

anterior margins weakly concave behind tip of snout and gently concave

between level of spiracles and outer corners; outer angles broadly rounded;

posterior margins evenly convex. Axis of greatest breadth o-g—1-3 times as far

from tip of snout as from posterior edge of disc. 'Tail with narrow lateral folds;

its length from middle of vent to origin of first dorsal fin 1-1-1-4 in length from

middle of vent to tip of snout. —

182 ANNALS OF THE SOUTH AFRICAN MUSEUM

TABLE 8. R. miraletus. Measurements expressed as permillage of the total length.

Number of specimens 24.

Character Mean Range

Total length : : : t , , I 000

Disc width . : , : ; , 683 630-724

Disc length ; : , : : : 5II 477-529

Snout to greatest disc width . ; : ‘ 263 237-299

Snout to middle of vent ; : : : 450 420-484

Middle of vent to 1st dorsal origin ; : 350 317-387

Snout length : : : : : : 113 94-136

Preoral length . , Q . : : 105 83-127

Prenasal length . : : : ; ; 84 66-99

Eye: longitudinal diameter . : ; : 35 29-40

Eye and spiracle ; : , 1 ; 51 45-56

Spiracle . ; : i : : ‘ 23 19-26

Interorbital distance . J f ; ; 43 38-48

Interspiracular distance : : ; : 62 51-72

Internasal distance. 4 L ¢ : 80 68-87

Mouth width . ; : ; : ; 82 67-88

Gill slit lengths: rst. ; ; A ‘ 20 17-23

BuGl ; ; : Y 20 18-23

5th ! : : 14 II-17

Distance between inner ends of gill slits:

RSEa : ; : ‘ 143 130-157

1st dorsal fin: height ) 5 : : 24 18-29

base length . ; ; : 56 47-65

end dorsal fin: height : : : 3 24. 19-27

base length . : ; : 59 45-69

Interdorsal space ‘ : : ; ‘ QI 10-28

Juvenile specimens with 2 thorns in front of orbit and 1 behind; o-2

thorns above spiracles; 3 nuchal thorns; 1 scapular thorn on each side; a row

of 18-23 thorns along mid-line of back and tail from scapular region to origin

of first dorsal and o—2 thorns in dorsal interspace. Sometimes a few small

spines situated semi-laterally on tail, otherwise dorsal and ventral surfaces

smooth.

Older specimens show an increase in the number of thorns around orbits

and on tail, but a reduction (to complete loss) of thorns along mid-line of

back and on scapular region. Adults typically with 4-8 thorns around inner

margins of orbits; o-2 median nuchal thorns; a row of 12-27 thorns along

mid-line of tail, from above axils of pelvics to first dorsal origin, and 1-2

thorns in dorsal interspace. Lateral rows of thorns on tail in 1-2 rows, generally

unequal in number on the two sides and ranging from 8-16 thorns in semi-

lateral rows and 0-13 in lateral rows. Dorsal surface with spinules on snout,

anterior margins to level of spiracles, mid-line of back, and some specimens

with a small group of stellate-based spines on posterior pectoral angle. Ventral

surface with spines on snout, internasal region and anterior margins of disc

to about level of nostrils.

Snout obtuse and not produced; its length in front of orbits 2-3~-3-1 times

as long as distance between orbits; its length in front of mouth 1-1-1-5 times

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 183

as great as distance between nostrils. Orbits 1-2-1-8 times as long as spiracles;

distance between orbits 1-1-1-4 times as great as length of orbit. Rostral

cartilage projecting from cranium as hard bar, with rostral appendices fused

to bar throughout their length; anterior rays of pectorals extending about

two thirds the distance forward from level of front of orbits towards tip of

snout.

Mouth slightly arched; nasal curtain fringed; expanded posterior margin

of nostril heavily fringed. Teeth arranged in 42—50 rows in upper jaw, usually

with a single median point in middle of jaw, more obtuse with shorter points

laterally.

Anterior lobes of pelvics fin-like and continuously connected with posterior

lobes along outer margin of fin.

Dorsal fins similar in shape and about equal in size; interspace between

dorsals usually small but distinct, up to 66% as long as base of first dorsal.

Number of precaudal vertebrae (Vprd) 47-52. Krefft (19682) gives the

following count for R. miraletus: Vtr 25-29; Vprd 44-52; V2' 70-81.

Colour

Brownish, with or without numerous small darker spots; a large blue-black

ocellus at the base of each pectoral, consisting of three definite rings of

colour. Ventral surface pale, with or without a small dark spot on tip of

snout.

Raja clavata Linnaeus, 1758

(Pl. 6A; Figs 10 A, B, C)

Raja clavata Linnaeus, 1758: 232. Gilchrist, 1922: 7. Fowler, 1936: 110; 1941: 360. Poll, 1949:

188, fig. 3. Hulley, 1966: 497, figs 1-3. Wallace, 1967: 35, figs 18, 19.

Raia clavata: Barnard, 1925: 64, pl. 4, fig. 2; 1947: 26, pl. 4, fig. 1. Von Bonde, 1933: 32.

Raja rhizacanthus: Thompson, 1914: 158. Norman, 1935: 40. Smith & Smith, 1966: 209, fig.

Rata rhizacanthus Regan, 1906: 3, pl. 3; 1908: 242. Gilchrist & Thompson, 1916: 286. Von

Bonde & Swart, 1923: 5. Smith, 1961: 67, pl. 3, fig. 71.

Raja capensis Miller & Henle, 1841: 151. Gray, 1851: 112. Bleeker, 1860: 58. Dumeéril, 1865:

540, pl. 12, figs 11, 12. Kner, 1865: 419. Gilchrist, 1902: 168. Thompson, 1914: 157.

Gilchrist, 1921: 27.

Raia capensis: Sauvage, 1891: 510. Von Bonde & Swart, 1923: 4.

Raia barnardi: Smith, 1961: 67, pl. 3, fig. 70.

? Raja maculata: Bleeker, 1860: 58. Gilchrist, 1902: 168. Thompson, 1914: 157.

? Raia maculata: Barnard, 1925: 71. Von Bonde, 1933: 32.

? Raia oculata: Von Bonde & Swart, 1923: 4.

Types

The holotype of R. rhizacanthus, a juvenile male (210 mm total length),

from the coast of Natal in 73 metres, in the collection of the British Museum

(Natural History).

The paratype of R. capensis in the collection of the Muséum National

d’Histoire Naturelle, Paris.

184 ANNALS OF THE SOUTH AFRICAN MUSEUM

L ‘Ocm

10-0cm

Fic. 10. Raja clavata.

A: rostral bar and rostral appendices.

B: external view of right clasper from the dorsal side.

C: lateral view of right clasper, opened to show structural features of the glans.

ap—apopyle; cf—cleft; fn—funnel; hp—hypopyle; pr—pseudorhipidion; ‘ps’—‘pseudosiphon’

(pocket) ; se—sentina; si—signal; sh—shield; sp—spike; st—sentinel.

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 185

Material

41 specimens of both sexes (171-790 mm total length), trawled between

Cape Columbine and Algoa Bay in 55-548 metres. 30 specimens preserved in

the collection of the South African Museum (SAM 22786-8, 23006-8, 24324-8,

24330-5, 24337-8, 24355-60, 24534-40).

Although the South African species R. rhizacanthus Regan was previously

thought to be distinct from the European ‘thorn-back’ skate R. clavata Linnaeus

because of differences in the spination and in the position of the vent (Norman,

1935), several authors (Barnard, 1925; Fowler, 1936, 1941) considered the

two to be synonymous. Comparative anatomical studies of the claspers (Hulley,

1966) and detailed morphometric measurements (Wallace, 1967) have con-

firmed that the South African specimens should be referred to R. clavata.

R. herwigi, R. maderensis, R. siraelent and R. clavata form a definite sub-group

of the genus Raja but at present the systematics of this complex of species,is

uncertain. R. clavata closely resembles R. straelent in proportional dimensions,

tooth count and number of vertebrae, but differs from this species in its intensity

and pattern of coloration, the degree of development of the shield in the

clasper glans and in its distribution. In the South Atlantic, R. clavata appears

to have a northern limit at 26° and is replaced by R. straeleni over the area

22°5 to 0°. |

Adult specimens can easily be recognized by the spinulose nature of the

dorsal surface of the disc and by the presence of enlarged buckler-like thorns,

especially in females (these thorns are absent in R. straeleni). Adults resemble

B. smithi in shape, but may be distinguished from this species by the number

of rows of teeth in the upper jaw, the structure of the rostral cartilage and rostral

appendices, and by the presence in R. clavata of lateral rows of thorns on the

tail. It should be noted that the dental formula for R. clavata given by Wallace

(1967) as 23—26 rows in the upper jaw, is inconsistent with previous investiga-

tions: Regan (1906) reported 36 rows in the type, Barnard (1925) gave 38-44

and Norman (1935) 36-44 rows. In the specimens examined by the author,

the range was found to be 36—45, which corresponds with the 36-44 rows in

European specimens (Clark, 1926). Juvenile specimens of R. clavata are generally

smooth (sometimes with spinules along the anterior margins of the disc) and

lack a lateral series of thorns on the tail. These facts, together with the lack of

a produced and pointed snout and a tooth count of more than 30 rows in the

upper jaw, separate this species from all other known South African skates,

except R. miraletus. However, R. miraletus possesses ocelli on the dorsal surface

of the disc.

Some juveniles of R. clavata possess a pair of ocelli at the bases of the

pectoral fins, which may lead to confusion with R. miraletus. In R. miraletus,

however, the ocellus is circular (or only very slightly ovate) and consists of

three definite rings of colour, while in R. clavata, the ocellus is markedly hori-

zontally ovate and is usually composed of a single dark ring, within which are

two (sometimes one) darker spots.

186 f ANNALS OF THE SOUTH AFRICAN MUSEUM

TABLE g. R. clavata. Measurements expressed as permillage of the total length.

Number of specimens 41.

Character Mean Range

Total length : : ; . : : I 000

Disc width : : : 4 5 : 731 679-773

Disc length : j : : 3 ‘ 529 481-569

Snout to greatest disc width . 4 : ; 308 280-336

Snout to middle of vent ‘ : . : 471 432-517

Middle of vent to 1st dorsal origin. y : 349 317-390

Snout length : , ‘ ‘ : : 133 120-155

Preoral length . ; : . ; : 123 106-142

Prenasal length . : : : : : 102 87-119

Eye: longitudinal diameter . ; : : 35 23-41

Eye and spiracle P : , : ; 49 45-56

Spiracle . ; ‘ : ¥ ; : 23 1O-27

Interorbital distance . : : ; ; 47 41-52

Interspiracular distance : : : : 70 65-75

, Internasal distance. : : ‘ : 84 76-92

Mouth width. : ‘ : ; : 82 77-91

Gill slit lengths: 1st. , : ; : 20

andy ie 4 ; : d 21

Fthy : 5 . P 14

Distance between inner ends of gill slits:

ESC) (us : ; : : 159 135-177

5th , : ; . 83 73-97

1st dorsal fin: height 5 : ; ; 29 19-39

base length . ; ; ‘ 58 45-71

2nd dorsal fin: height ; ; : : 30 21-37

base length . : q : 60 45-76

Interdorsal space : , F ; ; 19 0-35

Description

Disc about 1-3—1-4 times as broad as long, its width 1-3-1°5 in total length;

obtuse in front, with anterior angle in front of spiracles 80°-100°; anterior

margins weakly concave behind tip of snout and again at level of spiracles,

more deeply so in males than in females; outer and posterior angles broadly

rounded. Axis of greatest breadth 1-2—1-8 times as far from tip of snout as from

posterior edge of disc. Tail with moderately developed lateral folds on posterior

third; its length from middle of vent to origin of first dorsal fin 1-2-1-6 in

length from middle of vent to tip of snout.

Juvenile specimens with 2 thorns in front of orbits and 1 behind: 1 supra-

spiracular thorn and 1 pair of thorns between spiracles; 2-4 median nuchal

thorns; 2 scapular thorns on each side; a row of 24-30 thorns along mid-line

of back and tail from scapular region to origin of first dorsal fin; 1-2 thdrns

in dorsal interspace. No lateral rows of thorns on tail. Dorsal surface with

spinules on snout, interorbit, anterior margins of disc and along mid-dorsal

region of back and tail. Ventral surface smooth.

Older specimens show a reduction in number of preorbital, postorbital,

supraspiracular, interspiracular and scapular thorns, until these disappear in

the adult. Adults typically with no thorns around inner margins of orbits and

above spiracles, no nuchal spines, no scapular thorns and no enlarged thorns

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 187

along the mid-line of back; a row of 19-41 thorns along mid-line of tail from

above vent to origin of first dorsal fin; 1-2 thorns in dorsal interspace. A row

of lateral caudal thorns on each side (sometimes two), irregularly arranged.

Widely-spaced spinules on dorsal surface of snout, interorbit and mid-dorsal

region of back and tail. Ventral surface spinulose along anterior margins of

disc and around vent. Buckler-like thorns on dorsal and ventral surfaces of

disc, especially around vent, in some specimens, particularly females.

Snout pointed but not produced; its length in front of orbits 2-6—3-4 times

as long as distance between orbits; its length in front of mouth 1-3-1-7 times

as great as distance between nostrils. Orbits 1-3—2-0 times as long as spiracles;

distance between orbits 1-I1-1-4 times as great as length of orbit. Rostral

cartilage projecting from cranium as hard bar, with rostral appendices fused

to bar throughout their length; tips of pectoral fins falling short of appendices.

Mouth slightly arched; nasal curtain slightly fringed; expanded posterior

margin of nostril heavily fringed. Teeth arranged in 36-45 rows in upper jaw,

with large cusp in sexually mature males, but blunt and flat in juveniles and

females.

Anterior lobes of pelvics fin-like and continuously connected with posterior

lobes along outer margin of fin.

Dorsal fins similar in shape, with broadly rounded apices; first dorsal

usually slightly smaller than second; interspace between dorsals usually about

30% as long as base of first dorsal, but varying between 0% and 66%.

Number of precaudal vertebrae (Vprd) 45-53.

Colour

Colour variable, particularly in juveniles. Brownish or grey-brown, with

or without numerous irregular darker and lighter spots or with larger irregular

darker blotches. Juveniles generally with a dark, horizontally ovate ocellus at

the base of each pectoral, consisting of a single dark ring enclosing two (some-

times one) dark spots. Ventral surface pale, margins of pectorals generally

greyish, occasionally with some darker irregular patches.

Raja straelent Poll, 1951

(PG B: Figs 11 A, B,C)

Raja straelent Poll, 1951: 118, fig. 54. Krefft, 1968a: 66, pl. Va.

Types

The holotype, a mature male (620 mm total length), trawled in 100-110

metres at 13°05’S, 12°46’E, in the collection of the Institut Royal des Sciences

Naturelles, Brussels. 21 paratypes in the same collection.

Material

The holotype and paratypes, as well as 8 specimens trawled in 200-700

metres by R.V. Walther Herwig between 10°28’S and 22°03’S. Measurements

taken on 16 specimens including the holotype. Material from the R.V. Walther

Herwig located in the collection of the Institut fiir Seefischerei, Hamburg.

188 ANNALS OF THE SOUTH AFRICAN MUSEUM

R. straeleni forms part of the clavata-complex of species. The systematics of

this complex, which includes the species R. clavata, R. herwigi, R. maderensis and

R. straeleni is as yet uncertain. In the South Atlantic, R. straeleni has been found

in the region between the equator and about 22°S and is replaced by R. clavata

to the south of this limit.

‘ps

Mis

; A

Fic. 11. Raja straelen.

A: rostral bar and rostral appendices.

B: external view of right clasper from the dorsal side.

C: lateral view of right clasper, opened to show structural features of the glans.

ap—apopyle; cf—cleft; fn—funnel; hp—hypopyle; pr—pseudorhipidion; ‘ps’—‘pseudosiphon’

(pocket) ; se—sentina; sh—shield; si—signal; sp—spike; st—sentinel.

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 189

R. straelent very closely resembles R. clavata, and while there are noticeable

differences in colour pattern and intensity, which appear to be constant in

R. straeleni, there is only a slight difference in the external morphology of the

claspers of the two species — the shield in R. straeleni is more strongly developed,

suggesting a difference in the degree of development of the ventral terminal

cartilage. Morphometric proportions and vertebral counts are similar in the

two species.

On the basis of colour pattern, the taxonomic significance of which is at

present unknown, distribution and the structure of the shield, R. straeleni has

been retained for the moment as a separate species. Detailed investigations of

the skull and the nature of the clasper cartilages should elucidate this problem.

TABLE 10. R. straeleni. Measurements expressed as permillage of the total length.

Number of specimens 16.

Character Mean Range

Total length ; : ; é : I 000

Disc width 2 : : 3 : : 672 614-729

Disc length ; : : b é : 503 471-525

Snout to greatest disc width . : : : B29 2902-357

Snout to middle of vent ; : : : 465 439-500

Middle of vent to 1st dorsal origin. : : 343 326-383

Snout length : : : F : : 132 124-144

Preoral length . : : : : : 125 112-135

Prenasal length . ; ‘ , é é 102 93-109

Eye: longitudinal diameter . : ; : 40 33-49

Eye and spiracle ; i é : ; 55 48-60

Interorbital distance . : : : : 44 40-48

Interspiracular distance : ‘ : : 69 63-72

Internasal distance. Z : : : 88 84-92

Mouth width. ; : : i : 86 81-94

Gill slit lengths: 1st’. : P ‘ ‘ 21 17-28

ard x, ‘ : : : 23 17-28

Rea. . ‘ 3 5 16 14-22

Distance between inner ends of gill slits:

ESE( : : , , 163 150-180

1st dorsal fin: height : : : j 29 22-34.

base length . : : : 61 54-70

and dorsal fin: height ; : : , 27 23-32

base length . . 5 : 60 55-70

Interdorsal space : : : ‘ : 34 27-56

Description

Disc about 1-2—1°5 times as broad as long, its width 1-4—1-7 in total length;

obtuse in front, with anterior angle in front of spiracles about 95°—100°; anterior

margins weakly concave behind tip of snout and again at level of spiracles;

outer angles rounded, posterior margins evenly convex. Axis of greatest breadth

1-5—2-0 times as far from tip of snout as from posterior edge of disc. Tail with

lateral folds on posterior half; its length from middle of vent to origin of first

dorsal fin 1-2-1-5 in length from middle of vent to tip of snout.

1-2 thorns in front of orbits and 1-3 behind; o—1 thorn above each spiracle;

190 ANNALS OF THE SOUTH AFRICAN MUSEUM

4-5 median nuchal thorns; o—2 scapular thorns; a row of 20-52 thorns along

mid-line of back and tail, from about axis of greatest breadth to first dorsal

origin; o—6 thorns in dorsal interspace. A few lateral thorns on anterior region

of tail, becoming regularly arranged and hook-like posteriorly. Dorsal surface

in males entirely covered with small spines, but in females spinules only on

tip of snout, mid-dorsal region of back and tail, and anterior margins of disc.

Ventral surface with spines on tip of snout and along anterior margins of disc;

otherwise smooth.

Snout pointed but not produced; its length in front of orbits 2-8—3-2

times as long as distance between orbits; its length in front of mouth 1-3-1°5

times as great as distance between nostrils. Distance between orbits 0-8-1-4

times as great as length of orbit.

Mouth slightly arched; nasal curtain fringed; expanded posterior margin

of nostril heavily fringed. Teeth arranged in 36-45 (35-42: Poll, 1951) rows

in upper jaw, blunt and flat in smaller specimens, but sharp-pointed in adult

males.

Anterior lobes of pelvics fin-like and continuously connected with posterior

lobes along outer margin of fin.

Dorsal fins similar in shape with broadly rounded apices; first dorsal

usually smaller than second; interspace between dorsals 17°%-104% as long

as base of first dorsal.

Vertebral count: Vtr 24-28; Vprd 48-52; V2 73-78 (Krefft, 19682).

Colour

Dorsal surface brown to grey with numerous dark spots on whole of disc

except snout, pelvic fins and tail; spots regularly arranged. Ventral surface of

disc pale or mottled grey, sometimes with a darker margin along posterior

edges of pectorals and pelvics. Krefft (1968a) reports that in juveniles there is

no evidence of darker spots on the dorsal surface.

Raja robertsi n.sp.

(PL8. A; B; Figs 12¥AeB)

Type

A juvenile male (773 mm total length), trawled west of Cape Town by

R.V. Walther Herwig (Station No. WH 196/67: 33°51’S, 17°14’E) in 1350

metres, in the collection of the Institut fir Seefischerei, Hamburg.

Material

A single specimen, the holotype.

This species forms part of the radiata-group, the distribution and interrela-

tionships of which have been discussed by Krefft (1968). The group is charac-

terized by a lozenge-shaped disc, short tail and by the presence of a single mid-

dorsal row of stout spines, extending from the nuchal region (in some species

from the axils of the pelvics) to the origin of the first dorsal fin. Preliminary

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA IQ!

examination of the claspers of this group of species has revealed the presence

of a pseudosiphon in the outer dorsal wall of the clasper glans. Although the

type of R. robertsi is a juvenile male (clasper length 61 mm), the existence of

the pseudosiphon can already be detected in the still undeveloped clasper.

In the Atlantic, at least, the radiata-group appears to be divisible into

two sub-groups: the heavily-spined species, comprising the North-South

Atlantic pairs R. radiata—R. doellojuradoi and R. hyperborea—R. frerichsi: and the

sub-group comprising the North Atlantic species R. jensent and the South

Atlantic species R. georgiana, in which there are no enlarged stellate-based

thorns on the anterior margins and posterior angles of the pectorals. R. roberts:

is a member of this second sub-group.

Although it resembles R. georgiana in tooth count and spination of the

orbit, R. roberist can easily be separated from this species by the characteristic

dark coloration of its ventral surface. It most closely approximates to the North

Atlantic species R. jenseni in lacking large rostral thorns and in ventral colora-

tion, but differs markedly from this species in tooth count (56-66 rows in

upper jaw in R. jenseni (Bigelow & Schroeder, 1953): 42 rows in R. robertsz),

preoral length and scapular spination. Furthermore, there are fewer thorns in

the mid-line of the disc and tail in R. roberisz. |

3-0cm

Fic. 12. Raja robertsi.

A: rostral bar and rostral appendices.

B: external view of right clasper from the dorsal side.

ap—apopyle; “ps’—‘pseudosiphon’.

In tooth count and general coloration, R. robertsi falls close to R. badia

caught in 1270 fathoms in the Gulf of Panama (Garman, 1899). However,

R. badia belongs to the heavily spined sub-group (Berg, 1911 considers it to be

a possible synonym of R. hyperborea (Bigelow & Schroeder, 1953)) and therefore

differs in the spination of the disc.

This species has been named in honour of Mr. J. Douglas Roberts, who

by his kind generosity, made the study of the Walther Herwig material in

Hamburg possible.

192 ANNALS OF THE SOUTH AFRICAN MUSEUM ~*

TABLE 11. R. robertsi. Measurements expressed as permillage of the total length.

A single specimen, the type (773 mm total length).

Character Dimensions

Total length : ; A : , : : ; I 000

Disc width : : 3 : : : . : 729

Disc length , : : : A ; : 569

Snout to greatest disc width . : ‘ : ; : 428

Snout to middle of vent ‘ : : p : : 581

Middle of vent to 1st dorsal origin. , : : : 311

Snout length : : ; : : : F E 145

Preoral length . : : 4 : : : , 142

Prenasal length . : : s : : ; : 108

Eye: longitudinal diameter . : : : : ‘ 44

Eye and spiracle : ; : . : : 63

Interorbital distance . : ; ; , : 68

Interspiracular distance : ; , atone ‘ 107

Internasal distance. i d i : é ? 121

Mouth width. : : : 3 : ; ' 131

Gill slit lengths: 1st. ; ; : ‘ ; ‘ I5

Bra) J. : : : : : : 14

5th : : 2 : ‘ F II

Distance between inner ends of gill slits:

LES Bev 4 : , : : ; 215

5th ; ; , ; : ; 155

1st dorsal fin: height : : ; : ; iy

base length . : : : , : 45

2nd dorsal fin: height i : é : 15

base length . ; ; , ; 40

Interdorsal space : : : ih a : : 0)

Description

Disc 1-3 times as broad as long, its width 1-4 in total length; obtuse in

front, with anterior angle in front of spiracles 96°: anterior margins concave

just behind tip of snout and more deeply concave just posterior to level of

spiracles; outer angles somewhat pointed, posterior margins almost straight.

Axis of greatest breadth 3-4 times as far from tip of snout as from posterior

edge of disc. ‘Tail with fairly well developed lateral folds along its entire length;

its length from middle of vent to origin of first dorsal fin 1-9 in length from

middle of vent to tip of snout; its length from middle of vent to tip of tail 1-4

in length from middle of vent to tip of snout.

Spination:

Preorbital thorns . 1/1

Postorbital thorns . 1/1

Supraspiracular thorns . 1/1

Median nuchal thorns . 2

Scapularthorns yo ei eect ene

A single row of 21 (+ 1 scar) stout thorns extending from nuchal region

almost to origin of first dorsal. No lateral rows of thorns on disc or tail. 1 small

stellate-based spine on rostrum, no thorns on anterior margins and posterior

angles of pectorals; entire upper surface of disc and tail covered with close-set

spinules. Malar spines present. Ventral surface of disc and tail smooth.

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 193

Snout obtuse and not produced; its length in front of orbits 2-4 times as

long as distance between orbits; its length in front of mouth 1-2 times as great

as distance between nostrils. Distance between orbits 1-5 times as great as

length of orbit. Rostral cartilage extending from cranium as hard bar, with

rostral appendices fused to bar throughout their length; anterior rays of

pectorals extending half the distance forward from level of front of orbits

towards tip of snout.

Mouth slightly arched; nasal curtain not fringed; expanded posterior

margin of nostril fringed; no barbel-like lobe on inner edge of each nostril,

but a slight fleshy fold present. Teeth arranged in 42 regular rows in upper

jaw, with round bases and large, sharp posterior cusps.

Anterior lobes of pelvics fin-like and continuously connected with posterior

lobes along outer margin of fin.

Dorsals more or less similar in size with convex anterior margins and

rounded apices; first dorsal continuous with second.

Vertebral count: Vtr 32, Vprd 55, V2 87.

Colour

Dorsal surface of disc greyish with a few scattered darker patches over

disc and tail; pelvics somewhat darker. Ventral surface of disc uniformly grey

to black, tending to be darker at posterior margins of pectorals, tips of anterior

lobes of pelvics and on tail. A white inverted-triangular patch in median position,

with base line on level of pectoral girdle. Lower jaw and area about nostrils

white.

Clasper structure

The type is an immature male, in which definite structures in the clasper

glans cannot be identified. However, the claspers are of the short, spatulate

variety and a pseudosiphon can be distinguished in the outer dorsal surface of

the glans (fig. 12).

Raja radiata Donovan, 1807

(Pi JPA; Hig. 19)

Raja radiata Donovan, 1807: pl. 114.

Material

‘Two specimens, a male (573 mm total length) and a female (616 mm total

length), trawled west of Cape Town in 548-640 metres. Both specimens pre-

served in the collection of the South African Museum.

This species represents a new record for the eastern South Atlantic. R.

radiata is retained in the genus Raja because of the structure of the snout and

a precaudal vertebral count of 58-62. It can easily be distinguished from all

other South African species, except R. robertsi, by its extremely short tail, and

differs from R. robertst in the heavy spination of the disc and tail.

R. radiata forms part of the radiata-group of species, which are charac-

194 ANNALS OF THE SOUTH AFRICAN MUSEUM

terized by their lozenge-shaped disc, short tail and single mid-dorsal row of

stout thorns. The distribution and interrelationships of this group have been

discussed by Krefft (19680). |

In appearance and type of spination, R. radiata falls into the heavily-

spined sub-group of the radiata-complex. R. radiata is so closely allied to R.

doellojuradot, recorded from the western South Atlantic, that Bigelow &

Schroeder (1953: 255) point out that ‘no reliable criteria have been found to

distinguish the one from the other’. Pozzi (1935, 1936) and Norman (1937)

hold that in R. radiata the tail is longer, so that the vent is nearer to the tip of

the snout than to the end of the tail; the teeth are more numerous (38-46

rows in upper jaw in R. radiata; 31-34 rows in R. doellojuradoz); there are only

2 scapular thorns; and the spines on the disc are less well developed, there

being less than ro in the mid-line of the tail posterior to the axils of the pelvics.

Krefft (personal communication), in comparing six specimens of R. radiata

from the northern Atlantic with five specimens of R. doellojurado: from the

western South Atlantic, has found that there is no difference in the number

and distribution of the median thorns and no difference in spination except

in the number of scapular thorns. Whereas in R. radiata two large scapular

thorns exist, the third one being either vestigial or much smaller than the

outer ones, in R. doellojuradoi the three scapular thorns are of the same size.

In the South African specimens, tooth count favours their identity with

R. radiata, and while size at first maturity may be a questionable taxonomic

character, the size of the two specimens is well above the 530 mm total length

limit for R. doellojuradoi (Krefft, personal communication). Furthermore the

rajid distribution of the Atlantic suggests that the South African fauna, which

r 5-O0Ocm

Fic. 13. Raja radiata.

Rostral bar and rostral appendices.

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 195

contains a number of European species (Hulley, 1966; 1969) is more closely

related to that of the North Atlantic than to that of South America, i.e. a distri-

bution following the continental slopes from north to south seems more feasible

than an east—west distribution, which presupposes a crossing of the deep-sea

plane or a very old ‘Gondwanaland’ distribution. This seems all the more likely

since the generic and specific composition of the skate fauna of the eastern

South Atlantic differs markedly from that of the western Seuth Atlantic.

TABLE 12. R. radiata. Measurements expressed as permillage of the total length.

Number of specimens 2.

Character Male Female

Total length : : ‘ : : : . I 000 I 000

Disc width : ; ; ‘ : on GO4: Tif

Disc length : j : ; : 5 3 BOR 566

Snout to greatest disc leans : : 7 IN AOS 360

Snout to middle of vent ; : ; ROT 575

Middle of vent to 1st dorsal wees. : 3 Sy ital 326

Snout length ; : : : : : ae GO 149

Preoral length . : : : : ; se ED 148

Prenasal length . : : : ; § i (LG 112

Eye: longitudinal diameter . ; : : P 42 33

Eye and spiracle : : ; : : : 60 54

Spiracle . 3 é 3 : : ; 30 27

Interorbital Histaince 3 : : ; : : 65 WZ

Interspiracular distance ‘ ; : é <b LEOR 101

Internasal . ‘ : , : : : A 117 III

Mouth width . ‘ A : : : te aLIS 112

Gill slit lengths: 1st. ; F - : ; 13 12

Chit 0 ee : : : : 3 17 16

5th : . : 5 : 9 9

Distance between inner ends of gill slits:

TSEMA: : : : : R21 205

Be . z : : ey nara: 142

1st dorsal fin: height 2 : . ; : 23 QI

base length . : : ; : 40 36

and dorsal fin: height : ! : : : 20 19

base length . : : ; : 43 41

Interdorsal space é : : 5 : : 6 8

Description

Disc about 1-2—1°3 times as broad as long, its width 1-3 in total length;

obtuse in front, with anterior angle in front of spiracles about 100°; anterior

margins slightly concave close behind tip of snout, otherwise almost straight;

posterior angles much more broadly rounded than outer angles. Axis of greatest

breadth 1-7—2-1 times as far from tip of snout as from posterior edge of disc.

Tail with well-developed lateral folds beginning at axils of pelvics; its length

from middle of vent to origin of first dorsal fin 1-8 —1-9 in length from middle

of vent to tip of snout; its length from middle of vent to tip of tail 1-4-1°5 in

length from middle of vent to tip of snout.

Upper surface of disc with a median row of 18 large, conspicuous thorns

on radiate bases, extending from nuchal region to origin of first dorsal; 5-6 of

these anterior to axils of pelvics, and 13-14 laterally compressed thorns from

196 ANNALS OF THE SOUTH AFRICAN MUSEUM

axils of pelvics to origin of first dorsal, decreasing in size posteriorly. Mid-line

of back with 2-3 irregular rows of smaller thorns on stellate bases on either

side of the mid-dorsal row, extending almost from the nuchal region to first

dorsal origin, becoming sharper and recurved posteriorly ; small spines in dorsa]

interspace. Anterior, median and posterior areas of pectorals with scattered

thorns and smaller spines; thorns present on rostral projection, interorbit and

interspiracular regions and on sides of tail; naked areas confined to anterior

parts of pelvics. 1 large thorn in front of orbit and 1 behind; 1 thorn directly

above each spiracle, close to postorbital thorn; 2 nuchal thorns in mid-line;

2-3 scapular thorns, the third vestigial. Ventral surface of disc and tail com-

pletely smooth.

Snout obtuse and not produced; its length in front of orbits 2:1-2:4 times

as long as distance between orbits; its length in front of mouth 1:3 times as

great as distance between nostrils. Orbits 1-2—-1-4 times as long as spiracles;

distance between orbits 1:6—-2-2 times as great as length of orbit. Rostral

cartilage projecting from cranium as hard bar, with rostral appendices fused

to bar throughout their length; anterior rays of pectorals extending two-thirds

the distance forward from level of front of orbits towards tip of snout.

Mouth slightly arched, more so in the male than in the female; nasal

curtain not fringed; expanded posterior margin of nostril heavily fringed.

Teeth arranged in 37-39 regular rows in upper jaw, with round bases and large,

sharp posterior cusp.

Anterior lobes of pelvics fin-like and continuously connected with posterior

lobes along outer margin of fin. .

Dorsals similar in shape and about similar in size, with convex anterior

margins and broadly rounded apices; interspace between dorsals 14-23% as

long as base of first dorsal.

Number of precaudal vertebrae (Vprd) 58-62.

Colour

Upper surface brownish-grey, with scattered irregular black blotches,

especially on anterior limb of pelvic fin. Ventral surface white, with darker

spots and blotches on tail and pelvics, and black spot on tip of tail.

Raja ravidula nu. sp.

(Pligg AyBe Pigs)

Types

The holotype, a juvenile male (634°5 mm total length), trawled west of

Cape Town by R.V. Walther Herwig (Station No. WH _ 195/76: 33°490'S,

17°13’E) in 1000 metres, in the collection of the Institut ftir Seefischerei,

Hamburg. The paratypes, a juvenile male (605 mm total length) and a juvenile

female (631°5 mm total length), trawled west of Cape Town by R.V. Walther

Herwig (Station No. WH 194/67: 33°47’S, 17°14’E) in 1 000 metres, in the

same collection.

—

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 197

Material

The holotype and paratypes.

This species of deep-water skate differs from all known South African

species with less than 50 rows of teeth in the upper jaw in that a median row

of thorns along the back and tail is absent. In lacking a median row of thorns

it approximates to the European species R. fullonica, which has been recorded

as far south as Morocco (Murray & Hort, 1912). However, it may be dis-

tinguished from this species by its lower tooth count (39-44 rows in upper jaw

in R. ravidula; 58-68 rows in R. fullonica), lack of thorns on the tip of the rostrum

and by the number and arrangement of the nuchal spines.

TABLE 13. R. ravidula. Measurements expressed as permillage of the total length.

Character Type Paratype Paratype

3 3 2

Total length ; 4 : . : I 000 I 000 I 000

Disc width : ‘ : : : 561 578 570

Disc length ; : : : : 489 505 482

Snout to greatest disc width . ! : 291 314 312

Snout to middle of vent : : : 491 496 487

Middle of vent to 1st dorsal origin. é 384. 391 4.03

Snout length ; ; ; : ‘ 138 142 148

Preoral length . ; : ! ; 149 147 155

Prenasal length . : : : y 126 126 133

Eye: longitudinal diameter . ; : 38 38 38

Eye and spiracle ‘ ; : : 50 50 49

Interorbital distance . : : ; 32 30 30

Interspiracular distance i ‘ : 72 71 71

Internasal distance : : ‘ ; 82 84 84

Mouth width : ; : : FP 83 7

Gill slit lengths: st. : ; : 14 15 13

Sra)“ f ss : 15 18 14

Btls. : : : II 12 II

Distance between inner ends of gill slits:

WSEnI : : : 132 131 138

Bey pie : 4 3 go 87 QI

1st dorsal fin: height : j ‘ 29 oT 33

base length . : ; 48 50 43

and dorsal fin: height ‘ : 4 27 25 34.

base length . : : 48 45 50

Interdorsal space : ; sents 8 oO O

Description

Disc about 1-1—1-2 times as broad as long, its width 1-7—1°8 in total length;

maximum anterior angle in front of spiracles about 92°; anterior margins

weakly concave behind tip of snout and again at level of spiracles; outer and

posterior angles broadly rounded, posterior margins evenly convex. Axis of

greatest breadth 1-5-1-8 times as far from tip of snout as from posterior edge

of disc. Tail with moderately wide lateral folds on posterior third; its length

from middle of vent to origin of first dorsal fin 1-2-1-3 in length from middle

of vent to tip of snout.

198 ANNALS OF THE SOUTH AFRICAN MUSEUM

No thorns along mid-line of back and tail (paratype with 1 small thorn

in mid-line at about half tail length). A single series of thorns on each side of

mid-line, extending from nuchal region to first dorsal origin, widely spaced on

back, but, close-set and larger on tail. No thorns on rostral cartilage or between

dorsal fins. Entire dorsal surface of disc covered with blunt, flat, widely-spaced

asperities, tail with larger and more pointed spinules. Ventral surface of disc

and tail smooth.

Snout pointed and a little produced; its length in front of orbits 4-3-4:8

times as long as distance between orbits; its length in front of mouth 1-8 times

as great as distance between nostrils. Distance between orbits 1-2-1°3 in length

of orbit. Rostral cartilage projecting from cranium as hard bar, with rostral

appendices fused to bar throughout their length; tips of pectoral rays falling

short of appendices.

3-0 cm

Fic. 14. Raja ravidula.

Rostral bar and rostral appendices.

Spination Holotype 3 Paratype 3 Paratype 2

Preorbital thorns. : : , 3/3 4/4. 3/3

Postorbital thorns ; : ! 3/3 3/3 3/3

Supraspiracular thorns. A : 1/1 1/1 1/1

Interspiracular thorns . ; : 1/o o/o o/1

Median nuchal thorns , i 6 6 6

Lateral nuchal thorns . ; : 4/3+2 4/4 2+3/2+2

Scapular thorns . . : ; 2/2 o/t 3/2

Mouth slightly arched; nasal curtain slightly fringed; expanded posterior

margin of nostril heavily fringed. Teeth arranged in 39-44 rows in upper jaw,

blunt and flat with small posterior cusp. |

Anterior lobes of pelvics fin-like and continuously connected to posterior

lobes along outer margin of fin.

Dorsals similar in shape, with broadly rounded apices; first dorsal usually

slightly larger than second; interspace between dorsals 0°4-16:5% as long as

base of first dorsal.

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 199

Vertebral count Holotype § Paratype 3 Paratype 2

Vir. : ‘ : as : 33 30 30

Vprd. : : ; ; : 69 69 70

Ree. 3 : ; P : 102 99 100

Colour

Dorsal surface of disc pale grey, becoming darker along posterior margins

of pectorals and pelvics and at distal end of tail. Ventral surface of disc white,

but darker at margins of pectorals; pelvics white with dark blotches at axils,

on tip of anterior lobe and at anterior end of vent; tail greyish-brown becoming

mottled with white posteriorly.

Raja dissimilis n.sp.

(Pl. 10 A, B; Figs 15 A, B, C)

Types

The holotype, a mature male (640 mm total length), trawled west of

Cape Town by R.V. Walther Herwig (Station No. WH 194/67: 33°47'S, 17°14’E)

in I 000 metres, in the collection of the Institut fiir Seefischerei, Hamburg.

Two paratypes, females (424°5 mm, 501 mm total length), taken at the same

locality, in the collection in Hamburg.

Material

The type and paratypes.

R. dissimilis is unique among the southern Atlantic skates in that it shows

a marked reduction in size, to eventual loss, of mid-dorsal thorns from the

disc to the tail. In this respect and in tooth count, it resembles R. senta from the

northern Atlantic, but may be distinguished from this species in that lateral

rows of thorns are present on the disc and tail, and that it lacks the pale cross-

barring of the tail which is typical of R. senta (Bigelow & Schroeder, 1953).

In R. dissimilis, the ventral surface of the tail, from about the posterior edge of

the pelvics to about the origin of the first dorsal fin, is a uniform grey colour,

the tip of the tail being pale white. This peculiar tail coloration may result in

confusion with R. griseocauda from the Patagonian—Falkland region, but in this

species the tooth count is lower (30-36 rows in upper jaw in R. griseocauda

(Norman, 1937); 37-41 rows in R. dissimilis) and there are major differences in

ocular, scapular and caudal spination.

In morphometric dimensions, R. dissimilis approximates to R. leopardus,

but it differs from this species in its lower tooth count (52-70 rows in upper

jaw in R. leopardus), its peculiar caudal spination and in ventral coloration.

Its identity is confirmed by its higher vertebral count (R. dissimilis Vtr 29-30;

Vprd 65-69; VX 94-99; R. leopardus Vtr 31-33; Vprd 55-58; VX 88-go) and

by the presence of two well-defined slits in the inner dorsal border of the

clasper glans. By these facts, R. dissimilis is also distinguished from R. confundens,

which has a similar tooth count. In R. confundens the vertebral count is lower

(Vtr 28-30; Vprd 55-63; V2’ 84-92), and there is no distinct proximal slit in

the glans. Furthermore, the snout in R. dissimilis is comparatively longer and

the spination of the tail much less heavy than in R. confundens.

200 ANNALS OF THE SOUTH AFRICAN MUSEUM

2:0cm

Fic. 15. Raja dissimilis.

A: rostral bar and rostral appendices.

B: external view of right clasper from the dorsal side. |

C: lateral view of right clasper, opened to show structural features of the glans.

ap—apopyle; cf—cleft; rh—rhipidion; hp—hypopyle; sh—shield; sp—spike; st—sentinel.

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 20!

A ee! coat oan eT snout

B so. oo

VAN con am eae

i preota

Ae renasal

== p

A a internasal

A ae interorbit

SSS | SSeS

: 10 20

Te total length

Fic. 16.

Comparison of cranial measurements of R. confundens and R. leopardus after

the method of Hubbs & Hubbs (1953). In each diagram the base line

represents the range of the measurement, the vertical line represents the

mean, the open area represents the standard deviation on each side of the

| mean, and the solid area represents two standard errors on either side of

the mean.

A— Raja leopardus B— Raja confundens

TABLE 14. R. dissimilis. Measurements expressed as permillage of the total length.

Character Type Paratype Paratype

(501 mm) (424°5 mm)

Total length P : : : . 1000 I 000 I 000

Disc width : : : ~ : 609 571 593

Disc length 3 ; 3 ; ~ 495 503 516

| Snout to greatest disc width . : “1a 308 318 312

| Snout to middle of vent ; : / e506 485 502

Middle of vent to 1st dorsal origin. Be} by) 395 413

Snout length. 5 J ; ant3o 143 I51

Preoral length . : : ; elo 150 161

i Prenasal length . 3 Cee y: 126 141

Eye: longitudinal diameter . : : 35 39 40

Eye and spiracle ; : ; : 47 46 51

Interorbital distance . : ‘ : 33 ZI 36

' Interspiracular distance : : : 67 66 75

Internasal distance. é é 3 84 80 85

. Mouth width. j : : , 87 71 80

Gill slit lengths: rst. ; d : 14 14 13

Bedi... : , : 19 16 16

thy’. : é : 12 13 10

. Distance between inner ends of gill slits:

TSEne. : : : 139 140 141

EA As : ; ‘ 88 86 94.

Ist dorsal fin: height i : : 31 19 31

base length . : . 50 52 45

and dorsal fin: height : ; : 28 28 26

base length . : : 43 51 4I

Interdorsal space : : , : oO O O

202 ANNALS OF THE SOUTH AFRICAN MUSEUM

Description

Disc about 1-1—-1°2 times as broad as long, its width 1-6—1°8 in total length;

sharp-pointed in front with anterior angle in front of spiracles 88°—93°; anterior

margins concave just behind tip of snout and again at level of spiracles, more

so in adults than in juveniles; outer and posterior angles broadly rounded.

Axis of greatest breadth 1-5-1°7 times as far from tip of snout as from posterior

edge of disc. ‘Tail with moderately wide lateral folds on posterior half; its

length from middle of vent to origin of first dorsal fin 1-2—1-3 in length from

middle of vent to tip of snout.

Spination Holotype Paratype Paratype

(424°5 mm) (501 mm)

Preorbital thorns. : ; : 7/8 2/1 4/4

Postorbital thorns : : ; 2/5 2/2 3/2

Supraspiracular thorns. : : 1/1 1/1 1/1

Interspiracular thorns . : ; 1/1 1/1 1/1

Median nuchal thorns . 5 : 1/2 1/1 2/3

Scapular thorns . : ; 4 o/o 3/3 2/2

A series of 20-32 thorns on mid-line of back and tail extending from about

nuchal region almost to first dorsal origin in juveniles, but only extending

three-quarters of tail length in adults. Median series flanked on each side on

tail (juveniles) and on back and tail (adults) by a row of 16—35/17—38 thorns.

Thorns along mid-dorsal region of back and tail becoming smaller and more

widely-spaced posteriorly, until in larger specimens, thorns indistinguishable

from spinules in region in front of first dorsal origin. Juveniles with spinules

over whole of disc and tail, no large thorns on rostral cartilage; adults with

spinules on interorbit and on tail, otherwise naked, except for 2 rostral thorns.

Ventral surface of disc and tail smooth.

Snout pointed and a little produced; its length in front of orbits 4:1-4-6

times as long as distance between orbits; its length in front of mouth 1-7-1-9

times as great as distance between nostrils. Distance between orbits 1-1—1-3 in

length of orbit. Rostral cartilage extending from cranium as hard bar with

rostral appendices fused to bar throughout their length; anterior rays of pectoral

fins falling short of rostral appendices.

Mouth slightly arched, more so in adults than in juveniles; nasal curtain

fringed; expanded posterior margin of nostril heavily fringed. Teeth arranged

in 37-41 rows in upper jaw, blunt and flat in females, sharp-pointed in adult

males. .

Anterior lobes of pelvics fin-like and continuously connected with posterior

lobes along outer margin of fin.

First and second dorsals more or less similar in size and shape; first dorsal

continuous with second.

Vertebral count: Holotype Paratype Paratype

(424°5 mm) (501 mm)

Vir : : : 3 : cae 29 30

Vprd : : : : ; 67 65 69

Va , : : : : 98 04 99

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 203

Colour

Upper surface of disc and tail uniformly dark grey, tail mottled with white

posteriorly. Ventral surface pale, posterior margins of pectorals and pelvics

dark grey. Ventral surface of tail uniformly grey, becoming mottled with white

posteriorly; region below dorsal fins white.

Raja confundens n.sp.

(PI rr Ay Be iics 17 A,/ 5, C)

? Raja barnardi: Krefit, 1968a: 61, pls IIIc, IV.

Types |

The holotype, a juvenile male (378 mm total length), trawled off Cape

Columbine in 620 metres, in the collection of the South African Museum

(SAM 24411). The paratype, a male (478 mm total length), taken east of

Cape Point in 660 metres, in the same collection (SAM 24479).

Material |

22 specimens of both sexes (363-677 mm total length) including the

holotype and paratype, trawled between Walvis Bay and east of Cape Point

in 350-660 metres. Only 2 specimens preserved in the collection of the South

African Museum (SAM 24480, PN 3), the others discarded at sea.

10 specimens trawled by R.V. Walther Herwig in 200-500 metres between

Hondeklip Bay and Cape Frio also examined. These specimens in the collec-

tion of the Institut fiir Seefischerei, Hamburg.

R. confundens closely resembles the common South African west coast

species R. leopardus, with which it is taken in trawls, but it may easily be

distinguished from this species by its heavier spination pattern, particularly on

the tail, and by its lower tooth count. The head in R. confundens is shorter and

broader than in R. leopardus (cf. tables 15 and 16). ‘These differences have been

plotted graphically, according to the method of Hubbs & Hubbs (1953) and

as there are no overlaps of standard deviations (except interorbit) the differences

are significant (fig. 16). Furthermore, although the precaudal vertebral count

is similar in the two species, there is a small difference in the range of trunk

vertebrae. The identity of R. confundens is confirmed by its clasper structure.

The species identified as R. barnard:t by Norman (1935) was distinguished

from R. leopardus on the basis of a lower tooth count (40-42 rows in the upper

jaw), which corresponds to that of R. confundens, and because of a single row

of thorns along the back and tail. Krefft (1968a) on re-examination of the

type of R. barnard: has shown that there are more than 50 rows of teeth in the

upper jaw, while scattered lateral thorns in the juvenile type suggest the forma-

tion of lateral rows in the adult. There is no significant difference in tooth count,

spination, vertebral count and proportional dimensions between R. barnardi

and R. leopardus, and it is therefore held that these two species are synonymous.

Krefft (1968a) has referred four specimens obtained in the tropical east

Atlantic by the Galathea and Atlantide expeditions to R. barnardi. These specimens

are smaller than the type of R. barnardi, and yet possess a heavier spination

204. ANNALS OF THE SOUTH AFRICAN MUSEUM

pattern and a lower tooth count. Examination of the proportional measurements

and vertebral counts, suggests that these specimens should rather be referred

to R. confundens.

The strongly re-curved thorns along the back and tail may cause R.

confundens to be identified with R. caudaspinosa, but it differs markedly from

this species in tail length and in tooth count.

TAS

| 5.0 cm |

2.5cm

4.0cr

Fic. 17. Raja confundens.

A: rostral bar and rostral appendices.

B: external view of right clasper from the dorsal side.

C: lateral view of right clasper, opened to show structural features of the glans.

ap—apopyle; cf—cleft; dd—dermal denticles; hp—hypopyle; rh—rhipidion; se—sentina;

sh—shield; sl—slit; sp—spike; st—sentinel.

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 205

TABLE 15. R. confundens. Measurements expressed as permillage of the total length.

Number of specimens in the range 20.

Character Type Paratype Range

Total length : : : F : T1000 I 000

Disc width : ; : : : 595 596 558-668

Disc length . : 2 487 459 442-529

Snout to greatest disc iclttae , g 320 280 280-340

Snout to middle of vent ‘ : 444. 441 444-521

Middle of vent to 1st dorsal see : 413 397 360-413

Snout length : : 3 : 124 107 IOI—124

Preoral length . . ‘ : 3 136 118 107-137

Prenasal length . ; : : : III 94 81-111

Eye: longitudinal diameter . ‘ : 37 37 36-47

Eye and spiracle : . . ; 50 51 46-60

Spiracle . : ‘ : : 25 20 20-28

Interorbital Genes . : : . 36 36 36-44

Initerspiracular distance : : : 61 63 57-74

Internasal distance. . : : 69 75 69-85

Mouth width . 3 : : ‘ 69 69 69-95

Gill slit lengths: st. : ; : 18 17 14-24

Ore hs ; : : 19 18 15-25

5th : : 14 10 10-15

Distance between inner ends as call slits:

FSEs|). < ; ; ; 148 154 148-170

5th: : : . 82 86 82-101

1st dorsal fin: height : : ‘ 29 23 24-44.

base length . ; : 60 75 44-61

and dorsal fin: height ‘ : ‘ 32 22 19-38

base length . : : 60 69 41-61

Interdorsal space : : ; } oO ) 0-12

Description

The figures in parentheses refer to the range of variation for 20 specimens,

not including the type and paratype.

Disc about 1-2—1°3 (1:2-1°3) times as broad as long, its width 1-5~-1-7

(1°5-1°8) in total length; obtuse in front, with anterior angle in front of

spiracles 100° (99°—120°); anterior margins weakly concave just behind tip of

snout; outer angles broadly rounded, posterior margins evenly convex. Axis

of greatest breadth 1-6—1-9 (1-4-1-°8) times as far from tip of snout as from

posterior edge of disc. Tail with lateral folds along posterior third; its length

from middle of vent to origin of first dorsal fin 1-1 (1:1-1-4) in length from

middle of vent to tip of snout.

Spination Holotype Paratype Range

Circumorbital thorns . ‘ : 6/6 6/6 5/6-10/10

Supraspiracular thorns. : : 1/1 1/1 1/1—1/2

Interspiracular thorns . : : 1/1 1/2 1/1-2/2

Median nuchal thorns . é : 4 5 2-9

Lateral nuchal thorns . : 5 2/1 4+1/3 1/2-5-+2/3+5

Scapular thorns . ; : 2/2 2/2 2/2-3/3

| A row of 18-21 (17-24) thorns along mid-line of back and tail, from about

| nuchal region or just posterior to nuchal region to origin of first dorsal. Mid-

dorsal thorns flanked on each side by a single row of larger, recurved thorns,

206 ANNALS OF THE SOUTH AFRICAN MUSEUM

extending almost to first dorsal; smaller hooked spines situated laterally on

tail from about axils of pelvics to about half tail length; smaller flattened

asperities scattered laterally over tail. Stellate-based thorns on rostral cartilage

and along anterior margins of disc to outer angles. Some specimens with a

patch of spines on each posterior angle of disc. Ventral surface of disc and tail

smooth.

With increasing size, spination pattern becomes heavier and more complex,

so that 3 rows of thorns extend from nuchal region to origin of first dorsal and

a further row of semi-lateral thorns is developed on the tail, making 5 rows in

this region.

Snout rounded and not produced; its length in front of orbits 3-0-3-4

(2°5-3°3) times as long as distance between orbits; its length in front of mouth

1'6-1'g (1°4-2°0) times as great as distance between nostrils. Orbits 1-5-1°8

(1-1-2'0) times as long as spiracles; distance between orbits 1-0 (0°g-1°4) as

great as length of orbit. Rostral cartilage projecting from cranium as hard bar

with rostral appendices fused to bar throughout their length; tips of pectoral

rays falling short of rostral appendices.

Mouth slightly arched; nasal curtain fringed; expanded posterior margin

of nostril heavily fringed. ‘Teeth arranged in 44 rows in upper jaw (39-45),

broad and flat in females and juveniles, but sharp-pointed in mature males.

Anterior lobes of pelvics fin-like and continuously connected with posterior

lobes along outer margin of fin.

First and second dorsals confluent, sometimes with a small interdorsal

space; second dorsal about equal in size to first.

Vertebral count: Vtr 28-30; Vprd 55-63; V2 84-02.

Colour

Dorsal surface of disc and tail uniformly dark grey. Ventral surface with

irregular darker patches along posterior margins of pectorals and pelvics,

about vent and along tail.

Raja leopardus Von Bonde & Swart, 1923

(Pl. 12 A; Figs 18 A, B, C)

Raia leopardus Von Bonde & Swart, 1923: 7, pl. 20, fig. 2. Barnard, 1925: 74. Smith, 1961: 67,

7) tien 7a:

Raja leopardus: Norman, 1935: 44. Fowler, 1941: 390.

Raja barnardi Norman, 1935: 43, fig. 14; Fowler, 1941: 371.

Raia quadrimaculata (non Risso) Von Bonde & Swart, 1923: 5. Barnard, 1925: 70, pl. 4, fig. 5.

Raia lintea (non Fries) Barnard, 1925: 72.

Raia naevus (non Miller & Henle) Barnard, 1925: 72.

Types

The holotype of R. leopardus (247 mm total length), taken in 73 metres off

the coast of Natal, formerly in the collection of the Government Marine

Survey; now missing. 'T'wo juvenile specimens labelled ‘cotype’ in the collection

of the British Museum (Natural History).

}

;

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 207

The holotype of R. barnardi, a juvenile male (375 mm total length),

trawled by the Discovery off the west coast of the Cape Peninsula (34°00’S,

17°58’E) in 173-210 metres, in the collection of the British Museum (Natural

History).

Material

70 specimens of both sexes (269-957 mm total length) trawled from WNW

Liideritzbucht to east of Cape Point in 300-660 metres. 26 specimens preserved

ae \

4:0 cm

2:0 cm

/ Fic. 18. Raja leopardus.

| A: rostral bar and rostral appendices.

B: external view of right clasper from the dorsal side.

C: lateral view of right clasper, opened to show structural features of the glans.

i ap—apopyle; cf—cleft; dd—dermal denticles; hp—hypopyle; rh—rhipidion; sh—shield;

f sl—slit; sp—spike; st—sentinel.

208 ANNALS OF THE SOUTH AFRICAN MUSEUM

in the collection of the South African Museum (SAM 22478, 24341, 24421-2,

24453-5, 24475-7, 24481-4, 24486-92).

R. leopardus is the commonest skate taken by commercial trawlers on the

west coast of South Africa. It so closely resembles R. wallace: in external form,

that it is difficult to distinguish between the two on the basis of measurement

alone. However, there appears to be differences in the structure of the clasper

glans and in the precaudal vertebral count, but the two species may yet prove

to be synonymous.

R. leopardus has previously been confused with the North Atlantic species

R. naevus and R. lintea. Norman (1935) has pointed out the differences between

R. leopardus and R. naevus. R. leopardus may be distinguished from R. lintea by

having three rows of thorns along the mid-dorsal region of the back, in which

the thorns in the median row on the tail are the smallest. In R. lintea there is

always one row of spines along the back, while the median row on the tail

always has the largest thorns.

The species identified as R. barnardi by Norman (1935) was distinguished

from R. leopardus on the basis of a lower tooth count and the presence of a

single row of spines along the back and tail. Krefft (1968a) on re-examination

of the type of R. barnardi has found that there are more than 50 rows of teeth

in the upper jaw, while the presence of scattered lateral thorns on the tail of the

juvenile type suggests the formation of lateral rows in the adult. There is no

significant difference in tooth count, spination, vertebral count and proportional

dimensions between R. leopardus and R. barnardi, and it is therefore held that

the two species are synonymous.

It should be noted that the specimens identified as R. barnard: by Wallace

(1967) do not follow the type description for that species, and have been referred

to a new species R. wallacei.

Description

Disc about 1-1—1-3 times as broad as long, its width 1-5-1-7 in total length;

obtuse in front, with anterior angle in front of spiracles 100—110°; anterior

margins concave just behind tip of snout and again at level of spiracles, more

so in adult males than in females and juveniles; outer and posterior angles

broadly rounded. Axis of greatest breadth 1-4-1-9 times as far from tip of

snout as from posterior edge of disc. Tail with moderately wide lateral folds

on posterior third; its length from middle of vent to origin of first dorsal fin

I-I-1°5 in length from middle of vent to tip of snout.

Juvenile specimens with 4-5 thorns around inner margin of orbit; 0-1

thorn above spiracles; 1 pair small thorns between spiracles; 3-5 median

nuchal thorns; 1-2 scapular thorns. A row of 25-30 thorns along mid-line of

back and tail from nuchal region to origin of first dorsal fin. No interdorsal

thorns. Usually 3-6 small thorns situated semi-laterally on each side of tail at

about level of pelvics; remainder of tail with small fine spinules. Dorsal surface

with spines on tip of snout and anterior margins of disc to level of spiracles,

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 209

TABLE 16. R. leopardus. Measurements expressed as permillage of the total length.

Number of specimens 70.

Character Mean Range

Total length 3 : : , g ‘ I 000

Disc width j : : : : , 620 571-692

Disc length ‘ : ; : 510 481-543

Snout to greatest disc nade : : A 313 278-343

Snout to middle of vent : : : 4.90 435-553

Middle of vent to 1st dorsal oie ; F 374. 336-419

Snout length : ‘ : : , : 134 I1I—152

Preoral length . . h , : : 14! 122-169

Prenasal length . : : , ; : 115 94-139

Eye: longitudinal diameter . : : : 36 28-34.

Eye and spiracle : : ; ; : 50 46-56

Spiracle . ‘ : ‘ ‘ A 22 17-34

Interorbital Bi ence : : : : : 35 30-44.

Interspiracular distance : : : ; 60 48-69

Internasal distance . ; ; : 1 67 61-79

Mouth width. ‘ : 4 : ‘ 75 56-93

Gill slit lengths: rst. ‘ ; : : 14 11-24

sid ty : : : i 16 13-23

5th ‘ ‘ : ; II 7-21

Distance between inner ends of ell slits:

1 : : : : 147 135-176

Reh : : : ; 84 61-110

1st dorsal fin: height ; , : ‘ 29 21-44

base length . ; : : 54. 41-65

2nd dorsal fin: height : : : : 29 21-41

base length . , : ; 53 44-67

Interdorsal space ; : ; : : oO o-18

and wider spaced asperites over remainder of disc and pelvics. Ventral surface

with spines on tip of snout.

Older specimens with 5-13 thorns around inner margin of each orbit and

above spiracles; 1-2 pairs small inter-spiracular thorns; 4-9 median nuchal

thorns, usually with 1-2 lateral rows forming a triangular patch; 2-3 scapular

thorns. A median series of 19-29 widely-spaced thorns along the back and tail

to origin of first dorsal, becoming reduced in size posteriorly and flanked on

each side by 1 row thorns along back and 2-3 rows on tail; lateral and semi-

lateral rows of thorns not greatly enlarged. Dorsal surface of disc with spines

on snout, anterior margins of pectorals and sides of tail. Ventral surface with

spines on tip of snout and along anterior margins.

Snout slightly pointed but not produced; its length in front of orbits 2-8-4:4

times as long as distance between orbits; its length in front of mouth 1-5-2°3

times as great as distance between nostrils. Orbits 1-o—2-1 times as long as

spiracles; distance between orbits o-g-1-4 times as great as length of orbit.

Rostral cartilage projecting from cranium as hard bar with rostral appendices

fused to bar throughout their length; anterior rays of pectoral fins falling short

of rostral appendices.

Mouth slightly arched; nasal curtain fringed; expanded posterior margin

of nostril heavily fringed. Teeth arranged in 52—70 rows in upper jaw, with

large posterior cusp in males, but blunt and flat in juveniles and females.

210 ANNALS OF THE SOUTH AFRICAN MUSEUM

Anterior lobes of pelvics fin-like and continuously connected with posterior

lobes along outer margin of fin. First and second dorsals confluent, sometimes

with a small interdorsal space; second dorsal usually slightly smaller than first.

Vertebral count: Vtr 31-33; Vprd 55-58; V2 88-90.

Colour

Dorsal surface of disc uniformly brown to grey, sometimes with numerous

dark spots, especially in juveniles. Norman (1935) reports the presence of a

naevus-like ocellus at the base of each pectoral fin, particularly in adults. Ventral

surface uniformly pale or with irregularly arranged darker blotches and patches

on pectorals and pelvics.

Raja wallacei n.sp.

(Pl. 12 B; Figs 19 A, B)

Raja barnardi (non Norman) Wallace, 1967: 39, figs 20, 21.

T ype

The holotype of R. wallacei, an adult male (842 mm total length), trawled

at 34°10'S, 17°45’E in 292 metres, in the collection of the Oceanographic

Research Institute, Durban. The paratype, a female (489 mm total length)

Fic. 19. Raja wallacei.

A: external view of right clasper from the dorsal side.

B: lateral view of right clasper, opened to show structural features of the clasper glans.

ap—apopyle; cf—cleft; dd—dermal denticles; hp—hypopyle; pt— promontory; rh—rhipidion;

se—sentina; sh—shield ; sl—slit; sp—spike; st—sentinel.

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 2i1

taken SE Durban Bluff also in the same collection.

Material

The holotype and paratype. Wallace (1967) has examined 20 specimens.

Wallace (1967) confused this species with R. barnardi, which was taken at

the same locality. Comparison of the types of R. wallace: with R. barnardi has

revealed that in R. wallace: the precaudal vertebral count is higher, the snout

shorter, the interorbital distance greater than in R. barnardi. It should be noted

R. barnardi is now held to be synonymous with R. leopardus.

R. wallace: closely resembles the common west coast skate R. leopardus, but

may be distinguished from it by its greater number of precaudal vertebrae.

Furthermore there are slight differences in the structure of the clasper glans.

So closely are these two species related, however, that on the basis of measure-

ment alone it is difficult to separate the one from the other. They may yet

prove to be synonymous. #. wallace: seems to be characterized by the presence

of dark blotches on the tip of each pelvic fin.

Although the holotype of R. wallacei was taken in Cape waters, the present

author has not recorded this species during the survey. ‘The species would seem

to occur along the whole of the coastal region from the Cape to north of the

Limpopo River mouth (Wallace, 1967).

TABLE 17. R. wallacei. Measurements expressed as permillage of the total length.

Character Type Paratype

(B 155) (B 126)

Total length f ; : : é . 1000 I 000

Disc width : : : : ; : 625 557

Disc length ; : : , : 4 494. 466

Snout to greatest disc width . : 4 : 289 286

Snout to middle of vent 3 : ; : 472 448

Middle of vent to 1st dorsal origin. 3 ‘ 400 413

Snout length : 3 : E ‘ 2 96 112

Preoral length . ; : é : , 102 124

Prenasal length . é : : : : 73 99

Eye: longitudinal diameter . ; : : 36 40

Eye and spiracle ; : 4 : 5 55 56

Spiracle . : : i : : 25 21

Interorbital distance . ‘ : : : 42 37

Interspiracular distance : : ; : 62 62

Internasal distance : : : . é 69 70

Mouth width . 2 : X k ‘ 82 76

Gill slit lengths: 1st. : , : ; 20 14

grad... , ? : : aD 17

5th : : i : 15 13

Distance between inner ends of gill slits:

WSt | : : ; : 158 142

Gila oe : : : : 88 86

ist dorsal fin: height : ; : ‘ oy) 29

base length . ‘ : : 53 61

and dorsal fin: height 4 : . ; 27 27

base length . : E é 51 57

Interdorsal space eae : : : 20 Q

Pag ANNALS OF THE SOUTH AFRICAN MUSEUM

Description

Disc about 1-2—1°3 times as broad as long, its width 1-6—1-8 in total length;

obtuse in front with maximum angle in front of spiracles about 110°; anterior

margins concave just behind tip of snout and again at level of spiracles; outer

and posterior angles broadly rounded, posterior margins slightly convex. Axis

of greatest breadth 1-4-1-6 times as far from tip of snout as from posterior edge

of disc. ‘Tail with narrow lateral folds; its length from middle of vent to origin

of first dorsal fin 1-1-1-2 in length from middle of vent to tip of snout; its

length from middle of vent to tip of tail 1-1-1-3 times as long as distance from

middle of vent to tip of snout.

Spination Holotype Paratype

Circumorbital thorns . : 4 ; : 8/10 7/8

Supraspiracular thorns. ; : : 2/3 2/3

Interspiracular thorns . ; : : : o/o 1/1

Median nuchal thorns : : I 7

Lateral nuchal thorns . , : : 3 1/1 4/4

Scapular thorns . ; ; ; ; ; o/o 1/2

No median row of thorns in holotype, but paratype with 34 thorns along

mid-line of back and tail from about nuchal region to first dorsal origin; no

thorns in dorsal interspace. Median row (or mid-line) flanked on each side by

a single row of thorns from about nuchal region to axils of pelvics, increasing

to two rows on each side from axils to level of interdorsal space. Dorsal surface

with spines on tip of snout, along anterior margins of disc to outer angles, and

on tail. Ventral surface with spinules on tip of snout and anterior margins of

disc to level of mouth. Otherwise smooth.

Snout not produced, but with short terminal projection; its length in

front of orbits 2-3~-3-1 times as long as distance between orbits; its length in

front of mouth 1-5-1°8 times as great as distance between nostrils. Orbits 1-5—1-9

times as long as spiracles; distance between orbits 0-g—-1-2 times as great as

length of orbit.

Mouth slightly arched; nasal curtain fringed; expanded posterior margin

of nostril heavily fringed. Teeth arranged in 59-67 regular rows in upper jaw,

with round bases and sharp posterior cusp.

Anterior lobes of pelvics fin-like and continuously connected with posterior

lobes along outer margin of fin.

First dorsal larger than second and separated from it by a small but definite

space; interspace between dorsals about 39% as long as base of first dorsal.

Number of precaudal vertebrae (Vprd) 70.

Colour

Upper surface uniformly brown with scattered, irregular lighter spots.

Naevus-like ocellus at base of each pectoral. Lower surface pale, with a single,

dark blotch on tip of anterior lobe of pelvic fin.

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 213

Bathyraja smithi (Miller & Henle, 1841)

(Pl. 13; Figs 20 A, B, C; 21 A, B)

Raja smithi Miiller & Henle, 1841: 150, pl. 49, fig. 1. Gray, 1851: 112. Bleeker, 1860: 58.

Dumeéril, 1865: 553. Giinther, 1870: 467. Gilchrist, 1902: 168. Thompson, 1914: 159.

Raja smithi: Norman, 1935: 41. Fowler, 1941: 364.

Raia smith: Garman, 1913: 366. Von Bonde & Swart, 1923: 5.

Raia smithi: Barnard, 1925: 66, pl. 4, fig. 4.

non Raia smithi: Smith, 1961: 66, non pl. 3, fig. 68 (= C. parcomaculata).

Raja eatonit Giinther, 1876: 390; 1879: 166.

Raja eatont Giinther, 1880: 15.

Raia eatoni: Garman, 1913: 365.

Types

The holotype of R. eatoni, a male (26-5 in. (673 mm) total length), from

Royal Sound, and holotype of R. smithi, a dried skin, from South Africa, in

the British Museum (Natural History). A mature male specimen, labelled type of

'R. smithiu, from the Bosphorus, in the Muséum National des Sciences Naturelles,

Paris, does not belong to this species and should be referred to R. clavata.

100 mm By :

Fic. 20. Bathyraja smithit.

A: rostral bar and rostral appendices.

B, C: dorsal view of immature males, showing variation in spination.

214. ANNALS OF THE SOUTH AFRICAN MUSEUM

Material

10 specimens of both sexes (309-1 141 mm total length), trawled east of

Cape Point in 658-868 metres. 7 specimens preserved in the collection of the

South African Museum (SAM 24473, 15666).

Although previously included in the genus Raja, this species should now

be referred to the genus Bathyraja because of the nature of the rostral bar and

rostral appendices, the number of precaudal vertebrae, and the lack of a shield

in the clasper glans. Ishiyama & Hubbs (1968) also define the genus by the

presence of a pseudosiphon in the clasper glans, but this is present in species

of the radiata-group, and suggest that the genus Bathyraja is restricted to the

Pacific. However, Dr. G. Krefft (personal communication) suggests that

Bathyraja is a bipolar, antitropical genus.

B. smithii is most easily distinguished from all other known South African

species by its low tooth count (less than 30 rows in the upper jaw) and by the

lack of lateral rows of thorns on the tail. Adults resemble some specimens of

R. clavata in shape, but differ in tooth count and spination. It should be noted

5.0

2.0cm |

Fic. 21. Bathyraja smithii.

_ A: external view of right clasper from the dorsal side.

B: lateral view of right clasper, opened to show structural features of the glans.

ap—apopyle; hp—hypopyle; kn—knife; pj—projection; rd—ridge; ‘ps’—‘pseudosiphon’;

sl—slit.

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 215

that Wallace (1967) gives a tooth count of 23-26 rows for R. clavata. This is

inconsistent with previous work and with the investigations of this paper.

The dorsal surface of the disc of B. smithii is devoid of large thorns, except

in juvenile specimens, which resemble R. spinacidermis. However, R. spinacidermis

has a higher tooth count. The smoky-black borders to the ventral margins of

the pectoral and pelvic fins may lead to confusion with R. alba, but this species

can easily be distinguished by its produced and pointed sriout.

Although Giinther (1879) and Garman (1913) consider R. eatonii to be

closely allied to B. smithii, and Barnard (1925) regards R. eatonii as a sub-species,

Norman (1935) holds the two to be distinct on the basis of differences in spina-

tion and because of a longer snout in R. eatoni. Preliminary external examina-

tion of the claspers of the type of R. eatoni, has led me to synonymise the two

species, until an examination of the clasper cartilages can be made.

TABLE 18. B. smithii. Measurements expressed as permillage of the total length.

Number of specimens 10.

Character Mean Range

Total length : : ¢ 4 : d I 000

Disc width ; : ‘ ‘ : : 708 680-745

Disc length ; : A : : : 529 513-544

Snout to greatest disc width . é : : 324 302-338

Snout to middle of vent : : : : 517 498-543

Middle of vent to 1st dorsal origin. : : 355 335-376

Snout length : : : : : . 121 97-134

Preoral length. : : : : : 115 105-131

Prenasal length . : 4 : : : 89 82-99

Eye: longitudinal diameter . A ’ : 33 29-47

Eye and spiracle : : : : : 49 40-57

Spiracle . : ‘ : 2 : ‘ 26 19-30

Interorbital distance . : : : : 56 49-61

Interspiracular distance ie) ees : F 79 75-84

Internasal distance , d : ‘ 3 QI 84-98

Mouth width . : 2 3 : , 86 78-93

Gill slit lengths: 1st. . : : < 14 12-18

ord}: ‘ ‘ ‘ : 16 13-17

5th.’ : : : 14 12-18

Distance between inner ends of gill slits:

TStas : : : : 180 17I—190

5th . ; , 5 bs 122 109-136

1st dorsal fin: height : : ‘ : QI 14-29

base length . : ; : 38 32-50

and dorsal fin: height : ; i : 20 17-28

base length . : ‘ : 36 28-44

Interdorsal space 5 ; : : : 14 7-22

Description

Disc about 1:3-1°4 times as broad as long, its width 1-3-1°5 in total

length; obtuse in front, with anterior angle in front of spiracles 90°—100°:

anterior margins weakly concave posterior to snout in juveniles, but more

sinuous in adults and particularly concave at level of spiracles in mature males;

posterior and outer angles broadly rounded. Axis of greatest breadth 1-2-1-7

times as far from tip of snout as from posterior edge of disc. ‘Tail with moderately

216

ANNALS OF THE SOUTH AFRICAN MUSEUM

Location and collection numbers of South African rajid material.

TABLE 19.

Species Institution Type Coll. No.

C. parcomaculata . B.M.(N.H.) paratype 1935-7-14.1

R. albalinea. B.M.(N.H.) paratype 1935.7.14.4

R. spinacidermis . B.M.(N.H.) holotype 1935-7-19.7

R. ocellifera B.M.(N.H.) syntypes 1895.12.27.14

1905.6.8.14

R. rhizacanthus B.M.(N.H.) holotype 1905.6.8.13

R. leopardus . B.M.(N.H.) 2 ‘cotypes’ 1935-7-14.3

R. barnardi . B.M.(N.H.) holotype 1935.5.2.65

B. smithi B.M.(N.H.) holotype 1953.8.10.1

B. eatonu B.M.(N.H.) holotype 1876.3.23.21

R. pullopunctata . J.L.B.S. holotype —

paratype —

C. triangularis JcL.B,5. holotype —

paratype —

R. campbell. ORR holotype B 804

paratype B 859

R. wallace . OnE holotype B 155

paratype B 126

R. stenorynchus O.R.I. holotype B 186

R. springert . O.R.I. holotype B 909

paratype B 185

R. lanceorostrata . ORE holotype B 869

paratype B 868

R. doutret M.N.H.N. holotype 59, 41

R. capensis . M.N.H.N. paratype 1333

R. straeleni . I.R.S.N. holotype I.G. 16808; 99

20 paratypes I.G. 16808; 100-107

R. robertst LS-H. holotype 54/67

R. ravidula . I.S.H. holotype 50/67

2 paratypes 47a, b/67

R. dissimilis. iS. holotype 46a/67

2 paratypes 46b, c/67

R. confundens S.A.M. holotype 24411 ;

paratype 24479

B.M.(N.H.) — British Museum (Natural History), London.

J-L.B.S. — J.L.B. Smith Institute of Ichthyology, Grahamstown.

O.R.I. — Oceanographic Research Institute, Durban. :

M.N.H.N. — Muséum National d’Histoire Naturelle, Paris. |

I.R.S.N. | — Institut Royal des Sciences Naturelles, Brussels. a

1.S.H. — Institut flir Seefischerei, Hamburg.

S.A.M. — South African Museum, Cape Town.

wide lateral folds; its length from middle of vent to origin of first dorsal fin

1:3-1°6 in length from middle of vent to tip of snout. |

Juveniles with 1 thorn in front of orbit and 1-2 behind; 2 median nuchal

spines; 3—4 scapular thorns; a row of about 30 thorns along mid-line of back

and tail from nuchal region to first dorsal origin; 1 thorn in dorsal interspace.

Entire upper surface of disc and tail with widely-spaced spinules. Ventral

surface smooth.

Larger specimens with ocular, nuchal and scapular thorns absent; 14-19

large thorns along mid-line of tail from above vent to origin of first dorsal fin;

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA |

o-1 thorns in dorsal interspace. Dorsal surface of disc and tail spinulose, but

no lateral rows of thorns on tail. Ventral surface smooth.

Snout obtuse; its length in front of orbits 2-0-2-6 times as long as distance

between orbits; its length in front of mouth 1-1—1-4 times as great as distance

between nostrils. Orbits 1-o-1-4 times as long as spiracles; distance between

orbits 1-5—1-9 times as great as length of orbit.

Rostral cartilage elongate with delicate, slender bar, extending to tip of

snout without a segment; rostral appendix attached to extremity of rostral bar

on each side, with distal part free from lateral sides of rostral cartilage; radial

cartilages of pectorals extending anteriorly almost to rostral appendices.

Mouth slightly arched; nasal curtain not fringed; expanded posterior

margin of nostril heavily fringed. Teeth with a single large cusp, arranged in

24-28 regular rows in upper jaw.

Anterior lobes of pelvics fin-like and continuously connected with posterior

lobes along outer margin of fin.

Dorsal fins similar in shape and about equal in size, with convex anterior

margins and broadly rounded.apices; interspace between dorsals 22-40% as

long as base of first dorsal.

Number of precaudal vertebrae (Vprd) 68—71.

| Colour

Dorsal surface more or less uniformly greyish or brownish in preserved

specimens, sometimes with small white spots. Ventral surface white, with black

blotches between gill slits and around vent; anterior margins of pectorals

narrowly and posterior margins of pectorals and pelvics broadly coloured

black; ventral surface of tail smoky- black.

SUMMARY

This paper consists of a systematic revision at the species level of the

Rajidae of the west and south coasts of southern Africa. Five new species are

described, three of which are known only from deep water, while two species

are recorded for the first time in the South Atlantic. Keys to the southern

African rajid fauna are given.

Natural relationships between the species are evident, suggesting a

regrouping of the Raja species at the generic or sub-generic level. The signifi-

cance of these relationships will be dealt with in a later paper.

ACKNOWLEDGEMENTS

I am deeply indebted to Mr. S. X. Kannemeyer of the South African

Museum for his assistance during all phases of this work, and to Mr. P. Henry

of the Kaap-Kunene Group of Fisheries for the collection of many specimens.

I am extremely grateful to the late Prof. J. L. B. Smith and Mrs. M. M. Smith

of the J. L. B. Smith Institute for Ichthyology, Grahamstown, and to Dr. G.

Krefft of the Institut fiir Seefischerei, Hamburg, for access to their collections

and especially for their many valuable and inspiring comments. Also of the

218 ANNALS OF THE SOUTH AFRICAN MUSEUM

Institut fiir Seefischerei, I should like to express my special thanks to Dr. M.

Stehmann for his advice and recommendations, to Dr. F. Mombeck for the

collection of the Walther Herwig material, and to Miss Christa Liibben, not

only for the photographs of R. spinacidermis and R. straeleni, but also for her kind

help during my visit to Hamburg. I wish to acknowledge Dr. A. E. F. Heydorn

and Mr. J. Wallace of the Oceanographic Research Institute, Durban, for

the facilities offered to me during my stay, and Prof. R. Ishiyama of the Tokyo

University of Fisheries for much valuable advice.

My thanks are also due to: Dr. S. Springer of the United States National

Museum, for X-ray photographs of R. oregoni, R. mollis and R. garricki and Miss

P. Verity of the Nuffield Institute of Comparative Medicine for X-rays of

R. spinacidermis, Dr. W. ‘Templeman for X-rays of R. mollis, Drs H. Wormald

and R. Hindle of Durban, Dr. B. Hirschson of Cape Town and Mr. N. Peddie

of Kodak (South Africa) for assistance with X-rays; Drs P. H. Greenwood and

N. B. Marshall of the British Museum (Natural History), Dr. J. P. Gosse of

the Institut Royal des Sciences Naturelles de Belgique, Dr. M. Poll of the Musée

Royal d’Afrique Centrale, Tervuren, Dr. P. J. Castle, formerly of Rhodes

University, and Messrs M. J. Holden and C. N. Humphries of the Fisheries

Laboratory at Lowestoft, all for the loan of specimens.

The Trustees of the South African Museum thank the South African Coun-

cil for Scientific and Industrial Research for a grant in aid of publication.

REFERENCES

BARNARD, K. H. 1923. Diagnoses of new species of marine fishes from South African waters.

Ann. S. Afr. Mus. 13: 439-445.

BARNARD, K. H. 1925. A monograph of the marine fishes of South Africa. Ann. S. Afr. Mus. 21:

1-418.

BARNARD, K. H. 1947. A pictorial guide to South African fishes. Cape Town: Maskew Miller.

BicELow, H. B. & ScHROEDER, W. C. 1948. New genera and species of batoid fishes. 7. mar. Res.

7? 543-566.

BicELow, H. B. & SCHROEDER, W. C. 1950. New and little known cartilaginous fishes from the

Atlantic. Bull. Mus. comp. Zool. Harv. 103: 383-408.

BicELow, H. B. & ScHROEDER, W. C. 1953. Fishes of the western North Atlantic. Part 2.

Sawfishes, guitarfishes, skates and rays. Chimaeroids. Mem. Sears Fdn mar. Res. 1 (2):

i-—xv, 1-588.

BIGELow, H. B. & ScHROEDER, W. C. 1962. New and little known fishes from the western Atlantic.

Bull. Mus. comp. Zool. Harv. 128: 159-244.

BLEEKER, P. 1860. Over eenige vischsoorten van de Kaap de Goede Hoop. Nat. Tijdschr. Ned.

Ind. 21: 49-80.

CADENAT, J. 1960. Notes d’ichthyologie ouest-africaine. XXVII. Raja doutrei, espéce nouvelle

des eaux profondes des cétes du Sénégal. Bull. Inst. fr. Afr. noire (A) 22: 294-312.

Ciark, R. S. 1926. Rays and skates, a revision of the European species. Scient. Invest. Fishery

Bd Scotl. 1926: 1-66.

DumiriL, A. 1865. Histoire naturelle des poissons ou ichthyologie générale. 1. Elasmobranches, plagio-

stomes et holocéphales ou chimeres. Paris.

Donovan, E. 1807. The natural history of British fishes. 5. London: Rivington.

EKMAN, S. 1953. <oogeography of the sea. London: Sidgwick & Jackson.

Forster, G. R. 1965. Raja richardsoni from the continental slope of south-west England. 7. mar.

biol. Ass. U.K. 45% 773-777-

Fow er, H. W. 1925. Fishes from Natal, Zululand and Portuguese East Africa. Proc. Acad. nat.

Sct. Philad. 77: 187-272.

RAJIDAE OF WEST AND SOUTH COASTS OF SOUTHERN AFRICA 219

Fow.er, H. W. 1936. Marine fishes of west Africa based on the collection of the American

Museum Congo Expedition 1909-15. Bull. Am. Mus. nat. Hist. 70% i-vii, 1-1493.

Fow.er, H. W. 1941. Contributions to the biology of the Philippine archipelago and adjacent

regions. ‘The fishes of the groups Elasmobranchii, Holocephali, Isopondyli, and Ostarophysi

obtained by the United States Bureau of Fisheries steamer ‘Albatross’ in 1907 to 1910,

chiefly in the Philippine Islands and adjacent seas. Bull. U.S. nain. Mus. 100 (13) : i-x, 1-879.

FRANCA, F.. DA. 1959. Subsidio para um catalogo dos nomes vernaculos dos peixes marinhos de

Angola. Notas mimeogr. Cent. Biol. pise. 5: 1-37.

_ Garman, S. 1899. Reports on an exploration off the west coasts of Mexico, Central America

and South America and off the Galapagos Islands, in charge of Alexander Agassiz, by the

U.S. Fisheries Commission steamer ‘Albatross’ during 1891, Lieut. Commander Z. L.

Tanner U.S.N. commanding. XXVI. The fishes. Mem. Mus. comp. Zool. Harv. 24: 1-431.

GarMAN, 8. 1913. The Plagiostomia (sharks, skates and rays). Mem. Mus. comp. Zool. Harv. 36:

i-xil, 1-528.

Garrick, J. A. F. 1961. Studies on New Zealand Elasmobranchii. Part XIII. A new species of

Raja from 1 300 fathoms. Trans. R. Soc. N.Z. 88: 743-748.

Gitcurist, J. D. F. 1902. Catalogue of fishes from South Africa. Mar. Invest. S. Afr. 1: 97-99.

GitcurisT, J. D. F. 1922. Report for the year 1921. Annexure A, Annexure B. Rep. Fish. mar.

biol. Surv. Un. S. Afr. 23 7-59.

GitcuRisT, J. D. F. & THompson, W. W. 1916. A catalogue of the sea fishes recorded as

Natal. Ann. Durban Mus. 1: 255-431.

Gray, J. E. 1851. List of the specimens of fish in the collection of the British Museum. Part 1. Chondro-

pterygii. London: Trustees of the British Museum.

GUNTHER, A. 1870. Catalogue of the fishes in the British Museum. 8. Catalogue of the Physostomi,

including the families Gymnotidae, Symbranchidae, Muraenidae, Pegasidae and of the

Lophobranchii, Plectognathi, Dipnoi, Ganoidei, Chondropterygii, Cyclostomata, Lepto-

cardii in the British Museum. London: Trustees of the British Museum.

- GUNTHER, A. 1876. Remarks on fishes, with descriptions of new species in the British Museum,

chiefly from southern seas. Ann. Mag. nat. Hist. (4) 172 389-402.

GunTuHeEr, A. 1879. An account of the petrological, botanical and zoological collections made

in Kerguelen’s Land and Rodriguez, during the transit of Venus Expeditions carried out

by order of Her Majesty’s Government in the years 1874-75. Fishes. Phil. Trans. R. Soc.

168: 166.

GUnTHER, A. 1880. Report on the shore fishes procured during the voyage of H.M.S. Challenger

in the years 1873-1876. Rep. Voy. ‘Challenger’ 1873-1876 (Zool.) 1 (6): 1-82.

Huess, C. L. & Huss, C. 1953. An improved graphical analysis and comparison of series of

samples. Syst. Zool. 2: 49-57.

Husss, C. L. & Isuryama, R. 1968. Methods for the taxonomic study and description of skates

(Rajidae). Copera 1968: 483-491.

Huttey, P. A. 1966. The validity of Raja rhizacanthus Regan and Raja pullopunctata Smith, based

on a study of the clasper. Ann. S. Afr. Mus. 48: 497-514.

Huttey, P. A. 1969. The relationship between Raja miraletus Linnaeus and Raja ocellifera Regan

based on a study of the clasper. Ann. S. Afr. Mus. 52: 137-147.

Isntyama, R. 1952. Studies on the rays and skates belonging to the family Rajidae, found in

Japan and adjacent regions. 4. A revision of three genera of Japanese rajids, with descrip-

tions of one new genus and four new species mostly occurred in northern Japan. 7. Shimonoseki

Coll. Fish. 2: 1-34.

IsHtyamMa, R. 1958. Studies on the rajid fishes (Rajidae) found in the waters around Japan.

J. Shimonoseki Coll. Fish. 72 193-394.

Isuiyama, R. 1967. Fauna Japonica. Rajidae (Pisces). Tokyo: Biogeographical Society of Japan.

IsHryama, R. & Husss, C. L. 1968. Bathyraja, a new genus of Pacific skates (Rajidae) regarded

as phyletically distinct from the Atlantic genus Breviraja. Copeia 1968: 407-410.

Kner, P. 1865-1867. Reise der dsterreichischen Fregatte Novara um die Erde. Fische. Zoologischer

Theil. x (5): 1-433. Wien: K.-K. Hof- und Staatsdruckerei.

Krerrtr, G. 1968a. Knorpelfische (Chondrichthyes) aus dem tropischen Ostatlantik. A tlantide

Rep. 10: 33—76.

Krerrt, G. 19685. Neue und erstmalig nachgewiesene Knorpelfische aus dem Archibenthal des

Siidwestatlantiks, einschliesslich einer Diskussion einiger Etmopterus-Arten stidlicher Meere.

Arch. Fischereiwiss. 19 (1): 1-42.

220 ANNALS OF THE SOUTH AFRICAN MUSEUM

KrerrT, G. & LUspsen, C. 1966. Raja mollis Bigelow & Schroeder, 1950 (Batoidea, Elasmobranchii

Chondrichthyes), ein Erstfund im Nordost-Atlantik. Zool. Anz. 176: 385-395.

LackPEDE, B. G. E. 1803. Histoire naturelle des poissons. 5. Paris: Plassan.

LeIGH-SHARPE, W. H. 1920. The comparative morphology of the secondary sexual characters

of elasmobranch fishes. Memoir I. The claspers, clasper siphons and clasper glands.

F. Morph. 34: 254-265.

LEIGH-SHARPE, W. H. 1921. The comparative morphology of the secondary sexual characters

of elasmobranch fishes. Memoir II. 7. Morph. 35: 263-358.

LeIGH-SHARPE, W. H. 1922. The comparative morphology of the secondary sexual characters

of elasmobranch fishes. Memoirs III, IV, V. 7. Morph. 36: 191-244.

LEIGH-SHARPE, W. H. 1924. The comparative morphology of the secondary sexual characters

of elasmobranch fishes. Memoirs VI, VII. 7. Morph. 39: 553-577.

LEIGH-SHARPE, W. H. 1926. The comparative morphology of the secondary sexual characters

of elasmobranch fishes. Memoirs VIII, IX, X, XI. 7. Morph. 42: 307-368.

Linnaeus, C. 1758. Systema naturae. 10th ed. Holmiae: Laurentii Salvii.

Mutter, J. & HENLE, F. G. J. 1841. Systematische Beschreibung der Plagiostomen. Berlin: Veit.

Norman, J. R. 1935. Coast fishes. Part I. The south Atlantic. ‘Discovery’ Rep. 12: 1-58.

Norman, J. R. 1937. Coast fishes. Part II. The Patagonian region. ‘Discovery’ Rep. 16: 1-150.

PIETSCHMANN, V. 1906. Ichthyologische Ergebnisse einer Reise nach Island. Annin naturh. Mus.

Wien 21: 72-148.

Pott, M. 1949. Résultats scientifiques des croisiéres. du navire-école belge ‘Mercator’. IV.

Poissons. Mém. Inst. r. Sci. nat. Belg. (2) 33: 173-269.

Poti, M. 1951. Poissons. II. Sélachiens et chiméres. Résult. scient. Exped. océanogr. belge Atlant.

sud 4 (1): 12-154.

Pozzi, A. J. 1935. Comunicacion preliminar sobre una nueva especie de ‘raya’ de la costa atlan-

tico Argentina Raia doello-juradoi n. sp. Physis 11: 419-492.

Pozzi, A. J. 1936. Descripcién de la nueva especie de ‘Raya’ Raja doello-juradoi Pozzi. An. Mus.

argent. Cienc. nat. 38: 418-422.

Recan, C. T. 1906. Descriptions of new or little known fishes from the coast of Natal. Ann.

Natal Mus. 1: 1-6.

Recan, C. T. 1908. A collection of fishes from the coasts of Natal, Zululand and Cape Colony.

Ann. Natal Mus. 1: 241-255.

SAMUEL, C. T. 1963. Bottom fishes collected by R.V. Conch off the Kerala coast. Bull. Dep. mar.

Biol. Oceanogr. Unw. Kerala 1: 97-121.

SauvaGE, M. H. 1891. Histoire naturelle des poissons. Jn GRANDIDIER, A., ed. Histoire physique,

naturelle et politique de Madagascar. Paris: Imprimerie Nationale.

Situ, J. L. B. 1961. The sea fishes of southern Africa. 4th ed. Cape Town: C.N.A.

SmitH, J. L. B. 1964. Fishes collected by Dr. Th. Mortensen off the coast of South Africa in

1929, with an account of the genus Cruriraja Bigelow & Schroeder, 1954 in South Africa.

Vidensk. Meddr dansk naturh. Foren. 126: 283-300.

SMITH, J. L. B. & Smiru, M. M. 1966. Fishes of the Tsitsikama Coastal National Park. Johannesburg:

National Parks Board of Trustees.

STEHMANN, M. 1969. Vergleichend morphologische und anatomische Untersuchungen zur

Systematik der nordeuropaischen Rochen der Familie Rajidae. Ph. D. Thesis: University

of Kiel (unpublished).

TEMPLEMAN, W. 1965. Rare skates of the Newfoundland and neighbouring areas. 7. Fish. Res.

Bd Canada 22: 259-279.

TuHompson, W. W. 1914. Catalogue of fishes of the Cape Province. Mar. biol. Rep. Un. S. Afr.

2: 132-167.

TuHorPE, T. 1964. Record rays. Fld Tide 6 (3): 27.

VAN BruGcEN, A. C. 1965. Records and observations in the Port Elizabeth oceanarium in 1960.

Kool. Gart. Lpz. 34% 190-191.

Von Bonpbg, C. 1933. Report No. 10 for the year ending December, 1932. Annexure A.

Annexure B. Rep. Fish. mar. biol. Surv. Un. S. Afr. 10: 32-103.

Von Bonbe, C. & Swart, N. B. 1923. The Platosomia (skates and rays) collected by the S.S.

‘Pickle’. Rep. Fish. mar. biol. Surv. Un. S. Afr. 3 (Spec. Rep. 5): 1-22.

WALLACE, J. H. 1967. The batoid fishes of the east coast of southern Africa. Part III. Skates and

electric rays. Investl Rep. oceanogr. Res. Inst. 17: 1-62.

‘UI PUL WD UI 9[VIG “MoIA [esIOc[ ‘(paaouror

suy o1Aqod jo dy pure sszodsejo fapeur) sayjnop ‘y ‘g ‘ut pue wo UI B[BIG *MOIA [eso ‘(a]TeuIsy) vyvjnovU0IIDd “1 “VW

Plate 1

ea]

‘UI PUB UID Ul 9[eIg “MITA [eso] *(paAourer Jodseyo

pue uy o1ajod yy 8t1 fayeur) vynjsundoyng ‘y"q “Ul puv WD UT 9[BIG “MOTA [esIOd ‘(gyeur atuaanl) nynjoundoynd ‘YW

Plate 2

fea]

Ann. S. Afr. Mus., Vol. 55

Plate 3

Ann. S. Afr. Mus., Vol. 55

UID UT BTVOG *MOTA [esIO]

(

aTBWdy

)

psourdsppnvo

‘ul pur

‘UT PUY WD UT 9[BOG ‘MoTA |esIOC

(oyeun

)

psourdsppnva

Ann. S. Afr. Mus., Vol. 55 Plate 4

1ew.

(adolescent). Dorsal v

dermis

mact

Aaa

B. R. sp

In.

in cm an

Scale

1ew.

le). Dorsal v

juveni

A.? R. spinacidermis (

Plate 5

Ann. S. Afr. Mus., Vol. 55

‘Ul pUv WO UT OTVOG *“MoIA [es

10d

‘(ajeun) vgyy ‘yg

"Ul pUL WO UT 9IPIG ‘MTA [es1og] *(ayeur aprusAnf

)

Dq]D

Ann. S. Afr. Mus., Vol. 55 Plate 6

B. R. straeleni (male). Dorsal view.

A. R. clavata (female). Dorsal view. Scale in cm and in.

‘ul pue

WI UT 9[VOG “MIIA [esto *(ayeutoy) snyaypurw say ‘@ = “UL pue wo ur a[eog ‘Mata [essoq ‘(ayeur a[tusAnl) vypipvs say vy

Plate 7

Ann. S. Afr. Mus., Vol. 55

Plate 8

Ann. S. Afr. Mus., Vol. 55

WO UT Beg “MATA [eIUDA “FGA L ‘ds ‘u “spuagos *y *g

‘ul pue

‘UI PUL WO Ul 9feoG “MTA [esIOd "|qA L ‘ds ‘u “wspiagou yy “WV

Plate 9

Ann. S. Afr. Mus., Vol. 55

UT OTIS *MOTA [e.I}UD A °

ddAL

“‘pynpian.s

d a

‘Ul pue Wd

“UT PUL UD UT 9[VIG “MaTA [es1Oq “AAT,

DINpIADA

Plate 10

Ann. S. Afr. Mus., Vol. 55

2[VIG “MOTA [VIIUIA “AGA, ‘ds ‘u

‘symussip “Yq

‘UI pue WO UT

“UL PUL UID UT I[VIG “MIIA [esIOT "YALL ‘ds ‘u ‘synwassp -y vw

Plate 11

Ann. S. Afr. Mus., Vol. 55

Q[VIG *MOTA [VIIA

dd AL

‘suapunfuoa

Ul pue WO UT

‘UI pUv WD UI B[vVOG “MoTIA [eSIOC] “AqAL,

‘ds

suapunfuoa

Plate 12

Ann. S. Afr. Mus., Vol. 55

‘MIA TeSIO(, “Ad AL ‘ds ‘u “vasvyjom “yg

‘UT PUB WD UI aeog “MOTA TesIOd *(I]eUIaF) supundoa) “y-W

Ann. S. Afr. Mus., Vol. 55 Plate 13

B. smithit (female). Dorsal view. Scale in cm and in.

INSTRUCTIONS TO AUTHORS

Based on

CONFERENCE OF BIOLOGICAL EDITORS, COMMITTEE ON FORM AND STYLE. 1960.

Style manual for biological journals. Washington: American Institute of Biological Sciences.

MANUSCRIPT

To be typewritten, double spaced, with good margins, arranged in the following order:

(1) Heading, consisting of informative but brief title, name(s) of author(s), address(es) of

author(s), number of illustrations (plates, figures, enumerated maps and tables) in the article.

(2) Contents. (3) The main text, divided into principal divisions with major headings; sub-

headings to be used sparingly and enumeration of headings to be avoided. (4) Summary.

(5) Acknowledgements. (6) References, as below. (7) Key to lettering of figure. (8) Explana-

tion to plates.

ILLUSTRATIONS

To be reducible to 12cm x 18cm (19cm including caption). A metric scale to appear with

all photographs.

REFERENCES

Harvard system (name and year) to be used: author’s name and year of publication given

in text; full references at the end of the article, arranged alphabetically by names, chronologi-

cally within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year.

For books give title in italics, edition, volume number, place of publication, publisher.

For journal articles give title of article, title of journal in italics (abbreviated according to the

World list of scientific periodicals. 4th ed. London: Butterworths, 1963), series in parentheses,

volume number, part number (only if independently paged) in parentheses, pagination.

Examples (note capitalization and punctuation)

BULLOUGH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FiscHER, P.-H. 1948. Données sur la résistance et de le vitalité des mollusques. 7. Conch., Paris

88: 100-140.

FiscHER, P.-H., Duvat, M. & Rarry, A. 1933. Etudes sur les échanges respiratoires des littorines.

Archs Zool. exp. gén. 74: 627-634.

Konn, A. J. 19602. Ecological notes on Conus (Mollusca: Gastropoda) in the Trincomalee region

of Ceylon. Ann. Mag. nat. Hist. (13) 2: 309-320.

Konn, A. J. 19600. Spawning behaviour, egg masses and larval development in Conus from the

Indian Ocean. Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. Jn scHULTZE. L,

Koologische und anthropologische Ergebnisse einer Forschungsreise 1m westlichen und zentralen Siid-

Afrika. 4: 269-270. Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16: 269-270.

ZOOLOGICAL NOMENCLATURE

To be governed by the rulings of the latest International code of zoological nomenclature issued

by the International Trust for Zoological Nomenclature (particularly articles 22 and 51).

The Harvard system of reference to be used in the synonymy lists, with the full references

incorporated in the list at the end of the article, and not given in contracted form in the synonymy

list.

Example

Scalaria coronata Lamarck, 1816: pl. 451, figs 5 a, b; Liste: 11. Turton, 1932: 80.

POEM Oy Muti’ vile Rab P fae f

By Ly ¢ /. | =u

j d

iW . ia i pty

baie OAS FSS te thefnie WY 7

nal =

riImirints : va NATTA Sfeles 28 88) oe BIR AI oe ee ee ee e-em ee —e — - IT ROC OAILRA AL IAIQ Te Tipriojal AiO tk OL LA. ALLO RIWIAIQCLE ET LiALC a

a @ — oo om w ‘Ee w one Lee]

il om > DW = > Ee: > be >

ae Fe a N = % = —_ = e =

mm LDS = ™m z py m On 0c = ; m = m

w = rp) eee ” — wn = . nw

IWS Sa iyvusg rou B RAR ] BS MITHSONIAN INSTITUTION | NOILNLIZSNI NVINOSHIINS = S3 iyvua ru! B RAR ! ESE

4 = 2 ‘y eae S — MQ ~. paar =< =|

Te? 4a NE a \SEE : & o 5

5 ty 4 RV y = oO

ye Gy - = = S = 2 : 2

Y = *¢ = > = = > >

= 5 . 2 z z i 2 =

AN INSTITUTION NOILNLILSNI NVINOSHLINS S3INVYSIT LIBRARIES SMITHSONIAN INSTITUTION NOILALILSNI

BRARIES

»S

NOILNLILSNI

LIBRARIES

NOILNLILSNI

IWS S3IYvyusi NVINOSHLINS S3tYVvVugil LIBRARIES

INSTITU"

SONIAN INSTITUTION NOILNILILSNI SMITHSONIAN

INSTITUTION NOILOLILSNI

SAtYVYdIT LIBRARIES

INSTITUTION NOILNLILSNI

z c z - =

Oo reed [@) eae APR

— wo a & ee] tf 00]

= y, 2 = We 20 Ey, =

., = [| = SASS = AW 1,° 73

2 _ - KGS ~ ¥ ~

\ S SS

n n .

m AN m m

= eh Zz Ss m rm

AN INSTITUTION NOILNLILSNI NVINOSHLINS S3I1NVYSIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S31YY

NVINOSHLINS S31YVYdiIT LI

” z w = oD)

= = &: = @,: : : : :

eS = Oi m= = pe = i

Js Mai Ye: iO SS gs NNE 72: :

BX SEND a od Paes Mh hy Sai S Ss

y = SX SS = LM ee = rao ai = = ENS = = = =.

zs 5 : - se : 3 :

IWS S3 lYVYdiIT LIBRARI ES SMITHSONIAN INSTITUTION te LSM NINOSHLINS _LIBRARI ES SMITHSONIAN _INSTITUT

> > 2 z wp 2

ff) z 507 us = us ” us 7 lu Wn OD

= 4 = 4 = = re 2 = Uy ®

N = = = Bee =f = aet ~j < i fo 7 A

S < “oe < e Pe: Ee = “fle

Wr 3 oi = me = =! fre a

Yom = om. = aa re) \ = o = re)

NAN INSTITUTION NOILOLILSNI NVINOSHLINS S3Ily¥vug sis ee ee nck ee 1yVvYg

“ 5 - «. §& = S .: = a

, «0 — o XN = = On ed bE WY, —

= NYG = Pe) eee x Es) y S SAM

py 2 = 2 WAR 5 ~ 5 YY ® 5 2 NAR 5 fs

ee Se eee Oe ee =

4 z | “” i 7) a

JES. SMITHSONIAN INSTITUTION NOILALILSNI_NVINOSHLIWS S31YvYal1_LIBRARIE

- 3S z SX 2

. a7.) a3 Ws = “

we . y fi, pace o af Xs o =-

efUf 2 < 2WK = 2

“iy i o0 a NS m =.

_ “i Oo — e) ~ Oo

z —! = ta aol =

| I NVINOSHLIWS $3 IYVEGIT LIBRARIES SMITHSONIAN INSTITUTION _ NOILNLILS!

;

EEX Oo a Oo ee O

ae x = od = Giy, > =

Ss = au = Ait =

> SANG E > = sp =

= XA E = , = oe

oe 2 m Q° ; g

Pi ow honk — a) = wn =

FES (SMITHSONIAN INSTITUTION NOILNLILSNI_ NVINOSHLIWS | eee hs

= = = | < ; = > sr

4 > ar jar z

= 5 : : 3 = 5 ‘NY

a 7 n CW a a:

9 a3 . oO as ) i Ars

= & =e = =. =

a ine a 2 ee

NI_NVIN SAIMVNGIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILS!

Z oo Z 7 z uw

2) ex ” a w ae

i _< a < oe | <

y t = A = :

Fl Oo sis ro) a o =

= J = — z= 4

1Es INSTITUTION NOILNLILSNI NVINOSHLINS S3iuvudi 7 LIBRARIE

| = RG 2 & | = fi

a i oe te a

3 4 3 a 4 > ea

= ¥ = = = = =

- NS = = E =

ie. m Nonny 2 1 2 m

a w = wn —— w

LSNI_NVINOSHLINS S3IYVYEIT LIBRARIES SMITHSONIAN INSTITUTION | NOILNLILS

ae: = Re a a et) gee =

xs : SQL Z :

C4 ‘ : S ~ :

ro an = O = O ?

oe = = = = =

| 2 a Te a BP ee:

HES SMITHSONIAN INSTITUTION NOILMIILSNI_NVINOSHIINS S31UVYSIT_LIBRARIE

= = rae mm

; uJ uJ * uu

2. oe fo = = 2 NS 5 =

oGwZy c : ZR = =

ys/ = Zz = = Ae ag aN = =

eee ae Se:

pt is ad

LSNI— ‘SAlYVYGIT LIBRARIES SMITHSONIAN_INSTITUTION NOILNLILS

ee = a S | a =

YON = = Z —

) 23 5 ay & Yip a :

4 ~— vu? is 17) =

TES SMITHSONIAN INSTITUTION NOILNLILSNI_NVINOSHLINS, S31YVus!T

| s a 7" Sa = gh, =

— F2 4 = 4 z

A as f,

ws a 3 AD os D

SS me = 3 sae re - :

Se = é : Boies

LSNI_NVINOSHLINS LIBRARIES SMITHSONIAN _INSTITUTION none a

= =

w u; ut a: uJ Yr J

ce rw Ws. < if sh {E 4S) < iy OR <