as oi = . a

w = mes) w Co xy

A Fas Pig a y (oad +7, v

2 a 2 es) = ey

i z m rm ” | a2 Th

w” = ae) 77) , & be.

ITUTION NOILALILSNI Saiuvugid SMITHSONIAN _ INSTITU

“. = < = < < <= wal

NO = 2 4 z= 4 > SN

. % aL O el oO oa oO <

3 2 gy zZ é 2

2 = Z = z fe

> = > ; = CP Ss ,

= Ps Ww z= ap) *, Pas, Ww

YVR LIBRARI ES SMITHSONIAN INSTITUTION NOILNLIL NI NVINOSHLINS S3Z1YY?

si Bs ul M us :

—| , a Led

c as <

4 = cc

Ss) st ue

z J in

ITUTION NOILALILSNI NVINOSHLINS S3IYVYUSIT LIBRARIES

INSTITUTION NOILONLILSNI

INSTITUTION NOILNLILSNI

Salyvudiy

INSTITUTION

sa1uvuag

NVINOSHLINS S3ItYVYdITLIBRARI

AWHAIT_ LIBRARIES SMITHSONIAN INSTITUTION NOILALILSNI NVINOSHLINS, Sa1uvd

: < @ % = z | g a

Pst Leal —

. + 2 & =

. WD _ ON CN io)

AS 2 bine Oo ¢

= Zz, = “

=> = —

a 2 a 2

ITUTION NOILALILSNI_NVINOSHLINS S31YVYGIT LIBRARIES SMITHSONIAN INSTITUT

NOILNLILSNI NVINOSHLINS SSIYVYSIT LIBRARIES SMITHSONIAN INS

TIT

”

n = ” Zz # 2

We = : ts 2 uJ yy i=

< e SQ SS ve e & lp foe

me Boe cern? Oo — o

a Zz ae am Zz

YWUd bik! BRARIES _SMITHSONIAN INSTITUTION” NOLLALILSNI” NVINOSHLINS S314

2 S o S —

Ps anh a 5 a 5

> EY > a fad

>) a AN ra OD sta

rm Zz a z oo Z

oP)

—

SMITHSONIAN

NVINOSHLIWS

RSS

NVINOSHLINS

LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S31YY

SMITHSONIAN

—

a ” = "” eae

& uw es Wl Zz uy

S oe. Ges. a a

ie < % a < — <x

| Z a : Z a z iw i:

ITUTION _ NOILALILSNI NVINOSHLIINS _S3 lyYvuyd rhe LIBRARI ES ~INSTIT J

oO el OF oO a fo) ee SON

ee (oe) s oun ow raat o pe).

5 a I 5 0 5 0 (WY

= a e= - 2 "2 YG

rs cone KE ype — Set) ~— Or

Wwreve se Se Clr wre eee 8 ee a aS Ve SSS eee

z - z a z Jt

fe) es, Oo a = o

- ca = ea - 9]

“ae > = > =

E re = 2 = pa

-_— w —

ES Z o ae a

- NWINOSHLINS S3IY¥WYUSIT LIBRARIES SMITHSONIAN NOILNIL!

= “ = 2 ” z w

2 = 2 ce < =

S a ae - z =

no ” y, As S w = i

Eo 2 @ = Zz = z

PS ~ a S = S

NOILNLILSNI NVINOSHLINS S31yvudi7_LIBRAI

“J

BRARIES SMITHSONIAN _ INSTITUTION

LIBRARIES

NOILOLILSNI

LIBRARIES

NVINOSHLINS S3IY¥VudIT LIBRARIES INSTITUTION NOILNL!

INSTITUTION NOILNLILSNI

S3IYVUSIT LIBRARIES

c iz =

wo s o =

2B Ys ; 2 5

= WW a =

3 TENS. = =

BRAR 1ES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3i¥VesiT_ Li B RAI

“s ” a n = wn : eet

= < = y yy < = Ss :

Ny 2 5 = I: 5 a an

> = ~ >” = > = =

2 ” Pes A ” = u) |

_NVINOSHLINS Sa IyVug iT Li BRARI ES SMITHSONIAN _INSTITUTION Peet

g : g = %,* :

4a VSS < = < Wi fas = <

G WN. : a fe & :

NSasnt~ Oo SS = ro 2k fo) =

y, za j ol z =| = ad

IBRARIES SMITHSONIAN INSTITUTION NOLLNLILSNI_NVINOSHLINS” $31 uvugi7 LIBRA

, = oa aie

6 ifs = = Ce @

Ee; = F ie = 2

aa a = 8] —

= wd E 0 = =

= m ” m W m

= w = ” i a

NVINOSHLINS S31NVUEIT LIBRARIES, SMITHSONIAN INSTITUTION, NO!LM

Z 2 2 & § a

= 4 = Ss », os z= NS a

Oo 5 O MW ye US = re) eS S63

n Ny 2 Nx. *- = Dm Nw =

= - = z = Z

= — : pra = >

a 2 a . 2 = Zz

SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLIWS Saruvugi7_“-IBRA

w > ” > ee ” =

<3 = < = Ws ¢ i

ts Cc c QA < =

\ = = a = FQ pa aa

ene 5 1 2

pA —_ s\ lg"

OILALILSNI NVINOSHLIWS (S31NVYaIT LIBRARIES_ SMITHSONIAN INSTITUTION _ NOILNI

a S z =e Sake G

a = _ es =

— a

: - _ — = fs

ras wo ae wm’ my v?

‘

5 <

ANNALS

OF THE

SOUTH AFRICAN MUSEUM

VOLUME 52

= ‘5

ail

> NY

d Peasy

‘ 4

ieanin

PRINTED FOR THE

TRUSTEES OF THE SOUTH AFRICAN MUSEUM

1968-1969

i i 4 ; Y =< Wyo 6 PE

"THE RUSTICA PRESS, PTY.,

LIST OF CONTRIBUTORS

Besr, P. B.

A dolphin (Stenella attenuata) from Durban, South Africa (published March 1969)

CARRINGTON, A. J. & KeEnszy, B. F.

Pleistocene molluscs from the Namaqualand coast (published July 1969)

CiLuveErR, M. A.

Korillodoniops, a new scaloposaurid from the Karoo (published May 1969)

Grrrin, D. J. G.

Two new species of Achaeus (Crustacea, Decapoda, Majidae) from South Africa

(published October 1968)

HeEnpeEY, H.

See HENDEY, Q. B.

HENDEY, Q. B.

The Melkbos site: an Upper Pleistocene fossil occurrence in the south-western Cape

Province (published December 1968)

HEnpeEy, Q. B. & HENDEy, H.

New Quaternary fossil sites near Swartklip, Cape Province (published October 1968)

Rorrey, P. A.

The relationship between Raja miraletus Linnaeus and Raja ocellifera Regan based on a

study of the clasper (published March 1969)

KeEwnsLEy, B. F.

Decapod Crustacea from the south-west Indian ocean (published April 1969)

KEnSLEY, B. F.

See CARRINGTON, A. J.

PATERSON, N. F.

The behaviour of captive Cape rock ee asus lalandit ee Milne aeaee

(published November 1969)

PENRITH. M. J.

See TALBOT, F. H.

TALBOT, F. H. & PENnrRitTH, M. J.

~The tunas of the genus Thunnus in South African waters (published October 1968)

Page

I2!I

189

183

Us,

149

225

N GENERIC NAMES PROPOSED IN THIS VOLUME

oR. Namamurex Carrington & Kensley, 1969 (Muricidae), 197

___ Korillodontops Cluver, 1969 (Scaloposauridae), 187

" \

f =

—— z

‘

;

Acanthina

Acar

Acetes

Achaeopsis

Achaeus

Actaea

Reeth

Aloidis

Alpheus

Anancus

Anapagurus

Antidorcas

Arca

Re idickodon

Arctocephalus

Aulacomya

Auxis ..

Axius ..

Bathyergus

Brama

Calappa

Callianassa

Calliostoma

Calyptraea

Cambarus

Cancellaria

Canis ..

Carcinoplax ..

Carditella

Carditopsis

Ceratostoma .

Ceratotherium

Chamelea

Charybdis

Chlorotocus

Cistecephalus

Clanculus

Clinus

Conchoecetes

Connochaetes. .

Crocuta

Guna ..

INDEX TO GENERA AND SUBGENERA

(Synonyms in italics)

75:

154, 172;

56, 71, 94,

207;

94,

227,

F935

67, 68, 94,

201,

65, 94,

187,

36 ‘Oo,

210

198

190

I5I

154

188

203

256

15!

III

. 48, 96

210

Dardanus

Delphinus

Diceros

Diogenes

Donax

Dorippe

Eastonia

Ebalia

Ecklonia

Endothiodon .

Equus

Ericiolacerta

Ethusa

Eualus. .

Eumedonus

Eurynome

Euthynnus

Fasciolaria

Felis

Fissurella

Fusus ..

Galathea

Gastrana

Gennadas

Germo ..

Goneplax

Gonioneptunus

Gorgon

Herpestes

Hespererato

Heterocarpus

Hipparion

Hippolysmata

Hippopotamus

Hippotragus ..

Homarus

Homoioceras ..

Homola

PAGE

: 124, 126,

61, 62, 94, 102,

153, 248,

IQI,

210,

62, 94,

186,

151,

age Bi | 63, 64, 945 97

190, 191,

154,

152

129

103

252

213

151

222

I5I

227

187

103

187

161

154

155

I51

193

206

194.

153

218

167

Fe Os0t2

154,

61, 94,

54, 94,

248,

Hyaena

Hyastenus

Hystrix

Ictidodraco

Ictidostoma

Ictidosuchoides

Ictidosuchus ..

Inachus

Inachus |

Iphigenia

Jasus

Jaton ..

Katsuwonus ..

Kishinoella

Latiaxis

Latreutes

Leptochela

Leucosia

Loligo

Lophozozymus

Loxodonta

Lupocyclus

Lycaon

Lystrosaurus ..

Macropetasma

Macropodia ..

Makaira

Maurolicus

Meganteron ..

Mellivora

Merluccius

Merope

Mesochoerus

Metapenaeopsis

Munida

Namamurex ..

Nassa

Naucrates

Navicula

Neaxius

Nematopagurus

Neothunnus

Nikoides

Nursilia

INDEX TO GENERA AND SUBGENERA

; PAGE

48, 62, 94, 95

195, 197

a aA.

154, 168

151

2, 25

_ 152

94, 102

ro yy ge

- — 49; 64, 65

183, 187, 188

sie 154

75, 81, 152

154

152, 162

Oplophorus ..

Orchestia

Orconectes

Pagurus

Palicus

Palinurus

Pandalus

Panthera

Panulirus

Parapagurus ..

Parapenaeus ..

Paratergatis

Parathunnus

Patella

Penaeopsis

Penaeus

Periclimenes .

Petrolisthes

Philyra

Pilumnus

Plagusia

Platylambrus

Platypodia

Plesionika

Pomadasys

Pontocaris

Porcellana

Portumnus

Prionace

Processa

Prodelphinus . .

Proteles

Proterato

Pterorytis

Pylopagurus .

Raja

Ranina

Raphicerus

Redunca

Retropluma

Sardinops

Scaloporhinus

Scaloposaurus

Scomber

Scylacosaurus. .

Sergestes

Silpholestes

231-4, 2360) (24a ene—a,

PAGE

154, 169

246

248

153

i 152, 156

226, 252, 259, 260

‘ 170

255, 258-260

153

154

152, 156

oe 7

227

154

153, 165

152

256

152

155, 170

ae 2

155

153

152

me, 5

155, 172

134

201

198

153

137

61, 94, III

51, 94, 110

152, 158

2, 4, 25

184, 187

183

8, 10, 12

186

Buin 5:

187, 188

Simopithecus

Solenocera

Standella

Stegolophodon

Stenella ;

Steno ..

Struthio a

Stylodactylus

Surcula

Synalpheus

Syncerus

Tapinocephalus

Taurotragus ..

Tetrapturus

INDEX TO GENERA AND SUBGENERA

PAGE

117

eval: 15775

210, 222

113

wal ae) DT

126, 134

68, 94, 113

we a LG

193

Lipa L 5S

94, 104, 110

186

94, 109

Thalamita

Thunnus

Thynnus

Tozeuma

Tragelaphus ..

Tritonalia

Trochosuchus. .

Turris

Xanthias

Zorillodontops

PAGE

152

Licey)

piggy

155

94, 108

196, 222

186

193

152

183, 187

a yD

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 52 ~&Band

October 1968 Oktober

Part \?) a Deel

THE TUNAS OF THE GENUS THUNNUS

IN SOUTH AFRICAN WATERS

F. H. TALBOT & M. J. PENRITH

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

(Cash with order, post free)

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

(Kontant met bestelling, posvry)

OUT OF PRINT/UIT DRUK

I, 2(1, 3> 3» 7-8), BZ. 5» t.—p.1.), 5(2, 5> 7-9),

6(1, t.—p.1.), 71, 3). 8, g(1-2), 10(1-3),

I1(1-2, 7s EOpils)) 5 21, 24(2), 27) 31(1-3), 38,

44(4).

Price of this part/Prys van hierdie deel

R2.00

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

1968

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 TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN

WATERS.

PART I. INTRODUCTION, SYSTEMATICS, DISTRIBUTION AND

MIGRATIONS

F. H. TALsot

Australian Museum, Sydney*

M. J. PENRITH

South African Museum, Cape Townt

(With 8 figures)

[MS. received 19 August 1964]

CONTENTS

PAGE

Introduction : : : 5 ae |

Systematics ie ie ML Ha a

Distribution and migrations . 14

Acknowledgements . : Soa

References ta ae Ge

INTRODUCTION

In South Africa research on the tunas lagged far behind that in European

countries, the Americas, and especially Japan. Prior to 1945 tunas were con-

sidered by ichthyologists to be only rare migrants to South Africa.

Tunas were first recorded from South Africa by Giinther (1860). Other

early records were those of Gilchrist (1902) and Thompson (1918), and later

Barnard (1927), Biden (1930), Smith (1935) and Barnard (1939). All con-

sidered tunas to be rare in South Africa; only Biden suggested that they might

be present in greater quantities than indicated by the records.

In March 1941, tuna were reported to be common in False Bay (newspaper

reports), but the species was not identified. In December 1945, large numbers

of tuna reported to be small Thunnus thynnus (newspaper reports in which the

fishes were stated to have been identified by Molteno) were caught from the

shore near Cape Point, False Bay. Sport fishing recovered after the war, and by

1952 fishing for tuna off the Cape Peninsula was firmly established. Several

thousands of troll-caught tuna were being landed each summer season. Up to

* Formerly of the South African Museum, Cape Town.

t+ Research Officer, Council for Scientific and Industrial Research, seconded from the Institute

of Oceanography, University of Cape Town.

Ann S. Afr. Mus. 52 (1), 1968: 1-41, 8 figs.

2 ANNALS OF THE SOUTH AFRICAN MUSEUM

1960 there was no commercial exploitation (Talbot & Penrith, 1960). »

At the time of the beginning of the present survey (1960) the possibility

of subsurface populations being present was not known, tuna were thought to be

absent from the Cape in winter, and nothing was known of their biology from

this area. Molteno (1948) had published a useful list of species known and to be

expected from South Africa, with biological data based on work in other parts of

the world, and Smith (1949) considered the species known at that time, giving

a key for their identification in his treatise on South African marine fishes.

The present survey was conducted in two parts, as an eighteen-month

survey in the south-western Cape (1960-1) and a two-month survey in the

south-western Indian ocean (August 1962; February 1963). A ship, crew, and

equipment were supplied by Irvin & Johnson Ltd. for the Cape survey, and by

the South African Navy and the South African Council for Scientific and

Industrial Research for the Indian ocean. Scientists were supplied by the

South African Museum and the Council for Scientific and Industrial Research.

It was decided that of the four possible methods for obtaining tuna samples

(surface trolling, pole-and-line fishing, long-lining, and purse seining) only

long-lining would be used, as this method required the least ship modification.

Aims of the survey

It was decided that priority should be given to the following problems,

which were considered to be of particular interest:

(a) Systematics. In 1960 the taxonomy of tuna throughout the world was

very confused; for example, the Atlantic and Indo-Pacific yellowfin, which

were indistinguishable when placed side by side, were put in different genera.

This was the case with many of the tunas, and matters had reached such a

point, where, for clarity, the common names were being used rather than the

scientific. It was felt that specimens from the Cape, being situated at the

boundary between the Indo-Pacific and Atlantic oceans, could be of value in

solving some of these systematic problems.

(b) Distribution. As three very different water masses meet at the Cape,

attention was aimed at local distribution, and the relationship between distri-

bution and hydrographic factors.

South African waters. Troll-caught tuna fed mainly on pilchards (Sardinops

ocellata) and squid (Loligo reynaudi) (personal observations) and the only

published statements were by Horne (1959) on the finding of a grunter (Poma-

dasys operculare Playfair) in the stomach of a large Thunnus thynnus netted in

False Bay. ‘There was no knowledge of possible food preferences and competition

between species in this area.

_ (d) Breeding. Nothing was known about the breeding of the tuna found

off South Africa. The breeding times and size of maturity were unknown; it

was not even known whether tuna did breed in South African waters. From the

findings of sports fishermen it was assumed that the tuna left to breed elsewhere

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS 3

15° 1

13°

ee :

6 On '

1° 33 . 37°

i 35

We 19° 27°

Fig. 1. Area of survey. Each dot represents one station, except off Slangkop, Cape Peninsula,

where the solid line represents 51 stations.

during the southern winter (May—September), as no tuna were caught during

these months. It was hoped that detailed examination of the gonads could give

give some indication of when the breeding period occurred, and possibly

whether breeding took place locally or not.

(ce) Age and growth. It was proposed to use the length frequency method

of determining growth, and to examine scales or some part of the skeleton

(e.g. otoliths or vertebrae) in an attempt to find growth rings which could be

correlated with any modes found in the length frequency curves.

The results bearing on (a) and (b) above are reported on in this paper.

4 ANNALS OF THE SOUTH AFRICAN MUSEUM

Area of study

It was decided that, rather than cover a wide area in an overall search for

subsurface fishes, it would be more valuable to concentrate the survey in one

area, so that any changes during the year would be more comparable. The

survey was concentrated in a line running out magnetic west from Slangkop for

two reasons: firstly, surface tuna were known to occur in the area in large

numbers at certain times of the year (Horne, 1954), and secondly, the Univer-

sity of Cape Town research vessel John D. Gilchrist had a monthly series of

hydrographic and plankton stations on the same line. It was also known that

sudden changes in surface temperature occur along this line, a factor widely

known as a good indicator for tuna (Thompson, 1917; Powell eé¢ al., 1952;

Murphy & Shomura, 1955). Figure 1 is a map of the area surveyed, indicating

the stations at which long-lines were fished.

Sixteen cruises of between four and fifteen days’ duration were made.

During four of the cruises fishing was undertaken off the east coast, but on all

four occasions the fishing was severely hampered by storms. In addition to these

cruises, two cruises, each of a month, were made in the south-west Indian

Ocean, one in summer and one in winter. A complete list of stations is given in

Appendix A.

Description of methods used

One basket of gear with components and lengths is sketched in figure 2.

Each basket was 200 fathoms long. 8 mm Kuralon was used for the main line

Fig. 2. Diagrammatic representation of one basket of long-line as used in the present survey.

, Lengths in fathoms.

and droppers. In order to study depth distribution the normal catenary curve

of the line was accentuated by keeping the outer droppers short, and lengthening

the middle droppers.

Setting was begun before dawn (5.00—-6.30 hours) throughout the year,

as tuna seem to be mainly dawn and dusk feeders (Nakamura, 1949; Powell,

1950; Powell ef al., 1952). Bait used was whole pilchards (Sardinops ocellata),

frozen or lightly salted. Usually twenty baskets of long-line were set at a time,

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS 5

this taking some forty minutes.

At the completion of setting, a bathythermograph cast was made, water

samples and temperatures taken. Three water samples and temperatures were

made, at 0, 30 and 150 metres, using a Petterson-Nansen insulated bottle fitted

with a Negretti and Zambra thermometer. The two subsurface depths were

chosen as being at above the level of the shallowest and deepest hooks. Salinity

determinations were made by the Division of Sea Fisheries using a salinometer.

Hauling of the long-line was done at about 10.30 a.m., after the line had been in

the water for about 4-5 hours. The condition of each hook was listed as it was

hauled (baited, no bait, fish).

Immediately after completion of hauling the fish were all washed, spread

out on the deck, and the identification in the hook log checked. The fork-length

of all fish was then measured using a 2-5 metre long pair of calipers, and mor-

phometric measurements according to the method of Marr & Schaefer (1949)

were made on as many fish as possible. Measurements other than fork-length

were not made on albacore (T. alalunga).*

The gonads and stomach contents of all fish (other than shark-damaged

specimens) were removed and preserved. Formalin of 10% strength was used

so as to stop the process of digestion as rapidly as possible and to harden the

forage organisms which had become soft due to digestion.

Notes on sampling

Selectivity

No fishing gear is completely unselective, and the long-line is no exception.

The factors that might influence selection are discussed below.

(a) Hook size

The Japanese hook (size 3, 6 cm high, 3 cm wide) was chosen because it

was the middle of the size range offered. It was found, however, that hook size

is only of minor importance, as tuna from 5 to 400 lb were landed, as well as

other larger fish which had much larger mouths than any tuna (large sharks,

e.g. Prionace glauca and marlins, Makaira spp., of up to 1,460 lb), and excep-

tionally small-mouthed fishes such as Brama rai and a pilot fish Naucrates ductor

weighing less than 1 Ib. It is not possible to disprove selection due to hook size,

but the wide size range of fishes taken suggests it was not a major factor.

(b) Sampling depth

There was a very marked selection due to the sampling depth. Surface

trolling by game-fishing vessels off the Cape Peninsula results in a catch con-

sisting almost exclusively of yellowfin tuna (7. albacares), whereas in the survey

T. albacares accounted for only 11-6% of the total tuna catch, and in the Cape

* In South Africa T. alalunga has the common name longfin tuna. The name albacore is used

for the carangid Seriola lalandii.

6 ANNALS OF THE SOUTH AFRICAN MUSEUM

Peninsula area 8-8%. Conversely, the long-line took large numbers of other

tuna species that had only rarely been taken by surface trolling. ‘This makes it

difficult to attempt to assess the true relative abundance of the various species,

but subsurface long-lining did capture fair numbers of four species, whereas

surface trolling or pole-and-line fishing could be expected to take only two

species. Long-lining also in general captures larger fish than surface trolling;

T. alalunga of over 950 mm fork-length are rare using the latter method but are

commonly taken on long-lines. Presumably this is due to swimming depth of

various sizes of tunas.

By its nature a baited line will normally only take feeding fish. Sampling

error caused by this, and possibly also by seasonal changes in food require-

ments, could not be assessed.

Faults and advantages of the sampling method used

The major fault of long-lines for obtaining samples of tuna is that only the

subsurface populations are sampled. Surface trolling has shown much larger

populations of 7. albacares in the area surveyed than the results from the long-

lining.

Another disadvantage was the lack of any knowledge of actual depth of

capture. The work of Kamimura (1957) has suggested that most tuna are

caught at a depth shallower than the maximum to which the hook sinks. It was

not known in the present survey exactly how deeply the hooks did sink. Some

workers, especially at the U.S. Bureau of Commercial Fisheries at Hawaii, used

a rangefinder or radar to measure the distance between the buoys and, knowing

the length of the line between them, could calculate the catenary of the line.

This, however, does not take into consideration the effects of bowing or slanting

of the line due to subsurface currents or the action of hooked fish. It was planned

to attach Kelvin sounding tubes to certain of the droppers to find the catenary

of the line, but their use was discontinued owing to the very contradictory

results obtained (see also Murphy & Shomura, 1953). Graham & Stewart

(1958) have shown that the line may go deeper during hauling than it normally

hangs. Attempts to locate the line by means of echo-sounding were unsuccessful.

On one occasion off Port Elizabeth several of the hooks did touch the bottom,

bringing up bits of coral, from 78 to 80 fathoms, showing that the deepest hooks

at least reach this depth.

It was assumed that the outside hooks of the basket with 5-fathom droppers

fished at 15-20 fathoms, the 10-fathom droppers at 35-45 fathoms, and the long

centre dropper of 20 fathoms at 75-80 fathoms.

In general it can be stated that the long-line, in spite of the above dis-

advantages, is probably the best method of estimating the abundance and size-

range of large subsurface fishes.

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS i}

SYSTEMATICS

Starks (1910) was the first to notice the presence of a prootic pit in the skull

of certain of the Scombridae, and proposed the subfamily Thunninae to contain

the genera Thunnus, Auxis, Katsuwonus and Euthynnus, which possessed it.

Kishinouye (1923), after his careful anatomical studies on the Scombridae, on

the basis of the prootic pit and the subcutaneous blood system which he found

to be present in the fishes contained in Starks’s Thunninae, proposed that these

four genera (much subdivided by him), should be separated into the order

Plecostei, having equal rank with the order Teleostei, which would contain all

the other bony fishes. It is perhaps best to consider Kishinouye’s work as a

brilliant monograph on scombrid anatomy, but to ignore the final section on

classification. Berg (1940) in his definitive work unfortunately leaned heavily on

Kishinouye. Berg’s order Thunniformes (being Kishinouye’s Plecostei) is

removed from the suborder Scombroidei in the order Perciformes. This step is

not quite so drastic as that proposed by Kishinouye, but any attempt to separate

the tunas and mackerels into even separate families does not seem warranted.

Since 1950 several reviews of the Scombridae have appeared. Fraser-

Brunner (1950) proposed a straightforward classification, having two sub-

families, Gastrochismatinae and Scombrinae, and sinking many of Kishinouye’s

genera to subgeneric level. Ginsburg (1953) and Godsil (1954) agreed with

Fraser-Brunner that the Thunniformes and Scombroidei (of Berg) were so

closely related that they should be kept in one family. This has been done also

by Collette & Gibbs (19632, 19635), and has been followed here.

Many generic names have been suggested for the large tunas, and

T. alalunga, T. obesus, T. albacares and T. tonggol have all been placed in mono-

typic genera at various times. Most authors now consider that the genera

Germo, Neothunnus, Kishinoella and Parathunnus should be dropped (Fraser-

Brunner, 1950; Ginsburg, 1953; Godsil, 1954; De Sylva, 1955; Collette &

Gibbs, 1963a) but the names have remained in general use because they were

the names used by many tuna biologists in Japan and the United States. There

are a number of reasons for placing all the large tunas in a single genus, the main

one being the arbitrary nature of any divisions within the group, depending

upon what characters are used. The following choices of grouping indicate this.

Watson (1963) has shown on the basis of osteology that there can be two groups,

T. obesus being similar to T. thynnus and T. alalunga, while T. albacares is similar

to T. tonggol and T. atlanticus. De Sylva has suggested, also on the basis of

osteology, that TY. albacares and T. obesus are closely related. Beebe & Tee-

Van (1936) claimed on the basis of external morphology that T. obesus and

T. atlanticus are closely related. This indicates that such groupings give little

information on the phyletic history of the tunas. A single genus with several

species seems the best arrangement of the group.

8 ANNALS OF THE SOUTH AFRICAN MUSEUM

Genus THUNNUs South, 1845

Description: Body robust; abdomen without median groove for pelvic fins;

adults with body completely scaled, anterior scales larger, forming a corselet;

juveniles not necessarily fully scaled (Moore, 1951); interpelvic process bifid;

teeth small, conical, in a single row in each jaw; vomer and palatines toothed;

no strongly developed adipose eyelids; well developed lateral keel on each side

of caudal peduncle and two smaller keels on tail above and below hind end of

larger keel; gill-rakers 20-42 (both arches); dorsal fins almost contiguous;

dorsal spines 13-15, rays 22-23, some posterior rays separated as separate

finlets; anal rays 21-22, with posterior finlets as in dorsal; vertebrae 39, first

reduced and firmly articulated to the skull; well developed subcutaneous

blood-vessels on sides; skull with deep prootic pits.

The number of species in the genus is still in some doubt (Rivas, 1961;

Roedel & Fitch, 1961) but a number of authors such as Collette (1961),

Collette & Gibbs (1963a), Mather (1963a), and Watson (1963) consider that

there are six species. Here we provisionally recognize seven species, of which

five occur in South Africa waters.

Thunnus albacares (Bonnaterre)

Yellowfin

Scomber albacares Bonnaterre, 1788:140.

Neothunnus itosibi: Smith, 1935:207. Molteno, 1948:33.

Germo albacora: Molteno, 1948:20. Smith, 1949:299.

Germo itosibi: Smith, 194.9:299.

Thunnus albacares: Talbot & Penrith, 1960:1, 1961a:1, 1962:558, 19634a:617. Talbot, 1964:205.

De Jager, 1963 :589. De Jager e¢ al., 1963:11.

Liver without striations on its ventral surface, and with the right lobe

longer than the central and left lobes. Pectoral fins long, reaching to or beyond

the origin of the soft dorsal except in large individuals. Dorsal and anal lobes

exhibiting marked allometric growth and becoming extremely elongated in

large specimens, reaching even beyond the caudal fin. Gill-rakers 11 + 19-20.

(See table 1, for morphometric data.) Swim bladder present. A dense rod of

connective tissue running down the dorsal wall of the body cavity presses into

Thunnus albacares

Bork-leneth 32) 9) coy): 6. Oe 614 935 1145 1230 1505

Height D? Sea Tan ae 65 79 151 215 225 514

Heiphtvanal me V4). 68 a] 154 218 236 594

iheneth pectoral’ 75) 3 4) aya 190 266 318 315 371

Head a iis 5 (Ae? . 4 169 175 253 300 312 370

Snout— Dt eh He Sy ta eT 194 273 329 350 399

Snout— D+ Ons, Mae gee YG. 341 491 601 628 738

Snout— anal aro) Mey 1 yr ct eT Gy 371 531 654. 695 832

Deptaye Mamaee wie sie Vue 158 245 300 327 366

Length maxilla Fy (Malin igh 65 71 102 118 120 148

TABLE 1. Morphometric data for Thunnus albacares caught off the Cape Peninsula.

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS 9

the swim bladder. Body shape, when compared with T. obesus, less deep.

Reaches about 400 lb in weight. Colour royal blue on back, white or silver on

belly, separated by a lateral golden stripe some 5 cm wide in a fish of 1 metre.

Dorsal lobe dark at its base, golden yellow above this, with a white tip in large

specimens. Anal lobe similar to dorsal, but with a silver-white base. Dorsal and

anal finlets bright orange yellow, with or without a thin black border.

This species has an extensive synonymy, due in large measure to the

marked allometric growth of the dorsal and anal lobes. This aspect has been

discussed fully by Cunningham (1910), Kishinouye (1923), Beebe & Tee-Van

(1936), Schaefer & Walford (1950), Schaefer (1952), and particularly by

Royce (1961, 1965).

Following Collette (1961), Collette & Gibbs (19632), Royce (1961, 1965)

we recognize one world-wide species of yellowfin tuna.

A consideration of the taxonomic status of the bluefin tunas

We have decided not to consider the southern Indo-Pacific and Atlantic

bluefin tuna as subspecies of Thunnus thynnus, as suggested by Collette & Gibbs

(1963a), and previously by ourselves (Talbot, 1964; Talbot & Penrith, 19632).

We have done this after a careful consideration of the position in South African

waters, where these two forms overlap in distribution.

There are clearly at least three series of populations, one series being found

in the Atlantic and Mediterranean, another in the northern Indian and Pacific

oceans and a third in the southern Atlantic and Indo-Pacific. They can be dis-

tinguished from each other in almost all individuals on a number of characters,

including gill-raker counts, skeletal differences, and coloration.

The large bluefins found in summer at the Cape have the typical charac-

teristics of the populations known from the Atlantic Ocean, the clearest being

a high gill-raker count. The large populations of smaller bluefin found in the

area in winter have the characteristics of southern Indo-Pacific fishes (low

gill-raker counts and bright yellow peduncle keels). It is therefore likely that

these fishes belong to the North Atlantic and southern Indo-Pacific groups of

populations respectively and return to them. Until these fish are actually

marked and recovered this must remain an assumption.

Given this assumption there is a reasonable case for separating the bluefins

at the Cape into two species. Gill-raker counts, coloration, and skeletal

differences are clearly marked between the summer and winter populations.

No merging of differences or different combinations of characters are found,

although the two forms are sympatric at one time of year. Two isolating

mechanisms seem to have developed. There is a difference in breeding times,

with the southern fishes feeding at the Cape in the southern winter, with com-

pletely inactive gonads, so apparently breeding in the southern summer, and

with the northern fishes present in the Cape in the southern summer with again

completely inactive gonads, and apparently breeding elsewhere in the southern

winter. These characteristics indicate that different behavioural characteristics

IO ANNALS OF THE SOUTH AFRICAN MUSEUM

have developed which keep them genetically apart in spite of their sympatry

during some part of the year at the Cape. We consider that there is sufficient

genetic isolation, causing visible morphological differences, to justify accepting

the two bluefins found at the Cape as separate species, although the division of

the bluefins is clearly of more recent date than that of the other species.

The earliest name available for a southern ocean species is Thunnus maccoyiu

(Castelnau, 1877).

The position of the third bluefin, the northern Indo-Pacific species, at times

referred to as T. orientalis, is beyond the scope of this paper. Nakamura (1965)

appears to consider it to be no more than a subspecies of J. thynnus, a position

retained in the authoritative work of Gibbs & Collette (1967).

Thunnus thynnus (Linnaeus)

Atlantic Bluefin Tuna

Scomber thynnus Linnaeus, 1758: 297.

Thunnus thynnus: Barnard, 1927:798. Molteno, 1948:16. Smith, 1949:298. Talbot & Penrith,

1962:558, 1963a:637.

Entire ventral surface of liver covered with blood-vessels forming dense

striations. Pectoral relatively short, not reaching past the eleventh or twelfth

dorsal spine. Length of dorsal and anal lobes short. Gill-rakers (South African

specimens) 12 — 15 + 26 — 31, total 37 — 44 (see table 2). (Morphometric

data given in table 3.)

Gill-rakers 31 32) 33 34 \35 36 37 38 39°40 41) \)42RABRes

T. maccoyit I 2 3 5 I I

7. thynnus I I I

TABLE 2 (South African specimens only)

Colour of body dark blue-black above, with white below, in life or freshly dead

said to be separated by a yellow or iridescent blue band (Migdalski, 1958;

Tiews, 1963). Pectoral fin dark silver at base, dark grey or black distally with

red patches. Dorsal lobe dark, with yellow and red near tip. Anal lobe silver

with red and yellow as dorsal lobe. Finlets yellow with broad black edging.

Caudal fin dark with red or yellow posterior edge. Peduncle keels dark. Reaches

1,600 lb (Crane, 1936), and reputedly 2,000 lb. South African specimens were

all adult.

Thunnus thynnus

Fork-length . : d Q151 2671 Snout— D1 ‘ k 585 74.0

HeightsO2y Bi, 8s 362 4.22 Snout— D? ey | ae 1075 1345

Height anal ‘ ‘ 353 366 Snout—anal_ . : ‘ 1220 1537

Length pectoral .. 402 400 Depth Pele oi -0h i : 522 695

dp ef UPN 558 675 Length maxilla re 220 258

TABLE 3. Morphometric data for Thunnus thynnus caught off the Cape Peninsula.

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS ital

The distribution and taxonomy of Atlantic bluefins have been the subject

of much controversy. Storer (1867) differentiated western Atlantic fishes from

those on the European side under the name T. secundodorsalis. The validity of

secundodorsalis was questioned by Jordan & Everman (1926) and Godsil &

Holmberg (1950), but they did not formally suppress the name. Ginsburg

(1953), however, claimed the two populations were distinct on morphometric

and meristic grounds. His meristic characters show slight differences in means,

but it is not certain whether these would disappear with larger samples.

Rivas (1954) considered there to be one Atlantic species, with two separate

breeding populations racially or subspecifically distinct. Sella’s (1927, 1929)

suggestions that an exchange of fish took place across the Atlantic has now been

thoroughly proved by Mather’s tagging studies (1960, 1962, 1963). Tagging

returns indicate that JT. thynnus can cross from America to Europe, and is,

therefore, quite capable of swimming to the Cape.

Thunnus maccoyu (Castelnau)

Southern Bluefin Tuna

Thynnus maccoyii Castelnau, 1872:109.

Thunnus maccoyit: Nakamura, 1965:18. Penrith, 1967 :535

Thunnus thynnus (non Linnaeus) Talbot & Penrith, 1962:558.

Thunnus thynnus maccoyii: Serventy, 1956a:13.

Thunnus thynnus orientalis non Temminck & Schlegel, 1844:94. Talbot & Penrith, 1963a:630.

Talbot, 1964:201. De Jager, 1963:589. De Jager ef al., 1963:11.

Liver completely striated ventrally. Pectoral short, scarcely reaching the

origin of the second dorsal. Dorsal and anal lobes short. (Table 4 gives morpho-

Thunnus maccoyit

Fork-length . : : i : i ; 1045 1202 1235 1351

Head . : é : : : : 30°4 30°2 29°2 29°1

Snout— D! : : 3 : ; : 31°5 81-7 33-1 30°4

Snout— D? , : 3 , : 55°3 54°7 55°7 53°9

Snout anal. 3 : f : 60-2 60°3 63:4 58-7

Loe nn ee 26.7 26°5 26-2 26-9

Length pectoral . : : ; f : CLT 20°4 23°6 21-6

Height D? ; ; ; : : : 12°9 12-6 14°! 11-6

Height anal _ . : : : ; : é 13°3 13°0 13°6 11-6

Length maxilla a cs et on ee 122 ee: TOeT 11-3

Gill-rakers et) eth, 6s ee a ee 33 34. 33 34.

TABLE 4. Morphometric data for Thunnus maccoyii caught off the Cape Peninsula.

metric data.) Gill-rakers 30-38 in total, but usually 33-34 (see table 2). Swim

bladder variable, usually being better developed in large fishes. Colouring in

general similar to T. thynnus, but no trace of yellow on the dorsal and anal lobes

and the pelvic in fishes examined by us. The peduncle keels are usually bright

yellow, sometimes darker, but always with a translucent, almost hyaline, look,

unlike the solid black of T. thynnus. Finlets yellow with a thin black edge.

12 ANNALS OF THE SOUTH AFRICAN MUSEUM

Thunnus alalunga (Bonnaterre)

Longfin, Albacore

Scomber alalunga Bonnaterre, 1788:139 (based on a description by Cetti, 1777).

Thynnus alalonga: Giinther, 1860:366. Gilchrist, 1902:128. Thompson, 1918:108.

Germo alalunga: Smith, 1949:299. Molteno, 1948:29.

Germo alalonga: Barnard, 1927:799.

Thunnus alalunga: Talbot & Penrith, 1960:1, 1962:558, 1963a:609. Talbot, 1964:199. De Jager,

1963 :589. De Jager et al., 1963:11.

Ventral surface of liver densely covered with striations. Pectoral fins

sabre-shaped and elongate, reaching to or beyond the first anal finlet. Dorsal

and anal lobes low. Gill-rakers 8-9 upper, 19-21 lower, 28-30 total. Back dark

metallic blue, belly silver-white, separated by an iridescent blue band, which

disappears after the fish is gaffed. Pectoral fin black, ventral and dorsal lobes

dusky. First dorsal lobe colourless, anal lobe silver. Posterior edge of caudal fin

white, dorsal finlets pale yellow with dusky edges, ventral finlets dark. Reaches

80 lb, usually smaller. Flesh paler than other species. (See table 5.)

Thunnus alalunga

Fork-length : f ' ; 818 864 892 975 993

Height dorsal? Baran as Ch 93 III 118 128 128

Height anal : : : ‘ 04. 107 122 134. 130

Length pectoral : , : 352 387 4.09 397 408

Head : d : ; 233 254. 273 290 289

Snout— dorsal? f ? : 271 287 320 232 321

Snout— dorsal? : . : 482 502 574 573 583

Snout—anal_. : : : 514 535 610 631 631

Depth : ! 5 : : 203 220 257 Pai) 255

Length maxilla : . ; 90 97 104 12 105

Gill-rakers:

LI PEL Yo) ike Vee we ee 20 aI 21 QI 20

lower 4 : : f 10 9 9 9 10

Total , : : : 30 30 30 30 30

TABLE 5. Morphometric data for Thunnus alalunga caught off the Cape Peninsula.

This is considered to be one widely ranging species and while little is

known of its movements in the Indian or Atlantic oceans, it has been shown to

make trans-Pacific migrations (Blunt, 1954; Otsu, 1960; Uchida & Otsu,

1961; Clemens, 1961, 1963).

Thunnus obesus (Lowe)

Bigeye

Thynnus obesus Lowe, 1839:78.

Thunnus obesus: Talbot & Penrith, 1960:1, 1961a:1, 19615:240, 1962:558, 1963a:624. Talbot,

1964.:203. De Jager, 1963:589. De Jager et al., 1963:11.

Liver showing patches of peripheral striations. Pectoral fin one quarter to

one sixth of total length, reaching the end of the dorsal lobes, becoming relatively

shorter with increased size. Dorsal and anal lobes short compared with T. alba-

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS 13

Thunnus obesus

Fork-length : : : ah) 7S 1356 1483 1582 1784

Height D? : as 2 : 170 223 220 285 301

Height anal : : : 175 214 213 271 290

Length pectoral : : : 319 356 369 410 422

Head SE a oie oe! 911387 391 395 446 493

Snout—D? ont a ia 362 401 447 469 518

Snout— D? ‘ : : : 652 747 828 836 946

Snout—anal_. : : ; 652 853 924 937 _ 1057

Depth : : : : : 325 367 392 415 468

Length maxilla : : : 14! 152 172 182 199

TABLE 6. Morphometric data for Thunnus obesus caught off the Cape Peninsula.

cares. Gill-rakers 7-9 upper, 18-20 lower, 26-28 total. (See table 6 for morpho-

metric data.) Swim bladder with very deep paired pits anteriorly, and no

central thick connective tissue ridge pressing down into the swim bladder as in

T. albacares. In spite of the name the eye of the bigeye tuna is not much larger

than in the other tunas. The eye of 7. alalunga is actually greatest in relation to

fork-length (5:3-5:7). The eyes of both TJ. albacares (3:2-6:2%) and T.

maccoyi (3:1-41°%,) may occasionally reach the size range of T. obesus (3°5-

4°6%) at the same size. (See also Mather, 1963.)

Key to the genus Thunnus

The following key has been constructed to include all species which may be

found to comprise the genus Thunnus:

1 (a) Pectoral fin reaching beyond posterior margin of anal fin .. : 2

(6) Pectoral fin not reaching ie posterior margin of anal fin, sicigueh it eo

reach anal fin insertion

2 (a) Posterior edge of caudal fin ee a ese: “Sige secs: sspee! apace of i oe

and densely striated be T. alalunga

(6) No white edging on posterior edge of eel ne liver pale and. at most sien) Poa

striated (juvenile tuna) . é 3

3 (a) Liver without striations eres a Pm re e a ad sg Al

(6) Liver with distinct striations ventrally .. 6

4 (a) Total gill-raker count over 25 ae 27— 31) ; ‘all fas bright yellow “ag aaPae the

first dorsal T. albacares

(6) Total gill-raker cana les ite 2 = : Pee Ya we Re Me tee Mas

5 (a) Body deep (depth over 27% of fork- eae distance snout to anal fin more than

56% of fork-length. Small air bladder present . le T. atlanticus*

(6) Body slender (depth under 26% of fork- length): fesse ee to anal fin less than

60% of fork-length. No air bladder ae oe Ae T. tonggol*

6 (a) Total gill-rakers more than 38; often over 400 Tb weieee its ae T. thynnus

(6) Total gill-rakers 36 or less; usually less than 500 bere. he

7 (a) Ventral striations on liver in groups near margin. Peduncle keel peeks cage fin wih

red and yellow markings LAs ae i. be - ety, As abesus

(6) Ventral surface of liver fully seesi |. 8

8 (a) Peduncle keels bright yellow or hyaline; no “alse seeudires on ie an Deep

lateral pocket in pectoral region of roof of body cavity : T. maccoyit

(b) Peduncle keels opaque black, yellow markings on pelvic aoe Siallow lateral

depression in pectoral region of roof of body cavity .. a ay. T. orientalis*

* Not known from South African waters.

14 ANNALS OF THE SOUTH AFRICAN MUSEUM

DisTRIBUTION AND MIGRATIONS

Hydrography

Before the detailed distribution of tunas in the areas surveyed can be dis-

cussed, some knowledge of the physical environment is required. The coastline

of southern Africa is shown in figure 1, together with the positions of the stations

fished. The distribution of water masses in the region of the Cape of Good Hope

is complex and variable, although the broad pattern of the currents is known

(fig. 3).

Offshore the coasts are ringed by subtropical water ; inside this subtropical

water, the waters of the east and south coasts are warm and saline, whereas

the waters bounding the west coast are cold and of low salinity, although rich

in nutrients.

The major water masses in the South African region which were fished

during this survey are:

(a) South Atlantic Subtropical Surface Water

This water mass is part of the South Atlantic anti-cyclonic gyral. To the

south it is bordered by the West Wind Drift, from which it is separated by the

Subtropical Convergence at about 37°-40°S. Eastwards it is bordered by the

Agulhas Current south of the African continent and the surface waters of the

Indian ocean, but there is often no sharp boundary between these waters, and

the South Atlantic Subtropical Surface Water may move as far east as Cape

Agulhas or further when the force of the Agulhas Current is weak. The tempera-

ture and salinity decrease southwards; at 20°S they are of the order of 23°C;

36°/,,, while at 34°-35°S they have dropped to 15°-21°C; 35:°4-35°9°/,,

(Clowes, 1950).

(5) Benguela Current

The Benguela Current has recently been discussed in great detail by Hart

& Currie (1960), who have shown that the current is caused by local southerly

winds inducing an offshore movement of surface water in a north-westerly

direction, which in turn causes upwelling. This upwelled water is separated from

the north and west-flowing waters of the South Atlantic gyral by a convection

cell. It was formerly thought (Clowes, 1950) that the upwelled water had its

origin in the Antarctic Intermediate Water, but Hart & Currie have shown

that the upwelled water has its origin at 200-300 metres, far above the depth of

the Intermediate Current, and originates from Atlantic Central Water. The

water inshore of the convection cell is cold (9°-15°C), the temperature depending

largely on the degree of upwelling, and of relatively low salinity (34.°5-35-1°/,.);

although rich in phosphates, silicates, and nitrates. It is quite common during

calm weather in summer for the surface waters to warm up and a marked

thermocline to exist. At station 3425 in February 1958 Africana II, a research

vessel belonging to the Division of Sea Fisheries, found surface water of 17-55°C

and salinity 34°97°/,,, while at 30 metres, although the salinity remained con-

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS

‘suOT}IPUOD J9UTUUNS [evoIdA} JopuN volIpY YINOG puNoe sutd\sAs yUOIIND °*6 “BIZ

<____—_

DUCHY YINOS

<—____

16 ANNALS OF THE SOUTH AFRICAN MUSEUM

stant, the temperature had dropped to 12-14°C. Other stations in the area

showed the same effect (Div. Sea Fish. Ann. Rep. 1960).

The origin of this water is believed to be the sinking of surface water at the

Subtropical Convergence (Sverdrup et al. (1942) although Clowes (1950)

suggests that it is the direct result of mixing between surface subtropical water

and the underlying Intermediate Water. In the present survey only Atlantic

Central Water which contributes to the Benguela Current is involved. The

temperature range is 5°-12°C, and the salinity 34-3-35-1°/,, (Sverdrup e¢ al.,

1942). It lies below the surface waters at 150-500 metres (Hart & Currie, 1960).

(d) Agulhas Current

The Agulhas Current forms part of the South Indian ocean gyral, being

derived from the southern branch of the South Equatorial Current, flowing

down the west side of Madagascar, and down the southern African coast to the

tip of Africa. The currents in the Mozambique Channel are thought to be

variable, but may contribute at certain times of year to the Agulhas Current.

The speed of the current is considered to be 3-4 knots (African Pilot). When the

current reaches the Agulhas Bank, it swings away from the South African coast,

following the edge of the bank, moving in a south-westerly direction, and then

turning eastwards as the Agulhas Return Current.

There is indication that a varying amount of Agulhas Current water may

cross the Agulhas Bank, or round its tip, to mingle with the South Atlantic

Subtropical Water in the area of the survey, its temperatures generally being in

the region of 20°-24°C and the salinity 35-4-35:6°/,,, although both may

decrease near the coast, but are not known to fall below 17°Q; 35:3°/,,. The

inshore water between the Agulhas Current and the coast, where the current is

often northward flowing (Gilchrist, 1903), was not fished, as the water depth was

not sufficient for the type of gear used.

(e) South Indian Ocean Subtropical Water

This water mass is analogous with the South Atlantic Subtropical Water

and, although in general slightly less saline, is several degrees warmer. Discovery II

found a temperature of 26-58°C and salinity of 35-31°/,, at 21° 44’ E south of

the Cape in April (Clowes, 1950). In the south-west Indian ocean, tempera-

tures during this survey ranged from 18-7°-23:2°C in August 1962 and from

22-7°-26-7° in February 1963. This water is separated from the South African

coast by the Agulhas Current.

Clowes (1950) has discussed the hydrography of South African seas in

detail, and much of the above is based on his work, except that upwelled water

off the coast is considered to be Central Water (after Hart & Currie, 1960)

rather than Antarctic Intermediate Water. Clowes also gave T/S curves for

each 5° of longitude from 5°W to 45°E within latitudes 30°-35°S. The curves

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS 17

3 ee

20.5 201 13-5

Fig. 4. Temperatures of the surface water off the western Cape in March. Taken from data in the

Annual Report Division of Sea Fisheries, 1957-1958.

for the region 5°W to 15°E show clear differences from those for the region

25°—45°E, although crossing several times. Clowes considers the water west of

15°E to be purely Atlantic and that east of 25°E to be purely Indian. In the area

15°—25°E, however, there is a great deal of mixing, both between Atlantic and

Indian surface water, and between surface water and upwelled Central Water.

In winter the Agulhas Current is weakened and due to the prevailing winds

off the Cape at this time being westerly, the Atlantic water may penetrate

eastwards, often to or even beyond Cape Agulhas, and deflect the Agulhas

Current south and east (Buys, 1959; De Jager et al., 1963). In summer, however,

18 ANNALS OF THE SOUTH AFRICAN MUSEUM

the conditions are reversed; the Agulhas Current is flowing strongly, the pre-

vailing winds are southerly and easterly and tend to blow the surface water

away from the coastline along the west coast. This results in upwelling, and

inshore the South Atlantic Surface Water becomes mixed with Central Water.

A portion of the Agulhas water appears to continue into the Atlantic (Dietrich,

1935), and there is often in summer a broad band of very warm water (over

20°C) right round the Cape from the Atlantic to the Indian ocean outside the

cold upwelled water. Figure 4 shows surface temperatures and salinities pre-

vailing off the Cape in March 1958 (drawn from data in Div. Sea Fish. Ann.

Rep. 1960). It clearly shows in this case that, while there is cold upwelled

water close inshore from Cape Point westwards, there is very warm water out-

side, while salinities suggest that Agulhas water is only reaching Danger Point;

between it and Cape Columbine there is an area of lower salinities, suggestive

of mixed water between the Agulhas water and the Atlantic surface water.

Temperature Tolerance of Tuna

Much work in the field of temperature tolerance of tuna, especially

T. alalunga, has been done off the Californian coast. As long ago as 1917

Thompson (1917, 1919) suggested that small changes in water temperature

might cause marked changes in the distribution of oceanic fishes, as did Hubbs

& Schultz (1929) and Walford (1931). Statistics of the Californian tuna

industry show great fluctuations in catch; after several good years there are

one or two bad ones. In 1925 the catch of tuna was 25 million lb, and in 1926

it dropped to 3 million lb, for instance. Hubbs (1948) was the first to investigate

the problem, and concluded that T. alalunga had a temperature range of

15°5°-18:4°C, and that when this temperature range did not occur in the

fishing area during the tuna season, the fishery did not develop. Later workers

(Clemens, 1957, 1958, 1961; Craig, 1959, 1960; Powell & Hildebrand, 1950;

Powell et al., 1952; Neave & Hanovan, 1960; Radovich, 1960, 1961) found

slightly different limits, but all agreed that the absolute limits for a fishery to

develop were above 14-4°C and below 20°C. Occasional fish could still be

taken at temperatures above 20°C, but only one fish was recorded below the

minimum (Schaefers, 1953). Clemens (1958), Craig (1959, 1960), and Radovich

(1960, 1961) found that small fluctuations in surface temperature were of great

importance, since a slight rise above the normal, which would result in a north-

ward movement of the isotherms beyond their normal latitude during the tuna

season, would cause the fish to move north of their usual area. They found that

the best fishing area for the season could be accurately determined by means of

a pre-season temperature survey of the usual area, and so successful has this

forecasting proved in the last few years that Johnson (1960) could state: “The

value of the sea temperature as a forecasting tool of albacore (T. alalunga)

distribution has been amply demonstrated during the last four years.’ In the

Californian area similar temperature effects were found for all other tuna species.

Uda (19614, 6) has made similar observations for the western Pacific ocean.

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS 19

The temperatures quoted in literature are almost invariably surface, and

not ambient temperatures, because of the difficulty of obtaining the latter.

Even with fishes caught on long-lines whose depth is more or less known, catches

may be made while hooks are sinking or rising. Squire (1963) gives both surface

temperatures and those at the hook depth. Surface temperatures are used here,

T.obesus.

Lalbacares

I alalunga

14 15 eee i718 iS T30) eal 20°C Bs 24

Fig. 5. Percentage of successful stations for each species at various water temperatures (to the

nearest 1:0°C). Only temperatures at which more than five stations were occupied are given.

20 ANNALS OF THE SOUTH AFRICAN MUSEUM

but in appendix A subsurface temperatures are given for all stations where

known.

There seems to be a range of temperatures within which a species of tuna

is usually found in any one area, but occasionally it may also be found outside

this range. We are calling the normal range the ‘optimum range’, merely on

the basis of the fact that the fishes are usually found within it.

The correlation between catch and surface temperature is shown in the

histogram in figure 5.

T. alalunga

As shown above, TJ. alalunga off the Californian coast has a normal tempera-

ture range of 14-4°—20°C. In other areas, however, slightly different tempera-

tures may be preferred. Alverson (1961) found off British Columbia that the

population of north Pacific T. alalunga with which he was dealing had an opti-

mum temperature range of 13°3°-16-1°C (54°-61°F). In equatorial regions,

on the other hand, there are tuna fisheries taking 7. alalunga in water of up

to 21°C according to Rosa & Laevastu (1961) and Laevastu & Rosa (1963)

as shown by their histogram although their chart of fishing areas shows fishing

up to the Equator in surface temperatures over 25°C. According to Kishinouye

(1923) T. alalunga is found in the region of Japan in water of 13°—-24°C. Squire

finds a range of 11-5°-28-2°C.

Attempts have been made to find the optimum surface temperature (as

defined above) for 7. alalunga (as well as the other species present) during the

survey off the Cape. Unfortunately to do this requires a large number of

samplings at a wide range of temperatures, which the 111 stations occupied

during the present survey did not have. Observations in the centre of the range

were adequate, but this was not the case at the possible extremes. The correla-

tion between catch and temperature is shown in the histogram in figure 5.

T. alalunga was taken almost throughout the range of temperatures. Only

at temperatures above 22°C was this species never taken. The only stations

fished at temperatures higher than 22°C were off the east coast, however, where

T. alalunga was never found, perhaps for other reasons. On the basis of our

observations at the g1 stations at which this species was taken, T. alalunga has an

optimum range of 16°—22°C, and is occasionally taken in temperatures as low

as 14°C.

T. maccoyit

Relatively little work has been done concerning the temperature require-

ments of T. maccoyii1. None of the authors who have studied the populations off

the Australian coast have published temperature observations of this species,

except Robins (1963), who mentioned that adults are seldom found at tempera-

tures above 21°C, and ‘juveniles’ do not occur above 19°C.

By referring to figure 5 it may be seen that 7. maccoyii is only rarely taken

at temperatures above 19°C in the Cape area, the drop-off in catch rate above

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS 21

this temperature being very marked. The lower limit of 7. maccoyi is unfor-

tunately not clear; below 15-5°C it seems to become less common.

T. thynnus

According to Mather (1963) and Tiews (1963) YT. thynnus has a wide

temperature range of 12—25°C. In False Bay, Cape Town, this species is only

found in water above 20°C.

T. albacares

More observations on the temperature requirements of this species have

been made than for most other species, except perhaps T. alalunga. Kishinouye

(1923) gave a total range for this species of 14°-27°C, but gave as an optimum

temperature range 22°-24°C.

In the Pacific ocean, Kamimura & Honma (1963) have shown that the

best catches of 7. albacares are made in the equatorial area between 5°N and

10°S, with the centre of distribution just south of the Equator, where the

temperature range is 22°-26°C. Schaefer et al. (1963) agree that this species is

usually found in water above 20°C, but add that they may occasionally be

found in water below 20°C. T. albacares in the north-east Pacific is, like 7. ala-

lunga, affected by small changes in the surface temperature of its normal area,

moving polewards during periods of warm water both off California and South

America (Blackburn, 1960; Schaefer, 1961), and suggesting that 7. albacares

may avoid water which has a temperature of over 25°C.

Squire (1963) found that in the north-west Atlantic the yellowfin could be

found in water ranging from 18-4° to 28-8°C, with an average temperature at

which fish were caught of 25-2°C.

Tsuruta (1961) found T. albacares in the south-west Indian ocean in water

with a range of temperature of 19-0°-25-3°C.

Off the Cape (fig. 5), the histograms suggest that the sampling was con-

ducted at the lower end of the temperature range, since in all figures, both of

percentage catch rate and successful stations, the catch seems to be increasing

with increase in temperature. Little fishing was done in warmer water, however.

It is worth noting, too, that 7. albacares has a wide range of temperature toler-

ance, since off the Cape it was the only species to occur over the whole range of

temperatures fished (13-8°-24°C); there was, however, no sign of an optimum

temperature. There is some indication that the optimum temperature range of

T. albacares is higher in this area than that of the other species, an observation

that is in agreement with the findings of Squire (1963) for the north-west

Atlantic, and the tentative conclusions reached by De Jager (1963) and De

Jager et al. (1963) for the Cape. Figure 5 suggests that 7. albacares is normally

found only in water above 17-5°C.

T. obesus

While 7. obesus is in general a species of low latitudes, being found in its

greatest quantities just north of the Equator (Kamimura & Honma, 1963), it

22 ANNALS OF THE SOUTH AFRICAN MUSEUM

does have a very wide range of temperature tolerance. In the Japanese area

Kishinouye (1923) recorded that the temperature range was 13°—-25°C with an

optimum of 22°-24°C. Tsuruta (1961), in the south-west Indian ocean, found

this species in water ranging in temperature between 17°C and 25°C. Other

wide ranges of temperature recorded are 12°2°-28-8°C (54°-84°F) in the

eastern Pacific (Alverson & Peterson, 1963) and 13:5°-27-3°C in the north-

west Atlantic (Squire, 1963).

During the survey very few 7. obesus were taken, too few to give a true

picture of their temperature range in this area; also, like 7. albacares, there were

too few stations occupied in warm water, above 21-5°C.

In conclusion, it can be said that, although the tuna found off the Cape

have a wide range of temperature tolerance, they are usually found within a

restricted range of temperature, which in T. maccoyii is low (below 19°C) and

in 7. obesus and T. albacares higher (above 17:5°C). 7. alalunga, however, seems

to have a wide optimum temperature range.

Changes in the tuna population

Prior to 1960 the opinion was widely held that tuna were found off the

Cape during the summer months only (Horne, 1959), although whaling skippers

working off Cape Columbine in the winter of 1957 reported seeing tuna.

During the summer months (November to April) the sport fishing vessels took

large numbers of T. albacares and a small number of T. alalunga (Horne, 1959).

In some years small numbers of what were probably large T. thynnus were also

taken by beach seines at this time, as well as smaller specimens by anglers off

the rocks at Rooikrantz, near Cape Point (Cape Times, 31/12/1945, 5/1/1946,

and South African Museum records). In addition, two specimens of T. obesus

were caught in November 1959 by trolling off Slangkop, Cape Peninsula

(Talbot & Penrith, 19615).

It was therefore confidently expected that during the winter months long-

lining would produce few or no tuna, and doubts were expressed as to the

desirability of continuing the survey during the winter. It was thus extremely

surprising to find that during the winter period the catches, far from showing

any tendency to diminish, actually increased, though changing in species com-

position. The change is shown diagramatically in figure 6. Referring to figure 6

it may be seen that T. alalunga is the only species present off the south-western

Cape at all times of the year. During the months October to May, what may be

termed the summer population of tunas is present off the Cape. This is com-

posed of T. albacares, T. obesus, T. alalunga (in the main comprising small fish

of less than 950 mm fork-length of this species), and a small number of T. thynnus.

During May to November a different population of tunas is present, being

composed of 7. maccoyii and T. alalunga, but the latter species is more abundant

than during the summer due to the addition of large numbers of larger indivi-

duals of over 950 mm fork-length. During the winter occasional specimens of

T. albacares and T. obesus are found, but in such small numbers that they may be

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS 23

Cape West Coast

Spring Summer Autumn Winter

Cape East Coast

L=)

Fig. 6. Annual changes in tuna populations in the waters west of Cape Town and on the South

African south-east coast. Only Thunnus albacares and T. obesus were taken in the latter area. Height

of shaded area is proportional to the catch per hundred hooks of long-line gear, as shown by

our survey.

presumed to have been individuals which did not migrate from the Cape with

the main population.

Off the east coast, in the area surveyed, between Cape Agulhas and Cape

Recife, there was a marked difference both in the species composition and in the

numbers of tuna present in comparison with the western Cape. No T. alalunga,

24 ANNALS OF THE SOUTH AFRICAN MUSEUM

T. thynnus or T. maccoyit were found, nor have they ever been definitely recorded

from this area. Only two species, 7. obesus and T. albacares are definitely found

in this area. During summer T. obesus and presumably T. albacares were present,

but owing to bad weather only two stations could be fished in this area in

summer; at these stations four 7. obesus were taken. In late summer both species

were found in small quantities, while in winter, although stations were occupied

in warm water, no tuna of any sort were caught. At the end of April 1960 large

quantities of T. albacares and a smaller number of 7. obesus were caught between

Cape St. Francis and Cape Recife in water of 18°-20-3°C, although off Cape

Infanta, in water of similar temperature (18-3°C) no tuna were found.

In the south-west Indian Ocean so few fish were caught during the two

cruises on the 8.A.S. Vatal that no clear results could be obtained. Noteworthy

was the finding of 7. alalunga close inshore east of Durban, from where it had

been recorded only once previously (Molteno, 1948). It was also found that

there seemed to be a marked difference in the number of tuna present in

summer (February 1963) and in winter (August 1962). During winter tunas

were found in fairly small quantities on the stations east of Durban, T. alalunga

and 7. albacares being the species present. During the summer tunas were almost

completely absent in the area surveyed, only two T. albacares and one T. obesus

being found. As in the area further south, no 7. thynnus were obtained.

To our knowledge the first survey of the area south of Madagascar and east

of Durban was done in January 1961, by the R.V. Koyo-Maru belonging to the

Shimonoseki College of Fisheries, Japan. She obtained catch rates of 2-0% for

T. albacares, 1-1% for T. alalunga, and 0-2% for T. obesus (Tsuruta, 1961), a

total catch rate far in excess of the total catch rate for the February Watal cruise

of 0-3%. A possible reason for this seeming anomaly is patchy distribution.

Tsuruta’s data suggest small shoals very widely separated, and since the Koyo-

Maru used a far longer length of long-line than the Natal, thus sampling a far

greater area of the sea, this may partly account for the apparent difference in

results.

This area has, however, become a commercial fishing area since the survey

of the Koyo-Maru, and, judging by the aerial sightings of vessels off Durban

(Pinkerton, pers. comm.) and by landings of frozen tuna at Durban by Japanese

long-liners (Durban Customs, pers. comm.), most of the fishing in this area is

done in winter. It seems therefore that the indication given by the cruises on

S.A.S. Natal, that tuna are more plentiful in the south-west Indian ocean in

winter than in summer, is correct, although based on scanty evidence at

present.

The Distribution of Tuna in Relation to Hydrographic Conditions

Off the Cape there are both marked annual changes in tuna populations

and great annual changes in the hydrographic conditions of the surface water

masses. ‘The tuna populations change at the same time as the water masses,

suggesting a causal relationship. The main determinants of horizontal distribu-

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS 25

tion in a pelagic open-ocean fish such as tuna are probably the necessity to find

available food, the necessity to remain within suitable hydrographic conditions,

the necessity to satisfy the particular, usually more narrowly selected, hydro-

graphic conditions during the breeding season, and also behavioural factors

such as schooling behaviour and inherent migratory urges. It would be naive to

consider the spatial position of tuna to be anything but the result of a complex

inter-relationship of a number of these factors at any one time. Nevertheless,

given adequate feeding at a non-breeding season, temperature (which has been

considered by Yabe et al. (1963), to be the most important factor determining

the suitability of the hydrographic environment) may be closely linked to

distribution.

Food and feeding will be discussed in a later section, but it should be noted

that the Cape area, owing mainly to the upwelling of the water forming the

Benguela Current, is exceptionally rich in phosphates, nitrates and silicates,

and is biologically one of the richest areas of the oceans (Steemann-Nielsen,

1956), supporting vast quantities of potential tuna forage organisms such as

Sardinops ocellata, Maurolicus muelleri and Loligo reynaudt.

Tuna require a very uniform body of water within which to breed, with no

sudden changes or fluctuations in temperature. The temperature, too, must be

high (24°C or higher according to Yabe et al., 1963). These conditions are not

met with off the Cape.

If the catch rate of fish per hundred hooks is considered as a measure of

abundance, which over a series of stations fished at different times is a fair

assumption, the different water masses around the Cape have very different

populations of tuna. In figure 7 the populations have been expressed in number

caught per hundred hooks in the different water masses. The upwelled water

that forms the Benguela Current changes little in temperature from summer to

winter, although it is more clearly defined in summer because of the off-shore

wind systems at that time of year (fig. 3). The South Atlantic Subtropical

Surface Water changes considerably, however, from summer to winter in

temperature. Because of this, and the fact that species of tuna move from the

Cape at different times to breed the water masses have been considered sepa-

rately for summer and winter.

In summer there is much mixing of the Agulhas Current water (which

rounds the Cape under certain wind conditions) and the South Atlantic Sub-

tropical Surface Water in the area west of Cape Town where the main sampling

effort of the survey was concentrated (fig. 3). These surface waters have rather

similar temperature and salinity characteristics, and this, plus the problem of

mixed water, makes it difficult to decide in many cases in what water a station

was placed. They have therefore been considered together(fig. 7) in summer,

but in winter, where there is no rounding of the Cape by the Agulhas Current,

and the South Atlantic Subtropical Surface Water is much colder, they can be

clearly differentiated, and we have been able to consider them separately.

Summer has been considered to be from December to May inclusive, as these

26 ANNALS OF THE SOUTH AFRICAN MUSEUM

SUMMER

Agulhas Current (E Coast) |

Fig. 7. Catch rate of tuna in summer and winter in the various water masses sampled. Catch rate

is expressed as number of fishes taken per hundred hooks laid, and is stated above each histogram.

months showed warmer water (over 20°C) than the months June to November

(below 20°C) during the eighteen months of the survey.

The Benguela Current lies inshore of the warm Agulhas Current/South

Atlantic Subtropical Surface Water in summer, and is often so clearly demar-

cated from the latter that there is a change in surface temperature of as much as

5°C over a hundred yards, and also a change from greenish water with a high

plankton concentration to clear blue warm water. Only T. alalunga was found

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS ag |

in the Benguela Current in the few stations occupied in summer. In winter,

when T. maccoyii has reached the Cape, both this species and T. alalunga are

found in this water mass. A few T. obesus were also taken in it in winter, pre-

sumably being some of the very few fish of this species that remain through the

winter in the area. At 4 stations, however, all in summer, at the border of the

Benguela Current, and apparently just within it, both 7. albacares and T. obesus

were taken in quantity. Stations further into the Benguela water at the same

time gave none of these species. These stations were not included in figure 7,

and are mentioned here to indicate that on rare occasions the warmer water

species may move just into the borders of the Benguela, presumably to feed on

the rich stocks of forage species.

In the Agulhas Current/South Atlantic Subtropical Surface Water in sum-

mer the dominant species is 7. alalunga, and both T. albacares and T. obesus

are common (fig. 7). 7. maccoyit is absent from the Cape during summer; its

presence in this water mass (fig. 7) is due to a few specimens being taken on

7 and 8 December 1960, and on 21 and 22 June 10961, i.e. just outside the

limits of ‘summer’ as defined here. Although the two water masses are here

considered together, in comparing temperature/salinity curves at the stations

with those of Clowes (1950) there was an indication that Thunnus albacares and

T. obesus were mostly in water of Agulhas Current origin, while 7. alalunga

was found in both waters, but mostly in South Atlantic Subtropical Surface

Water. The only specimen of T. thynnus was taken in this water mass in Decem-

ber 1960. From the catches of commercial long-line boats which began operating

after this survey it has become clear that at least in some summers T. thynnus is

very much more common than this single record indicates. In the summer of

1963-4, as well as in subsequent years, a large number of this species was also

taken by sports fishermen in False Bay. As they were using a technique new in

the area (chumming and whole bait), it is possible that this species may have

been here in other years.

In the South Atlantic Subtropical Surface Water which lies offshore from

the western Cape in winter the dominant species are T. alalunga and T. maccoyit.

The occasional finding of T. albacares and T. obesus in this water mass in winter

suggests that a small proportion of these predominantly summer fish do not leave

but overwinter at the Cape.

Migrations of South African Tunas

Tunas are known to occur off Angola (Molteno, 1948; Schaefer & Walford,

1950; Vilela & Monteiro, 1959) and Zanzibar (Copley, 1947; Morrow, 1954;

Williams, 1956, as well as the south-west Indian ocean (Tsuruta, 1961). Off

Angola T. albacares, T. alalunga and T. obesus have been recorded, but no

T. thynnus, although this species is known from St. Helena Island (Molteno,

1948). J. albacares is the dominant species off Zanzibar, together with a few

T. alalunga, but Williams (pers. comm.) has not recorded T. obesus or T. thynnus.

In the south-west Indian ocean off Durban T. obesus, T. albacares and T. ala-

ANNALS OF THE SOUTH AFRICAN MUSEUM

‘sypjuaiso “[, puv ndossvum - 7 ‘snuudyz snuunyy, euny uyanyq jo sotoads atqeqosd 99TY} SY} JO UOTNGISIP P[IOM “g “SIT

Asenigay wuayzos| 9,¢Z -—-—-. =.

4{snbny wiuayjos; 9.¢¢ — —_—

“sieyueiso snuudyy 1 NN

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS 29

lunga are found (Molteno, 1948; Tsuruta, 1961; and the present survey).

The major annual changes in the tuna populations of the Cape have

already been discussed, but which ocean the South African populations are

derived from, and what their migrations are in these oceans has always been in

doubt. The work of Rivas (1955), Clemens (1961), Hamre (1962) and Mather

(1962) have shown that tuna are capable of long migrations at a relatively high

average speed (up to 3-5 knots). Mather has shown that T. thynnus, of both

large and small size, cross the Atlantic. Long distance migrations of albacore

across the Pacific have also been shown to occur by the U.S. Fish and Wildlife

Service (Otsu & Uchida, 1963).

(a) Thunnus thynnus

: Thunnus thynnus has never been recorded on the Indian ocean side of the

Cape, and it seems probable that the small number of this species which occurs

off Cape Town during summer have their origin in the Atlantic.

(b) Thunnus maccoyit

T. maccoyi had never been recorded in the Atlantic until this survey, and is

an Indo-Pacific species. The fishes of this species found off the Cape normally

have bright yellow peduncle keels, and the only other population of bluefin

tuna which is stated in the literature to have this feature occurs in the waters

off Australasia (Serventy, 1956a). Mimura (1961) described the main fishing

areas for this form, off the Australian west coast, mentioning two areas, the

‘Old ground’ 20°-30°S, 100°-110°E, and the ‘New ground’ 10°-18°S, 115°-

120°E. In addition to these, Laevastu & Rosa (1963) show a third area,

32°—-38°S, 80°-95°E, which they state is fished for ‘bluefin’. Mimura states that

T. maccoyit is found only from October to April in the ‘New ground’, and Sep-

tember to March in the ‘Old ground’. It is significant that this is the time

during which T. maccoyii is absent from the Cape. It is possible that the fish

occurring off the Cape may have their origin off the Australian coast. The

results of tagging of T. alalunga in the north Pacific (Otsu & Uchida, 1963) and

of J. thynnus in the Atlantic and Pacific (Hamre, 1963; Mather, 1963a) have

shown that an annual migration of this magnitude is not impossible.

Subsequent to this survey, however, T. maccoyii has been found to occur in

the south Atlantic at Vema seamount (31° 38'S 8° 20’E) (Penrith (1967) and

later personal catches from the R.V. Thomas B. Davie belonging to the University

of Cape Town), and Shingu (1967) has shown that T. maccoyii is found over a

wide band of the south temperate zone, but probably avoids the warm water of

the Agulhas Current.

The distribution of T. alalunga around the South African coast shows a

strange break in continuity. The species is common off Angola, St. Helena

Island, Walvis Bay, Stompneus Bay (outside the Benguela Current), and from

30 ANNALS OF THE SOUTH AFRICAN MUSEUM

Cape Columbine to Danger Point. The species is then, as far as is known, absent

from waters off the South African south-east coast, but reappears off Durban

and, although found by Tsurutu (1961) to be present well off the African coast

to 35°S, is not plentiful in this area (Kataoka, 1957; Laevastu & Rosa, 1963).

It seems therefore very reasonable to assume that the population of T. alalunga

found off the Cape is of Atlantic origin, and is the south-eastern limit of a large

population which is known in the South Atlantic from 10°-35°S, according to

Laevastu & Rosa (1963).

(d) Thunnus obesus and T. albacares

The origin of T. obesus and T. albacares is less clear. ‘The two species are

common off Angola and are also found in the south-west Indian ocean off

Durban. Mention was made earlier of the extremely large catch of T. albacares

made at three stations between Cape St. Francis and Cape Recife at the end of

August 1960, together with smaller numbers of 7. obesus. It is felt that these

catches were highly significant for two reasons. Firstly, August was

approximately one month before the two species began to appear off the Cape

Peninsula in the long-line industry catches (Bacon, pers. comm.) ; and secondly,

the large numbers of fish obtained (almost half the 7. albacares obtained during

the entire survey were taken during the three days) suggests a compact migra-

ting population rather than a widespread feeding population. It seems, there-

fore, in the absence of any valid evidence to suggest an Atlantic origin for

IT. albacares and with it T. obesus, that they are probably of Indian ocean

origin. The presence of fair numbers of T. albacares off Durban in winter and

their almost complete absence in summer cannot be explained. It might be

speculated that their absence is due to their having migrated southward owing to

high temperatures found in the region in summer (similar to the situation off

North America, as noted by Radovich, 1961), and that they are actually part

of the same population as occurs off the Cape. Alternatively, however, they

could have migrated northward to spawn during the summer.

One further point must be discussed in connection with the migration of

T. albacares. It is common knowledge among Cape anglers that the best catches

of 7. albacares off the Cape Peninsula are made at the beginning and end of the

season; during the middle of the season catches in this area tail off somewhat,

and at the same time the fish are sometimes found at Mossel Bay. This, in the

opinion of Horne (1962) suggests an Atlantic origin for the species. The species

is said to come from the Atlantic, reach the Peninsula, then migrate up the east

coast to Mossel Bay, and then turn and come back to the Peninsula. If this

were true one would expect large numbers of the fish to be caught off Mossel

Bay, not the small numbers that are actually found. It is far more likely that

large compact migratory schools (such as were found off the east coast during

this survey) should reach the Cape (and since they are large and near the surface

are easily spotted during trolling) and then break up, so that anglers’ catches

decline. Some of these fish will feed on the Agulhas Bank, where they often are

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS 31

seen during summer by trawler crews (Barker, pers. comm.). It is probable that

some of these fish enter Mossel Bay. At the end of summer the migratory schools

will re-form, and again anglers’ catches off the Cape will show a marked

increase.

Genetic exchange between Atlantic and Indo-Pacific Tunas

The waters to the south of the Cape are probably the only place where

genetic exchange between Atlantic and Indo-Pacific populations of T. alalunga,

T. albacares, and T. obesus could take place.

As has been shown, the closely allied T. thynnus and T. maccoyii occur at the

Cape, but these are found at different seasons, and neither breeds in the area.

The other three species, T. alalunga, T. albacares, and T. obesus do not show,

in the adult form at least, any great differences between Atlantic and Indo-

Pacific populations. This suggests (although does not prove) that there is some

genetic interchange between the Atlantic and Indo-Pacific populations. While

the migratory patterns discussed above are thought to be the major ones, it is

probable that there are smaller migrations of T. obesus and T. albacares from the

Atlantic and T. alalunga from the Indian ocean to the Cape, and contact

between Atlantic and Indo-Pacific ocean populations takes place.

The possibility of both Atlantic and Indian ocean tunas being present

off the Cape in summer, at least, is strengthened by the finding, in the water off

the Cape in summer, of marlins which are indigenous to the Indo-Pacific

(Makaira indica and Makaira audax) as well as a marlin known only from the

Atlantic (Tetrapturus albidus) (Talbot & Penrith, 1962, 19635). The problems

of the origin and migration of the Cape tuna population can be elucidated only

by the inauguration of a large scale tagging programme on the tunas off the

Cape.

ACKNOWLEDGEMENTS

This survey would not have been possible without the generous help and

co-operation received from Irvin & Johnson (Pty.) Ltd. who supplied all

equipment, boat and crew for the eighteen-month Cape survey. Many persons

associated with Irvin & Johnson assisted the museum in various ways, but Mr.

C. 8S. Milford, Mr. K. McKay and Mr. C. Linnigar deserve special thanks. The

high degree of co-operation received at sea from Skipper G. Barker and the crew

of the M.T. Cape Point contributed much towards the success of the survey.

That portion of the survey conducted in the south-west Indian ocean was

done from S.A.S. Natal by courtesy of the South African Navy. The co-operation

of the hydrographer, Capt. J. K. Mallory, and the captain and crew of S.A.S.

Natal is appreciated. The cost of the long-lining equipment used on S.A.S.

Natal was covered by a grant from the Council for Scientific and Industrial

Research as part of the South African participation in the International Indian

Ocean Expedition.

32 ANNALS OF THE SOUTH AFRICAN MUSEUM

Mr. N. Bacon, manager of the Atlantic Tuna Corporation (Pty.) Ltd.

and Mr. C. Horne, a keen angler and author of several books on angling in

South Africa, gave much help and information.

We are grateful to Dr. J. R. Grindley and Mr. 8. X. Kannemeyer of the

South African Museum, Mr. A. J. de Freitas of the Instituto de Investigagao

Cientifica de Mocambique and Mr. W. Nussey of Cape Town for assistance at

sea.

Mr. F. Williams, formerly of the East African Marine Fisheries Research

Corporation in Zanzibar, kindly made unpublished data on TY. albacares

available.

The Division of Sea Fisheries analysed all water samples and loaned many

journals not available in the museum library.

We are grateful to the South African Marlin and Tuna Club for making

arrangements for us to measure large numbers of yellowfins caught during their

championships.

We are grateful to Professor J. H. Day, Dr. B. B. Collette and Dr. R. H.

Gibbs for reading and criticizing 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

ALVERSEN, D. L. 1961. Ocean temperatures and their relationship to albacore tuna (7. germo)

distribution in waters off the coast of Oregon, Washington, and the province of British

Columbia. 7. Fish. Res. Bd Can. 18: 1145-1152.

ALVERSON, F. G. & PETERSON, C. L. 1963. Synopsis of biological data on bigeye tuna Parathunnus

stht (Temminck & Schlegel, 1844). F.A.O. Fish. Rep. 6: 482-514.

BARNARD, K. H. 1925. A monograph of the marine fishes of South Africa. I. Ann. S. Afr. Mus.

21: 1-418.

BARNARD, K. H. 1927. A monograph of the marine fishes of South Africa. II. Ann. S. Afr. Mus.

21: 419-1065.

BARNARD, K. H. 1939. A note about yellowfin tunny. Rep. S. Afr. Mus. 1938: 12.

Breese, W. & TEE-VaN, J. 1936. Systematic notes on Bermudan and West Indian tunas of the

genera Parathunnus and Neothunnus. Zoologica, N.Y. 21: 177-194.

Berc, L. 8. 1940. Classification of fishes, both recent and fossil. Trudy zool. Inst., Leningr. 5:

87-517. (In Russian, with complete English translation.)

Biven, C. L. 1930. Sea angling fishes of the Cape. London: Oxford University Press.

BLACKBURN, M. 1960. Scripps tuna oceanography research (STOR) program: final report,

June 21, 1957—June 30, 1960. Scripps Instn Oceanogr. Ref. 60-50: 1-80.

Biunt, C. E. 1954. Two mid-Pacific recordings of recoveries of Californian tagged albacore.

Calif. Fish Game 40: 339.

BONNATERRE, J. P. 1788. Tableau encyclopédique et méthodique des trois régnes de la nature . . .6. Ichthy-

ologie. In Encyclopédie méthodique. Paris.

Buys, M. E. L. 1959. The South African pilchard (Sardinops ocellata) and maasbanker (Trachurus

trachurus). Hydrographical environment and the commercial catches, 1950-1957. Investl

Rep. Div. Fish. Un. S. Afr. 37: 1-176.

CasTELNAU, F’. Dz. 1872. Contribution to the ichthyology of Australia. Proc. zool. Acclim. Soc.

Vict. 1: 29-247.

Certi, F. 1777. Anfibi e pesci di Sardegna. Sassari: Piattoli.

CHABANAUD, P. 1930. Sur I’aisselle de Ja pectorale de divers poissons scombroides. Bull. Soc. zool.

Fr. 55% 142-150.

CiEmens, H. B. 1957. Logging the longfins. Outdoor Calif. 18(11): 6-7.

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS 33

Ciemens, H. B. 1958. Albacore newsletter. Calif. Dept. Fish Game, Marine Research Operations.

4 Pp.

Ciemens, H. B. 1961. The migration, age and growth of Pacific albacore (Thunnus germo)

1951-1958. Fish Bull. Calif. 115: 1-128.

Ciemens, H. B. 1963. A model of albacore migration in the North Pacific Ocean. F.A.O. Fish.

Rep. 6: 1537-1548.

Crowes, A. J. 1950. An introduction to the hydrology of South African waters. Invest] Rep. Div.

Fish. Un. S. Afr. 172 1-26.

Co.teTTE, B. B. 1961. A preliminary review of the tunas of the genus Thunnus. Paper presented

to the Pacific Tuna Biology Conference, Honolulu, Hawaii, 14-19 August 1961, (VI-I):

30 p. (Mimeo.)

Co..eTTE, B. B. & Grsss, R. H. 1963a. A preliminary review of the fishes of the family Scom-

bridae. F.A.O. Fish. Rep. 6: 21-32.

Co.tetTe, B. B. & Gress, R. H. 19630. Preliminary field guide to the mackerel- and tuna-like fishes

of the Indian Ocean (Scombridae): 1-48.

Copsey, H. 1947. Fish records and observations. Nature E. Afr. 2: 9-10.

Craic, W. L. 1959. Albacore newsletter. Calif. Dept. Fish Game, Marine Research Operations. 3 p.

Craic, W. L. 1960. Albacore newsletter. Calif. Dept. Fish Game, Marine Research Operations. 3 p.

CRANE, J. 1936. Notes on the biology and ecology of giant tuna, Thunnus thynnus (L.), observed

at Portland, Maine. Zoologica, N.Y. 21: 207-212.

CUNNINGHAM, J. T. 1910. On the marine fishes and invertebrates of St. Helena. Proc. zool. Soc.

Lond. 1910: 86-127.

De JAGER, B. 1963. Synopsis of biological data on bluefin tuna Thunnus thynnus orientalis (T. & S.)

1942, longfin tuna Thunnus alalunga (Bonnaterre) 1788, yellowfin tuna Thunnus albacares

(Bonnaterre) 1788, and bigeye tuna Thunnus obesus (Lowe) 1839 (South Africa). F.A.O.

Fish. Rep. 6: 588-607.

DE JAGER, B. Nepcen, C. S. & Van Wyk, R. J. 1963. A preliminary report on tuna in west

coast waters of South Africa. S. Afr. Shipp. News Fish. Ind. Rev. 18 (4, 5 & 6).

De Syitva, D. P. 1955. The osteology and phylogenetic relationships of the blackfin tuna

Thunnus atlanticus (Lesson). Bull. mar. Sci. Gulf Caribb. 5: 1-51.

Dretricu, G. 1935. Aufbau und Dynamik des siidlichen Agulhas-Stromgebietes. Veréff. Inst.

Meersk. Univ. Berl. (A) 272 1-79.

FRADE, F. 19314. Données biométriques sur trois espéces de thon de |’Atlantique orientale.

Rapp. P.-v. Réun. Cons. perm. int. Explor. Mer. 70: 117-126.

FRADE, F. 19315. Neothunnus albacora (Lowe, 1839). Faune ichthyol. Atlant. N. 8.

FRASER-BRUNNER, A. 1950. The fishes of the family Scombridae. Ann. Mag. nat. Hist. (12) 3:

131-163.

Gisss, R. H. & Co..ette, B. B. 1967. Comparative anatomy and systematics of the tunas,

genus Thunnus. Fishery Bull. Fish Wildl. Serv. U.S. 66: 65-130.

GitcurisT, J. D. F. 1902. Catalogue of fishes recorded from South Africa. Mar. Invest. S. Afr.

1: 97-163.

Giucurist, J. D. F. 1903. Currents on the South African coast as indicated by the course of drift

bottles. Mar. Invest. S. Afr. 22 155-165.

Ginspurc, I. 1953. The taxonomic status and nomenclature of some Atlantic and Pacific popula-

tions of yellowfin and bluefin tunas. Copeia 1953: 1-10.

GopsiL, H. C. 1954. A descriptive study of certain tuna-like fishes. Fish Bull. Calif. 97: 1-185.

Gopsi_, H. C. & Hotmperc, E. K. 1950. A comparison of the bluefin tunas, genus Thunnus,

from New England, Australia and California. Fish Bull. Calif. 77: 1-55.

GrauaM, J. J. & STEwart, D. G. 1958. Estimating maximum fishing depth of longline gear with

chemical sounding tubes. Spec. scient. Rep. U.S. Fish Wildl. Serv. Fish. 285: 1-16.

GREAT BRITAIN. HyDROGRAPHIC OFFice. 1954. Africa pilot. Part III. 11th ed. London: Admiralty.

GUNTHER, A. 1860. Catalogue of the acanthopterygian fishes in the collection of the British Museum 2.

London: British Museum.

Hamre, J. 1963. Tuna tagging experiments in Norwegian waters. F.A.O. Fish. Rep. 6: 1125-1132.

Hart, T. J. & Curr, R. I. 1960. The Benguela Current. ‘Discovery’ Rep. 31: 123-298.

Hetpt, H. 1927. Contribution 4 l’étude des races de thons (Thunnus thynnus (L)). Annls Sin

océanogr. Salammbé 4: 1-54.

Horne, C. 1954. Fisherman’s Eldorado. Cape Town: Timmins.

34. ANNALS OF THE SOUTH AFRICAN MUSEUM

Horne, C. 1959. Big game fishing in South Africa. Cape Town: Timmins.

Horne, C. 1962. Is there a tunny route to the Cape? Tight Lines 3(3) : 3-4.

Husss, C. L. 1948. Changes in the fish fauna of western North America correlated with changes

in ocean temperature. 7. mar. Res. 7: 459-482.

Husss, C. L. & Scuuttz, L. P. 1929. The northward occurrence of southern forms of marine life

along the Pacific coast in 1926. Calif. Fish Game 15: 234-240.

Jounson, J. H. 1961. Sea temperatures and the availability of albacore (Thunnus germo) off the

coasts of Oregon and Washington. Paper presented to the Pacific Tuna Biology Conference,

Honolulu, Hawaii, 14-19 August 1961, (V-6): 14 p. (Mimeo.).

Jorpan, D. S. & Evermann, B. W. 1926. A review of the giant mackerel-like fishes, tunnies,

spearfishes, and swordfishes. Occ. Pap. Calif. Acad. Sci. (12): 8-25.

Kamimura, T. & Honma, M. 1963. Distribution of the yellowfin tuna Neothunnus macropterus

(T. & S.) in the tuna longline fishing grounds of the Pacific Ocean. F.A.O. Fish. Rep. 6:

1299-1328.

Kataoka, A. 1957. At the middle Indian Ocean about result of tuna longline fishing examina-

tion. 7. Shimonoseki Coll. Fish. 6: 93-100.

KisH1InouyE, K. 1923. Contributions to the comparative study of the so-called scombroid fishes.

J. Coll. Agric. imp. Univ. Tokyo 8: 293-475.

Laevastu, T. & Rosa, H. 1963. Distribution and relative abundance in tunas in relation to their

environment. F.A.O. Fish. Rep. 6: 1835-1851.

Linnaeus, C. 1758. Systema naturae . . . Editio decima. 1. Holmiae: Salvii.

LowE, R. T. 1839. A supplement to a synopsis of the fishes of Madeira. Proc. zool. Soc. Lond. 7:

76-92.

Marr, J. C. & ScoaerFer, M. B. 1949. Definitions of body dimensions used in describing tunas.

Fishery Bull. Fish Wildl. Serv. U.S. 543 241-244.

Martner, F. J. 1960. Recaptures of tuna, marlin and sailfish tagged in the western North Atlantic.

Copera 1960: 149-151.

MarTuHER, F. J. 1962. Newsletter, Co-operative game-fish tagging program. Woods Hole. 1962, Nov.

MatTuer, F. J. 1963. Tunas (genus Thunnus) of the western North Atlantic. III. Distribution and

behaviour of Thunnus species. F.A.O. Fish. Rep. 6: 1159-1161.

MicpaAtskI, E. C. 1958. Angler’s guide to the salt water game fishes, Atlantic and Pacific. New Jersey.

Mimura, K. 1961. Studies on the Indomagura. Paper presented to the Pacific Tuna Biology

Conference, Honolulu, Hawaii, 14-19 August 1961, (1-3): 7 p. (Mimeo.)

Mo ttTeEno, C. J. 1948. The South African tunas. Cape Town: South African Fishing Industry

Research Institute.

Moorg, H. L. 1951. Estimation of age and growth of yellowfin tuna (NV. macropierus) in Hawaiian

waters by size frequencies. Fishery Bull. Fish Wildl. Serv. U.S. 52: 133-149.

Morrow, J. E. 1954. Data on dolphin, yellowfin tuna, and little tuna from East Africa. Copeza

1954: 14-16.

Murpnuy, G. I. & SHomura, R. S. 1953. Longline fishing for deep swimming tunas in the central

Pacific Jan.—_June 1952. Spec. scient. Rep. U.S. Fish Wildl. Serv. Fish. 108: 1-32.

Mourpny, G.I. & SHomurA, R. S. 1955. Longline fishing for deep-swimming tunas in the central

Pacific, August-November 1952. Spec. scient. Rep. U.S. Fish Wildl. Serv. Fish. 137: 1-42.

NAKAmovRA, H. 1949. The tunas and their fisheries. Spec. scient. Rep. U.S. Fish Wildl. Serv. Fish.

82: 1-118.

Nakamura, I. 1965. Relationships of fishes referable to the subfamily Thunninae on the basis

of their axial skeleton. Bull. Misaki mar. biol. Inst. 8: 7-38.

NEAVE, F. & Hanovan, M. G. 1960. Seasonal distribution of some epipelagic fishes in the Gulf

of Alaska region. 7. Fish. Res. Bd Can. 17: 221-233.

NiEsEN, E. S. 1956. Measuring the productivity of the sea. In Bruun, A. F. et al., eds. The

Galathea deep sea expedition 1950-1952: 53-64. London: Allen & Unwin.

Otsu, T. 1960. Albacore migration and growth in the North Pacific Ocean estimated from tag

recoveries. Pacif. Sci. 14: 257-266.

Otsu, T. & Ucuipa, R. N. 1963. Model of the migration of albacore in the North Pacific Ocean.

Fishery Bull. Fish. Wild. Serv. U.S. 63: 33-44.

PenritH, M. J. 1967. The fishes of Tristan da Cunha, Gough Island and the Vema Seamount.

Ann. S. Afr. Mus. 48: 523-548.

PowELL, D. E. 1950. Preliminary report on 1950 North Pacific albacore tuna exploration of the

John N. Cobb. Comml Fish. Rev. 12: 1-7.

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS 35

PowELL, D. E., ALVERSON, D. L. & Lrvincstong, R. 1952. North Pacific albacore tuna explora-

tions, 1950. Fishery Leafl. Fish Wildl. Serv. U.S. 4022 1-56.

PowE Lt, D. E. & HILDEBRAND, H. A. 1950. Albacore tuna explorations in Alaskan and adjacent

waters, 1949. Fishery Leafl. Fish. Wildl. Serv. U.S. 376: 1-33.

RADOVICH, J. 1960. Redistribution of fishes in the eastern North Pacific Ocean in 1957 and 1958.

Rep. Calif. coop. oceanic. Fish. Invest. 7: 168-171.

RapovicH, J. 1961. Relationships of some marine organisms of the northwest Pacific to water

temperatures particularly during 1957 through 1959. Fish Bull. Calif. 112: 1-62.

Rivas, L. R. 1954. A preliminary report on the spawning of the western North Atlantic bluefin

tuna (Thunnus thynnus) in the Straits of Florida. Bull. mar. Sci. Gulf Caribb. 4: 302-322.

Rrvas, L. R. 1955. A comparison between giant bluefin tuna (Thunnus thynnus) from the Straits of

Florida and the Gulf of Maine with reference to migration and population identity. Proc.

Gulf Caribb. Fish. Inst. Sess. 72 1-17.

Rivas, L. R. 1961. A review of the tuna fishes of the subgenera Parathunnus and Neothunnus (genus

Thunnus). Annali Mus. civ. Stor. nat. Giacomo Doria 62: 126-148.

Rosins, J. P. 1963. Synopsis of biological data on bluefin tuna Thunnus thynnus maccoyit (Castelnau)

1872. F.A.O. Fish. Rep. 6: 562-587.

ROEDEL, P. M. & Fitcu, J. E. 1961. Taxonomy and nomenclature of the Pacific tunas. Paper

presented to the Pacific Tuna Biology Conference, Honolulu, Hawaii, 14-19 August

1961, (VI-2): 19 p. (Mimeo.)

Rosa, H. & Latvastu, T. 1961. World distribution of tunas and tuna fisheries in relation to

environment. Paper presented to the Pacific Tuna Biology Conference, Honolulu, Hawaii,

14-19 August 1961, (V-1): 7 p. (Mimeo.)

Royce, W. F. 1961. A morphometric study of yellowfin tuna Thunnus albacares (Bonnaterre).

Paper presented to the Pacific Tuna Biology Conference, Honolulu, Hawaii, 14-19 August

1961, (III-6): 79 p. (Mimeo.)

RussELL, F. 8S. 1934. Tunny investigations made in the North Sea on Col. E. T. Peel’s yacht “‘St.

George’”’, summer 1933. Pt. I. Biometric data. 7. mar. biol. Ass. U.K. 192 503-522.

ScHAEFER, M. B. 1952. Comparison of yellowfin tuna of Hawaiian waters and of the American

west coast. Fishery Bull. Fish Wildl. Serv. U.S. 52% 353-373-

ScHAEFER, M. B. 1961. Report on the investigations of the Inter-American Tropical Tuna

Commission for the year 1960. Rep. inter-Am. trop. Tuna Commn 1960: 40-107.

ScHAEFER, M. B., BROADHEAD, C. C. & ORANGE, C. J. 1963. Synopsis on the biology of yellowfin

tuna Thunnus (Neothunnus) albacares (Bonnaterre) 1788 (Pacific Ocean). F.A.O. Fish. Rep.

6: 538-561.

ScHAEFER, M. B. & WaAtForpb, L. A. 1950. Biometric comparison between yellowfin tunas

(Neothunnus) of Angola and of the Pacific coast of Central America. Fishery Bull. Fish Wildl.

Serv. U.S. 512 425-443.

ScHAEFERS, E. A. 1953. North Pacific albacore tuna exploration, 1951. Comml Fish Rev. 14(5):

I-12.

SELLA, M. 1927. Is the tuna an international wanderer ? Can. Fisherm. 14: 81-82.

SELLA, M. 1929 (Publ. 1952). Migrations and habitat of the tuna (TJ. thynnus) studied by the

method of the hooks, with observations on growth, on the operation of fisheries, etc. Spec.

scient. Rep. U.S. Fish Wildl. Serv. Fish. 76: 1-20 (Translation).

SERVENTY, D. L. 1956. The southern bluefin tuna Th. th. maccoyii (Gastelnau) in Australian waters.

Aust. 7. mar. Freshwat. Res. 72 1-43.

SHINGU, C. 1967. Distribution and migration of the southern bluefin tuna. Rep. Nankai Reg.

Fish. Res. Lab. 25: 19-36.

SMITH, J. L. B. 1935. New fishes from South Africa. Rec. Albany Mus. 4: 358-364.

SmiTH, J. L. B. 1949. The sea fishes of southern Africa. Cape Town: Central News Agency.

SouTH AFrRicA. DIvIsioN OF FISHERIES. 1960. Twenty-ninth annual report for the period 1st April, 1957

to 31st March, 1958. Pretoria: Government Printer. (Reprinted, in part, from Commerce ©

Industry, Nov. 1959.)

SouTH, J. F. 1845. Zoology. Thunnus. In Encyclopaedia metropolitana. 25: 620-622. London: Rose &

Rose.

Sguire, J. L. 1963. Thermal relationships of tuna in the oceanic north-west Atlantic. F.A.O.

Fish. Rep. 6: 1639-1657

STARKS, E. C. 1910. The osteology and mutual relationships of the fishes belonging to the family

Scombridae. 7. Morph. 21: 77-99.

36 ANNALS OF THE SOUTH AFRICAN MUSEUM

StTorER, D. H. 1867. A history of the fishes of Massachusetts. Cambridge & Boston.

SvERDRUP, H. U., Jounson, M. W. & FLemine, R. H. 1942. The oceans: their physics, chemistry and

general biology. Engelwood Cliffs, N.J.: Prentice-Hall.

Tasor, F. H. 1964. The South African tunas. In Proceedings of the symposium on scombroid fishes.

1: 187-209. Mandapam Camp: Marine Biological Association of India. (Symposium series

I).

Ta.sot, F. H. & Penritu, M. J. 1961a. East coast Thunnidae. CCTA/CSA Symposium on

Thunnidae, Tunny (60) ro: 1-11.

Tasot, F. H. & Penriru, M. J. 19615. Thunnus (Neothunnus) albacares and Thunnus (Parathunnus)

obesus on the south and east coasts of Africa. CCTA/CSA Symposium on Thunnidae, Tunny

(60) 4x2 1-16.

TatsoTt, F. H. & PEenritH, M. J. 1961c. Thunnus obesus Lowe in South Africa. S. Afr. JF. Sct.

57: 240.

TasotT, F. H. & PenrirH, M. J. 1962. Tunnies and marlins of South Africa. Nature, Lond. 193:

558-559-

Tasort, F. H. & Penrith, M. J. 1963a. Synopsis of biological data on species of the genus Thunnus

(sensu lato) (South Africa). F.A.O. Fish. Rep. 6: 608-646.

Tasot, F. H. & Penrith, M. J. 1963+. The white marlin Makaira albida (Poey) from the seas

around South Africa. S. Afr. 7. Sct. 59: 330-331.

TEMMINCK, C. J. & SCHLEGEL, H. 1850. Pisces. Pt. 5-6: 73-112. In StEBOLD, P. F. von. Fauna

Japonica .. . Lugdoni Batavorum.

Tuompson, W. F. 1917. Temperatures and the albacore. Calif. Fish Game 3: 153-159.

Tuompson, W. F. 1919. The occurrence of the albacore north of San Francisco. Calif. Fish Game

53 203-204.

Tuomeson, W. W. 1918. Catalogue of fishes of the Cape Province. Rep. Fish. mar. biol. Surv. Un.

S. Afr. 42 74-177.

Trews, K. 1963. Synopsis of biological data on bluefin tuna Thunnus thynnus (L.) 1758 (Atlantic

and Mediterranean). F.A.O. Fish. Rep. 6: 422-481.

TsuRUTA, S. 1961. Morphometric characters of yellowfin tuna Neothunnus macropterus (Lowe) in

the south-western waters of the Indian Ocean. 7. Shimonoseki Coll. Fish. 11 3'71-390.

Ucuiwa, R. N. & Orsu, T. 1961. Analysis of sizes of albacore occurring in various Pacific fisheries

—a preliminary report. Paper presented to the Pacific Tuna Biology Conference, Honolulu,

Hawaii, 14-19 August 1961, (VII-11): 10 p. (Mimeo.)

Upa, M. 19614. Cyclical fluctuation of the Pacific tuna fisheries in response to cold and warm

water intrusions. Paper presented to the Pacific Tuna Biology Conference, Honolulu,

Hawaii, 14-19 August 1961, (V—7): 5 p. (Mimeo.)

Upa, M. 19615. Localized concentrations of tunas in the eddies of oceanic fronts. Paper pre-

sented to the Pacific Tuna Biology Conference, Honolulu, Hawaii, 14-19 August 1961,

(V-8): 3 p. (Mimeo.)

VILELA, H. & Monteiro, R. 1959. Sobre atunes de Angola. Bolm Pesca 12(64): 11-54.

WALFoRD, L. A. 1931. Northward occurrence of southern fish off San Pedro in 1931. Calif. Fish.

Game 17: 401-405.

Watson, M. E. 1963. Tunas (genus Thunnus) of the western North Atlantic. I. Key to the species

of Thunnus based on skeletal and visceral anatomy. F.A.O. Fish. Rep. 6: 1153-1154.

WitiiaMs, F. W. 1956. Preliminary survey of the pelagic fishes of East Africa. Fishery Publs colon.

Off. 8: 1-68.

YABE, H., Yasuta, Y. & UEYANAGI, S. 1963. Comparative distribution of eggs, larvae and adults

in relation to biotic and abiotic environmental factors. F.A.O. Fish. Rep. 6: 979-1009.

‘TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS

aor = ae) —_ es a= OZI = = — o&G.91 OOI eyuejuy odey "aS ae 9/gz

o-GSb SG — «s.% gQ. 0.06 OZI oG.G& s.G1 Gh.c1 19-41 cP urewUNOWY FIGVT, “M’N'M 1S L/a%

g-o1 &1 — 86 @g. 8 O31 gz-S& 18.6 LVv.v1 gQb-F1 cE purjsy] uosseq ‘"M'N'M 0& L/1%

— _— — — — — OZI 6V.G6 1-1 oG.G1 gS.G1 L& ouIquinjon ode = "MN 6z L/oz

L.ov 2% — — — 4.ob FS Lv.GG 6.11 10-91 gP-gI o& doysury]s M Qs L/61

0-0§ Lz — G.br — G.G1 06 6V.G6 o.c1 16.G1 9G.1 cP doysuels M Le L/g1

6.6 Ol — Q-G — G.9g g01 €v.G6E Go.€1 o&.91 o4.91 GL doysurys M gz L[L1

6-9 € — 9% — <G.G 9& — — — 0g-9I oS doysur|g M Go L/S1

g-cb 9 — &16 — 6.11 oGt 9v.G& o&.G1 1%-L£1 o€-41 G9 doysur|s M Va 9/E1

6.61 G — 661 — — 9& 67.GE6 q.11 Ga-L1 O€-L1 GV doysurys M £% 9/1

9-6 ie = a(0) Q:% — 1-9 Prt LV.G& Gg.11 16-41 Goe.L1 o& yuiog oder M hd 9/11

— —- — — a — ov LV.GG 11-11 €&-61 O1-1% Gz stouesy “1G oder S 1% 9/S

— — — — — — Vil gv.S& 66.11 30-16 VO-.1% cS euskuyy S 0% 9/V

— — — — — — Vil LvV.G& LG.%1 16.06 g6.12 G9 JUIO YIVAIOZ "M’S'S 61 g/€

0-8% a& — — — 0-82 VII £1.G66 99-11 06.61 Sg.F1 ov puvjsy uosseq ‘M'N’'M _ QI C/E

— — —_— — — — of oo-G& $g.s1 &%.G1 gg-GI 09 suiquinjon sdey “M'N'M AI G/z

— — — — — — 9& g&-GSE 9g-o1 2&.VI 99-91 o& ouriquinjory oden = “M’N'M QI G/1

Vago. ly 6.6 6.6 9-9 0-01 of go-S& 19-11 9-61 &g.G1 GE pues] Uosseq "M'N'M SI blot

6.6 2 — — 86 I.t 06 69.96 fo.€1 60-g1 I1-gI oS doysur[s M V1 b/gz

Gaga ~£ Gal Sed.) Qed ed VG 6S.S66 16.91 61.91 V¥z-gI of doysur|s M G1 b/Le

L.9 v Let — — og 09 GG.G& o9G.z%1 62.61 o&-61 GG jUIOg 1osueqd M oI b/z

6.§ 9 — — &.& — ogi — — — of.0z zQ oplooy odeyn S II £/o0&

6.¢ 9 ¢. — -% — ogi G€.GE ov.O1 Z%9-gI 0g-gI 09 oplrooy odey S OI &/g

Oe % — f= OS Org 99 £L&.GE Gi.11 €g.00 09.8% cP gyrooy ode S 6 €/9z%

g-4 G — — G1 0.9 99 GG.G& 96.61 1.06 EV.0% GL doysurys ‘MSM 8 S/o1

0.€ z — — Gr G1 99 LV.G& GE.cI 61-06 06-06 cP doysurys M L 6/6

g-4 G — — — gL 99 LvU.G& €&.c1 61.06 61-06 cv doysur|g M 9 £/6

v.6 L — — — 4.6 99 gb.GE 99-6 08.0% g&.0% ov yuIOg 1osueq M G &/L

G.L G — — —/ gf 99 Lb.GG§ €o.11 — _ 0.0% GE doysurys M 14 &/9

1.6 9 G.1 — — gf 99 bG.GE og-E1 sag-61 VO-.0% 6L doysurys MSM & €/S

6.8 G 6.8 — — — 09 oF.G& QI-dI 09-91 3g-QI &v doysurys M j &/P

— — — —- — — gv —— —= — ob.0z ov doysue[s M I z/91

0961

(ou DONS WG a A aT wost Wor WO

ss ——_ prs wo — —— SOTTIA yrevurpury suuvog ‘ON eq

[210,.L, % ul yoyery syooRT) §=Aqturjeg)§= oanjesodway, 1aye\4 =: POUR STIC] ne)

ANNALS OF THE SOUTH AFRICAN MUSEUM

6.6 z €.€ —_ — —_

bea re L.1 — — —

c.L 6 — — ¢.L =

g:-41 gt 9-9 — — oO-jl!

€-98 or §&.€ — — oo

6.6 6 Bots — — 9-9

z-6 Ir §&.§ — — 9G

orci. 61 “Geo — G.a1

Sey Ae > Of Ol 1 — Gz

GeGG ees (Gat 29. — v.&

&.L lee crcl Gil — 4L.0

Leis — 9 — — 3b 6.9

S-GB GI — 9-G — £.01

E-Q% OG @Q: g.G — 0.01

G20le sco, Gal “nc — 36.41

rae) L — ¥.P — gl

0-06 Lz — ¢¢G G¢.¢ 6.91

€.€1 gI — Gs — 9-01

6.€ z — 4L.1 === hat

¢.L 6 g- E.E — §.§

6.§ z Ea — — L.1

0-91 vz — L.g — §.6

£.¥1 9 9-7 — — G.6

QatG- 1G - Gel — 136 — 6-18

G.Gg 62 6. 6.1 — §.a1

o-Gr1 | Gr — — — 2.61

ree) fe — g-!I — +.¥

0.6% && — .P — 9.¥z

L.9gt bo SG. — 3.05 —

@-GSr 61 4g. — oGr —

g-62 th F.1 — 0.93 —

oe ONE SS Ag AT

[e301 % Ur yore)

09 1€.GE

OZI PeeGS

O31 a

06 —

O31 =

06 —

OZ! =

OZI =

OSI —

OZI =

oG1 —

ol —

US =

OSI §¢.G¢g

OSI ¢.G§

Pir oG.G&

06 —

OSI bG.SE

09 —

O31 eb . SE

09 —

oS1 1$.G¢

ov —

99 a

Vil 6¢.c§

VII PG .GE

Vil €¢.c§

Vil 1¢.¢§

06 bV.G§

OZI ov. G§

oS! LV.GE

Pest) WO

syooRT = Aqrunyes

69-61

69.1

3G.s1

ov. SI

Lg.&1

96.61

GS.gI

ZQ. 11

go.11

Wost

6.61

66.61

wos

oinjersdula J, INVA

SOTTIN

20ue}sIq,

aytoay oderyn

ayio0y oder)

Aeg nox]

doysur[s

doysurys

ureyunow 219% “NN M

puvys] ussseq ‘M‘N'M

surquinjory sdey

doysurys

quiog odey

yutog aden

yuiog odey

juIog Josueq

juIog Josueqg *M

yuiog aden

doysur[s

doysur]s

doysurys

doysurjs

doysur[s

auiquinjory ede -M'N'M

ourquinjor) ede) “M"N"M

Sees

SSS Se

SSeeesoe4

puvys] ussseq "NU NUM

aproay eden “a's's

aploay aden “TSS

siouvi1y 1g aden “a's

yIvupue'yT SuLivog

COlCO CO! © OF 69 OE CO CaN 6S OrcOKO co OC OFC AN ©

L onl

™ mem n~nonrm

anon |

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS

CimMDMDomMmOaoaranaramamae none onrwmmnanaonrnmnuouonododrnd onrod

Loy ip)

Pia erie isa Ope. ene) iaes

zee

om oce)

= =

22s

mal |

cP puvysy uosseq ‘MA'N’M _ OO! g/1z

GY doysur|s M 66 G/9z

0% doysurys SXM g6 G/Sz

OdI doysur[s M L6 G/bz

09 doysurjs SKM 96 G/&z

0% yuiog odey M G6 G/zz

GG juIog Josueqg M v6 G/1z

ov juIog Jasueq M £6 G/oz

09 doysurys M 26 b/Ss

09 doysur[s M 16 b/€z

cc doysur[s M 06 b/z

ov doysurys M 6g b/1%

ov doysur[s M 88 b/1z

gl doysurys M Lg b/oz

gI doysur[s M 98 b/oz

cv doysurys M Gg b/61

cP doxysur[s M VE v/61

GP puvys] Uosseqd "M’N'M ~~ &8 b/g1

cP puv|s] Ussseq "M’N'M 28 b/g

09 suIquinjory odey M 1g D/L1

09 suIquinjor adery M 0g b/L1

ov auIquinjor) oder) M 6L b/g

ov uIeJUuNoyy 2IG%T, “M'N'M ~ gh 6/81

ov puvys] ussseq MNM ZL 6/L1

ov puvrysyT uosseq MN'M 94 6/41

ov doysuryg M GL 6/91

GY doysurys M PL £/91

oS JUIOg Josueqd SX M el 6/G1

oS JuUIOg Josueqd SXM ol €/G1

GG juIog Josueqd "M’'S ifs 6/1

GG yulog 1osueq "MS of 6/P1

cP doysurys M 69 6/€1

of doysur|s M 89 6/G1

oo doysur[s M L9 6/11

Gv doysurls M 99 1/6

GP doysurys M G9 1/6

ov juIog 1asueq S 140) 1/g

ANNALS OF THE SOUTH AFRICAN MUSEUM

L.g1 II C.1 C.1 — 9-1 99 — — — 0g. G1 Of doysur[s SXM III 6/S

Q-Ir 6 coal Gr I — 0.6 gl — — — 09.G1 Co doysur]s "MSM OI! 6/€

Wolhte = —— ee = Santa OA SOAS) 140) — — — ov. QI o§ doysurys M 601 6/z

0-9 ¢ — BI — 9-90 bg — — — 0c-QI ov puvys] uasseq "MNM gol g/o&

Q-%o ~=s« GI — 1-61 — &G.6 tg — ov.€1 Z1-91 61-9g1 oS suIquinjor oder) “MS Loi 9/6%

1-61 QI — std — 6.11 40) — — — 06.G1 CV suIquinjory oder M gol 9/ge

GaSe = VG == 626 — 9g.61 ZOI — 6b.z1 99-41 6.41 Ov qulog oder "M’S Cor 9/92

Z-O1 II — — — &.OI gol — — — 09-41 or doysur[s NXM Fol 9/Sz

amt = hs — — — GROG gol — —— — o§.91 cE doysuejs M So1 g/bz

g:0rF bP — 6.61 — _ 6.92 901 — Ib.61 96-91 66.91 CP puvysy uasseq “M'N'M ZOI 9/z

g-1b Gb 6. G-er 6. Gane gol — QG.61 16-61 33.61 CP suIquuinjory ode'y M IOI g/ze

1961

Ye ON 24d AG AA wost wot WO

pesn no ———__ aw yrewpuey Suleog ‘ON 21eq

[210], % ul yoven syooRy Aquyeg oinjesoduray, 1aeajA 90 Ue} SI “ary

ji ghG@ GecoyG ~*~ == =" qulog yaeAIOZ”

Ot aol oer. a= * (juIog 1ayInq) Aeg yoFT ,0S 41 ,6% ,o& asnoyiYyst sutquinjon aden

7,80 Q1 ,gS ,&& Aweoroy S.Jevapeyy) ureyUNOyP[ 2[qz,J, ,S0 81 ,9% 66 * asnoyiYst pueysy uosseq

FOG 401,16 ove ; asnoyYys] yUIOg adery ,oV $6 ,60 ,.vE -° asnoy Ys oproaxy adery

BOG Veal VG SsmounUet stouely 19 oder /81 .61 ,g& .VE ° asnoyYSI] yUIOg 1a8ueq

PUOnae 00. PG °° : " speopT eushuyy ,61 .gt ,60 VE - * asnoyrYysy doxsuejs

a $ a Ss

[918 p9}SI] SYAVUIPUL] IY) JO suOTIIsOg

“pepsosed jou ore (939 “YSYTIIG ‘sysreys) Ysy J9y0 “YDI¥9 oY) UI paysI] s1e vuN} ATUC

‘YAVUIPULT OY Wo Jassoa 94) Aq poi99}s ssIN0d oY) IIe UDAIS sSuTIvaq 2 i

snsago J, = “Ag 2dos0nwm snuunyy, = “Ag Ssasvovqjnu-T = “q'X ‘pSunpojo 7

"19-0961 ysvo0o ade 1s9M pur YINOs oy} IFO Adams vun} oy} SuLAINp ju20g adp/p 1/u1 Aq potdnss0 suo1e}s Jo ae ce

(1) vy xtpueddy

TUNAS OF THE GENUS THUNNUS IN SOUTH AFRICAN WATERS

‘Eg-zQ61 ‘UvIDO ULIPUT 1soM YINOs dy} UT AdAINS VUN) oY} SULINpP 770" *S’y"S Aq potdnos0 suone)s jo ArewuNg

(H) y xipueddy

= — — — — O31 B+Qz /GS oe /9V ,6c (2) hI z/Qz

8 I — Qg: — O31 T+ QZ / GS ,o& /QV 62 Vi z/9z

— ~ — — — OsI 6.1% ,00 ,o& /&V ,9z G1 2/G%

a — — — — O31 6.Qz /1l ov ,95 Qe oI 3/GS

o— — — — -— O31 1.Qo /SV GV /O1 Lz II 2/1S

8 I 8: — — OZI 0. Vz /11 bv /0G 18 OI 2/61

9 I — g- — OSI 0.0% / GS G6 /90 o& 6 o/G1

£961

6.11 VI — 6-8 GE Od! 6.1% , ov 16 ,00 ,0& (2)g 9/&z%

z-6 11 — z-6 — OG1 6.16 ,oV 16 ,00 ,0& 8 9/2

— — — — — OZI 0. €z /VV ,&& /GS Qe L g/%o-1%

— — — — — OSI 1.&% ,16 ,6& 795 ae 9 g/oz-61

_ — — — — OZI Q- 3s ,VV ov / VG Qa G 9/61

= = ass = = O31 L.az /61 bP VP gz v 8/81

— — — — — O31 0-gI / GV ,&V 7Oo o16 (e)& Q/G1

G.z 6 — L.1 8: OZI 0-QI /9¥ &V /QS If € g/S1

1.0 G — Q. 6.& OI 1.06 /Q1 G6 /O6 ,o& re g/a1

o-6 II — 0.S av OZI 1.1% /OV ,o& , £0 ,o& I 9/11

zo61

ad AA AT

T?19.L gS os ee ee Se ee ee SxyOOH] JICJING a S NTI 97eq

% Ul yoier) *du93.1978 \A UOT}ISOg u0l}e1S

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. (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)

BuLLoucy, 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. 19604. 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. 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

y ‘|

Mini, Mitigk

i j

We a he

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 52 ~ ~°&Band

October 1968 Oktober

Paktvge2 deel

NEW QUATERNARY FOSSIL SITES NEAR

SWARTKLIP, CAPE PROVINCE

Q. B. HENDEY & HELENE HENDEY

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

(Cash with order, post free)

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

(Kontant met bestelling, posvry)

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.1.), Ws B))5 8, g(1-2), 10(1-3),

11(1—2, 7> t=P.12)s 21, 24(2), 275 Bu G=2)|5 38,

44(4).

Price of this part/Prys van hierdie deel

R1.75

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

1968

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

NEW QUATERNARY FOSSIL SITES NEAR SWARTKLIP,

CAPE PROVINCE

Q. B. HENDEY

South African Museum, Cape Town

HELENE HENDEY

(With 7 plates and 2 figures)

[MS received 10 September, 1967]

CoNnTENTS

PAGE

Introduction . . ike PoSNAg

Geological Bccociations of the esis See oe is ee

Archaeological associations of the fossils . . 45

The fossils recovered ie ee eae Ag

Oricinvol the tosses ee) eS De oS 4s

The faunal assemblages . . Se Sa BO

Systematic description of faunal ae Pps (fete “Bt

Whe posteranialremams: . . . . +>. 68

Ace.of theassemblagesy, 3) = Ff. se «| JO

Gonclusion ita mar eee cel Pale! et Ah pty SFE

Summary . Le ea (a a 2

ee eteemen sips 8 ie eI oe A Me |

References st A mone os. eR ye

INTRODUCTION

During April, 1966, students from the Department of Geology, University

of Cape Town, discovered an horizon containing fossil bone in the sand- and

limestone cliffs near Swartklip on the False Bay coast. Subsequent investiga-

tions revealed the existence of fossil-bearing deposits at three localities—Site I,

the most prolific occurrence, and Sites II and II Extension, two separate

exposures of a single horizon (fig. 1).

Earlier, Singer & Fuller (1962) had reported on an assemblage of fossil bone

from a fallen block of the cliff-face in the Swartklip (Zwartklip) area. This

discovery is herein referred to as the ‘Singer/Fuller Occurrence’.

All the fossil material recovered from these occurrences is almost certainly

at least broadly contemporary, and it is the purpose of the present report to

place on record the material from the three new sites which had been recovered

up to July, 1966. The fossils, which are housed in the South African Museum

(Natural History), Gape Town, have in many cases been incompletely classi-

fied, largely because of the lack of adequate comparative material. It is hoped

Ann. S. Afr. Mus. 52 (2), 1968: 43-73, 7 pls., 2 figs.

ANNALS OF THE SOUTH AFRICAN MUSEUM

‘(9g QIVE ‘ou y90Yg AdAING “S11, WoIy paidepy) ‘d"+D ‘dippaeMg svau sojIs [ISSOJ MOU JO uoneso] Surmoys dep ‘1 ‘OL

dIALYVMS

SHyID stit2t

"UyX3 I

8 TL sais

LV9I70M

SS 8 a Se MOSES TTS DTIC NT eS 2

NEW QUATERNARY FOSSIL SITES NEAR SWARTKLIP, C.P. 45

to refer back to these specimens as detailed osteological studies of their modern

and fossil counterparts progress.

GEOLOGICAL ASSOCIATIONS OF THE FOSSILS

The geology of the Swartklip deposits has already been recorded in some

detail (Singer & Fuller, 1962). The fossils at the new sites occur in an irregular

horizon of unbedded calcareous sand and grit, at approximately 60 feet above

sea-level. At Site I an extensive talus slope has developed from beneath the

small overhang in which the fossil horizon is exposed, and it was from the

rubble of this scree that most of the fossils were recovered (pl. 1A, B). The

deposits are, for the most part, incompletely consolidated, and their instability

limited the collection of zn sztu fossils.

The fossiliferous horizon at Site I is clearly distinguishable from the overlying

and underlying horizontally bedded deposits (fig. 2), and it appears to fill

what was once a low cave. Gomminuted marine shells occur in great quantities

throughout the deposits incorporating the fossil horizon, and they are clearly

marine in origin, although they may well have been redistributed in part by

subsequent wind action. Their age is uncertain, but Singer & Fuller consider

that they cannot post-date the ‘Riss/Wiirm’ interglacial. It is unlikely that they

are older, since they would have had to withstand marine erosion during the

high sea-level of that period, which is improbable in view of their relatively

unconsolidated nature. The fossil deposits are clearly younger, but geological

evidence can provide no closer estimate of age.

Overlying the horizontally bedded sands are discontinuously developed

horizons of calcrete (‘calcareous tufa’ of Singer & Fuller) and calcareous sands,

partially consolidated reddish-brown sand and unconsolidated white sand, all

of which reflect the more recent geological and pedological history of the

area. Shells of land snails occur within these deposits, which are aeolian in

origin, and the calcareous and ferruginous characters reflect chemical changes

within the soil body subsequent to its accumulation. The calcrete has resulted

from the induration of the calcareous sands.

The geology at Sites II and II Extension is similar in most respects.

Singer & Fuller concluded that the fossils recovered by them probably

came from a crevice in the cliff formation and that the occurrence post-dated

the cliff sands, but pre-dated the calcrete. Observations at the new sites confirm

the former speculation, but do not necessarily support the latter. At none of the

sites is there evidence for a vertical opening through the deposits which has

been sealed by the calcrete. At Site I there is a strong suggestion that the ‘cave’

opening was in a gully to the east of the fossiliferous deposit. The formation

of the calcrete, therefore, had no direct effect on the ‘cave’ opening, and the

fossils do not necessarily pre-date its formation.

ARCHAEOLOGICAL ASSOCIATIONS OF THE FOSSILS

A single undiagnostic silcrete flake was recovered from the talus slope at

Site I, but no material of an archaeological nature was found definitely asso-

ANNALS OF THE SOUTH AFRICAN MUSEUM

*[ A3Ig Jo UOTIIIS YOIDYG *S “Ol

“pups ‘pappaq’ jjpjuozIsOH

Buluado

jowibiO ¢ <—_

e ° .

° . . .

e s e ° . «

“416 puo puns papp2q Ajjoyuoz toy

_ a5 San gee, atts NE a Oueke’ pw a Cie hae) ‘

——/

Sp en cane poe: DO EES

AEE a > = os

pups — UMOIq -ySippay —_

. °

3 =

Do WWD

“3

u) pudDS UDI|O2D

2|D25

271UM paz,OP!OSUO .UN Aq

dow! Xosddy

spy wosy ADMD uUIDIsZAO

NEW QUATERNARY FOSSIL SITES NEAR SWARTKLIP, C.P. 47

ciated with the fossils at this locality.

At Site II Extension, approximately 2 feet above the fossils, is a clearly

defined horizon, apparently an old land surface, from which several flakes

and numerous marine shells were recovered. This midden dates probably

from the Late Stone Age, and may be indirectly associated with the fossils

(uede infra).

From Site II a complete ostrich-egg, with a single perforation on its sur-

face, was recovered (pl. 1C), and this resembles the ‘water-containers’ known

from Late Stone Age contexts, and still used today by the Bushmen of the

Kalahari Desert. The perforation is irregular and appears to have been made

by being chipped away with a sharp object. This is the technique commonly

employed in the manufacture of ostrich-egg water containers (Maggs, 1966),

although in the present instance the perforation has not been rounded out.

If indeed this specimen is a water-container, its association with the fossils

might be coincidental, having been buried into the fossil horizon from the

Late Stone Age living floor immediately above.* These containers were, and

still are, buried when being stored.

None of the fossils show any evidence of human interference, neither

artificially produced fractures (spiral fractures, punctate marks, flaking, etc.),

nor wear through use having been observed.

There is, therefore, no certain evidence for associating the fossils with

human activity.

THE Fosstts RECOVERED

The fossils were, in general, extremely well preserved. This is accounted

for by the alkali nature of the deposits, and by the original protection afforded

them in the caves from natural destructive agencies such as wind, water and

temperature changes.

Damage to the specimens has been largely post-fossilization, and occurred

mainly as a result of the collapse of deposits from the cliff-face, and their

subsequent exposure to the atmosphere. Absorption of salts from the sea air,

and their later crystallization within the fossils has resulted in lamination and

cracking of some specimens.

The fossils are not heavily mineralized, and this rendered the preparation

of those specimens in well consolidated matrix very difficult.

They occurred most commonly towards the base of the horizon, with

several pockets of concentration, and usually lay horizontally in the deposit.

Several instances were noted where elements of the skeleton were in articu-

lation or only slightly separated. Since no systematic excavation was under-

taken it was not possible to assess the frequency with which this occurred. This

feature is a further indication of accumulation under protected conditions.

* The living floor was not actually detected at Site II, but since it and Site II Extension are

exposures of the same horizon on either side of a small spur, it is almost certainly present.

48 ANNALS OF THE SOUTH AFRICAN MUSEUM

ORIGIN OF THE FOSssILs

The lack of positive cultural associations, taken in conjunction with the

small size of the caves, indicates that some agency other than man was respon-

sible for the bone accumulations. The nature of the fossils described by Singer &

Fuller (1962) prompted the observation that, ‘they could well have been

deposited on the floor of a small cave or rock shelter used as a lair’. This view

is supported by observations made in the present study.

In southern Africa an animal lair containing bones immediately suggests

that the porcupine (Hystrix africaeaustralis) is, or was, the occupant. In the pre-

sent instance this possibility was discounted, since the assemblages contained

no gnawed bones. Porcupine lairs contain a high proportion of bones which

show gnaw marks (Hendey & Singer, 1965).

Alternatives were sought among the Carnivora. One of the most striking

features of the assemblages was the variety of carnivores represented. In the

mammalian fauna of a region the numbers and types of carnivores form a

relatively small percentage of the total, and this balance is usually reflected in

the faunal assemblages of fossil sites (table 1). The situation at Swartklip

suggests that the lairs were inhabited by a number of carnivores over a period

of time, and that the assemblages contain the remains of animals which died

there.

TABLE |

THE VARIETIES OF CARNIVORA FROM SITE I COMPARED WITH THOSE FROM

OTHER SITES IN THE CAPE PROVINCE

Open Sites Cave Sites*

Swartklip

Site 1 Melkbos MHopefield A.K. 1 “Aa 2

Animal lairs .. ike x x

Human activity evident ie x x x

Varieties of carnivora : 7 3 12 6 2

Other mammals (excl. micro-

fauna) <6 aes 8 14 38 1] 11

Carnivora as % of total $e 47 18 24 35 18

* Hendey & Singer, 1965.

Perhaps the most significant evidence in support of the ‘lair theory’ is

provided by the official name, Wolfgat, of the area in which the sites are

located (fig. 1). This Afrikaans word, which literally translated is ‘wolf-hole’,

but in free translation means ‘hyena-lair’, is an example of the naming of

localities after animals, an extremely common practice amongst the early

European settlers in South Africa (Sargent, 1954). The use of the term ‘wolf’

for hyenas is now largely an anachronism, but it does persist in the common

Afrikaans names of these animals, i.e. ‘gevlekte wolf’ or ‘tierwolf’ =Crocuta

crocuta; ‘strandwolf’ = Hyaena brunnea; ‘aardwolf’ = Proteles cristatus.*

* Often classed in a separate family, the Protelidae, but undoubtedly considered a true

hyaenid by early non-scientific observers.

NEW QUATERNARY FOSSIL SITES NEAR SWARTKLIP, C.P. 49

Dart (1949) states that ‘most South African carnivora, such as the lion,

jackal and spotted hyena, actually avoid caverns and live out in the veld,

killing their prey and consuming it in the open country. Two South African

types, namely the leopard and brown hyena, are attracted by the protection

of rock shelters and of fissures. . . .. Remains of the brown hyena occur in

both the Site I and Site II assemblages, and since it is also well known as a

scavenger on sea-shores, it is the animal considered most likely to have given

rise to the name ‘Wolfgat’. Furthermore, the presence in the assemblages of

isolated bones of large animals, such as the rhinoceros, suggests that at least

one of the inhabitants of the lair was a scavenger.

The suggestion that the brown hyena was one of the ‘bone-collectors’

must, however, be accepted with reservations. Dart (1956) has gone to great

lengths to prove that it does not accumulate bones in its lairs, and that it does

not defecate in its lairs. Both Sites I and II Extension have yielded coprolites,

and although the animal responsible for these was not identified, they were of

a size sufficient at least to suggest the brown hyena.

The aardwolf also inhabits burrows, and although it is unlikely to have

been responsible for the bone accumulations, being largely insectivorous, it is

reported also to be a carrion eater (Miller, 1954; Maberly, 1963).

In any event, it is certain that within historic times there was a lair of a

‘hyaenid’ in the area where the fossil sites are located. Whether or not this

can be related to the fossil occurrences is a matter for conjecture.

Speculations on other possible occupants of the lairs have proved equally

inconclusive. The African hunting dog (Lycaon pictus), recorded from Site I,

does not normally inhabit lairs, but the females do occupy burrows during the

natal and early post-natal periods.

The lion (Felis leo), although not an occupier of lairs today, must be con-

sidered as a possibility simply because of the number of individuals recorded.

‘Remains of the lion are extremely rare at fossil sites in South Africa, whereas

at least five individuals are represented in the relatively small Swartklip

assemblages.

Fragments of ostrich-eggshell were recovered at all three sites, and unless

these are all from broken water containers, which is considered highly unlikely,

their presence suggests that eggs formed part of the diet of one of the occupants.

The presence of fossils at and near the top of the fossiliferous horizon is

an apparently anomalous situation in the lair theory. There would, however,

have been some fall of sand from the walls and roof of the caves, and with the

accumulation of sand, bones and other debris in the lair, the occupying animal

would have been forced to clear the floor from time to time. In the clearing

of new living space, bones would have been thrown up and in some cases

reached roof level. This regular disturbance would also account for the com-

pletely amorphous character of the sand constituting the fossiliferous horizon.

It is possible that the lair at Sites II and II Extension was occupied at

the time that the land surface on which the midden accumulated was in

50 ANNALS OF THE SOUTH AFRICAN MUSEUM

existence (vide supra). If this was the case, there may well be some tenuous link

between human activity and the fossils, such as, for example, the scavenging of

food remains from nearby human occupation sites by the animal inhabiting

the lair at that time.

THE FAUNAL ASSEMBLAGES

An analysis of the material recovered confirms the observation of Singer &

Fuller (1962) on the great diversity of faunal types represented in a relatively

small total assemblage from a limited area (table 2).

Of the new sites, Site I was the most productive (table 3).

TABLE 2

Site IT Singer/Fuller

Site I Site II Extn. Occur.

Class MAMMALIA

Order ARTIODACTYLA

Family Bovidae

Redunca cf. arundinum

Hippotragus cf. leucophaeus

cf. Connochaetes sp.

Antidorcas marsupialis

australis n. subsp. :

Raphicerus sp... he

Family Hippopotamidae

cf. Hippopotamus sp.

Order PERISSODACTYLA

Family Rhinocerotidae

Diceros simus out

Incertae sedis

Family Equidae

Equus sp. He of a x

x x x XX X

x x x

x x

x x &X x

x X

Order CARNIVORA

Family Hyaenidae

Hyaena brunnea .. a an 4

Family Felidae

(?) Felis serval .. si oe x

Felis leo x

Family Canidae

Lycaon pictus

cf. Canis sp. :

Canis cf. mesomelas

Family Viverridae

(?) Herpestes ichneumon .. se x

Family Mustelidae

Mellivora capensis ee ie ?

Aonyx sp. ae ae shy

Incertae sedis (+1) .. oe a x

NT ee

x X

xX X x

Order RODENTIA

Family Bathyergidae

Bathyergus suillus

Incertae sedis (3) ..

xX X

NEW QUATERNARY FOSSIL SITES NEAR SWARTKLIP, C.P. BE

Site II Singer/Fuller

Site I Site II Extn. Occur.

Class AVES

Order STRUTHIONIFORMES

Family Struthionidae

Struthio australis .. ae ey < x oe

Class REPTILIA

Order CHELONIA

Incertae sedis th ats an s<

TABLE 3

THE FOSSIL MATERIAL RECOVERED FROM THE THREE SITES

some) =

ma 5 =

a ee ela 3 = 4 = a »

oy fare mane eas ae |S “| Sb 1s

6, act th meer peas ot ) S zy 4

oe her ieee |e) S|. oe 3 Sei ees Sy | dee

a FEL) SUS © fy 2 OS

ei|ee|eele=| B | gs | & | bE] &

oe RStar laser | 2S hhp SL Nagle eae es |S

Site FE) ~... ied Pe 270 45 700 25 1630 —_ 199 —_— 7

Site IT .. si at 1 8 100 12 119 | 33 1 —

Site II Extn. _.. =“ 6 3 13 3 34 — 2 — 5

Totals .. es Ey 288 56 813 40 1783 ] 234 1 12

SYSTEMATIC DEscCRIPTION OF FAUNAL TYPES

The specimen numbers are those in the accession registers for the Swart-

klip area in the Subdepartment of Quaternary Palaeontology, South African

Museum, and all comparative material referred to is in the collections of this

museum.

Unless otherwise stated the material described is from Site I. Measure-

ments are in millimetres.

Where possible the nomenclature follows that of Ellerman ef al. (1953).

An exception to this is the use of the generic name Felis for the lion, in place

of Panthera (Hershkovitz, 1959).

Family Bovidae

Tribe Reduncin

Redunca cf. arundinum Boddaert

Reedbuck

At Site I a minimum of six individuals were represented, while there was

at least one from both Sites II and II Extension.

52 ANNALS OF THE SOUTH AFRICAN MUSEUM

KW 376—(¢4) (table 4, pl. 2A, B)—This is the skull of an adult, with the

braincase almost intact, but having lost the tympanic bullae, most of the basi-

occipital, the sphenoid, the zygomatic arches, and parts or all of other delicate

features from the base of the skull. Both horn cores have lost part of their

distal extremities. The frontals are largely intact above the supraorbital

foramina, but those parts forming the roof of the orbits have been lost.

TABLE 4 :

DIMENSIONS OF 6 Redunca arundinum SKULLS COMPARED WITH THOSE

OF TWO FOSSIL SPECIMENS FROM SWARTKLIP

Fossil Specimens Modern R. arundinum (6)

No.

ZW 376 ZW 1316 | Mean Range Measured

Postorbital constriction uy c75:0 76-0 69-5 — 1

Max. length across occipital

condyles a ei: me 593 58-8 57-2. 56-5-57a7 4

Max. skull width in occipital

plane .. a ms ee 10265 102-1 98-2 94-7-101-0 4

Height from most superior point

of foramen magnum to

lambda* Me oe 2 35-0 36-8 37-5 36-°8-38-6 4

Dimensions ( Ant./Post. bh 35°9 35°5 35-1 31-9-38-0 4

of horn

pedicle Transverse an 38-0 34-0 34-3 32-2-36-6 4

Skull width at horn pedicles .. 102-4 105-6 99-7 95-4-102-0 4

* Point on median line contact of occipital and parietal.

KW 1316—( 3) (table 4, pl. 2C)—The skull of an adult from Site II. It corre-

sponds closely in size and morphology to ZW 376. Except for the horn cores,

of which only the bases remain, it is more complete than ZW 376, and lacks

only the facial region, the zygomatic arches, the tympanic bullae and rims of

the orbits.

The braincases of these two specimens have distinctly flattened dorsal

surfaces, with the nuchal crests not as strongly arched as in the comparative

specimens. This morphological difference is not clearly expressed by the skull

dimensions, the width/height ratio being 1: 2-8 in the fossil specimens and

1: 2°6 for the mean of the comparative series. It has clearly resulted from an

upward expansion of the braincase on either side of the median line. In all

other observable respects the two fossil skulls were similar to those of the

comparative series.

KW 74—() (tables 5 & 6, pl. 2D) —The almost intact skull of a young adult,

with the M? incompletely erupted. It lacks the nasals, the premaxillae and

parts of the maxillae, and is slightly damaged in the parietal and occipital

NEW QUATERNARY FOSSIL SITES NEAR SWARTKLIP, C.P. 53

regions. All the molars are present and well preserved, as is the left P*. The

specimen ZW 712, a fragment of right maxilla with the P?, P®, and P* belongs

almost certainly to ZW 74.

TABLE 5

DIMENSIONS OF 2 Redunca cf. arundinum sKULL FROM SWARTKLIP (Zw 74)

Postorbital constriction o: a dye ee. a8 ae vA ia ven AChe?

Length between external occipital protuberance and bregma .. te ae oe Ors

Width between external alveolar margins at midpoints of M? .. e: ae Sey ENE

Height from basioccipital to most superior point of occipital .. is ae ap, O64

No female R. arundinum skulls were available for comparison, and conse-

quently no conclusions could be drawn as to the relative size and morphology

of the fossil specimen.

In addition twenty-two isolated upper teeth and maxillary fragments

were assigned to Redunca. These included deciduous teeth. The permanent

teeth were, without exception, longer than those of the comparative specimens,

while the breadths were greater than, or near the top of the range of the

comparative series (table 6). The six modern R. arundinum skulls examined

were those of males with horn lengths ranging from 24 cm to 30 cm along the

front curve. They were, therefore, above average size for the species (Ward,

1899).

TABLE 6

DIMENSIONS OF UPPER TEETH OF Redunca arundinum COMPARED WITH THOSE OF

FOSSIL SPECIMENS FROM SWARTKLIP

Fossil Specimens Modern R. arundinum (3)

No.

ZW 74 ZW 712 ZW 80 ZW 392 | Mean Range Measured

p2 fl — 7-0 — — 6-4 5°5-7-2 2

Ub — a-9 — — a2 4-8-5-4 2

ps 1 — €9°5 — — 8-6 7-9-9: 1 4

b —- 8-0 — — 8-5 8-3-9-0 4.

ps ] 11-0 10-9 — — 9-1 8 -3-9-7 a

b ‘ 9-4 | — — 8-7 7-8-9-1 4

1 cl6-5 — — he=o 12-2 10-4-13-8 6

M? < bm 12-8 — — 10-7 EL 2 10-7-11-7 3

bd Pt — — 9-4 10-9 10-1-12-2 5

] 17-9 _ — 18-0 ba-9 12-1-15-0 6

Bee. bm... 12-1 a — 3-9 10-9 9-8-11-7 6

bd i, 10-8 — — 8-0 10-2 9-5-10-8 4

] ai the? — 17-9 — UGICS: 13-3-17-3 6

Seibm... 10-4 — ji Ree — 10-0 8-8-10-8 5

bd a 9-2 — O= 7 — 8-5 7-8-9-1 5

_. Sixteen mandibles, mandible fragments, and isolated lower teeth of

Redunca were also included in the assemblages, but since no mandibles of the

modern reedbuck were available for comparison, they were not studied in

detail.

The available evidence suggests that the Swartklip Redunca is at least

54 ANNALS OF THE SOUTH AFRICAN MUSEUM

subspecifically distinct from the extant form. In view of the inadequate range

of comparative material, the degree of difference could not be accurately

assessed, and the fossil material is therefore only tentatively assigned to

R. arundinum.

Tribe Aippotragim

Eippotragus cf. leucophaeus Pallas

Blue Antelope

The remains of at least four individuals were recovered at Site I.

The tentative conclusion on the presence of the recently extinct blue

antelope was based on three specimens, which belong almost certainly to a

single individual.

ZW 355—(tables 7, 8 & 9g, pl. 5A, B)—A right mandible with a complete set

of moderately worn cheek teeth.

KW 375—An incomplete left mandible with P,— Msg.

KW 357—RP?.

The incisal alveolar region of ZW 355 is missing, as is much of the ascending

ramus, and that part of the mandible in the region of the angle. The teeth and

remainder of the corpus are well preserved. The teeth are clearly hippo-

tragine, but differ significantly from those of the extant roan (H. equinus) and

sable (H. niger) antelopes. They do, however, closely resemble the teeth of the

type specimen of H. problematicus Cooke (1947). This specimen is from Bloem-

bos, near Darling, about 50 miles north of Swartklip.

In H. problematicus (SAM 661A & B) the premolar series is proportionately

longer than the molar series, the premolars resembling those of the roan

antelope, while the molars are intermediate between those of the roan and

sable antelopes (Cooke, 1947). This feature is also exhibited by the Swartklip

specimen (table 7).

TABLE 7

THE LENGTHS OF THE CHEEK TEETH SERIES OF THE SWARTKLIP Hippotragus,

COMPARED WITH THOSE OF OTHER HIPPOTRAGINES

H. problematicus* H. equinus* —_H1. niger*

ZW 355 (S.A.M. 661A) (Mean) (Mean)

P, — P, length Ne oe ayer 53-0 51-0 45-0

M, — M, length Ay) Ags 70-2 71-0 78-0 68-0

* Cooke, 1947.

In general the teeth of ZW 355 correspond closely to those of the H. pro-

blematicus type specimen in size, although the M, and M, are slightly narrower

(table 8). Morphologically the differences are also slight. The anterior accessory

cusp of the P, of SAM 661A is seen on the lingual surface as a clearly defined

vertical pillar, whereas in that of ZW 355 the base of the cusp broadens out

NEW QUATERNARY FOSSIL SITES NEAR SWARTKLIP, C.P. 55

posteriorly and is not as well defined as an ‘accessory’ feature. The P, of

ZW 355 has a small accessory cusp developed on the anterior part of the buccal

surface, and this is not present on the P, of SAM 661A. In ZW 355 the anterior

enamel folds are more strongly developed than those of SAM 661A. In the M,

of ZW 355 the external accessory pillar between the lobes is separated from

the folds of buccal enamel, whereas in SAM 661A it has been incorporated.

Usually this might be ascribed to different degrees of wear, but the teeth of

the two specimens have very similar crown heights.

TABLE 8

DIMENSIONS OF THE LOWER TEETH OF THE SWARTKLIP Hippotragus, COMPARED

WITH THOSE OF THE TYPE SPECIMEN OF H. problematicus

ZW 355 S.A.M. 661A

Sa Re 15-0 14-8

athe 11-9 11-6

ane 17-0 17-4

b 14-0 14-3

px J! 18-6 19-1

NG 14-8 14-1

l 20-0 19-3

Mj ~bm 13-1 13-7

bd 12-1 cl2-0

l 22°5 23-0

Mj ~bm 11-9 13-4

bal aa: 12-0 12-8

eee 28-7 30-1

M,t ~bm .. 11-2 12-5

ba 10-8 11-8

* Maximum dimensions midway up crown.

{ Dimensions at occlusal surface.

The two specimens differ most widely in the relative sizes of their mandi-

bular corpora, the Swartklip specimen being considerably more robust (table

g). This is possibly a reflection of a difference in sex.

Cooke (1947) suggested that H. problematicus may prove to be a synonym

of H. leucophaeus Pallas, which became extinct in the Cape Province at the end

TABLE 9

DIMENSIONS OF THE MANDIBULAR CORPUS OF THE SWARTKLIP Hippotragus,

COMPARED WITH THOSE OF THE TYPE SPECIMEN OF H. problematicus

ZW 355 S.A.M. 661A

Height below midpoint of M,* .. c62-0 —

Height below midpoint of M,* .. 53-4 44-0

Height below midpoint of P,* .. 40-0 30-9

Distance between post. end of

mental foramen and ant. end of

P, ee at aN a 46:8 40-0

* Measured on lingual surface.

56 ANNALS OF THE SOUTH AFRICAN MUSEUM

of the eighteenth century. Later he expressed reservations on this question

as a result of information supplied him by the late Robert Broom (Cooke,

1955), but as far as can be determined Broom’s opinion remains unsubstan-

tiated.

Nevertheless, in view of the geographical position of Swartklip and the

relatively recent age of the fossils (vide infra), it is considered that sufficient

justification exists for tentatively assigning the Swartklip hippotragine material

to H. leucophaeus.

The specimen ZW 357 (RP?) is considerably larger than the corresponding

teeth of the roan and sable antelope comparative specimens, having a maximum

length of 18-3 and a maximum breadth of 22-1 mm.

Tribe Alcelaphini

cf. Connochaetes sp.

The minimum number of alcelaphines represented was three. All the

specimens recovered were assigned to the genus Connochaetes (sensu lato) (pl. 5D).

In general the material showed greater affinities to C. gnou, rather than

C.. (Gorgon) taurinus, but owing to the lack of adequate comparative information,

no conclusion on the specific rank of the fossil form was reached.

Tribe Antzlopini

Antidorcas marsupialis australis n. subsp.

Diagnosis

An Anitidorcas differing from the extant A. marsupialis principally in the

size and shape of the horn cores of the male. They are notably more slender

in the fossil form, being intermediate in size between those of the male and

female of the modern form. Although lyrate in shape they show no indication

of an inward curve towards the tip. They rise vertically above the orbits as

in the modern form, but then have only a slight backward sweep. The brain-

case of the fossil specimens is slightly longer and lower than that of the extant

springbok, but in general its morphology is similar. A notable exception is

that the basioccipital and body of the sphenoid of the new subspecies lie in a

plane with the inferior margin of the occipital condyles, whereas in the modern

springbok the sphenoid curves upwards fairly sharply immediately anterior of

the basilar tubercles.

Morphologically the teeth of the modern and fossil forms are indisinguish-

able, although those of the new subspecies tend to be rather narrow.

Material

KW 66—(3) (table 10, pl. 3)—The skull of an adult consisting of a complete

braincase and frontlet. The left horn core is intact, but only the base of the

right core is preserved. The braincase has suffered some slight superficial

damage, and delicate parts such as the tympanic bullae have been lost. The

left horn core, from which the extreme tip is missing, is damaged on the

GS Lome Ome 99 I 8°69 = 03992 ae 8°89 “ "* spejotzed ssoroe YIPIM *xeyy

6°¢68-8:L2L L:6L I 6:69 mS — — 0°64 ** sassa001d PIOjsSVUl 1% YIPIM [[NySs “Key

0:LP-9:°Ib 9ST I 0:0 aie 0:99 = 0: PP sajApuoo *di900 ssoioe yISUST “xXe]/[

DoW) 9204 22) I 8:09 as §599 = 0°02 = = "* Burseiq pue

souvieqnjoid ‘di990 *}x9 useMJEq YISUIT

8°Sb-O'Ib GS: tr I 6°9E — 0:02 — ¢-OF souvieqnjor1d *dio90 [eusayxe Jo do} 03

UINUSeUL USUIVIOJ JO aseq WOT, IYSIO-Y

€°6&-6°GG 1:64 6 GEST 6°L6 6°66 6°16 0:F<6 osroasuesy, | aporped us0y

T:07-0°&& GLE 6 L:6c-0°bI 6°81 G: 66? 0:46 8°86 L:8¢ ie "* ysog/ywy f jo suotsuawq

9 le ir Ue@ I Corsi 6 6:°1&-0°S6 0°82 0:10? 0:10 0-619 0°6l PERRO AIEAOUT EMIS TET RTO

6 8°6£-6:0L §&°GL & 9° 19-46 Sees EVE 0:92 0-792 8°L9 soporped UroY 38 YIPIM [NAS “xe

NEW QUATERNARY FOSSIL SITES NEAR SWARTKLIP, C.P.

‘ON "ON |

P supidnsivu 5 Syvidnsinu "v

MO@MONIAdS NYAGOW AHL JO ASOHL HLIM

G#xVdWOD dITHLUVMS WOUd ‘dsqns ‘u szvysno syvidnsivU svdopYUup AO STINAS AHL JO SNOISNAWIG

Ol ATav.L

538 ANNALS OF THE SOUTH AFRICAN MUSEUM

anterior surface of the proximal end. There is, however, no indication of

transverse ridges on this, or any of the other horn cores.

ZW 402—(3) (table 10, pl. 4A)—The skull of an adult consisting of the

frontlet, left and right horn cores and incomplete braincase. The tip of the

left horn core is missing. The base and sides of the braincase are extensively

damaged, and approximately fifty fragments of bone from this region were

not restored to the skull. The largest of these fragments includes the basi-

occipital with the right occipital condyle, and the sphenoid. This specimen

compares remarkably closely with ZW 66 in all observable respects.

KW 67a—(8) (table 10, pl. 4B) —The right horn core and part of the frontal

of an adult. It corresponds closely with ZW 66 and ZW 402 in all observable

respects.

KW 1317—(3) (table 10)—An incomplete and heavily weathered skull from

Site II, with only the base of the horn cores and part of the frontal and parietal

remaining. The dimensions of the horn cores correspond more closely with

those of the comparative specimens than do any of the others.

KW 48—(2)—The only female specimen known from Swartklip, having come

from the Singer/Fuller Occurrence, and referred to Antidorcas marsupialis or

Gazella wellst Cooke by them. The specimen consists of an incomplete left horn

core and part of the frontal. Owing to its condition, accurate measurements

were not possible, but it appears to fall within the range of variation of the

female springbok.

KW 68—A fragment of the roof of a braincase, very similar to the corre-

sponding parts of ZW 66 and ZW 402.

KW 69, LW 70 & KW 71—Horn core fragments, indistinguishable from the

corresponding parts of the more complete specimens.

KW 151 & KW 161—LM3 and incomplete RM, indistinguishable from the

M®? of the modern springbok.

AW 1571—LM! from Site II Extension, somewhat smaller than the M? of the

comparative specimens.

KW 81 & KW 82—(table 11, pl. 4C) Incomplete left mandibles, with the P;

lost from both specimens. As is characteristic of Antidorcas there is no Py. In all

respects these specimens correspond closely to the comparative series.

Discussion

The minimum number of individuals represented in the Site I assemblage

was four, while there was at least one from both Site II and II Extension.

Three subspecies of Antidorcas marsupialis Zimmerman are generally

recognized: marsupialis in the Transvaal, Orange Free State and south of the

Orange River; angolensis in the Kaokoveld and northern Namib Desert; and

hofmeyrt in Great Namaqualand and the Kalahari Desert (Roberts, 1951;

Ellerman et al., 1953).

NEW QUATERNARY FOSSIL SITES NEAR SWARTKLIP, C.P. 59

TABLE 1]

DIMENSIONS OF THE LOWER TEETH OF THE FOSSIL Antidorcas FROM

SWARTKLIP COMPARED WITH THOSE OF THE MODERN SPRINGBOK

Modern springbok

(3 specimens)

ZW 81 ZW 82 | Mean Range

P ] ef — 8-6 8-9 7-7-10-0

b he — 5:0 5:3 4-4-6 -0

1 - 13-0 12°5 13-1 11-7-14-0

M, bmi C2 6-1 6-4 6-8 6-5-7-0

bd ae 6-7 6-9 7-8 6-9-8-2

] ee 14-5 14-9 14:9 12-8-16-4

M, bar =: 6-2 ya 7°4 7-+3-7-4

bd : 6-4 4-3 Eat 7:3-8-4

1 : 20-4 21-4 DROZ | 22-7326

Me= bn.) -: 5°5 7-0 2 6-7-7-8

bd “4 Dad eo 7-6 6-9-8 -3

M,-M,; 48-1 47-6 50:0 46-1-53-0

* M, of ZW 81 incompletely erupted.

Unfortunately the localities from which the comparative specimens came

were not all known, but at least three of the males were from South West

Africa. These were, therefore, probably either angolensis or hofmeyri, while most

of the others were almost certainly marsupialis. Blaine (quoted by Roberts,

1951) in his definition of angolensis stated that the horns of this subspecies are

not as stout at the base as those of marsupialis, while the skull is, ‘relatively

longer, narrower and vertically less deep’. This description could apply to the

new subspecies, but Thomas (1926) in his definition of hofmeyri stated that it

has the ‘same long heavy horns’ of angolensis. Consequently since the horns of

these two subspecies are apparently similar, and since the comparative series

included specimens of at least one of them, the characteristics defined by

Blaine are obviously not of a degree sufficient to make the horns of angolensis

comparable in size to those of the new subspecies.

Clearly then the horns of the males of the extant subspecies, although

variable, differ significantly from those of the fossil specimens. These are gracile,

not heavily ridged, do not have the marked sweep posteriorly, and do not appear

to recurve inwards towards the tip. They show a considerably greater diver-

gence in character to those of the extant subspecies than these show relative to

one another.

Gentry (1964) has shown the importance of the transverse compression

of horn cores in the classification of gazelles, and applying his method of

quantitatively expressing this feature, it was found that the horn cores of the

Swartklip Anizdorcas correspond very closely to those of the modern springbok

(table 12). In this respect at least, the horns of the modern and fossil forms are

similar.

The angulation of the sphenoid relative to the basioccipital in the modern

60 ANNALS OF THE SOUTH AFRICAN MUSEUM

springbok is a feature which relates directly to the size and orientation of the

horns. Gentry (1964) has indicated the relationship between the degree of

development of the basilar tubercles and horn size in gazelles, and in the

slender-horned Swartklip Antidorcas these tubercles are indeed less pronounced

than in the modern springbok. With the marked backward sweep of the horns

in the latter, an adaptation of the angle of attachment of the neck muscles to

the skull would be necessary to compensate for the relative posterior shift in

weight of the horns. This requirement could be met by the downward move-

ment of the basioccipital relative to the sphenoid. The basioccipital of the new

subspecies is somewhat more rectangular than that of the modern springbok

and this feature, together with the slightly shorter and lower braincase, are

probably related expressions of the same factor.

TABLE 12

THE TRANSVERSE COMPRESSION* OF THE SWARTKLIP Antidorcas

HORN CORES COMPARED WITH THAT IN A SERIES OF MODERN SPRINGBOK

Transverse diameter

——___—_—_—_————— x 100

Antero-posterior diameter

Mean Range No. measured

Swartklip specimens .. 80-85 73 6-84 -0 4

Modern A. marsupialis 3 80-56 75 - 7-87 +3 8

Modern A. marsupialis 9° 83-47 76 -9-90-7 3

* Gentry, 1964.

Recently Gentry (1966) has recorded from Beds I and II at Olduvai

Gorge, a series of gazelline horn cores (his Group (B): 65-66, pl. 2C) which,

as far as can be determined, compare closely to those of the Swartklip Anti-

dorcas. He refers these specimens tentatively to Gazella, but concludes that they

are unlike the horn cores of any living gazelle. The illustrated specimen

(1953, SHK II, 285) in lateral view is very similar to the Swartklip horn cores,

and its dimensions (28-2 < 23-1 mm) fall within the range of variation of the

Swartklip specimens. If this Olduvai Gorge form is indeed the same as that from

Swartklip, it indicates a wide temporal and geographical range for the animal.

It is also almost certainly present at the Elandsfontein (Hopefield) site, where a

number of horn cores and dentitions of this and other Antilopini have been

recovered. i

The taxonomic status of the Swartklip Antidorcas does, in part, hinge on

the question of the geographical distribution of this genus. The extant sub-

species of the springbok are plains-dwelling animals confined largely to the

inland plateau and the western coastal region of southern Africa (Ellerman

et al., 1953; Roberts, 1951). The natural pattern of their distribution has been

extensively disrupted by human settlement, and while the springbok may have

occurred in the south-western Cape during the seventeenth century, no his-

torical records to this effect could be located. If it was indeed present, there

NEW QUATERNARY FOSSIL SITES NEAR SWARTKLIP, C.P. 61

will certainly be no adequate descriptions of the physical character of the local

form available, and reference to the fossil record must therefore be made. The

Swartklip fossils do indicate the presence of a springbok differing slightly from

the subspecies occurring further north.

In the light of available evidence it is concluded that the Swartklip Anti-

dorcas is at least a geographical variant of the extant species, and in view of the

location of the sites, the taxon Antzdorcas marsuptalis australis is proposed. How-

ever, if the suggested wider relationships of this form can be convincingly

demonstrated, its elevation to the species level will need to be considered.

Tribe: Neotragint

Raphicerus sp.

The minimum number of neotragine antelope represented in the Site I

assemblage was six, while a single individual was represented at Site IT.

The two extant South African species of Raphicerus, namely campestris

(steenbok) and melanotis (grysbok), cannot be satisfactorily distinguished on

osteological grounds. Both are found in the south-western Cape Province today,

and the Swartklip specimens show no significant differences from the living

forms.

Singer & Fuller (1962) stated that the Raphicerus specimens recovered by

them had closer affinities to the recent forms than to the fossil form from the

Elandsfontein site.

Family Bovidae: Incertae sedis

Almost one hundred skull fragments and isolated teeth, mostly incomplete,

from the Site I assemblage were not classified, but none suggested the presence

of an antelope other than those already listed.

Family Hippopotamidae

cf. Hippopotamus sp.

A single unerupted (? deciduous) tooth, doubtfully ascribed to the hippo-

potamus, was included in the Site I assemblage.

Family Rhinocerotidae

Diceros (Ceratotherium) stmus Burchell

White Rhinoceros

Remains of the rhinoceros were rare in all the assemblages, and only a

single unerupted RP? (ZW 192) from Site I was identified positively as belong-

ing to Diceros simus. This species was not recorded in the south-western Cape

Province during historic times, but has been recorded from the Langebaanweg,

Hopefield and Melkbos fossil sites.

62 ANNALS OF THE SOUTH AFRICAN MUSEUM

Diceros sp.

The only other cranial remains of the rhinoceros recovered were four tooth

fragments. Owing to their condition it was not possible to assign them to any

species.

Six elements of the postcranial skeleton were recovered from Sites I and

II, including a radius and ulna (ZW 292 and ZW 1120) found in close asso-

ciation at Site I. |

It is generally considered impossible to distinguish the white from the

black rhinoceros on the basis of their postcranial skeleton (Hooyer & Singer,

1960), but the specimens from Site I compared closely in size with corresponding

elements of the skeleton of a white rhinoceros in the South African Museum

(SAM 21379). However, the postcranial material recovered may belong in

part, or entirely to the black rhinoceros (Diceros bicornis), which was recorded

in the area during historic times, and which is commonly found at other fossil

sites in the area.

Family Equidae

Equus sp.

No cranial remains of this genus were recovered, but several elements of

the postcranial skeleton were included in the Site I assemblage. ‘They repre-

sent the remains of at least two individuals, an adult and a juvenile.

Family Hyaenidae

Hyaena brunnea 'Thunberg

Brown Hyena

The minimum number of individuals represented in the assemblages was

three.

ZW 394—A right premaxilla, probably of an immature adult, with only the

I? remaining in position. This tooth is fully erupted, but unworn.

KW 1311—(table 13, pl. 6A, B)—An incomplete skull of an immature indi-

vidual from Site II, with associated right and left mandibles (ZW 1312 &

ZW 1313). The facial part of the skull is largely intact, but most of the brain-

case has been lost. The teeth preserved are as follows:

Left: Roots of I? and di’; I? and I? about to erupt; root and damaged crown of

dc; P!; dp?; dp*; dp*; P* about to erupt; M! just erupted.

Right: As above, except that the I! is intact, only the posterior root of the dp?

remains, and the P! has been lost.

The right mandible (ZW 1312—pl. 6C) has the following teeth present:

I,; dc; dp.; dps; dp,; and M, partially erupted.

The left mandible (ZW 1313) is incomplete, and only the dp, and part

of the dp, are preserved.

NEW QUATERNARY FOSSIL SITES NEAR SWARTKLIP, C.P. 63

TABLE 13

DIMENSIONS OF THE TEETH OF AN IMMATURE Hyaena brunnea,

SPECIMEN (ZW 1311) FROM SWARTKLIP

Pp! dp? dp? dp* M!

] b l b l b l b 1 b P1 — dp*

meme OF Sot. Nene 2) 241 1323 9-6 | 14-2 el2Z3)* 1674" 51-5

* At midpoint of tooth.

No skulls of immature brown hyenas were available for comparative

purposes, but the permanent teeth present corresponded closely with those of

the adult comparative specimens.

A fragment of right maxilla (ZW 141) with the dp? and dp? from Site I,

closely resembled the corresponding part of ZW 1311 in all observable respects.

Family Felidae

? Felis serval Schreber

Serval Cat

At least two individuals were represented in the Site I assemblage.

KW 110—(table 14, pl. 7CG)—An incomplete skull of a small felid consisting

of the braincase, the posterior part of the palate and left maxilla, and the left

carnassial. This tooth was somewhat shorter than those of modern serval cats

(Roberts, 1951), but since it is poorly preserved the dimension given may be

inaccurate. An isolated felid P* from the Site I assemblage (ZW 183) has a

length of 14-0 mm, which is within the range of variation for this species.

TABLE 14

DIMENSIONS OF THE SKULL OF A ? Felis serval FROM SWARTKLIP

COMPARED WITH THOSE OF MODERN SPECIMENS

Felis serval subsp.*

3 adult 3

ZW 110 3 2? -— 3

Min. Max.

Intertemporal constriction | 30-8 — 30-0 29-5 35-0 g2°5 35-2

Braincase width .. wd O22 — 49-7 52-5 56-0 52:5 52-6

Bullae— greatest diameter | c24:5 26:5 23-5 PES) Delete 26°5 25-0

P* length .. oe | e12-0 16-0 13-0 13-9 15-3 14-0 14-8

* Roberts, 1951.

KW 377—The anterior part of the snout of a small felid, with the LI‘, LI?,

RI?, RI? and RC.

Owing to the lack of comparative material these specimens are only

doubtfully referred to Felis serval.

64. ANNALS OF THE SOUTH AFRICAN MUSEUM

Felis leo Linnaeus

Lion

A minimum of four individuals was represented in the Site I assemblage,

and one in the Site II assemblage.

KW 100—(pl. 7A)—An incomplete skull consisting of the braincase and part

of the frontals. The occipital is intact except for that part in the region of the

left mastoid process. The right tympanic bulla is slightly damaged and the

left completely absent. The right temporal is complete and both the articular

groove and postglenoid process are present. The basioccipital and sphenoid

are largely intact. The left side of the braincase is damaged and most of the

parietal and temporal have been lost.

This specimen was morphologically indistinguishable from the skulls of

the modern lion examined, and in size fell within the range of variation of the

comparative series.

KW 101—An incomplete braincase lacking the top of the nuchal crest, the

sagittal crest and immediately adjacent parts. It compares closely in all obser-

vable respects with ZW 100, but is less well preserved.

KW 131 (& <W 174)—An incomplete left maxilla with the dp*, dp* and the

M! just erupting. The C, P? and P* are present but unerupted. The unerupted

teeth and the M! were similar in all respects to those of the comparative series.

KW 106—(pl. 7B)—A left maxillary fragment with the dp? and dp%, and the

unerupted P*.

Eight other isolated deciduous or unerupted permanent teeth were

recovered, in addition to a fragment of right maxilla (ZW 144), which belongs

almost certainly with ZW 106.

The deciduous teeth show all the characteristics of those of the modern

lion (Broom, 1949).

Few elements of the postcranial skeleton were recovered. A first phalanx

(ZW 237), two fragments of a right radius (ZW 350 and ZW 526) and the

distal end of an ulna (ZW 1209) were the only postcranial remains of adults

recovered, the last-mentioned from Site II. A single metapodial (ZW 261)

of an immature individual came from Site I. None of these specimens, with

the possible exception of the ulna, was of unusual size.

Family Canidae

Lycaon pictus ‘Temminck

African Hunting Dog

Two isolated teeth, a LI? (ZW 137) and an incomplete RI? (ZW 177),

were assigned to this species. They were indistinguishable in all respects from

the corresponding teeth of the comparative series.

NEW QUATERNARY FOSSIL SITES NEAR SWARTKLIP, C.P. 65

cf. Canis sp.

KW 1314—A left maxillary fragment with the P* and M! from Site II. While

the teeth resemble those of Canis morphologically, they fall beyond the size

range of variation of the two extant South African jackals (C. mesomelas and

C. adustus).

KW 108—A canine which is significantly larger than those of the two extant

jackals, but appreciably smaller than that of Lycaon pictus.

Canis cf. mesomelas Schreber

Black-backed Jackal

This was the most commonly represented carnivore at Site I, the remains

of at least six individuals being present.

The tentative diagnosis of the specimens was based principally on the

size of the lower carnassials, and the carnassial: molar ratio (Ewer, 1956a),

which excluded the possibility of the material belong to the side-striped jackal

(C. adustus) (tables 15, 16).

TABLE 15

DIMENSIONS OF THE LOWER TEETH OF Canis cf. mesomelas FROM SWARTKLIP

COMPARED WITH THOSE OF A SERIES OF TEN C. adustus AND FOURTEEN C. mesomelas

Swartklip specimens C. adustus* C. mesomelas*

i No.

Mean Range measured | Mean Range Mean Range

P e aa == 2 Oe 2-9-4-0 4-0 3:-44-8

7 Yb 332) — 2 233 eV) 27 2-3-3 -2

Pp - G20 oF oF l—o-9 3 7-6 de Do 2 8-0 7-2-9-0

pani als, 4-0 3-84-4 3 2 | 2-8-3 -5 3e3 2-4-3-8

P ti 9-4 9-1-9-7 4 8-6 8-0-9-0 8-8 7-1-10-0

cane 2, segs dy2 | oaacls 4 orre 2-8-4-4 aD 3-0-4-0

P ‘e 1029" 10= 5-123 4 9-8 9-3-10°6 10-3 9-2-11-5

tc b ape) Gee C4 4 ane 3:7-4-7 4-3 3-6-4:8

M 4 20-5 19-6-21-0 5 16-2. 15-0—17-4 18-8 16-5—-20-7

aah es oo) sot oo 5 6-1 5:1-6°8 7-2 6-5-8-2

M i 327) 3-0-9 -4 6 leas 8-7-10-0 8-5 7-4-9-6

eb 6-7 6-0-7-1 6 ved 5:2-7-0 6-3 5:7-7-0

M it 4°5 — 1 4-6 49 —o 0 AD. 3:5-4-9

Bab Se 4-] = 1 S10¥) 3-5-4:5 3-8 3:3-4-3

Eee, |... 35-9 34-6-38-0 3 39°0) 3L-7—36-4 | 34-6 $1-5—37-1

Mar Mi br) es. 2:46 2-35-2-58 3 1-73 1-60-1-81 Boel, 1-99-25 7

* Ewer, 1956a.

Although available evidence indicates the presence of C. mesomelas in the

assemblages, it may prove on closer examination that all the canid material

from the Swartklip sites belongs to an extinct member of the family.

66 ANNALS OF THE SOUTH AFRICAN MUSEUM

TABLE 16

DIMENSIONS OF THE UPPER TEETH OF Canis cf. mesomelas FROM SWARTKLIP

COMPARED WITH THOSE OF A SERIES OF TEN C. adustus AND FOURTEEN C. mesomelas

Swartklip specimens C.. adustus* C. mesomelas*

No.

Mean MRange measured | Mean Range Mean Range

P 5:0 — 1 3-6 2:9-4-1 4+4 3-8-5-0

b 4-8 —- 1 — — — —

iO “ 6:4 6:0-7-1 5 4-4 3-8-5-0 a3 4-6-6°3

b 5:2 4:-9-5-8 5 — — — —

pl i 5:0 — 1 4-] 3-4-4-9 5:0 4-1-5-8

b 3°5 — 1 TAS: 2:2-2-8 3-0 2-2-3°5

pe at 9°5 — 1 UST, 7-2-9-0 8-6 7:4-9-6

b 4-2 — l — — — —

ps of 10-5 — 1 9-2. 8 -5-9-8 9-6 8-7-11-0

b 4-9 — 1 — —- — _

ps i. 18-5 — l 14-4 12°8-15-6 | 17-1 t5=2isat

b 9-7 — l 6-7 5:5-7+7 7:8 6-8-8-7

M1 4 11-0 — ] 11-7 10-4-12-7 11-8 10-3-13-7

b 15-0 — 1 12-5. 10-7-13°3 | 13-7 11-7-15°-5

We hi 7°3 — 1 7:7 7-0-8:-6 6:8 6-2-7-8

b 10-9 — 1 10-4 9-3-11-8 | 10-5 9-8-11-5

* Ewer, 1956a.

Family Viverridae

? Herpestes ichneumon Linnaeus

Egyptian Mongoose

A single specimen (ZW 111) (pl. 7D) was doubtfully assigned to this

species. It is the crushed skull of a large viverrid which lacks the braincase, the

left P!, P*, M! and M? and the right P?. No comparative material was available

and the identification is based on published information (Ewer, 19565; Roberts,

1951). The teeth of the specimen compare fairly closely in size with those of

the extant H. ichneumon (table 17).

TABLE 17

DIMENSIONS OF THE TEETH OF A ? Herpestes ichneumon FROM SWARTKLIP

COMPARED WITH THOSE OF THE MODERN FORM

ZW 111 Herpestes ichneumon*

C—M? length .. c41-4 38-0 39-1 38-8 40-0

P* length aie cl10-3 11-0 1-9 10-7 11-3

* Roberts, 1951.

As far as can be determined this species has not been recorded in the

south-western Cape Province during historic times, but is known to occur in

the coastal region further east.

NEW QUATERNARY FOSSIL SITES NEAR SWARTKLIP, C.P. 67

Family Mustelidae

? Mellivora capensis Schreber

Honey Badger

A left maxillary fragment (ZW 142) from Site I was doubtfully assigned

to this species. It is intact up to and including the most anterior part of the

orbit. Only the P? remains, but the alveolus of the C and the anterior part of

the alveolus of the P® are preserved.

It corresponds closely in size and general morphology with modern

Mellivora capensis skulls, and also with fossil specimens from Swartklip (ZW 1,

Singer/Fuller Occurrence) and Elandsfontein. However, in ZW 142 the P? is

situated posterior to the C and is set parallel to the margin of the maxilla,

whereas in all the comparative specimens it is at an angle to this margin and

overlaps with the posterior margin of the C on its lingual surface.

The alveoli of the G and P? were comparable in size to those of the com-

parative specimens, although the P? was large compared with the mean of the

comparative series (table 18).

TABLE 18

DIMENSIONS OF THE P? OF THE ? Mellivora capensis FROM SWARTKLIP COMPARED

WITH THOSE OF A SERIES OF MODERN AND FOSSIL SPECIMENS

Mellivora capensis

i Qe

poe

Hopefield specimens Modern specimens

ZW 142 ZW 1 E.C.14 20916A 8640D | 3325 19940 3326

efi GES Wei ioee SE leiesesey (A524 Bes 1696. 1558

b 4-4 3-8 3°95 4-0 3-4 3-8 4-3 3-4

CARNIVORA: Incertae sedis

The following specimens were not classified:

KW 178—An incomplete left mandible with an incisor, canine and premolar

of a small carnivore of uncertain affinities.

KW 145—An incomplete left mandible of an immature individual, with an

unerupted canine and premolar, probably belonging to a small felid.

AW 109—The anterior part of a right mandible with an incomplete deciduous

dentition, belonging to a small carnivore.

<£W 136—A canine (?) of a carnivore of uncertain affinities, with an extremely

long root relative to the crown height.

AW 1315—An incomplete canine, possibly belonging to a lion.

Family Bathyergidae

Bathyergus suillus Schreber

Cape Dune Mole

The remains of at least two individuals were present in the Site I

assemblage.

68 ANNALS OF THE SOUTH AFRICAN MUSEUM

ZW 123—An incomplete right mandible with the incisor, both premolars and

the first of the molars.

ZW 398—An incomplete left mandible with only the first of the premolars

remaining.

KW 1112—A fragment of left mandible.

KW 190—A fragment of a left upper incisor.

These specimens were indistinguishable from the corresponding parts of

Bathyergus sullus, which occurs commonly in the area today.

RODENTIA: Jncertae sedis

Three incisors (ZW 169, ZW 1205 and ZW 1206) representing three

distinct types of rodent other than Bathyergus were recovered at Site I.

Family Struthionidae

Struthio australis

Ostrich

In addition to the egg and eggshell fragments already mentioned, the

following bones of the ostrich were recovered from Site I:

ZW 208—The shaft of a tibia.

KW 714—The shaft of a femur. |

KW 948—An incomplete right metatarsal of an immature individual.

KW 1124 A, B, C—Fragments of the distal end of a femur.

Morphologically the specimens were indistinguishable from the corre-

sponding parts of the extant ostrich, but the tibia shaft was somewhat longer

than those of the comparative specimens.

CHELONIA: Inceriae sedis

A single fragment of carapace (ZW 1208) was recovered at Site I.

THE PosTCcRANIAL REMAINS

Approximately 86% of all the identified postcranial bones recovered came

from Site I (table 2), and a summary of the analysis of this material (excluding

non-mammalian remains) is given in table 19.

One of the most striking features emerging from this analysis was the

scarcity of postcranial remains relative to the number of individuals repre-

sented. In the case of the lion, for example, the disparity was very marked.

The appendicular skeleton of the lion is made up of approximately 110 bones,

excluding the innominates, patellae and sesamoids. With four individuals

represented at Site I, theoretically at least 440 bones should have been

NEW QUATERNARY FOSSIL SITES NEAR SWARTKLIP, C.P.

O€Es | OF9T

6961 | O€9T

0946

bile

| Ee

‘eo 5.

la Qu

> 5

ee ct.

nN e)

(a)

sqra

dVIQI1I9 A,

GREGG ele GGm | 699M) wan seg) 1 Le

6 0€ | 9F | 99 6 |) Ol

G le teilG La)

Be ie Ne Ih [Gl Sra at

I [mete C

I I I

v ¢ L 0c | ¢

Cs a a

el sl2lFlerls|els

el Boe 22 hecelee 9) oslo

Bee 20 al ses eee! come

oe aie Fe, 2 | [= es

eleas Celica Nag

P| & s | 2 | 6

g | Et

_—_— | Ss |S —— | | | | TEE

—_— [OO OS | | | TTT *"

** CIOATUICS TIYIOC

‘souoq 930[duI09 9}3¥01pur AjIZessa09U JOU Op sJaquINNy - a0A/

‘* STV.LOL

IUPTIOIIOOUIYY

sepinby

* 08] $4)44q—Seplay

GaNINAALAGNN

VILNAGOa

VIALOVGOILAV

VIALOVGOSSIAdd

VUAOAINAVD

| | | | | | | | |_| f "_

aeyjoyed 29 sprouresag

sosureud

sjerpodrjour

porenussopipuy)

I ALIS WOUd SNIVNAA IVINVYOLSOd

6] FTav,L

NVITIVNAVAW FHL

oeindyvog

SdNOUD TVOIODOTOOZ

70 ANNALS OF THE SOUTH AFRICAN MUSEUM

recovered. There were in fact only parts of three bones recovered—less than

1% of the theoretical total. Other figures calculated on this basis ranged up to

20% in the case of the appendicular skeleton of antelope.

The presence of incomplete skeletons at fossil sites is quite common, and

many factors might be responsible. Recent work by Brain (1967) is of special

interest in the present instance. He has demonstrated by careful field obser-

vation that disproportionate occurrences of bones and parts of bones can

result from the destructive chewing of small carnivores—domestic dogs in the

study undertaken by him. Since carnivores are thought to have been responsible

for the fossil assemblages at Swartklip, they were probably also responsible

for the destruction of much of the original assemblage. The main objection to

this theory is that none of the bones showed signs of tooth marks. The possibility

that subsequent weathering has removed all traces of these is considered

unlikely, in view of the excellent state of preservation of the specimens.

Clearly allowance must be made for the method of collection, which

resulted in only part of the entire deposit being handled. However, this factor

alone cannot adequately account for the persistent shortage of postcranial

remains of all categories.

In addition there was almost certainly some selective collecting of bone

by the inhabitants of the lairs.

It is probably a combination of these factors which has given the assem-

blage its present character.

A second feature of the postcranial assemblage was the ‘completeness’

of the bones. A frequently characteristic feature of bone accumulations resulting

from the activities of man is the highly fragmented state of the bone (cf. Dart,

1957, for details and references). There are two basic reasons for man’s deliber-

ate fragmentation of bone—firstly, to obtain pieces suitable for the manu-

facture of tools, and secondly, to remove all edible soft tissue. Although the

bones from the Swartklip assemblages were often incomplete, this was the

result mainly of post-fossilization damage, and as a whole the assemblages

lacked large numbers of small bone fragments and splinters. In this respect at

least the assemblages had the appearance of animal lair residues, rather than

human occupation site debris (Hendey & Singer, 1965: 212).

AGE OF THE ASSEMBLAGES

The fauna of the sites is essentially ‘modern’ in character. No extinct

genera are present, and most of the forms are indistinguishable from, or closely

related to extant species. The fauna clearly post-dates that of the Elands-

fontein site, which is usually termed ‘late Middle/early Upper Pleistocene’

(Boné & Singer, 1965), and corresponds most closely with that of the Melkbos

fossil site (Hendey, in press).

Singer and Fuller (1962) suggested that the assemblage described by them

was late Upper Pleistocene, and the present study tends to confirm this,

although the possibility that the fauna is post-Pleistocene cannot be discounted.

NEW QUATERNARY FOSSIL SITES NEAR SWARTKLIP, C.P. Fics

The application of the name ‘Wolfgat’ to the area in which the sites are

located, suggests the possibility that the lairs in which the fossils accumulated

were still being occupied after the arrival of the first settlers from Europe in

1652. It is uncertain when this name was first used but it is most likely to have

been during the late seventeenth or early eighteenth centuries. However, even

by 1649, before the first permanent European settlement, the indigenous fauna

of the Cape Peninsula area was much depleted (Sargent, 1954), and it is con-

cluded from the variety of faunal types represented, that at least part of the

assemblages accumulated in prehistoric times.

All the available evidence points to the Swartklip fossils being late Upper

Pleistocene or Recent in age.

CONCLUSION

The importance of the Swartklip sites lies chiefly in the fact that they have

provided the largest range of later Quaternary fossils known from the south-

western Cape Province to date. Information on the Quaternary fauna of this

region is now coming from the Early or Middle Pleistocene deposits at Lange-

baanweg, the Middle/Upper Pleistocene site at Elandsfontein and the more

recent Melkbos and Swartklip sites. Consequently there is now a proven

potential in this limited geographical region for establishing a succession for

the Quaternary fauna of southern Africa, a situation unparalleled elsewhere

in the subcontinent.

Furthermore the fossils at Swartklip come from sealed deposits where the

danger of admixture of earlier and later elements is excluded. This is not the

case with the other major sites of the region. The excellent state of preser-

vation of the specimens makes detailed morphological studies of them possible,

and there is no doubt that the future exploitation of the sites, and a closer

examination of the material already recovered, will provide a great deal of

unique and valuable information.

They are potentially the most important fossil occurrences of their kind

discovered in southern Africa in recent years.

SUMMARY

The fauna of three recently discovered fossil sites in the south-western

Cape, South Africa, is described, accounts being given of nineteen mammalian

types, including one new subspecies, Antidorcas marsupialis australis.

_ Brief accounts of the geological and archaeological associations of the

fauna are given, and the suggestion made that the fossils accumulated in

carnivore lairs.

It is concluded that the assemblages are late Upper Pleistocene or Recent

in age.

ACKNOWLEDGEMENTS

We are indebted to Dr. E. Middlemost of the University of Cape Town,

who first drew our attention to the sites, and to Dr. Middlemost, Mr. R. R.

72 ANNALS OF THE SOUTH AFRICAN MUSEUM

Inskeep and students of the Department of Geology, University of Cape

Town, who were responsible for much of the original collecting of material.

Members of the Western Cape Local Centre of the South African Archaeo-

logical Society and the University of Cape Town Archaeology Field Club

undertook most of the later collecting, and we are especially grateful to Miss E.

Speed, who arranged these collecting trips.

Prof. L. H. Wells of the University of Cape Town kindly offered opinions

and suggestions on some of the material. Mr. C. E. Gow of the South African

Museum took the photographs of the specimens, and our thanks go also to him.

Much of the preparation of specimens was competently undertaken by Mr.

M. Jacobs.

Finally we wish to thank Dr. Ronald Singer and Dr. E. C. Olson of the

University of Chicago, who kindly checked the manuscript and made many

useful suggestions.

This study was supported in part by the U.S. Public Health Service grant

no. GM 10113 to Dr. Singer, and the National Science Foundation grant

no. B 2453 to Dr. Olsen.

REFERENCES

Bonet, E. L. & SincER, R. 1965. Hipparion from Langebaanweg, Cape Province and a revision

of the genus in Africa. Ann. S. Afr. Mus. 48: 273-397.

Brain, C. K. 1967. Hottentot food remains and their bearing on the interpretation of fossil

bone assemblages. Scient. Pap. Namib Desert Res. Sin 32: 1-11.

Broom, R. 1949. Notes on the milk dentition of the lion, leopard and cheetah. Ann. Transv.

Mus. 21: 183-185.

Cooxg, H. B. S. 1947. Some fossil hippotragine antelopes from South Africa. S. Afr. F. Sci. 43:

226-231.

Cooke, H. B. S. 1955. Some fossil mammals in the South African Museum collections. Ann. S.

Afr. Mus. 42: 161-168.

Dart, R. A. 1949. The predatory implemental technique of Australopithecus. Am. 7. phys. Anthrop.

(n.s.) 7: 1-38.

Dart, R. A. 1956. The myth of the bone-accumulating hyaena. Am. Anthrop. 58: 40-62.

Dart, R. A. 1957. The osteodontokeratic culture of Australopithecus prometheus. Mem. Transv.

Mus. 10: i—vili, 1-105.

ELLERMAN, J. R., Morrison-Scott, T. C. S. & Hayman, R. W. 1953. Southern African mammals,

1758-1951: a reclassification. London: British Museum (Natural History).

Ewer, R. F. 1956a. The fossil carnivores of the Transvaal caves: Canidae. Proc. zool. Soc. Lond.,

126: 97-119.

Ewer, R. F. 1956). The fossil carnivores of the Transvaal caves: two new viverrids, together

with some general considerations. Proc. zool. Soc. Lond. 126: 259-274.

Gentry, A. W. 1964. Skull characters of African gazelles. Ann. Mag. nat. Hist. (13) 7: 353-382.

Gentry, A. W. 1966. Fossil Antilopini of East Africa. Bull. Br. Mus. nat. Hist. Geol. 12: 45-106.

(Fossil Mammals of Africa, No. 20.)

HEnDEY, Q. B. The Melkbos site: an Upper Pleistocene fossil occurrence in the south-western

Cape Province. (In press: Ann. S. Afr. Mus.)

NEW QUATERNARY FOSSIL SITES NEAR SWARTKLIP, C.P. 73

HEnpDEY, Q. B. & SincErR, R. 1965. The faunal assemblages from the Gamtoos Valley shelters.

S. Afr. archaeol. Bull. 20: 206-213.

HERsHKOVITZ, P. 1959. Status of names credited to Oken, 1816. 7. Mammal. 30: 289-301.

Hooyer, D. A. & SincER, R. 1960. Fossil rhinoceroses from Hopefield, South Africa. Zool.

Meded., Leiden 37: 113-128.

MaBeERLy, C. T. A. 1963. The game animals of southern Africa. Johannesburg: Nelson.

Maccs, T. M. O’C. 1966. An unusual implement and its possible use. S. Afr. archaeol. Bull.

2I: 52-53.

Miter, W. T. 1954. Wild life of southern Africa. 2nd ed. Pietermaritzburg: Shuter & Shooter.

Roserts, A. 1951. The mammals of South Africa. Johannesburg: Central News Agency.

SARGENT, J. U. 1954. Place names and the fauna of the Cape before 1800 A.D. Rep. Dep. Nat.

Conserv. C.G.H. 11: 38-45.

SINGER, R. & FuLLER, A. O. 1962. The geology and description of a fossiliferous deposit near

Zwartklip in False Bay. Trans. R. Soc. S. Afr. 34: 205-211.

Tuomas, O. 1926. On mammals from Ovamboland and the Cunene River. Proc. zool. Soc. Lond.

1926: 285-314.

Warp, R. 1899. Records of big game. London: Rowland Ward.

aay ,

nt t

yo ‘ 4 Pee auwi

t Ag Ay Sia

’ “ar JO (ae ah vi

. Te Se ay

ital

}

> Ane

y i)

uv,

¥ emt

i j

« »

(

tas 4 M

ie i

Ff Wea

are ed

« at i

‘ 7

id ¢ A

¢ yi) i i ih

heal 1

1 ay

Diy

ini ;

1 f

‘

. }

j ba

é,

eek 8 } tut

3 ch ri

ne Sait

a i

SPHIONI

Case sel on

Bog seo co

=x eA Oo -Q !

o ¢ rs

3.55 02 oe

(ian N= B

6 S¢ ral 534 n

N s) Og : (e)

7a ase SES fs)

—

> St oc eee

f Gc Ciccone enor

Z sre os HH

ah ar go, F

=) ra m™ ¢ 4 Dm

= % oO A. TO Ss

= S Solent 64> <a orn =

aes jo) 6) e=ihre, 64 N

fal ne oO ”

a OW VUDGy oF

4 oy Sd oO ane)

S Sat Ss Res

¢ : <5 joa) O

PII

: . 52.

Ann. S. Afr. Mus., Vol

pied

of the Redunca

and frontal views

Lateral

skull ZW 376.

& B.

Dorsal view of the Redunca skull ZW 1316.

Dorsal view of the Redunca skull ZW

74:

Plate ILI

1€

1 views of tl

1pita

rontal and occ

ventral

Antidorcas skull ZW 66.

Lateral

K i

. igh ,

‘

i aa ae

Tah

EY’

Pla n. S- Afr. Mus., Seal?

Plate II

\&B. Lateral and frontal views of the Redunca

skull ZW 376.

Dorsal view of the Redunca skull ZW 1316.

Dorsal view of the Redunca skull ZW 74.

mtal and Occipital views of the

Ann. S. Afr. Mus., Vol. 52.

SFHINI

Frontal view of the Antidorcas skull ZW 402.

Medial view of the Antidorcas horn core ZW 67a.

Buccal view of the Antidorcas mandibles ZW 81 and ZW 82

BN.

Plate V

a 526

non. S. Afr. Mus., Vol

‘206 AAZ, Ul rejoutoad SNO[BUIOUL 9]0\)

‘006 MZ pue LoS MZ ‘GLI MZ Se[qtIpueuL sajapys0UU0’) BY JO MATA [eoong|

‘OLE MZ, 2IGIpueu snianiydoy IU) JO MATA [eoong

‘CCS MATZ aIqtipueu snépuyoddiyy 2u) JO MOIA Jesnyjo00 pure yeoone

MZ ue IA | [ | | d

INCHES

Py, Afr. Mus., Vol. 52.

“Ban. S- 4

Plate III

\& B. Lateral and

skull ZW 376.

frontal views of the Redunca

Dorsal view of the Redunca skull ZW 1316.

Dorsal view of the Redunca skull ZW 74.

ral, rontal and occipital views of the

ZW 66.

Ann. S. Afr. Mus., Vol. 52. Plate |

MMM Uy

A & B. Dorsal and ventral views

of the Hyena skull ZW

I3IT.

C. Buccal view of the Hyena

mandible ZW 1312.

SIH ON/

Plate Vil

mS. Afr. Mus., Vol. 52.

‘III MZ [4s uoulnauya saysaqiayy i jo MoIA [elo] eT ‘qd

‘OTL MZ [MAS [Va4as SUA i Jo MOIA [SLOG] vy

‘QO! MZ &l[txeu | SUA JO MOIA [BOON “gq

‘OO! MZ [OAs 09] SUA IY} Jo MOTIA [P19 eT Nf

ms ganna as ens

ved)

SERRE &

;

}

pote

seta):

aN, ae i Me ; ; Te es NA, fi oy,

i ce i. &% A Ah

a ad, Snes \ Ci As ae a

: i Lia ad r

y &

” airs 1s A

on z

i Aue: 7, :

|) ‘a i - d i Wie 4 te > ie

‘ NO gee) ak ieee TD rks Bit a, ye >

a A v af Ne

fe *

af r

3 F -

~~ f

+ >

“i a 184

So de

Lut 4 if

a ‘

pe r

‘ tw

* Cape

Pete a

¥ a

ra -

’ * ‘

sf ‘> egret

4 as a

Plat Ann. 5- Afr. Mus., Vol. 52.

“a,

Plate VII

Ann. S. Afr. Mus., Vol. 52-

a a 7 ‘»

= * +

ufc.¥

=v u

\& B. Dorsal and ventral views

of the Hye skull y\ a

ryt :

ie oe

qa 5x

\ Ry \ wo H i

, \\

éadq

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. (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)

BuLioucuH, 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 Schulize, 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

)

ave ‘ i i Wi 4 i ri ia

el EE RY ALL RS tena

a ar i it Pat li OWT

‘ndvlaicalths, lini mt } he Hy,

uh FAM DENY Hi) 1h one Ww

i) ah as ne A a

| | hn teat hee ‘ne basinal sft

nin oa vd i Lian : a HoH ue iat ‘wt Bo

(i ‘id oe

)

é * 4 vi] Ate) \

Ne aah WATER ABS Rea te eda ie 2 Pa | Eth d

t ‘ ae fy t

/ a Pos § wih yg ‘ake De vat ei in ie

iH } + iw) ee We | Rulactes i, te

F Wy

07.6%

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 52 7° °&Band

October 1968 Oktober

Bare gi Deel

TWO NEW SPECIES OF ACHAEUS

(CRUSTACEA, DECAPODA, MAJIDAE) FROM

SOUTH AFRICA

D.4J.,G.. GRIFFIN

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

(Cash with order, post free)

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

(Kontant met bestelling, posvry)

OUT OF PRINT/UIT DRUK

I, 2(1, 3, 5, 7-8), 3(1-2, 5, tp), 5(2, 5, 7-9),

6(1, t.—p.i.), AGS Bs 8, g(1-2), 10(1-3),

T1(1-25 77,5) t:-P:1.).\) 20242) 27aens ts) ena

44(4).

Price of this part/Prys van hierdie deel

R1.50

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

1968

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

TWO NEW SPECIES OF ACHAEUS (CRUSTACEA, DECAPODA,

MAJIDAE) FROM SOUTH AFRICA

D. J. G. GRIFFIN

Australian Museum, Sydney, Australia

(With 4 text-figures)

[MS. received 15 January 1968]

CONTENTS

PAGE

Introduction O57 0. {63° 75

Systematic account .. 76

Discussione NM) 30 0655-7 5086

Summary ©5202)" 86

Acknowledgements .. 87

Riclerenees, Movi as.n | 87

INTRODUCTION

The small, long-legged spider crabs of the genus Achaeus, belonging to the

majid subfamily Inachinae, have most recently been reviewed by Griffin &

Yaldwyn (1964), who considered that about 20 species were known. A more

detailed count (Griffin, unpublished) shows that 27 species of good status are at

present placed in the genus; two of these have recently been described from the

south Atlantic by Forest & Guinot (1966). Three others are of uncertain status.

Two of these are from South Africa and were discussed by Barnard (1950):

“Achaeus cf. lorina Adams and White’ and ‘Achaeus cf. affinis Miers’. Re-examina-

tion of the material studied by Barnard shows that both these are distinct,

unnamed species of Achaeus. The first is similar in some ways to a few species of

the genus Macropodia. Examination of the holotype of A. lorina, of a specimen of

that species from the Philippines and of one specimen from the series recorded by

Rathbun (1911) from Amirante in the Indian Ocean permits a further diagnosis

of A. lorina and reveals that Rathbun’s material is probably referable to A. /issi-

Jrons (Haswell). Barnard’s ‘A. cf. affinis’ is similar to A. brevirostris (Haswell).

In the present paper the new species are described and figured and com-

pared with other species of Achaeus. The terminology and system of measure-

ments used follow Griffin & Yaldwyn (1964). In general, carapace length is

abbreviated as c.l.

Ann. S. Afr. Mus. 52 (3), 1968: 75-87.

76 ANNALS OF THE SOUTH AFRICAN MUSEUM

SYSTEMATIC ACCOUNT.

Achaeus spinosissimus u.sp.

(Figs 1, 2, 4a, b)

Achaeus cf. lorina: Barnard, 1950: 22-23, fig. 3g.

[non] Inachus lorina Adams & White, 1848.

Holotype: Male, c.l. 10-0 mm, off Hood Point (East London), South Africa

(S. Afr. Mus. reg. A8309) —South African Museum, Cape Town.

Description: Carapace elongate subtriangular (length 1-3 times width),

narrowed anteriorly, not markedly constricted behind orbits, branchial regions

swollen, lateral margins and dorsal surface armed with numerous spines and

spinules and a few scattered short hairs; regions moderately well defined. Sur-

face of carapace, sternum, abdomen and third maxillipeds coarsely granular.

Rostrum of two very short, slender, blunt spines separated by a narrow slit.

Supraorbital eave bearing midway along a long, slender, weakly curved,

acuminate spine directed outwards, upwards and forwards; eave otherwise

unarmed. Postorbital region with about four short spines laterally, one longer

than others, and two or three dorsal spinules near lateral margin. Eyestalks

stout, a narrow process extending above cornea terminating in a small tubercle;

a small pointed spinule on anterior surface close to cornea; cornea large, ovoid,

obliquely terminal.

Hepatic region not greatly expanded, with four or five spines on margin,

one longer spine on dorsal surface posteriorly and some spinules anteriorly.

Dorsal surface of carapace with two prominent short, blunt tubercles in

mid-line, one spinous gastric and a blunt one on tumid cardiac region; a low

swelling behind cardiac tubercle. Protogastric regions each with a curved spine

just in front of hepatic regions with one or two spinules at its base. Branchial

regions with six spines on, or close to, margin; eight or nine spines and spinules

on outer slope of each mesobranchial region and a very low swelling on meta-

branchial region above last legs.

Basal antennal article very slender, armed with four equidistant, broad-

based, long, subacute spines mostly directed outwards situated towards lateral

border. First segment of flagellum short, with an outwardly-directed terminal

spine. Second segment long, with a small spinule laterally. Flagellum of mode-

rate length, with a few long hairs.

Antennular fossae large, longitudinally subovate; basal segment of anten-

nules bearing medially a row of seven prominent spinules; inter-antennular

spine slender, projecting downwards as an obtuse, triangular process; anterior

process of epistome extremely slender.

Epistome about as wide as long with several spinules laterally, two behind

antennal article and others near anterolateral angle of mouthfield. Pterygosto-

mian regions with a spine midway along lateral border. Several spinules ven-

trally on branchial regions.

Third maxillipeds almost meeting in midline, spinous and hairy. Ischium

TWO NEW SPECIES OF ACHAEUS

Fig. 1. Achaeus spinosissimus n.sp. Holotype, male, carapace, dorsal aspect.

78 ANNALS OF THE SOUTH AFRICAN MUSEUM

bearing spines in two oblique rows, one row of two or three laterally and a

central row of seven similar spines; medial edge finely toothed, overlaid by long

hairs. Merus narrow, subovate, with four spines near medial edge, three in a

central row and two smaller ones at anterolateral angle; several long hairs

arising from medial edge. Palp long and stout, dactyl as long as carpus and pro-

podus together, carpus and propodus each with a spine near distal border on

medial surface; long hairs arising from medial surface and tips of all three

segments.

Chelipeds long and stout, with numerous spines and spinules more or less

in longitudinal rows, and long sparse hairs. Ischium with several spines. Merus

subcylindrical, swollen, bearing numerous spinules and curved spines largest

along outer lateral surface where five, larger than others, stand in a row. Carpus

subcylindrical, with numerous curved spines which are generally longest

laterally. Chela about half total length of cheliped, compressed, fingers not much

shorter than palm which is deep, dorsally weakly convex and ventrally strongly

convex; dorsal surface and dorsal part of outer surface with short, curved spines

and blunt tubercles in poorly defined rows in proximal two-thirds; a few, similar

but smaller, spinules and tubercles on ventral part of outer surface and along

ventral surface; distal part of outer surface smooth. Fingers stout, very widely

gaping for about proximal two-thirds; fixed finger with a very large, apically

truncate tooth proximally filling gape, its straight apex spinulate, remainder of

inner edge denticulate, strongly concave proximally, obtusely angled at distal

end of gape. Dactyl with large truncate tooth, smaller than, and just beyond,

that on fixed finger, remainder of inner edge irregularly dentate. Chela with

long hairs dorsally, ventrally, on outer surface of palm distally, along both

fingers and filling gape.

Ambulatory legs very long, slender, with curled hairs arising singly on

dorsal surface of propodi and long hairs, especially on distal two-thirds of

propodi and on dactyls; first leg the longest, remainder decreasing to last;

bases of all legs and ischia of first with some spinules ventrally, meri of all with

a terminal dorsal spine; dactyls almost straight in first pair and unarmed,

second with a single, strong, subterminal, curved spinule ventrally, dactyls

of third and fourth legs short, weakly curved with several small denticles and

two curved, subterminal spines ventrally, the distal the longer.

_Sternum with spinules in transverse rows opposite base of each leg and two

on each side along margin of abdominal fossa anteriorly at base of chelipeds.

Abdomen in male of six segments, segments 6 and 7 coalesced. All segments

wider than long, first segment the longest, second very short, remainder sub-

equal in length, third and last a little longer than others. Abdomen widest

about middle of laterally convex third segment, lateral margin concave to just

beyond base of last segment, then tapering rapidly, apically rounded but with

surface deeply concave and appearing bilobate in ventral view. Surface with a

broad medial elevation distally in segments 1-5 and proximally in last, each

bearing a pair of long hairs. Third segment with two oblique rows of three spines

TWO NEW SPECIES OF ACHAEUS 79

on each side of midline on swollen lateral surfaces. Proximal part of last segment

laterally swollen.

Male first pleopod moderately stout, uniformly tapering apically, curved

outwards, terminally blunt; aperture terminal, a narrow slit at end of groove

along medial surface; lacking hairs except for several at base laterally.

Measurements: Carapace length 10-0 mm, carapace width 7-1 mm, rostral

length 0-5 mm, rostral width 0-5 mm, cheliped length 17-5 mm, chelar length

8-4 mm, chelar height 3-6 mm, dactyl length 5:1 mm, first ambulatory leg

length 26-2 mm.

Remarks: This new species agrees with Adams & White’s original description

and figures only in a number of general features, including shape of the cara-

pace, presence of two tubercles in the midline, presence of a larger tooth near

the base of each finger of the chela, setose anterior ambulatory legs and curved

dactyls on the posterior ambulatories.

Examination of drawings of the holotype (a male, c.l. 11 mm (approx.),

‘Eastern Seas’, in the British Museum (Natural History)) of Inachus lorina by

Dr. A. L. Rice and the availability of a specimen from the United States

National Museum collections—1 ovigerous female, c.l. 10°5 mm (reg. no.

49837), Albatross Philippine Expedition 1907-1910, Sta. 5355, North Balabac

Strait, 44 fms—which is certainly conspecific with the holotype, show that the

diagnostic features of A. lorina include the weakly bilobed, apically rounded

rostrum, acute but not markedly expanded hepatic regions, single gastric eleva-

tion, two small tubercles side by side on elevated cardiac region, a small spine

or tubercle close to posterior margin above last ambulatory leg, supraorbital

eave without spines or tubercles, eyestalks with a sharp tubercle, basal antennal

article with a spine centrally and an apical spine. The Philippine specimen

possesses a small spinule at the base of the basal antennal article and there is an

oblique row of three low tubercles on the branchial regions medially extending

forward from opposite the cardiac prominence; neither of these features are

apparently obvious in the holotype. The third maxillipeds are of the usual form

in this genus and are spinous and hairy; there is one longitudinal row of spines

on the ischium, the surface of which is weakly excavate immediately lateral to

the spines; there are two spines not far from the lateral edge, one proximal, one

distal; the merus is also weakly excavate centrally and lateral to this there are

three sharp spines; the medial and anterolateral edges bear sharp spines and the

carpus and propodus each bear a single, slender spine. The slender chelipeds

are spinous and hairy, long spines and hairs arising from the ventrolateral edge

of the merus, the dorsomedial surface of the carpus and the ventral edge of the

palm of the chela; there are long hairs and shorter spines along the dorsal sur-

face of the merus and palm of the chela; fringes of long hairs continue on to the

dorsal edge of the dactyl and the ventral edge of the fixed finger. The ambula-

tory legs are extremely long and slender and the propodus and dacty]l of all

bear very long hairs; the dactyl of the last ambulatory leg is weakly falcate with

a double row of short spines on the proximal two-thirds and three longer back-

80 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 2. Achaeus spinosissimus n.sp. Holotype, male: a, right cheliped; b, left third maxilliped;

c, left first ambulatory dactyl; d, left fourth ambulatory dactyl; e, front of carapace, ventral

aspect; f, abdomen.

TWO NEW SPECIES OF ACHAEUS 81

wardly curved spines, equally spaced, distally, the last distant from the tip;

there is a short spine just beyond the penultimate large spine.

Thus, A. spinosissimus differs from Achaeus lorina notably in the presence of a

supraorbital spine, larger spines on the dorsal surface of the carapace, particu-

larly the protogastric ones, more numerous spines on the hepatic and branchial

margins, in the form of the rostrum and in the presence of much larger spines

on the basal antennal article. As Barnard (1950) states, this species agrees

reasonably well with the description of Achaeus lorina given by De Man (1902)

of specimens from Ternate and Halmahera in Indonesia. However, none of the

most notable features of this species are present in the holotype of Achaeus lorina.

The material reported on by De Man may be conspecific with A. spinosissimus.

Examination of one specimen —an ovigerous female, c.l. 5-g mm (reg. no.

1912:2:10:82), Percy Sladen Trust—Sealark Expedition to the Indian Ocean

Sta. E14, Amirante, 34 fms—in the collections of the British Museum (Natural

History) from the series recorded by Rathbun (1911:244) as Achaeus lorina shows

that this identification by Rathbun was incorrect. The specimen is definitely not

conspecific with A. lorina but possesses the general features of the carapace, orbit,

maxillipeds and cheliped of A. fissefrons (Haswell) (see Griffin & Yaldwyn, 1964:

38-41, figs. 1-8) except that the eyestalks are much longer, the postorbital

spinules are minute and the palm of the chela possesses fewer spinules on the

dorsal and ventral surfaces than does A. fissifrons. The fact that the specimen is

a female without ambulatories makes precise determination difficult. Positive

identification of Rathbun’s series thus awaits further study.

Among the known species of Achaeus, A. spinosissimus appears to be most

closely related to those such as A. inimicus Rathbun, A. akanensis Sakai, A.

anauchen Buitendijk, A. fissifrons (Haswell) and A. cadellt Alcock, in which the

supraorbital eave possesses 1-3 large spines. From these it is distinguished by its

more spiny carapace and chelae, the form of the rostrum and the presence of

long spines on the basal antennal article. The closely approximated rostral

spines are reminiscent of species of Macropodia in which, however, the rostrum is

nearly always much longer. The species of Macropodia most similar to A. spino-

sissimus is M. formosa Rathbun (see Barnard, 1950: 17, figs 2¢-2).

Achaeus barnard1 n.sp.

(Figs 3, 4¢-g)

Achaeus cf. affinis: Barnard, 1950: 19-20, figs 3d—.

[non] Achaeus affinis Miers, 1884 (= A. brevirostris (Haswell)—see Griffin & Yaldwyn, 1964:

46-48).

Holotype: Male, c.l. 8-5 mm, chelipeds and legs missing, off Cape Morgan,

South Africa (S. Afr. Mus. reg. A1392) —South African Museum, Cape Town

(this is the specimen figured previously by Barnard).

Paratypes : wo males, c.l. 6-7, 5:5 mm, 1 female (ovig.) c.l. 7-0 mm, same data

as for holotype—South African Museum, Cape Town.

82 ANNALS OF THE SOUTH AFRICAN MUSEUM

Description: Carapace elongate subtriangular,, narrowed anteriorly, not

markedly constricted behind orbits, branchial regions swollen, lateral margins

and dorsal surface with a few tubercles, regions well defined. Surface of cara-

pace, sternum, abdomen and third maxillipeds coarsely granular.

Rostrum of two short, slender, subacute lobes separated apically by a very

narrow, V-shaped slit.

Supraorbital eave with up to six or seven very small, sharp spinules

anteriorly on dorsal surface near margin. Postorbital region unarmed. Eye-

stalks stout, a narrow process extending above cornea terminating in a small but

prominent tubercle; anterior surface with a prominent rounded or subacute

lobe midway along; cornea large, ovoid, obliquely terminal.

Hepatic region not greatly expanded, with a small tubercle at summit aid

one or two tubercles or spinules in front of this.

Branchial regions smooth or with three very low tubercles laterally in a

shallow arc, one anteriorly, one just forward of cardiac prominence and one

close to posterior margin on metabranchial region above last leg; anterolateral

margins with two or three small tubercles, posterior margins laterally with some

minute spinules.

Dorsal surface of carapace with four low tubercles in midline, one far back

on mesogastric region, two small tubercles side by side surmounting tumid

cardiac region and a low tubercle on posterior slope.

Protogastric regions smooth or with a small, low tubercle on each side just

in front of marginal hepatic tubercle.

Basal antennal article with surface weakly convex, oblique, smooth or with

up to four very small tubercles in a row centrally. First segment of flagellum

stout, short, a small, apically curved spine laterally midway along and two

smaller spines ventrally, one midway along and one apically; second segment

almost three times as long. Flagellum of moderate length, with a few long hairs.

Antennular fossae large, longitudinally subovate, distal part of lateral edge

slightly outwardly splayed; basal segment of antennules with a row of spinules

medially; interantennular spine and anterior process of epistome slender, the

former projecting down as an obtuse lobe.

Epistome slightly longer than wide, with one or two spinules just lateral to

opening of green gland, otherwise smooth.

Third maxillipeds almost meeting in midline, spinulous and hairy. Ischium

bearing spinules in two oblique rows, one laterally and one centrally bordering a

broad, shallow groove; medial edge coarsely toothed. Merus narrow, subovate,

with four or five spinules centrally in proximal half, a similar number of spinules

close to medial margin and three spinules on anterolateral angle. Palp long and

stout.

Chelipeds long and stout in male, short and slender in female, tuberculate

or spinous. Ischium tuberculate. Merus subcylindrical, bearing three sharp

tubercles dorsally in both sexes, two proximal, one distal and a row of tubercles

ventrally which are blunt in male and become small distally but sharp in female

TWO NEW SPECIES OF ACHAEUS

Fig. 3. Achaeus barnardi n.sp. Holotype, male, carapace, dorsal aspect.

84 ANNALS OF THE SOUTH AFRICAN MUSEUM

and of equal length throughout. Carpus medially bearing about five sharp

spinules. Chela in male about half length of cheliped, compressed, robust, palm

dorsally weakly convex and ventrally strongly convex, outer surface granular,

dorsal and ventral surface with a few small spinules proximally, without spines

or tubercles; fingers as long as palm, stout, very widely gaping for proximal half,

fixed finger with a narrow, apically truncate tooth proximally almost filling

gape, remainder of inner edge strongly concave in proximal half, obtusely

angled at distal end of gape; dactyl with a small truncate tooth, slightly shorter

than, and just beyond, that on fixed finger, a broader, minutely dentate tooth

near distal part of gape, distal half of inner edge irregularly dentate; a few

short hairs on both fingers extending into gape. Chelae slender in female, fingers

with inner edges adjacent for entire length.

Ambulatory legs very long, slender, with curled hairs arising singly on

dorsal surface of propodi, long hairs on distal two-thirds of propodi and on

dactyls; first leg the longest, remainder decreasing to last, all legs without spines;

dactyls almost straight in first and second pair, strongly falcate to semicircular

in last two legs, third dactyl with strong spinules ventrally along whole length,

last dactyl with strong spinules for distal half to two-thirds.

Sternum with tubercles in transverse rows along each sternite; anterior

margin of sternum subtruncate, bearing spinules close together.

Abdomen in male of six segments, segments 6 and 7 coalesced. All segments

wider than long, last segment the longest, almost as long as wide, first slightly

shorter, second the shortest; fifth also short, third and fourth subequal, about

half length of last. Abdomen widest about middle of laterally convex third seg-

ment, lateral margin concave to just beyond base of last segment, then tapering

to subtruncate, weakly concave distal edge. Surface with a broad medial eleva-

tion distally in segments 1-5 and proximally in last segment. Third segment with

small spinules in three ill-defined longitudinal to oblique rows on swollen

lateral surfaces; proximal part of last segment laterally swollen.

Male first pleopod moderately stout, uniformly tapering apically and out-

wardly curved, terminally blunt; aperture subterminal, a broad slit at end of

groove on medial surface; lacking hairs.

Measurements (paratype male): carapace length 6-7 mm, carapace width 4:8

mm, rostral length 0-5 mm, rostral width (at base) 0-9 mm, chelar length

5:0 mm, chelar height 1-5 mm, dactyl length 3-0 mm, first ambulatory leg

length 21-0 mm.

Remarks: This species is very similar to A. brevirostris (Haswell) (of which A.

affinis Miers is a synonym). Comparison with the series from Australia examined

by Griffin & Yaldwyn (1964: 46-47) shows the following differences. In A.

brevirostris the carapace is narrower, the rostral lobes are somewhat broader,

usually more widely separated and blunter, the supraorbital lobe lacks spinules

or tubercles, the cardiac elevation is usually very much more prominent, the

branchial margin is without tubercles or spinules, the merus and carpus of the

cheliped in both sexes are tuberculate but without spines, the fourth ambulatory

TWO NEW SPECIES OF ACHAEUS 85

Fig. 4. Achaeus spinosissimus n.sp. Holotype, male, left first pleopod (a, b): a, abdominal aspect;

b, tip, sternal aspect. Achaeus barnardi n.sp. (c—g): c, holotype, male, left first pleopod, abdominal

aspect; d, tip of same, sternal aspect; e, abdomen of holotype; f, fourth left ambulatory dactyl,

posterior aspect, of male paratype, c.l. 6-7 mm; g, right chela, outer aspect, of male paratype.

dactyl is falcate but not semicircular and possesses spinules ventrally along the

whole length, and the sternum in the male is anteriorly weakly convex and bears

only minute spinules or lacks spinules altogether. Lastly, the male first pleopod

is almost straight, not curved distally.

Barnard’s description of this species is in error only in the following particu-

lars. He did not mention the mesogastric tubercle or the spinules on the

anterior part of the supraorbital eave and he described and illustrated the third

ambulatory dactyl in mistake for the fourth. He also stated that the male first

86 ANNALS OF THE SOUTH AFRICAN MUSEUM

pleopod resembled that of Macropodia falcifera. In that species, however, the

distal part is rather abruptly bent, not smoothly curved as in this species.

The new species is named for the late Keppel Harcourt Barnard, former

director of the South African Museum, Cape Town, who made so many notable

contributions to what is known of South African Crustacea.

Discussion

The total number of species of Achaeus in South African waters is four.

These are A. spinosissimus and A. barnardi, described in this report, A. lacertosus

Stimpson and a species which is almost certainly A. laevioculis Miers (A. cf.

laevioculis of Barnard). Re-examination of the three specimens which Barnard

identified tentatively as this species and comparison with Miers’s (1884: 520,

pl. XLVI, figs A, a) description does not suggest that the South African species

is distinct. However, the following remarks are necessary: Barnard figures one

specimen with the rostrum bearing denticles on the anterior margin —the other

two specimens have the rostrum medianly divided apically, the lobes rounded

and entire with a submarginal fringe of hairs; the hepatic margin bears a few low

tubercles; the anterior tubercle near the medial margin of the branchial regions

has one or two small tubercles near by; the lateral margin of the branchial

regions possesses spinules anteriorly; there is a very low tubercle near the

posterior margin above the base of the last legs; the meri of the chelipeds have

one dorsal tubercle and several tubercles ventrally, especially along the ventro-

lateral edge; the carpi of the chelipeds possess a few tubercles dorsally; and the

palm of the chela has tubercles on the dorsal and ventral edges distally.

Of these four species the two described in this report are restricted to South

African waters, A. laevioculis is known from the western Indian Ocean and A.

lacertosus is widespread in the Indo-Pacific, from Australia and Japan through

the Gulf of Siam and the Gulf of Martaban to India and the Iranian Gulf.

SUMMARY

A re-evaluation of the material from South African waters, of two species

of majid spider crab belonging to the genus Achaeus discussed recently by Barnard

shows that they are distinct and previously unnamed species. The two species

are described and figured and compared with closely related species. Additional

descriptive notes are given on A. lorina (Adams & White) which is known with

certainty only from Indonesia and the Philippines and the material recorded

from the western Indian Ocean by Rathbun as A. lorina is considered to belong

to the widespread A. fissifrons (Haswell). The new species appear to be confined

to South Africa; one other species, A. laevioculis Miers, is found in other parts of

the western Indian Ocean whilst A. lacertosus Stimpson is found throughout most

of the Indo-West Pacific.

TWO NEW SPECIES OF ACHAEUS 87

ACKNOWLEDGEMENTS

I wish to thank the Director and Mr. B. Kensley of the South African

Museum, Cape Town, for making this material available. I am grateful also to

Dr. A. L. Rice, British Museum (Natural History), and Dr. R. B. Manning,

United States National Museum, for information about material in their

collections and for the loan of comparative material and Dr. J. C. Yaldwyn,

Australian Museum, for helpful discussion and for commenting on the

manuscript.

REFERENCES

Apams, A. & Wuire, A. 1848. Crustacea. In Apams A. The zoology of the voyage of H.M.S. Sama-

rang ; under the command of Captain Sir Edward Belcher, during the years 1843-6. London: Reeve &

Benham.

BARNARD, K. H. 1950. Descriptive catalogue of South African decapod Crustacea (crabs and

shrimps). Ann. S. Afr. Mus. 38: 1-837.

Forest, J. & Gurnot, D. 1966. Résultats scientifiques des campagnes de la Calypso. 7. Campagne

de la Calypso dans le Golfe de Guinée et aux Iles Principe, SAo Tomé et Annobon (1956).

16. Crustacés décapodes: Brachyoures. Annis Inst. océanogr., Monaco. (n.s.) 44% 23-124.

GrirFin, D. J. G. & Yatpwyn, J. C. 1964. A record of the majid brachyuran genus Achaeus

from New Zealand with notes on the Australian species. Trans. R. Soc. N.Z. Zool. 6: 33-51.

Man, J. G. DE 1902. Die von Herrn Professor Kiikenthal im Indischen Archipel gesammelten

Dekapoden und Stomatopoden. Jn KUKENTHAL, W. Ergebnisse einer zoologischen For-

schungsreise in den Molukken und Borneo. Abh. senckenb. naturforsch. Ges. 25: 467-929.

Misrs, E. J. 1884. Crustacea. Jn BriTIsh MUSEUM (NATURAL HISTORY). Report on the zoological collec-

tions made in the Indo-Pacific during the voyage of H.M.S. “Alert? 1881-2: 178-332, 513-575.

London: British Museum.

Ratusun, M. J. 1911. The Percy Sladen Trust expedition to the Indian Ocean in 1905. Marine

Brachyura. Trans. Linn. Soc. Lond. (2, Zool.) 14: 191-261.

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. (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)

Bu.Lioucu, W. S. 1960. Practical inveriebrate 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., DuvAL, 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

ae NE |

;

¢ ‘ ] u : ; 5 Wy \ iy 5

“4 ; oY ae 7

R f y

CUP? t

ih i vy

aie sol) Beas Bs ae, id Ns ‘eesnen

} bith hy Dart Sadan Gk |

K j sai hi we hiah ean) - Wh nh

mi ’ (et! Wen why i malt Rew Ra,

ve y ‘ ones A cay a mn RE Aha |

) } an uy \ whet U} hi ah at fi lovable e

rel eet BA her? a sa

f Tal PARTS |, ; \ 1 \ a4

'y ath eh) mn Day yed yh

i F Tae i 4 A ‘ eu hy 1)

1) Wy "SF. a me sil PRE

X Pe Lin ie Lrira nid t¢ PT “ai

oe ‘

fi pureed eh: NAY Oto coe AR W meray

, | hie he etek, ; i er ie rein | bis uli mi

; ‘ 7 | te oy By thy tie VedtR a a fir wiht

; J

i i fr ~ ¥ i \ ‘

5, f A) i 7 ate vty ¥)

ra iea Se. } Or

' iattvon hs Mean |

107. 68

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 52 ~~ °&Band

December 1968 Desember

Part. 4. Deel

THE MELKBOS SITE: AN UPPER PLEISTOCENE

FOSSIL OCCURRENCE IN THE

SOUTH-WESTERN CAPE PROVINCE

EN HSO Niay

MAR 2 0 1969

LIBRARIES

Q. B. HENDEY

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

(Cash with order, post free)

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

(Kontant met bestelling, posvry)

OUT OF PRINT/UIT DRUK

1, 2(1, 3, 5) 7-8), 3(1-2, 5, tp), 5(2, 5, 7-9);

6(1,, t=p:i2),) 7 (050 3) os G2): eerOUT =a).

11 (1-2, 7> t.—p.i.), ai, 24(2), 275 31(1-3), 38,

44(4).

Price of this part/Prys van hierdie deel

R1.50

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

1968

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 MELKBOS SITE: AN UPPER PLEISTOCENE FOSSIL

OCCURRENCE IN THE SOUTH-WESTERN CAPE PROVINCE

Q. B. HENDEY

South African Museum, Cape Town

(With plates 8-10 and 3 figures)

[MS received 10 September 1967]

CONTENTS

: PAGE

Introduction. . : 789

Geological associations aie rare ae OceUrKENCce, |). 91

Archaeological associations of the fossils . 5 Og

The faunal assemblage : : i : AG?

Systematic description of the ees : : : uN 205

Discussion of the fauna of Sue soh NOMS AA cS ORe Sea G )

Acknawledsememisiy) paeiret i ys inie Wur Mehy ya DEF

Summary . : ; : : : ‘ , Fine By)

References. : : f ‘ : ; : nr Gy!

INTRODUCTION

In the collections of the South African Museum (Natural History), Cape

Town, is an assemblage of fossils recovered from a site near the coast north of

the town of Melkbosstrand (fig. 1). The first recorded discovery of fossils at this

locality was made in 1956 by Mr. J. Rudner of Cape Town. Thereafter small

collections were made from time to time, and much of the material recovered

found its way to the Anatomy Department of the University of Cape Town. In

1962 this material was presented to the South African Museum and forms the

nucleus of the assemblage presently to be described.

The limits of ‘the Melkbos site’, as it has become known, are not clearly

defined, fossil and cultural material having been recovered from an area about

14 miles long and a few hundred yards wide, beginning about 3 miles north of

the mouth of the Sout (Klein Sout) River. The fossils have been exposed by

erosion in, and associated with, an horizon of calcareous sandstone (‘calcrete’)

and sand in the area, which forms part of the Sandveld region (Talbot, 1947)

(pl. 8A).

Pleistocene fossil occurrences are known from several localities along the

Table Bay and False Bay coasts. These include Milnerton and Ysterplaats

(Broom, 1909; Cooke, 1955), Paarden Eiland, Strandfontein and Swartklip

(Singer & Fuller, 1962; Hendey & Hendey, 1968). The Melkbos site has been

one of the most productive of these occurrences, and the South African Museum’s

collection at present includes over 600 specimens from the site. However, its

true wealth can never be accurately assessed, since it is easily accessible and

Ann. S. Afr. Mus. 52(4), 1968: 89-119, 3 pls, 3 figs.

gO ANNALS OF THE SOUTH AFRICAN MUSEUM

= = —-— REFERENCE

f Nea ‘ . 2 TERTIARY &

’ Sed er times sagt Wieenh oman t

= “ TS COU USN ye ett Ra UE, QUATERNARY DEPOSITS

f \

a7 PLC g Poca 7 eee eee ieee ese eh Oe ee

“7 x s t

> ‘——————— PRE-MESOZOIC ROCKS

a Sah Me) a / Vise he

sii p eae a aN ny

rod ae ical Banh J ee

ye eas , \ ‘

Ay Ne pee aes Bae Cee a a a eee ee ee ee eee

rea ety 1 s |.

L ’ } = J

nie cae ji Ss a

1 foe FD won

oar of OS SS eS eee Sa aad

(oo == — eee

‘

ea Se ee sa aa se

I ee a mR

fol ee we _ re

SS a ee

eter —— — SS ae ae ae eee eee

=, D

SST el

we Se a Ee eS eee

A sae ee ee ee a

“i

Ode ety se) A = — ee a ee

KBOSSTRAND=— Se es i sa eee

Se ee ee ei a i

ROBBEN OMRON Ae eT ; i ss BE see

ISLAND Sit pa Da ed Se NS oe ae = a

‘

eS = a a

os —

MILNERTON

STRANDFONTEIN

BAY

MILES

Fig. 1. The location of the Melkbos fossil site.

THE MELKBOS SITE: AN UPPER PLEISTOCENE FOSSIL OCCURRENCE QI

extensive collecting by visitors to the area takes place. In addition, the fossils are

rapidly destroyed by weathering once they have been exposed.

The purpose of this report is to place on record the presence of the site,

its relationship to other south-western Cape fossil sites and to give a preliminary

account of the more important specimens thus far recovered.

GEOLOGICAL ASSOCIATIONS AND NATURE OF THE OCCURRENCE

The deposits in the area can conveniently be divided into two categories.

Firstly, there are Recent aeolian sands, which have resulted from the

weathering and erosion of pre-existing deposits. These are, in places, vegetated

and fairly stable, but elsewhere are in the form of mobile barchan dunes. Where

they have been stabilized they are sometimes found in association with Late

Stone Age middens. They are discontinuously developed and clearly post-

date the period of fossil accumulation.

The second category of deposits are the Pleistocene sands, calcrete and

ferricrete, with which the fossils are associated. They have a long and complex

history, the sands probably having been laid down during the Tertiary, and

been transported and redeposited several times since. No systematic excavations

have been undertaken in the area, and the sequence of events suggested below

is based only on surface observations.

In the area between the fossil occurrence and the sea, a fairly extensive

platform is exposed, almost certainly ‘wavecut’, with an elevation of about 20

feet above sea-level. Krige (1927) refers this platform to his ‘Minor Emergence’,

stating that ‘the overgrown flats at Melkbosch Strand suggests a terrace of the

right order’. Assuming correspondence to the European Pleistocene sea-level

changes, and in the present instance there is no reason to believe that such a

correlation is invalid, this terrace is Late Monastirian (= Tyrrhenian IIT)

in age (Zeuner, 1959; Oakley, 1964). 7

The fossiliferous deposits at Melkbos overlie the 20-foot terrace, and there-

fore post-date it (fig. 2).

Mabbutt et al. (1955) recorded a succession at Bok Baai (fig. 1) which

includes a calcrete, almost certainly equivalent to that at Melkbos, overlying

‘Minor Emergence’ beaches and related cliff slopes.

The Early Wiirm interstadial, with which the Late Monastirian is corre-

lated (Oakley, 1964), has been assigned a chronometric date of about 40,000

years B.Pp. (Emiliani, 1961). The Chatelperronian industry of the Upper

Palaeolithic of Europe has been related, in time, to this interstadial and the

industry has been dated to about 32,000 B.c. at Arcy sur Cure, France (Oakley,

1964). The South African Middle Stone Age/Second Intermediate industries,

with which the Melkbos fauna has suggested associations (vide infra), has been

dated variously between 44,000 B.P. to 2540 B.c., with ‘the Middle Stone Age

in sensuo [sic] stricto. . . later than 40,000 years B.P.’ (Deacon, 1966).

The Melkbos fauna is therefore considered to be no earlier than 30,000 to

40,000 years B.P. No upper age limit was determined, but a date of late Upper

92 ANNALS OF THE SOUTH AFRICAN MUSEUM

20’ terrace developed

POeanie¢n: (parts of VUPPER PE eS iOG Ee

Accumulation on 20° terrace of deposits

with fossils and M.S.A. artefacts

Sea level below

present mean

2. Latter part of UPPER PLEISTOCENE

Erosion and exposure of

fossils and artefacts

AT TELE. ee ae

a. REGENT

Fig. 2. Diagrammatic representation of the suggested sequence of events at the Melkbos site.

THE MELKBOS SITE: AN UPPER PLEISTOCENE FOSSIL OCCURRENCE 93

Pleistocene is proposed for the fauna. This will serve adequately to place it

chronologically, until such time as a more definite chronometric date is

established.

It was not evident from the surface observations whether single or multiple

fossiliferous horizons exist. The fossils appeared to be preserved only in associa-

tion with the calcrete, i.e. the original lime-rich horizon of the Upper Pleistocene

palaeosol. A few very poorly preserved specimens were found in the more acid

horizons of the palaeosol, which are now marked by the exposures of ferricrete.

Gradual concentration of the calcium carbonate in the lime-rich horizon,

and induration following its exposure, has given rise to the calcrete in its present

form. The outer surfaces of the exposures are extremely hard, but the degree of

induration diminishes away from the exposed areas (cf. Du Toit, 1917: 12).

The calcrete is, therefore, not strictly speaking Pleistocene in age since some of

its characters are still being developed, but it had its beginnings at the time of

the development of the soil body overlying the 20-foot terrace and after the

fossils were already in place.

The site is exposed to both south-easterly and north-westerly gales which

are prevalent in the region in summer and winter respectively, and wind erosion

of the area is perennial and severe. The exposed fossils suffer extensive sand-

blasting, and develop a deep red or dark brown colour and a very characteristic

glazed patination, reminiscent of the ‘desert varnish’ of rocks in arid regions.

Bone fragments litter the site, but diagnostic material is much less common

(pl. 8B).

ARCHAEOLOGICAL ASSOCIATIONS OF THE FOSSILS

No positive association between cultural material of a known period and

the fossils has been demonstrated.

The fossils predate the Late Stone Age middens which often occur on or

within the Recent aeolian sands. They have been found in surface association

with undoubted L.S.A. artefacts, but this is considered to be fortuitous.

No Early Stone Age material has been recorded from the site.

Undiagnostic silcrete flakes have on occasion been found during the exca-

vation of fossils in the unconsolidated deposits, and on rare occasions there have

been ‘flakes’ of dubious authenticity found embedded in the calcrete. The for-

mer, at least, were considered to be genuine archaeological associations. Several

finely worked bifacial points of a rather small size, and flakes with prepared

striking platforms, all in a silcrete similar to that of the excavated material

have been recovered from the surface of the site. This lithic material,

undoubtedly dates from the Middle Stone Age or Second Intermediate, and an

association between it and the fossils is suggested.

THE FAUNAL ASSEMBLAGE

The nomenclature, with a few exceptions, follows that of Ellerman et al.

(1953).

94 ANNALS OF THE SOUTH AFRICAN MUSEUM

Class MAMMALIA

Order CARNIVORA

Family Hyaenidae

Hyaena cf. brunnea

Family Felidae

Felis leo aff. spelaea

Family Canidae

Canis cf. mesomelas

Order PINNIPEDIA

Family Otaritidae

? Arctocephalus pusillus

Order PROBOSCIDEA

Family Elephantidae

? Loxodonta africana

Order PERISSODACTYLA

Family Rhinocerotidae

Diceros simus

Diceros bicornis

Family Equidae

Equus sp.

Order ARTIODACTYLA

Family Hippopotamidae

Hippopotamus amphibius

Family Bovidae

Syncerus sp.

Tragelaphus cf. strepsiceros

Taurotragus oryx

Redunca arundinum

cf. Hippotragus sp.

Connochaetes sp.

Raphicerus sp.

cf. Antidorcas sp.

Order RODENTIA

Family Bathyergidae

Bathyergus suillus

Class REP TILIA

Order CHELONIA

Incertae sedis

Class AVES

Order STRUTHIONIFORMES

Family Struthionidae

Struthio australis

THE MELKBOS SITE: AN UPPER PLEISTOCENE FOSSIL OCCURRENCE 95

SYSTEMATIC DESCRIPTION OF THE FAUNA

Class MAMMALIA

Family Hyaenidae

Hyaena cf. brunnea 'Thunberg

Brown Hyaena

The only hyaenid remains recorded from the site are an incomplete

mandible (Mb 116) and an incomplete maxilla (Mb 117), which belong almost

certainly to the same adult individual (pl. 9). Both specimens are from the left side.

Mb 116—(table 1)—The mandible has lost that part of the corpus posterior

to the P,, and has suffered extensive post-fossilization weathering. The I, root,

incomplete C, P, and P,, and the almost intact P, are preserved. It exhibits

characteristics of the genus Hyaena as defined by Ewer (1954), although it differs

slightly from the H. brunnea mandibles of the comparative series.

TABLE |

DIMENSIONS OF MELKBOS HYAENA CF. BRUNNEA TEETH AND MANDIBLE, COMPARED WITH THOSE

OF A FOSSIL SPECIMEN FROM ELANDSFONTEIN AND TWO SERIES OF MODERN SPECIMENS.

Modern Modern

E’'f’tein H.. brunneat H. brunnea*

Mb 116 16686 Mean Range Mean _ S.D.

Sing al sisi ae sat Wigs Wgie

5 1323) Sh 519-3 17-16 «1-14

b Sy Ml ee 12-6 13-5 12 Dy 2502 IGFs hits 35 8

P, f alveolar 1 a i 16-7 17-3 14-7 = =13-4-15-3 _

alveolar b ar: oe 8-0 o5 UES 6-7— 8-4 —- ——

122i fiat! a sor a ZL-0 JIA 9 Ue 20:9 20°3-21-5 | 21-17 0-40

b fs c15-0 15-0 14-4 13-6-15-0 14-40 0-60

i ee ae oa E2309, 24-7 2D ie A 2D Oh 23539 O56

b oe ae on El3e7 14-2 13-4 12-9-13-7 13-30) 10- a

P,—-P, alveolar 1... ue c60-0 61-6 56:8 56-6-57-0 _ —-

Diastema length ... ah 8-7 Sez Ope OES =a Rt — —

Corpus height below P,_... 40-0 41-0 36°3 34-1-38-5 — —

Corpus width below P,__ ... 21-0 24-1t 18-7 17-4-19-5 — _

Distance between LP, and

RP, ‘en a a c58 -0 c62-0 D739) 5o-0-59>1 _ -—

+ Sample of four specimens.

* Ewer & Singer, 1956; sample of thirteen specimens.

t Measurement somewhat high due to a pathological condition in this region of the mandible.

The diastema is appreciably shorter and the P, and P, slightly longer than

those of the comparative specimens. However, the P, length corresponds closely

to the mean of a series of 13 measurements given by Ewer and Singer (1956) for

HI. brunnea.

The corpus is very robust, and falls beyond the size range of the compara-

tive series. ‘The estimated inter-corporal width falls within the range observed

in HZ. brunnea.

Mb 117—(table 2)—The maxilla has only the damaged crown of the P? pre-

served, while little more than the roots of the P? and P* remain. It shows greater

96 ANNALS OF THE SOUTH AFRICAN MUSEUM

differences from the H. brunnea comparative series than does the mandible,

although it clearly belongs to Hyaena rather than Crocuta.

TABLE 2

DIMENSIONS OF MELKBOS HYAENA cf. BRUNNEA UPPER TEETH, COMPARED WITH THOSE OF A

SERIES OF FOUR MODERN SPECIMENS.

Modern

Mb 117 Hyaena brunnea

Mean Range

P1—P# alveolar length ae us 2s bh, 80-0 712. [3 O=7aeb

f alveolar 1 ee He 2 se a 17-5 15-5 15 eae

2 alveolar b a Me me a ue 9-0 8-3 7:8- 8-8

1 Re xa a wes ne ho, 22-9 23-2 22-6-23-8

bm se ae pie os. ae we cl4-4 14-7. 13-0-15°5

Ba Spycde 14s 1 ee bute aA eh gee eaeta a acm 15°5 15-8 15-0-16-8

alveolar 1 By aes ss beth a 33-0 32-8 30-0—-34-2

4 ) alveolar b nese oe so ae i) 18-6 19-3 18-0-20°5

The length of the anterior alveolus of the P* expressed as a percentage of

the total alveolar length of this tooth, is 39-4, as against a mean of 43:6 for the

HZ. brunnea series and 29-4 for the C. crocuta series.

The alveolus of the P! is small, almost circular in shape and overlaps the

distal end of the C alveolus on its lingual side. The position of the P? relative to

the C of the comparative specimens was variable, but none exhibited an over-

lapping of these teeth. With the C set back so far relatively, the canine fossa is

very pronounced, and descends steeply from the infra-orbital foramen to the

alveolar margin. These features, taken in conjunction with the relatively short

mandibular diastema indicate an overall foreshortening of the anterior region

of the snout. The lengths of the cheek tooth rows are greater than those of the

comparative specimens, indicating that the foreshortening is confined to the

canine /incisor region.

The P? is both longer and broader than those of the comparative series,

while the P? falls within the size range of variation. The latter differs from the

P? of the comparative series in that an anterior accessory cusp is present in the

form of a small but clearly defined projection from the cingulum.

Discussion

Owing to the condition of the fossil specimens, the fact that they repre-

sent only a single individual, and in view of the limited number of comparative

specimens, there are insufficient grounds for separating the Melkbos Hyaena

specifically or subspecifically from the extant H. brunnea.

The possibility that it is not an isolated aberrant specimen is lent some

support by a recently discovered Hyaena mandible and P® (16686 and 16687)

from the Elandsfontein (Hopefield) site. These specimens exhibit some of the

THE MELKBOS SITE: AN UPPER PLEISTOCENE FOSSIL OCCURRENCE 97

unusual features observed in Mb 116 and Mb 117. H. brunnea has already been

recorded from the Elandsfontein site (Ewer & Singer, 1956), but the new

specimens differ from those previously described in much the same way as those

from Melkbos differed from the comparative series. The mandibular corpus of

16686 is extremely robust and corresponds closely in size to that of Mb 116. The

diastema is slightly shorter than that of Mb 116 (table 1). The P® (16687)

has a small anterior accessory cusp identical in development to that of Mb 117.

The teeth of the Elandsfontein and Melkbos specimens are similar in size.

Family Felidae

Felis leo aff. spelaea Goldfuss

The lion is represented in the assemblage by four specimens, which belong

to at least two individuals. Postcranial remains include the distal half of a right

radius (Mb 561) and the proximal half of a left fourth metacarpal (Mb 576).

Cranial remains include a fragment of a right mandible (Mb 603) and an incom-

plete left mandible (Mb 143).

Mb 145—(pl. 10; table 3) —This is the right mandible of an adult individual

which lacks the ascending ramus, the angular process and those parts of the

corpus anterior to the P,. The P, and M, are largely intact and well preserved.

The P, has lost the most mesial portion of the anterior cusp. In spite of

this, its estimated length was significantly greater than the length of correspond-

ing teeth in a series of 17 modern lion specimens. The breadth falls within the

range observed in this series, although it is greater than the mean.

The M, shows slight wear on the shearing blades. Its length is slightly

above the maximum of the comparative series, and the estimated breadth is

near the upper limit of the observed range.

Morphologically the teeth are indistinguishable from those of the modern

lion, although the ‘talonid’ of the M, is slightly more pronounced in the fossil

specimen. |

On discovery the specimen was embedded in a block of calcrete with only

the lingual surface of the corpus and the M, exposed. This area has been heavily

weathered, and the mandibular canal and anterior root of the M, have been

exposed. The vertical wall of the masseteric fossa and inferior margin of the

corpus posterior to the M, have been lost.

The corpus is extremely robust, but differs most notably from those of

the comparative specimens in the region of the masseteric fossa. The superior

and anterior walls of the fossa are steep, while the inferior wall becomes pro-

gressively more U-shaped posteriorly, with the development of a ridge of bone

along the buccal surface of the inferior margin. On a line 20 mm behind the

posterior end of the M,, the height measured from the top of this ridge to the

inferior margin of the mandible is + 26:0 mm. The mean for corresponding

measurements in 10 comparative specimens was 18-4 mm, with a maximum of

only 22-0 mm. Towards the posterior limit of this ridge and immediately below

ANNALS OF THE SOUTH AFRICAN MUSEUM

“SUIIOJ J[VUIS, pue IGIL], YIOG Jo s}UsWIOINSeaUL {(7/8[—-99B]) PAOJULS 2 sUTYMECy

= =m = Co. = ep Saype 19° Gr == LCG 8 "+ TAT Mojaq yysI0y snd1z0;)

= = = 98-¢G 8-6G-8-ZE OF-Gr == 8:2 ae "sg Mopeq 1ystoy snds0pH

9 G:SI-Z-Ol ZI-Fl 61-1 G:91-Z-ZI 8I-F1 = 0-919 qf

l 0-SE-Z-9% §§-OE CS: 9-0§-L:SZ Z1°8z = 0-1g 1 UW

6 §-9I-F-Ol ¥2-§I oS: I ¢-SI-8-01 $S°SI 0-SI+ 6:FI pq

== = == 68-0 G:Gl=L28 09-01 O-II+ 6-II wq< 'd

6 0-S&-6:22 §1-0§ L0°% C-62-1°&Z C8-92 0:GZ+ 0:2g2 I

poimsvour ssury uray ‘as oSue yy uesy 609 GIN SPI IW

4000] agS 03] “+1 JUIDO}SIIT 0a] SYaY Westy UIIpOy SO”qAPW

‘NIVLIGNA WOU HLAAL PAV IAdS OFT SITAd AO SAIUAS V AUNV ‘SNAWIDAdS NUAGOW

Lj SO Sdragds Vv dO ASOHL HLIM GHaVdNOO ‘AIGIGNVW GNV HLAAL OFT SITAY SOMMTAW AO SNOISNAWIAG

¢ F1av

THE MELKBOS SITE: AN UPPER PLEISTOCENE FOSSIL OCCURRENCE 99

it, there arises a horizontally projecting ridge which curves downwards in the

direction of the missing angular process. Although there was considerable

variation in the nature of this region of the mandible in the comparative speci-

mens, none showed the pronounced development of the features observed in

Mb 143. This suggests that the masseter muscle was very powerfully developed,

a factor which is in keeping with the relatively large size of the specimen.

Mb 603—(table 3)—This poorly preserved specimen, a fragment of a right

mandible, retains only a damaged P,. Although accurate measurements of this

tooth are not possible, it appears to be larger than the mean of the comparative

series.

Mb 561—(table 4) —This specimen, the distal portion of a right radius, is poorly

preserved and has lost the styloid process and much of the brachioradialis

tuberosity. Morphologically it is indistinguishable from the radius of the modern

lion, but was significantly larger than the three comparative specimens avail-

able. It corresponds closely in size to a series of four felid radii from the Elands-

fontein site, which are thought to belong to a large extinct lion (vide infra).

TABLE 4

DIMENSIONS OF MELKBOS FELIS LEO RADIUS, COMPARED WITH THOSE OF A SERIES OF THREE MODERN

LIONS, AND A SERIES OF FOUR FOSSIL SPECIMENS FROM ELANDSFONTEIN.

Elandsfontein | Modern F. leo

Mb 561 Mean Range Mean Range

Transverse diameter* 40-0 41:0 37-0-44-0 31-5 28-4—34-0

A/P diameter* Ae DN) 2 3) 205-230 AMET, © PN IO8s

* Measured immediately above brachioradialis tuberosity.

Mb 576—A metacarpal fragment, which although heavily weathered, clearly

belonged to a lion of substantial proportions.

Discussion

The Melkbos lion, while being of robust proportions, cannot on available

evidence be separated specifically from the extant F. leo. The few previously

described fossil lions from South Africa (Broom, 1939 & 1948; Ewer, 19562)

are known only from fragmentary remains, and their specific rank is open to

some doubt. In respect to size and the development of the M, ‘talonid’, the

Melkbos lion is similar to the ‘Panthera shawi’ from Kromdraai (Ewer, 1956a),

but the remains of both forms are too scanty to allow conclusive comparisons.

It is probable that a close relationship exists between the Melkbos lion and

the as yet undescribed form from Elandsfontein, which is appreciably larger

than the extant form.

The relatively abundant lion remains from Swartklip (Hendey & Hendey,

1968) represent a form indistinguishable from the extant F. leo, and it is probably

at least subspecifically distinct from that from Melkbos. Possible relationships

100 ANNALS OF THE SOUTH AFRICAN MUSEUM

of both these forms to the recently extinct Cape Lion (Felis leo melanochaitus

Smith) were considered, but owing to the lack of information on the osteology

of the latter, the question remained inconclusive.

The relatively large size of the Melkbos lion invites speculation on its

possible relationship to the ‘giant’ felines of the Pleistocene of the Northern

Hemisphere. These include the giant ‘jaguar’ of North America (Felis atrox)

(Merriam & Stock, 1932; Simpson, 1941; Kurtén, 1965a), the ‘tigers’ of the

Asian Pleistocene (F. tigris subspp.) (Hooier, 1947) and the European cave

lion (F. leo spelaea) (Dawkins & Sanford, 1866—72). It is with the latter that the

Melkbos lion is tentatively identified, although the basis for comparison is

limited by the fragmentary remains from Melkbos. Both the size and character

of the teeth, and the overall size of the animals suggest a close relationship

between the two forms (fig. 3, table 3).

BREADTH

22 24 26 28 30 32

LENGTH (2

o Felis /eo spelaea Goldfuss (Dawkins & Sanford, !868)

x Melkbos lion

Fig. 3. The length/breadth relationship of the P, of the Melkbos lion, compared with those of a

series of Felis leo spelaea specimens from Britain.

The great distance between Melkbos and the known limit of distribution of

F’. leo spelaea makes it necessary to express reservations on the identification. The

question can clearly not be resolved until such time as a greater number of

more complete specimens are known from South Africa. However, even the

extant subspecies of large felines are not easily distinguished on osteological

grounds alone. ‘This is true of the African and Indian lions (Todd, 1966) and the

tigers of south-east Asia (Hooijer, 1947). In the case of the latter there is a

THE MELKBOS SITE: AN UPPER PLEISTOCENE FOSSIL OCCURRENCE IO!

marked decrease in size towards the periphery of the area of distribution. This

is a parallel of the situation observed in many carnivores during the Pleistocene,

where a marked decrease in size occurred with the passage of time (Kurtén,

19655). The variations in the size of carnivores in both time and geography

obviously limit the use of this factor alone as a taxonomic discriminant.

Family Canidae

Canis cf. mesomelas Schreber

Black-backed Jackal

The jackal is represented in the assemblage by a single M, (MB 556),

which is well preserved and in a moderately advanced state of wear. Morpholo-

gically this tooth is not distinguishable from the corresponding teeth of the two

extant South African jackals, C. mesomelas and C. adustus. However, metrically it

corresponds more closely to the former, being only slightly larger than the

maximum of a range given for this species by Ewer (19560) (table 5).

TABLE 5

DIMENSIONS OF THE M, OF CANIS CF. MESOMELAS FROM MELKBOS, COMPARED WITH A SAMPLE

OF TEN C. ADUSTUS AND FOURTEEN C. MESOMELAS SPECIMENS (EWER, 1956B).

C’. adustus C’. mesomelas

Mb 556 Range Mean Range Mean

M | ne ae as ~ 21-0 17°4-15:0 16:2 20-7-16°5 18-8

SD: 23 cae ms one se7 6-8— 5-1 6-1 8-2— 6:5 72

In a study of the C. mesomelas remains from the Elandsfontein site, Ewer and

Singer (1956) found that several of the fossil teeth exceeded the known size

range of the modern species.

Since then newly discovered material from the Elandsfontein and Swart-

klip sites has suggested the presence of a jackal in the south-western Cape during

the Pleistocene of somewhat greater proportions than the extant C. mesomelas.

It may prove that the Melkbos specimen has affinities with this extinct

form, but at present there is no adequate reason for separating it from the

species mesomelas.

Family Otariidae

? Arctocephalus pusillus Schreber

Cape Fur Seal

A single fragment of the distal end of a humerus (Mb 546) is tentatively

ascribed to this species. Only the medial epicondyle and part of the trochlea

are preserved. It compares closely in size and morphology to the humeri of

female A. pusillus specimens. The proximity of the site to the coast in an area

where this seal occurs today, lends some support to its suggested affinities.

102 ANNALS OF THE SOUTH AFRICAN MUSEUM

Family Elephantidae

?Loxodonta africana Blumenbach

African Elephant

Three fragments of a cheek tooth (Mb 513 A & B, and Mb 514) are the

only elephant cranial remains recorded from the site. Their condition precludes

positive diagnosis, but there is no reason to believe that they belong to a form

other than L. africana.

Two fragments of a right scapula (Mb 78 & Mb 515), belonging almost

certainly to the same bone, compare closely in all observable respects to the

corresponding parts of scapulae of L. africana.

In view of the fragmentary nature of these remains, they are only tenta-

tively ascribed to L. africana.

The recently described lower molar of ‘Archidiskodon’ cf. transvaalensis

(Hendey, 1967), recovered near the mouth of the Klein Sout River, has no

apparent associations with the main Melkbos fossil site.

Family Rhinocerotidae

Diceros stmus Burchell

White Rhinoceros

The White Rhinoceros was identified in the assemblage from a series of

upper teeth of a single immature individual.

Mb 511A—(table 6) —LMz?, lacking the roots and in which the protoloph and

metaloph have been reconstructed. The crochet is slightly worn, but the crista

is still below the grinding surface. All the outer cement covering has been lost,

but some remains in the medi- and postfossettes.

TABLE 6

DIMENSIONS OF DICEROS SIMUS TEETH FROM MELKBOS, COMPARED WITH THOSE OF TWO MODERN

SPECIMENS.

Modern D. simus

Mb 511A Mb 511B 21379 21381

M 1 xe se ae mee 66:1 — 69-0 71-4

2) b a shes ae Bee c45-0 — 41-0 45-5

M | Me Nee sae as, — c67-0 69-6 61-5

2 |) ay. nae ah nee — c42°5 44-5 43-5

Mb 511B—(table 6)—LM1}, lacking the roots and outer cement covering, and

in which the ectoloph, protoloph and distal portion of the metaloph have been

reconstructed. The medifossette is separated from the prefossette by the uniting

of the crista and crochet.

Mb 511C—LP%, very poorly preserved, and either completely unworn or in a

very early stage of wear.

Mb 511D—LP’, and least well preserved of the teeth.

THE MELKBOS SITE: AN UPPER PLEISTOCENE FOSSIL OGCURRENCE 103

In size and morphology this series of teeth corresponds closely to those of

two modern D. simus specimens, and the degree of wear indicates an order of

eruption normally found in the rhinoceros, i.e. M1, M?, P®, P* (Cooke, 1950).

A skull fragment (Mb 591) was also tentatively assigned to D. simus.

Diceros bicornis Linnaeus

Black Rhinoceros

This species is represented in the assemblage by an incomplete mandible

(Mb 125). It consists of a large part of the right corpus and the symphyseal

region. The right P, and P, are largely intact, and the RM, 1s partially pre-

served. Only the roots of the other right molars, and the LP, and LP, remain.

The symphyseal region is of the characteristic D. bicornis type and in size (table

7) and morphology the teeth are indistinguishable from those of the modern

species.

TABLE 7

DIMENSIONS OF DICEROS BICORNIS TEETH FROM MELKBOS, COMPARED WITH THOSE OF A SERIES

OF FIVE MODERN SPECIMENS.

Modern D. bicornis

Mb 125 Mean Range

P.-M, alveolar length os a fe 285-0 272-0 253 -0-297-0

P l oe ss a3 ~~ ne 35-0 57-0 33-0- 39-1

3\Lb ee es 2. ue c31-0 og 30-8— 34-6

Pp 1 aos 23: sa Ree ak aye 44-6 41-9- 46-9

== b Ea Ju: sae a2 ae c34-0 39-6 34-3- 38-0

Diceros sp.

Two tooth fragments (Mb 179 & Mb 583) and two mandible fragments

(Mb 154 & Mb 427) were not identified as to species.

Rhinoceros postcranial remains are abundantly represented in the assem-

blage, but owing to the difficulty of distinguishing the two extant African species

on the basis of their postcranial skeletons, no attempt was made to categorize

these specimens specifically.

Family Equidae

Equus sp.

Equid remains are rare at the site, and only three teeth and five elements

of the postcranial skeleton have been recovered.

Mb 120—A left M?, the best preserved of the teeth, which corresponds closely

in size to the M? of the Equus helmet type specimen (Dreyer & Lyle, 1931), but

which does not exhibit the same complexity in the enamel pattern of the occlusal

surface.

Mb 121 © Mb 137—Right upper and lower premolars, respectively, and both

poorly preserved.

104 ANNALS OF THE SOUTH AFRICAN MUSEUM

In view of the present unsatisfactory classification of South African fossil

Equidae, and the number and condition of the Melkbos specimens, no specific

diagnosis is proposed.

Family Hippopotamidae

Hippopotamus amphibius Linnaeus

The hippopotamus is represented in the assemblage by an incomplete Msg,

of which only the hypoconid and entoconid are preserved.

Family Bovidae

Tribe BOVINI

Syncerus sp.

One of the most commonly represented bovids at the site is a large syncerine

buffalo.

Material

Mb 610—An incomplete and highly fragmented skull, which consists of the base

of the horn cores and frontal region, nasals and part of the left maxilla. Although

observations on the skull characters were limited by the condition of the speci-

men, it proved to be considerably more robust than the Syncerus caffer compara-

tive specimens (table 8).

The nasals correspond fairly closely in size to those of Homoioceras specimens

from the Elandsfontein site. They are flattened rather than arched, and when

articulated with the single remaining maxillary fragment it was apparent that

in general appearance and proportions, the facial region of this specie is

similar to that of ‘Homotoceras’. :

TABLE 8

DIMENSIONS OF SYNCERUS SP. SKULL FROM MELKBOS, COMPARED WITH THOSE OF SYNCERUS

CAFFER SPECIMENS.

Syncerus caffer*

3 e

MAA.

Mb 610 Mean Range No. 35445

measured

Maximum width across

nasals ... ab .. +86-0 63-8 54-5-72-0 6 c60-0

Maximum A/P width of

horn core bosses ... +247-0 2207) 207 -0—230-0 5 156-0

Minimum distance be-

tween left and right

bosses ae ‘ai c25-0 29-7 27-0-34-0 3 c77:0

Maximum distance be-

tween left and right

bosses ... os re 033-0 48-3 41-0-52-0 2 —_

* Specimens in collections of S.A. Museum (Nat. Hist.)

THE MELKBOS SITE: AN UPPER PLEISTOCENE FOSSIL OCCURRENCE 105

However, the morphology of the frontal region corresponds more closely

to that of Syncerus. The roof of the skull above the frontal sinuses is heavily

rugose, far more so than in S. caffer, but clearly quite unlike the condition in

‘Homoioceras’, in which the bases of the horns are not bossed. The median ‘valley’

separating the bosses is narrower in the Melkbos specimen than in any of the

S. caffer comparative specimens, which is a factor in keeping with the greater

overall size of the former. The anterior edge of the horn cores above the orbits

parallels the condition in Syncerus exactly, the only difference being in degree of

development. The antero-posterior diameter of the horn cores could not

be measured accurately, since their posterior margins have been lost, but the

remaining part of the left boss apparently comes close to the posterior point of

recurvature. It measures 247 mm, 17 mm more than the maximum of the

S. caffer comparative series. The males of this series were all hunting trophies

and almost certainly above average size for the species.

There were two features in which the horn cores differed from the compara-

tive specimens. The bosses did not display the marked dorsal antero-posterior

convexity present in Syncerus, and secondly, although little of the horn cores

beyond the bosses remains, it is clear that they do not have a pronounced down-

ward sweep of the horns beyond the bosses. Roberts (1951) states that, ‘as we

go northward in the continent the horns of Buffaloes dip downward less and

less until . . . they are on the top of the skull with only an outward and upward

trend’. Consequently, the sweep of the horns in Mb 610, although not typical of

the southern African S. caffer, is apparently the normal condition in their

northern relatives.

Mb 15, and associated fragments Mb 89, 95 © 96—The horn core fragments of a

-second individual, which parallel in all observable respects the features of

Mb 610, indicating that the latter is not a single atypical specimen.

Mb 608 & Mb 609— (tables 9 & 10)——Right and left mandibles of a single indi-

vidual, the former with a complete set of cheek teeth, and the latter with the

M,, M, and incomplete M;. The corpora of both halves are incomplete.

TABLE 9

DIMENSIONS OF SYNCERUS SP. MANDIBULAR CORPORA FROM MELKBOS, COMPARED WITH THOSE

OF A SYNCERUS CAFFER SPECIMEN.

Mb 67 Mb 144* Mb 608 M 147t

Maximum height below M, apg Ai Pepsin 89-0 67-6

es ee sion M, 43-2 c36-0 c47-0 30-7

Maximum height below P, 64-4 ams hae 57-2

aR cal te P, 32-5 i pane OE yaaa

* Immature individual.

t Large g Syncerus caffer.

106 ANNALS OF THE SOUTH AFRICAN MUSEUM

Mb 67—(table 9) —-The almost intact corpus of a right mandible lacking all the

teeth.

Mb 144—(tables 9 & 10)—A left mandible, lacking that part of the corpus

anterior to the M,, but otherwise intact. The M,, M, and M, are well preserved

but incompletely erupted.

Mb 4—(table 10)—An incomplete right mandible with the M, and M,

preserved.

Mb 220—Incomplete left mandible with poorly preserved M, and Ms.

Mb 465—(table 10)—An incomplete left mandible with P,, P, and M,, and

associated M, (Mb 464A) and M, (Mb 464B).

In addition to the above material, ten isolated lower and four upper

teeth belonging to this genus were recovered.

A large number of elements of the postcranial skeleton were also

recovered, but they were not studied in detail.

Morphologically the lower teeth of the Melkbos buffalo are indistinguish-

able from those of the extant S. caffer.

The premolars were similar in size to those of the single comparative

specimen available, but the molars were, in general, significantly larger (table

10). The mandibular corpora of the fossil specimens were considerably more ro-

bust than that of the single comparative specimen, which belongs to a large male.

The greatest size difference was in the region posterior to the premolar series.

None of the lower teeth approached those of ‘Homotoceras’ in size, but the posterior

parts of the corpora corresponded fairly closely to those of the smaller “Homozoce-

ras’ specimens from Elandsfontein.

Discussion

Published records of the fossil buffaloes of southern Africa are very limited,

although a large amount of material is available for study. The Pleistocene

longhorned forms from this region are now almost invariably referred to

‘Homotoceras baini’, in spite of the fact that the original description of this species

(‘Bubalus’ bain Seeley, 1891) is very inadequate, and the generic designation

(Bate, 1949 & 1951) is based on inference rather than actual study of available

material.

The relationship between the southern African ‘Homoioceras’ and the

extant Syncerus is not at all clear, and consequently speculations on the affinities

of the Melkbos buffalo to these two forms are somewhat premature at this stage.

There is, however, little doubt that the Melkbos material will be of great

importance in any future consideration of the phylogeny of the African buffa-

loes. The admittedly inadequate comparisons with S. caffer and ‘Homoioceras’

specimens made in the course of the present study, suggest that it is “inter-

mediate’ between the two forms in many respects. A more detailed comparative

study must inevitably be hampered by the limited amount and condition of the

cranial material from Melkbos, but a clearer appreciation of the characteristics

107

THE MELKBOS SITE: AN UPPER PLEISTOCENE FOSSIL OCCURRENCE

‘saffv sniaquds P IBIe'T »

“UMOAD JO 9Seq IVIU UdYL} SUOISUSUIP WINWIIXeU YI99} Joy}O [[e Uo {pazdnso Ajo}a[duIOOUT 499} JOUIS VUNIXLUT ONT} JOU 9Iv UBATS sUOIsSUSUIIC (1)

ne LEE EnInIIE SSIS GUESS

L:61 61 6°ch 6-02 2:61 0-08 PSI §:91 8:h22 |] 8-9l 8-22 6:ZI LIZ 6 O-91 | °° - Sel PIN

a die ee ee =. ie Se oe see €:F19 G-619 SS a a se “= £GG FIN.

ee SS oe at Sa eee 6'FI 2:02 —- — pron ee "* £09 GW

— — 0:-0Fr Se aS <= —- — —- = —- = ps ee (1) GIN

0:022 — 0-069 | 1:02 4:61 66€ I-I1Z 0:02 0-S¢9 a Sc aan me er pat ne (x) PPI IN

a a oe ee Oe O:L19 1:2 6'SI 81 a a a Cor IN

GANG CG Le = ae Bese ee = ae — — — — —- = ue uN A $9b IW

— 8-12 I:1Z 2:88 a eo a a ae ee V FOr IW

0:029 L:1% Z:SP 0:62 ES L-E 0G «6S1S «(GEE I‘LI SZ 9:GI 9:02 8:01 S:ST | °° us "* 809 GIN

pa eu = +] jis jelgh meg fea q I qe 4

STAT a TANI WI | Wea on or

a ra) UE un ncn SS aa

‘NUWIOAdS YAVAVID SANFONAS V dO ASOHL HLIM GauVdWOD ‘sOmMTAN WOU HLAAL WAMOT ‘dS SQYFINAS JO SNOISNAWIG

Ol AIaVvL,

108 ANNALS OF THE SOUTH AFRICAN MUSEUM

and status of the Elandsfontein buffalo will enable a more objective analysis of

the Melkbos material to be made.

It was considered that adequate grounds exist at present for assigning this

material to the genus Syncerus. The size differences alone cannot justify an

alternate conclusion, and while other critical differences may be revealed when

more complete skulls are known, the available evidence suggests that the

material belongs to what is merely an extremely robust buffalo of the Syncerus

type.

Family Bovidae

Tribe TRAGELAPHINI

Tragelaphus cf. strepsiceros Pallas

The specimens tentatively assigned to this species are few in number, and

almost all are poorly preserved. They are, however, of particular interest since

they provide an additional record of an apparently extinct form of koodoo,

which inhabited the south-western Cape in later Pleistocene times.

Material

Mb 134—The incomplete horn cores and frontlet of a large male, rather poorly

preserved and reconstructed in part. The horn cores are robust, their diameter

at the pedicle being comparable to those of the largest of the modern Greater

Koodoo (T. strepsiceros) specimens available. Although only a part of their

original length is preserved, it is evident that the helices had a considerably

shorter pitch than those of the modern species. This feature is also apparent in

the koodoo horn cores recovered at the Elandsfontein site.

Mb 24—The fragment of a left horn core, also with a tighter spiral than that of

T. strepsiceros.

Mb 388—The fragment of a left frontal and horn core.

Mb 135 A-E— (table 11) —A series of left upper cheek teeth (P?—M®?), recovered

in association with Mb 134, and belonging almost certainly to the same large

male. Comparison with the upper cheek teeth of three male T. strepsiceros

specimens revealed that morphologically the teeth are indistinguishable, but

that they are significantly shorter, with the exception of the P*. A similar dis-

parity in tooth size is observed in the Elandsfontein koodoo material.

Discussion

The limited amount and condition of the material available precludes the

possibility of establishing the specific status of the Melkbos koodoo. The

differences from the extant T. strepsiceros, which it shares with material from

the Elandsfontein site, strongly suggest that it belongs to a form at least sub-

specifically distinct from the modern southern African koodoo. It clearly has

greater affinities with this form than with the Lesser Koodoo (T. imberbis.)

A detailed study of the Elandsfontein material, which is more abundant

and, in general, better preserved, should resolve the question of the classifica-

tion of this fossil form.

THE MELKBOS SITE: AN UPPER PLEISTOCENE FOSSIL OCCURRENCE

TABLE I|1

109

DIMENSIONS OF MELKBOS TRAGELAPAUS CF. STREPSICEROS UPPER TEETH, COMPARED WITH

A SERIES OF THOSE OF THREE MODERN KOODOO.

Tragelaphus strepsiceros

Mb 135 Mean

ps ] on fa fe ihe 14-8 17-5

b ae iz. Ls vo 16-0 16-7

ps 1 ay Pt: cai ue 14-0 14-8

b af ae Ak ve 16-3 17-6

1 18-4 TAN

Mi< bm 20:2 20-6

bd 19-9 20:3

] 23-4 26:5

M?2< bm 23°] SVS)

bd Dilias] 22-4

1 25:0 28-7

M?< bm Oa a 22:

bd 17-8 21-1

Family Bovidae

Tribe TRAGELAPHINI

z T aurotragus oryx Pallas

Eland

Range

16-9-18-0

eS IAD OS)

22-0 24°5

21-7-23-3

28-0-30- 1

22-20-2329

20-8-21-3

The eland is poorly represented in the assemblage, and only four specimens,

none of which show appreciable differences from the extant form, were assigned

to this species.

Mb 70—(table 12) —The complete horn cores and frontlet of a large adult

individual, and the best preserved of all the antelope cranial material recovered

at the site.

TABLE 12

DIMENSIONS OF MELKBOS TAUROTRAGUS ORYX HORN CORES AND FRONTLET.

Skull width at horn pedicles

A/P diameter

Base of right horn core :

Transverse diameter

Distance between tips of horn cores

Maximum length of left horn core

Mb 589—A horn core fragment.

Mb 11—A mandible fragment with a single damaged molar.

Mb 10—A mandible fragment with incomplete M, and M,.

Mb 70

170-0

77-0

66-0

350-0

510-0

I1IO ANNALS OF THE SOUTH AFRICAN MUSEUM

Family Bovidae

Tribe REDUNCINI

Redunca arundinum Boddaert

Reedbuck

The reedbuck is the most commonly represented antelope in the assem-

blage. Eighteen horn cores, mostly incomplete, were recovered, and these did

not differ in any observable respect from those of the extant species (table 13)

Also recovered were an incomplete M! or M?, and a fragment of mandible

(Mb 215A & Mb 123).

TABLE 13

DIMENSIONS OF MELKBOS REDUNCA ARUNDINUM HORN CORES, COMPARED WITH THOSE OF A

SERIES OF FOUR MODERN SPECIMENS.

Melkbos specimens Modern R. arundinum

No.

Mean Range measured Mean Range

A/P diameter oe aS: 33:4 28-0-38-0 6 35-1 31-9-38-0

Transverse diameter Heh 33-8 30-0-38:-7 3 34°3 32-2-3656

This material may well have affinities with the reedbuck from the Swart-

klip sites (Hendey & Hendey, 1968), which, although on the basis of horn cores

is indistinguishable from the modern species, appears to differ from it in other

respects. However, at present there are no adequate grounds for separating the

Melkbos reedbuck from R. arundinum.

Family Bovidae

Tribe HIPPOTRAGINI

cf. Hippotragus sp.

Two specimens in the assemblage were doubtfully ascribed to the genus

Hippotragus.

Mb 122—A right dp, in an advanced state of wear, and which corresponds

fairly closely in most respects to the dp, of a mandible from the Elandsfontein

site (12209) that has been tentatively ascribed to Hippotragus.

Mb 1—(table 14)—An incomplete right mandible with the M, and M,, roots

of the P, and M,, and alveoli of the P, and P3. The ascending ramus and the

most anterior part of the corpus have been lost. In size the corpus and teeth are

similar to those of a Syncerus caffer specimen (M 147, see tables 9 and 10), but the

teeth are hippotragine rather than syncerine in character. The mesial enamel

plication of the M, and Msg, a characteristic of hippotragines but not present in

Syncerus, does not extend through the entire crown height, and would have

become less pronounced as wear on the teeth progressed and would have dis-

appeared completely even when an appreciable amount of the crowns remained.

In all the hippotragine comparative specimens examined, this accessory fold

extended through the entire crown height.

THE MELKBOS SITE: AN UPPER PLEISTOCENE FOSSIL OCCURRENCE Lit

TABLE 14

DIMENSIONS OF MELKBOS CF. HIPPOTRAGUS sP. TEETH AND MANDIBLE.

Mb | Mb 1

Height of corpus below M, .... Ea 8 T3PD ly pcaae4as

Width of corpus below M, _.... oe ae c33°3 M,< bm. 45-7 @20-8)t

Height of corpus below M, ... wf ah 57 tbhd-4 15-6. (19-0)

Width of corpus below M, _..... t. i 34-0 fab atten? 25

Height of corpus below P;_.... wz = 48-3 M;< bm 14-5 (18-0)

Width of corpus below P; _... 4 SA 25-9 Ubd — f5-50(¥7 738)

* Measurements at the occlusal surface. P,-M;

= Maximum dimensions near base of crown. alveolar 1 161-7

A giant hippotragine is known from the Pleistocene of Africa (Hippotragus

gigas Leakey, 1965), and this species is thought to be present in the Elandsfon-

tein assemblage. It is possible that the specimen Mb 1 has affinities with this

form.

Family Bovidae

Tribe ALCELAPHINI

cf. Connochaetes sp.

A total of six horn core fragments and seven teeth from the assemblage were

tentatively assigned to the genus Connochaetes (sensu lato). Attempts to place this

material specifically were inconclusive.

Family Bovidae

Tribe NEO TRAGINI

Raphicerus sp.

The neotragine group of antelopes is represented in the assemblage by a

single poorly preserved horn core (Mb 29). It is indistinguishable in all observa-

ble respects from the horn cores of the extant species, R. campestris and R.

melanotis, both of which occur in the south-western Cape today.

Family Bovidae

Tribe ANTILOPINI

cf. Antidorcas sp.

A pair of incomplete horn cores (Mb 177 E and F) were doubtfully ascribed

to the genus Anitidorcas. Little of the right horn core is preserved and observations

were confined to that from the left (Mb 177E).

The antero-posterior axis of the horn core is at an angle to the median line

of the skull, the orientation in this respect being as in A. marsupialis. In the verti-

cal sense its angulation begins as in A. marsupzalis, but unlike this species it does

not curve posteriorly, remaining instead remarkably straight along its entire

112 ANNALS OF THE SOUTH AFRICAN MUSEUM

length. In this respect it corresponds more closely to the form of the horn cores

of A. marsupialis australis from Swartklip (Hendey & Hendey, 1968).

It also corresponds closely in size to the horn cores of the males of the Swart-

klip subspecies, having an anteroposterior diameter of c27-5 mm and a trans-

verse diameter of c24:0 mm at the pedicle. The transverse compression of the

core expressed as a percentage (Gentry, 1964) is 87-3, which is within the range

of variation encountered in the modern A. marsupialis. In vertical section the

inter-frontal plane is almost identical to that of the Swartklip specimen ZW 67a.

Only the marked straightness of this specimen is not in keeping with the

horn cores of Antidorcas, and it is possibly a feature of an isolated aberrant

individual, but until other specimens are recovered, the tentative diagnosis

must stand.

Family Bovidae

Incertae sedis

In addition to the material already described, sixteen incomplete teeth,

two incomplete mandibles and two horn core fragments were recovered. None

of these specimens suggest the presence of a form other than those already listed.

No attempt was made to classify the large number of bovid postcranial

bones recovered at the site.

Family Bathergidae

Bathyergus suillus Schreber

Cape Dune Mole

An almost complete skeleton (Mb 173) and an isolated lower incisor

(Mb 49) of Bathyergus were recovered at the site. Although the condition of

these specimens resembled that of the other fossils, it is possible that they are

recent intrusive elements in the assemblage. The dune mole is common in

the area today, and skeletons, or parts thereof, of recently dead individuals have

been recovered at the site. It is possible that such remains which become buried

may take on the appearance of the other fossils simply by becoming

discoloured. ait

Intrusive elements at surface sites such as Melkbos can often not be detected

without resort to chemical analyses, and since none were carried out in con-

junction with the present survey, the status of these questionable elements in the

assemblage remains uncertain.

Class REP TILIA

Order CHELONIA

Incertae sedis

Fifty fragments of carapace, probably all belonging to land tortoises, were

included in the assemblage, but no attempt was made to classify them.

THE MELKBOS SITE: AN UPPER PLEISTOCENE FOSSIL OCCURRENCE 113

Class AVES

Family Struthionidae

Struthio australis

Ostrich

The ostrich is represented in the assemblage by two first phalanges (Mb 496

A & B), which are indistinguishable in all observable respects from those of the

extant species.

DIscuUSSION OF THE FAUNA

There are a number of sites in the south-western Cape Province which have

yielded mammalian fossils, and while the study of the fauna of these sites is still

in progress, it is apparent that in time they span the greater part of the Quater-

nary. Most attention has been focused upon the fossil occurrences at Lange-

baanweg, Elandsfontein (Hopefield), Melkbos and Swartklip, and the fauna of

these sites is now moderately well known (table 15).

The Langebaanweg sites are both the earliest and potentially the most

important of them all. They date probably from ‘the earliest phases of the

Pleistocene’ (Boné & Singer, 1965), and have yielded a number of archaic

forms, such as Hipparion, Anancus, Stegolophodon, an agriotheriine bear and a

nimravine. The fossils have been recovered in the course of the mining of phos-

phate by the African Metals Corporation (Singer, 1961), and most of the

material recorded to date has come from three quarries, viz. Baard’s, ‘E’ and

‘C’ Quarries. The deposits in Baard’s Quarry are apparently largely fluviatile,

those in ‘E’ Quarry are estuarine and in ‘C’ Quarry there is a mixture of marine

and estuarine sediments. The mammalian faunas from the three quarries are

apparently at least broadly contemporary. The fossils have no known archaeolo-

gical associations.

The fauna of the Elandsfontein site is the best known in the region (cf.

Singer, 1962 for references), and the geology and archaeology of the site have

recently been the subject of intensive study (H. J. Deacon and J. Wymer, in

preparation). The fauna includes a number of extinct genera and species, but

modern forms, or forms only subspecifically distinct from them, are well repre-

sented. The date usually used in reference to the fauna of this site is late Middle /

early Upper Pleistocene (Boné & Singer, 1965), but it is possible that later

elements occur in accidental association with the main body of fossils (Inskeep

& Hendey, 1966). Artefacts of three industrial complexes have been recovered

at the site, with the main faunal element apparently associated with an evolved

Acheulian (‘Fauresmith’) industry (Howell & Clark, 1963; J. Wymer, pers.

comm.).

| The sites at Swartklip, which are late Pleistocene or Recent in age, have

yielded a fauna made up almost entirely of modern forms (Hendey & Hendey,

1968). These sites have no certain cultural associations, but tenuous links with

the Late Stone Age and historic times (Post-1652 a.p.) have been suggested.

II4 ANNALS OF THE SOUTH AFRICAN MUSEUM

TABLE 15

THE FAUNA OF THE PRINCIPAL FOSSIL SITES OF THE SOUTH-WESTERN CAPE PROVINCE.

Class MAMMALIA

Order ARTIODACTYLA

Family Hominidae

*Flomo sapiens rhodesiensis

Family Cercopithecidae

*Simopithecus oswaldi hopefieldensis

Order PHOLIDOTA

Manis sp. ..

Order CARNIVORA

Family Mustelidae

Mellivora capensis ..

* Aonyx sp.

Family Canidae

Subfamily Caninae

Canis mesomelas

Canis adustus

*Canis sp.

Subfamily Simocyoninae

*Tycaon pictus magnus

Lycaon pictus

Family Ursidae

Subfamily Agriotheriinae

*cf. Agriotherium sp.

Family Viverridae

Herpestes ichneumon

Family Felidae

Subfamily Felinae

Felis serval . .

Felis caracal

* Felis sp.

Felis leo a

*Feiis leo aff. spelaea

Subfamily Machaerodontinae

* Megantereon cf. gracile

Subfamily Nimravinae

* Dinofelis sp.

Family Hyaenidae

Hyaena brunnea

* HTyaena sp. ses

*Crocuta crocuta spelaea

Order PINNIPEDIA

Family Otariidae

? Arctocephalus pusillus

*cf. Arctocephalus sp.

Langebaan- | .&

weg 5.

Tia oh

r 7

s =|

3 3

i) BN ees

Os] 8 |) SaaS

2) 3) 3 | See

Paicancae: |= | -

| ey | ty | eee ee

ael(SiFlaleaia

x x

x jock ave

cf. ch.

xX xX

4 cf.

x Aiea

<x | Chae

Kt OX

Kaul) oes: es

THE MELKBOS SITE: AN UPPER PLEISTOCENE FOSSIL OCCURRENCE I T5

Table 15 — cont.

Langebaan-

Elandsfontein (Hopefield)

Swartklip

_———_ |__| ————— | ———— |

Order PROBOSCIDEA

Family Gomphotheriidae

* Anancus sp. ‘ ee an ie 5% x

Family Elephantidae

* Stegolophodon sp.

* Stegodon sp. ‘

*cf. ‘Archidiskodon’ a ee ae ae

*“Archidiskodon’ broom... nyt Ne as x

*‘Toxodonta’ zulu .. er ee a a x

? Loxodonta africana uy ss a: it x

Order PERISSODACTYLA

Family Rhinocerotidae

Diceros bicornis

Diceros simus

* ? Diceros sp.

Family Equidae

* Hipparion albertense baardi - a RIS SQ. eR

*Fquus plicatus .. ds ea Bi ) ee x

* Equus helmei os be shy a OTM @ <

Equus sp. .. ae is us ste vy <i

Order ARTIODACTYLA

Family Suidae

* Mesochoerus lategani

* Mesochoerus paiceae

* Tapinochoerus meadowsi a - ae

*Incertae sedis (2) oe ‘! ae a8 x

Family Hippopotamidae

Hippopotamus amphibius .. vs 4 Biv by (Gl Bd lok «| Che

Family Giraffidae

*Tibytherium olduvaiense .. i a Se (ee KEE

*Giraffa gracilis... ae ae or 3 chy nek:

Family Bovidae

*Tragelaphus cf. strepsiceros ih 8 am ee ex

Taurotragus oryx .. a a oy. a8 xlhix

*“Homotoceras bainit’ : x

* Syncerus sp. fe a a ee eA x

Redunca arundinum ae - ae of SX

Redunca cf. arundinum 4 i Ge 4. x

* Redunca sp. ate y ere) ure) nx

* Hippotragus cf. lewcophacus a 4 be Xx

* Hippotragus spp. ay ne si Ss chal x4, x

*Damaliscus sp... Be we v om ley e©

* Damaliscus sp. ;

cf. Connochaetes sp.

*Tunatoceros cf. mirum

* Megalotragus sp. ... Ae ssa ie

cf. Oreotragus sp. ... a Bia ome cer ee i’ @

Raphicerus sp. ae us BN St a

OO“

OO)

x OK OK

x XK &

x xX XK XK

116 ANNALS OF THE SOUTH AFRICAN MUSEUM

Table 15 — cont.

)

Langebaan- | &

weg a

et ore

- =)

O| f) 6 ee

2) S| Slee ieee

ge) ae} ey

a1 2 | 2 |e

s/O|B|a| ala

Family Bovidae (cont.)

* Antidorcas marsupialis australis

* Antidorcas sp. Kee

*Gazella cf. wellsi A hi Bee x

*Gazella spp. ae a se a eal OX <I

* Incertae sedis (+1) so nN ae Ee a es

Order CETACEA

Incertae sedis x

Order LAGOMORPHA

Lepus capensis aps He is ee x

Lepus sp. ... 5 ae ats st ae x

Order RODENTIA

Family Bathyergidae

Bathyergus suillus x | ee

Georychus cf. capensis

Family Hystricidae

Aystrix cf. africae-australis x

Family Muridae

Otomys cf. saundersiae x

Parotomys cf. brantsi 2 e is

Incertae sedis (+1) 2 <5 oh Py x x

Class REP TILIA

Order CHELONIA

Incertae sedis (+1) a a ft re x % x < x 4

Class AVES

Order STRUTHIONIFORMES

Struthio australis eh ae e Wi x | Eee are

Incertae sedis (-++1) tee o ae ae aie

Class CHONDRICHTHYES

Several genera and species te a 2 wu x

Class OSTEICHTHYES

Several genera and species a 7 AA ate EG

* Extinct forms.

The Melkbos fauna has essentially the same ‘modern’ character as that

fron. Swartklip, but it includes elements such as the lion, hyaena and koodoo,

which have affinities to forms from Elandsfontein. In addition, the buffalo and

? hippotragine antelope from Melkbos are extinct forms not represented in the

Swartklip assemblage, and it is concluded that the Melkbos fauna predates that

from Swartklip. The archaeological associations of these two sites, while too

dubious to be of real value, do at least suggest an age difference in the right

THE MELKBOS SITE: AN UPPER PLEISTOCENE FOSSIL OCCURRENCE EV?

order. Since none of the extinct genera of the Elandsfontein fauna (Simopithecus,

Mesochoerus, Megantereon, etc.) occur at Melkbos, the fauna of the latter site

almost certainly post-dates the main Elandsfontein fauna.

These four sites have yielded widely different numbers of specimens,

ranging between about 600 from Melkbos to about 20,000 from Elandsfontein,

and they have been studied with varying degrees of thoroughness. Nevertheless,

it is concluded that there are sufficient grounds for considering Melkbos to be

intermediate in age between Elandsfontein and Swartklip, and that these three

occurrences date from the latter part of the Quaternary. The Langebaanweg

fauna takes a more isolated position in the earlier part of the Pleistocene.

ACKNOWLEDGEMENTS

Over the years many persons have presented fossils they had collected at

the Melkkos site to the South African Museum. Without their interest and

generosity this report would not have been possible. In this connection I wish

particularly to mention Prof. R. Singer of the University of Chicago and Messrs.

J. Rudner, H. G. A. Craye, R. R. Inskeep and G. Hoehn, all of Cape Town.

Prof. Singer kindly assisted and encouraged in many ways and I am

especially grateful to him.

I wish also to thank Prof. L. H. Wells of the University of Cape Town for

many helpful discussions.

Dr. E. C. Olson of the University of Chicago kindly read the manuscript

and offered many useful suggestions.

The conclusions reached in this report do not necessarily reflect the views

of any of these people.

Finally I wish to thank my wife who was a constant source of

encouragement.

This study was supported in part by the U.S. Public Health Service grant

no. G.M. 10113 to Prof. R. Singer, and the National Science Foundation grant

no. B2453 to Dr. E. C. Olson.

SUMMARY

The location, geological and archaeological associations, and fauna of a

fossil site north of the town of Melkbosstrand, Cape Province is discussed. The

remains of eighteen types of mammals, one bird and one tortoise are described or

mentioned. It is concluded that the site dates from the latter part of the Upper

Pleistocene, that it post-dates that at Elandsfontein (Hopefield), and is earlier

than the Swartklip fossil sites.

REFERENCES

- Bate, D. M. A. 1949. A new African fossil long-horned buffalo. Ann. Mag. nat. Hist. (12) 2:

396-398.

Bate, D. M. A. 1951. The Pleistocene fauna of two Blue Nile sites. The mammals from Singa

and Abu Hugar. Fossil Mammals Afr. 2: 1-28.

118 ANNALS OF THE SOUTH AFRICAN MUSEUM

Bont, E. L. & Sincer, R. 1965. Hipparion from Langebaanweg, Cape Province and a revision

of the genus in Africa. Ann. S. Afr. Mus. 48: 273-397.

Broom, R. 1909. On the evidence of a large horse recently extinct in South Africa. Ann. S. Afr.

Mus. 7: 281-282.

Broom, R. 1939. A preliminary account of the Pleistocene carnivores of the Transvaal caves.

Ann. Transv. Mus. 19: 331-338.

Broom, R. 1948. Some South African Pliocene and Pleistocene mammals. Ann. Transv. Mus.

21: 1-38.

Cooke, H. B. S. 1950. A critical revision of the Quaternary Perissodactyla of southern Africa.

Ann. S. Afr. Mus. 31: 393-479.

Cooke, H. B. S. 1955. Some fossil mammals in the South African Museum collections. Ann. S.

Afr. Mus. 42: 161-168.

Dawkins, W. B. & SANForD, W. A. 1866-1872. A monograph of the British Pleistocene Mamma-

lia. x. British Pleistocene Felidae. Palaeontogr. Soc. | Monogr.] [1866-72]: 1-194.

Deacon, J. 1966. An annotated list of the radiocarbon dates for sub-Saharan Africa. Ann. Cape

prov. Mus. 5: 5-84.

Dreyer, T. F. & Lyte, A. 1931. New fossil mammals and man from South Africa. Bloemfontein:

Nasionale Pers.

Du Tort, A. L. 1917. Report on the phosphates of Saldanha Bay. Mem. geol. Surv. S. Afr. 103: 1-38.

ELLERMAN, J. R., Morrison-ScoTt, T. C. S. & Hayman, R. W. 1953. Southern African mammals,

1758-1951; a reclassification. London: British Museum (Natural History).

EMILIANI, C. 1961. Cenozoic climatic changes as indicated by the stratigraphy and chronology of

deep-sea cores of globigerina-ooze facies. Ann. N. Y. Acad. Sct. 95: 521-536.

Ewer, R. F. 1954. Some adaptive features in the dentition of hyaenas. Ann. Mag. nat. Hist. (12)

7: 188-194.

Ewer, R. F. 1956a. The fossil carnivores of the Transvaal caves. Felinae. Proc. zool. Soc. Lond.

126: 83-095.

Ewer, R. F. 19565. The fossil carnivores of the Transvaal caves. Canidae. Proc. zool. Soc. Lond.

126: 97-119.

Ewer, R. F. & SINGER, R. 1956. Fossil Carnivora from Hopefield. Ann. S. Afr. Mus. 42: 335-347.

Gentry, A. W. 1964. Skull characters of African gazelles. Ann. Mag. nat. Hist. (13) 7% 353-382.

HENDEY, Q. B. 1967. A specimen of ‘Archidiskodon’ cf. transvaalensis from the south-western Cape

Province. S. Afr. archaeol. Bull. 22: 53-56.

HENDEY, Q. B. & HENDEy, H. 1968. New Quaternary fossil sites near Swartklip, Cape Province.

Ann. S. Afr. Mus. 52: 43-73.

Hooyer, D. A. 1947. Pleistocene remains of Panthera tigris (Linnaeus) subspecies from Wanhsien,

Szechwan, China, compared with fossil and recent tigers from other localities. Am. Mus.

Novit. 1346: 1-17.

Howe ., F. C. & Criark, J. D. 1963. Acheulian hunter-gatherers of sub-Saharan Africa. Publs.

Anthrop. Viking Fund 36: 458-533. (Howe Lt, F. C. & Bour.iRE, F. eds. African ecology and

human evolution.)

INsKEEP, R. R. & HENDEY, Q. B. 1966. An interesting association of bones from the Elandsfontein

fossil site. Actas del 5 Congreso Panafricano de Prehistoria y de Estudio del Cuaternario, Santa Cruz

de Tenerife, 1963, 2: 109-123. (Publnes Mus. arqueol., Tenerife 6: 109-123.)

Krice, A. V. 1927. An examination of the Tertiary and Quaternary changes of sea-level in

South Africa. Annale Univ. Stellenbosch (A) 5: 1-81.

KurtTeEn, B. 1965a. The Pleistocene Felidae of Florida. Biol. Sci. Bull. Fla St. Mus. 9: 216-273.

KurteEn, B. 19655. The Carnivora of the Palestine caves. Acta zool. fenn. 107: 1-74.

Leakey, L. S. B. 1965 Olduvai Gorge 1951-1961 1. Cambridge: University Press.

MassutTtT, J. A., RUDNER, J. & SIncER, R. 1955. Geomorphology, archaeology and anthropology

from Bok Baai, Darling district, Cape Province. S. Afr. archaeol. Bull. 10: 85-93.

Merriam, J. C. & Stock, C. 1932. The Felidae of Rancho La Brea. Publs Carnegie Instn 422:

1-231.

THE MELKBOS SITE: AN UPPER PLEISTOCENE FOSSIL OCCURRENCE II9

OAKLEY, K. P. 1964. Frameworks for dating fossil man. London: Weidenfeld & Nicholson.

Roserts, A. 1951. The mammals of South Africa. Johannesburg: C.N.A.

SEELEY, H. G. 1891. On Bubalus bainii (Seeley). Geol. Mag. (3) 8: 199-202.

Smmpson, G. G. 1941. Large Pleistocene felines of North America. Am. Mus. Novit. 1136: 1-27.

SINGER, R. 1961. The new fossil sites at Langebaanweg, South Africa. Curr. Anthrop. 2: 385-387.

SINGER, R. 1962. Simopithecus from Hopefield, South Africa. Biblthca primat. 1: 43-70.

StncErR, R. & FuLier, A. O. 1962. The geology and description of a fossiliferous deposit near

Zwartklip in False Bay. Trans. R. Soc. S. Afr. 36: 205-211.

TALBoT, W. J. 1947. Swartland and Sandveld. Cape Town: Oxford University Press.

Topp, N. B. 1966. Metrical and non-metrical variation in the skulls of Gir Lions. 7. Bombay nat-

Hist. Soc. 62: 507-520.

ZEUNER, F. E. 1959. The Pleistocene period. London: Hutchinson.

ee cS eR

ahs tohin

by,

Ann. S. Afr. Mus., Vol. 52 Plate 8

A—A view of the site showing recent dunes (background and left), and exposures of calcrete.

B—A typical surface on the site showing bone fragments and blocks of calcrete (top left).

Ann. S. Afr. Mus., Vol. 52 Plate 9

A—Buccal view of Hyaena cf. brunnea maxillary fragment (Mb 117). (Scale in centimetres

and inches.)

B—Lingual view of Hyaena cf. brunnea mandible fragment (Mb 116).

Ann. S. Afr. Mus., Vol. 52 Plate 10

Buccal view of Felis leo aff. spelaea mandible fragment (Mb 143).

j my) ; he, ie dle.

5 : ‘ R ;

i) n : eh t » , a 7 ht Pein

i : , : , oth me : bi Pies

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. (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, 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)

Bu.LLoucu, 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., Duva, M. & Rarry, A. 1933. Etudes sur les changes respiratoires des littorines.

Archs Kool. 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. 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. Polyphacophora, 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 coronalta Lamarck, 1816: pl. 451, figs. 5 a, 6; Liste: 11. Turton, 1932: 80

lori 2a. Lagictle tl hen Gono aie heabe Be ali

| A shirk eae ; /! ie sh senlint he Bird Leese ead (Peay y ae! Pata

can it Pers } Pas Ge iit “a pani é ais * ere inatiie

' ef Beige P Leahy Wd aod,

Fi gy oy j & rey (py, vant bey a ir arte rent he ‘

GN Se ay tat t

ne crite ee manors taagaher, abet freaihonss ura ol

pT TS

a Soda ede. aktetlbe aims

Ms hey fr Pee at dee

” » “ j

. vies etn

Hy hay ye rE)

f ( 17 ‘My Lao ishowmee

Ws smu hes ch NAM Nae a

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 52 °&Band

March 1969 Maart

Part (ii 50m Deel

A DOLPHIN (STENELLA ATTENUATA) FROM

DURBAN, SOUTH AFRICA

PETER B. BEST

Cape Town Kaapstad >

aN) HSO Ni4 p>

JUL 16 1969

C/BRARIES

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

(Cash with order, post free)

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

(Kontant met bestelling, posvry)

OUT OF PRINT/UIT DRUK

I, 2(1, 3> D> 7-8), ais, 5> t—Pas) 5(2, 5» 7-9)»

6(1, alone 70; 3). 8, g(1-2), 10(I-3),

ri(ies, 7> tee) 2i, 24(2), 27> 31(I-3), 38,

44(4).

Price of this part/Prys van hierdie deel

R1.85

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

A DOLPHIN (STENELLA ATTENUATA) FROM DURBAN,

SOUTH AFRICA

PETER B. BEstT

South African Museum, Cape Town

(With plates 11-17 and 1 figure)

[MS. received 15 January 1968]

CONTENTS

PAGE

Introduction : ‘ Sap oi

External appearance . =) DOE

Skeleton : ; 6 Re

Additional notes : 5) Be

Summary : : tee

Acknowledgements. =| 13H

References : : He Rev

INTRODUCTION

On 4 July 1964, a female dolphin seven feet and half an inch long was

stranded on Battery Beach, Durban, South Africa. Mr. J. Wallace of the

Oceanographic Research Institute recognized the unusual appearance of the

specimen and so preserved it entire in a deep-freeze. With the kind co-opera-

tion of the Institute, this dolphin was eventually presented to the South African

Museum (catalogue number 35515) in September 1966, where on examination

it proved to be a specimen of the narrow-snouted dolphin Stenella attenuata

(Gray, 1846), the first recorded from South Africa for nearly 100 years.

EXTERNAL APPEARANCE

Colour pattern. Mr. Wallace took several black and white photographs of the

specimen while it was still fresh, and also made a pencil sketch of the distribu-

tion of pigment. Although the dolphin was subsequently in a frozen state for

over two years, and was in fact pure black on arrival at the museum, immersion

in water rapidly caused most of the original colour pattern to reappear. The

distribution of pigment, therefore, has been accurately recorded, but less

confidence can be placed in the colour tones attributed to these pigmented

areas after two years in cold storage.

The dolphin was basically a dark slate-grey above and a light slate-grey

below, the border between the two being fairly sharply defined in the anterior

half of the body but less defined in the area behind the dorsal fin (fig. 1). A light

grey band began just behind the apex of the melon and ran back above the eye,

I2I

Ann. S. Afr. Mus. 52 (5), 1969: 121-135, 7 pls, 1 fig.

122 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 1

where it joined a similar-coloured band from the angle of the gape. This com-

mon band then passed back above the flipper, subsequently spreading out to

cover the entire ventral surface as far as the base of the tail. There was no

darkening or lightening of the grey pigment along the centre of the belly. The

upper border of the light grey area continued along the body from the side of

the head in a gradually downward-curving sweep, which in the region between

flipper insertion and dorsal fin levelled out more than half-way down the side.

Thereafter it curved up sharply at the level of the dorsal fin, the grey colour

from each side meeting over the back about half-way between the dorsal fin and

the tail. The portion of the back just anterior to the tail-stock, however, was

dark slate-grey, giving the area of light grey pigment behind the dorsal fin the

appearance of an upward-curving band. The remainder of the upper and lower

surfaces of the tail flukes was dark slate-grey.

Both surfaces of the flippers were similarly coloured, and a dark streak ran

from the flipper insertion to just below the angle of the gape, whence it extended

as a black margin along the lower jaw to meet the corresponding band of the

other side towards the jaw tip. The eye was ringed with dark slate-grey, and

this extended forward as a narrow streak along the base of the melon. From

each side of the apex of the melon a light grey streak ran back to the lateral

margin of the blowhole, narrowing all the way.

The throat was chiefly light slate-grey in colour, but with yellowish-pink

mottling extending as a widening band from each angle of the gape, forming a

‘bridle’ round the throat. The extreme tips of both upper and lower jaws were

yellow-grey flecked with black, as were the inner margins of the lips and the

palate, where the black flecks were larger. The tongue was a bright pink.

Small white flecks, mostly elongated in an anterior-posterior direction,

and on average about three-sixteenths of an inch in diameter, were scattered

over most of the body surface. They were less abundant on the throat and belly

as far as the genital aperture, but were particularly numerous on the sides

behind the dorsal fin and on the ventral caudal peduncle. There were no such

markings on the rostrum or either side of the flippers and tail flukes, but there

were a few at the base of the dorsal fin. A narrow area on the back from head to

dorsal fin was also largely devoid of flecks.

Although the colour pattern in many Stenella species is either unknown or

poorly described, this specimen was clearly separate from forms which are

A DOLPHIN (STENELLA ATTENUATA) FROM DURBAN, SOUTH AFRICA 123

typified by a dark stripe down the side, such as S. euphrosyne, S. lateralis, S. caeru-

leoalba and S. styx. It also seemed distinct from S. plagiodon, which has a pure

white belly and no stripe from the flipper to the angle of the mouth, and

S. graffmani, which is coal black all over except for many whitish-grey spots,

particularly on the back. Among other forms in which the colour pattern is

known in some detail, the specimen most closely resembled S. frontalis, as

described by Fraser (1950), and S. attenuata, as described by Nishiwaki, Naka-

jima & Kamiya (1965).

The colour description of the Atlantide dolphin, identified by Fraser (1950)

as §. frontalis, agreed closely with the Durban specimen in both the distribution

of major pigment zones and the colour values assigned to them. Thus the

Aitlantide dolphin was described as black above and a dark grey below, the

boundary between the two zones following the same curve down the body as

in the Durban specimen. The colour of the dorsal fin, tail flukes and flippers

was identical in the two animals, and the Atlantide dolphin also had a dark

stripe from the flipper to the angle of the mouth, and a dark ring round the eye

which extended forward to the base of the snout. Similar irregular light grey

flecks covered the back, and white flecks were present on the belly.

The two specimens differed, however, in the colour of the snout and lower

jaw. Whereas the Atlantide dolphin had the distal third of the snout whitish-

pink in colour, only the extreme tip of the snout of the Durban dolphin was light

in colour, being a yellowish-grey flecked with black. Similarly, the forward end

of the lower jaw was described as white with dark spots posteriorly in the

Ailantide dolphin, whereas the Durban specimen only had the extreme tip of the

lower jaw yellowish-grey flecked with black. In both dolphins, however, the

light colour anteriorly was continued as a narrow band along the margin of

upper and lower jaws as far as the angle of the gape.

The band of yellowish-pink mottling on the throat of the Durban specimen

was absent in the Atlantide dolphin.

Apart from these differences in the colour of the snout and lower jaw, the

only other features of body coloration that differed were the presence of abun-

dant dark spots on the belly and a narrow band of lighter grey mid-ventrally,

both of which were present on the Atlantide dolphin but not seen on the Durban

dolphin.

Nishiwaki, Nakajima & Kamiya (1965) described a number of dolphins

from Arari, Japan, which they identified as S. attenuata. The colour pattern as

figured and described for these dolphins agreed in almost every detail with that

of the Durban specimen. The only discernible difference appeared to be in the

extent of light pigment on the snout and lower jaw, which was greater in the

Arari dolphins but at the same time was considerably less than in the Aflantide

dolphin. Nevertheless, this was a variable feature, Nishiwaki et al. remarking

that such white tips to the snout were absent in foetuses and new-born calves.

Consequently both the Arari and Durban dolphins differed from the Aflantide

specimen in the extent of white on the snout, in having a ‘bridle’ of light

124 ANNALS OF THE SOUTH AFRICAN MUSEUM

mottling, and in lacking dark spots on the belly and a lighter mid-ventral area.

Body proportions. In general shape the Durban dolphin resembled many other

species of Stenella and Delphinus in possessing a slender, streamlined body, a

narrow beak clearly separated from the rest of the head, small curved flippers

and a centrally-placed, recurved dorsal fin. A series of 32 body measurements

were made of the Durban specimen, and these, expressed as percentages of

the body length, have been compared with similar measurements recorded

for the Ailantide dolphin (Fraser, 1950) and for the Arari dolphins (Nishiwaki

et al., 1965) in table 1.

All but two measurements of the Durban dolphin fell within the range of

the Arari dolphins, but in nearly every case they were nearer the lower end of

the range. The two exceptions were the length of the dorsal fin base and the

width of the tail flukes, the first of which was less and the second of which was

greater than that recorded for any Arari dolphin. The base of the dorsal fin in

the Durban dolphin was prolonged posteriorly as a narrow ridge along the

caudal region, and so its actual length was extremely hard to judge. For this

reason the posterior end of the base was arbitrarily fixed as that point imme-

diately below the tip of the dorsal fin, and so this measurement was probably

not directly comparable to that recorded for the Arari or Atlantide dolphins.

The relative width of the tail flukes in other dolphins is known to increase with

the size of the animal, and as the Durban specimen was larger than any mea-

sured at Arari this probably accounts for the greater width of the tail flukes. In

most respects, therefore, the Durban dolphin could be considered as indis-

tinguishable in body proportions from those collected at Arari.

Ignoring the length of the dorsal fin base, the Atlantide dolphin apparently

differed from both the Durban and Arari dolphins in having a shorter beak

(measurement number 3) and an anus situated farther forward. However, the

position of the latter was fixed differently in the Atlantide specimen, the measure-

ment being taken from the notch of the flukes rather than from the tip of the

snout, and this may have introduced some disparity. Nevertheless, the Atlantide

dolphin seems to be clearly separated from the Durban and Arari dolphins by

its shorter beak.

SKELETON

Skull. The skull and mandibles of the Durban dolphin are shown in plates 11

and 12. ,

The right premaxilla extended back about 15 mm farther than the left,

and was considerably wider at its termination. Unlike the Atlantide dolphin,

neither premaxilla came into contact with the nasal bones posteriorly, there

being a clear wedge of the maxillary bone 4 to 5 mm wide between the tip of

the right premaxilla and the nasal. As in the Aiélantide dolphin, the prenarial

triangle was concave and the apical portions roughened. The lateral margins

of the triangle were also elevated above and overhung the adjacent areas of the

maxillae, particularly at the level of the anterior margin of the superior nares,

*(1x9} 99s) dy uy Mojeq AT[voNI9A yuIOd 0} uy [esIOp Jo UOHJISUT JOLI9}UL WOTF DOULISI[ y “POP1OII JOU JUOUIIINSvOUT JO 911K "YISug] [210}

Woy snue O} saxyNY JO YoU, IouvjsIp sutjovsjyqns Aq pouiyjqg , *.stoddipy jo dy 0} ynous jo dy, sour stp wosy Jaddiy jo yysus] sunoenqns Aq Peureyqo ;

a a

' ‘j 6 ; yi sus] ‘Q,OyMoTg ‘ze

WIPIM “poymorg “1g

yisug] ‘surusdo jeuy ‘¢9¢

9A00I3 AOS 3 SuIpNpoUur yyBue] “Ws [eyTUsS-ouy 6%

ae Ps. ha e : yisuz] “1s [eyru98-ouy “97

yisusy “ys Areurureur yor +z

manele Ws Arvuruew jySrY -9z

ysug] ‘ok *CZ

" aystey ‘Aq “47

yojou jo yjdap ‘soyny prey, *¢z

aod JO119}uUe UO yuIOd jsarv9U 0} Yo}0u ine TU, °ZZ

Zi . y : ; i. : g SSS ehatroni chai © YIP ‘sayny [rey “[Z

jie : : ’ : : ‘ : aseq jo y)sug] ‘uy [esIoq °*(~z

" * yysroy “uy [es1oq “G]

pm wnurxew “taddipq “gy

dy 0} eyprxe ‘soddyq +z]

Oe ot UOTIIISUT JOLIO}Ue sodety A 9]

snue jv ‘YIN *C|

(nous wosy joj ¢) UNUIXeU “YIN “ZT

° . . ° . . . . . . eyixe 12 at phiey Tes

zaddn puodaq mef ramoz Jo uotoaforg “zy

JOYMOTY JO 91}U99 0} 949 Jo aayUIDH “TT

odes jo a[8ue 0) 949 Jo anjuary ‘¢]

snue JO 31}U99 0} ynouS Jo dry,

ainjiode [e}1U98 Jo 913U90 0} ynous jo dry,

" * uy [es1op jo dry 0} ynous jo dry,

ge JO UONJOSUI IO1I9}Ue 0} ynOUuS jo dry,

odes jo a[d8uv 0} ynous jo dry,

9TOYMOT 07 ynous Jo dry,

uojout Jo xodv 0} ynous Jo dry,

I ane JO 91}U99 0} ynNOUS jo dry,

9 é 5 : ° 5 ; : O % A 0 yisu9] [P10],

FNFNOMAANNNOCOOCOYTONnrMNrOOHT OO HT TODO O

° ° . ° ° . ° . .

eRe OMe eNO (CSE COBEN ICO: SCO COLO Oe eS St Ot ©

= ON

SoS CQ =

OV N

le 8°¢L-8: CL €°bL OVLsG GL 20:02

= SS SOR ™

Sse tL OO

—

BNOTHOON DS

i=)

homes |

i=)

—_—

—

56 PP £ 6

GO6T “78 72 THeMTYSIN] OS6I “soserq

(njonuayyn *s)) (s1j0juoLf Ss) urydjop JUIUIOINSvIYI

surydjop weay urydjop ueqing

apyunyy

‘uOsTIedUIOD OF DJVNUIYID *s) pure Sipmuot ‘S jo oro a cop pec om jo syusuainsvou Apog

_ — —— — ee ae —— — <—_ — _—

126 ANNALS OF THE SOUTH AFRICAN MUSEUM

where they formed distinct promontories. The smooth rostral portions of the

premaxillae were arched more prominently than either their roughened lateral

portions or the adjacent maxillae. At the base of the rostrum the lateral margins

of the maxillae were raised into a roughened ridge, as described for the Atlantide

dolphin.

The palate conformed to the normal Stenella pattern in lacking a pair of

longitudinal grooves. At its extreme anterior tip the palate was concave, and in

profile curved slightly downward. The central region, however, was noticeably

concave, bearing a strong median groove, but this closed up about 30 mm

anterior to the hindmost tooth. The palate was then almost flat, but thereafter

the centre rose rapidly as a prominence bearing the pterygoid bones. The ptery-

goids were closely apposed in the midline, their posterior margin being roughly

bracket-shaped. The ventral surface of each bone was raised into a slight ridge

which diverged posteriorly and became less prominent on the pterygoid wing.

These ridges did not seem to be as well developed as in the Atlantide dolphin, and

were placed nearer the midline. The inner edge of each tooth alveolus was raised

up as a little cone-like ridge, and these formed a more or less continuous band

along the whole tooth-row, being marked off anteriorly from the rest of the

maxilla by a shallow groove.

The lower jaws were not ankylosed at the symphysis. The mandibular

rami in lateral view seemed to be more curved than those figured for the

Atlaniide dolphin, and the tips were more tapered.

The dental formula, expressed in such a way as to indicate the normal-

sized teeth and the smaller anterior teeth, was

R 3/ +14, 37-04

38-2 37+2,

which hardly differed from that of the Aélaniide dolphin, and fell well within the

38-45

36-43

A series of 20 measurements were made of the skull and mandibles of the

Durban dolphin (table 2).

In a comparison of the skull of the Ailantide dolphin with 30 other simi-

lar specimens, among them the types of Delphinus attenuatus, Steno capensis,

Delphinus doris, Delphinus frontalis and Delphinus fraenatus, Fraser (1950) was able

to separate the skulls into two groups from the shape of the rostrum distally:

this was best expressed by the width of the rostrum at a quarter its length from

the tip. One of these groups, including the types of frontalis, fraenatus and doris

and the Ailantide specimen, could be distinguished by a distinctive range in six

other measurements from the second group which included the types of attenua-

tus and Steno capensis. In this way it was possible to state that in the frontalis group

the rostrum at the middle, the zygomatic, preorbital, postorbital widths and the

premaxillary width proximally were all greater than in the attenuata group, and

the height of the mandible at the coronoid process also seemed to be greater in

the frontalis group.

range of

given for the Arari dolphins.

0-dS-6:9P

bv 9I-L: Fl

6:°98-€-18

G:6S-¢: 9F

0-FI

6-8

SOON ©

ide)

9:¢6S-1 -6P

OF risse or

8:°S8-¢6:F8

Gof

TLG-0¢

"8:41

JS 38-S8

"Ls: S8

60S

ENOMIG

"¥P] 10 1YSII Jay}OyM ‘usUIOads Yove JO; SuIpvor WNUIIxeYy 7

8:81-0:91

6°1S-S-6P

9: 1S-€-6P

8:°68-1:F8

6:06-8:68

€:6S-0: 8h

9:-CS-L: 8h

G:6l 18

806

L0G

Lov 19

LVE

LIG

916

sisAydutés Jo y}SuaT

(7) ‘op

Oy eyo hyper

jo dt} 03 Mor y100} Jo pus sJopurFT

(7) op

(a) dYysI9Yy Prouos0;)

(T) ‘op

(a) Yy suey s[qrpuey]

(T) ‘op

(wD) eyiIxewoid

ug di} 0} Mor Y}00} Jo pus sJopuTFT

JYSroy “essojy jesoduia J,

*" * yysuay] “essoy teroduia J,

Roskieid ae pus 0} winsysor jo dry,

sorvu Jo1odns

‘jue O01 Windsor jo dry,

aes Ajpeurxoid

eypixeulsid jo yIpIM wWnNuTxeyy

; : Wie ena

YIPIM oneulosA47

EE | eMES LOM S04 |

RR AERIS ARINC KOI GT

yisus] € yIpIMm wins4s0yy

s9yo}0U. [e]Iq107;Ue

jo UOJ} UL WU 99 YIPIM winIjs0y

Q[PpIuU 7e YIPIM wanI1sOYy

oseq 1e YIPIM wINIsOY

: : ; : YISUZ] WINIySOY

YISUgyT [210,],

UISIeUL

‘06

‘6!

‘81

14!

———

L-91 @%-6I-9-§1 -~

Z-6h G+ §S-1-9F

$-6F ae a

bbl 0°91-%- SI

9-F1 + a a

C-F8 0-88-9-18

b°b8 ee ae

66h O-SS-O-LPL .

DO Goaesn Mele OY

= — Rr 6ll

= — 121

= — 9-69

= — G+L9

Osfl F Ol-e-cp 9-81

GEG. 2 TeGe-G8ces [6

bch = L-Gh-L-68 = 2 GP

L-Gh 1-SP-L-0h F-SP

0-88 0-Zi-9:Sg §-0F

— — 6:8

Zbl =Ss«dL-91-Z-SI «9 ST

— = €-ZI

9:16 6:82-0:04 1°22

P20G = {6-10-)9G= G46

001 — 001

uvoyl osuey ueop|

“70 49 IHEMTYSIN

996] ‘s9SeI,7

(eus[>H{ 1S)

sisuadpo “32 *s)

OG6I ‘s9Se1y

dnois

syoquosy *s

OS6I ‘tse.

dnois

pyonuayyo *S'

C961

(ojonuajyv *S")

surydjop 1esiy

urydjop

ueqing

JUSWOINSeo]I

‘uostreduros 104 sisuadva ‘Jo “¢' pure symjuolf -s “vjonuayD Ss Jo as0y} YIM UTYdjop ueqinq ey} Jo s}uoUIomnsvour 14s

Z TIGVL

128 ANNALS OF THE SOUTH AFRICAN MUSEUM

When the measurements of the skull and mandibles of the Durban speci-

men, expressed as percentages of the skull length, were compared with those

recorded for the frontalis and attenuata groups by Fraser (1950), it was imme-

diately obvious that the specimen had a greater affinity to the attenuata group.

Of 18 of the measurements, nine fell exclusively within the range recorded

for the attenuata group, while five were shared by both groups. Only two mea-

surements (numbers 17 and 18) fell exclusively within the range of the frontalts

group, both being measurements referring to the mandible. However, two of

the skull proportions used by Fraser to distinguish between the two groups, the

zygomatic width and the rostrum width at three-quarters of its length, fell

between the ranges for frontalis and attenuata.

When the skull measurements of seven dolphins from Arari (Nishiwaki e¢

al., 1965) were compared with the ranges recorded for the frontalis and attenuata

groups, a similar overlap was found between the ranges of both groups in

13 of the 18 measurements. These included four of the proportions found

by Fraser to be of diagnostic value in separating attentuata from frontalis: the

rostrum width at the middle, the postorbital width, the premaxillary width

proximally and the coronoid height of the mandible. Three additional measure-

ments could be related only to the attentuata group, including the zygomatic

width, while two were referable only to the frontalis range: the preorbital width

and the rostrum width at three-quarters of its length.

The close similarity in external appearance between the Durban and

Arari dolphins has already been established. Nevertheless, nine of the skull

proportions of the Durban dolphin fell outside the range recorded for the

Arari dolphins. These included three of Fraser’s diagnostic measurements, the

postorbital and zygomatic widths and the premaxillary width proximally.

Consequently it appeared that the Durban skull bore a closer similarity to the

specimens in Fraser’s attenuata group than to the Arari dolphins. Neither the

Durban nor the Arari dolphins, however, fully conformed with the skull pro-

portions found by Fraser to be typical of frontalis or attenuata, the Arari speci-

mens in fact appearing to occupy an intermediate position between the two

groups.

Fraser (1966) has listed the skull proportions of a series of Stenella cf.

capensis from around St. Helena in the South Atlantic. Twelve of the proportions

recorded for the Durban dolphin, for which there were comparable data from

St. Helena, fell within the range of the S. cf. capensis specimens. The only

exception, the length of the mandibular symphysis, was a measurement which

Fraser had stated to be only really useful for distinguishing species in which it is

widely divergent. Seven of the 12 proportions in fact fell within the standard

deviation of the means recorded for the St. Helena dolphins. Unfortunately,

however, the external appearance of the St. Helena dolphins was unknown,

but there seemed to be a close similarity between the skulls from this locality

and the Durban specimen. It should be added that all of the skull proportions

of S. cf. capensis straddled the ranges recorded for the attenuata and frontalis

A DOLPHIN (STENELLA ATTENUATA) FROM DURBAN, SOUTH AFRICA 129

groups by Fraser (1950), even in the six measurements considered by Fraser to

be of value in distinguishing between /ronialis and attenuata.

Vertebral column. There were 78 vertebrae, the vertebral formula being as follows

—cervical 7, thoracic 16, lumbar 19 and caudal 36. This contrasted with the

number given by Fraser (1950) for the Aélantide dolphin, 81, composed of 7

cervical, 15 thoracic, 23 lumbar and 36 caudal, but was in close agreement with

the formula given for the Arari dolphins (Nishiwaki et al., 1965). According to

these authors individual variation in the vertebral formula was very little, the

typical column consisting of 7 cervical, 15-16 thoracic, 18-19 lumbar and 37

caudal, or a total of 78 vertebrae.

The epiphyses were fused to their centra throughout the vertebral column,

indicating that the Durban dolphin had attained full physical maturity.

The atlas had a breadth of 84 mm between the outer margins of the articu-

lating surfaces. The lateral processes were 25 mm long, and the length of the

neural spine from its tip to the nearest point on the anterior edge of the neural

canal was 47 mm. The neural spine bore a prominent central ridge about 3 mm

high (maximum), and on each dorsal extremity of the articulating surfaces

there was a sharp spine 4-5 mm long which projected towards the anterior edge

of the neural canal (pl. 13).

The axis was fused with the atlas, only the neural arch being free laterally,

as in both the Aflantide and Arari dolphins. The axis had a short, pointed lateral

process to each side which was directed obliquely backwards and slightly

upwards. These processes measured 6—7 mm in length, and so were somewhat

shorter than those described for the Ailantide dolphin (13 mm).

Unlike the latter specimen, both the neural arch and centrum of the third

cervical vertebra were free of the axis in the Durban dolphin. Such also

appeared to be the case in the Arari dolphins. Cervical vertebrae 3 to 6 all had

small neural spines, whereas the neural arch of cervical 7 was not complete but

had a gap mid-dorsally and thus no spine: this feature is recorded for the

Ailantide dolphin but not for the Arari dolphins. Nevertheless, evidence given by

Fraser (1950) indicated that an incomplete neural arch to cervical 7 has been

recorded for Stenella graffmani and Delphinus delphis, and so has little taxonomic

significance. There were short lateral processes from the base of neural arches

3, 4 and 7, with only the very slightest indications of one on vertebrae 5 and 6.

The process on cervical 7 was the largest of all, and the whole of the lateral wall

of the neural arch was greatly widened. Lateral processes from the side of the

centrum were clearly present in cervicals 5 and 6, with the faintest trace of one

on cervical 3. Cervical 7 bore a lateral process low down on the right hand side

of the centrum. As in the Aflantide and Arari dolphins, the lateral process on

cervical 6 was the largest, being directed anteriorly and 10-12 mm in length.

Cervical 7 also bore a facet on the centrum for the articulation of the first rib.

The neural spines of the remainder of the vertebral column sloped back-

wards as far as lumbar 8 (31 in the column), the ninth being vertical (pl. 14).

130 ANNALS OF THE SOUTH AFRICAN MUSEUM

From lumbar 10 to caudal 15 (57 in the column), the spines were directed

anteriorly, but caudal 16 was vertical. From there till caudal 25, which was the

last vertebra bearing a distinguishable neural spine, they again sloped back-

wards. These facts agreed closely with the description for the Atlantzde dolphin

(Fraser, 1950).

The height of the neural spines increased rapidly from thoracic 1 to 15,

but then remained more or less constant at 83-85 cm until lumbar 13, when

their height decreased, slowly at first but rapidly after caudal 19. The spines on

the thoracic vertebrae tended to be larger than in the Aélantide dolphin, those

on the first, second, third, tenth and last respectively being 33, 43, 51, 70 and

83 mm, as compared to 29, 37, 47, 59 and 76 for the Atlanizde dolphin.

Metapophyses were absent on the first two thoracic vertebrae, but slight

prominences appeared in the relevant position on the third. This contrasted

with Fraser’s (1950) statement that metapophyses were lacking from the first

three thoracic vertebrae of the Atlantide specimen. From thoracic 4 to 7 the

metapophyses were well developed and separate from the prezygapophyses,

but from thoracic 8 onwards these two processes were fused and have been

termed metapophyses thereafter. They were developed as far back as lumbar

7 (30 in the column), and remained only as traces from lumbar 8 to 17. Meta-

pophyses reappeared, however, on lumbar 18, and persisted as separate pro-

cesses until caudal 17 (59 in the column). From caudal 18 onwards the processes

from each side tended to fuse together, and the last vertebra bearing a trace of a

metapophysis was caudal 23 (65 in the column, as in the Atlantide dolphin). The

Durban specimen, therefore, seemed to have developed metapophyses on more

vertebrae than the Atlantide dolphin, the central area of column where they

persisted only as traces being ten vertebrae as opposed to 22 vertebrae (Fraser,

1950). The Arari dolphins seemed to have a similar greater development of

metapophyses from the plates figured by Nishiwaki et al. (1965).

Transverse processes on the first four thoracic vertebrae were almost

straight transverse, but from thoracic 5 onwards they projected obliquely back-

wards, particularly so in the caudal region. The last traces of transverse pro-

cesses were found on caudals 18 and 19 (60 and 61 in the column). Fraser

(1950) found the last traces of transverse processes on the 61st and 62nd verte-

brae of the Ailantide specimen.

Arising close to the tip of the transverse process on the last lumbar vertebra

there was a prominent ridge which ran obliquely backwards. On successive

posterior vertebrae this ridge was situated closer to the centrum, and from caudal

7 onwards placed on the body of the centrum, rising dorsally to occupy a position

either side of the base of the neural arch. The last trace of this ridge was found

on caudal 22 (or 64 in the column). On caudals 2 to 6 the ridge was accom-

panied by a small anterior projection to the leading edge of the transverse

process. A similar ridge and process has been described for the Ailantide dolphin

(Fraser, 1950).

Vertically perforating foramina for blood vessels started on caudal 13

A DOLPHIN (STENELLA ATTENUATA) FROM DURBAN, SOUTH AFRICA Ig!

(55 in the column), as compared to the 56th vertebra of the Aflantide specimen.

The first chevron bone was found immediately below the centrum of the

44th vertebra in the column, but appeared from its position and structure to

consist of two adjacent chevron bones fused in an anterior-posterior direction.

Consequently, the first caudal vertebra was considered as number 43 in the

column. This first chevron bone was not fused in the midline, as has also been

recorded for the Arari and Ailantide dolphins. The last three pairs of chevrons

also consisted of two separate halves (as in the Arari dolphins), and were

located on the 7oth to 72nd vertebrae in the column. Asa result, there must have

been a total of 29 chevron bones, the first consisting of two fused together. This

was Closer to the total of 28 bones recorded for the Arari dolphins than to that

of 24 recorded for the Aélantide dolphin, though in the latter case the smallest

of the chevrons appeared to be absent.

Ribs. The Durban dolphin had 16 pairs of ribs, two of which were free: according

to Fraser (1950), the Atlantide specimen had 15 pairs of ribs, one being free,

and the Arari dolphins were recorded as having 15 to 16 pairs, one or two of

which were free (Nishiwaki e al., 1965). The first six pairs of ribs were double-

headed, numbers 7 to 14 possessing only a tubercle, as in the Aélantide dolphin.

The first pair of ribs was very strongly compressed, their transverse breadth

being 14—24 mm but their thickness only 5 mm. The second pair was similarly

compressed but to a lesser extent, the corresponding measurements being 14-15

mm and 5 mm. As described by Fraser (1950) for the Ailantide dolphin, each of

the remaining ribs had a sharp, posteriorly-projecting keel on the dorsal surface

which extended from the tubercle to the front where the rib bent round laterally.

However, by the 13th rib this had become a mere trace, and was absent from

the last three ribs.

Sternum. The sternum of the Durban dolphin consisted of two parts, the manu-

brium fused with the first mesosternal element, and the second mesosternal

element as a separate, smaller bone (pl. 15). The third mesosternal element

appeared to be absent. This contrasted with the sternum described for both the

Ailantide and Arari dolphins, in which the manubrium, first and second meso-

sternal elements were fused, and in which the third mesosternal element was

present as a separate bone. The manubrium of the Durban dolphin had a

central foramen measuring 5:5 by 3°5 mm, similar to that figured for the Arari

dolphins, whereas the Ailantide dolphin had no trace of a foramen. The measure-

ments of the sternum were as follows: manubrium breadth in front of the first

rib articulations 48 mm, behind these 65 mm, and between the ends of the

lateral processes 76 mm. Behind these processes the manubrium was 38 mm

wide. In general shape, therefore, the manubrium was very similar to that

of the Aflaniide dolphin, but was relatively wider behind the lateral processes.

The first mesosternal element was about 50 mm long and 19 mm wide at the

middle, and the second mesosternal element 35 mm long and 17 mm wide.

Both elements were also relatively wider than in the Aélantide dolphin.

132 ANNALS OF THE SOUTH AFRICAN MUSEUM

There were only nine pairs of sternal ribs present in the Durban dolphin,

as against 11 pairs in both the Arari and Afélantide dolphins. This was pro-

bably correlated with the reduction of one mesosternal element in the Durban

dolphin. All but the last two pairs of sternal ribs bore faceted ends which

indicated that they were probably articulated with the sternum.

Hyoid bones. The basihyal was only partly fused to the thyrohyals, the basihyal

length being 31 mm and its width about 28 mm. There was no median notch

to the basihyal anteriorly as in the Atlantide dolphin, and in fact the anterior

edge of the basihyal had a slight, V-shaped projection (pl. 16), a feature which

was also figured for the Arari dolphins (Nishiwaki et al., 1965). The basal

breadth of the thyrohyal was 19 mm and its length 56 mm. The stylohyal length

was 71 mm and its thickness 12 mm.

Pelvic bones. The pelvic bones were simple and rod-like in shape, measuring

69 mm in length with a transverse width of about 7 by 4 mm (pl. 16).

Scapula. The scapula was small and very similar in general outline to that figured

for the Arari dolphins. The shape of the acromion, however, was different

(pl. 17), its distal margin being straight and forming a much acuter angle

with the axis of the acromion than that figured for either the Aitlantide or Arari

dolphins, although there is apparently considerable intra-specific variation

in this feature (Fraser, 1950). The head of the coracoid was longer than that

figured for the Arari dolphins, and the dorsal margin was more curved. The

measurements of the scapula were as follows:

Greatest antero-posterior length—151 mm.

Height from anterior margin of acetabulum to upper margin of scapula —

103 mm.

Greatest length of acromion—37 mm.

Greatest breadth of acromion—27 mm.

Greatest length of coracoid—30 mm.

Fore limb. The flipper was of the same general outline as those figured for the

Aitlantide and Arari dolphins. The measurements of the flipper bones were as

follows:

Greatest length of humerus—51 mm.

Greatest breadth of humerus distally —37 mm.

Greatest length of radius—67 mm.

Greatest breadth of radius distally —33 mm.

Greatest length of ulna —62 mm.

Transverse breadth of proximal row of carpals—61 mm.

There were three bones in the proximal carpal series, identified as the

scaphoid (in contact with the radius), the lunar (in contact with both radius

and ulna) and a bone in contact with the ulna which appeared to consist of the

A DOLPHIN (STENELLA ATTENUATA) FROM DURBAN, SOUTH AFRICA 133

cuneiform fused with the pisiform (pl. 17). Fraser (1950) stated that the pisi-

form was completely absent in the Atlantide dolphin. The distal series of carpals

also consisted of three bones, the trapezium (in contact distally with the first

metacarpal), the unciform (in contact distally with the third and fourth meta-

carpals) and a bone between the two in contact with the second and third meta-

carpals which seemed to represent the fusion of the trapezoid with the magnum.

In the Atlantide dolphin the trapezoid was still present as a separate but very

small element, whereas in the Arari dolphins the distal carpal series seemed to

consist of only three bones as in the Durban dolphin.

Distal to the carpals were five metacarpals, of which the first is fused with

the first phalanx in the right flipper only (pl. 16). The number of phalanges

was as follows; first digit one, second digit nine, third digit six, fourth digit two

and fifth digit one. Alternatively, if all the joints distal to the carpals were

considered to be phalangeal, the formula became I:2, II:10, III:7, 1V:3 and

V:2, which compared with I:2, Il:9, II1:6, [1V:3 and V:2 for the Atlantide

dolphin and I[:2, II:9, I1I:7, 1V:3 and V:2 for the Arari dolphins. Intra-

specific variation in the phalangeal formula, however, is known to be con-

siderable for other odontocete species.

ADDITIONAL NOTES

The Durban dolphin was an adult female in anoestrus. There was no milk

in the mammary glands, which from a histological examination proved to be in

a resting state. The diameters of the uterine cornua differed considerably, the

left horn (24 mm) being larger than the right horn (17 mm). This was asso-

ciated with a similar difference in the size of the ovaries, the left one (3-87 g)

being considerably larger than the right (1.93 g—preserved weights). As dis-

cussed by Ohsumi (1964), members of the genus Stenella seem to ovulate exclu-

sively from the left ovary until about nine corpora have accumulated, when the

right ovary may begin to function. This effect is responsible for the greater

development of the left side of the reproductive tract in the Durban dolphin, for

macroscopically the left ovary appeared to contain the scars of about 16 cor-

pora whereas the right had none. There was no trace of a functional or recent

corpus luteum, and no enlarged Graafian follicle was found.

The contents of the stomach consisted of a mass of small fish bones and

one squid beak. 3

A post-mortem examination of the lungs revealed that the dolphin was

suffering from a form of necrotizing pneumotitis at the time of death. An ulcer,

probably of traumatic origin, was present in the lining of the first stomach.

SUMMARY

Evidence from the skull proportions of the Arari and St. Helena dolphins

indicates that the frontalis and attenuata groups distinguished by Fraser (1950)

are not as exclusive as was at first thought, and this has further complicated the

identification of dolphins apparently belonging to the frontalis/attenuata complex.

134. ANNALS OF THE SOUTH AFRICAN MUSEUM

In external appearance the Durban dolphin was closest to the Arari

dolphins, although the appearance of S. cf. capensis from St. Helena is at present

unknown. In skull proportions, however, the Durban specimen was rather

different from those from Arari, was more closely related to Fraser’s attenuata

group, and was apparently closest to S. cf. capensis from St. Helena. Data from

the post-cranial skeleton is difficult to evaluate because so little is known of the

amount of intra-specific variation within any of the relevant species. Indeed,

the skeleton of S. cf. capensis is completely unknown. In most features the post-

cranial skeleton of the Durban dolphin resembled that figured for the Arari

dolphins. The vertebral formula, free third cervical vertebra, greater develop-

ment of metapophyses, the shape of the hyoid bones and arrangement of the

carpal series in the fore limb all showed a greater similarity to the Arari speci-

mens than to the Atlantzde dolphin. As in the skull, however, the skeleton of

the Durban dolphin had certain other characters apparently unlike either the

Arari or Atlantide specimens, notably in the development of the sternum and

associated sternal ribs.

Until more information on the external appearance and skeleton of the

Stenella species around St. Helena is available, it seems that the Durban dolphin

must be referred to Stenella attenuata (Gray, 1846). As such, it is the first specimen

of the species to be recorded from South Africa since the type skull of Steno

capensis was sent to the British Museum in 1865 by E. Layard of the South

African Museum (Gray, 1866). This skull, the locality of which was recorded

as the Cape of Good Hope, has subsequently been included in Prodelphinus

attenuatus by Flower (1885) and True (1889), and in the Stenella attenuata group

by Fraser (1950). To date, these two specimens represent the only records of

S. attenuata from South African seas.

ACKNOWLEDGEMENTS

I am indebted to the Trustees of the Oceanographic Research Institute,

Durban, for presenting the dolphin to the South African Museum, so enabling

me to examine it, and also for their permission to publish this paper.

I must also thank Professor C. J. Uys of the Pathology Department,

University of Cape Town Medical School, for undertaking a post-mortem

examination of material collected from the dolphin, and Mr. Sidney Kanne-

meyer of the South African Museum for preparing the skeleton and photo-

graphing it.

Dr. F. C. Fraser of the British Museum (Natural History) was kind enough

to read and comment on this 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.

REFERENCES

FLtower, W. H. 1885. List of the specimens of Cetacea in the zoological department of the British Museum.

London: British Museum.

Fraser, F. C. 1950. Description of a dolphin Stenella frontalis (Cuvier) from the coast of French

Equatorial Africa. Atlantide Rep. 1: 61-84.

A DOLPHIN (STENELLA ATTENUATA) FROM DURBAN, SOUTH AFRICA 135

Fraser, F. C. 1966. Comments on the Delphinoidea. Jn Norris, K. S., ed. Whales, dolphins and

porpoises: 7-31. Berkeley; Los Angeles: University of California Press.

Gray, J. E. 1846. On the cetaceous animals. Jn Richardson, J. & Gray, J. E.,eds. The zoology

of the voyage of H.M.S. Erebus and Terror. 1: 13-53. London.

Gray, J. E. 1866. Notices of a new genus of delphinoid whales from the Cape of Good Hope, and

of other cetaceans from the same seas. Proc. zool. Soc. Lond. 1865: 522-520.

NisHtwakt, M., Nakajima, M. & Kamrya, T. 1965. A rare species of dolphin (Stenella attenuata)

from Arari, Japan. Scient. Rep. Whales Res. Inst., Tokyo 19: 53-64.

Oxsumt, S. 1964. Comparison of maturity and accumulation rate of corpora albicantia between

the left and right ovaries in Cetacea. Scient. Rep. Whales Res. Inst., Tokyo 18: 123-148.

TRuE, F. W. 1889. Contributions to the natural history of the cetaceans. A review of the family

Delphinidae. Bull. U.S. natn. Mus. 36: 1-191.

Ann. S. Afr. Mus., Vol. 52° Plate 11

a. Skull, dorsal view. b. Skull, lateral view. c. Skull, ventral view.

Ann. S. Afr. Mus., Vol. 52 Plate 12

a. Mandibles, lateral view, external aspect. b. Mandibles, lateral view, internal aspect.

c. Mandibles, dorsal view.

Ann. S. Afr. Mus., Vol. 52 Plate 13

a. Cervical vertebrae, dorsal view. b. Cervical vertebrae, lateral view. c. Cervical vertebrae,

ventral view.

Ann. S. Afr. Mus., Vol. 52 Plate 14

Bererrrevect tate ty ety reys

a. Thoracic vertebrae, lateral view. 6. Lumbar vertebrae, lateral view. c. Caudal vertebrae,

lateral view. d. Chevron bones.

Ann. S. Afr. Mus., Vol. 52 Plate 15

Ribs, sternum and sternal ribs.

Ann. S. Afr. Mus., Vol. 52 Plate 16

a. Hyoid bones. b. Pelvic bones. c. X-ray photograph of right flipper.

Ann. S. Afr. Mus., Vol. 52 Plate 17

a. Scapulae. 6. Left and right flippers.

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, 6, 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)

BuLttoucuH, 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. Polyphacophora, 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.

;

tS as

——— ee ee

nants int vr

eee ‘apn rr

sre Ga

dear are same

pantie ae ke

apne

/-GO

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 52 ~#Band

March 1969 Maart

Part... 0, Deel

THE RELATIONSHIP BETWEEN RAJA MIRALETUS

LINNAEUS AND RAJA OCELLIFERA REGAN BASED

ON A STUDY OF THE CLASPER

Pp. A. HPUELEY

Cape Town Kaapstad

HSOWN

eet ‘4,

JUL 10 1969

Pe Ue Le ae

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

(Cash with order, post free)

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

(Kontant met bestelling, posvry)

OUT OF PRINT/UIT DRUK

I, 2(1, 3, 5, 7-8), 3(1-2, 5, tpl), 5(2, 5, 7-9)s

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

11(1-2, 7> teats) 21, 24(2), 27> Sui(t=3)) 38,

44(4).

Price of this part/Prys van hierdie deel

R1.50

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 RELATIONSHIP BETWEEN

RAJA MIRALETUS LINNAEUS AND RAJA OCELLIFERA REGAN

BASED ON A STUDY OF THE CLASPER

P. A. Huey

South African Museum, Cape Town

(With 3 figures)

[MS. received 27 September, 1967]

CONTENTS

PAGE

Introduction i OHeN a ac) nds, «Tz

Clasperistmucturemy ):) ) . s. .130

External characters .. .... «144

Discussion . 2 Said LS ERRORS (1

Summary . : : ; : . 146

Acknowledgements . . . . 146

References ITE Od | eR EINE 6

INTRODUCTION

Bigelow & Schroeder (1953) have pointed out that several South African

skates seem to be indistinguishable from European species from the correspond-

ing latitudinal belt in the northern hemisphere. These species are Raja pullo-

punctata Smith, 1964 and Raja batis Linnaeus, 1758; Raja rhizacanthus Regan,

1906 and Raja clavata Linnaeus, 1758; and Raja alba Lacépéde, 1803, which

is recorded as such from both regions. These pairs seem to represent cases of

masked bipolarity (Ekman, 1953). Besides the brief references given by Norman

(1935), little work has been done on the comparison of the specimens from the

two regions. On the basis of the clasper structure of mature males, Hulley

(1966) has shown that R. rhizacanthus Regan is synonymous with R. clavata

Linnaeus, while R. pullopunctata Smith is distinct from R. batis Linnaeus.

A more interesting relationship is shown by the pair R. miraletus Linnaeus,

1758 and R. ocellifera Regan, 1906. R. miraletus (fig. 1) has been recorded from

the Mediterranean (Italy, Adriatic, Nice, Tunis, Tripoli: Clark, 1926), and

this species extends southwards along the north-west African and west African

coasts, being recorded from Morocco (Clark, 1926), Cape Bojador (Murray &

Hyjort, 1912), Rio de Oro, Cape Blanc to Cape Verde (Pellegrin, 1914), Senegal

(Cadenat, 1950) and from Cape Lopez in Gabon to Baie les Tigres in Angola

(Fowler, 1936; Poll, 1951). The species R. ocellifera is known to extend from

137

Ann. S. Afr. Mus. 52 (6), 1969: 137-147, 3 figs

138 ANNALS OF THE SOUTH AFRICAN MUSEUM

@ R. miraletus a

@ R. ocellifera

Fig. 1

Map showing the distribution of R. miraletus and R. ocellifera.

False Bay and the Agulhas Bank northwards to Natal (Barnard, 1925; Smith,

1961). It has not been recorded along the west coasts of South Africa and South

West Africa. R. ocellifera has also been reported from the Kerala coast, in

southern India (Samuel, 1963).

Norman (1935) has pointed out that R. ocellifera from South African waters

is closely related to R. miraletus from the Mediterranean and the west coast of

Africa, so much so that specimens taken from Cape St. Blaize, near Mossel Bay,

have been identified as R. miraletus by Boulenger (Von Bonde & Swart, 1923)

RAJA MIRALETUS LINNAEUS AND RAJA OCELLIFERA REGAN 139

and as R. ocellifera by Norman (1935). Furthermore, specimens obtained by

both the Discovery and by the Belgian South Atlantic Expedition off the Congo

and Angola, approximate to the South African form, but have been referred to

R. miraletus on the basis of a somewhat longer tail, a longer snout, narrower

interorbital distance, fewer teeth in the upper jaw, and a circular ocellus.

The genus Raja, however, shows great variability in its morphological

characters, which vary with age and/or sex in the individuals within a species,

so that a more critical method of examination needs to be employed. The

purpose of this paper is to re-examine the relationship between R. miraletus and

R. ocellifera in the light of the structure of the claspers of mature males.

Leigh-Sharpe (1920-6) has shown that the species of the genus Raja may

be recognized by the morphological structure of their claspers, and Ishiyama

(1958) has pointed out that the clasper structure is the most reliable basis for

the systematics of this group, since both the external and internal structures of

the claspers are species-specific without exception. On this basis, Hulley (1966)

has employed the structure of the clasper for comparing geographically separa-

ted species. However, differences in the external and measurable characters of

the species must also be taken into account.

CLASPER STRUCTURE

Comprehensive anatomical studies have been carried out on the claspers

of European and Japanese rajids, so that the nomenclature used in this paper

is in accordance with Petri (1878), Jungersen (1899), Huber (1901), Leigh-

Sharpe (1920-6) and Ishiyama (1958).

The claspers of three adult specimens of R. ocellifera, trawled in Algoa Bay

and Jeffrey’s Bay, one adult specimen of R. miraletus obtained by the Belgian

South Atlantic Expedition (Station 15: 05°50’ S, 11° 32’ E, WNW Banana,

in 210 metres), and two specimens of R. miraletus from the north-west coast of

Africa (Walther Herwig Station 125/1964; 08° 32’ N, 14° 09’ W, in 44-48

metres) were examined. Although the latter two specimens were small and

had only just reached maturity, the external characters of the claspers could be

recognized.

External anatomy

The shape of the claspers is the same in both R. miraletus and R. ocellifera

(fig. 2A). The clasper is comparatively short: proximally, it is cylindrical, but

from about two-thirds its length, it becomes dorso-ventrally flattened, especially

on the ventral surface. It tapers gradually towards its distal end and terminates

in a pointed, spatulate tip. The ventral surface is pale, while the dorsal surface

of the clasper tones with the general coloration of the upper surface of the disc.

The apopyle is situated dorsally, some little distance behind the vent,

and the appendix groove arising at this point runs laterally outwards towards

the margin of the clasper, before turning to run parallel with this margin,

almost to the distal end of the organ (fig. 2A).

140 ANNALS OF THE SOUTH AFRICAN MUSEUM

ap.

2.0 =|

cm.

| |

2.0cm_

2.0 cm 1

hp. hp.

sl. sh. si. sh,

Si. Si,

sl}.

sl. rh. rh.

st. se. st.

Fig. 2

A. External view of the right clasper from the dorsal side.

B. R. ocellifera: \ateral view of the right clasper opened to show the structural features of the glans.

C. R. miraletus: lateral view of the right clasper opened to show the structural features of the glans.

ap.—apopyle; cm.—pecten; hp.—hypopyle; rh.—rhipidion; se.-sentina; sh.—shield; si.—signal; sl.—slit; st._sentinel

_ Adjacent to the groove and situated at a point about midway along the

length of the clasper, there is a row of four to six small, sharp protuberances,

the pecten. This is the result of the projection of the outer lateral margin of the

dorsal terminal 2 cartilage (fig. 3B) through the skin of the clasper. Although

the pecten is present in both R. ocellifera and the Angolan specimen of R.

miraletus, it was found to be absent in the two Walther Herwig specimens. This

is probably due to the fact that these specimens are only just mature and no

calcification of the clasper cartilages has taken place.

Apart from a difference in the size of the sentinel, the structures of the

clasper glans were the same in all specimens examined (figs 2B, 2C). It should

be noted that the slight differences in these figures is due to the fact that the

clasper has been more fully opened in R. ocellifera.

RAJA MIRALETUS LINNAEUS AND. RAJA OCELLIFERA REGAN 141

The spade-shaped sentinel is strongly developed and is situated at the

distal end of the glans, slightly dorsal to the line of the appendix groove. It is

covered by epithelium, but its outer lateral margin is knife-like and exposed.

The spike lies ventrally just beneath the sentinel, which completely obscures

it from sight. A sentina is present. The fleshy pad, situated in the mid-region

of the organ and ventral to the appendix groove, is the signal. This is capable

of some rotational movement about the longitudinal axis of the clasper. The

fleshy pad extends distally as a plate-like structure, the rhipidion. The ventral

shield is narrow, but is well developed, especially on its inner lateral margin,

where it is somewhat raised into a longitudinal ridge. The shield extends from

the proximal region to about three-quarters the length of the glans. In mature

specimens, the shield is covered by pleated epithelium (figs 2B, 2C), but this is

so thin along the outer lateral margin that the sharp edge of the ventral ter-

minal cartilage is exposed. Two pouches, the slits, are present in the dorsal

surface of the concavity of the clasper glans. They are shallow depressions, but

are well demarcated from each other by a cartilaginous shelf. The proximal

slit has a raised inner margin, due to the prominence of the axial cartilage in

this region.

Internal anatomy

The following descriptions and relevant diagram (fig. 3) are based on the

dissection of the claspers of R. ocellifera.

Axial, dorsal marginal and ventral marginal cartilages (figs 3A—E)

The axial cartilage is stout and cylindrical at its junction with the b,-

cartilage, but it becomes broader and dorso-ventrally flattened from about

half of its length to the distal end, terminating in a broadly rounded tip. On

the ventral surface of the axial cartilage, there is a short, shallow groove, which

runs from almost the flat tip of this cartilage to about one-quarter its length

from the distal end (fig. 3E). This groove accommodates the accessory 2

cartilage.

The marginal cartilages are more heavily calcified than the axial and can

easily be distinguished from it. The marginal cartilages border the appendix

groove dorsally and ventrally, and are somewhat scrolled. The origin of the

dorsal marginal cartilage is more proximal than that of the ventral marginal,

so that the interspace between the two marks the position of the apopyle. The

dorsal marginal extends proximally for about half the length of the axial, and is

tightly bonded to this cartilage, except at its distal end. Here the dorsal marginal

turns laterally outwards and terminates in a blunt point (fig. 3A). At a point

_about two-thirds along its length, the dorsal marginal possesses a step, into

which fits the dorsal terminal 2 cartilage (fig. 3A).

_ The ventral marginal cartilage is bonded to the axial along the whole of

its inner lateral margin. It is ventrally concave, so that its outer lateral margin

is raised into a thin ridge, which runs along the whole of the appendix groove.

Distally, the ventral marginal cartilage is obliquely truncate.

142 ANNALS OF THE SOUTH AFRICAN MUSEUM

: eat ngs es

d.mg.

a.tr.1

ax.

| 2.0 cm |

E H

Fig. 3

R. ocellifera: cartilages of the terminal group of the clasper.

A. axial and dorsal marginal (dorsal view) ; B. dorsal terminal 2 (dorsai view) ; C. dorsal terminal

2 cartilage in situ (dorsal view) ; D. dorsal terminal 1 (dorsal view) ; E. axial and ventral marginal

(ventral view); F. accessory terminal 1 (ventral view); G. accessory terminal 2 (ventral

view); H. accessory 1 and 2 cartilages in situ (ventral view); I. ventral terminal (dorsal view) ;

J. ventral terminal (ventral view) ; K. accessory terminal 3 (dorsal view).

ax.—axial; d.mg.—dorsal marginal; v.mg.—ventral marginal; d.tr.2-dorsal terminal 2; a.tr.1—

accessory terminal 1; a.tr.2—accessory terminal 2.

RAJA MIRALETUS LINNAEUS AND RAJA OCELLIFERA REGAN 143

Dorsal terminal 1 cartilage (fig. 3D)

This is a flatly-rounded, shield-like cartilage occurring on the dorsal side

of the clasper. It is curved proximally and elongated into a blunt point along

its outer lateral margin. This outer lateral margin fits into the slight groove in

the dorsal terminal 2 cartilage (fig. 3B). From this point, and running in a

curve, is a slight shelf, which serves as the area of insertion of the m. dilatator.

Distally, the cartilage is somewhat pointed on its inner lateral margin. This is

slightly curved and wraps around the body of the clasper.

Dorsal terminal 2 cartilage (figs 3B, 3C)

The dorsal terminal 2 cartilage is dorsally convex, with a broadly rounded

anterior end. This end fits into the step of the dorsal marginal cartilage (fig. 3C).

Along the distal edge of its outer lateral margin, the dorsal terminal 2 is serrate,

the serrations being raised into small, sharp points. These are associated with

the development of the pecten. Distal to the dorsal terminal 2 cartilage and in

the covering of the connective tissue, there is a cartilaginous ‘rib’, which joins

the axial at about two-thirds along its length (fig. 3C). This rib forms the shelf

which separates the two dorsal slits.

Ventral terminal cartilage (figs 31, 3J)

This is a J-shaped cartilage, which is attached to the distal end of the ventral

marginal cartilage. It is somewhat curved medially, so that it runs around the

inner lateral side of the axial cartilage. The ventral terminal is bluntly rounded

anteriorly, but posteriorly and distally there is a small point. The elongated

region of the cartilage possesses a shelf and a broad ridge on its dorsal side,

while ventrally the cartilage is slightly concave. The shelf on the dorsal side,

which is associated with the shield, is thin and has a knife-like outer lateral edge.

Accessory 1 cartilage (figs 3F, 3H)

The cartilage is S-shaped, with its distal end curved laterally outwards

and expanded into a thin, flat plate. This is the sentinel of the clasper glans.

The proximal end of the accessory 1 cartilage is dorso-ventrally flattened and

terminates in a blunt point. This point and the adjacent portions are attached

to the distal end of the ventral marginal cartilage. There is a slight ridge along

the inner lateral side of the accessory 1 cartilage. This fits into a groove on the

outer lateral edge of accessory 2.

Accessory 2 cartilage (figs 3G, 3H)

This cartilage is attached to the inner, truncate edge of the ventral marginal

cartilage (fig. 3H). It has a broad anterior end and is more or less cylindrical in

shape. It curves laterally outwards to form a hook at its distal end. This hook is

the spike of the clasper glans. Along its outer lateral edge, the accessory 2

cartilage possesses a slight groove, into which fits the accessory 1 cartilage (fig

3H).

Accessory 3 cartilage (fig. 3K)

The accessory 3 cartilage consists of a single cartilaginous element, which

manifests itself in the clasper glans as the signal. Ventrally, this cartilage is flat,

144 ANNALS OF THE SOUTH AFRICAN MUSEUM

but dorsally, it is raised into an anteriorly directed ridge. The accessory 3 is

attached to the distal end of the ventral marginal cartilage, more ventrally

than the attachment of the accessory 1 and 2 cartilages.

The cartilaginous elements of the clasper glans of R. miraletus were examined

by X-ray photography. The structure of these elements proved to be identical

to that of R. ocellifera, as given above.

EXTERNAL CHARACTERS

While the clasper structure may be used as a means of comparison between

geographically separated species, it is as well to examine the external and

measurable characters of the species concerned. As already pointed out, South

African specimens have been referred to R. ocellifera Regan rather than to

R. miraletus Linnaeus, because of differences in the length of the tail, the length

of the snout, the interorbital distance, the number of rows of teeth in the upper

jaw and the nature of the ocellus.

‘Table 1: A comparison of the external and measurable characters of specimens of R. miraletus

from the Mediterranean, Sierra Leone and Angola, with specimens of R. ocellifera from

South Africa.

R. miraletus R. ocellifera

Mediterranean Sierra Leone Angola South Africa

Tail length in disc

width nei I°l vr I*2-1°4

Middle of vent to ist

dorsal origin in Not given 1-2-1913 Ei We I*I-1°4

snout to middle of

vent

Snout length in disc 5°0-5°7 5°0-5°2 5°5-5°6 5°1-6°1

width

Interorbit in disc width 22-0-24-0 16°6-18°5 16-8-20°4 14°O-17°1

Interorbit in snout 4.°2-4.°4 3° Bon 3°5-3°6 PIPER N,

length

Interorbit in internasal PAO) 5 1G 5 2/0 1°$-2°3 1-7

No. of teeth in upper

_ jaw 40-42 45-47 47-52" 42-50

No. of precaudal

vertebrae Not given 49-51 52-54. 49-53

*Poll (1951)

The following comparison (table 1) of these characters is based on 24

specimens of R. ocellifera (disc width 84-354 mm) taken in South African waters

from False Bay to Algoa Bay; 2 specimens of R. miraletus (disc width 305,

343 mm) trawled off the coast of Angola; and 2 specimens of R. miraletus (disc

width 196, 204mm) taken off the coast of Sierra Leone (Walther Herwig Station

125/1964). Measurements given by Clark (1926) for Mediterranean R. miraletus

have also been employed.

As has been pointed out (Norman, 1935), there is a difference in the

RAJA MIRALETUS LINNAEUS AND RAJA OCELLIFERA REGAN T45

ocellus. In R. miraletus, the ocellus is circular in shape, while in R. ocellifera, it

tends to be horizontally ovate. The ocellus of R. miraletus consists of a number of

concentric rings of colour: a blue-black centre spot, surrounded by a dark

black ring, which is in turn surrounded by a complete brownish-yellow outer

ring. In R. ocellifera, the outer brownish-yellow ring is usually incomplete and

consists of a number of spots of the colour, which are narrowly united.

Discussion

The structural features and lay-out of the clasper glans are the same in both

R. miraletus and R. ocellifera. The two dorsal slits are shallow in both species and

the sentinel completely covers the spike. The slight difference in the size of the

sentinel is probably due to an age difference, since the shape of the sentinel and

the nature of the corresponding cartilage (accesory 1) are the same in both

species. Furthermore, the presence of the comb confirms that the claspers of

R. ocellifera are identical to those of R. miraletus.

Since the cartilaginous elements of the claspers are directly related to the

external characters of the glans, it would be expected that these would be

similar. This is true at least for the accessory 1 and 2 cartilages, where the

author is unable to detect any appreciable difference. As these cartilages are

species-specific without exception (Ishiyama, 1958), it is reasonable to assume

that the species R. ocellifera Regan is synonymous with the species R. miraletus

Linnaeus.

There are differences in the external and measurable characters of the

two ‘species’ (table 1). The tail is comparatively shorter in R. ocellifera than in

R. miraletus, as shown by the measurement of the tail length in the disc width.

In R. ocellifera this measurement varies from 1-2 to 1:4, while in R. miraletus,

it is more or less constant at 1-1. However, Ishiyama (1952) has pointed out

that a more reliable method for comparing tail length is the determination of

the number of precaudal vertebrae. These counts are not always useful in the

separation of species. In R. ocellifera, the count is from 49 to 53, and in R.

miraletus, from 49 to 54, so that the number of precaudal elements in R. ocellifera

is within the range for R. miraletus. This is confirmed externally by the measure-

ment of the distance from the middle of the vent to the 1st dorsal origin (R.

ocellifera 1-1-1-4; R. miraletus 1- 1-1-3).

The snout appears to be somewhat longer in R. miraletus. It would seem

from table 1 that there is a slight decrease in the snout length from the Mediter-

ranean specimens to the west African specimens of R. mzraletus. Similarly, the

interorbital distance appears to be less in R. miraletus. These two differences are

particularly noticeable in the measurement of the interorbital distance in the

snout length.

The author is unable to detect any difference in the number of teeth in

the upper jaw, since the range for R. miraletus covers that of R. ocellifera. The

most easily distinguishable difference between the two ‘species’ is the nature of

146 ANNALS OF THE SOUTH AFRICAN MUSEUM

the ocellus. This has been shown above. However, the systematic importance

of such a character is unknown.

It is therefore concluded that these external and measureable differences

are probably of secondary importance in the systematics of this group, and that

since the structures of the claspers are identical, and the number of precaudal

vertebrae is similar, R. ocellifera and R. miraletus should be regarded as two

extremes of one varietal species.

This being the case, it is interesting to note the distribution of the species.

It is present along the whole of the west African coast, from the Mediterranean

to the southern border of Angola, absent from the region Walvis Bay to Cape

Point, and again present along the east coast of South Africa. It is suggested

that the cold water of the Benguela Current, in the region Walvis Bay to Cape

Point, acts as a temperature barrier. Since some gene flow must occur from

west to east and vice versa, the species probably moves off-shore. A similar

occurrence is found in R. clavata (Hulley, 1966: fig. 8), but in this temperate

species, it is warmer water which acts as the barrier.

SUMMARY

After detailed studies of the anatomy of the claspers of the males, and

consideration of the external features, it is concluded that Raja ocellifera Regan

is synonymous with Raja miraletus Linnaeus.

ACKNOWLEDGEMENTS

My thanks are due to Dr. G. Krefft of the Institut fiir Seefischerei in

Hamburg and Mr. J. P. Gosse of the Institut Royal des Sciences naturelles de

Belgique for the loan of the R. miraletus specimens, and to the late Prof. J. L. B.

Smith of Rhodes University for specimens of R. ocellifera. 1 am also grateful to

Mrs. E. Cluver and Mr. 8S. X. Kannemeyer of the South African Museum for

assistance with the X-ray photographs.

The Trustees of the South African Museum thank the Council for Scientific

and Industrial Research for a grant in aid of publication.

REFERENCES

BARNARD, K. H. 1925. A monograph of the marine fishes of South Africa. Ann. S. Afr. Mus. 21:

1-418.

BicELow, H. B. & ScHROEDER, W. C. 1953. Fishes of the western North Atlantic. Pt. II. Saw-

fishes, guitarfishes, skates and rays. Chimaeroids. Mem. Sears Fdn mar. Res. 1 (2) : i-xv, 1-588.

CADENAT, J. 1950. Poissons de mer du Sénégal. Init. afr. 3: 1-345.

Ciark, R. S. 1926. Rays and skates, a revision of the European species. Scient. Invest. Fishery Bd

Scotl. 1926: 1-66.

Ekman, S. 1953. Coogeography of the sea. London: Sidgwick & Jackson.

Fow er, H. W. 1936. The marine fishes of west Africa based on the collection of the American

Museum Congo Expedition, 1909-1915. Bull. Am. Mus. nat. Hist. 70: 1-1493.

Huser, O. 1901. Die Kopulationsglieder der Selachier. Z. wiss. Zool. 70: 592-674.

RAJA MIRALETUS LINNAEUS AND RAJA OCELLIFERA REGAN 147

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.

IsHtyAMA, R. 1952. A revision of three genera of Japanese rajids, with descriptions of one new

genus and four new species mostly occurring in northern Japan. 7. Shimonoseki Coll. Fish. 2:

I-34.

IsHtyAMA, R. 1958. Studies on the rajid fishes (Rajidae) found in the waters around Japan.

J. Shimonoseki Coll. Fish. 72 193-394.

JunceErsen, H. F. E. 1899. On the appendices genitales in the Greenland shark Somniosus micro-

cephalus and other selachians. Dan. Ingolf-Exped. 2 (2) : 1-88.

LeicH-SHARPE, W. H. 1920-6. The comparative morphology of the secondary sexual characters

of elasmobranch fishes. Mems I-XI. 7. Morph. 34: 254-265 (1920); 35: 263-358 (1921);

36: 191-244 (1922); 39? 553-577 (1924); 42 307-368 (1926).

Murray, J. & Hjort, J. 1912. The depths of the ocean. London: Macmillan.

Norman, J. R. 1935. Coast fishes. Pt. I. The South Atlantic. ‘Discovery’ Rep. 12: 3-58.

PELLEGRIN, J. 1914. Missions Gruvel sur la c6éte occidentale d’Afrique. Annis Inst. océanogr.,

Monaco 6 (4): 1-100.

Petri, K. R. 1878. Die Copulations-organe der Plagiostomen. Z. wiss. Zool. 30: 288-335.

Poti, M. 1951. Poissons. II. Sélachiens et chimeéres. Résult. scient. Exped. océanogr. belge Atlant.

sud. 4 (1): 12-154.

SAMUEL, C. T. 1963. Bottom fishes collected by R. V. ‘Conch’ off the Kerala coast. Bull. Dep.

mar. Biol. Oceanogr. Univ. Kerala 1: 97-121.

SmiTH, J. L. B. 1961. The sea fishes of southern Africa. 4th ed. Cape Town: C.N.A.

Von Bonne, C. & Swart, D. B. 1923. The Platosomia (skates and rays) collected by the SS.

‘Pickle’. Rep. Fish. mar. biol. Surv. Un. S. Afr. 3 (5): 1-22.

i tb vaya:

MeN sani! ae

i ye iy

| rf ‘ 2

Peay

| > : tos ;

% Pie V

i} A y i

% Ae ithe SW

an iM Ry

if 5 bs iy

7 fo % ey f

y a i ; rf ; y ¥ ih i An

| Phy tes )

i iy ts Vinal "

¢ é i ¥ 4 } far \

1 } ) ih Re th | Hiei :

) bese iar a

A Dirt i "

| | ee |

L) | / ij y i }

i A (Abe aes id pate {

i SM a's oe { se te ee

: , : trRD aa

ii ‘ A : t

f | v: : 4 ry

{

| { '

| i

f r A

‘ Wh /

: ‘

f \

‘| . { i

f /

ta!

i ’

i} L

} * \

} (

Ds ~

| ; ) \ i Mien

ae } - ae) I

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 43 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)

BuLLoucu, 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. Polyphacophora, Gastropoda marina, Bivalvia. In Schultze, L.

Koologische und anthrepologische 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

ITEC PIL, aa

ast t Nit ba

} q fou ae

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 52 ~ ~°&Band

April 1969 April

Party (sory Deel

DECAPOD CRUSTACEA FROM THE

SOUTH-WEST INDIAN OCEAN

B. F. KENSLEY

Cape Town Kaapstad

ie 2.0 1969

Ligrarie2

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

iy Alls By D> 7-8), BiB Fy PSOE) BR a) 8)

6(1, t.—p.i.), 7(1, Me 8, g(1-2), 10(1-3),

11(1-2, 7, t.—p.1.), 21, 24(2), 27,° 31(1=3), 36,

44(4).

Price of this part/Prys van hierdie deel

R2.10

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

DECAPOD CRUSTACEA FROM THE

SOUTH-WEST INDIAN OCEAN

B. F. KENsLEY

South African Museum, Cape Town

(With 16 figures)

[MS. received 30 Fuly 1968]

CONTENTS

PAGE

Introduction Se iS iowa) FAQ

Station Hist oye 4s) > ss (150

Species Sty. Bo We 8S 4 IGE

Systematic discussion Bidens EGG

Distibution -Capsa 6. eo. FG

SumiMaryy Gay en? os lS 896

Acknowledgements . . . 179

eterenees. = We) 29 te 3. (ISO

INTRODUCTION

The material dealt with in this paper comes from several sources. The

greatest proportion was collected on the seventh cruise of the R/V Anton Bruun,

in 1964, as part of the International Indian Ocean Expedition. The station

numbers of the Anton Bruun are designated by the letters BRU, while catalogue

numbers of the Zoology Department, University of Cape Town, are designated

either as NAD (off the Natal coast) PED (off Mocambique coast), MDD (off

the south-western coast of Malagasy Republic), or WBS (Walter’s Shoal).

This latter is a shallow area about 400 nautical miles south of the Malagasy

Republic, and about 600 miles off the coast of South Africa. A final report of

the expedition was published by the U.S. Program in Biology, I.1.O.E., in

1965, to which body I am indebted for the opportunity of examining the

material.

Other sources of material include the collections made by the R/V Vema

(station Vema 6) off the Natal coast, the John D. Gilchrist (station GIL 56-58)

off Natal, and the S.A.S. Natal (station NGY) off Natal.

Type material is housed at the South African Museum, and catalogued

with South African Museum (S.A.M.) numbers. All other material is in the

Zoology Department of the University of Cape Town.

149

Ann. S. Afr. Mus. 52 (7), 1969, 149-181, 16 figs.

150 ANNALS OF THE SOUTH AFRICAN MUSEUM

Abbreviations

Sampling gear:

AD— Agassiz Dredge

CG— Campbell Grab 0.6 m?

Substrata:

c— coarse

Calc. Alg.—Calcareous Algae

Crl—coral

d—dark

f—fine

G— gravel

gn—green

Station Number Date

GIL 56 13.7.59

GIL 58 13.7.59

NGY 21 17.5.58

NGY 17 16.8.58

NGY 59 12.8.58

NGY 63 13.8.58

Vema 6 23.4.58

BRU 356B _29.7.64

BRU 356] 29.7.64

BRU 357B __30.7.64

BRU 358C —_30.7.64

BRU 363P _5.8.64

BRU 363 W_ 12.8.64

BRU 370G_18.8.64

BRU 371E 18.8.64

BRU 371F 18.8.64

BRU 371G __19.8.64

BRU 372C 19.8.64

BRU 372G 19.8.64

BRU 372L _19.8.64

BRU 372Q 22.8.64

BRU 373B —_22.8.64

BRU 381 A-C 30.8.64

BRU 389G __ 8.9.64

BRU 390C 8.9.64

BRU 390E _—8..9.64

BRU 390G __8.9.64

BRU 390H 8.9.64

BRU 390L _9.9.64

BRU 390N 9.9.64

BRU 390P _9.9.64

BRU 390R _ 9.9.64

BRU 390S 9.9.64

BRU 391B _—‘9.9.64

BRU 391C 9.9.64

BRU 391F 9.9.64

BRU 391H 9.9.64

BRU 391J —-9.9.64

BRU 392 F _10.9.64

BRU 392G 10.9.64

BRU 392H 10.9.64

BRU 392 K__10.9.64

MT— Menzies Trawl

RD—Rock Dredge

G.O.—Globigerina Ooze

Sy — grey

M—mud

R—rock

S—sand

Sh—shell

St—stone

STATION LIsT

Position Depth (m)

29°53’S/31°06’"E 20

29°53’S/31°04’E gu

30°47°S/30°29’E 44,

29°53’S/31°04’E 38

29°58’S/31°27'E: 49

30°47’S/30°27’E 36

29°46'S/31°17’E 110-130

29211735 / 31737 18

29°10’S/31°51’E 43

29°11’S/32°02’E 70

29°21°S/31°58’7E 370

23°17'S/43°33’E 12-25

23°19’S/43°36’E 82

24°40’S/35°28’E 347

24°46'S/35°20’E 132

24°46'S/35°18’E 110

24°53’S/34°56’E 55

24°46'S/24°50’E 22

24°53’S/34°56’E 55

25°07'S/34°34’E 2

25°57°S/33°02’E 42

26°00’S/33°05’E 135

33°13’S/43°51’E 38-46

29597 (S| Dao, 700

29°45’S/31°40’E 440

29°42’8/31°38’E 350

29°38’S/31°36’E 200

29°37 (S| ale aoees 175-200

29°39'S/31738 4 150

29°3475/3.1°39 FE: 115

29°34’S/31°39’E 118

29°35’S/31°42’E 138

29°29’S/31°45’E 86

29°29'S/31°45’E 86

29°26'S/31°46’E 7

29°21 S/31°357E 57

29°31’S/31°35’E 57

29° Gis) aloo 35

2954 7S Sirol Ei 18

29°13°9/31°314E 18

29°19"S/31°26°E, 38

Bottom Gear

Calc. Alg.

G DnNNn

(ous es

ADMNDRnRNRNnRN

Lo)

—

DECAPOD CRUSTACEA FROM THE SOUTH-WEST INDIAN OCEAN

* — new record for the southern African region

ovig — ovigerous

juv — juveniles

Species 3} Q

BRACHYURA |

Achaeopsis spinulosus Stimpson 1 3 ovig

Achaeopsis thomsoni 2 1 ovig

(Norman) 1 _

Achaeus laceriosus (Stimpson) 7 2 ovig

Achaeus cf. affinis Miers — 1

— 3 ovig

Actaea rueppellit (Krauss) ] ~

Calappa lophos (Herbst) l —

Carcinoplax longimanus — 1

(de Haan) 3 ---

Charybdis cf. annulata — 1 ovig

(Fabricius)

Conchoecetes artificiosus (Fabricius) 8 6

Dorippe lanata (Linnaeus) 1 3 ovig

1 4 ovig

Ebalia (Lithadia) barnardi

Stebbing 1 —

Ebalia (Ebalia) tuberculata Miers 4 2 ovig

1 fie

2 1

Ebalia (Ebalia)tuberculosa f.

postulans (Stebbing) 1 —

Ebalia (Ebalia) tuberculosa f.

scandens Stebbing 3 2

l l

5 1 ovig

4 pa

Ebalia sp. 1 —~-

* Ethusa sinespina n.sp. — 1 ovig

~ 1

1 Bie

Eumedonus granulosus

MacGilchrist 1 —

Eurynome aspera (Pennant) 8) 1 ovig

Goneplax angulata (Pennant) 5 3

Gonioneptunus africanus (Shen) 10 17+2 ovig

2 1 ovig

Homola barbata (Fabricius) — 1

Hyastenus spinosus Milne-

Edwards 4 1 ovig

D) ob

Inachus cf. dorsettensis (Pennant) 1 —

Inachus guentheri (Miers) 2 1 ovig

1 3 ovig

Inachus sp. — —

Leucosia marmorea Bell — 1 ovig

Species List

15]

Catalogue number

NAD 11B

PED 20 Y-Z

PED 7B

PED 16N

PED 16 N

NAD 3 Z

NAD 3 X

MDD 1 D

NAD 31 C

PED 8 Y

NAD 8 L

NAD 62 L

NAD 51 E

NAD 35 E

NAD 26 R

NAD 55 N-Q

NAD 11 E

NAD 45 N

NAD 45 L

ABD 1C

PED 8 V

NAD 14M

ABD 15 K

ABD 8 Z

ABD 1B

ABD 15 L

S.A.M. A12648

S.A.M. A12649

NAD 45 K

PED 16 Z

NAD 38

NAD 31 E

NAD 35 D

PED 19N

WSS 3 D

NAD 62 B

PED 7C

ABD YA

PED 7C

NAD 11B

ABD 9A

NAD 62 K

152 ANNALS OF THE SOUTH AFRICAN MUSEUM

Lophozozymus dodone (Herbst)

Lupocyclus tugelae Barnard

Macropodia formosa Rathbun

* Nursilia dentata Bell

* Palicus sexlobatus n.sp.

* Paratergatis longimanus Sakai

Philyra globosa (Fabricius)

Philyra globulosa Milne-

Edwards

Pilumnus hirsutus Stimpson

Pilumnus longicornis Hilgendorf

Pinnotherids (Unidentified)

Platylambrus quemvis Stebbing

Platypodia cf. granulosa (Rippell)

Portumnus mcleay: Barnard

Ranina ranina (Linnaeus)

* Retropluma planiforma n.sp.

Thalamita woodmasoni Alcock

Thalamita sp.

Xanthias tuberculidens Rathbun

? Xanthias sp. (Immature)

Xanthids (Unidentified)

ANOMURA

PAGURIDEA

Anapagurus hendersoni Barnard

Dardanus arrosor (Herbst)

Dardanus euopsis (Dana)

Dardanus setifer (Milne-

Edwards)

eee |

i imee | eno |

aah | see es

7+2 ovig

1 ovig

| |

Station

— | | ef

Catalogue Number

NAD 3 W

PED 16 K

NAD 18 T

PED T9)a5

WSS 3E

PED 6 Z

S.A.M. A12642

PED 19M

NAD 51 C

NAD 40 K

NAD 73 B

NAD 51 A

NAD 86 D

NAD 75 E

PED 12 Q-R

PED 16 F-G

PED 8 X

NAD 80 X-Y

NAD 3 Y

MDD 1 C

PED 19K

NAD 45 J

PED 16 H

Sf NAD 62 A

NAD 18 R

WSS 3 G

PED 12 N

NAD 64 W

PED 16 E

NAD 31 D

S.A.M. A12643

NAD 3R

NAD 35 G

NAD 45 R

ABD 14 L

NAD 62 T

NAD 268

PED 16 R-T

NAD 20 V

NAD 3 N

DECAPOD CRUSTACEA FROM THE SOUTH-WEST INDIAN OCEAN 153

3 2 Juv Station Catalogue Number

Diogenes brevirostris Stimpson I 1 — SIZE PED 12S

Diogenes costatus Henderson 2 — — NGY 59 NAD 18 Y

(356 J NAD 86 A

| 392 H NAD 73 K

SU 2S | NAD 75 H

many 391 J NAD 628

391 F NAD 58 J

391 C NAD 51 N

(391 H NAD 67 C

Diogenes custos (Fabricius) 5 2 — 392 H NAD 73 J

* Nematopagurus gardineri Alcock 1 --- — 390 S NAD 45 P

* Nematopagurus squamichelis

Alcock 2 1 — 370 G ABD 8 V

Pagurus spinulentus (Henderson) 5) 3 — 392 K NAD 75 J

GIL 56 NAD 24 D

| NGY 59 NAD 18 V

1 — — 356 J NAD 87 W

13 7 ovig — 391 J NAD 62 N

Pagurus sp. 1 — — 372 C PED 12 V

1 — — 390 S NAD 45 Q

— —- 1 370 G ABD 8 W

— 2 ovig — 372 G PED 16 R-T

Between Pagurus & Pylopagurus 1 — — 372 G PED 16 R-T

Parapagurus pilosimanus Smith 2 1 — 358 C ABD 1E

? Pylopagurus sp. 1 —— — 371 E PED 2G

GALATHEIDEA

Galathea dispersa Bate Ze ] — 357 B NAD 208

— 1 ovig — 371G PED 8U

2 1 ovig — 372 L PED 19 L

— — 1 371k PED 23 D

Galathea intermedia Liljeborg 5 5 — NGY 59 NAD 208

— — 2 390 G NAD 33 N

— 1 — NGY 21 NAD 3Q

l 4 -— 372 G PED 17 W

Munida sanctipauli Henderson — 1 ovig — 373 B PED 23 Z

— 1 oo 389 G ABD 13 L

Munida semoni Ortmann 1 1+ 1 ovig 1 372 L PED 19Q

— 1 ovig — 373 B PED 23 C

— 1+2 ovig — Sil 1s PED 6 X

2 1 | — 372 G PED 16C

10 12 — 390 P NAD 40 L

— 2 — S91 NAD 51G

7) 2 — 390 S NAD 45 B

11 6 — 390 H NAD 31 A

Munida cf. semoni Ortmann — 1 ovig — STE PED 6 X

*Petrolisthes militaris (Heller) 5 3) — 372 G PED 16A

1 — ABD 82 M

Porcellana dehaanii Krauss — — 2 356 B NAD 80 Z

Porcellana streptocheles Stimpson 2 Ds — 357 B NAD 91 J

— 1 — 356 B NAD 82 M

3 — — NGY 21 NAD 3 P

— ~- l SD PED 127

4 6 — 372 G PED 16A

THALASSINIDEA

*Axius (Neaxius) sp. — — 357 B NAD 91 N

| -

1 — 356 B NAD 81 A

154 ANNALS OF THE SOUTH AFRICAN MUSEUM

3 Q Ju Station Catalogue Number

Callianassa sp. 1 — — 390 H NAD 31 B

Callianassa sp. — | — 390 H NAD 31 B

MACRURA

PENAEIDEA

Acetes erythraeus Nobili 5 3 2 NGY 21 NAD F

2 13 — Ce yas | NAD 73 F-H

-- 1 — BO7 KK NAD 75 N

*Gennadas propinquus Rathbun 1 —- — 563.P ABD5A

Macropetasma africana (Balss) l 10 — 392 NAD 73 E

Metapenaeopsis adamanensis

(Wood-Mason) — 1 — 390 S NAD 45°,

Metapenaeopsis cf. stebbingi Nobili) — — 1 392 H NAD 73 F-H

Parapenaeus fissurus (Bate) 1 — — 390 P NAD 40 R

—- l — S91/C NAD 51M

— — 1 390 L NAD 35 N

Penaeopsis rectacuta (Bate) — 1 —- 370 G ABD 8 T

Penaeus japonicus Bate 1 a -— 592i NAD 73 D

Sergestes prehensilis Bate 1 — _- 390 C ABD 14R

Solenocera africanum Stebbing 1 -- ~ 390.2 NAD 408

— 1 — 390 H NAD 31 G

Solenocera ? pectinata (Bate) 1 — — 390 P NAD 40 R

Solenocera sp. ?] —- as 390 L NAD 35 Q

CARIDEA

Alpheus frontalis Milne-Edwards 390 G NAD 33 K

* Alpheus nonalter n.sp. sev. sev. —- 390 H NAD 31 H,

S.A.M. A12650-1

sev. sev. — 390 P NAD 40 M

391 C NAD 51 H-J

l — — Si 2a PED 19 P

* Alpheus waltervadi n.sp. 4 2 ovig 6 381 A-C WSS 2 Y,

S.A.M. A12646-7

Alpheus sp. (damaged) 1 _ — 391 J NAD 62 V

1 = — NGY 59 NAD 20 Q

390'P NAD 40 N

392K NAD 75 P

390 N NAD 43 B-C |

Chlorotocus crassicornis (Costa) — l 3 390 L NAD 35 M |

— 1 — 390 C ABD 14 T

1 1 ovig — 390 P NAD 40 U |

— l — 372 L PED 19 Y-Z |

Eualus ctenifera (Barnard) 24 2 ovig —- 381 A-C WSS 2 Z

* Heterocarpus woodmasoni Alcock — | — 370 G ABD 8 U

Hippolysmata vittata Stimpson a 1 = 356 B NAD 81 B-E

Latreutes mucronatus (Stimpson) 5 3 ovig sev. 356 B NAD 81 B-E

. It -—- — 372 C PED, 12: U7

— — S12 PED 16 X-Y

Leptochela pugnax de Man os | —- 392 F NAD 72 P

Leptochela robusta Stimpson —- 1 ovig — 371 E PED 2 C

Nikotdes cf. danae Paulson ] l — 356 J NAD 86 F

* Oplophorus spinicauda Milne-

Edwards — il — 363 P ABD 5C

Periclimenes (Periclimenes) sp. ?3 o — 356 B NAD 81 B-E

Periclimenes sp. Pal — — 373 B PED 22 U

DECAPOD CRUSTACEA FROM THE SOUTH-WEST INDIAN OCEAN 155

3 2 Juv Station Catalogue Number

* Plesionika cf. acanthonotus

(Smith) 2 — — 390 P NAD 40 T

f 1 — 390 H NAD 31 F

Plesionika martia (Milne-

Edwards) — ?] — 392 Et NAD 73 F-H

2 — — 358 C ABD 1 D

Pontocaris cataphracta (Olivi) _- 1 ovig — S127 PED 19 W

1 1 ovig —- 372 G PED 16 U

1 = _ 391 F NAD 58 H

1 1 ovig —- Sw |] NAD 62 W

1 — ~ 390 P NAD 40 Q

3 — —— 391 C NAD 518

Poniocaris lacazei (Gourret) 2 — — 390 C ABD 14 P

— 1 -—— 390 L NAD 35 P

Processa austroafricana Barnard 2 4+5 ovig 1 390 L NAD 35 L

— 1 ovig — 390 P NAD 40 V

— — 1 356 B NAD 81 B-E

Processa sp. — 1 ovig — 372 G PED 16 X-Y

Stylodactylus bimaxillaris Bate — 1 ovig — 372 L PED 19 Y-Z

— — 1 390 P NAD 40 W

Synalpheus anisocheir Stebbing — 3+3 ovig 4 NGY 21 NAD 4B

?1 — — NGY 59 NAD 20 Q

Synalpheus jedanensis de Man — 1+2 ovig —— 372 G PED 16 V-W

Synalpheus cf. jedanensis de Man —- 1 — 357 B NAD 91M

Tozeuma armata (Paulson) ?1 — _- LINE: PED 8 Z

SYSTEMATIC DIscussION

BRACHYURA

Family Parthenopidae

Eumedonus granulosus MacGilchrist, 1905

, Fig. 1 a—b

Eumedonus granulosus MacGilchrist, 1905: 253. Rathbun, 1911: 259. Flipse, 1930: go. Barnard,

1954: 96.

Previous records: Amirante, Persian Gulf, Zanzibar, Delagoa Bay.

Material: 1 g, carapace length (including rostrum) 4 mm, carapace breadth

(including lateral spines) 3-8 mm. Station BRU 372 G. Depth, 55 metres.

Remarks: The present specimen appears to differ from the original description

of E. granulosus only in the degree of granulation of the carapace. From Mac-

Gilchrist’s description and figures (latter in Illustrations of the zoology of the

RILM.S. ‘Investigator’, 1907, plate 77, figs 2, 2a), it would seem that the whole

integument is granulated, whereas in the present specimen it is granulated

only in the antero-lateral region. The specimen recorded by Barnard (1954)

has the carapace granulated in the mid-region only. The carapace grooves of

the type are more distinct than in this specimen. Body proportions are similar.

Unfortunately, all the pereiopods and chelipeds are missing. The differences

mentioned may be due to the immaturity of the specimen, as the type measures

fe 5X BE 5 om.

I 56 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 1. Eumedonus granulatus MacGilchrist

a. Carapace. b. First pleopod, male.

Family Xanthidae

Paratergatis longimanus Sakai, 1965

Paratergatis longimanus Sakai, 1965a: 98, fig. 1; 19655: 128, fig. 16.

Previous records: Sagami Bay and near-by Japanese islands.

Material: 1 9, carapace length 12:0 mm, carapace breadth 19-0 mm. Station

BRU 372 L. Depth, 112 metres. 2 $34, carapace lengths 10-6 mm, 6-0 mm,

carapace breadths 17-0 mm, 7:0 mm. Station BRU 390 P. Depth 118 metres.

2 34, carapace lengths 12:0 mm, 5:0 mm, carapace breadths 19-0, 8-o mm.

Station BRU 391 C. Depth 86 metres.

Remarks: There can be no doubt that this is the same species as that recorded

by Sakai (19654, b). This would appear to be the first record of this monotypic

genus outside Japanese waters.

Family Palicidae

Cympolidae: Rathbun, 1918: 182. Sakai, 1939: 607.

Palicidae: Holthuis & Gottlieb, 1958: 104.

Palicus sexlobatus n.sp.

Fig. 2 a—e

Description: Carapace wider than long, dorsally convex, granular, with larger

scattered tubercles, margins crenulate. Fronto-orbital margin with 2 pairs of

spines, inner 2 more slender, longer than outer spines, set slightly lower than

latter. 3 supra-orbital teeth, innermost broadest, outer 2 acutely triangular.

External orbital tooth largest of 6 antero-lateral teeth, latter decreasing in size

posteriorly. Postero-lateral margin of carapace concave at origin of 5th pereio-

DECAPOD CRUSTACEA FROM THE SOUTH-WEST INDIAN OCEAN 157

pod. (Latter dorsal in position.) Posterior margin with 6 separated flattened

lobes. Gastric region tuberculate, with anterior row of 4 larger transverse

tubercles, 2 posterior transverse tubercles. Gastric region separated from cardiac

and branchial regions by well-defined grooves. Cardiac region with row of 4

transverse flattened tubercles, inner 2 larger. Branchial regions with scattered

tubercles. Lower orbital margin formed by 2 broad rounded crenulated lobes.

#72 PL bee. 5

Solve “NAR ennues Maer

Fig. 2. Palicus sexlobata n.sp.

a. carapace, dorsal view. 6. Antero-ventral region of carapace. c. Right cheliped. d. First pleopod,

male. e. Second pereiopod.

I 58 ANNALS OF THE SOUTH AFRICAN MUSEUM

Distal portion of basal joint of antenna wider than proximal part, with 3 small

ventral tubercles. 2nd and grd joints of peduncle slender, equal in length. Eye

wider than stalk, latter with 3 dorsally visible lobes. 4th joint of endopod of

maxilliped 3 with external curved flattened portion. Right cheliped stout,

(left missing), upper surfaces of propodus, carpus, merus tuberculate. Finger

and thumb flattened, tips overlapping, cutting edge entire. Pereiopods stout,

meri granulate, upper surface with 2 large spines, more anterior of which

largest, plus several smaller flattened spines. Lower surface with many small

spines. Carpus half length of merus, slightly flattened, proximally with a

flattened rounded lobe on upper edge, distally with an acute flattened spine.

Propodus and dactylus flattened, lower edge of latter entire. Abdomen with 7

segments, first 2 very short, with prominent transverse raised ridge. 3rd segment

twice length of 2nd, also with raised ridge, all 3 ridges dorsally visible.

Material: 1 3, holotype, S.A.M. A12642, carapace length (including rostral

spines) 8:6 mm, carapace breadth 10 mm. Station BRU 371 F. Depth 110

metres.

Remarks: This species would seem to be most closely related to P. investigatoris

(Alcock), but differs in the arrangement of the larger tubercles of the carapace,

also in the posterior carapace ridge. The latter has 6 flattened lobes in P.

sexlobatus, 8 in P. investigatoris. The ambulatory pereiopods differ in that the

propodi and dactyli of P. ¢nvestigatoris are denticulate on their lower edges,

entire in P. sexlobaius. This appears to be the first record of this cosmopolitan

genus from the Mocambique Channel.

Family Retroplumidae

Alcock & Anderson, 1894: 180. Gill, 1894: 1044. Doflein, 1904: 29. MacGilchrist, 1905: 266.

Tesch, 1918: 29. Rathbun, 1932: 33. Sakai, 1948: 606.

Retropluma planiforma n.sp.

Fig. 3 a-g

Description: Garapace more or less flattened, almost naked, slightly granular.

Carapace divided into 3 parts by 2 transverse carinae. Entire carapace margin

crenulated. Anterior portion sloping forward to slender apically rounded

rostrum. Latter slightly shorter than basal antennular peduncle joint. Supra-

orbital border smoothly contoured, unarmed. External orbital angle a rounded

forwardly projecting lobe. Antero-lateral border sloping obliquely outward

from external orbital lobe to more anterior lateral carapace lobe. Latter is

extension of anterior transverse carina. Lateral margin of middle carapace

portion with a convex rounded lobe. Second carapace carina slightly curved.

Carapace wider than long (excluding rostrum). Postero-lateral angles rounded.

Eyestalks free, directed laterally, upper surface granular. Base of eyestalks

wider than cornea. Infra-orbital spine prominent, forwardly directed, margins

crenulated, reaching to middle of second joint of antennal peduncle. Latter

DECAPOD CRUSTACEA FROM THE SOUTH-WEST INDIAN OCEAN 159

Fig. 3. Retropluma planiforma n.sp.

a. Carapace, dorsal view. b. Carapace, ventral view. c. Abdomen, female. d. Abdomen, male.

e. First pleopod, male. f: Right chela. g. Left chela.

half length of basal joint. Flagellum of antenna with 19 or 20 joints. Basal

antennular joint globular, inflated, granular. 2nd joint arising at antero-mesial

corner of basal joint. Bases of antennae, antennules and infra-orbital spines in

a line. Chelipeds equal in length, right stouter than left, palm higher, equal in

length to finger. Palm of left chela shorter than finger, upper borders of both

chelae smooth. Fingers wide at base, compressed. A gap between fingers of

right chela. Fingers of latter with 3 (possibly 4) teeth, on sharp cutting edge

160 ANNALS OF THE SOUTH AFRICAN MUSEUM

Cutting edge of left chela only slightly denticulate. All pereiopods fringed with

feathered hairs. Ambulatory pereiopods (i.e. 2, 3, 4) long and slender, 2nd

pair longest. Dactyli slender, slightly curved, equal in length to outer margin

of propodus of same pereiopod. Anterior margins of propodi and carpi, anterior

and posterior margins of meri finely denticulate. Distal denticles of carpi and

proximal denticles of propodi particularly well developed. 5th pereiopods

reduced, almost dorsal, prominently fringed with feathered hairs. Entire 5th

pereiopod equal in length to merus of 4th. Abdomen of male and female

triangular. Abdomen of female 7-jointed, 7th longest, apically rounded. 6th

segment with lateral notch, distal blunt tooth, a raised ridge between the

notches. Abdomen of male 5-jointed, 3rd joint largest (consisting of fused

segments 3-5), 6th joint with posteriorly directed lateral tooth, also with

raised ridge. 7th segment apically rounded. Pleopods of female sometimes

protruding from beneath abdomen.

Material: Holotype, $S.A.M. A12644, paratypes, S.A.M. A12643, A12645.

2 carapace 6 carapace

length breadth length breadth Station | Depth (m)

Holotype 7:0

Paratype 5°4

6-8

Paratype 5:8

ze) BRU 390 H 200

4°35

Remarks: The genus Retropluma is represented by 5 species, viz. notopus (Alcock

& Anderson, 1894), from the eastern Indian Ocean, chuni Doflein, 1904, from

the Andaman Islands, plumosa Tesch, 1918, from the Kei Islands in the Banda

Sea, and denticulata Rathbun, 1932, from Japan. Retropluma eocenica Via Boada,

1959, has been recorded from the Eocene of Spain. The present species differs

from notopus and denticulata in having rounded lobes on the lateral margins of

the carapace and an apically rounded rostrum. R. notopus has no lateral extru-

sions, neither has denticulata. The rostrum is apically bifid in notopus, a mere

point in denticulata. R. planiforma differs from chuni in that while the latter has

acute lateral carapace spines and a tapering rostrum, the former has rounded

lobes and a rostrum with parallel sides. R. planiforma most closely resembles

plumosa in the rounded carapace lobes, the shape of the male abdomen and

the granulate character of the integument, especially that of the appendages.

The main differences include the shape of the rostrum (tapering in plumosa)

the externo-orbital angles (angular in plumosa, rounded in planiforma), the

denticulate nature of the ambulatory pereiopods in planiforma (granulate but

not noticeably denticulate in plumosa). The antennal flagellum has fewer

segments (25 in plumosa, 19-20 in planiforma). A very obvious difference is the

lack of specialized hairs found on the appendages in plumosa. These hairs are

scattered amongst the plumose hairs and are inflated and membranous. R.

planiforma has only plumose hairs.

DECAPOD CRUSTACEA FROM THE SOUTH-WEST INDIAN OCEAN 161

Family Dorippidae

Ethusa sinespina n.sp.

Fig. 4 a—c

Ethusa spp. Alcock, 1896: 281-286. Doflein, 1904: 27-32.

Description: Carapace slightly longer than broad, finely and evenly granulate,

anterior portion with fine scattered hairs. Frontal indentation forming angle

of about 60°. Front bilobed, 4-toothed, inner teeth slightly longer than outer,

smooth rounded indentation separating the 2 teeth. Supra-orbital angle acute.

External orbital spine acute, directed slightly outward, not reaching to tip of

outer frontal spine. Carapace regions not very well demarked, cardiac and

branchial regions with indistinct delimiting grooves. Branchial regions very

slightly convex in dorsal view. Antennal peduncle slightly longer than frontal

spines. Eyestalks stout, movable. Efferent branchial canals ending just behind

frontal notch. Chelipeds small, finger and thumb equal in length to palm.

Finger and thumb separated by gap, each with 4 small teeth on cutting edge.

Dactylus of 3rd leg longer than propodus, equal in length to merus. Abdomen

7-jointed, 3rd and 4th joints widest. Distal segment apically rounded.

Fig. 4. Ethusa sinespina n.sp.

a. Carapace, dorsal view. b. Carapace, antero-ventral region. c. Right cheliped.

162 ANNALS OF THE SOUTH AFRICAN MUSEUM

Material: 1 9, paratype, S.A.M. A12649 carapace length, 5-8 mm, carapace

breadth 5:2 mm. Station BRU 390 S. Depth 138 metres. 1 9, ovigerous,

Holotype, S.A.M. A12648, carapace length 7-5 mm, carapace breadth 6-8 mm.

Station BRU 390 E. Depth 350 metres. 1 9, carapace length 4 mm, carapace

breadth 3-5 mm. Station BRU 358 C. Depth 370 metres.

Remarks: These specimens are most closely related to E. zurstrassent Doflein,

recorded by the Valdivia from the East African coast, but differ from this species

in that there is no minute spine between the pairs of frontal spines. The external

orbital spine is not dorso-ventrally flattened. The grooves defining the carapace

regions are not so well defined as in zurstrassent.

Family Leucosiidae

Nursilia dentata Bell, 1855

Fig. 5 a-e

Nursilia dentata Bell, 1855: 309. Alcock, 1896: 260. Rathbun, 1911: 203.

Description: Entire carapace finely and evenly granular, roughly hexagonal.

Front with large raised supra-orbital lobes, with single smaller ventro-lateral

exorbital tooth. Indistinct mid-dorsal carapace carina, stretching from base of

supra-orbital lobes to 1st median spine in cardiac region. 4 blunt prominent

spines in median cardiac region, most anterior of which largest, flanked by

Fig. 5. Nursilia dentata Bell.

. Carapace, dorsal view. b. Carapace, lateral view. c. Abdomen, male. d. First pleopod, male.

e. Right chela.

DECAPOD CRUSTACEA FROM THE SOUTH-WEST INDIAN OCEAN 163

pair of blunt knobs on posterior gastric region. Lateral carapace margin slightly

upturned, with 5 teeth, most posterior of which largest. Between largest lateral

tooth and median line, 4 or 5 slight knobs. Abdomen of 4 segments, first 2

narrow, 3rd largest, latter three times longer than broad, with single forwardly

directed median spine near distal margin; distal segment triangular, apically

rounded. Palm of left chela slightly inflated, two-thirds length of fingers. Latter

slender, with about 25 large and small teeth on cutting edge.

Previous records: Andaman Sea, off Ceylon, Madras coast, off Maldives, Sey-

chelles, Malabar coast, Cargados Carajos.

Material: 1 3, carapace length 7 mm, carapace breadth 8 mm. Station BRU

971 F. Depth 110 metres.

ANOMURA

Family Paguridae

Nematopagurus squamichelis Alcock, 1905

Fig. 6 a—d

Nematopagurus squamichelis Alcock, 1905: 113, pl. 12, fig. 1.

Description: Carapace breadth three-quarters that of length, broadest in bran-

chial region. Latter clearly defined dorsally by 2 ridges. Cervical groove distinct.

Rostrum lacking, frontal margin smoothly curved. Eyes much wider than eye-

stalks, latter stout, one-third of carapace length. Ocular scales tiny. Chelipeds

equal in length, right stouter than left. Tips of finger and thumb corneous,

entire hand and fingers and carpus of chelae covered with flat imbricating

squamae. Vas deferens of male protruding on right side, slender and elongate,

ending in coiled filament. Vas deferens of left side protruding as short conical

papilla.

Previous records: Andaman Sea.

Material: 2 33, carapace length 7-5 mm, 8-o mm, 1 9, carapace length 5-5 mm.

Station BRU 370 G. Depth 347 metres.

Nematopagurus gardinert Alcock, 1905

Fig. 6 e-h

Nematopagurus gardineri Alcock, 1905: 110, plate 12, fig. 2.

Description: Carapace breadth half length. Cervical groove and ridges defining

branchial regions well marked. Rostrum lacking, frontal margin smoothly

rounded, tiny supra-antennal spine present. Ocular scale minute. Eyes wider

than eyestalks. Chelipeds thickly pilose, especially on outer surface of propodus

and carpus. Right chela slightly longer, much stouter than left. Cutting edges

and tips of finger and thumb corneous. Upper edge of carpus and propodus

spinous, denticulate. Outer surface with median longitudinal row of spines.

Previous records: Maldive Islands.

164 ANNALS OF THE SOUTH AFRICAN MUSEUM

Material: 1 3, carapace length 4:8 mm. Station BRU 300 S. Depth 138 metres.

Remarks: Easily distinguished from the previous species by the pilose and

spinous nature of the chelipeds.

Fig. 6. Nematopagurus squamichelis Alcock.

a. Carapace, and anterior appendages. b. Left chela and carpus. c. Right chela and carpus.

d. Sternum and 5th pair pereiopods, male.

Nematopagurus gardineri Alcock.

e. Carapace and anterior appendages. f. Left chela and carpus. g. Right chela and carpus.

h. Sternum and 5th pair pereiopods, male.

DECAPOD CRUSTACEA FROM THE SOUTH-WEST INDIAN OCEAN 165

Family Galatheidae

Munida semoni Ortmann, 1894

Munida semoni Ortmann, 1894: 24, plate 1, fig. 4. Barnard, 1950: 491, fig. gac.

Material: see page 153 of species list.

Remarks: The present specimens agree closely with semoni, but differ in the

following respects: the 2nd abdominal segment has 8 spines on the anterior

margin, the posterior portion of segments 2 and 3 with only 1 setose transverse

groove (2 in semonz), 4th joint of maxilliped 3 with only 2 strong spines (3 in

semoni). In smaller specimens, there are sometimes 2 spines on the anterior

margin of the 3rd abdominal segment.

Petrolisthes militaris (Heller, 1862)

Fig. 7 a-d

Petrolisthes militaris (Heller), Miyake, 1943: 56. Haig, 1964: 357.

Description: Carapace length (including rostrum) equal to breadth. Frontal

margin broadly triangular, apically rounded, base (between supra-orbital

teeth) just less than twice length, margin finely crenulated. Prominent supra-

Fig. 7. Petrolisthes militarts (Heller).

a. Carapace. b. Left cheliped. c. Basal joint of antennule. d. Basal

jomts of antenna.

166 ANNALS OF THE SOUTH AFRICAN MUSEUM

orbital spine present. Epibranchial spine present. Just posterior to latter, small

spine in antero-branchial region. Lateral branchial margin with 2 or 3 largish

spines, anterior to which, 2 or 3 minute spines. Gastric and branchial regions

covered with transverse rugae. Basal joint of antennule with distal toothed keel.

ist free peduncular joint of antenna with distal slightly flattened denticulate

lobe. Chela longer than carapace. Merus, carpus, propodus, and dactylus

of cheliped covered with flattened rugae. Merus with large crenulated tooth on

inner distal angle. Carpus with 5 crenulate flattened teeth on anterior margin,

distal 2 apically blunt. Posterior margin of carpus with 3 small distal spines.

Proximal posterior margin of chela slightly denticulate, denticles decreasing

in size distally. Meri of ambulatory pereiopods with flattened rugae. Propodi

cylindrical, about twice length of carpus. Dactylus with 3 sharp spines on lower

margin, ending in acute curved talon.

Distribution: Widespread throughout Indo-Pacific region.

Material :

G carapace @ carapace

length length Station Depth (m)

4°5 BRU 372G 55

5:0

6-0

oOo Ou ph PA

Se © Go ©

Remarks: These specimens would appear to constitute the most south-westerly

record for the species.

THALASSINIDEA

Family Axiidae

Axius (Neaxius) sp.

Fig. 8 a-b

Description: Rostrum triangular, breadth equal to length, apically notched,

margin with 4 spines. Prominent exorbital spine, region between latter and

base of rostrum smooth. Lateral margin of anterior carapace region with 13

to 14 spines. Very prominent cervical groove. Anterior two-thirds of flattened

part of carapace with scattered spines. Telson slightly broader than long,

without transverse carinae.

Material: 1 3, carapace length 4 mm, overall length 9:5 mm. Station BRU

357 B. Depth 70 metres. 1 ovigerous 9, dimensions as ¢. Station BRU 256 B.

Depth 18 metres.

Remarks: As many of the appendages are missing, it is difficult to be definite

DECAPOD CRUSTACEA FROM THE SOUTH-WEST INDIAN OCEAN 167

about the specific position of these specimens. Axius acanthus var. mauritianus

Bouvier, 1914, has been recorded from Mauritius. This species has a telson

with 2 transverse carinae, poorly developed lateral carapace spines, and no

scattered carapace spines. The present specimens lack the telson carinae, while

possessing scattered spines on the carapace, and well-developed antero-lateral

carapace spines.

Fig. 8. Axius (Neaxius) sp.

a. Carapace. b. Telson and uropod.

PENAEIDEA

Family Penaeidae

Gennadas propinquus Rathbun, 1906

Fig. 9

Gennadas propinquus Rathbun, 1906: 907. Barnard, 1950: 634.

Gennadas scutatus Kemp, 1910: 178, non Bouvier, 1908: 42.

Gennadas scutatus indicus Balss, 1927: 259.

Previous records: Indian Ocean, off Hawaii.

Material: 1 3, carapace length (including rostrum), 9 mm. Station BRU 363 P.

Depth 1225 metres.

Fig. 9. Gennadas propinquus Rathbun. Petasma.

168 ANNALS OF THE SCUTH AFRICAN MUSEUM

CARIDEA

Family Pasiphaeidae

Leptochela pugnax de Man, 1920

Fig. 10 a—c

Leptochela pugnax de Man, 1920: 26. Kemp, 1925: 255. Barnard, 1958: 6 (In L. robusta).

Description: Rostrum slender, reaching beyond eyes to 2nd antennular peduncle

segment. Small antennal spine present. Rostral carina not continued posteriorly

along carapace. Dactylus of 2nd pereiopod with about 19 spines, finger of

propodus with about 21 spines. 5th abdomimal segment not dorsally carinate,

unarmed. Pleurae of 3rd, 4th, 5th abdominal segments ventrally rounded, each

with small tooth in anterior region. 6th abdominal segment with long ventro-

lateral spine about two-thirds from anterior end, followed by several setae.

Posterior margin of 6th segment with prominent lateral spine. Anterior part

of telson with 1 pair of dorsal spines, 1 pair of lateral spines at about the mid-

point. Telson with 5 pairs of apical spines.

Previous records: Maldives, Andamans, Nicobars, Mergui Archipelago.

Material: 1 9, carapace length 4 mm, overall length 13 mm. Station BRU

392 F. Depth 35 metres.

Remarks: The present species closely resembles L. robusta Stimpson, which has

sex

SA)

aS]

= Tareditvett ess WEEE EVA EES Sty

= aay Beaty EN AGT CAR INLECS ANDES oy

- ES re EEG ee

othe! v=

tr mt

Fig. 10. Leptochela pugnax de Man.

a. Lateral view. b. First chela. c. Telson.

DECAPOD CRUSTACEA FROM THE SOUTH-WEST INDIAN OCEAN 169

been recorded from Inhambane, but differs in the possession of an antennal

spine, and one pair of lateral telson spines and not two as in robusta.

Family Oplophoridae

Oplophorus spinicauda Milne-Edwards, 1883

Fig. 11

Oplophorus spinicauda Milne-Edwards, 1883: plate 29. de Man, 1920: 48. Chace, 1940: 184.

Description: Postero-lateral angle of carapace lacking tooth. and, 3rd, 4th

abdominal segments ending in long spines. Telson terminating in end piece,

latter armed laterally with spines.

Previous records: East coast of U.S.A., West Indies, off Morocco, north of

Malagasy Republic, off Indian coast, Philippines, Hawaii.

Material: 1 ? 9, carapace length (excluding rostrum) 5:5 mm. Station BRU

363 P. Depth 1225 metres.

—m

PETROL A ek Y Pad

oath? 886 ome On Soy hot een Lia

CORD Cn TOI Cari cb Uncorked aR ah Ay

PSB ORBIT I pate tithe

——at

wee

Fig. 11. Oplophorus spinicauda M.-Ed.

Telson.

170 ANNALS OF THE SOUTH AFRICAN MUSEUM

Family Pandalidae

Heterocarpus woodmasoni Alcock, 1901

Fig. 12

Heterocarpus woodmasoni Alcock, 1901: 108. de Man, 1920: 156.

Description: Rostrum one and a half times longer than carapace, 10 dorsal,

7 ventral teeth. Rostral carina produced almost to posterior margin of cara-

pace, posterior portion indistinct. Carapace with prominent. post-antennular

and post-antennal carinae, starting with prominent antennal and branchio-

stegal spines respectively. Antennal scale two-thirds carapace length. Ist

pereiopods equal in length to 3rd maxillipeds, ending in minute dactyl. 2nd

pereiopods both chelate, right shorter than left. Right chela slightly larger than

left, carpus of former with 12 segments, carpus of latter with about 20 segments.

grd, 4th, 5th pereiopods similar, with slender dactyls. Abdominal segments

1 and 2 dorsally smooth, 3rd with flattened hook-like tooth, 4th, 5th, 6th dor-

sally smooth. Telson elongate, tapering, equal in length to outer branch of

uropod, apically pointed, with 2 pairs of subapical spines.

Previous records: Bali Sea, Makassar, Kei Islands, Madura Straits, Andaman

Sea.

Material: 1 9, carapace length (excluding rostrum) 9 mm, overall length

43 mm. Station 370 G. Depth 347 metres.

Remarks: This appears to be the most southerly record of this very distinctive

Indian Ocean species.

Fig. 12. Heterocarpus woodmasoni Alcock.

Lateral view.

Plestonika acanthonotus (Smith, 1882)

Fig. 13 a—b

Pandalus acanthonotus Smith, 1882: 61, plate 13, figs 10, 11.

Plesionika acanthonotus (Smith), de Man, 1920: 105. Holthuis, 1951: 62, fig. 13 b-t.

Description: Rostrum two-thirds carapace length, compressed, spines variable

(11/5, 12/4, 12/5). Prominent antennal and branchiostegal spines. Basal joint

DECAPOD CRUSTACEA FROM THE SOUTH-WEST INDIAN OCEAN 171

of antenna with prominent spine on lower distal angle. Antennal scale seven-

eighths carapace length, rostrum two-thirds antennal scale length. Carapace

and abdomen with minute scales. Ist pereiopod with microscopic dactyl, very

slender, propodus half length of carpus, 2nd pereiopods equal in length, reaching

beyond antennal scale. Chela about one-eighth length of carpus. Latter con-

sisting of about 21 jointlets. Merus and ischium equal in length, each slightly

more than half length of carpus. Dactyl of 3rd pereiopod one-third length of

propodus, merus reaching to end of antennal scale. Posterior margin of merus

with 11 spines. 4th pereiopod similar to 3rd. 5th pereiopod longest, midpoint of

carpus reaching to end of antennal scale. Merus armed with 8 spines. Pleuron

of 5th abdominal segment with tooth on postero-ventral angle. Telson with

3 pairs lateral spines, slightly shorter than inner branch of uropod, latter

slightly shorter than outer branch. Outer margin of outer branch with tooth

some way behind apex.

Previous records : East coast of U.S.A., off Portugal, Spain, Brazil, Angola.

Material: 1 3, carapace length (excluding rostrum) 6 mm, 1 9, carapace length

5 mm. Station BRU 390 H. Depth 175-200 metres. 2 3g, carapace lengths

8 mm, 7:5 mm. Station BRU 390 P. Depth 118 metres.

Remarks: From de Man’s 1920 key to the genus Plesionika, this specimen falls

into the group characterised by a rostrum shorter than the carapace, 2nd

pereiopods equal in length. The species in this group, viz. hypanodon Doflein,

ST EVES

Fig. 13. Plesionika acanthonotus (Smith).

a. lateral view. b. Endopod of pleopod one.

172 ANNALS OF THE SOUTH AFRICAN MUSEUM

and brevis (Rathbun) bear little resemblance to the present species, which more

closely resembles the ortmanni Doflein, longipes (Milne-Edwards), sindot (Rath-

bun) group, in spite of the short rostrum. This species most closely resembles

P. acanthonotus (Smith). The rostra of the 2 species are very similar, being

shorter than either the carapace or the antennal scale. The variation in number

of the rostral teeth in this species is similar to that of acanthonotus. ‘The shape of

the telson, abdomen, and antennular peduncle also agree closely. The pereio-

pods are similar in both species, including the number of jointlets of the carpus

of the 2nd pereiopods. The mouthparts are identical. The antennal scale of

this species appears proportionally more slender than acanthonotus, while the

endopod of the 1st pleopod in the male differs in shape. It seems unusual that

such a typically Atlantic species should occur in the Indian Ocean. Without

further material it is not possible to be more definite regarding the status of

this species, but if not acanthonotus, it is certainly very closely related.

Family Processidae

Processa sp.

Fig. 14

Description: Rostrum apically bidentate, reaching to posterior part of orbit.

Lateral process of basal joint of antennule smoothly rounded, with a slightly

elongate inner rounded angle. Antennal scale apically rounded, spine on outer

margin slightly longer than apex. Palm of chela of right 1st pereiopod almost

twice length of finger and thumb. Dactylus of left 1st pereiopod about one-

third length of propodus. Postero-inferior angle of pleuron of 5th abdominal

segment with a small tooth.

Material: 1 ovigerous 9, carapace length (including rostrum) 4-5 mm, overall

length 16 mm. Station BRU 372 G. Depth 55 metres.

Remarks: This specimen closely resembles P. australiensis Baker, 1907, in the

shape of the carapace, rostrum, telson and 5th abdominal pleuron, but differs

in possessing a smoothly rounded process on the basal antennular joint. P.

australiensis has this process with a spine on the outer angle. The 2nd—5th

pereiopods appear to be more slender than in australiensis. Without more

material it is difficult to give this specimen definite status.

Family Alpheidae

Alpheus nonalter n.sp.

Fig. 15 a—d

Description: No supra-orbital spines. Rostrum reaching about two-thirds along

basal antennular peduncle joint. Latter with swollen base, spine on external

distal angle reaching just beyond end of joint. 2nd antennal peduncle segment

3 times length of 3rd. Antennal scale reaching beyond antennular peduncle.

Larger chela with palm about 4 times longer than broad. No notch at base of

DECAPOD CRUSTACEA FROM THE SOUTH-WEST INDIAN OCEAN 173

Fig. 14. Processa sp.

a. Peduncle of antennule. 6. Tip of antennal scale. c. Chela

of first right pereiopod. d. Dactyl and propodus of left first

pereiopod. e. Antero-lateral portion of carapace. f. Pleuron,

5th abdominal segment. g. Telson.

dactyl, but shallow notch at base of fixed finger. Latter longer than movable

finger. Both chelae granulous. Smaller chela with fingers half to two-thirds

length of palm. Latter cylindrical, six times longer than wide. Movable finger

slightly balaeniceps-like. Meri and ischia of both chelae with inner margin

denticulate, former with spine on inner distal angle. Carpus of 2nd pereiopod

of 5 jointlets, 2nd joint two-thirds length of 1st, equal in length to jointlets

3, 4, 5 together. Pereiopods 3, 4, 5 with simple flattened dactyls. Propodi,

carpi, and meri equal in length. Ischia with single ventral spine. Telson length

twice basal width.

Material: 3 ovigerous 99, carapace lengths (including rostrum) 8-0, (holotype,

S.A.M. A12650), 8-0, 9:0 mm. 7 gd, carapace lengths, 6:0, 7:0, 7°4, 7°5

174. ANNALS OF THE SOUTH AFRICAN MUSEUM

(paratype, S.A.M. A12651), 8-0, 8-0, 8-0 mm. 18 damaged specimens.

Station BRU 390 P. 118 metres. 1 9, carapace length 7-5 mm. 1 damaged.

Station BRU 391 C. 86 metres. 4 99, carapace length 7-5, 8-0, 8-0, 8-o mm.

4 3d, 6:0, 6:0, 7:0, 8-0, 15 damaged, 4 juveniles. Station BRU 390 H. 175-200

metres.

Remarks: The unarmed meri of the third pereiopods, the balaeniceps-like

smaller chela of the male, the simple lanceolate-like dactyls of the last 3 pairs

of pereiopods and the lack of supra-orbital spines place this species in the

Brevirostris group of de Man (i911). This species is most closely related to

A. acutocarinatus de Man, and A. macrosceles Alcock & Anderson. It differs from

Fig. 15. Alpheus nonaliter n.sp.

a. Lateral view. b. Anterior carapace and appendages, dorsal view. c. Large chela. d. Small chela.

DECAPOD CRUSTACEA FROM THE SOUTH-WEST INDIAN OCEAN 175

A. acutocarinatus in that it lacks a prominent post-rostral carina and the tubercle

or tooth just behind the orbital hood. It differs from A. macrosceles (which has a

subcylindrical chela) in having a slightly flattened larger chela. The smaller

chela of A. macrosceles is not balaeniceps-like, as in this species; both fingers meet

when shut.

Alpheus waltervadi n.sp.

Fig. 16 a—c

Description: Rostrum reaching to middle of 1st antennular peduncle segment,

supra-orbital spines slightly shorter. 2nd antennular peduncle segment twice

length of ist, 13 length of 3rd. 1st antennular peduncle segment basally broad,

with external spine reaching to end of segment. Apical spine of antennal

scale reaching beyond 3rd antennular peduncle segment. Larger chela of Ist

pereiopods with distinct deep notch in upper margin of palm behind dactyl.

Notch followed distally by raised flattened lobe, distal end of which armed with

forwardly directed tooth. Latter overhangs dactyl. Carpus one quarter length

of merus, latter triangular in cross section, distally ending in 3 lobes at the

angles, margins unarmed. Smaller chela only slightly shorter than other,

fingers equal in length to palm. Spine at distal end of palm overhangs dactyl.

Latter with concave inner edge. Carpus of 2nd pereiopods consisting of 5

jointlets, 1st largest. Ischium only slightly longer than merus. Dactyls of pereio-

pods 3, 4, 5, biunguiculate. Carpi two-thirds length of propodi, latter armed

ventrally with 8 spines. Carpus with distal lobe overlapping propodus, other-

wise unarmed, merus with flattened spine on ventral distal margin. Telson

14 times longer than basal width, with 2 pairs lateral spines dividing the

appendage into thirds.

Material: 2 ovigerous 99, carapace lengths 5-5 (paratype, S.A.M. A12647),

4:1 mm (holotype, S.A.M. A12646). 4 3g, carapace lengths 3:4, 4:0, 4:1,

4-1 mm. 6 juveniles, 5 damaged, station BRU 381 A-C, 38-46 metres (Walter’s

Shoal).

Remarks : According to de Man’s 1911 classification of the genus Alpheus (taken

from Coutiere, 1899), this species belongs to the Megacheles group, characterized

by the presence of supra-orbital spines, a grooved and notched first chela,

unarmed meri of the 3rd pereiopods and biunguiculate dactyls for the last

three pairs of pereiopods. This species is very closely related to the following:

A. hailstonei Coutiere, and its varieties assimulans de Man, laetabilis de Man,

and to A. paradentipes Coutiere. It differs from hailstonei in that the larger chela

is proportionally more robust and shorter and differently shaped. The supra-

orbital spines are not as pronounced as in hailstonez.

It differs from hailstonei var. assimulans in the detailed shape of the larger

chela, the stockier meri of the third pereiopods and in the 4th joint of the carpus

of the 2nd pereiopods being only half as long as the 2nd. (These are equal in

assimulans.) It differs from hailstonei var. laetabilis in lacking the spinous merus

I 76 ANNALS OF THE SOUTH AFRICAN MUSEUM

of the larger chela characteristic of this variety. It differs from paradentipes, which

it most closely resembles, in the shape of the larger chela, in having much less

prominent supra-orbital spines, and no spines on the carpi of the last three

pairs of pereiopods. Further investigation may well prove this species to be

synonymous with an already established one. This is, however, a preferable

situation to incorrectly assigning them to an already established species.

Fig. 16. Alpheus waltervadi n.sp.

a. Lateral view. b. Anterior carapace and appendages, dorsal view.

c. Large chela. d. Small chela. e. Telson.

DIsTRIBUTION

In the following table the geographical distribution of all the species in

this collection is given. All are automatically included in the Indo-Pacific

region. The South African region includes the area from Cape Point east to

DECAPOD CRUSTACEA FROM THE SOUTH-WEST INDIAN OCEAN N77

about Durban. The Atlantic region includes the area west from Cape Point.

It can be seen from this table that the majority of animals have Indo-Pacific

affinities, while only about 21 species occur in all three regions.

Species Indo-Pacific South African Atlantic

BRACHYURA

Achaeopsis spinulosus

Achaeopsis thomsont

Achaeus lacertosus

Achaeus cf. affinis

Actaea rueppelli

Calappa lophos

Carcinoplax longimanus

Charybdis cf. annulata

Charybdis variegata

Conchoecetes artificiosus

Dorwppe lanata

Ebalia barnardi

Ebalia tuberculata

Ebalia tuberculosa f. postulans

Ebalia tuberculosa f. scandens

Ethusa sinespina

Eumedonus granulosus

Eurynome aspera

Goneplax angulata

Gontoneptunus africanus

Homola barbata

Hyastenus spinosus

Inachus cf. dorsettensis

Inachus guentheri

Leucosia marmorea

Lophozozymus dodone

Lupocyclus tugelae

Macropodia formosa

Nursilia dentata

Palicus sexlobatus

Paratergatis longimanus

Philyra globosa

Philyra globulosa

Pilumnus hirsutus

Pilumnus longicornis

Platylambrus quemvis

ISAS OK DSI

7K eS 2S

PO Va RE IRI RR IR OR OD RE EEK, REI Ge Da ee a eX

I 78 ANNALS OF THE SOUTH AFRICAN MUSEUM

Species

Platypodia granulosa

Portumnus mcleayr

Ranina ranina

Retropluma planiform

Thalamita woodmasoni

Xanthias tuberculidens

ANOMURA

PAGURIDEA

Anapagurus hendersoni

Dardanus arrosor

Dardanus euopsis

Dardanus setifer

Diogenes brevirostris

Diogenes costatus

Nematopagurus gardinert

Nematopagurus squamichelis

Pagurus spinulentus

Para-pagurus pilosimanus

GALATHEIDEA

Galathea dispersa

Galathea intermedia

Munida sanctipaul

Munida semoni

Porceilana dehaanu

Porcellana streptocheles

MACRURA

PENAEIDEA

Acetes erythraeus

Gennadas propinquus

Macropetasma africana

Metapenaeopsis adamanensis

Metapenaeopsis stebbingi

Parapenaeus fissurus

Penaeopsis rectacuta

Penaeus japonicus

Sergestes prehensilis

Solenocera africanum

Solenocera pectinata

Indo-Pacific

OOK OK IG OOD Re GEE OK DK eX eee

OP OO eK IX XX

South African

Atlantic

DECAPOD CRUSTACEA FROM THE SOUTH-WEST INDIAN OCEAN 179

Species Indo-Pacific South African Atlantic

CARIDEA

Alpheus nonalter

Alpheus waltervadi

Alpheus frontalis

Chlorotocus crassicornis

Eualus ctenifera

Heterocarpus woodmasoni

Hippolysmata vittata

Latreutes mucronatus

Leptochela pugnax

Leptochela robusta

Nikoides cf. danae

Oplophorus spinicauda

Plestonika cf. acanthonotus

Plesionika martia

Pontocaris cataphracia

Pontocaris lacazet

Processa austroafricana

Stylodactylus bimaxillaris

Synalpheus anisocheir

Synalpheus jedanensis

Tozeuma armata

MeO CEI KO eK OKO KT

ne ees

SUMMARY

A collection of brachyuran, anomuran and macruran decapod Crustacea

is described. The material is from the south-west Indian Ocean, i.e. off the

coasts of Portuguese East Africa, Natal and Malagasy Republic, and includes

approximately 110 species, of which 15 are new records, and 5 previously

undescribed.

ACKNOWLEDGEMENTS

Thanks are due to the Zoology Department of the University of Cape

Town, and particularly to Mr. J. Field, for making the material available.

Thanks are also due to Mr. C. Berrisford and Miss L. Joubert for preliminary

identifications of some of the material and to Dr. M.-L. Penrith for helpful

comments and criticisms.

The Trustees of the South African Museum are grateful to the Council for

Scientific and Industrial Research for a grant towards the publication of this

paper.

180 ANNALS OF THE SOUTH AFRICAN MUSEUM

REFERENCES

Aucock, A. 1896. Materials for a carcinological fauna of India. 2. The Brachyura Oxystoma.

F. Asiat. Soc. Beng. 65: 134-206.

Aucock, A. 1901. A descriptive catalogue of the Indian deep-sea Crustacea, Decapoda Macrura and Anomala,

in the Indian Museum, being a revised account of the deep-sea species collected by the Royal Indian

Marine Survey ship ‘Investigator’. Calcutta: Indian Museum.

Aucock, A. 1905. Catalogue of the Indian decapod Crustacea in the collection of the Indian Museum. 2.

Anomura. Calcutta: Indian Museum.

Atcock, A. & ANDERSON, A. R. 1894. Natural history notes from H.M. Indian Marine Survey

steamer ‘Investigator’, Commander C. F. Oldham, R.N., commanding. Series 2, no. 14.

An account of a recent collection of deep-sea Crustacea from the Bay of Bengal and Lac-

cadive Sea. 7. Asiat. Soc. Beng. 63: 141-185.

BAKER, W. H. 1907. Notes on South Australian decapod Crustacea. Part 5. Trans. R. Soc. S.

Aust. 31: 173-191.

Bauss, G. 1927. Macrura der Deutschen Tiefsee-Expedition. 3. Natantia, Teil B. Wiss. Ergebn.

dt. Tiefsee-Exped. ‘Valdivia’ 23: 245-275.

BARNARD, K. H. 1950. Descriptive catalogue of South African decapod Crustacea. Ann. S. Afr.

Mus. 38: 1-824.

BARNARD, K. H. 1954. Notes sur une collection de crustacés décapodes de la région Malgache.

Mem. Inst. scient. Madagascar (A) 9: 95-104.

BARNARD, K. H. 1958. Further additions to the crustacean fauna-list of Portuguese East Africa.

Mem. Mus. Dr. Alvaro de Castro 4: 3-27.

BELL, T. 1855. Horae carcinologicae, or notices of Crustacea. A monograph of the Leucosiadae,

with observations on the relations, structures, habits and distribution of the family; a

revision of the generic characters; and descriptions of new genera and species. Trans. Linn.

Soc. Lond. 24: 277-314.

Bouvier, E. L. 1908. Crustacés décapodes, provenant de campagnes de |’‘Hirondelle’ et de la

‘Princesse Alice’ 1886-1907. Résult. Camp. scient. Prince Albert I. 33: 1-122.

Bouvier, E. L. 1914. Sur la faune carcinologique de Vile Maurice. C. r. hebd. Séanc. Acad. Sci.

Paris 159: 698—704.

Cuace, F. A. 1940. Plankton of the Bermuda Oceanographic Expeditions. 9. The bathypelagic

caridean Crustacea. Zoologica, N.Y. 25: 117-210.

CouTiErE, H. 1899. Les Alpheidae. Annls Sci. nat. 9: 1-560.

DoF LeEIn, F. 1904. Brachyura. Wiss. Ergebn. dt. Tiefsee-Exped. ‘Valdivia’ 6: 1-314.

Fuipse, H. J. 1930. Oxyrrhyncha. Parthenopidae. Siboga Exped. monogr. 39c?: 1-96.

G1, T. 1894. A new bassalian type of crab. Am. Nat. 28: 1043-1045.

Hate, J. 1964. Papers from Dr. Th. Mortensen’s Pacific Expedition 1914-16. 81. Porcellanid

crabs from the Indo-West Pacific, Part 1. Vidensk. Meddr dansk naturh. Foren. 126: 355-386.

HELLER, C. 1862. Neue Crustaceen, gesammelt wahrend der Weltumseglung der k.k. Fregatte

Novara. Zweiter vorlaufiger Bericht. Verh. zool.-bot. Ges. Wien 12: 519-528.

Ho.ruuss, L. B. 1951. The caridean Crustacea of tropical west Africa. Atlantide Rep. 2: 7-187.

Hotruuss, L. B. & Gortriies, E. 1958. An annotated list of the decapod Crustacea of the Medi-

terranean coast of Israel, with an appendix listing the Decapoda of the eastern Mediter-

ranean. Bull. Res. Coun. Israel 7B: 1-126.

Kemp, S. 1910. Notes on the Decapoda in the Indian Museum. 1. Rec. Indian Mus. 5: 173-181.

Kemp, S. 1925. On various Caridea. Notes on Crustacea Decapoda in the Indian Museum. 17.

Rec. Indian Mus. 27: 249-343.

Maccitcurist, A. C. 1905. Natural history notes from the R.I.M.S. ‘Investigator’ Capt. T. H.

Heming, R.N. (retired) commanding. Series 3, no. 6. An account of some of the new and

some of the rarer decapod Crustacea obtained during the surveying season Wein gee

Ann. Mag. nat. Hist. (7) 15: 233-268.

DECAPOD CRUSTACEA FROM THE SOUTH-WEST INDIAN OCEAN 181

Man, J. G. de. 1893. Report on the podophthalmous Crustacea collected in the year 1891 by

Dr. H. Ten Kate in some islands of the Malay Archipelago. Notes Leyden Mus. 15: 284-311.

Man, J. G. de. 1911. The Decapoda of the Siboga Expedition. Part 2. Family Alpheidae. Siboga

Exped. monogr. 3ga!: 1-465.

Man, J. G. de. 1920. The Decapoda of the Siboga Expedition. Part 4. Families Pasiphaeidae,

Stylodactylidae, Hoplophoridae, Nematocarcinidae, Thalassocaridae, Pandalidae, Psali-

dopodidae, Gnathophyllidae, Processidae, Gylphocrangonidae and Crangonidae. Siboga

Exped. monogr. 3g9a*: 1-318.

Miine-Epwarps, A. 1883. Recueil de figures de crustacés nouveaux ou peu connus. Paris.

Mrvaxg, S. 1943. Studies on the crab-shaped Anomura of Nippon and adjacent waters. 7. Dep.

Agric. Kyushu imp. Univ. 7: 49-158.

OrTMANN, A. E. 1894. Crustaceen. Jn Semon, R. Zoologische Forschungsreise in Australien und den

malayischen Archipel. 5: 1-80. Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 8: 1-80.

Ratusun, M. 1906. The Brachyura and Macrura of the Hawaiian Islands. Bull. U.S. Fish

Commn 23: 827-930.

Ratuepwn, M. 1911. Marine Brachyura. Trans. Linn. Soc. Lond. (2, Zool.) 14: 191-261.

Ratusun, M. 1918. The grapsoid crabs of America. Bull. U.S. natn. Mus. 97: i-xxii, 1-461.

Ratusun, M. 1932. Preliminary descriptions of new species of Japanese crabs. Proc. biol. Soc.

Wash. 45: 29-38.

Sakal, T. 1939. Studies on the crabs of Japan 4. Brachygnatha, Brachyrhyncha. Tokyo: Yokendo.

Sakal, T. 1965a. On two new genera and five new species of xanthoid crabs from the collection

of His Majesty the Emperor of Japan, made in Sagami Bay. Crustaceana 8: 97-106.

SAKAI, T. 1965). The crabs of Sagami Bay. Honolulu: East West Centre Press.

Smiru, S. I. 1882. Reports on the results of dredging . . . on the east coast of the United States...

by the U.S. Coast Survey Steamer ‘Blake’. . . 17. Report on the Crustacea. Part 1. Deca-

poda. Bull. Mus. comp. Zool. Harv. 10: 1-108.

Tescu, J. J. 1918. The Decapoda Brachyura of the Siboga Expedition. 1. Hymenosomidae,

Retroplumidae, Ocypodidae, Grapsidae and Gecarcinidae. Siboga Exped. monogr. 39c!:

1-148.

<r ied

Ran

my fg

a |

me at) ;

! u

i |

, |

POP a

Rigs

i } a

1 |

‘

. 2

[

{

f }

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)

Bu.tLoucuH, 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. 19604. 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. 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 ful! 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.

y (mae Yoh ay ihn eeu keane. ue

er ; eee aT Al Hi Wy eth ao Dabiueacid siete:

— =—s

————

ah i

AP ae Re tyrant

7-68

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 52 °&Band

May 1969 Mei

Part: Deel

<ORILLODONTOPS, A NEW SCALOPOSAURID

FROM THE KAROO

M. A. CLUVER

Cape Town Kaapstad

aN HSOW;, yy

AUG 20 1969

C/BRARIED

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, tpl), 5(2, 5, 7-9),

G(r, t=pt), 7(1, 3), 8 o(t-2), 1e(r—3);

Duis; 7» tpl.) 5 2i, 24(2), 27) Sul (bieey 38,

44(4).

Price of this part/Prys van hierdie deel

Ro.50

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

ZORILLODONTOPS, A NEW SCALOPOSAURID FROM THE KAROO

M. A. CLUVER

South African Museum, Cape Town

(With 2 figures)

[MS. received 11 September, 1968]

CONTENTS

PAGE

Introduction . : : : = fo9

Description of the material : Ee kos

Discussion. : : : BES cols

Acknowledgements. : p s . F668

Summary . - - : PPSEGS

References . 3 : ‘ Biers

INTRODUCTION

The South African Museum has recently acquired the major part of the

fully articulated skeleton of a small therocephalian reptile (S.A.M. No. K 1392),

collected by the late Mr. J. J. Pansegrouw at Edenville in the Orange Free

State, from Middle Beaufort beds, most probably those of the Lystrosaurus zone.

Preparation has revealed the entire right side of the skull and most of the right

shoulder girdle and limb. Also most of the vertebral column and parts of the

pelvic girdle and right hind-limb have been exposed. The specimen is embedded

in a soft, green mudstone matrix.

DESCRIPTION OF THE MATERIAL

The skull has suffered some lateral compression, but it nevertheless seems

that in its natural state it was relatively deep and narrow. From the dorsal,

supraoccipital border of the foramen magnum to the tip of the premaxilla it

measures 44 mm, making it one of the smallest therocephalians yet recovered

from the Karoo. The short temporal fossa and relatively large orbit both

measure 10 mm in length. The preorbital length is 19 mm, and is thus a little

less than half of the total skull length.

Lateral compression appears to have resulted in an artificial median

crest above the orbit and temporal fossa. In figure 2 allowance has been made

for this and the dorsal midline has been drawn slightly lower than in the actual

specimen. It does not, however, seem likely that a flat, intertemporal plate

(as in Scaloposaurus) was ever present. One side of the intertemporal region

has been damaged, but there are no indications of a pineal foramen in the

rest of the region. The parietal separates the squamosal from the postorbital,

183

Ann. S. Afr. Mus. 52 (8), 1969: 183-188, 2 figs.

184 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fic. 1. Zorillodontops gracilis, gen. et sp. nov. (S.A.M. No. K1392). Lateral view of preserved

portions of the skeleton x 8. The unshaded areas represent bone impressions in the matrix.

AT—atlas. AX-—axis. CL—clavicle. COR—coracoid. FIB—fibulas HUM—humerus.

IL—ilium. ISC—ischium. PA—proatlas. PC—procoracoid. PUB—pubis. RAD —radius.

SCA—scapula. TIB—tibia. UL—ulna.

which forms only a short part of the dorsal border of the temporal fossa. ‘The

squamosal descends fairly far ventrally towards the quadrate. A triangular

postfrontal lies wedged between the postorbital and the frontal. The postorbital

bar is complete but slender.

The prefrontal is smali and does not separate the nasal from the lacrimal.

This condition has been found in Scaloporhinus angulorugatus (Boonstra, 1953).

At the anterior extremity of the snout the sutures between the premaxilla,

septomaxilla and nasal are not clear.

The premaxilla carries six incisors, the first two being indistinct and the

last one very small. The premaxillary alveolar border meets the maxillary

alveolar border smoothly, and there is no premaxillary ‘step’. There are two

canines, the first one small and separated by a diastema from the second,

large canine. There is a clear diastema between the last incisor and the anterior

canine. Eight small postcanine teeth follow closely on the canine. The post-

canines are shorter and broader than the incisors and there appear to be

minute anterior and posterior cusps on some of them. Unfortunately, the

posterior three teeth have been badly damaged and show no details.

The occiput has been laterally compressed and displaced so that in lateral

view the supraoccipital, exoccipital and opisthotic are visible. The tabular

lies above the post-temporal fenestra, closely apposed to the squamosal. Beneath

the opisthotic the stapes is visible, abutting against the inner surface of the

quadrate.

The epipterygoid is a flat sheet of bone, expanded ventrally but not

dorsally, and is not hour-glass shaped.

A complete ring of approximately 18 extremely delicate scleral plates is

present within the orbit, giving an indication of the size of the eyeball. In the

type specimen of Scaloporhinus angulorugatus scleral plates are also visible.

A NEW SCALOPOSAURID FROM THE KAROO 185

The lower jaw is drawn up tightly inside the maxilla, and no mandibular

teeth could be exposed. The dentary is slender and smoothly curved and the

coronoid process projects into the temporal fossa. The large angular wing

projects below the level of the dentary and bears a system of prominent radiating

ridges. The surangular and articular are also partly visible, but show few details.

The small size of the new specimen suggests the possibility of its being an

immature individual. However, the dermal bones of the skull appear to be

tightly knit and the scleral plates fully ossified. Furthermore, the posterior

canine is fully developed and relatively powerful.

So!

AT. PA. ART N= JUG. LAC. DEN. MAX.

GA TANG. SEL

Fic. 2. Korillodontops gracilis, gen. et sp. nov. (S.A.M. No. K1392) X 2. Lateral view of skull

with occipital distortion uncorrected.

ANG—angular. ART — articular. AT —atlas. DEN—dentary. EPT—epipterygoid. FR —frontal.

jJUG—jugal. LAC— lacrimal. MAX — maxilla. NAS—nasal. OOT—opisthotic. PA—proatlas.

PAR— parietal. PFR-—prefrontal. PM—premaxilla. POW—postorbital. POF —postfrontal.

PTF—post-temporal fenestra. Q—quadrate. SA—surangular. SCL—scleral plates. SMX—

septomaxilla. SO—supraoccipital. SQ—squamosal. ST—stapes. TAB—tabular.

A fairly broad and robust proatlas is present on each side between the

atlas and the side of the foramen magnum. The atlas shows few details, but the

axis bears its characteristic enlarged neural spine. From the eighth vertebra

back the spines are decreased in height and become posteriorly slanted. Beyond

this there is no sudden transition between the cervical and thoracic portions

of the vertebral column. Little can be seen of the column posterior to the fifth

or sixth thoracic vertebra. Four ‘lumbar’ vertebrae, relatively more robust

than the thoracic vertebrae, are preserved.

Twenty ribs of the right side can be made out, the first two of which are

short and meet the sixth and seventh vertebrae. Anterior to the ilium there are

three short, stout ‘lumbar’ ribs.

Most of the right side of the pectoral girdle has been preserved. The

scapula appears to have been slender dorsally but it broadens ventrally to

186 ANNALS OF THE SOUTH AFRICAN MUSEUM

meet the large procoracoid and coracoid (the iatter slightly displaced). The

ventral part of the clavicle lies up against the anterior edge of the scapula and

procoracoid. The proximal portion of the right humerus is preserved in articula-

tion with the slightly distorted glenoid cavity. This portion is only moderately

expanded. Distally, the humerus is indicated as an impression, but it is possible

to estimate the length as 25 mm. Radius and ulna are present, proximally

indicated as bone impressions. The length of the ulna is approximately 19 mm.

The manus was exposed but is incomplete. A radiale and ulnare are clearly

present, as are what I take to be two centrals. There is an indication of a fourth

or fifth distal, but no sign of the first three distals. All five metacarpals are

present, followed by five elongated proximal phalanges similar to Ericiolacerta

(Watson, 1931). The second and third digits have a further anterior, similarly

elongated phalanx each, but no other phalanges or terminals could be found.

Posteriorly the expanded blade-like right ilium is preserved, but the right

pubis and ischium have been lost. Preparation, however, exposed the inner

surfaces of the left ischium and pubis, indicating their general outline. ‘The

obturator foramen can be made out as a notch in the posterior edge of the

pubis, closed off posteriorly by the ischium. There is a fairly strong tibia and

a thin, reduced fibula alongside it. Of the hind-foot, only a few small, scattered

bones and what appear to be two phalanges are left.

DiIscussION

In the following discussion the division of the Therocephalia used by

Haughton & Brink (1954) has been followed.

Owing to its small size, slender, smoothly curved dentary, single large

canine and small anterior canine, short temporal fossa and lack of a raised

temporal crest, our skull does not correspond to any members of the Prister-

ognathidae, Lycosuchidae or Alopecodontidae. It is also manifestly distant

from the specialized Whaitsiidae, Lycideopsidae and Euchambersidae.

Among the Ictidosuchidae, both Ictidosuchus and Ictidodraco have one

canine. Broom (1920, 1932) found two anterior canines in Ictidosuchoides but

Boonstra (1934) could find only two unerupted precanines. This family is,

moreover, characterized by a long temporal fossa and a raised intertemporal

crest.

The Akidnognathidae have a small anterior canine tooth and a post-

frontal, but the skull is much larger. In Akidnognathus the first canine follows

immediately on the powerful incisors, whereas in our specimen a considerable

diastema separates the incisors and canines. Ictidosaurus resembles our specimen

more in this respect, but is from the Tapinocephalus zone and is much larger.

There is, moreover, a marked premaxillary ‘step’, which occurs also in Scylaco-

saurus. Trochosuchus has only four postcanine teeth.

The new specimen must therefore fall under the Scaloposauridae.

A NEW SCALOPOSAURID FROM THE KAROO 187

It should be noted that Romer (1956, 1966), divides the Scaloposauridae

of Haughton & Brink (1954) into several new families which, together with

the Lycideopsidae and Ictidosuchidae, he places in another infraorder, the

Bauriamorpha. According to Haughton & Brink (1954), the Scaloposauridae

can be defined as follows:

‘Skull small. Snout usually long. Intertemporal bar fairly broad. Postorbital

bar either complete or incomplete, but always feeble. Secondary palate

in process of development. Teeth numerous and pointed, with one or two

small canines in front of the main canine. Pineal foramen present or

absent’ (p. 141).

The presence of a postfrontal is uncommon in this family, but has been

found in Ictidostoma hemburyi (Broom, 1932), which is from the Endothiodon

zone. For the rest, the small size, slender dentary, weak postorbital, small

precanine tooth, short temporal fossa (and possible absence of a pineal foramen)

are characteristics found within the scaloposaurids.

Stlpholestes yackae (Broom, 1948) from the Cistecephalus zone is only slightly

larger than the new specimen, and has the same general proportions. There

is an enlarged canine and a small anterior canine, but there are thirteen post-

canines (as against eight postcanines in our specimen) and the first canine is

not separated from the incisors and large canine by diastemas. The postorbital

is figured as being complete (although Broom was uncertain of this). Further

points of difference are the large prefrontal and the apparent absence of a

postfrontal in Silpholestes.

Among the other scaloposaurids Scaloporhinus (Boonstra, 1953) has a

similar large ornamented angular and short prefrontal, while Ictzdostoma

(Broom, 1912, 1932) resembles our specimen in its dentition and the presence

of a postfrontal. Ericiolacerta (Watson, 1931) is also from the Lystrosaurus zone.

On the basis of these similarities and differences, it seems justified to

regard our specimen as representative of a separate genus in the family

Scaloposauridae, and I propose for it the name ortllodontops gracilis, gen. et

sp. nov. orillodoniops is probably related to Silpholestes and can be taxonomically

defined as follows:

Class Reptilia

Subclass Synapsida

Order Therapsida

Infraorder Therocephalia

Family Scaloposauridae

ZORILLODONTOPS n.gen.

Diagnosis: Skull small, lacrimal meets nasal, postfrontal present, postorbital

bar complete but slender, 6 incisors, 2 canines of which posterior one is enlarged,

8 postcanines, reflected lamina large with radiating ridges.

188 ANNALS OF THE SOUTH AFRICAN MUSEUM

Zorillodontops gracilis n.sp.

Holotype: Skull with major part of postcranial skeleton, S.A.M. No. K1392.

Horizon and locality: Lystrosaurus zone, Edenville, Orange Free State, South

Africa.

Diagnosis: As for genus.

If Romer’s (1956) classification is followed, Xorillodontops would be included

in the family Silpholestidae of the infraorder Bauriamorpha.

ACKNOWLEDGEMENTS

I am indebted to Dr. P. J. Rossouw of the Geological Survey of South

Africa for information regarding the geological horizon from which the fossil

was recovered, and to Mr. D. Pansegrouw, who presented the specimen to

the South African Museum.

The Trustees of the South African Museum thank the South African

Council for Scientific and Industrial Research for a grant in aid of publication.

SUMMARY

The skull and major part of the skeleton of a small, new Lystrosaurus zone

therocephalian, <orillodontops gracilis, is described. It is concluded that

Korillodontops is a scaloposaurid and is unusual in having a postfrontal. The

new genus is in several respects similar to Szlpholestes, a small scaloposaurid

from the Cistecephalus zone.

REFERENCES

Boonstra, L. D. 1934. A contribution to the morphology of the mammal-like reptiles of the

suborder Therocephalia. Ann. S. Afr. Mus. 31: 215-267.

Boonstra, L. D. 1953. A new scaloposaurian genus. Ann. Mag. nat. Hist. (12) 6: 601-605.

Broom, R. 1912. On some new fossil reptiles from the Permian and Triassic beds of South Africa.

Proc. zool. Soc. Lond. 1912: 859-876.

Broom, R. 1920. On some new therocephalian reptiles from the Karoo beds of South Africa.

Proc. zool. Soc. Lond. 1920: 343-355.

Broom, R. 1932. The mammal-like reptiles of South Africa and the origin of mammals. London:

Witherby.

Broom, R. 1948. The skeleton of a very small therocephalian. Ann. Transv. Mus. 21: 39-41.

Haucuton, 8. H. & Brink, A. S. 1954. A bibliographical list of Reptilia from the Karoo Beds

of Africa. Palaeont. afr. 2: 1-187.

Romer, A. S. 1956. Osteology of the reptiles. Chicago: University Press.

Romer, A. 8. 1966. Vertebrate paleontology. 3rd ed. Chicago: University Press.

Watson, D. M. S. 1931. On the skeleton of a bauriamorph reptile. Proc. zool. Soc. Lond. 1931:

1163-1205.

INSTRUCTIONS TO AUTHORS

Based on

CONFERENCE OF BIOLOGICAL EDITORS, COMMITTEE ON FORM AND STYLE. #960.

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. (73 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)

BuLiLoucuH, 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. 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, b; Liste: 11. Turton, 1932: 8o.

- ‘

;

}

, a tee a '

: - /

}

| :

1 Oe te Bis %, rise Rb A Wee

i h LPR Sa VIN ay wh igh Ai

}

Si ; ; ] ey | |

OC RN Bart at ree who arits 1

' it f bv eh ice i US ae et eee) wed

? : F v a; ed ys HEB ida j

: / va) iy Gacy eRe aia

\ / i j ft :

; \ my. 8 hsv

fs Diu '

" s . wae

‘ : 4 : By |

| | | mer bry partite

| f : oH ey

te } “y * Hh QUAI

a , wi ,

. ‘

ere wy a UA saclay ll ei

r . ’ ; ; 0) Bae ta dh Be wey ea vy ae .

‘a bt) | ) m

mn fiat hs WS tng lg

y Fa

\ fhe

| a’ s)

5 ; 4

hes f A A ee

j » Oba e iia va

Hi i ‘i } mn 4; { thy ah wy) , he nif Servet if

; t koh Ee kine o> CAG

: md ; PME Han ht pa ae Pe vane + ht “ae tai

| ; 1A NE I RS RM as ae

| |

}

: ; i

| : (

| j

| ) 7

ao /

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 52 ~ #&2Band

July 1969 Julie

Part 9 Deel

ANN HSON,; 4p

OCT8 1969

PLEISTOCENE MOLLUSCS FROM THE

NAMAQUALAND COAST

A.J. GARRINGTON & B. F. KENSLEY

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—p..), 5(2, 5, 7-9),

6(1, t—pi.), 71, 3), 8, 9(1-2); ro0(1—3);

I1(1-2, 7s t.—p.i.), aI, 24(2), 275 31(I-3), 38,

44(4).

Price of this part/Prys van hierdie deel

R3.00

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

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST

A. J. GARRINGTON & B. F. KENsLEy

South African Museum, Cape Town

(With plates 18 to 29 and 11 figures)

[ MS. received 1 October 1968 |

CONTENTS

PAGE

Introduction . : : ~ 169

Succession ‘ é : . 190

Systematic discussion . : lot

Acknowledgements. : 5 BED,

Summary. ‘ , : OED

References ; : 3 228

INTRODUCTION

In the course of an examination of the Tertiary to Recent sediments of

the Namaqualand coast, being carried out by one of the authors (A.J.C.), a

collection of fossil molluscs was assembled from the Pleistocene horizons

encountered in the area.

The purpose of this paper is to introduce and describe some twenty species

from this collection, including forms new to the South African palaeontological

literature.

In the case of a number of these species, the present geographic ranges of

the genera to which they are assigned suggests that, in the Lower Pleistocene

at least, the Namaqualand coast was washed by waters markedly warmer than

those which presently bound this shore.

In the areas examined, the local Archean metamorphics are overlain by

a succession of marine, fluviatile and terrestrial sediments of Tertiary to

Recent age.

The marine elements of this succession extend some three to five miles

east of the present coastline and are the result of a series of geologically fairly

recent transgressions by the sea on to the land.

The oldest sediments present, fluviatile clays, silts and sands occupying a

system of partly confluent channels deeply incised in the gneissic bedrock, are

succeeded by isolated remnants of a profoundly indurated phosphatic siltstone

and sandy clay horizon carrying an abundance of shell casts and moulds.

Overlying these fossiliferous silts is a succession of compact, poorly

cemented, marine sands and gravels associated with a series of high-level

beaches, developed during Pleistocene fluctuations of sea-level resulting from

the periodic wasting and renewal of high latitude, continental ice-sheets.

That this succession of marine sands is of Pleistocene age seems certain. In this

area, where there is no evidence of tectonic instability in Pleistocene-Recent

189

Ann. S. Afr. Mus. 52 (9), 1969: 189-223, 12 pls, 11 figs.

190 ANNALS OF THE SOUTH AFRICAN MUSEUM

times, the series of fossil beaches, testifying to repeated marine transgression

during which sea-levels reached altimetric maxima of 75-90 (oldest), 45-50,

29-34, 17-21, 7-8, 5 and 2 (youngest) metres, accords very nearly with the

well documented succession of Pleistocene strands along the Atlantic coast of

Morocco (Biberson, 1963; Butzer, 1966).

Although the coarse, porous, marine sands of the 75-90 m transgression

have, so far, proved to be barren of animal remains, the succeeding beaches

are abundantly fossiliferous. ‘The oldest of these shell-bearing beds, laid down

during the 45-50 m inundation, has yielded more than 120 species of molluscs

of which only seventeen of the twenty forms here described, plus the previously

recorded Chamelea kriget Haughton and Fissurella robusta Sowerby, are not found

living in the seas surrounding the coasts of South Africa. |

The stratigraphic correlation alluded to earlier, the high proportion (84%)

of species extant in the present seas, the low degree of lithification and the

extraordinary ‘fresh’ appearance of the invertebrate remains together militate

against the acceptance of an age other than Pleistocene for these sands and

gravels laid down during the 45-50 m transgression. |

Comparison with the Mediterranean and Moroccan marine Pleistocene

successions, based solely on altimetric evidence, suggests a date equivalent to

the Maarifian (Morocco) or Milazzian (Mediterranean) for this high sea

stand.

SUCCESSION

Thickness Suggested age

Loose surface sand ee rh ae oe o-4 m

— unconformity —

2m Transgression complex

Highly fossiliferous sands and gravels; berm sands. . 2—7-+m

Recent

— unconformity —

7-8 m Transgression complex

Stabilised berms; calcareous, often highly garneti-

ferous marine sands over well-developed boulder

gravels .. aS3 fe a H, nat ae 3-5-+m

— unconformity —

Terrestrial sands

' Sheet wash deposits .. st 4 is a 2-7 m

— unconformity —

17-21 m Transgression complex

Coarse sands and grits over massive, basal boulder

gravels .. ae bis se, ae Ae ke 10-15 m

— unconformity —

29-34 m Beach

Thin, discontinuous, shelly gravel beach .. a, uncertain

— unconformity —

45-50 m Transgression complex

Aeolianite and coarse regressive facies overlying

locally fossiliferous, fine-grained, transgressive sands 5-25-+m Lower Pleistocene*

— unconformity —

Upper Pleistocene

(to Recent in part)

Middle Pleistocene?

— unconformity —

5 m Transgression complex

_ Slightly calcareous sands and shelly granule gravels 2—3m

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST IQI

75-90 m Transgression complex

Regressive aeolianite overlying coarse, marine sands

and thin gravels 3E * ots ute ay 35-+-m Basal Pleistocene?

— unconformity —

Fossiliferous phosphatic siltstones a sg as o-I m Pliocene

— unconformity —

Fluviatile beds

Linear deposits of clays and clayey sands. Unfossili-

ferous 2h :

— unconformity —

Basement gneiss .. ae te ke af a

1 Subdivision of the Pleistocene after Butzer, 1966.

Details of the succession given above and dating of the component horizons are based on the

results of the geological survey at present being carried out by A.J.C. It is hoped that a fuller

description of the geology and palaeontology of this area will be given at a later date.

up to 20m Mio-Pliocene ?

Archean Complex

The fossil species described in this paper were collected from the three

localities indicated in figure 1. With the exception of a new species of Fissurella,

all were recovered from sands deposited during the 45-50 m marine

transgression.

The extreme dissimilarity in lithological character of sediments from the

45-50 m and the 17-21 m inundations is a notable feature. Typically, the

shelly fore-beach deposits of the 45-50m transgression are constituted of

extremely fine-grained, moderately rounded and sorted sands and silts. Fine

to laminar bedding structures are apparent and the faunule includes a strong

complement of sand-dwelling bivalves of the families Mactridae, Tellinidae and

Veneridae. The sedimentary characteristics of these beds, which locally yield

an abundance of anew species of Donax, and the prevalence of burrowing bivalves

point to a quiet depositional environment with low rates of both provenance

and accumulation of sediment.

Conversely, the extremely coarse sediments of the 17-21 m transgression,

granular, well sorted and rounded, frequently cross-bedded and with a paucity

of sand-dwelling bivalve species, suggest that a more rigorous, abrasive

environment with a high rate of sediment accumulation was current during

the deposition of these beds.

In at least one species, it is suggested that adaption to the profound shift

in the nature of the environment, inferred from the change in lithology, has

progressed so far and has engendered such physical change that the final product

of these adaptive trends should be accorded a separate, specific identity.

SYSTEMATIC DiscussION

Key to abbreviations

E5= leneth

B. = breadth

D. = greatest diameter

A. = altitude

W. = width

192 ANNALS OF THE SOUTH AFRICAN MUSEUM

Location 1.

Oranjemund

x— Location 2.

Swartlintjies

Fm.

Hondeklip (o

20 miles

Hondeklip Bayt ; Location 3.

a «7 Strandtontein

Fm,

Brand’s Bay

Access roads.

MOjOr ———

minor ——-——

Lambert’s Bay?

ATLANTIC

OCEAN

Vredenburg if

Pizistocene Molluscs

from the

Namaqualand Coast”

Locality Sketch

Cape Town

Fig. 1

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST 193

Family Turridae

‘Turris’ nigrovitta n.sp.

Pl. 18

Location I. 45-50 m Transgression complex.

Fore-beach environment.

Description

Protoconch eroded, possibly 14 whorls. Postnatal whorls 5. 1st whorl with

7 axial ribs, 5 spiral lirae, 2nd whorl with 8 axial ribs, 5-6 lirae, 3rd whorl

with 8 axial ribs, 7-8 lirae, 4th whorl with 7 axial ribs, 7-8 lirae, body whorl

with 7-8 axial ribs, spiral lirae obscured by erosion. Axial ribs broad. Lower

body whorl and anterior canal with about 11 lirae. No sinus on outer lip.

Inner area of outer lip with 8 or 9 elongate plicae. Anterior canal subequal to

aperture, latter equal in length to spire.

Named from location — ‘Swartlintjies’, Afrikaans — ‘black ribbons’.

Material

Eialotype: S:A.M. Ki443°.L..27mm B. 12mm

Paratype: S.A.M. K1444 20mm 8mm

Remarks

With only the shell available, it is difficult to decide on generic status for

this species. The main diagnostic features relating to the shell are the shape

of the outer lip sinus, and the anterior canal. As there is no trace of a sinus in

the present shells, choice of a subfamily is made almost impossible. The axial

sculpture is similar to that found in Turris saldanhae Barnard and Surcula scalaria

Barnard, but there are no other similarities. Without more material being

available, the species is placed in the genus ‘Turris’ merely for convenience.

Family Fasciolariidae

Fasciolaria sp.

Piss

Location 1. 45-50 m Transgression complex.

Beach environment.

Description

Aperture slightly shorter than spire. Shell 54 whorls, protoconch and first

3 whorls eroded. 9-10 spiral lirae commencing on latter part of 3rd whorl,

I1—12 on 4th whorl, 12-15 on 5th whorl. Outer lip with 28-2 lirae, internally

plicate in the larger shell. Exact number of lirae on outer lip difficult to deter-

mine as lirae divide into two at upper and lower portions of whorls and each

may again divide. Latter divisions not as obvious in middle portion of whorl

194 ANNALS OF THE SOUTH AFRICAN MUSEUM

where erosion has occurred. Profile of whorls smoothly convex, lower two

with very slight shoulder. No columella pleats present, no axial sculpture

apart from irregular growth lines.

Maiterial

S.A.M. Ki445 L. 61-0mm_ B. 26-5 mm (outer lip broken)

S.A.M. K1446 66-0 mm 330mm (very worn)

Remarks

In general shape, these shells are typical of the genus Fasciolaria and

resemble F. lugubris Reeve very closely. There are, however, several points of

difference. The columella is not as sinuous anteriorly as in F. lugubris. The

latter usually has several columella pleats in the region of the anterior canal

while near the suture of the body whorl with the preceding one, there are

usually several small pleats on the columella. In the present shells there is

only a slight nodule on the upper columella. The sculpturing differs from F.

lugubris, which usually has the outer lip with 22 broad spiral lirae, alternating

with finer lirae. The broad lirae are each made up of 5-7 fine lirae while the

narrow lirae are made up of 2-3 finer ones. The 4th whorl has 9 broad lirae.

The present species, however, has 28-32 lirae on the outer lip, 11-12 on the

4th whorl, while each lira is made up of 2 finer ones, or, if the lirae are split,

4 each.

The present shells are perhaps a form of F. lugubris which has been recorded

from the Pleistocene deposits of the Saldanha Bay area, but without more

material being available, their status is difficult to determine.

Fusus fauret Barnard, 1959

7 Pl. 18 |

Fusus faurei Barnard, 1959: 94.

Location 1. 45-50 m Transgression complex.

Beach environment.

Material

One specimen: L. 42:2 mm__siB.. 20:5 mm

Previous records

Alive, off Cape Point.

Remarks

The present shell agrees almost exactly with the description of F. faurei

Barnard, the only difference being in the number of axial ribs. The penultimate

whorl of F. faure: Barnard has 14-15 axial ribs, while the present shell has 12-14

ribs. The earlier whorls also have fewer ribs. The spiral sculpture is, however,

almost identical and there can be no doubt that this shell is a form of F. faurei

Barnard.

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST 195

Family Nassidae

Nassa litorafontis n.sp.

Pie ie

Location 3. 45-50 m Transgression complex.

Early transgression beach.

Description

Protoconch 1-14 whorls, 44-5 post-natal whorls, first four post-natal

whorls with strong axial sculpture, about 20—24 axial ribs on 1st whorl, crossed

by 4 spiral lirae. and whorl with 25-30 axial ribs, 4 main lirae, very fine

intermediate lirae starting between major ones. 30-35 ribs on 3rd whorl, 6—7

lirae. 4th whorl with about 28 axial ribs, becoming broader and rounder and

less distinct, finally obsolete on body whorl. Latter with about 25-27 lirae on

outer lip, upper ones broad, lower ones narrow. Axial ribs in general stronger

on the lower part of the whorls than on the upper. Aperture slightly shorter

than spire. Inner margin of outer lips smooth except near junction with

columella, where a single nodule is followed by a very low, smooth ridge.

Columella smooth; just below junction with body whorl, an acute raised ridge,

just opposite the nodule of the outer lip. Anterior end of columella carinate.

Outer lip excavate at anterior junction of columella.

Named from location—coastal farm Strandfontein —‘Strandfontein’,

Afrikaans — ‘beach fountain’.

Material

Holotype: S.A.M. Ki441 L. 28-:omm B. 12°5mm

Paratype: S.A.M. K1442 22-0 mm 10-0 mm

Remarks

In general shape, the present species most closely resembles one of the

South African representatives of the genus NV. bicallosa Smith, recorded from

the Natal coast, although the figure given with the original description (Smith,

1876) is of a more squat shell. The sculpture of axial whorls is similar but

N. bicallosa has fewer ribs (13-17, as opposed to up to 35 in W. lziorafontis). In

N. bicallosa there is both an external parietal callus and an internal callus,

while the columella is denticulate, and the outer lip plicate. N. litorafontis has

a single callus, a smooth columella, and no plications on the outer lip. In the

latter species, the single ridge near the fusion of the outer lip and the columella

is more developed than in W. bicallosa.

Latiaxis sp.

Pl 20

Location 2. 45-50 m Transgression complex.

Fore-beach environment.

196 ANNALS OF THE SOUTH AFRICAN MUSEUM

Description

Shell broken, very eroded, 4 whorls extant. Aperture longer than existing

spire. Profile of whorls with prominent angular shoulder. Sutures undulate.

Very faint indications of axial ribs on earlier whorls. Strong parietal callus,

narrow umbilicus. Rostrum present, but very worn.

Material

S.A.M. Ki447 L. 62:5mm_ B. 37mm (outer lip missing)

Remarks

This specimen agrees with Thiele’s (1929) definition of the genus. Specific

status, however, cannot be accorded this shell, due to its very incomplete state.

In general shape it resembles Latzaxis tortilis H. & A. Adams, which has been

recorded once in the South African region, from west of Cape Point. None of

the characteristic axial and spiral sculpture of this species is discernible in the

present shell. The undulations of the sutures may seem to indicate that there

were fairly strong axial ribs present. The genus appears to be characteristic of

the Indo-Pacific region. Barnard (1959) remarked on the unusualness of the

occurrence of L. tortilis from the west coast of the Cape Peninsula. This record

from the Namaqualand coast wou!d serve as yet another indication of the

presence of a warm-water fauna on this coast during the period under discussion.

Family Muricidae

Tritonalia bonaccorsii n.sp.

Pl. 19

Location 1. 45-50 m Transgression complex.

Back-beach environment.

Description

Spire slightly longer than aperture. Protoconch of 2-24 whorls, 4-44

post-natal whorls. First 3 whorls bicarinate, the 2 carinae formed by 2 strong

spiral lirae. 4-5 fine lirae between upper carina and suture. 3rd whorl with 3

fine lirae between the 2 major lirae, 2 fine lirae between lower carina and suture.

Cancellate sculpture on first three whorls formed by axial ribs crossing lirae.

Ist whorl with 10-12 axial ribs, 11-12 on 2nd and 3rd whorls. Lower carina

on last whorl not stronger than rest of lirae, body whorl therefore unicarinate.

Body whorl with 9 lirae above the carina, 16-19 below. Outer lip with about

8 plicae on inner margin. Short open anterior canal, strong parietal callus,

very slight umbilicus, slight rostrum with 3 worn squamae.

Named after donor, Dr. G. Bonaccorsi.

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST 197

Material

Holotype: S.A.M. K1436 L. 34-2mm _ B. 17mm

Remarks

The strong spiral sculpture, which forms a cancellate decoration with the

axial ribs, is characteristic of several genera of the Muricidae. The tall spire,

egg-shaped aperture, very convex body whorl, and short anterior canal bent

to the left would seem to indicate the genus T7rztonalia. The present shell lacks

the strong axial sculpture of 7. decussata Gmelin or T. fasciata Sowerby of

west Africa and 7. kienert (Reeve) and T. scrobiculata (Dunker) of the South

African region. Similarly, T. bonaccorsit lacks the axial sculpture of T. purpuroides

(Reeve), known from the Pleistocene -deposits of Saldanha Bay. The early

whorls of this latter species, however, resemble the earlier whorls of the present

shell in possessing a cancellate sculpture, and two strong spiral lirae per whorl.

T. sperata (Cossman) of the east coast has a much longer anterior canal; this,

and a nodulose axial sculpture, separates it from the present shell. The latter

most closely resembles 7. puncturata (Sowerby), recorded from the east coast

as well as from the west coast of the Cape Peninsula. This latter species

occasionally shows a tendency towards carination, particularly in the earlier

whorls, where the sculpture is cancellate. The aperture of T. puncturata is

usually more elongate, while the penultimate and body whorl never have so

strong a carina as in T. bonaccorsit.

There is a superficial similarity between the present shell and Cancellaria

lyrata (Brocchi), but the former lacks the columella pleats and prominent spined

axial ribs of the latter species. There is also a slight, superficial similarity

between Latiaxis rosaceus Smith and the present shell but the former is more

squat, and has blunt, nodulose axial ribs. From the above, it is obvious that

some doubt exists as to the generic status of this shell. Without more material

being available, greater certainty is not possible.

NAMAMUREX n.gen

Description

Shell of 4 or 5 whorls. Varices 3 per whorl, connecting with those of

previous whorls. No sculpture between varices. Outer lip with 8 submarginal

nodules, one prominent erect tooth near base of nodules. Aperture longer than

spire. Anterior canal open, slightly shorter than rest of aperture.

Generic name in part a contraction of the general location — Namaqualand.

Discussion

The presence of strong varices, a well-marked anterior canal, a toothed

outer lip, and a prominent tooth on the lower outer lip, make it seem certain

that this shell is a member of the family Muricidae.

The presence of a spine on the outer lip is characteristic of several genera

198 ANNALS OF THE SOUTH AFRICAN MUSEUM

of the Muricidae, viz. Acanthina, Ceratostoma, Pterorytis, Jaton.

The genus Acanthina is known living from the west coast of North and

South America, especially from the Californian coast. Apart from the presence

of the spine on the outer lip, the present shells bear little resemblance to any

of the forms of Acanthina, all of which are rather Thais-like.

The genus Ceratostoma, as defined by Vokes (1964), includes the species

C. nuttalli (Conrad) from the Californian coast, which the present shells most

closely resemble. C. nuttallc has three well-developed almost foliose varices per

whorl, and in addition, single elongate axial nodules between the varices. The

absence of these nodules in the present shells, as well as the absence of the

prominent spiral ridges, excludes these shells from the genus Ceratostoma. ‘The

latter genus invariably has a closed anterior canal, quite different from

Namamurex. The genus Ceratostoma is a Pacific form, being found on the west

coast of North America, and in the Japanese region (Hall, 1959).

The genera Jaton and Pterorytis, both of which have a denticulate outer

lip, are typical Atlantic forms. The former, having 3 or 4 wide transverse

plications separated by deep, slightly striated grooves, contains the single

species, 7. decussata (Linnaeus) from west Africa (vide Vokes, 1964: 21).

Pterorytis, defined as being fusiform, with 6 prominent recurved foliated ribs,

and having a closed anterior canal, is known only from the Miocene and

Pliocene of the Atlantic coastal plain of the United States. The present shells

are thus eliminated from both these genera, and as they conform with none of

the recognized general of the Muricidae, it is necessary to erect a new genus,

as defined above.

Namamurex odontostoma n.sp.

Pls 20-21

Location 2. 45-50 Transgression complex.

Fore-beach environment.

Description

Shell of 4 or 5 whorls, protoconch eroded. 3 prominent varices per whorl,

continuous with those of previous whorls. Outer lip, at formation of varix,

having 8 slightly raised nodules on inner margin. Near base of latter nodules,

a prominent erect spine. Latter spine also visible on two earlier varices, slightly

embedded in following portion. A shallow groove, becoming almost a line,

running from tooth along following portion of whorl. Profile of earlier whorls

with slight shoulder, later whorls becoming smoothly rounded. No sculpture

other than faint growth lines.

Material

Holotype: S.A.M. K1437 L. 55-omm B. 30-0mm

Aperture L. 36°4mm

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST 199

Other specimens: L. 54°0mm_ B. 36-0mm

64°5 mm 40°5 mm

51-omm 35°5 mm

39:0 mm 25-0 mm

(Latter two specimens with 4 varices per

whorl on earlier whorls, varices not

meeting, shoulder between varices

almost joining to form a ridge.)

S.A.M. 9925 L. 68:omm OB. 43:0mm

(Collected by Haughton, very eroded,

3 varices per whorl.)

Family Eratoidae

Subfamily Eratoinae

Hespererato oppenheimeri n.sp.

Pl. 19

Location 3. 45-50 m Transgression complex.

Early transgression beach.

Description

Shell small, sub-conical, inflated posteriorly, attenuated anteriorly.

Posterior extremity of outer lip angular and projecting.

Spire small, eroded in the present specimens, but quite definitely pro-

jecting above the terminal whorl; apparently non-granulate.

Surface smooth, possibly glossy in life; non-sulcate.

Outer lip sharply incurved, attaining its greatest thickness in the central

region; externally thickened at margin. Incurved marginal area bearing a

series (czrca 10) of equal, regularly spaced, plicate ridges, giving a denticulate

aspect to outer lip.

Posterior canal very poorly defined, virtually obsolete; ill-defined oblique

ridge marking the left border.

Aperture slightly curved, rather narrow posteriorly, becoming considerably

wider anteriorly before being constricted by the terminal flexure of the

columella lip.

Anterior edge of columella lip obliquely truncated, bounded by two

closely spaced terminal ridges.

A series (circa 8) of elongate subequal plicae, regularly spaced, traverses

the columella lip of the labrum. Plicae approximately at right angles to the

axial plane of the shell, their alignment markedly dissimilar to the oblique

trend of the terminal ridges, tending to be discontinuous along the central

portions, forming two parallel rows of elongate columella and labral denticles.

Between the most anterior of these denticles and the terminal ridges, a distance

200 ANNALS OF THE SOUTH AFRICAN MUSEUM

of approximately one fifth the length of the inner lip, the columella is smooth

and non-denticulate. The series of plicae persists to the posterior end of the

columella lip, being there less strongly developed, the columellar denticles

becoming subordinate to those on the labrum.

Extreme anterior area of shell squarely terminated, no obvious concavity,

fossula obsolete.

Named after the late Sir Ernest Oppenheimer, past Chairman of Anglo

American Corp. S.A.

Material

Holotype: S.A.M. Ki428 L. 12mm _ B. 8mm

Paratype: S.A.M. K1429 13 mm g mm

Other specimen: 10mm 7mm

Remarks

The general shape of the shell, the numerous transverse denticles arming

the columella and labrum, the character of the aperture, which is not axial,

and the form of the spire, which projects above the terminal whorl, suggest

that the present species should most properly be assigned to the sub-family

Eratoinae (Eratoidae) rather than to the genus Persicula (Marginellidae), or to.

the sub-family Triviinae (Eratoidae).

In his monograph on the Eratoinae, Schilder (1933) recognizes four genera

and three sub-genera. Regarding the generic position of the present species

within the sub-family, the following characters of this species appear to be of

diagnostic importance:

(1) dorsum smooth;

(11) fossula almost obsolete;

(111) two terminal ridges on columella lip;

(iv) columellar denticles transverse, quite definitely not parallel with terminal

ridges;

(v) anterior columellar denticles develop some distance away from the

terminal ridges.

Characters (i)—(iv) suggest that the present species best fits into the Benne

Hespererato Schilder.

It is notable that the hiatus between the terminal ridges and the columellar

denticles (v) seen in the present species is a character held in common by a

number of species within the genus Hesperato and one which is not apparent in

any of the three other genera in this sub-family. Similarly, a sinuous columellar

outline and an outer lip bearing fewer but larger denticles are features borne

by several species within this genus.

Of the 99 fossil and living species listed by Schilder, none is exactly similar

to the present form. The strongest resemblance is to H. vitellina (Hinds), a

Californian form, and the type species of the genus. Representatives of this

genus are known from Miocene to Recent and range from the west coast of

the Americas to France and Italy.

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST 201

Fig. 2. Ventral views: a. Calyptraea viridarena n.sp. Holotype. b. C. helicoidea (Sowerby). c. C.

aurita striata n.subsp. Holotype. d. C. aurita Reeve.

It is of interest to note that the local contemporary form Proterato (P.)

sulcifera (Sowerby) is a member of the sub-genus Proterato Schilder, found only

in the Indo-Pacific and Australasian areas. The presence of apparently generi-

cally different forms in Lower Pleistocene and Recent seas around the South

African coasts and the surprising distribution of the genus Hespererato (west coast

of America, Antilles, and Europe) may perhaps suggest that the phylogenetic

aspects of one, at least, of these generic divisions are not of profound significance.

Family Calyptraeidae

Calyptraea aurita striata n.subsp.

Pl. 22 & Fig. 2

Location 2. 45-50 m Transgression complex.

Fore-beach environment.

202 ANNALS OF THE SOUTH AFRICAN MUSEUM

Description

Basal circumference circular. Protoconch eroded. Post-natal whorls

indistinct, possibly 3. External sculpturing consists of fine, oblique, radiating

ribs. Shelf loosely reflexed upon itself at the columella, forming a wide elongate

umbilicus. Margin of the shelf slightly incised at its fusion with the shell

periphery.

Material

Holotype: S.A.M. Ki433 D.t9o°6mm A. 8-7 mm

Other specimens: 24°8mm I0‘O mm

24:0 mm 11-5 mm

Remarks

The only difference between the present shells and C. aurita Reeve lies

in the external sculpture. The usual sculpture of C. aurita consists of irregular,

radial rows of hollow, squamiform prickles. This subspecies lacks these prickles

but has very fine radial ribs. While the present subspecies is seen to be fairly

common, no examples of C’. aurita have been recovered from any of the locations

visited. C.. aurita has been recorded alive from the warm waters of Algoa Bay

and St. Francis Bay, while fresh dead shells have been reported from Struis

Bay, Cape Recife and Cape St. Blaize.

Calyptraea viridarena n.sp.

Pi 2o%sechiee

Location 2. 45-50 m Transgression complex.

Fore-beach environment.

Description

Basal circumference of shell circular. Protoconch eroded. Post-natal

whorls 3-4. External sculpture consisting of strong oblique radiating ribs,

close together. Internal shelf reflexed on itself at junction with columella,

forming a slit-like umbilicus. Reflexed portion about 0: 3-0-4 of length of shelf

at its widest. Outline of shelf curved. Shelf somewhat incised at peripheral

junction.

Name descriptive of the colour characteristic of the enveloping sands.

Material

Holotype: S.A.M.Ki499 D) 27 -Gimaa YU. ror aims

Other specimens: 27°5 mm 12-0 mm

29°9mm 13-0 mm

22-5 mm 9°5 mm

20'0 mm 7°-0mm

Remarks

The present species differs significantly from living southern African

representatives of the genus. The sculpturing is very similar to that of C.

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST 203

helicoidea (Sowerby), but with the radiating ribs relatively stronger and closer

together. The whorls are less clearly demarcated than in C’. helicordea, which is

a proportionally higher shell, with a more obvious spire and with a better

developed protoconch. The following measurements indicate the relative

differences in height of shells of similar size of the two species:

C.. helicoidea D.27-0mm: A. 14-0mm

23-0mm I1°3mm

C.. viridarena n.sp. 27°5 mm I2-Omm

22°5 mm 9°4mm

The structure of the shelf and columella more closely resembles that of C.

chinensis (Linnaeus) than C. helicozdea. In the latter, the reflexed portion is very

short and forms a wider umbilicus than in the present species, while there is

no incision at the fusion with the periphery, as in the present species.

Family Trochidae

Clanculus murray n.sp.

Pl; 2oné Pig..9

Location 2. 45-50 m Transgression complex.

Fore-beach environment.

Description

Shell trochoid in form, somewhat depressed; periphery angular, accen-

tuated by strong carina, carinate sculpture lending a turreted aspect to the

outline.

Eroded protoconch nucleus, plus apparently four whorls.

Suture non-canaliculate, umbilicus open and deep.

A single carina on the early whorls, located slightly below the midline,

accompanied on the penultimate and last whorls by a second carina, developing

from the suprasutural lira on the penultimate whorl and forming the shell

periphery on the last whorl. Carinae equal to sub-equal in strength; where

subequal, the upper, non-peripheral member always the stronger. Carinae

composed of closely spaced oval to sub-rectangular granules with long axes

sub-parallel to the vertical axis of the shell; granules slightly inclined on earlier

whorls. Granulate radial lirae apparently present on all whorls; granules sub-

circular on earlier whorls becoming more elongate along the vertical axis and

eventually oval to sub-rectangular on the final whorl. In the largest specimens,

areas between lirae with raised threads. On the 2nd whorl, approximately 5

lirae and a single carina; on the grd, 5-6 lirae and 2 carinae; 4th whorl 4-5

supra-carinate lirae, 2 carinae separated by 2-3 lirae with approximately 7

lirae on base of shell.

Specimens show a tendency for the body whorl to drop away from the

periphery of the preceding whorl.

204. ANNALS OF THE SOUTH AFRICAN MUSEUM

Outer lip of aperture strongly plicate, bearing 10-15 dentiform plicae.

A prominent, ridged denticle present at the anterior end of the inner lip,

protruding into the aperture; a second, smaller denticle, also pointing into the

aperture on the posterior portion of the inner lip. Two small denticles, often

elevated on a narrow ridge running into the umbilicus, on the median portion

of the inner lip, pointing away from the aperture. On the larger specimens,

dentiform plicae present on the margin of the umbilicus.

Named after Dr. L. G. Murray, Consulting Geologist, Anglo American

Corp. S.A.

Material

Holotype: -S:A.M:; - Ki4g4)) 1. 49-6 mn) AS wae ema

Paratype: S.A.M. K1435 I7-O mm I4-Omm

Remarks

The trochoid form, open umbilicus and denticulate aperture suggest that

this species should be assigned to the genus Clanculus, having strong affinities

with the local species C. miniatus (Anton).

The present species differs from all the South African representatives of

the genus in having a generally depressed form and a strongly turreted outline.

The presence of strongly developed carinae distinguishes the present species

from the non-carinate C. atricatena Tomlin and C. puniceus (Philippi), both of

which exhibit a more complex and pronounced plication of the aperture lips.

The non-canaliculate suture, amongst other features, serves to distinguish the

present species from C. waltonae Sowerby, and C. mixtus Smith. In C. miniatus,

carinate forms are common. However, such forms carry only a single carina

°o Clanculus miniatus (Anton)

From S.A.Museum collection.

e Clanculus murray! n.sp.

Gradient of field boundary =1

> 3

Greatest diameter mm.

Fig. 3. Comparison of altitude/diameter ratios in Clanculus miniatus (Anton)

and C. murrayi n.sp.

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST 205

on the last whorl, at the shell periphery. In some of the specimens of C.. miniatus

examined, the mid-whorl granules on the last whorl are more strongly developed

than their neighbours but are never sufficiently prominent to produce a bicari-

nate aspect. It is notable that in the present species, where the carinae are

subequal, the mid-whorl member is always the stronger.

The depressed form of the present species in comparison with C. miniatus

is illustrated by figure 3 and is possibly linked to the fewer whorls borne by this

new species.

C’. villanus Phillipi, reported from Angola by Paes-da Franca (1960), is

another elevated member of the genus having a denticulate aperture but lacking

the single, large, prominent tooth on the anterior area of the columella lip.

The new form shows little affinity with the west African representatives

of the genus (C. kraussi Philippi, and C. guineensis Gmelin) noted by Nickles

(1950), both of which lack a prominent denticle on the inner lip and a strong

peripheral carina.

Calliostoma depressa n.sp.

Pl 28

Location 2. 45-50 m Transgression complex.

Fore-beach environment.

Description

Shell a wide cone, apical angle greater than 90°, profile straight. Protoconch

smooth, 2 whorls, 44-5 post-natal whorls. Profile of body whorl smoothly

rounded. Sculpture consisting of very fine growth lines and spiral lirae. Latter

formed by single strand of knobs connected by a low ridge. Body whorl with

19 lirae (counted at aperture). Penultimate whorl and preceding whorl with

7 lirae, first two postnatal whorls with 4-5 lirae, latter all of equal strength.

Each lira separated from the next by a definite regular space. Each knob

distinct, not touching the following or preceding one. Umbilicus closed, colu-

mella smooth.

Material

Holotype: S.A.M..Ki440 D. 21mm A. 15mm

Paratype: S.A.M. K1453 37 mm 30 mm

Remarks

Compared with the known living forms of this genus from southern

Africa, this species is a much flatter shell with fewer whorls, most of the other

species having 6-8 post-natal whorls. The sculpture, spiral lirae with knobs, is

typical of the genus. The present species, in general shape, most resembles

Calliostoma eucosmia Bartsch, particularly the ‘low’ form mentioned by Barnard

(1963). The latter, however, is a more elevated shell, with the lirae much

closer together and usually having strong lirae alternating with weaker ones,

206 ANNALS OF THE SOUTH AFRICAN MUSEUM

the knobs crowded on each other. C. eucosmia usually has a definite shoulder to

the last whorl, although some forms approach the present species in having an

almost smoothly rounded body whorl. The latter type of body whorl is charac-

teristic of C.. multiliratum (Sowerby), but the sculpture of this species is distinctive

and very different from C’. depressa. In the latter only a suggestion of a shoulder

is present where the outer lip joins the preceding whorl. C. africanum Bartsch

has sculpturing similar to the present shell, but with the knobs closer together.

In the older shells of C. africanum the knobs tend to coalesce to form a solid

raised lira. In C. depressa, the knobs remain distinct. There is little similarity

between this shell and any of the west African, Angolan, or Mocgambique

species. Apart from C. perfragile Sowerby, which has been taken alive from the

west coast of the Cape Peninsula, the genus is not known from the west coast

of South or South West Africa. Further north the genus is again encountered

in warm Angolan waters, where it is represented by C. granulatum.

Family Fissurellidae

Fissurella glarea n.sp.

Pl. 24

Location 3. 29-34.m Beach.

Description

Margin of shell egg-shaped, anterior end narrower than posterior. Lateral

margins ventrally convex. Foramen elongate oval, lateral margins slightly

concave, situated anterior to midline. Sculpture only of concentric irregular.

growth lines. No sign of radial sculpture. Dark radiating bands (circa 23-26)

imposed on lighter background, stretching from foramen to margin.

Name descriptive of the gravelly nature of the enveloping sediment.

Material

Holotype: S.A.M. K1438 L. 39‘-0mm OB. 27-5mm

Paratype: S.A.M. K1439 76-0mm 59°5 mm

Remarks

The following are the species of Fissurella which have been recorded from

the South African region:

F. robusta Sowerby, known only as a fossil from the Pleistocene deposits of the

west coast and Algoa Bay.

F.. mutabilis Sowerby, known living from the west coast, False Bay, east coast,

Madagascar, etc. Fossil from Pleistocene of Algoa Bay and Little Brak River.

F. natalensis Krauss, known living from the east coast, fossil from the Pleistocene

of Algoa Bay.

The present species most closely resembles F. robusta, of which several

undamaged specimens are available. This resemblance is most apparent in the

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST 207

young stages, older shell being obviously different. ‘The marginal outline is

very similar, the foramen is in the anterior portion of the shell in both species,

and the sculpturing is similar. There are, however, several differences. The

foramen in F. robusta has slightly convex sides, whereas in the present species,

the sides are slightly concave (this portion of the shell is particularly subject

to erosion, resulting in a variety of shapes, the true form being apparent only

in unworn specimens). The profile of F. robusta, from posterior margin to

foramen, is never straight. In younger specimens it is slightly convex, while in

older specimens, the shells are distinctly arched, almost ‘humped’. The posterior

profile of the present species is never curved. F. robusta is a proportionally

higher and heavier shell than the present species, which seems, if anything, to

become flatter with age. Young specimens of F. robusta occasionally show faint

radial lines. These are not apparent in the present species. The dark radial

bands of the outer layers of the shell of the present species are not apparent in

F. robusta, except in one specimen, which has a slightly curved posterior profile.

F. glarea differs from F. mutabilis Sowerby, which has radial costae, and

no convexity of the margins, i.e. the entire margin touches a flat substrate. The

shape of the foramen also differs. F. mutabilis does not have the dark radial

bands of the present species.

F. natalensis is a rugose shell with almost nodulose radial costae, and with

fine concentric lines, subordinate to the radial sculpture; all very different

from F. glarea. Further differences include the overall shape of the shell (F.

natalensis rests on the anterior and posterior ends of the shell), and the shape of

the foramen.

-F’, dubia Reeve, from Mocambique, also has radial costae and the foramen

is placed more or less centrally.

F. tanner Verril, from Angola, has fine radial lines and an almost circular

foramen. Both F. coarctata King and F. nubecula Linnaeus of west Africa have

strong radial sculpture.

Family Arcidae

Arca avellana Lamarck, 1819

Pl. 25

Arca avellana Lamarck, Barnard, 1961: 192; 1964: 369. Boshoff, 1965: 109.

Arca acuminata Krauss, 1848: 14.

Navicula kraussii (Philippi), Cox, 1930: 154.

Location 3. 46-50 m Transgression complex

Early transgression beach.

Description

Cardinal area flat, triangular, the angle at the umbo go°, marked with

divergent ligament grooves. Umbo eroded in all except youngest valve.

208 ANNALS OF THE SOUTH AFRICAN MUSEUM

Material

L. 44:00 mm (broken valve)

29°6mm

28-2 mm

28-0 mm

20°0 mm

19°6 mm

Previous records

Living: Inhaca Island, Inhambane, Bazaruta, Durban.

Fossil: Pleistocene: Redhouse, Zwartkops River.

Post Pliocene: Inhambane.

Pliocene: Zwartkops.

Remarks

Cox (1930) described an abnormal right valve of Navicula krausii (Philippi)

from the Pliocene deposits of Zwartkops, near Port Elizabeth. The present valves

(with the exception of the smallest) all agree with Cox’s description, in hinge

details, sculpture, and general shape. The smallest valve is a typical Arca avellana

without any distortion. All the complete valves show, to some degree, the

septum-like platform in the anterior corner, as was noted in Cox’s specimen.

The ventral marginal area is narrow and concave, forming a large byssal gape.

Boshoff (1965) notes that as this species frequents crevices, the valves are often

distorted.

This species, in its present geographic distribution, is restricted to the warm

waters of the Indian Ocean.

Arca (Acar) halmyrus n.sp.

Pl. 24

Location 3. 46-50 m Transgression complex.

Early transgression beach.

Description

Shell longer than high, umbo anterior to middle of shell. Fairly prominent

posterior ridge present. Well-developed concentric and radial sculpture.

Nearer the umbo, radial ribs more prominent than concentric lines. Anterior

7 or 8 ribs strong, increasing in thickness ventrally. Posterior 6 or 7 radial ribs

similarly prominent, the largest forming the posterior ridge. Where concentric

ridges cross anterior and posterior ribs, distinct knobs formed. Radial ribs of

mid-region fine, an almost cancellate sculpture formed with the concentric

growth lines. Intermediate radial ribs developing as growth proceeds. Cardinal

area very narrow, 5 narrow, oblique grooves in posterior region, ending just

posterior to umbo. Hinge slightly curved dorsally with 12 or 13 anterior teeth,

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST 209

the most anterior 4 or 5 being oblique. Median teeth reduced and almost

absent in some. 10 or I1 posterior teeth, almost all oblique. Inner surface of

valves with fine radiating lines. Inner margin of shell crenulate.

Named from the location — ‘Brak’, Afrikaans — ‘brack’ river.

Material

Holotype: S.A.M. Ki448 L. 26mm A. 15mm

Other specimens: 23 mm II mm

21mm 12mm

16mm 8 mm

Remarks

These shells are placed in the subgenus Acar as defined by Thiele (1929),

as there is a posterior ridge present, the sculpture being knobbed, the umbo

anterior to the midline of the shell, the dorsal plane small, the median hinge

teeth weak/rudimentary and the outer teeth oblique. The present species

differs from Arca (Acar) plicata Dillwyn, which it most strongly resembles, in

the following respects: the present shell is flatter than A. (A.) plicata and has

well-developed anterior and posterior radiating, nodulose ribs which A. (A.)

plicata lacks. In the region of the posterior ridge, A. (A.) plicata usually has

prominent squamae instead of the nodules of the present species. In the hinge

area, the dorsal plane of A. (A.) plicata, which is wider than in the present

species, has 10 or more oblique/elongate grooves, the more medial ones extend-

ing the entire length of the hinge area; A. (A.) halmyrus however, has only about

5 oblique grooves, none extending beyond the umbo. The angle of the posterior

ribs radiating from the umbo differs considerably in these species.

Family Condylocardiidae

Carditella calipsamma n.sp.

Fig. 4

Location I. 45-50 m Transgression complex.

Back-beach environment.

“a4,

Ma, PL

Ct,

Weg,

Ss ‘iS

Fig. 4. Carditella calipsamma n.sp. Holotype. a. Right valve in external view. b. Right valve in

internal view. c. Hinge of left valve.

210 ANNALS OF THE SOUTH AFRICAN MUSEUM

Description

Valves equilateral, prodissoconch eroded. 23-27 radial ribs with low

nodules. Latter oval near apex, becoming transversely rectangular nearer

margin. Grooves between ribs becoming wider and shallower towards margin.

Very fine concentric growth lines present, more apparent in grooves on younger

portion. 4—5 growth lines per nodule. Ventral inner margin of shell crenulate,

mantle line uninterrupted. Left valve with two cardinal teeth, anterior tooth a

low ridge, posterior tooth prominently triangular. Right valve with 2 cardinal

teeth, separated by a deep triangular pit. Low ridge present in front of anterior

tooth. Each valve with a single anterior and posterior lateral tooth.

Name descriptive of the local lithology.

Material

Holotype: S.A.M. Ki4g49 L. 11:-gmm A, 12:3mm

Paratype: S.A.M. K1450 9°2mm 9°2mm

Other specimens: 12°8mm 12°8mm

12°5 mm 12°7 mm

11-8 mm 12°2mm

Remarks

The sculpture of this species resembles that of Carditella similis Jaeckel &

Thiele, Carditopsis dartevelle: Nicklés and to some extent Cuna gambiensis Nickles.

C. gambiensis differs from the present species in that the grooves separating the

radial ribs are relatively narrower. The dorsal profile of C. gambiensis is more

acute (go° as compared with circa 110°). The west African species also appears

to lack lateral teeth (see Nicklés, 1955). The present species is more closely

related to Carditopsis dartevellec Nicklés, 1952, recorded from the Quaternary

deposits of Gabon, but this differs in possessing fewer rays (14-16) and in being

a much smaller shell (2-5 x 2:6 mm being the largest). The adductor muscle

scars of the present species are not as obviously reniform as those in C. dartevellet.

Carditella similis differs from the present species in possessing fewer ribs (20-21),

in the older nodules being more obviously rectangular and flattened and in

being a smaller shell (3-5 7:3 mm being the largest).

Family Mactridae

Standella (Eastonia) namaquensis n.sp.

Pl. 26 & Fig. 5

Location 1. 45-50 m Transgression complex.

Back beach environment.

Description

Ovate, inequilateral, equivalve, transversely elongate. Umbones small,

slightly incurved, located anterior to mid-line of shell. Anterior end rounded,

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST 2t

Fig. 5. Standella (Eastonia) namaquensis n.sp. Hinge details: a. Right valve. b. Left valve.

posterior end subtruncate. Dorsal outline, posterior to umbones, very slightly

concave. Valve interior moderately concave; posteriorly the concavity

diminishes, the posterior termini becoming flattened, with a narrow gape. A

weak dorsal ridge runs from the umbo to the posterior end of the shell, fairly

close to the postero-dorsal angle. Lunule poorly defined. Outline non-crenulate.

Sculpture of concentric growth lines, raised and sub-lamelliform on the

ventral periphery. Radial sculpture of numerous, very fine, plicate ribs, irre-

gularly spaced, slightly sinuous, often showing random bifurcation and fusion,

frequently discontinuous across growth ridges. The radial and concentric

sculpture together produce a very fine, highly irregular, reticulate pattern.

Posteriorly, the surface becomes slightly rugose, the growth lines correspondingly

more sinuous.

Hinge plate robust, carrying a chondrophore projecting backwards but

torted towards the median area of the ventral margin. Distal outline of chon-

drophore smoothly rounded. Hinge of right valve with 2 elongate, posterior

lateral teeth; the larger ventral tooth erect, the dorsal tooth suberect. 3

anterior dental elements present; a thin cardinal lamella, running the length of

the anterior margin of the chondrophore, leaning slightly towards the mid-line

to cover a small segment of that hinge component, a narrow lateral aligned

close to the dorsal margin of the hinge plate and a large, erect, apparently

bilobed, cardio-lateral tooth. This last element appears to be compounded of a

PI Ps ANNALS OF THE SOUTH AFRICAN MUSEUM

small, dorsally situated, trigonal cardinal fused to a much larger, longer, sub-

trigonal, lateral tooth. The suture marking this fusion is clearly evident when

viewed from the anterior aspect. The compound cardio-lateral tooth is separated

from the cardinal lamella by a narrow, triangular pit.

In the holotype, the cardinal lamella has broken away but the scar is

apparent. The lamella is preserved on a second specimen (paratype S.A.M.

K1425).

Hinge of left valve with a single, elongate, posterior lateral ridge extending

to the hinge plate margin. In front of the chondrophore, 2 anterior cardinal

teeth fused to form a A-shaped compound tooth which extends approximately

half way across the hinge plate. In the holotype, the A-shaped tooth is suberect,

the crest deflected some 20~30° from the vertical; in smaller specimens the

degree of deflection is somewhat greater. The two branches of the A-tooth

diverge at about 30°, the posterior branch overhanging the chondrophore. A

narrow, linear depression separates this compound tooth from a single, large,

erect lateral.

Pallial sinus deep, sub-parallel; terminus well rounded, extending to the

level of the extreme posterior edge of the chondrophore.

Examination of other specimens from the type locality indicates that the

outline and decoration show some variation. In the smaller specimens, the

valves are less elongate and slightly donaciform. In some specimens the

sculpture is apparently muted with concentric growth lines dominant. Examina-

tion of such forms indicates that the fine, radial elements of the reticulate

sculpture have been preserved in depressions in the somewhat rugose surface

of the extreme posterior areas of the valves. It is supposed that the destruction

of these delicate plicate ribs is the result of abrasion encountered during trans-

port of the valves along the beach. It is notable, in this context, that discrete,

separated valves are more profoundly affected in this respect than the holotype,

recovered with both valves together.

Named from the general location — Namaqualand.

Material

Holotype: 5.A.M. Kige4 L. 66mm A. 43mm W-. 28)aam

(both valves)

Other specimens: 51 mm 34 mm

43 mm 31 mm

Paratype (hinge fragment): K1425

Remarks

As regards dentition, the present species more closely resemble Lamy’s

figure (1918: 383) of Eastonia rugosa (Helbling) than it does his figure (p. 382)

of Standella pellucida (Chemnitz). The two branches of the A-shaped compound

cardinal tooth in the left valve as noted above, diverge at an acute angle, the

posterior branch overhanging a segment of the chondrophore. Despite the

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST 213

degree of fusion, it is suggested that the compound anterior cardio-lateral tooth

in the right valve of the present species more nearly resembles the closely

apposed anterior cardinal and lateral teeth of EF. rugosa than it does the con-

tinuous, bilobed, anterior cardio-lateral element in S. pellucida.

For the above reasons the present species is assigned to the sub-genus

Eastonia.

S. (E.) namaquensis is unlike the only South African representative of the

sub-genus, viz. Standella (Eastonia) solanderi (Gray). (Barnard (1964) suggests

that S. nicobarica (Gmelin) falls outside the limits of the sub-genus.) In S. (£.)

solandert, which is a far smaller form bearing a strong radial ornament of

relatively regular, continuous ribs, the A-shaped tooth occupies the full width

of the hinge-plate; the divergence of the branches of this tooth being greater

than in the present species.

The similarity in hinge detail between S. (E£.) namaquensis and the west

African representative S. (£.) rugosa has been noted above. The two species

differ as regards decoration, the last named, according to Nickles’ figure

(1950: 210), possessing a strong radial ornament of regular, closely spaced

costae. In this figure, the terminus of the pallial sinus attains the level of the

anterior edge of the chondrophore.

In Lamy’s figure (p. 383) of S. (Z.) rugosa, the anterior lateral tooth in the

left valve appears oblique, not erect as in the present species.

S. (£.) rugosa is a Mediterranean species extending into the Atlantic ocean

along the Portuguese and west African coasts. It is known from the Pliocene

and Pleistocene of France.

Nicklés (1950: 210) suggests, by his terminology, that S. senegalensis

Philippi is not to be included in the sub-genus. The outline and decoration of

this shell is markedly different from the present species.

The Philippine forms S. (Merope) plicatilis (Deshayes) and S. (Merope)

capillacea (Deshayes), according to Deshayes (1854: 69), have an irregular,

reticulate sculpture of fine plicate ribs like that possessed by the present species.

In M. plicatilis however, the pallial sinus is trigonal, terminating at an obtuse

angle.

Family Donacidae

Donax haughtonzi n.sp.

Pls 27-28 & Figs 6-9

Location 2. 45-50 m Transgression complex.

| Fore-beach environment.

Description

Shell equivalve, inequilateral, greatest pre-umbonal length longer than

greatest post-umbonal length. Shell anteriorly smooth, posterior ridge not very

distinct, with fine concentric lines sometimes posteriorly present; rest of shell

214 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig. 6. Donax haughtoni n.sp. Hinge details: a. Right valve. b. Left valve.

° Donax(Iphigenia) rogersi Haughton

« Donax haughtoni n.sp.

WW UY HuU2| |EUOqUUN-}SOd }sa}eEDUD

40 60 80

Greatest pre-umbonal length mm.

Fig. 7

surface smooth. Inner margin smooth. Antero-ventral margin smoothly rounded,

postero-ventral margin with a slight corner in older valves. Hinge of left valve

with 2 cardinal teeth with triangular pit between. Anterior tooth slightly

broader than posterior. Hinge of right valve with a low anterior ridge and a

prominent, broad cardinal tooth. Latter apically bifid. Rectangular flattened

process, posterior to umbo in both valves, at least three times longer than wide.

Named after Dr. 8. H. Haughton.

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST 215

Fig. 8. Donax haughtoni n.sp. Diagrammatic cross-section of valve to

illustrate distance d—greatest width, hinge excluded.

Material

Holotype: S.A.M. Ki1430 L. 97-0mm A. 61:0mm

Paratype: S.A.M. K1431 70°5 mm 44°0 mm

(A complete range of valves from 20mm to 97mm in length, in the

South African Museum.) |

Remarks

This species most closely resembles Donax (Iphigenia) rogersi Haughton,

from the coastal deposits of Doornbaai. Haughton’s material in the South

African Museum, however, includes both D. (J.) rogerst and D. haughtont.

Haughton (1931: 36), states that the smaller shells obtained from The Point,

Van Rhynsdorp, have the umbo well in advance of the midline of the shell

but that the umbo’s position ‘accords well with that in equivalent growth

stages of the larger shells from Doornbaai and Alexander Bay, and these

specimens fall therefore within the limits of the species’. With more material

now available it would seem that this smaller and more elongate form men-

tioned by Haughton is the present species and not a young form of D. (Z.)

¢ Donax (Iphigenia) rogersi Haughton

¢ Donax haughtoni n.sp,

A-B Mean Width/Length

‘WA Bode BBuly BuIpnioxa-UpIM 4sSojeBUD

Greatest length mm,

Fig. 9. Comparison of greatest length/greatest width (hinge excluded) ratios in Donax (J.) rogersi

Haughton and D. haughtoni n.sp.

216 ANNALS OF THE SOUTH AFRICAN MUSEUM

rogerst. Figure 7 graphically illustrates the relationship of valve proportions in

the two species. It can be seen that in D. haughtoni the proportion between the

greatest pre-umbonal length and the greatest post-umbonal length is fairly

constant. Specimens of D. (J.) rogersi, although more variable, usually have a

more equilateral shell. There is a distinct separation of the two species in the

region of the graph where shells of similar size are recorded. The elongate

anterior region of D. haughioni is thus an almost constant feature of the species.

Other differences between the species include the strength of the posterior

ridge, hinge details, and relative thickness of the shell.

In D. (I.) rogerst the posterior ridge is far more pronounced than in D.

haughton. This results in a relatively wider shell. In an attempt to illustrate

this graphically, total shell length was plotted against the width of the shell.

This latter parameter, the distance ‘d’, was measured at right angles to a line

connecting the ventral and dorsal margins but excluding the immediate hinge

areas, as the teeth especially are subject to erosion. Figure 9g illustrates this

relationship. This character would seem to be rather variable, especially in

D. rogerst. Almost all the D. (I.) rogers: lie above a line representing the mean for

D. haughton, thus showing that D. rogers: usually has a relatively wider shell.

Although Haughton in his description of D. (J.) rogers: stated that the

central tooth of the right valve was not bifid, examination of less eroded material

shows that this tooth is indeed bifid, but not as markedly so as in D. haughtoni.

The rectangular process just posterior to the cardinal teeth in D. (Z.)

rogersi is about twice as long as wide. All the specimens of D. (J.) rogersi are

massive shells, much thicker than shells of comparable size of D. haughtont.

The stratigraphical distribution of the two forms is perhaps the most telling

point in favour of according a specific identity to Haughton’s ‘young growth

stages of D. (I.) rogerst’.

In fore-beach environments associated with the 45-50m strand line,

D. haughtont has been recovered in extraordinary abundance (see pl. 28).

Here, owing to this super-abundance and to the excellent preservation, a very

large number of individuals may be examined and a complete range of growth

stages recovered. From this wealth of material, not a single specimen of D. (J.)

rogersi was located. A large percentage of the individuals examined were

grouped in the largest size range, length 80 mm to 95 mm, which probably

represents the dimensions normally attained by mature adults of the species.

Growth stages from length 20 mm upwards to the adult length were examined

and found to exhibit no major variation in form, differing from one another in

size alone. If, as has been suggested previously, D. haughtoni is merely the juvenile

form of D. (I.) rogerst and that only mature or senescent forms of this latter

species are characterised by a thickening of the shell, an apparent shift of the

umbones to a median position and the development of a strong posterior ridge,

then a few at least of such ‘mature’ forms might reasonably be expected to

appear in the assemblage. None was apparent.

In the transgressive gravels and the coarse back-beach sediments of the

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST 217

45-50 m strand, representatives of the genus were not commonly found, those

located were usually fragmental. Of these, the few entire individuals and

occasional! shards collected could all be assigned to D. haughtont. Fragments or

entire valves of D. (/.) rogerst were not recorded.

At the localities investigated, the species D. (J.) rogerst is found associated

with the extremely coarse, granular to pebbly, back-beach deposits relating

to the 17-21 m sea stand. At this elevation numerically large concentrations of

individuals of this species are met with (pl. 28), all such individuals being

_ ascribed to the species D. (J.) rogersi; not a single individual of the proposed

' D. haughiont was noted. The degree of preservation of the shells is rather poor,

much of the material being chipped or badly broken and often of a fragile,

powdery or flaky consistency. The fractured nature of the shells, from coarse

breakage to fine comminution, may be attributed to the extremely rigorous

environment which existed during the time of entombment and is testified to

by the coarse grain size of the sediments. The powdery, fragile condition of the

shells points to subsequent leaching of carbonate by solutions percolating

through the very coarse, highly permeable, enveloping granular sediment.

It is suggested that this compounding of rigorous depositional environment

and post-depositional leaching is, in large measure, responsible for the virtual

absence of young growth stages of D. (J.) rogerst in the assemblages.

It seems apparent that there are stratigraphical as well as purely mor-

phological! differences between D. (J.) rogersi and the proposed D. haughtoni —the

former flourishing at the time of the 17-21 m marine transgression, the latter

being prevalent during an earlier encroachment by the sea on to the land.

It is possible that D. haughiont eventually gave rise to D. (/.) rogersi, the

change being conditioned by a profound shift in environment indicated by the

widely different character of the sediments in which the two species are found.

As suggested in the opening remarks, the nature of the 45-50 m trans-

gression sediments may indicate a slow accumulation of very fine grained

material producing an environment favourable to sand dwelling bivalves and

one allowing fairly thin-walled, elongate shells to flourish.

In the case of those sediments comprising the 17-21m beach, the

abundance of coarse, granular horizons up to 5m above the basal gravel

suggests that molluscs inhabiting the littoral during those times were obliged

to withstand a rigorous, abrasive environment, characterized by a rapid

accumulation of coarse sediment.

In this habitat, the thin-shelled form would be under stress and liable to

breakage. Accordingly, by a process of selection, communities containing

individuals tending towards a more robust and resistant form became dominant.

The shell of the average individual became thicker, with a powerful hinge

placed at the mid-dorsal point of the shell, producing a structure more suited

to withstand vertical stresses in a plane at right angles to the surface of the shell.

Similarly, stresses parallel with the short axis of the shell were countered by

the development of the strong posterior ridge.

218 ANNALS OF THE SOUTH AFRICAN MUSEUM

The placing of D. rogerst in the subgenus of Iphigenia must now be con-

sidered in the light of this new information. If Donax (Donax) haughtoni gave

rise to the later D. (J.) rogersi, it would seem unlikely that the same subgenus

(Iphigenia) would also occur in the Miocene deposits of Venezuela, where

D. (1.) olssont Hodson, 1931, was recorded. The subgenus Iphigenia is also

represented by the living D. (J.) rostrata Romer, D. (J.) laevigata (Gmelin) and

D. (I.) truncata (Monterosato) in west Africa.

In this particular case, it would seem that the division of the southern

African members of the genus Donax is an artificial one and that D. rogersi

probably does not fall into the subgenus Iphigenia.

Paes-da Franca (1960) records the discovery at Baia dos Tigres, Angola,

of a single valve, 77mm long, of Donax serra Chemnitz. The accompanying

plate portrays a sub-donaciform shell having a smooth, non-crenulate ventral

outline. The angle between the postero-dorsal and antero-dorsal margins is

rather high, certainly greater than that normally exhibited by D. serra. The

arena behind the dorsal ridge is fairly smooth, prominent crenulation of the

growth lines in this region, so typical of D. serra, being apparently absent. It is

suggested that Paes-da Franca has recovered a valve of D. haughton, probably

washed out of a local Pleistocene horizon and not a specimen of D. serra.

Family Tellinidae

Gastrana rostrata n.sp.

Pl. 29 & Fig. 10

Location 1. 45-50 m Transgression complex.

Fore-beach environment.

Description

Shell moderately large, thin-walled, elongate, inequilateral, equivalve.

Anteriorly rounded, posteriorly strongly rostrate. Median areas of valves

vaulted, posteriorly the degree of concavity reduced, the rostrate regions

becoming flattened along a plane separating the valves, forming a small gape.

Dorsal outline behind umbones concave; ventral outline smooth.

Umbones small, not inflated, slightly forward of the median dorsal area;

external surface of shell hardly visible in interior view. A weak dorsal ridge

running from the umbonal region to the posterior ventral angle.

Lunules moderately long, narrow, and well-defined; ligament mainly

external.

Sculpture of concentric growth lines which become raised to sub-lamelli-

form towards the periphery, also microscopic radial striae, entirely subordinate

to the concentric sculptural elements.

Hinge areas extremely small in relation to the size of the shell.

Left valve with 2 cardinal teeth, very unequal in size. Posterior tooth

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST 219

small, erect, lamelliform, situated very close to the anterior end of the ligament

and obliquely set. Anterior cardinal tooth centrally situated on hinge plate,

relatively large, wedge shaped, squarely set pointing just anterior to the median

portion of the ventral outline. A deep, narrow groove running along the upper

surface, giving a profoundly bifid appearance to the tooth. 2 raised processes,

separated by this groove, situated on the extreme distal portion of the tooth.

=H é

= -

Ss z

at geet ee ie”

= ap id . 2

a, al Sk Ee ke faa)

=e “oe Stel a hese >

2. : . oe

> 2 . - =

Fe ee ° e - .? tte

pom eh ee me aoe > 8 ” = S ~

Se . s

: rs

= : :

== “eg 5 :

. -

G ‘.

iz ©

—

Fig. 10. Gastrana rostrata n.sp. Hinge details: a. Right valve. b. Left valve.

Terminal area of this large cardinal not projecting beyond the hinge plate

and obliquely truncated, sloping posteriorly. Cardinal teeth separated by a

deep triangular pit.

Right valve with 2 moderately large, divergent cardinal teeth; sub-equal,

the anterior member somewhat broader; both erect, narrowly wedge-shaped

and patently bifid, each with a longitudinal groove on the upper surface. A

broad, deep triangular pit, accommodating the large cardinal tooth of the

left valve, separating the two teeth.

Pallial sinus deep, passing beyond the level of the hinge; sides not parallel,

the dorsal margin sloping steeply downwards; terminus poorly rounded.

Ventral margin running close to, and parallel with, pallial line.

Material

Holotype: S.A.M. K1426 L. 67-0mm A. 39°5mm

Paratype: S.A.M. K1427 64°5 mm 43°5 mm

220 ANNALS OF THE SOUTH AFRICAN MUSEUM

Paratype: )S:A:M.>, 351445 (| 1 /62-0mmma VAS ber oO mann

Paratype: S.A.M. K1446 =66-0mm 43°0 mm

Remarks

The presence of two cardinal teeth in each valve, the absence of laterals,

the gross inequality of the cardinal teeth in the left valve, the bifid nature of

the teeth and the form of the ornament and the pallial sinus suggest that this

new species should be assigned to the genus Gastrana Schumacher.

The present species shows affinities with G. abildgaardiana (Spengler) and

with G. matadoa Gmelin, all three forms having a similar pallial sinus form and

a basically similar dental design. 7

The former species is an inhabitant of the southern and eastern coasts of

South Africa and is recorded from the local Pleistocene deposits; Sedgefield

(Martin); Knysna (Schwarz) and Saldanha Bay (Haughton). The latter lives

along the west African coast as far south as the Congo.

G. matadoa and the present species are dissimilar in regard to shape and

ornament, the former being generally oval in outline and decorated with more

massive concentric ridges, becoming very strongly developed on the peripheral

parts of the shell. In the left valve of the west African species, the central

tooth is relatively larger, projects slightly beyond the hinge-plate margin and

carries a broad shallow trench rather than a deep, narrow groove on the upper

surface.

The sculpture of G. abildgaardiana is very similar to that of the present

species. The two forms show definite differences in shape (G. abildgaardiana

being ovato-cuneiform) and in dentition. In Spengler’s species the triangular

central cardinal tooth in the left valve is a relatively massive structure, very

broad and occupying a larger area of the hinge region. In this large tooth,

which projects well beyond the hinge margin, the bifid condition is only

poorly developed.

In the right valve the subequal cardinals show only poor bifidity. In both

G. matadoa and G. abildgaardiana the umbones are moderately inflated.

The difference in shape and size of the dental elements suggests that the

present form is specifically different from both G. abildgaardiana and G. matadoa

and not simply an aberrant rostrate form of one or the other. In this context

it is interesting to note that the present species is to be found in abundance in

certain localities at the horizon noted above. From these localities, G. abildgaar-

diana has not been recovered; a single, small, left valve of G. matadoa has,

however, been found amongst the hundreds of individuals of G. rostrata. ‘This

valve bears the ovato-cuneiform outline and the strong, concentric sculpture

ridges of the contemporary G. matadoa as figured by Nicklés (1950: 220, fig. 426).

A broad, shallow groove runs along the upper surface of the central cardinal

tooth which projects beyond the hinge-plate margin.

Deeply abraded, exceptionally thick-walled and heavy, ovate specimens

of G. abildgaardiana are found in the very coarse sediments of the 17-21 m beach.

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST aoe

Family Aloididae

Aloidis palaegialus n.sp.

Pl. 25 & Fig. 11

Location 3. 45-50 m Transgression complex.

Early transgression beach.

Description

Inequilateral, umbo in front of middle. Inequivalve, right valve larger

than left, strongly vaulted. Both valves posteriorly with blunt carina stretching

from umbo to margin, posteriorly slightly beaked. Antero-ventral margin

smoothly rounded. Ventral edge of right valve very thin, overlapping left

valve, mainly in posterior region. Sculpture of not very regular, fine, concentric

ridges, becoming stronger ventrally. Umbonal region almost smooth. Internally,

Fig. 11. Aloides palaegialus n.sp. Hinge details: a. Right valve. b. Left valve.

222 ANNALS OF THE SOUTH AFRICAN MUSEUM

overlapping portion of right valve clearly demarcated from rest of valve.

Anterior and posterior muscle scars present, pallial lime uninterrupted, no

pallial sinus; posteriorly, pallial line turns upwards at right angles, to meet

the posterior adductor muscle scar. Hinge of right valve with prominent

triangular tooth, a low ridge running anteriorly. Just posterior to the strong

cardinal tooth, a deep ligamental pit, set well under the umbo. Hinge of left

valve with an anterior triangular pit (to accommodate the cardinal tooth of

the right valve) just posterior to which, a large, spatulate, triangular ligament

support.

Material

Holotype: S.A.M. Ki451 L. 15:5mm A. 9:8mm

Paratype: S.A.M. K1452 16-2 mm I0‘Omm

Remarks

This species is very similar in shape to Aloides striatissima Lamy, known

living from Angola to Senegal and as a fossil from the Pleistocene of Gabon

(Nickles, 1952). The concentric sculpture appears identical in the two species.

In most of the available descriptions of A. striatissima mention is made of very

fine radiating lines, particularly on the anterior portion of the valves. The

present species bears no sign of radiating sculpture. No figures or description

of the hinge of Lamy’s species is at present available.

ACKNOWLEDGEMENTS

The authors wish to express their appreciation of the assistance rendered

by the Management of De Beers Consolidated Mines Ltd., Namaqualand

Venture. They are indebted to Dr. G. Bonaccorsi, who kindly donated the

holotypes of Standella (FE) namaquensis and Tritonalia bonaccorsu and to Mr. S.

Kannemeyer (South African Museum) from whose photographs plates 18-27

and 29 are derived.

The Trustees of the South African Museum thank the South African

Council for Scientific and Industrial Research for a grant in aid of publication.

SUMMARY

During the course of a geological survey of the Tertiary to Recent sediments

of the Namaqualand coast, a suite of fossil molluscs was recovered from

Pleistocene horizons at three locations along this coastal strip. The paper

describes some twenty species and subspecies from this collection, seventeen of

which are new, the remaining three being new records for the west Coast.

The generic affinities of the new species and the present geographic ranges

of the previously recorded forms, here described, suggest that, in Lower

Pleistocene times at least, the waters bounding this coast were considerably

warmer than those presently washing this shore.

PLEISTOCENE MOLLUSCS FROM THE NAMAQUALAND COAST 223

REFERENCES

BARNARD, K. H. 1959. Contributions to the knowledge of the South African marine Mollusca.

Part II. Gastropoda: Prosobranchiata: Rhachiglossa. Ann. S. Afr. Mus. 45: 1-237.

BARNARD, K. H. 1961. Revised list of South African late Tertiary and Pleistocene marine

Mollusca. Trans. R. Soc. S. Afr. 36: 179-196.

BARNARD, K. H. 1963. Contributions to the knowledge of South African marine Mollusca.

Part IV. Gastropoda: Prosobranchiata: Rhipidoglossa, Docoglossa. Tectibranchiata.

Polyplacophora. Solenogastres. Scaphopoda. Ann. S. Afr. Mus. 47: 201-360.

BARNARD, K. H. 1964. Contributions to the knowledge of the South African marine Mollusca.

Part V. Lamellibranchiata. Ann. S. Afr. Mus. 47: 361-593.

BIBERSON, P. 1963. Palaeoclimatic variations of the Atlantic Pleistocene. Publs Anthrop. Viking

Fund 36: 417-447.

Bosuorr, P. H. 1965. Pelecypoda of Inhaca Island, Mocambique. Mems Inst. Invest. cient. Mogamb.

(A) 7: 65-206.

Butzer, K. W. 1966. Environment and archaeology. Chicago: Aldine.

Cox, L. R. 1930. An abnormal Navicula from South Africa. Proc. malac. Soc. Lond. 29: 154-155.

Cox, L. R. 1939. Depédsitos tercidrios e post-pliocenos do distrito de Inhambane. Estudo

paleontolégico . . . de moluscos. Moluscos miocénicos, pliocénicos e post-pliocénicos de

Mocambique— Miocene, Pliocene and post-Pliocene Mollusca from Mozambique. Bolm

Servs Ind. Minas Geol., Lourengo Marg. 3: 21-58 (Portuguese), 65-103 (English).

DesuayEs, G. P. 1854. Descriptions of new species of shells from the collection of Hugh Cuming,

Esq. Proc. zool. Soc. Lond. 1854: 62-72.

Hatt, C. A. 1959. The gastropod genus Ceratostoma. F. Paleont. 33: 428-434.

Haucuton, 8. H. 1926. On some new Mollusca from Tertiary beds in the west of the Cape

Province. Trans. R. Soc. S. Afr. 132 159-162.

Haucurton, 8. H. 1931. The late Tertiary and Recent deposits of the west coast of South Africa.

Trans. geol. Soc. S. Afr. 34: 19-57.

Hopson, F. 1931. Some Venezuelan mollusks. Part I. Bull. Am. Paleont. 16: 1-94.

Krauss, F. 1848. Die stidafrikanischen Mollusken. Stuttgart: Ebner & Scubert.

Lamarck, J. B. P. A. DE M. be. 1819. Histoire naturelle des animaux sans vertébres . . . l’ exposition

des principes fondamentaux de la zoologie: Paris.

Lamy, E. 1918. Révision des Mactridae vivants du Muséum d’Histoire Naturelle de Paris.

J. Conch., Paris 63: 291-411.

NArpInI, S. 1937. Molluschi delle spiagge emerse del Mar Rosso e dell’Oceano Indiano. Parte II

(Lamellibranchi). Palaeontogr. ital. 37: 225-278.

NickEs, M. 1950. Mollusques testacés marins de la cote occidentale d’ Afrique. Paris: Lechevalier.

Nick.Es, M. 1952. Mollusques du Quaternaire marin de Port-Gentil (Gabon). Bull. Dir. Mines

Géol. Afr. equat. fr. 5: 74-101.

Nickies, M. 1955. Scaphodes et lamellibranches récoltés dans l’Ouest africain. Atlantide Rep.

32 93-238.

Paks DA FrAnNcA, M. DE L. 1960. Contribuicgéo para o conhecimento da fauna malacoldgica de

Angola. (Terceira nota.) Mems Fta Invest. Ultramar. 15: 9-40.

SCHILDER, F. A. 1933. Monograph of the subfamily Eratoinae. Proc. malac. Soc. Lond. 20: 244-283.

SmiTH, E. A. 1876. A list of marine shells, chiefly from the Solomon Islands, with descriptions

of several new species. 7. Linn. Soc. Zool. 12: 535-562.

THIELE, J. 1929. Handbuch der systematischen Weichtierkunde. 1. Jena: Fischer.

Voxes, E. M. 1964. Supraspecific groups in the subfamilies Muricinae and Tritonaliinae.

(Gastropoda: Muricidae.) Malacologia 2: 1-41.

Pienct

8 a ae

nH ham

Ree Mr ;

Ann. S. Afr. Mus., Vol. 52

(Note: on all plates, one division of the scale = 1 cm.)

a. ‘Turris’ nigrovitta n.sp. Holotype

b. Nassa litorafontis n.sp. Holotype

c. Fasciolaria lugubris Reeve

d. Fusus fauret Barnard

Plate 18

Ann. S. Afr. Mus., Vol. 52 Plate 19

a. Tritonalia bonaccorsii n.sp. Holotype

b. Latiaxis sp.

c. Hespererato oppenheimeri n.sp. Holotype

Ann. S. Afr. Mus., Vol. 52 Plate 20

Namamurex odontostoma n.gen. et sp. Holotype

Ann. S. Afr. Mus., Vol. 52

Plate 21

Namamurex odontostoma n.gen. et sp. Size range

Top left — Haughton’s specimen

Bottom left— Holotype

Ann. S. Afr. Mus., Vol. 52 _ Plate 22

a. Calyptraea aurita striata n. subsp. Dorsal view, Holotype

b. Calyptraea viridarena n.sp. Dorsal view, Holotype

Ann. S. Afr. Mus., Vol. 52 Plate 23

a. Clanculus murrayi n.sp. Holotype, upper right

Paratype, upper left

b. Calliostoma depressa n.sp. Holotype

Ann. S. Afr. Mus., Vol. 52

a. Fissurella glarea n.sp. Size range, Holotype bottom left

b. Arca (Acar) halmyrus n.sp. Holotype

Plate 24

Ann. S. Afr. Mus., Vol. 52 Plate 25

a. Arca avellana Lamarck

b. Aloidis palaegialus n.sp. Holotype, top

Ann. S. Afr. Mus., Vol. 52

Standella (Eastonia) namaquensis n.sp. Holotype

Plate 26

Ann. S. Afr. Mus., Vol. 52 Plate 27

a. Donax haughtoni n.sp. Holotype, external view of right valve

b. Donax rogersi Haughton. External view, right valve

c. Donax haughtoni n.sp. Holotype, internal view of left valve

d. Donax rogersi Haughton. External view of left valve

Plate 28

Ann. S. Afr. Mus., Vol. 52

10n 2

sp. Locat

1 Haughton. Location 1

Donax haughtoni n

Bottom. Donax rogers

Top

Ann. S. Afr. Mus., Vol. 52 Plate 29

Gastrana rostrata n.sp.

a. Dorsal view, Paratype

b. External view, Holotype

c. Internal view, Paratype

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. (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)

BuLtoucH, 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., Duvau, 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. Polyphacophora, 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.

| i

| A

} )

;

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 52° #2Band

November 1969 November

Part 10 Deel

AQIHSONI gy

at]

jan 26 19/0

LisRARies

THE BEHAVIOUR OF

CAPTIVE CAPE ROCK LOBSTERS,

JASUS LALANDII (H. MILNE EDWARDS)

NELLIE F. PATERSON

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, tpl), 5(2, 5, 7-9);

6(1, tpt) 71; 3); 8, g(1-2), 10(1-3),

2) 7s tpt) 2I, 24(2), 27) Z1(E—3), 38,

44(4).

Price of this part/Prys van hierdie deel

R2.50

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 BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS,

JASUS LALANDIT (H. MILNE EDWARDS)

NELLIE F. PATERSON

South African Museum, Cape Town

(With 4 figures and 15 tables)

[MS. received 12 September 1968]

CONTENTS

PAGE

Introduction ; : , 4 : 225

Material and methods . : ; 2 a 226

General observations . : : : 220

Locomotor and feeding activities . : Se V DBTS

Bedysis =! . \ 7. F : d : 232

Regeneration of appendages . : : E226

Cannibalism ; : ; ; 2 oa

Average rate of growth : 3 : SF 2RG)

Intermoult periods. : : : e240

Moulting frequencies and annual rate of growth 243

Ovigerous setae . : : : : Al

Reproduction . ‘ ‘ ‘ : 2A,

Attempts at artificial insemination. : 5 EVI)

Histology of the reproductive organs. Je 2EO

Fertilization y : : é : a 252

Incubatory period : é : : a4:

Discussion . ; é A ; : mm ess

Summarys. 0. 4 ee : ; . 260

Acknowledgements : ; ; 2p 20k

References . : : : 2 : Sead

Appendix . : : f : : «| 7264

INTRODUCTION

Various problems concerning the habits and the biology of Fasus lalandii

have been investigated by Gilchrist (1913, 1918), Von Bonde & Marchand

| (1935), Von Bonde (1936), Matthews (1962), Heydorn (1965, 1966), and

| Lazarus (1967). However, the results of field and laboratory observations

| indicate that clarification of some aspects is still desirable.

225

Ann S. Afr. Mus. 52 (10), 1969: 225-264, 4 figs, 15 tables

226 ANNALS OF THE SOUTH AFRICAN MUSEUM

It has generally been accepted that the adults moult once a year, but

differences between the moulting frequencies of feral and captive animals are

suggested in several reports. In certain holding experiments, Gilchrist (1913)

showed that two ecdyses a year occurred in some mature females, and recent

observations (Paterson, 1968; 19692) have confirmed the prevalence of biannual

ecdyses in captive animals.

There is little information regarding the mating behaviour in 7. lalandii,

and the method of insemination has been a matter of speculation. Based on the

nature of the membranes of freshly-laid eggs and on the supposed absence of

external spermatophores, Von Bonde (1936) concluded that fertilization

probably takes place in the oviducts. This opinion has received some support

from Fielder (1964c) and Heydorn (1966), but it has been noted (Paterson,

1968) that the external genital apparatus of 7. Jalandii resembles that of some

Scyllaridae, in which family external spermatophores have been described by

Matthews (19542). It has also been implied (Paterson, 1968) that the mating

behaviour of 7. lalandi and the Natal rock lobster, Palinurus gilchristi Stebbing,

may prove to be similar and that in both species the eggs are probably fertilized

externally.

In pursuit of further information on the general behaviour of 7. lalandu,

various specimens were maintained at the Sea Point Aquarium from March

1966 to July 1968. During the course of the study, a soft, transparent, external

spermatophoric mass was detected behind the gonopores of one experimental

female. It is considered that this fact contributes towards an understanding of

the process of fertilization in the Cape rock lobster.

MATERIAL AND METHODS

As soon as possible after their capture, the animals were placed in glass-

fronted tanks filled with aerated sea water operating from a closed circuit.

The largest experimental tank measured 72 in. X 69 in. X 48 in. and had a

built-in rocky background. Frequent use was also made of ten smaller tanks,

each of which was approximately 42 in. X 20 in. X 18 in. in size and had a

fair depth of sand and loose stones on the bottom. In addition, a large exhibition

tank, measuring 102 in. X 48 in. X 72 in. and accommodating a variable

number of mature and immature rock lobsters, was also available for general

observations.

_ From time to time, adult males and females were generously supplied by

Dr. A. E. F. Heydorn, Mr. G. G. Newman and their colleagues, who collected

the animals by hand during routine diving operations off Robben Island in

Table Bay. Random specimens in each batch of animals were dissected to check

whether the gonads were mature. In most of the experiments the sexes were

paired and, as the females were generally smaller than the males, they appeared

to be suitable for mating. More information concerning the habits of captive

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS 227

animals was gained from 24 juvenile females, which were selected from a large

number of immature animals introduced into the Aquarium in July 1966.

To avoid overcrowding, only five pairs of mature animals were placed in

the largest experimental tank and usually two pairs of adults and no more than

a dozen juveniles were kept in the smaller tanks. The animals were handled

carefully and as seldom as possible, yet several of them lost some of their

appendages at times when there was no outside interference. Most of the

maimed animals were replaced by more perfect specimens, but it is probable

that, had they been retained, regeneration of the injured limbs would subse-

quently have taken place.

The animals were maintained successfully on a daily meal of pieces of

stockfish, Merluccius capensis Castlenau, supplemented by quantities of the

mussels, Aulacomya magellanica (Chemnitz), various species of Patella, and the

kelp, Ecklonia.

At 6 p.m. each day the electric lights in the Aquarium were regularly

switched off and all the tanks were in total darkness until 9 a.m. on the following

day. In an endeavour to make the conditions as normal as possible, the artificial

illumination of the rock lobster tanks was dispensed with and it was found that

sufficient subdued natural light reached the tanks from the outside windows of

the building.

After their introduction into the tanks, the animals were inspected at least

once or twice during the day, and visits of varying duration and frequency were

made at night when torchlight was used for the observations. During these

visits the habits of individual animals were studied.

Using steel calipers, the total cephalothoracic length (T.C.L.) was

measured in millimetres from the tip of the rostrum to the postero-dorsal edge

of the carapace. Heydorn (1965) has concluded that the males and females

may become sexually mature when the T.C.L. is 6-0 to 6:5 cm and 7-0 cm

or more respectively. In the present study few small males were handled and

most of the adult females were at least 70 mm in T.C.L.

Exuviation, which usually took place at night, was seen on a number of

occasions. Records of the ecdyses of individual animals were used to determine

the growth rate and the length of the intermoult periods.

As few of the captive females produced eggs, attempts were made to

inseminate several specimens by artificial means. In another inquiry into the

possible intromission of sperm, the histology of the reproductive organs was

examined. Sections were cut of part of the ovary and the entire oviduct of an

ovigerous female and of portions of the testis and the vas deferens of a mature

male. The sections were 10 uw thick, and alternate slides were stained with

Mallory’s triple stain or with Heidenhain’s iron haematoxylin and eosin.

Smears of seminal matrix, fixed either in osmic acid fumes or in Zenker’s fluid

and stained with Heidenhain’s haematoxylin, were used as an additional check

for possible traces of spermatophoric material in sections of the female

organs.

228 ANNALS OF THE SOUTH AFRICAN MUSEUM

GENERAL OBSERVATIONS

After becoming acclimatized to their surroundings, most of the rock

lobsters established territorial rights to certain rock niches or corners in the

tanks and adopted a belligerent attitude towards other animals attempting to

usurp their chosen retreats. The individuals varied in their habits from day to

day and, while some remained dormant for long periods, others in the same

tank were often fairly active. Some degree of activity was therefore frequently

observed during the day and at night in one or more of the tanks.

Most of the specimens devoted some considerable time to the diligent

cleaning of the appendages or various parts of the surface of the body. The

antennulary flagella seemed to receive most attention, being repeatedly drawn

through the apposed medial setose margins of the third maxillipeds. The latter

were then rubbed together and apparently cleaned after the antennulary

flagella had been released.

Occasionally some of the legs of one side were hooked over and drawn as

far as possible along the antennal flagellum of the same side, the surface of

which was presumably brushed by the spines and setae on the dactyls. The

antennal peduncles, the eyes, and parts of the cephalothorax were most

frequently cleaned by the dactyls of the second or the fourth pereiopods.

In both sexes the dactyls of the fifth legs were employed to scrape the

surfaces of the abdomen and the pleopods. As remarked by Gilchrist (1913),

Von Bonde & Marchand (1935), Von Bonde (1936), and Fielder (19645),

ovigerous females habitually used the chelae of the fifth legs to groom the -

incubation chamber. The loosened debris was probably cleared by the subse-

quent vigorous flapping of the pleopods, which at times was also instrumental

in releasing swarms of larvae.

When not engaged in grooming, many of the animals rested on the

bottom of the tanks, usually supported by the flexed tail-fan and the posterior

legs, while the other legs swayed leisurely to and fro. Such limb movements —

appeared to be rather aimless, but they were doubtless of consequence in

maintaining the stream of respiratory water entering the branchial chambers

at the bases of the legs.

Occasionally some individuals were observed drawing the anterior legs

through the sand, and several specimens seemed to be fastidious about their

resting-places. Some of the occupants of the smaller tanks dug pits in the sand

in which they later rested. Fairly heavy loose stones lying in front of the bottom

rock recesses in the large tank were often forcibly removed, probably to make

the retreats more accessible.

LOCOMOTOR AND FEEDING ACTIVITIES

From June to September 1966, a study was made of the locomotor and

feeding activities of ten pairs of adult rock lobsters which had been in captivity

for two months before the observations began. The T.C.L. of the animals

229

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS

‘Iaquiojdeg pue ounf ul ssouysep jo sinoy ose19Av DY} JUIseIda1 spueq yoxrIq dy], *}USsSoId sjeuMTUe JO JoquINU 9Y} JO seseyUsII0d

sv passaidxa oie speurtue 9AT}0¥ Jo sjyunoo AyInoy sy J, ‘wd F pue & usaMjoq Aep yors Ajre[nso1 porddns sem poog ‘speuirue jo sured say poure}

-u00 puv ‘ulgh x ‘ur69 x ‘uTad pornsvow LI yUR], ‘s[euttuY jo sired OM} poure}UOd yore puv “UIgI X “UIOS X “ULSP noge poinsvow GI puv ZI syuLy,

9961 szaquioajdag O& 03 sunf{ I ported 9Y} 4107 sxajsqo] yOor y[npe jo sdno1ss sa14} Jo AIIATINV IOJOUIODOT BY} UI suOT}eIIeA ApINOY IY], “1 “Oly

Ldas

Cir CC — FN

L SYNOH

fe)

Siu

Ges

Gere

Ov Oo

Gv z

OSG O

py)

SG >

Go x

@recvsver®@ LI ANVIL

e------e@ Sb YANVL : SL

eo» dl ANVIL

ANNALS OF THE SOUTH AFRICAN MUSEUM

230

(S82[ 4G} DU} OFUL

peonpo.nUr sem pooy UsYM UOOUIEIe 9Y} SUTINp JUSprAS seM ARIAT” Jo JOAI] YSty Y “Uosoid speurrue Jo JaquNu oY} Jo so8vyUsII0d sv possoidxo

ose are syunod ATMoy ey], “I ainsy ut se porsod sures oY) Joy pue speurTUY sUTES dy} Jo AWANDV SuIpsay 9y}) UT sUONEIIeA A]Inoy, oyT, * “Oy

G e

We ®

he °

We .

i :

1 . e

~ "i emit =

Ol ‘\ we bs <o t ; ° e

\ . e

’ U ' . s

. ' Wo. P

\ ‘ ios e

\ ? 1 e 2

\ é i] . e

oe » r) tt. e

LY ‘ 1 oe t

\ ‘ ieee '

\ A ' e A

\ 1 oY i

‘ ' i) es A

‘ \ ine ‘

\ U n i

Od : : '

\ ’ \

\ , ' YU

° J °

\ i ' '

\ e-~ ' '

\ r) i} 4

‘ ) ‘ 1

GZ _ Ul 1 !

\ ? ; u

A ’ ' 1

ry ’ t

ay ' '

Cee ; t

oe ° a

1 t

| ’

t i

Jom

i !

GE st

' {

’

@escovvec® Zl YNWL

Oe=ri-s--@ CLANVL

2 —_.* -Gl INVE

O¢e

ONIGSS3 30, SSoV LN sodac

O :

“)

PUTED,

itp)

THE .BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS 231

ranged from 78 to 122 mm in the males and from 83 to 100 mm in the females.

Exploratory movements away from an established niche were regarded as

locomotor activity, while actual seizure and consumption of food were recorded

as feeding activity. :

Intermittent observations of the tanks were made at least twice an hour

- during the day and at night, and the hourly counts of the animals active were

expressed as percentages of the total number of animals present. These percen-

tages were used to determine the mean hourly variations in the locomotor and

the feeding behaviour during the four months in question (figs 1, 2).

From the above figures a comparison may be made between the hourly

activities of 18 animals located in two of the smaller tanks (nos 12, 15) and the

larger experimental tank (no. 17). The results suggest that the four animals in

each of the two smaller tanks were generally more active than the ten animals

in the larger tank. Although the animals in the smaller tanks were by no means

cramped, they probably disturbed one another more frequently than did the

animals in the larger tank, in which some specimens remained quiescent in

their rock shelters for relatively long periods.

The graphs shown in figures 1 and 2 may also be used to illustrate the

hourly levels of activity over a period of four months. Three main peaks of

locomotor and feeding activity are indicated, one occurring before sunrise,

another between 3 and 4 p.m., and a third after sunset.

The normal rhythms of foraging and feeding were doubtless disturbed by

the daily introduction of fish into the tanks between 3 and 4 p.m. Consequently,

the marked afternoon increase in activity appears to be anomalous. If, however,

the responses to artificial feeding are excluded, a bimodal rhythm seems to be

suggested in both the locomotor and the feeding activity.

A somewhat similar bimodal cycle has been reported by Kubo & Masuda

(1964) in the feeding habits of captive specimens of Panulirus japonicus. It was

established that the twilight peaks roughly corresponded with those found in

fishing experiments, but the dawn peak of the latter was two hours later than

the corresponding peak in the captive animals.

On the other hand, in captive Fasus novaehollandiae (lalandit), Fielder (1965)

observed that the maxima for both the locomotor and the feeding activity

occurred at dusk. This may imply that the normal rhythm is unimodal in this

species.

Following procedures suggested by Sutcliffe’s (1956) experiments on

captive Panulirus argus, the hourly counts of activity in 18 to 20 animals formed

the basis of a comparison between the day and the night locomotor and feeding

activities (table 1). The observations covered the same period of four months

indicated in figures 1 and 2, but, in addition to the previously mentioned

18 animals, another pair was maintained in a separate tank during June and

July. The means are represented as percentages of the quotients of the total

number of animals present and the number of active animals observed in

hourly counts throughout the whole period.

232 ANNALS OF THE SOUTH AFRICAN MUSEUM

TABLE I

Summary of counts of the diurnal and the nocturnal locomotor and feeding activities of 18 to

20 adult rock lobsters for the period June to September 1966. The means are expressed as

percentages of the total number of animals present.

No. of No. of No. of Mean % of No. of Mean % of

Time hourly animals animals locomotor animals Seeding

counts present Soraging activity Seeding activity

Day 149 2850 414 I4°5 194. 6-8

Night 194. 3748 1093 29°2 423 [i-3

Applying Hoel’s (1962) formula for testing the difference between two

proportions to the data in table 1, an assessment was made of the respective

day and night proportions of activity relative to the number of animals present.

The values of z were 14:1 and 6-2 with regard to the locomotor and

feeding activities respectively. This means that in both foraging and feeding

the difference between the day and the night activity was significant at the

Tip leviel:

It has been concluded that, although the daily introduction of food during

the afternoon affected the diurnal counts, there was an increase in both types

of activity at night. |

EcpysIs

Most of the animals were soft to the touch about two weeks before ecdysis,

but in a few large males proecdysis apparently began nearly three weeks before

the animals moulted. During this time the animals tended to become gradually

more quiescent, but they had occasional periods of restiveness. Feeding was

usually suspended for approximately five to seven days before ecdysis and was

resumed in four to eight days after the event.

As has been indicated in Panulirus by Travis (1954) and other investigators,

the first obvious sign of proecdysis was the appearance of a resorptive line

running somewhat obliquely backward and downward along the branchio-

stegite to its posterior articulation with a condyle on the last thoracic epimeron.

This line is probably more noticeable in species of Panulirus than in fF. lalandt1,

in both sexes of which a longitudinal pale streak is commonly present on the

branchiostegite more or less in the position of the suture along which calcium

resorption takes place during proecdysis.

About four to seven days before ecdysis the resorptive line was most clearly

seen by torchlight at night. It resembled a shining, greasy-looking marking

along each branchiostegite with an additional, short, broad band extending

downward at the junction of the branchial and the prebranchial chambers.

Even though the resorptive line became progressively more marked, it was less

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS | 233

obvious in the adult males than in mature females, but was very distinct in

both sexes of juvenile animals.

As the time for exuviation approached, the animals usually became very

restless and the body had a somewhat deformed appearance owing to an

apparent loosening of the parts at the junction of the thorax and the abdomen.

_ This was often evident about 24 hours before ecdysis and was associated with

the gradual exposure of the first abdominal tergum and a widening of the gap

between the carapace and the flange on the first abdominal segment. The

abdomen was then frequently held in the fully extended position, but the

animals were still able to move about fairly rapidly, although rather clumsily.

About two to three hours before exuviation in the juveniles and 45 to

75 minutes in the adults, the animals seemed to be more perturbed and were

noticeably hostile towards other rock lobsters in the tank. As they roamed

about, the antennae were held aloft and the pleopods, which were often pendent

at this stage, were swung leisurely to and fro. During short periods of rest, the

antennulary flagella were drawn vigorously through the apposed margins of

the third maxillipeds. The pereiopods and even the eyes were moved restlessly

in various directions, and at intervals the first abdominal segment was thrust

forward below the posterior margin of the carapace. Travis (1954) has suggested

that similar activities in Panulirus argus probably help to loosen the old exo-

skeleton from the underlying new one.

Immediately before ecdysis began, the animal became more agitated and

attempted to brace itself on the bottom of the tank or over a loose stone,

occasionally exhibiting convulsive movements. If, as frequently occurred, it

happened to be disturbed by other individuals at this juncture, it was still able

to evade them by darting swiftly backward through the water.

As in other species, the movements in 7. lalandit were accompanied by the

gradual and extraordinary swelling of the thoraco-abdominal intertergal

membrane which, according to Drach (1939) and subsequent observers, is

effected by water absorption and increased pressure in the pericardial pouches.

The distension of the membrane seemed to be less rapid in the juveniles

than in the adults. In the former the membrane began to stretch at least

80 minutes before ecdysis, while in the adults it was usually first visible about

30 to 40 minutes before actual exuviation. When the dorsal membrane became

markedly stretched, the animal seemed to be less sensitive to light. It settled

down fully extended, supported on the tips of the pereiopods and the edge of

the tail-fan and often with the antennae directed forward.

The preliminary step of exuviation was the detachment of the lower

margin of the old branchiostegite from the new exoskeleton, as a result of which

the posterior border of the old carapace was slightly raised. Thereafter, the

dorsal intersegmental membrane became progressively more taut until even-

tually the posterior edge of the new carapace was clearly visible through the

membrane. At the same time, the bracing action of the limbs and the somewhat

telescopic movements of the abdominal segments probably created additional

234 ANNALS OF THE SOUTH AFRICAN MUSEUM

pressure required for the implementation of ecdysis.

Rupture of the fully stretched membrane was sudden, and the old carapace

began to rise quickly and almost imperceptibly until it was tilted forward at an

angle of about 90 degrees. After the eyes had been withdrawn, the animal

seemed to find purchase against the anterior and lower parts of the old exo-

skeleton while the antennae, the legs, and the successive abdominal segments

were freed from their previous casings. Almost simultaneously, the sternum and

the old endoskeleton were released and, with a sudden convulsive movement,

the animal extricated itself from its former skeleton.

Most of the young rock lobsters were inclined to dart out from the right

side of the exuviae, but in the adults the procedure seemed to be more normal

and they emerged dorsally through the wide gap between the raised carapace

and the first abdominal tergum of the old exoskeleton. Once the animal had

moulted, it appeared to be unusually excited and very sensitive to light, and its

rather flexible antennal flagella were swept through the water with considerable

force. |

Ecdysis usually took place at night, but was also recorded before noon in

two adult males and one female. The exact time of ecdysis was somewhat

variable, and in observations of 11 juveniles and seven adults (fig. 3) it ranged

from 22 minutes before sunset in the largest specimen to over six hours after

sunset in one immature female.

In both the juveniles and the adults exuviation was generally completed

in three to five minutes and there appeared to be no direct correlation between

the size of the animal and the duration of ecdysis. Nor did the time of the year

have any obvious effect on the period required for emergence. Eight of the

observations below were made in spring, six in summer, two in autumn, and

two in winter. It may also be remarked that one of the summer ecdyses was

that of a female (T.C.L. 77 mm) which had produced eggs in captivity 140

days before this particular ecdysis.

The mean duration of exuviation in the 18 animals was 4:4 + 0-95 minutes.

This average includes the exceptional time of seven minutes recorded in one

immature animal (T.C.L. 58 mm) which required short periods of rest between

frenzied efforts to free its legs. During the struggle, the old exoskeleton was

torn asunder, and in several unsuccessful ecdyses of adult animals the exuviae

were found to be similarly dismembered.

The main difference between ecdysis in 7. lalandii and the process described

in other Decapoda is in the time taken for the withdrawal of the animals from

their exuviae. Drach (1939) observed that the period required in Brachyura

was 10 to 30 minutes, while in Panulirus it seems to vary with the species.

Several accounts show that it is about 15 minutes in P. japonicus (Nakamura,

1940), three to five minutes in small specimens and five to ten minutes in

somewhat larger individuals of P. argus (Travis, 1954), 11 to 18 minutes in

P. interruptus (Lindberg, 1955), and eight to ten minutes in P. cygnus (Thomas,

1966). In figure 3 it is evident that the normal duration of ecdysis in the

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS 235

+43

HOURS RELATIVE TO SUNSET (0)

40 50 60 Tie) 80 90 100

TEGsiEs SNe MM

Fic. 3. Starting times of ecdysis in 18 animals, plotted to the nearest minute before

or after sunset (0). The duration of ecdysis is indicated by the symbols shown in

the key. The T.C.L. of the individuals ranged from 43 to 100 mm.

juveniles of 7. lalandii is comparable to that observed by Travis (1954) in young

forms of P. argus, but in the adults it is shorter than in other palinurids.

During 1967 some records were also kept of ecdysis in a number of males

of 7. tristant exhibited at the Sea Point Aquarium. The exuviation of three

specimens with cephalothoracic measurements of 83 mm, 105 mm, and

119 mm occurred about two hours after sunset and took four minutes, five

minutes, and seven minutes respectively. The process was similar to that of

J. lalandiu, except that before and during ecdysis the animals were more com-

posed than moulting specimens of the local species and withdrawal from the

exuviae was effected with a minimum of effort. Such deliberate action was,

however, consistent with the habits of these fairly large rock lobsters which were

generally more placid than the captive juveniles and adults of 7. lalandit.

236 ANNALS OF THE SOUTH AFRICAN MUSEUM

Several investigators have stated that during ecdysis in palinurids the old

exoskeleton fractures along the resorptive lines but, as in Panulirus argus (Travis,

1954), this does not eventuate in 7. lalandii or in 7. tristan. The sutures along

the branchiostegites are softened by calcium resorption and, although they are

easily broken while handling the exuviae, they normally remain intact and

simply serve as pliable seams along which the side flaps of the otherwise rigid

exoskeleton are able to bend slightly when the carapace is raised during

ecdysis.

From an examination of exuviae, it is further evident that the carapace

becomes detached from the thoracic epimeral plates and that it is hinged at

the base of the epistoma. As Drach (1939, 1950) has shown in his comprehensive

studies of ecdysis, withdrawal of the animal is also facilitated by extensive

resorption of the medial elements of the endoskeleton. The resultant gaps in

the old endoskeleton are not only essential for the speedy release of the sternum

and the thoracic appendages, but their presence also accounts for the obvious

weakness of animals in the so-called ‘soft old shell’ stage.

All newly shed exuviae are provided with a thin slimy lining which,

according to Drach (1939) and Passano (1960), is formed during proecdysis

from a membranous layer adjacent to the epidermis. As these authors have

suggested that this lining may serve as a lubricant for the easy withdrawal of

the animal, it seems likely that a deficiency of mucilage may be one of the

causes of death during ecdysis. Failure to extricate the appendages and the

exposure of some of the gills and epipodites appeared to be common reasons for

fatalities during or shortly after ecdysis in some of the present animals.

In most of the captive animals the new integument seemed to be fairly

hard about three weeks after ecdysis, but the time required under natural con-

ditions is not known. Drach (1939) has shown that some of the calcium necessary

for the reconstruction of the new exoskeleton is derived from food intake and

from reserves in the digestive gland (hepatopancreas). He has also concluded

that calcium absorption from the sea water, particularly during the first stages

of the moulting cycle, is of great importance in the consolidation of the new

exoskeleton. A calcium deficiency in the external medium would therefore be

detrimental to postecdysial animals in that hardening of the exoskeleton would

be retarded. Chemical tests of the water in the rock lobster tanks indicated,

however, that the calcium level was satisfactory.

REGENERATION OF THE APPENDAGES

Renewal of appendages, which had been lost either before or shortly after

the animals were captured, also accompanied ecdysis in some adult and

juvenile specimens. The amputation of an injured pereiopod invariably took

place at the fracture plane between the basipodite and the ischiopodite, and the

extremity of the stump was then sealed by an obvious black scab.

At the next ecdysis, all the segments of the limb had been reformed, but

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS 237

they were always shorter and thinner than normal. Dissections of several

animals which died during exuviation showed that the soft regenerating limbs

were neatly folded within the stumps of the old limbs.

Occasionally an antennal flagellum, which had fractured close to the

peduncle, was partly regenerated during one intermoult period, but usually it

was no more than half its normal length. In one specimen, in-which the antenna

had apparently been severely injured, a short bifurcated antennal flagellum

was produced when the animal moulted. In another specimen, which had lost

the right middle pereiopod and the left antennal flagellum, two ecdyses were

required before the limb had regained its natural proportions and strength. The

left antennal flagellum, however, only became subequal to the one on the

other side after three ecdyses.

This fairly rapid replacement of injured appendages in 7. lalandi is

compatible with Bradstock’s (1950) observations of autotomy and autospasy

of the pereiopods and autospasy of the antennae of 7. edwardsi (lalandit), in

which an antennal flagellum became fully regenerated within a year.

CANNIBALISM

From time to time cannibalism of recently-moulted juveniles occurred in

the large exhibition tank. It was common when the animals were somewhat

crowded and a greater competition for food was likely.

In one instance, a young, newly-moulted female (T.C.L. 52 mm) was

relentlessly attacked by two other rock lobsters of a similar size and died within

half an hour. One assailant immediately tore out the eyes and at the same time

presumably destroyed the brain, while the second cannibal confined its atten-

tion to the posterior end of the body. This observation is of interest inasmuch as

Carlisle (1953) has reported that prawns kept under crowded conditions devour

the eyestalks of dead specimens before eating other parts of the body.

Cannibalism was not observed among immature and adult experimental

animals which were less confined and probably better nourished than the

other captive rock lobsters. In some of the tanks, however, the exuviae were

occasionally partly eaten before they could be removed.

From all accounts, the consumption of exuviae is usual in Crustacea and

is not confined to captive animals. Its occurrence in the latter may indicate an

inadequate supply of their natural food which could lead to cannibalism in

hungry animals and result in a reduction of the stock. Drach (1939) has

indicated that certain species eat exuviae as a source of calcium for accelerating

the hardening of the exoskeleton after ecdysis. He has, however, also remarked

that some animals will consume exuviae, not because of a need for calcium,

‘but because no other food is available.

AVERAGE RATE OF GROWTH

The exoskeleton was allowed to harden for about three weeks before the

average growth rate was determined by comparing the premoult and the

238 ANNALS OF THE SOUTH AFRICAN MUSEUM

postmoult cephalothoracic measurements of individual specimens. The moult

increments in T.C.L. for the juvenile females, the adult males, and the adult

females are considered separately in tables 2 to 4.

TABLE 2

Average growth rate per moult in juvenile females

Mean

TCL. No. of Increase increase Mean %

im mm specimens im mm in mm increase

TABLE 3

Average growth rate per moult in adult males. No specimens with a

T.C.L. of 80 to 89 mm were available

Mean

Increase increase Mean %

in mm im mm increase

TCD, No. of

in mm specimens

Jo=19 7 1°53 °° 2°4 oie

90-99 9 2°0-3°0 Br, 2°6

100-109 8 I*0-3°5

110-119 3 2°0-3°0

TABLE 4

Average growth rate per moult in adult females

Mean

No. of Increase increase Mean %

specimens increase

100-109

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS 239

An examination of the above percentage increases in length in the various

size Classes shows that the growth rate decreases as the animals become larger.

A somewhat similar gradation in the growth rate of a greater number of

specimens of 7. novaehollandiae has been reported by Fielder (19642).

The increase in the T.C.L. of the adults was usually 2-0 or 3-0 mm at each

ecdysis, but in some individuals the increments fluctuated at successive ecdyses

and. occasionally little or no difference in size was noted. In a few of the adults

the growth of the body was probably moderated by the necessary regeneration

of one or more injured appendages, but relatively low values were also common

in a number of perfect specimens.

Although there is no relationship between the T.C.L. and the growth

increments in the present small samples, an attempt was made to examine the

difference between the mean increments of the adult males and females on the

one hand and the adult females and the juvenile females on the other hand. The

relevant statistics are summarized in table 5.

TABLE 5

Estimated mean increases in T.C.L. per moult in captive animals, together with

the standard deviations and the ¢ values

No. of | Mean increase Standard

Specimens in mm deviations t values

From the statistical information in table 5, it has been calculated that:

1. In the ratio of the adult males to the adult females, with 65 degrees of

freedom, a ¢ value of 1-83 is significant at the 10% level only. Thus, no appre-

ciable difference in the growth increments of the captive adult males and

females was apparent.

2. In the comparison between the adult females and the juvenile females,

with 69 degrees of freedom, a ¢ value of 2°5 is significant at the 5% level. It is

therefore probable that the T.C.L. moult increments of the juvenile females

slightly exceeded those of the adult females.

It may be remarked that the increments shown in tables 3 and 4 are lower

than those in Heydorn’s (1966) field observations, in which the natural growth

rate of 19 tagged adults of 7. lalandi ranged from 2:0 to 8-5 mm per moult.

None of the present experimental adults increased by more than 3-5 mm in

T.C.L. after any one ecdysis, and a diminution of the growth rate was evident

at the consecutive ecdyses of several animals. Some confirmation of Lindberg’s

(1955) suggestion that growth is probably retarded in captive animals may

240 ANNALS OF THE SOUTH AFRICAN MUSEUM

therefore be indicated.

At the same time, some of the length increments recorded by Heydorn

(1966) appear to be greater than would be expected at a single ecdysis. His

results are obviously based on the conclusion that there is an annual ecdysis

in the adults of 7. lalandii, but it was found that biannual ecdyses were com-

mon in the present captive animals. If it should later be established that there

is more than one ecdysis a year in the field, it is probable that the average

growth rates of captive and free-living rock lobsters may prove to be more

compatible than is at present apparent.

INTERMOULT PERIODS

From May 1966 to July 1968, 109 ecdyses (table 6) were recorded at

different times of the year in 75 adults, comprising 28 males (T.C.L. 71 to

122 mm) and 47 females (T.C.L. 66 to 102 mm). During much the same

period, 38 ecdyses (table 6) occurred in 18 immature females (T.C.L. 26 to

69 mm).

TABLE 6

Summary of the number of seasonal ecdyses in captive rock lobsters

from May 1966 to July 1968

Specimens Autumn Winter Spring Summer

Adult males

Adult females

Immature females

Of the 75 adults, 18 males and 29 females either died or were abandoned

after their first ecdysis. ‘Thirteen of the 29 females, including one with a T.C.L.

of 66 mm, had been collected in berry during the winter breeding season, and

all of them moulted the following spring or summer after the eggs had hatched.

As indicated in a brief account of the ecdyses of captive rock lobsters

(Paterson, 1969a), the remaining 28 adults were maintained for a year or

longer (table 7) and usually moulted in autumn or winter and again in spring

or summer.

TABLE 7

Summary of the plural ecdyses in individual adult

rock lobsters from May 1966 to July 1968

Four ecdyses | Three ecdyses | Two ecdyses

in 24. months | in 18 months | in 12 months

Males 0) I 9

Females I 3 14

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS 241

Most of the females which survived two or more ecdyses failed to spawn,

but three of them moulted in captivity before producing eggs and each had a

second ecdysis five or six months later.

In 26 of the 28 animals enumerated in table 7 definite spring/summer to

autumn/winter and autumn/winter to spring/summer intermoult periods

occurred (tables 8, 9). The two exceptional animals were females, in one of

which (T.C.L. 72 mm) there was a short spring to summer intermoult period

of 119 days and in the other (T.C.L. 85 mm) a winter to early spring inter-

moult period of 105 days. These two intermoult periods have been omitted in

tables 8 and 9.

TABLE 8

Duration of the spring/summer to autumn/winter intermoult period in adult males and females

No. of intermoult periods Intermoult periods in days Mean duration

TG.L. of intermoult

im mm Males Females Males Females periods in days

12-79 I I 154 201 E/7 5

80-89 O 4 e) 176-220 IQI°2

90-99 I I 147 58 180-0

100-109 I fo) 207 fo)

TABLE 9

Duration of the autumn/winter to spring/summer intermoult period in adult males and females

No. of intermoult periods Intermoult periods in days Mean duration

meee. 1] of intermoult

im mm Males Females Males Females periods in days

70-79 O 3 O 120-185 148-3

80-89 ) 6 O 143-188 158°3

go—99 3 4 166-198 146-166 164°1

100-109 4. 2 161-186 150, 188 171°6

IIO-119 I O 173 )

Although the number of examples is small, it seems obvious that in the

adults the interval between the spring/summer and the autumn/winter ecdyses

is the longer of the two intermoult periods.

Ecdysis was irregular in the immature females, but some correlation

242 ANNALS OF THE SOUTH AFRICAN MUSEUM

between the size of the animals and the length of the intermoult periods is

indicated in table tro. 7

TABLE I0

Intermoult periods in the various size classes of juvenile females

No. of Intermoult Mean duration

intermoult periods of intermoult

periods in days periods in days

2 Tp ek 76°5

ee ee 87°7

85, 102 93°5

87-223 157°8

152-234. 185-2

In the 50 to 59 mm size class short periods of 87, 120, and 127 days were

referable to three animals measuring 53 to 57 mm in T.C.L. The other animals

in this and in the 60 to 69 mm size class underwent two ecdyses a year, one in

spring or summer and the other in autumn or winter. Two distinct intermoult

periods were thus recognized in several animals (tables 11, 12).

TABLE II

Duration of the spring/summer to autumn/winter intermoult

period in juvenile females

Intermoult Mean duration

TKO EG No. of periods of intermoult

in mm specimens in days periods in days

50-59 3 203-292 Q11°7

60-69 3 152-234 195°3

TABLE 12

Duration of the autumn/winter to spring/summer intermoult

period in juvenile females

Intermoult Mean duration

CLL. No. of periods of intermoult

in mm specimens in days periods in days

50-59 2 141,151 146

60-69 I 155

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS 243

As in the adults, there is a difference between the mean duration of the

two intermoult periods in the juvenile females, but too few specimens were

handled for a true assessment to be made of the length of either period. In

agreement with Fielder’s (1964a) observations on 7. novaehollandiae, however, a

progressive lengthening of the intermoult period is indicated (table 10) as the

animals increase in size. Fielder (1964a) has also ascertained that in

33 fF. novachollandiae, measuring from 5:0 to 5:9 cm in T.C.L., the average inter-

moult period was 137 days from July to December and 158 days from January

to June.

Mowu.LtTING FREQUENCIES AND ANNUAL RATE OF GROWTH

No observations were made of the behaviour of captive immature male

rock lobsters, but some information was gained regarding the consecutive

ecdyses of several juvenile females and adult males and females.

Although the records are somewhat meagre, a tentative estimate of the

mean annual frequency of moulting may be obtained from a consideration of

the duration of the intermoult periods in the various size classes (tables 13, 14).

In addition, the probable annual increase in T.C.L. has been appended to each

size Class in tables 13 and 14.

In collating the data, the two short intermoult periods of 119 and 105 days,

which were omitted in tables 8 and g, have been included in table 13 in the

mean intermoult periods of the 70 to 79 mm and 80 to 89 mm size classes

respectively.

TABLE 13

Mean frequency of moulting and annual increase in T.C.L. in captive female rock lobsters

Mean Mean frequency | Mean increase | Annual increase

HGeL. No. of intermoult of moulting per moult im DCE.

in mm Specimens in days per annum in mm in mm

20-29 2 76°5 4°8 2°4 TES

30-39 3 87°7 4°2 2°6 10°9

40-49 - 93°5 3°9 2°8 10°9

50-59 8 157°8 2°3 2°5 58

er ehiitle rants se EW asa Liar nib oh is seer) §

a ee ee 18 43

| ene oh ee iy as sh aea. ee

_ | |

100-109 2 169°0 2°2 I°7 3°79

244 ANNALS OF THE SOUTH AFRICAN MUSEUM

TABLE 14

Mean frequency of moulting and annual increase in T.C.L. in captive adult male rock lobsters.

No specimens were available in the 80-89 mm size class

Mean Mean frequency | Mean increase | Annual increase

tM Ops bp No. of intermoult of moulting per moult in. TCL,

mmm Specimens in days per annum mmm in mm

70-79 I 15470 2°4 24 5°8

90-99 4 yp ee PAI aaa 5:0

100-109 5 179°8 BIG) psig 4°6

ee ee eS) ee ee

110-119

The moulting frequencies seem to be rather variable, but it may be

concluded that, as the animals increase in size, they undergo fewer ecdyses a

year. It is also apparent that there is a corresponding diminution in the annual

growth rate.

The above results indicate that the smallest rock lobsters may moult at

least four times a year and that, correlated with the lengthening of the inter-

moult periods in the larger size classes, the number of ecdyses is reduced to

three and then to two a year.

Two or more ecdyses a year occurred in only ten of the experimental

juvenile females, and direct observations suggest that the transition from three

to two annual ecdyses may take place in animals with T.C.L. measurements of

53 to 57 mm. The occurrence of biannual ecdyses in the 60 to 69 mm size class

is of interest and may denote that the moulting rhythm of young animals

approaching puberty is similar to that of the adults.

A higher moulting frequency in immature than in mature specimens of

Jasus has been reported by Hickman (1945) and Bradstock (1950). More

recently, Fielder (19642) has found that in 7. novaehollandiae the average number

of annual ecdyses is three in specimens with a rostrum-carapace length of

5:0. to 7-9 cm and two in the 8-0 to 8-9 cm size class.

As shown in table 7, two ecdyses a year took place in 28 adults which

survived their first ecdysis in captivity. It may also be remarked that, in the

few animals which moulted more than twice, the cyclical sequence (Paterson,

19694) was sufficiently regular to warrant the conclusion that a definite and

recurrent moulting rhythm is established in individual rock lobsters.

It was suggested by Gilchrist (1913) that the adults of 7. lalandi probably

moult once a year, the males in spring (September and October) and the

females in mid-winter. In four captive females, however, he observed that two

of them moulted annually, while two ecdyses a year occurred in the other two

animals. Except for the omission of the annual dates, the table and comments

published by Von Bonde & Marchand (1935) are similar to Gilchrist’s (1913)

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS 245

findings and provide no further information as to the frequency of ecdysis in

mature females.

More recent observations by Matthews (1962) and Heydorn (1965) on

populations of adult rock lobsters off the west coast of South Africa have not

determined that there is more than one ecdysis a year. From his investigations

on the biology of 7. lalandi found off Robben Island, Heydorn (1966) has

concluded that most of the adult females moult in late autumn or early winter

(May and June) before mating begins and that there is a marked increase in

the ecdysis of adult males during late spring and summer (November to

February).

It has also been inferred by Heydorn (1966) that captive ‘unfertilized

females undergo an untimely moult four or five months after the normal

annual moult’. Nevertheless, biannual ecdyses were indicated in both sexes of

the present captive animals (tables 7-9), and a spring/summer ecdysis took

place, not only in unmated females, but also in females which had previously

been in berry.

Investigations by Crawford & De Smidt (1922), Nakamura (1940),

Sutcliffe (1953), Lindberg (1955), George (1962), and Sheard (1949, 1962)

have also proved that the mature females of various species of Panulirus undergo

two ecdyses a year, one before mating and the other after the eggs have hatched.

It is even more pertinent to the present study that Grua (1964) has reported

the occurrence of winter and summer ecdyses in the females of Fasus paulensis.

OVIGEROUS SETAE

After it was found that the captive adults moulted twice a year, more

attention was paid to the condition of the ovigerous setae. During a period of

two years, 70 observations were made of the setal arrangement at the successive

ecdyses of 33 females whick either possessed ovigerous setae when they were

collected or acquired them at their first ecdysis in captivity.

If, as is detailed in the appendix, the presence and absence of the ovigerous

setae are respectively represented by plus and minus signs, the following

definition of the setal history relative to the biannual ecdyses is obtained:

Autumn /winter ecdyses: + = 31; — =5

Spring/summer ecdyses: + = 3; — = 31

In an analysis of 37 pairs of ecdyses referable to the 33 females (see

appendix), the sequence of setal (++) and a-setal (—) ecdyses was found to be:

Many to a Oar ie 2342 aia 2 laa

The above summaries show that, whereas 31 of the females possessed

ovigerous setae at the autumn/winter ecdysis, only three females produced

setae at the spring/summer ecdysis. It is also apparent that in most of the

females there was an almost regular biannual development and loss of the

ovigerous setae.

246 ANNALS OF THE SOUTH AFRICAN MUSEUM

There is no doubt that the ovigerous setae are acquired when the females

attain sexual maturity. Consequently, if the alternation of the setal and the

a-setal phases observed in captive females is normal, it suggests that in a large

population of rock lobsters there may sometimes be a number of mature females

which lack ovigerous setae. Such females could be described as being in a state

of sexual rest and would resemble immature animals if the setae had been shed

at an ecdysis following the incubation of the eggs.

Little is known about this aspect of 7. lalandi in the field, but Bradstock

(1950) has concluded that a relatively high percentage of the females of

J. edwardsui retain the ovigerous setae after the eggs have hatched. In Panulirus,

on the other hand, accounts by Nakamura (1940), Sutcliffe (1953), and George

(1962) indicate that the females moult and shed the ovigerous setae after the

breeding season.

In F. lalandi large ovaries containing brick-red or orange eggs are typical

of females possessing long ovigerous setae. It was therefore of interest that

apparently mature ovaries were observed in three unmated females after a

lengthy period of captivity and upon renewal of the ovigerous setae at the

winter ecdysis. By contrast, in a number of unmated and _ post-ovigerous

females, which lost the ovigerous setae at the spring/summer ecdysis, the

ovaries were narrow and only slightly folded and the eggs were small and white.

Apart from some observations by George (1958) and Fielder (19640), the

correlation between the condition of the ovaries and the setal arrangement in

the Palinuridae seems to have been hardly investigated. The significant results

of experiments performed by Charniaux-Cotton (1960) on Orchestia gammarellus

may, however, be mentioned in this connexion. In this amphipod it has been

shown that the oostegites revert to the juvenile state and have shorter ovigerous

setae when the ovaries are inactive. It has been concluded that the ovigerous

setae of Orchestia are temporary sexual characters associated with the incubation

of the eggs and that their growth is induced by an ovarian hormone which is

secreted during vitellogenesis.

It seems reasonable to suppose that similar conditions may prevail in some

palinurids and that the absence of ovigerous setae is connected with the

immature state of the eggs between one reproductive cycle and the next. At

such a time, hormonal control of the external secondary sexual characters

would be precluded and, if an ecdysis ensues shortly after the eggs have hatched,

it is unlikely that ovigerous setae will be formed at this ecdysis. It thus appears

that, even though the moulting frequencies of experimental animals suggest

that the mature females of 7. lalandiit may moult twice a year, only one incu-

batory period will be possible if the ovigerous setae are shed at one or the other

of the two ecdyses.

Grua (1964) has concluded that the damaged setae of post-ovigerous

females of Jasus paulensis are replaced at a summer ecdysis which is quite

distinct from the one that precedes mating. The present observations indicate,

however, that in most captive 7. lalandii females the ovigerous setae are renewed

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS 24°77

only at the premating ecdysis. There is also a strong probability that the latter

takes place in late autumn or winter in some females (31 in the present study)

and in spring in other females (3 in the present study).

If this opinion is correct, it may explain why one of the captive females in

Gilchrist’s (1913) experiments spawned on two different occasions after the

second ecdysis, while another female spawned after the first ecdysis. As the

moulting cycles of these two females are more or less comparable with the

present findings, it is surmised that the ovigerous setae were developed at the

second ecdysis in Gilchrist’s first female and at the first ecdysis in the second

female.

Although most of the females of 7. lalandii are apparently equipped for

carrying eggs after the autumn/winter ecdysis, it has been shown that a consider-

able number of the captive females discarded their ovigerous setae at the

spring/summer ecdysis. It may be noted, however, that two females, which

were captured in berry towards the end of winter, did not moult until the

‘following autumn and the ovigerous setae were again renewed and not lost.

As these females were included in an independent tagging experiment, they

have been omitted from the above considerations. Nevertheless, if their

behaviour was normal, it introduces the possibility that in some females the

post-incubatory ecdysis may be omitted and that the premating ecdysis may

occur at slightly different times in consecutive years.

REPRODUCTION

Most of the mature females were provided with apparently suitable part-

ners, but mating was not observed and only five females produced eggs. Two

of the females were soft when collected and, as they became ovigerous in

11 and 18 days after their capture, it was assumed that mating had taken place

in captivity some considerable time after they had moulted in the field.

The other three females moulted in captivity and spawned respectively in

9g, 20 and 63 days after ecdysis. Two of them (T.C.L. 72, 79 mm) were of

interest in that they spawned in spring and were consistent with Heydorn’s

(1966) conclusion that the smaller females are usually in berry later in the year

than the larger femaies.

These few results suggest that the interval between ecdysis and egg-

production is variable. The period of 9 to 20 days recorded in four of the females

is probably more normal than the delay of 63 days observed in the fifth

individual. At all events, it was noticed that most postmoult females appeared

to avoid the males for several days, and it is surmised that the females are

_ probably too weak to spawn immediately after ecdysis. Contrary to the opinions

of some observers, it has been inferred that ecdysis and spawning are not

necessarily closely associated in 7. lalandit.

Retrogression of the ovaries and autolysis of the eggs have been suggested

as probable causes of sterility in several species of captive decapods. Resorption

248 ANNALS OF THE SOUTH AFRICAN MUSEUM

of yolk from the eggs of Homarus americanus has been cited by Allen (1895) and

has also been reported in Orconectes limosus (Cambarus affins) by Andrews

(1906), in Diogenes pugilator by Bloch (1935), and in J. lalandii by Heydorn

(1966).

It is possible that similar processes may have contributed to the negative

results obtained in most of the present breeding experiments. Nevertheless, in

at least three females the ovaries were large and no yolk resorption was apparent

when they were examined macroscopically after an ecdysis at which long

ovigerous setae were developed. As two of the females were in captivity for over

a year and had survived three ecdyses, it seems that, although they did not

spawn, the sexual cycle had not been markedly affected. No sections of the

ovaries of these animals were prepared, and it is not known if the eggs were

fully mature. Until further information on this aspect is available, a decline in

the fertility of the females during long periods of captivity cannot be

excluded.

From current observations, it is also doubtful if young females become

mature in captivity. A few small females were maintained for two years, by

which time the T.C.L. was 69 to 71 mm. Even though females smaller than this

have occasionally been collected in berry, none of the young experimental

females acquired ovigerous setae and the ovaries were immature. A likely

explanation is that maturation of the ovaries may be delayed or inhibited by a

lack of factors essential to vitellogenesis.

Another reason for the persistent sterility in most of the captive adult

females may have been the approximate coincidence of their ecdyses with those

of the males. As the captive males also moulted twice a year, it is feasible that

they are subject to periods of sexual impotence and that, unless they moult

some time before the females are furnished with ovigerous setae, mating will

not take place.

When soft-bodied males were placed in tanks containing females, which

had moulted and were known to have ovigerous setae, no eggs were produced.

Assuming that metabolic changes correlated with moulting had affected the

male sexual behaviour, it is possible that the proper stimulus for mating was

lacking.

The association of freshly-collected, hard-shelled males and females, which

had either recently moulted or which moulted in the presence of the males, was

equally ineffectual. Although the females possessed ovigerous setae, no sexual

activity was displayed.

As most of the females moulted successfully but failed to reproduce, some

of them may have been either too large or too small for the available males. In

one experiment, however, a female (T.C.L. 79 mm), which had moulted in

isolation, spawned six days after a male (T.C.L. 100 mm) was introduced into

the tank. Similarity in the size of the partners may, therefore, not be an essential

attribute to successful mating, but intercourse between small males and larger

females seems unlikely. |

—

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS 249

ATTEMPTS AT ARTIFICIAL INSEMINATION

When it became apparent that reproduction had failed in most of the

experiments during the first half of 1966, the following four rather crude

attempts were made to inseminate a number of freshly-collected mature

females. All the females were soft when captured and the semen was removed

directly from the vasa deferentia of mature males.

I. Spermatophoric material was smeared on the sterna behind the

gonopores of three females (T.C.L. 82, 82, 83 mm).

2. Seminal substance was pressed with a spatula into the gonopores of

three females (T.C.L. 75, 77, 80 mm).

The above females were captured on 26 June 1966 and were isolated for

about three weeks to ensure that they had not mated recently in the field.

In the next two experiments, injection syringes with blunted fine and

coarse needles were used to introduce spermatophoric matrix into the oviducts

of four females. ‘These females were isolated for nine days after their collection

on 26 July 1966.

3. Pure spermatophoric material was injected into the oviducts of two

females (T.C.L. 82, 84 mm).

4. As some difficulty was experienced in drawing sufficient quantities of

the viscous, transparent spermatophoric mass into the syringes, a dilution of

equal parts of seminal material and sea water was injected into the oviducts of

two females (T.C.L. 73, 77 mm).

In each experiment the treated animals and control females of similar.

sizes and condition of sheil were placed in separate tanks. They were observed

for at least a fortnight, but none of them produced eggs.

All the females seemed to be unusually restless for several days after treat-

ment. They moved about, either with the abdomen fully extended and with the

pleopods pendent, or with the abdomen tightly flexed and simulating the

ovigerous posture.

The somewhat rough handling of the specimens appeared to have no ill

effect. One female, which moulted in captivity before the experiments, survived

two further ecdyses. Another female moulted 36 hours after the injection of

diluted seminal fluid into the oviducts and had a second ecdysis six months

later. The other females had obviously moulted in the field before their capture

and, although most of them were maintained for several months in tanks

containing males, reproduction did not take place.

Assuming that artificial insemination would have induced spawning,

one reason for the negative results could have been the time lapse between the

_ winter ecdysis of the females and the beginning of the experiments. Nevertheless,

previous comments (p. 247) show that 7. lalandiz females are capable of repro-

duction some time after they moult.

It has also been suggested (p. 246) that in mature females a period of sexual

quiescence occurs after the breeding season and coincides with the immature

condition of the ovaries and the loss of the ovigerous setae. All the females in

250 ANNALS OF THE SOUTH AFRICAN MUSEUM

these experiments possessed long ovigerous setae, and it was assumed that they

had not been in captivity long enough for the production of any radical changes

in the state of the ovaries. Physical contact and sexual stimulation by hard-

shelled males may, however, be essential to successful reproduction in this

species.

Several tests indicated that the spermatophoric substance remains soft and

simply disintegrates after being immersed in sea water for a few days. It is

therefore probable that most of the spermatophoric material which was applied

to the sterna of the first three females was washed off the surface soon after the

animals were returned to the tanks.

Another factor which seemed to militate against the possible fertilization

of the internal eggs of the females used in the other experiments was the rapid

expulsion of the injected seminal matter through the gonopores, presumably by

internal pressure either within the body cavity or in the oviducts. Thus, even

if the females had been capable of reproduction at this juncture, it is doubtful

if sufficient spermatophoric material was retained in the oviducts to impregnate

the eggs.

No valid conclusions concerning the method of fertilization in 7. lalandu

can be drawn from these simple experiments. However, if true copulation takes

place, it is still uncertain how sperm intromission is effected. It is also question-

able if there is adequate provision for the storage of spermatophoric material

in the oviducts. On the other hand if fertilization is external, the consistency

and the rapid deterioration of the exposed contents of the vasa deferentia

suggest that the eggs must be laid shortly after mating.

HIsTOLOGY OF THE REPRODUCTIVE ORGANS

Histologically the gonads of both sexes of 7. lalandii seem to be practically

identical with those of 7. novaehollandiae (Fielder, 19645). In both species the

testes are intricately coiled tubes and the parts of the vasa deferentia are sharply

differentiated. In the present material they are divisible into three regions, the

first and third of which are similar to the proximal and distal vas deferens of

F. novaehollandiae.

The slender, convoluted proximal portion of the vas deferens (fig. 4.4) is

oval in section and its lumen is reduced to a narrow, keyhole-shaped slit by the

arrangement of the columnar epithelium into small cells at the poles and

elongate, very narrow cells along the two sides.

The middle and distal parts of the vas deferens are wider than the first

part and each has a thick muscular wall. The slightly coiled middle region

(fig. 4B) is lined by a villiform glandular epithelium which is produced into a

conspicuous typhlosole. The latter underlies a surface streak, called the hyaline

line by Matthews (1951, 1954a, 19545), along which the wall appears to be

somewhat thinner and less muscular. Some similarity is evident between

sections of this region and those of the enlarged swollen part of the vas deferens

251

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS

vitelline membrane

ra ANAS

sc HRGRRUCZCOT ES

wu |

Se ey

sOGonc orp.

voa:7,

g oy

TS di 228!

spermptozoa~ gage Nor SY)

p Je] ~S<> apR Pes

PET TR

M7 matrix

ps ae

2 ee

rN See Soe

s epDreen ee MEDEN ys bs

OSNG s cier oka)

7 hye NY os,

LET SANS i E

Ee D

Fic. 4. Diagrammatic sections of the organs of a mature male (T.C.L. 110 mm) and an

ovigerous female (T.C.L. 85 mm).

A. Transverse section through the proximal portion of the vas deferens.

B. Transverse section of the middle region of the vas deferens.

C. Transverse section of the distal part of the vas deferens.

D. Longitudinal section passing through the junction of the ovary and the oviduct.

PAS] ANNALS OF THE SOUTH AFRICAN MUSEUM

of Panulirus penicillatus depicted by Matthews (1951).

The structure of the longer, straighter, and more cylindrical tube (fig. 4C),

which passes to the gonopore, is comparable with that of the distal part of the

vas deferens of 7. novaehollandiae (Fielder, 19645). The epithelium is flatter than

that of the middle region, and the typhlosole is small and conical in section.

The wide lumen is filled with a reticulated, folded matrix, the sinuous edges of

which are charged with spermatozoa. The definite spermatophoric wall and

the ‘putty-like’ basis of the matrix, described in P. penicillatus (Matthews, 1951),

seem to be lacking. It is believed by Matthews (1951, 1954a, 19545) and Fielder

(19645, 1964c) that most of the spermatophoric matrix is secreted by the

typhlosole. In the present sections the matrix is absent from the middle part

of the vas deferens (fig. 4B), but occurs in the other two regions of the duct

and also in the lumen of the testis.

In sections of the testis and the vas deferens and in smears of the sperma-

tophoric matrix of mature males the spermatozoa are small spherical bodies

measuring 6-8 to 13-5 w in diameter. As in Palinurus elephas (vulgaris) (Bloch,

1935), each spermatozoon has a distinct cupular nucleus and a globular

capsule. It seems unlikely that the spermatozoa are immature but, even in thin

smears, the characteristic radiating processes are indeterminate. Compara-

tively few spermatozoa are furnished either with one or with four to six delicate

filaments measuring about 7 to 11 » in length, and it is possible that some of

these are artefacts. Heydorn (1965) has, however, detected up to five spines

projecting from the spermatozoa of 7. lalandii, and Matthews (1951) has

mentioned the presence of rayed spermatozoa in Panulirus penicillatus.

From a number of ovigerous females examined, the one selected for

microscopic study of the gonads contained mature eggs in the ovaries and the

oviducts. In section, the ovary (fig. 4D) presents some of the features of the

ovaries of 7. novaehollandiae and 7. lalandii, described respectively by Fielder

(19645) and Heydorn (1966). Numerous oocytes with a maximum diameter of

120 pm are visible. Each is enclosed in a delicate vitelline membrane and a

layer of follicle cells, and in many of the oocytes the large nucleus contains a

conspicuous nucleolus.

The mature eggs in the ovary and in the oviduct measure from 400 to 500 »

in diameter. In addition to the thin vitelline membrane, each is surrounded by

a thicker envelope, the chorion (fig. 4.D), which is cribriform in section. It has

been shown by Herrick (1911) that the ripe eggs of Homarus americanus are

invested by a chorion which is secreted by the egg follicle. Bloch (1935) has also

suggested that, as the mature ovarian eggs of Diogenes pugilator are furnished

with both a vitelline membrane and a chorion, the latter is not a product of

the oviduct. Von Bonde’s (1936) conclusion that fertilization is internal in

F. lalandu is based on the supposition that the ‘chitinous’ egg membrane is

secreted by the lower part of the oviduct, but his interpretation of the egg

membranes may be somewhat erroneous.

The external eggs of 7. lalandit are approximately 750 » in diameter and

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS 253

are enclosed in an additional cuticular membrane which is continuous with the

egg-stalk and is believed by several authorities to be secreted by the tegumental

glands of the pleopods.

Although the residual mature eggs of ovigerous females are probably

destined to be resorbed, it seems logical to assume that, if intromission takes

place, traces of the seminal material will remain for some time in the oviducal

lumen. In the present serial sections of the oviduct, however, no spermatozoa

or any substance resembling the spermatophoric matrix could be detected.

In support of Von Bonde’s (1936) assumption that the eggs of F. lalandit are

fertilized in the proximal part of the oviduct, Fielder (1964c) has proposed that

the spermatozoa could be lodged in the folds of the oviducal wall after the

dissolution of the spermatophoric matrix. This hypothesis has been accepted

by Heydorn (1966) but has not been confirmed in the present investigation.

FERTILIZATION

Various accounts of mating and reproduction in macrurous Reptantia

have shown that spermatophoric masses are deposited by the male on the

sternum of a recently-moulted, soft-bodied female. It has also been suggested

that in Panulirus the sperm are released when the female uses the chelae of

the fifth pereiopods to scarify the hardened spermatophoric mass on her

sternum.

In the absence of resistant external spermatophores in 7. lalandii, it appears

that the main function of the chelae is to groom the abdomen and the pleopods

in both unmated and ovigerous females. Sexual dimorphism of the fifth pereio-

pods of this species was not apparent in small females with a T.C.L. of 26 mm.

As the females increased in size, however, the process of the propodite, which

forms an essential part of the chela, gradually became more pronounced.

Failure to detect the usual palinurid external spermatophores has been one

of the most puzzling features in the reproductive biology of 7. lalandii, and it

has often been assumed that fertilization is internal. Microscopic examination

of the reproductive organs and attempts to induce spawning by artificial

insemination suggest that this supposition is unlikely.

Dissections of a number of mature females at different times of the year

proved that no spermathecal enlargements occurred on the oviducts. It was

also observed that, in freshly-collected, soft females possessing long ovigerous

setae, the oviducts were distended with mature eggs awaiting extrusion.

Moreover, the male has no intromittent organs, and no support has been found

for Von Bonde’s (1936) theory that the eggs are fertilized in the oviducts.

The following observation is considered to be significant in that it seems to

indicate how fertilization is accomplished externally. It concerns an experiment

with afemale (T.C.L. 89 mm) and two hard-shelled males (T.C.L. 90, 100 mm).

The smaller male had been in captivity for a month and the larger male was

a recent introduction.

254 ANNALS OF THE SOUTH AFRICAN MUSEUM

On 25 July 1967, 36 days after the female had moulted, she was observed

in a crouched position about an hour after sunset. Mating had probably just

taken place, and it appeared that spawning was imminent. The abdomen was

strongly flexed and the tail-fan was fully expanded beneath the cephalothorax.

An immediate examination of the female’s sternum behind the gonopores

revealed the presence of a fairly thick layer of colourless, gelatinous matter

resembling the spermatophoric matrix.

Stained smears of some of the material, which had been kept overnight in

sea water, were somewhat unsatisfactory, but the nature of the deposit was

established by the identification of a number of spermatozoa. These were

comparable in size to the spermatozoa observed in fixed preparations of the

contents of the vasa deferentia. It was also of interest that several freshly-laid

eggs were found in the vessel in which the material had been stored.

In trying to evade capture, the female had obviously been greatly disturbed

at a critical period and much of the extraneous matter was dislodged in

handling her. By the following morning all traces of the hyaline deposit had

disappeared and, so far as is known, no more eggs were produced.

Although this isolated incident requires further confirmation, it suggests

that a soft, transparent spermatophore is placed on the female’s sternum

during the brief act of mating. If, however, the eggs are extruded almost

immediately after its application, the spermatophore will virtually be obscured.

(A short report of this opinion has been submitted in a letter to the Editor of

the South African Fournal of Science, Paterson, 1960).)

In several females it was observed that a considerable number of eggs were

still adherent to the sternum shortly after spawning had taken place. Pre-

sumably, these eggs were embedded in the remains of the colourless, viscous

spermatophoric mass. It is therefore probable that the eggs are fertilized on

the female’s sternum before they pass into the incubation chamber, where they

become attached by stalks to bundles of the ovigerous setae on the endopodites

of the pleopods.

INCUBATORY PERIOD

The period of incubation, during which the eggs are carried externally and

cared for by the females, was studied in 22 specimens, 19 of which were captured

in berry. As it was noted that in many of the broods some larvae emerged

earlier than others, a distinction has been made between the start and the

completion of hatching (table 15). The average interval between spawning and

the first signs of larval eyes in the incubating eggs is also indicated.

It was observed that the incubatory period differed somewhat in individual

females, but its duration may depend upon the size of the animal and on the

number of eggs carried. Although no regular pattern was evident, clearance of

the pleopods was generally effected sooner in some of the smaller than in the

larger females in each size class.

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS 255

TABLE 15

Summary of the development of the larval eyes and the length of the incubatory period in the

eggs of captive females

Development of Completion of

; larval eyes Start of hatching hatching

mC... 0 OR)

in mm specimens No. of Mean in No. of Mean in No. of Mean in

days days days days days days

60-69 I 32 44. 81

79-79 5 31-37 51-81 77-102

80-89 7 20-37 COT 61-81 62-2 78-102 94°2

——. ——_}———— +2°6 9 ~————] 4+ 11-4 =/————| +84

90-99 8 28-37 51-81 85-102

100-109 I 32 67 go

The eyes of the developing larvae in the eggs of females which were

collected in berry were apparent in 29 to 32 days. This compares with 28 to 37

days observed in the eggs of the present females which spawned in captivity.

It is therefore reasonable to suppose that most of the ovigerous females had

been captured shortly after the eggs were laid. On this premise, it has been

concluded that the incubation of the broods was generally completed in

approximately three months. Occasionally, however, the development of some

embryos was retarded, and a few females moulted in spring or summer before

all the larvae had hatched.

Gilchrist (1913) considered that the eggs of 7. /alandii are carried for about

five months, but in his holding experiments a shorter period of two to four

months is indicated. The latter time is more consistent with an incubatory

period of three to four months reported by Von Bonde & Marchand

(1935). )

The length of the incubatory period also appears to be variable in species

of Panulirus. While Allen (1916) estimated that in P. interruptus hatching was

completed in nine or ten weeks, a period of 18 days was recorded for P. argus

by Crawford & De Smidt (1922). Contrasted with these findings, it is thought

that the embryonic development lasts about a month in P. japonicus (Terao,

1929; Nakamura, 1940) and in P. argus (Sutcliffe, 1952).

DIscussION

Most of the important biological and ecological facts concerning 7. lalandi

have been acquired either by regular experimental trawling or tow-netting

(Gilchrist, 1913, 1918; Von Bonde & Marchand, 1935; Matthews, 1962;

2 56 ANNALS OF THE SOUTH AFRICAN MUSEUM

Lazarus, 1967) or by direct underwater observations of the animals (Heydorn,

1965, 1966). A few aquarium experiments have also been performed by some

of these investigators.

The primary object of the present study was an inquiry into the mating

behaviour and the process of fertilization, but several side-issues have also

proved of interest. Some of these are discussed in an attempt to correlate the

behavioural patterns of captive 7. lalandit with the data obtained in the field

by other workers.

Locomotor and feeding activities

As in other Decapoda, the Cape rock lobsters are more active at night

than during the day, and the suggestion of a bimodal rhythm in the locomotor

and feeding activities is partly substantiated by reports that the highest catches

of rock lobsters are obtained by local fishermen just before daybreak.

Foraging and feeding were also observed in some captive animals during

the day, and it seems unlikely that all the animals in thickly populated areas

will remain strictly nocturnal in their habits. Active feeding on the sea bed has

been observed during the day by Heydorn (1966), and some measure of

diurnal activity also appears to be implied in the successful day-time hauls of

rock lobsters described by Gilchrist (1913, 1918) and Von Bonde & Marchand

(1935).

It is well known that rock lobsters are easily caught in traps or hoop nets

baited with dead fish and, while the captive animals became accustomed to

feeding on a daily supply of pieces of fresh stockfish, live fish and other active

animals are probably not included in their normal diet.

A few small klipfish, Clinus superciliosus (Linnaeus), and a number of crabs,

Plagusia chabrus (Linnaeus), lived unharmed for several months in a tank

containing eight juvenile rock lobsters. It was only when the klipfish became

moribund after an overnight failure of the water and air circuits that they were

partly eaten by the rock lobsters. On other occasions, some klipfish and crabs

were introduced into tanks containing adult rock lobsters and, although the

crabs were often seen perched on the carapace of the rock lobsters, no attempts

were made to attack or dislodge them.

The cardiac stomach of freshly-collected animals usually contained

fragmented mollusc shells. Sometimes pieces of the exoskeleton of rock lobsters,

the byssus threads and complete shell valves of small mussels, Aulacomya

magellanica (Chemnitz), were also observed. It is therefore likely that their

natural food consists mainly of sedentary organisms such as Aulacomya, which are

said to be plentiful in rock lobster grounds. Omnivorous feeding is suggested

by the preference shown by captive animals not only for lamellibranchs, but

also for small limpets and kelp. They were, however, also attracted to the dead

bodies and the exuviae of other rock lobsters, and cannibalism, which only

occurred among crowded immature specimens, is probably resorted to when

other food is scarce.

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS 257

Moulting frequencies

Based on field investigations, it has been assumed that in F. lalandi there

is an annual ecdysis which takes place in late spring or summer in the males and

in late autumn or early winter in the females. Nevertheless, Gilchrist (1918)

recorded the capture of a number of soft-bodied males in winter and soft

females in summer. Moreover, in a study of 7. novaehollandiae, Hickman (1945)

remarked that, although most of the males moulted in spring, some catches

taken during autumn and winter contained a few soft males. It is likewise

pertinent that a winter ecdysis is indicated in the males of 7. tristani, 17 of which

were maintained for several months at the Sea Point Aquarium during 1967.

Five of the specimens died between 15 June and 23 August just before ecdysis

and, of the remaining animals, one moulted in May, five in July, three in

August, and three in September.

Biannual ecdyses were common in the captive adult males and females

of 7. lalandi, and a similar frequency seems to be implied in Fielder’s studies of

Jj. novaehollandiae. He established (1964a) that animals with a rostrum-carapace

length of 8-0 to 8-g cm moulted twice a year. It was also noted (1964c) that

four females moulted between August and October but, like most of the present

specimens, they did not reproduce.

Ecdysis and spawning

As it is generally agreed that ecdysis of the females precedes spawning, the

prevalence of a spring/summer ecdysis in the captive females is of interest, more

particularly because in some females it was correlated with the loss of the

ovigerous setae and the diminution of the ovaries. It has therefore been inferred

that the annual period of sexual activity in the females is characterized by the

development of long ovigerous setae and is followed by an ecdysis which marks

the beginning of a period of sexual quiescence.

The presence of the ovigerous setae, either at the autumn/winter ecdysis

or at the spring/summer ecdysis, suggests that two overlapping incubatory

periods are possible in a given population of 7. lalandii, one occurring from

winter to spring in some females and the second from spring to summer in other

females. This supposition is partly confirmed by Gilchrist’s (1913) observations

of captive animals and by his conclusion (1918) and that of Matthews (1962)

that spawning takes place later in some areas than in others.

Two distinct breeding communities are likewise suggested by Heydorn’s

(1966) field observations which have shown that, while the larger females

reproduce in winter, some smaller ovigerous females may still be found during

the following summer. The existence of separate spawning populations has

also been considered by Lazarus (1967) as a possible explanation of the occur-

rence of ‘winter’ and ‘summer’ hatching peaks associated with the distribution

of the phyllosoma larvae of 7. lalandit.

As embryonic development probably lasts about three months, the presence

of ovigerous females throughout several seasons obviously indicates that the

258 ANNALS OF THE SOUTH AFRICAN MUSEUM

females are not all fertile at the same time and that the pre-mating ecdysis of

the females must likewise be staggered over five or six months.

The relatively high percentages of ovigerous females recorded in the field

during spring therefore seem to be significant. They may support the theory

that the population is roughly divisible into two groups, one of which spawns

earlier than the other. The numbers of animals in the two groups may be

disproportionate but, if the incubatory periods of some females in each sub-

division happen to coincide in spring, an increase in the percentages of

ovigerous females collected at that time will be expected.

The cycles of ecdyses recorded in the field do not altogether disprove this

opinion, since minor peaks of ecdysis seem to be evident in spring and summer

in some of the areas so far investigated. As the pre-mating ecdysis of the females

is undisputed, it may be postulated that the general patterns of moulting and

ovigerous females should be more equable than is apparent in most field

observations, some of which hardly reflect the ecdyses of the females that

spawn in spring. Moreover, the spring/summer ecdysis frequently recorded in

captive unmated and post-ovigerous females has not been corroborated in the

field, but it is conceivable that it corresponds to the post-incubatory ecdysis

described in species of Panulirus.

Mating

Several investigators have indicated that in some captive Decapoda the

ovaries degenerate and that reproduction is consequently suspended, yet

Gilchrist (1913) demonstrated that mating recurred annually in a few experi-

mental specimens of 7. lalandit. In later observations, however, Gilchrist (1918)

found that after a year in captivity a number of females became sterile. He

tentatively suggested that spawning might occur biennially, but this seems

unlikely. Most of the present females were also infertile, although post-mortem

examinations of a few specimens, which had been captive for some time, showed.

that the ovaries looked mature and apparently had not been permanently

impaired.

It has been surmised that the sexual activity of captive males was limited

by their biannual ecdyses and that some were therefore unsuitable for mating

with the available females. The choice of physiologically adjusted partners is

doubtless a deciding factor in effective breeding experiments. When selecting

hard-shelled male rock lobsters, however, it is difficult to determine whether

they have already moulted in the field or whether they will moult during the

reproductive period of the females.

Gilchrist (1913) gave no particulars concerning the males used in his

experiments, but it is possible that the successful results were obtained by the

annual introduction of new hard-shelled males. No real support for this conjec-

ture was evident in the present study. It has been shown (p. 247) that five

females produced eggs after mating with fresh hard-shelled males, but it must

be noted that four of the females were also fairly recent acquisitions. The fifth

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS 259

female spawned in spring after four months of captivity, but other females,

which had been held for even longer periods, remained barren. It is therefore

possible that reproduction in some of the females was inhibited by factors in the

environmental conditions to which they were subjected.

It has usually been assumed that the annual reproductive cycle in

f. lalandii is actuated by the ecdysis of the female and that successful mating is

dependent upon the soft condition of the female. Judging by the present

observations, this conclusion seems doubtful. An unusually long interval of

63 days occurred between ecdysis and spawning in one female. In four other

individuals the eggs were laid in 9 to 20 days after ecdysis and while hardening

of the exoskeleton was in progress.

Variations in the time between the ecdysis of the female and spawning are

also evident in other reports on the biology of 7. lalandiz. An analysis of Gilchrist’s

(1913) experiments indicates that one female spawned on two occasions

shortly after ecdysis. In other females the interval between ecdysis and mating

was about 14 days and, when the time of mating was not recorded, there were

periods of 38 to 47 or 48 days between ecdysis and egg-production. Matthews

(1962) has also suggested that spawning probably takes place about three

weeks after the females have moulted.

From these accounts, it would seem that mating in 7. lalandit is possible

when the exoskeleton has begun to harden. This conclusion is, however, at

variance with the views of Von Bonde & Marchand (1935), who considered that

mating and oviposition are accomplished shortly after the female has moulted.

In this connexion, it is of interest that Sutcliffe (1952, 1953) has observed that

in Panulirus argus mating may take place when both the males and the females

are hard shelled.

Fertilization

Elsewhere (Paterson, 1968) it has been suggested that the male genital

apparatus of 7. lJalandit resembles that of the Natal rock lobster, Palinurus

gilchristi Stebbing. There is also a marked difference between the male gonopores

of these two species and those of the east coast rock lobster, Panulirus homarus

(Linnaeus), in an account of which firm external spermatophores have been

described by Heydorn (1966).

Hard external spermatophores comparable to those of Panulirus have not

been detected in 7. /alandii, but some importance is attached to the discovery of

a colourless gelatinous mass behind the gonopores of one of the present captive

females. As this substance contained spermatozoa and resembled the trans-

parent viscid contents of the vasa deferentia of mature males, it is believed that

in 7. lalandi a soft external spermatophore is deposited on the female’s sternum.

It is also anticipated that a similar external spermatophore will be found in

Palinurus gilchristi.

The latter expectation seems to be reasonable enough, because, according

to Brocchi (1875) and Bloch (1935), one investigator (Coste, 1860) reported

260 ANNALS OF THE SOUTH AFRICAN MUSEUM

the presence of external spermatophores in the langouste, Palinurus elephas

(Fabricius). In addition, some early workers were convinced that true copulation

did not take place in Palinurus and other Macrura. It was postulated that intro-

mission of sperm would hardly be effective in fertilizing the large number of

eggs produced by each female.

A similar argument would be equally cogent in speculations regarding the

method of insemination in F. /alandi but is no longer considered to be necessary.

The present conviction that a soft external spermatophore occurs in this species

confirms previous conjectures (Paterson, 1968) that fertilization is external and

not internal, as has formerly been supposed. The spermatophore of 7. lalandii

apparently lacks the ‘putty-like’ matrix described in species of Panulirus, and

the fact that it is transparent and transient probably explains why it has

hitherto been overlooked.

SUMMARY

Some aspects of the behaviour of juvenile females and adult males and

females of the Cape rock lobster, Fasus lalandii (H. Milne Edwards), were

studied in holding experiments at the Sea Point Aquarium, Cape Town, from

March 1966 to July 1968.

Shortly after their introduction into the tanks, the animals established

territorial rights to particular corners or rock niches and defended them

against intruders.

Reference is made to some of the grooming operations frequently observed.

Direct observations of 18 to 20 adults over a period of four months suggested

the presence of a bimodal diurnal rhythm in the locomotor and feeding activi-

ties. The normal behaviour was probably affected by the daily introduction of

food during the afternoon but, even so, the levels of activity were generally

higher at night than during the day.

Ecdysis, which usually took place at night, was observed in eleven juveniles

and seven adults of 7. lalandi and also in three males of F. tristan. An account

is given of the process which is essentially similar to that described by several

authors in Panulirus. Actual exuviation was completed in three to five minutes

in 7. lalandi and in four to seven minutes in 7. tristant.

No significant difference was found in the average growth rates of the adult

males and females, but the mean increments in the T.C.L. of the juvenile

females were probably slightly higher than those of the adult females. As the

animals increased in size, a progressive decrease in the growth rate was observed.

The mean moulting frequency decreased from 4:8 to 2:0 per annum as the

animals became larger. Biannual ecdyses, one in autumn or winter and the

other in spring or summer, occurred in the 60 to 69 mm T.C.L. size class of

juvenile females and in both sexes of the adults.

A study of the ecdyses of 33 mature females revealed that the ovigerous

setae were frequently developed at the autumn/winter ecdysis and were shed at

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS 261

the spring/summer ecdysis. Only three of the females acquired ovigerous setae

at the spring/summer ecdysis. The loss of the ovigerous setae was correlated

with a diminution in the size of the ovaries.

Apparently compatible males and females were selected for breeding

experiments, but most of the females failed to spawn. The coincidence of the

ecdyses of the captive males and females and degeneration of the ovaries may

have contributed to the persistent sterility in the adult females. Maturation of

the ovaries also seemed to be delayed or inhibited in young captive females

and was probably occasioned by the unnatural environmental conditions.

Mating and spawning were not observed, but four females produced eggs

in g to 20 days after ecdysis. In a fifth female, which reproduced in spring, there

was a period of 63 days between ecdysis and spawning. It has been concluded

that the soft condition of the female is not necessarily a prelude to mating and

egg-production.

The assumption that fertilization is internal in 7. lalandi has not been

confirmed. Attempts to inseminate several mature females gave negative results,

and serial sections of the ovary and the oviduct of an ovigerous female showed

no traces of spermatophoric material.

In one female, which had probably just mated and which seemed about to

spawn, a transparent seminal mass, resembling the viscous contents of the vasa

deferentia of mature males, was found applied to the sternum behind the

gonopores. It is believed that a soft, transient spermatophoric mass occurs in

fj. lalandiu and that the eggs are fertilized externally on the female’s sternum

shortly after mating takes place.

ACKNOWLEDGEMENTS

The writer is greatly indebted to the Director of the Division of Sea

Fisheries, Cape Town, for permission to use the facilities at the Sea Point

Aquarium; to Dr. A. E. F. Heydorn of the Oceanographic Research Institute,

Durban, and to Mr. G. G. Newman and his colleagues at the Division of Sea

Fisheries for supplying the experimental animals and for useful discussions of

some of the problems involved in their maintenance; to Mr. S. E. Hope, Divi-

sion of Sea Fisheries, for checking the mathematical calculations; to Mr. G. G.

Newman for valuable suggestions concerning the interpretation of the statistical

results; and to the South African Council for Scientific and Industrial Research

for financial support during the prosecution of this study.

The Trustees of the South African Museum thank the C.S.I.R. for a

grant in aid of publication.

REFERENCES

ALLEN, B. M. 1916. Notes on the spiny lobster (Panulirus interruptus) of the California coast.

Univ. Calif. Publs Zool. 16: 139-152.

ALLEN, E. J. 1895. The reproduction of the lobster. 7. mar. biol. Ass. U.K. 4: 60-69.

262 ANNALS OF THE SOUTH AFRICAN MUSEUM

ANDREWS, E. A. 1906. Egg-laying of crayfish. Am. Nat. 40: 343-356.

Biocy, F. 1935. Contribution a étude des gamétes et de la fécondation chez les crustacés

décapodes. Trav. Stn zool. Wimereux 12: 181-279.

Brapstock, C. A. 1950. A study of the marine spiny crayfish Jasus lalandii (Milne-Edwards)

including accounts of autotomy and autospasy. Zoology Publs Vict. Univ. Coll. 7: 1-38.

Broccui, P. 1875. Recherches sur les organes génitaux males des crustacés décapodes. Annls

Sci. nat. Zool. (6) 2: 1-131.

CaRLISLE, D. B. 1953. Studies on Lysmata seticaudata Risso (Crustacea Decapoda). III. On the

activity of the moult-accelerating principle when administered by the oral route. Pubbl.

Staz. zool. Napoli 24: 279-284.

CHARNIAUX-CoTTon, H. 1960. Sex determination. Jn Waterman, T. H., ed. The physiology of

Crustacea, 1: 411-447. New York & London: Academic Press.

CRAWFORD, D. R. & DE Sint, W. J. J. 1922. The spiny lobster, Panulirus argus, of southern

Florida: its natural history and utilization. Bull. Bur. Fish. Wash. 38: 281-310.

Dracu, P. 1939. Mue et cycle d’intermue chez les crustacés décapodes. Annls Inst. océanogr.,

Monaco 19: 103-391.

Dracu, P. 1950. Les étapes évolutives de l’endosquellette chez les crustacés décapodes. C. r. hebd.

Séanc. Acad. Sci., Paris 231: 1563-1565.

FreLper, D. R. 1964a. The spiny lobster, Fasus lalandei (H. Milne-Edwards), in South Australia. I.

Growth of captive animals. Aust. 7. mar. Freshwat. Res. 15: 77-92.

FIELDER, D. R. 19645. The spiny lobster, Fasus lalandei (H. Milne-Edwards), in South Australia.

II. Reproduction. Aust. 7. mar. Freshwat. Res. 15: 133-144.

FIELDER, D. R. 1964c. The process of fertilization in the spiny lobster Fasus lalande: (H. Milne-

Edwards). Trans. R. Soc. S. Aust. 88: 161-166.

FIELDER, D. R. 1965. The spiny lobster, JFasus lalandei (H. Milne-Edwards), in South Australia.

III. Food, feeding, and locomotor activity. Aust. 7. mar. Freshwat. Res. 16: 351-367.

GeorcE, R. W. 1958. The status of the ‘white’ crayfish in Western Australia. Aust. F. mar.

Freshwat. Res. 9: 537-545-

GeorcE, R. W. 1962. Description of Panulirus cygnus sp. nov., the commercial crayfish (or spiny

lobster) of Western Australia. 7. R. Soc. West. Aust. 45: 100-110.

GitcuRisT, J. D. F. 1913. The Cape crawfish and crawfish industry. Mar. biol. Rep., Cape Tn

I: I-45.

GitcurisT, J. D. F. 1918. Crawfish investigations, including experimental hauls, artificial

rearing, and migratory movements of the Cape crawfish (asus lalandii). Mar. biol. Rep.,

Cape Tn 4: 1-43.

Grua, P. 1964. Le cycle sexuel et l’évolution des soies ovigéres des femelles de Jasus paulensis

(Palinuridae). Jn Carrick, R., Holdgate, M. & Prévost, J., eds. Biologie antarctique (premier

symposium, Paris, Septembre 1962): 471-480. Paris: Hermann.

Herrick, F. H. 1911. Natural history of the American lobster. Bull. Bur. Fish. Wash. 29: 149-408

Heyporn, A. E. F. 1965. The rock lobster of the South African west coast, Fasus lalandi’

(H. Milne-Edwards). I. Notes on the reproductive biology and the determination o%

minimum size limits for commercial catches. Investl Rep. Div. Sea Fish. S. Afr. 533 1-32.

Heyporn, A. E. F. 1966. The biology of the South African rock lobsters Fasus lalandi and

Panulirus homarus with notes on Jasus tristani. Ph.D. thesis, University of Cape Town.

Hickman, V. V. 1945. Notes on the Tasmanian marine crayfish, Fasus lalandii (Milne Edwards).

Pap. Proc. R. Soc. Tasm. 1945: 27-38.

Hoe, P. G. 1962. Elementary statistics. New York & London: Wiley.

Kuso, I. & Masupa, T. 1964. On the feeding pattern of the Japanese spiny lobster, Panulirus

japonicus (V. Siebold). Bull. Fap. Soc. scient. Fish. 30: 14-20.

Lazarus, B. I. 1967. The occurrence of phyllosomata off the Cape with particular reference to

Fasus lalandii. Investl Rep. Div. Sea Fish. S. Afr. 63: 1-38.

LINDBERG, R. G. 1955. Growth, population dynamics, and field behavior in the spiny lobster,

Panulirus interruptus (Randall). Univ. Calif. Publs Zool. 59: 157-248.

MattHews, D. C. 1951. The origin, development and nature of the spermatophoric mass of

the spiny lobster, Panulirus penicillatus (Olivier). Pacif. Sci. 5: 359-371.

Matruews, D. C. 1954a. The development of the spermatophoric mass of the rock lobster,

Parribacus antarcticus (Lund). Pacif. Sci. 8: 28-34.

THE BEHAVIOUR OF CAPTIVE CAPE ROCK LOBSTERS 263

MattTHeEws, D. C. 1954). A comparative study of the spermatophores of three scyllarid lobsters

(Parribacus antarcticus, Scyllarides squammosus and Scyllarus martensii). Q. Fl microsc. Sci. 95:

205-215.

MattTHEws, J. P. 1962. The rock lobster of South West Africa (asus lalandii) (Milne Edwards).

Size frequency, reproduction, distribution and availability. Investl Rep. mar. Res. Lab. S. W.

ajr. 72 1-61.

NAKAMURA, 9. 1940. Ecological studies on the spiny lobster, Panulirus yaponicus (V. Siebold), with

special reference to its conservation. 7. imp. Fish. Inst., Tokyo 34% 101-113.

Passano, L. M. 1960. Molting and its control. In Waterman, T. H. ed. The physiology of Crustacea,

I: 473-536. New York & London: Academic Press.

Paterson, N. F. 1968. The anatomy of the Cape rock lobster, Fasus lalandi (H. Milne Edwards).

Ann. S. Afr. Mus. 51: 1-232.

Paterson, N. F. 1969a. The moulting frequency in captive adult Cape rock lobsters, Fasus

lalandu (H. Milne Edwards). S. Afr. F. Sci. 65: 72-74.

Paterson, N. F. 1960). Fertilization in the Cape rock lobster, Fasus lalandiit (H. Milne Edwards).

S. Afr. F. Sct. 65: 163-164.

SHEARD, K. 1949. The marine crayfishes (spiny lobsters), family Palinuridae, of Western

Australia, with particular reference to the fishery on the Western Australian crayfish

(Panulirus longipes). Bull. Commonw. scient. ind. Res. Org. 24°73 1-45.

SHEARD, K. 1962. The Western Australian crayfishery 1944-1961. Perth: Paterson Brokensha.

SuTCLmFFE, W. H. 1952. Some observations of the breeding and migration of the Bermuda spiny

lobster, Panulirus argus. Proc. Gulf Caribb. Fish. Inst. 4: 64-69.

SuTcLIFFE, W. H. 1953. Further observations on the breeding and migration of the Bermuda

spiny lobster, Panulirus argus. F. mar. Res. 12: 173-183.

SutTcuirFe, W. H. 1956. Effect of light intensity on the activity of the Bermuda spiny lobster,

Panulirus argus. Ecology 37: 200-201.

TERAO, A. 1929. On the embryonic development of the spiny lobster, Panulirus japonicus

(V. Siebold). Fap. 7. Zool. 2: 387-449.

Tuomas, L. R. 1966. Moulting behaviour of the Western Australian crayfish Panulirus cygnus

George (Decapoda Reptantia). Crustaceana 112 111-112.

Travis, D. F. 1954. The molting cycle of the spiny lobster, Panulirus argus Latreille. I. Molting

and growth in laboratory-maintained individuals. Biol. Bull. mar. biol. Lab., Woods Hole 107:

433-4590.

Von Bonpbe, C. 1936. The reproduction, embryology and metamorphosis of the Cape crawfish

(Jasus lalandii) (Milne Edwards) Ortmann. Investl Rep. Fish. mar. biol. Surv. Div. Un. S. Afr.

6: 1-25.

Von Bonpve, C. & Marcuanp, J. M. 1935. The natural history and utilization of the Cape

crawfish, kreef or spiny lobster, Fasus (Palinurus) lalandii (Milne Edwards) Ortmann.

Bull. Fish. mar. biol. Surv. Div. Un. 8. Afr. 13 1-55.

264

ANNALS OF THE SOUTH AFRICAN MUSEUM

APPENDIX

Summary of the incidence of ovigerous setae in captive females.

No. 1966 66/67 1967 67/68 1968

aS oo

Dn iS)

++t++4+t4+4+4+4+4+44+

| +++

ioe)

++t++++t++ +4441

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. (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)

BuLLoucnH, 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. 19604. 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. 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, b; Liste: 11. Turton, 1932: 80.

—

FESS

fee

' wn

1) Ray ee

be }

(Pheer

=A — bs ho r 4 4

Sa nc ip OT epg : eee sor tere ee eo —— —— <= See a eee EB =

—tuvuart LIBRARIES _ SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S31UY)

> ra > = AD & ea Ey

> ke > rat D> ° Fel

4 a o = as a

STITUTION NOILNLILSNI S3IYVYGIT LIBRARIES SMITHSONIAN

Le. =F a Z =) ae ye a

ik NAS 5 S fee = ‘2 AS 3 =

| SS = iff > oe & Ge ae 5

mA 2° Yi), = a" s ne F

| > GQ = > : = ae =, :

th i Pd _ ae 77) by = w .

H1YVEAIT_LIBRARIES SMITHSONIAN _ INSTITUTION NOLLNLILSNI_NVINOSHLIWS Sa luvus

=z ‘ tu! d lJ a . “ay rie iy

: ze 3 z 2 ON

= <% re < at: NS Sa

= a = = GS MQ «

mee faa) -_— ae) wa. * rae)

j = ne = onl = a

STITUTION NOILNLILSNINVINOSHLINS S3INYVYU9IT LIBRARIES SMITHSONIAN

= aye = i = s yi

— w fe, ad ow ae ag o

5 Ps) NY NN = 29 Be GE. Se)

> IS Gs p ty V4. >

E 3 ANGE . &E Gr Li = \

D m \s mH os ee ee -

”) m Noo S m “” m

= wn re = on, < n

Yvagia LIBRARIES SMITHSONIAN INSTITUTION NOILMIILSNI_ NVINOSHLINS, $3

y ae = sures = TZ < =

, <= —s Y, fy

BS NN = S = GUY s 2

2 SIN 2 2 8 YG? Ee

E Wo 2 = ANC Gy agai = Zz

ee ne A oe : = }

"STITUTION NOILALILSNI_NVINOSHLINS SA1YVYIT_ LIBRARIES INSTITUTI

o = ow = ee =f

G os fe

«<f Cc . c= a Cc < ei, c

& =~] Co 4 feed |

= fo) es o m: 5S ine

= 5 - met a z |

AtYVeGIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS SJIYVUE

il cS = = : S a a7 re 4

to a= a pa w a6 ~

“a = tf, y a) = bo ‘ a and

; EG * E a E

“~ - Wy stig ra =

ip Z “lif? os = m 2 Louis a

ons = wn 5 ea it

NOILNLILSNI NVINOSHLINS S31YVYGIT LIBRARIES SMITHSONIAN INSTITUT!

op) ; P28 (9) a wes wn = ae. a

= S Sl hae aes = Gs

+ pth fe = = a ae . =i = 2 :

ZO ESI RE TCDS

z @ = = “py E bias =

> G = > os = tua = wy

Rl Sy pee 7p) ama ” Bi aes ” _

“AIYVESIT_ LIBRARIES SMITHSONIAN INSTITUTION NOILOLILSNI_NVINOSHLIWS Saluva

z= uw z ul 2 XS, as | re

S on “S - 4 SNe

= a = = ores YQ cc

| rs) =i A 3 Pa rs) So Neuss

Zz - z -! Pe * et =

STITUTION | NOLLALILSNI_NVINOSHLINS |S31YVYGIT LIBRARIES SMITHSONIAN INSTITUTIO

30 'NG Oo eet) Oo = fe) i za

Zemm = - : : - G%®

i S| aa = ci i Le A

ne) cy - oma ke wiles me % 3 Fiat ype

Wingsh> wn mM <2) aes w rm

a—

NVINOSHLINS S3IYVYGIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI_NYINOSHLINS

NOILNLILSNI SJIMVUSIT LIBRARIE

)

“SN

SMITHSONIAN

NOLLMLILSNI

NOILNLILSNI

LIBRARIES

NOILNLILSNI

LIBRARIES

NOLLALILSNI

INSTITUTION NOILALILSNI NVINOSHLINS S3IYVYEIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILS?

UTION

UVUgItT

INSTITUTION, NOILNLILSNI

SJIYVUSIT LIBRARIES §

INSTITUTION

Sal

NOILNLILSNI NVINOSHLIWS SMITHSONIAN NOILALILSNI

S S3INVYUGIT LIBRARIES

SSIYVYSIT LIBRARIES SAINYVUEITLIBRARIE

SMITHSONIAN

NVINOSHLINS

‘

‘°

NVINOSHLINS S3IYVYEIT LIBRARIES SMITHSONIAN INSTITUTION NOILNLILS!

NVINOSHLINS S3IYVUdtIT

SMITHSONIAN

NVINOSHLIW

SMITHSONIAN

NVINOSHLIWS

LIBRARIES SMITHSONIAN INSTITUTION NOILNLILSNI

) : Ls a WW a ran

oe oc = o . = oc

a << oe x % = <

lead c er fod Pa m~

~ = — fea)

ean oO oO _

=! —- = ad

NOILALILSNI SJINVYIT LIBRARIES SMITHSONIAN INSTITUTION NOILALILSNI” NVINOSHLINS. S3INVU@I1 LIBRARIE

qd!71 LIBRARIES

INSTITUTION NOILNLILSNI

ace faa]

= 5

a z : z =

| w \ co tf — w S = a

z Ygg, : Gy: E = We 5 5

| rong” SS: > YP Ana ram ie WS = >

| = , KS _ e7 ss Ie = RN ‘ WS = e

en SN si (ee 2 Xa 2 a

Li B RARI ES SMITHSONIAN INSTITUTION NOILLALILSNI _ NWINOSHLIWS ot! B RARI ES SMITHSONIAN INSTITUTION NOILNLILSA

<= Bit = gt, = aR» 5 ee = = =

@ : z i @y = = 2 ua 2 =

WY 2 |B G YP 4s % g 2 eS! 2 ae

or" 2 = ae les 2 = = = S

“S > i i = >" = SS Ss = >

ee alee ” sees ee 7) z ” : = Be z

NOILNLILSNI_ preva LIBRARIES SMITHSONIAN pNOILNLILSNI_ NVINOSHLINS LIBRARIE

7 ans | 39 NG WwW Vy tiJ e WwW