INSTITUTION NOILIILILON

S3iy¥vudl

LIBRARIES

NVINOSHLIWS

LIBRARIES SMITHSONIAN

ff.

NOILNLILSNI

LIBRARIES

NOILALILSNI

NVINOSHLINS S3!¥vVadll

SMITHSONIAN

NOILNLILSNI

INSTITUTION NOILNLILSNI

LIBRARIES

NOILNLILSNI

IBRARIES SMITHSONIAN

NVINOSHLIWS

OILNLILSNI

LIBRARIE:

NOILNLILSNINVINOSHLIWS

INSTITUTION

SSIYVYUSIT LIBRARIES

LIBRARIE

NOILALILS!

S3'yvVudIT LIBRARIES INSTITUTION NOILN

INSTITUTION NOILNLILSD

INSTITUTION NOILMLILSI

SJIMYVUGIT LIBRARIE

i z=

ow )

\y 0) =

Se Qe i -

oS, SENN |

LY = D

tc ae Oy m ; =

wn =

SMITHSONIAN INSTITUTION NOILONLILSNI NVINOSHLINS id B RA

ria = ~ Faas 22) ae

x ‘. = < = < ~

pS 3 = =

S Ya = 2 & ay 5

YM ~ Wee ow np, wn ”

RY ED \ So) 7 O

: NS ee = =

NVINOSHLINS Sa IYVYE!IT LIBRARIES SMITHSONIAN INSTITUTION Pili. .

: Z3 -- is

ul re wl = us

On = ce he = f&

ee = oe aS ox

a ame - Ps. co

~ as 23 me _

SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3luvudll LIB RA

7 z & z is .

2 z ) Lay a

se 2 ae = =

8) = P = a

m ” m ” =

wn = a Pye = n

Pe 1yVvuad eS BRARI ES SMITHSONIAN INSTITUTION |

= .s = a

o hig a o Tp ie : ws

O MYyiy = O Tr O

Zz Ci = 2 = , 2

> 2 = Ye = ; >

= 77) oa an bones

SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLIWS ~~

rs f | Z “ ar z

4 ” Ass tu

“s = 2 WN 2 4

= < = \ ee: =

4 ra + %N XX a =i

ye Oo _ O = O

. z = S Sy a >

NVINOSHLINS S3IYVYaIT LIBRARIES SMITHSONIAN INSTITUTION NOILAL

Sega: : aes :

= w = L, w =

Ya 5 : 5 Gy = 5

a i = = = =

WY 2 5 Q* m —g

_ op) am WM =o

SMIT SONTAN _ INSTITUTION NOLEALIZSN( — SVINOSHLIWS: &S 3 lYyYvualt as

ae z pee = me oe

2 \o i G&? : Zz

no ~~. wn fou 7) wv 7)

r BANE 6 Gf his. i Oo Be

8 AN WS Re ae = = =

= SS >’ > >" =

a ‘ae Z 7) a a

NVINOSHLIWS Sa 1YVug (1 LIBRARIES SMITHSONIAN _INSTITUTION NOILAL

Yyg = = YU * Z

NS : : = Ye: :

x SS oc Cc oc IY. eps SG Rag

NO 5 z mh Aroha 5

on ra: ‘o) sors 2D om

baits _ Ve ARSE os

: : eB& =

= oO za eS oo

S pe = Sion

S3JiIY¥vVYgIT LIBRARIES SMITHSONIAN INSTITUTION

S a Ss ae Pale Neel

Ss w ae ow

= a aa Meee

od > = aces

: ic, z

Zz ys Zz ys

MITHSONIAN INSTITUTION NOILALILSNI NVINOSHLINS S3lYvudit

NVINOSHLIWS

SMITHSONIAN

NVINOSHLIWS

IWINOSHLINS SAldVdd!T LIBRARIES _INSTITUTION

ef za z

pass oe) wo

oc mn y eis)

. a, %

< | 2 ‘2

m a ae

aa | Zz Zz

LIBRARIES SMITHSONIAN

SMITHSONIAN INSTITUTION NOILALILSNI NVINOSHLINS S3I1YVuUal)

on z i z

: ie O

> = zs s

& > ae =

Be) Ee ra

_ a = ee

0 2 ss ae:

pos laVd dl T_LIBRARIES SMITHSONIAN INSTITUTION

g i Z we g z Aa

Aa EAN ee = a

2 Gy * 2 :

> wn = HD is Ss .

Fa Cp) ‘S z YW”) >

MITHSONIAN INSTITUTION NOILMLILSNI_NVINOSHLINS saluvud!y

li at i

10 ul A

= = = :

Coq

ar ac S ac

, S es 6 a

i 2 oa 2 Stee

WWINOSHLINS (S31YVYAIT LIBRARIES SMITHSONIAN INSTITUTION

a pe

\ Oo as © -

We § 2 © Yi &

WE a = is WY,

SAS Se = cS Lebo

SNS Sa m a m

5 = w = ep)

SMITHSONIAN INSTITUTION NOILALILSNI NVINOSHLINS S3I1uYvVudtt

oe eS :

= {X = (f% = zs

Z Nw 5 Wy? =

2] SS WS — W Ye Uy ”) Nn

ot aN SANS Oo a Me ale Oo

Es AS i n= =)

= WN > = >

a a 2 a 2

BVINOSHIIWS Saldvaai LIBRARIES SMITHSONIAN INSTITUTION

: au. ys on

= = — Vly, ah

— (oa Uy pA nomad

. a x Uji Uj ai

a = fod pS

mM _ co. “i —

Sas oO ea i@)

4 SF ples

ai! fe y

: < fy Lif

a mo “if

Zz we ¢

NOILNLILSNI NVINOSHL1I

ie ae ‘>

. wo

2 >

= 3

Z HOG

LIBRARIES SMITHSON

cee wn

ean =

Zz C es

fo) SS Nogaro

e AN eg

i Wo =

=> RS Ss >

w : 2

NOILNLILSNI_ NVINOSHL

9) = ;

uu 7)

Fe a

< c

o o

en fe)

a Z

LIBRARIES SMITHSON

CF Zz

ov oO

ow ax

Po) ra Ve ¢

> = LY,

i = iy

m oe jf

Ww =

NOILNLILSN! NVINOSHL

w rds

= ae

= ht piel Ze

Se 4 )

i JL ?

2 “yy ie

> 77, =

Fe 7)

LIBRARIES SMITHSON

Z G2 |

” ud é;

a a Wit ie

Ge ow 49 Witz 5

. o Vy

NOILNLILSNI NVINOSHL

Zz ~ >

5 x

= \. S

— Ps) ee

tie = S&S.

EI BRARIES SMITHSO?

ze ep)

<< =

Fe =

Oo ao

Tp) 7)

ae O

i AN =

NN >

2) ae

NOILALILSNI NVINOSH:

(ep) Pd Ae

a nl eae

ac — WS

a = AS

(oa) Sm, NSN

Pa O SS

: ANNALS OF THE

S67% H AFRICAN MUSEUM

VOLUME 84

ANNALE VAN DIE

SUID-AFRIKAANSE MUSEUM

BAND 84

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

VOLUME 84 BAND

WRC:

S B:0.8:9'2)

FE)

ut é

5 S

“Vou nw WS

THE TRUSTEES OF THE DIE TRUSTEES VAN DIE

SOUTH AFRICAN MUSEUM SUID-AFRIKAANSE MUSEUM

CAPE TOWN KAAPSTAD

1981

e..

¥e

\ ; ie |

toe Lie

bm oe

e —a

= ¢ al

. 2 = i ws x e

7 as 7 :

4 a . :

<3 - Es F a

3 fi 3 oe

ie 6 -

= + ry

ad —_

= i

hie = \

ee ie = z

‘oe =

att

ao —

= 2 3 P

- Le

= -

i ; ‘ ' Me

= 2 = ~

——)

x = =

}

= =

vi i

= ad : ts

4 i

f ¥

{

SET, PRINTED AND BOUND IN THE REPUBLIC OF

"THE RUSTICA PRESS (PTY.) LTD., WYNBERG.

C914

LIST OF CONTENTS

BARNARD, J. L. see KARAMAN, G. S.

GRINE, F. E.

Cragievarus kitchingi Brink, 1965: a subjective junior synonym of Diademodon

tetragonus Seeley, 1894 (Reptilia, Therapsida). (Published July 1981.) ........

GRINE, F. E.

A new composite juvenile specimen of Australopithecus africanus (Mammalia,

Primates) from Member 4, Sterkfontein Formation, Transvaal. (Published July

WOT) B'S Cicer ane cacy See Meee) Wee, Steel eat oe ae eee ee ee

GOSLINER, T. M. & GRIFFITHS, R. J.

Description and revision of some South African aeolidacean Nudibranchia (Mollus-

CasGasiropoda)= (eublished uly OSI een eerener rye ea eee. ota oe

GRIFFITHS, R. J. see GOSLINER, T. M.

HENDEY, QO. B.

Palaeoecology of the late Tertiary fossil occurrences in ‘E’ Quarry, Langebaanweg,

South Africa, and a reinterpretation of their geological context. (Published July

NO SiES Mee EE eRe yee sc 52 ols y Soe g Ptah ARV oo hielo Se ead ead

KARAMAN, G. S. & BARNARD, J. L.

The synonymization of Triodos K. H. Barnard with Ampelisca Kroyer (Crustacea,

Amp mipoda) a (kublisnedeAueust 19815) ei = ase sae ene a 5 soe

Kine, G. M.

The postcranial skeleton of Robertia broomiana, an early dicynodont (Reptilia,

Therapsida) from the South African Karoo. (Published July 1981.) ...........

KLEIN, R. G.

Ungulate mortality and sedimentary facies in the late Tertiary Varswater Forma-

tion, Langebaanweg, South Africa. (Published July 1981.)...................

Page

169

105

203

233

OF THE SOUTH AFRICAN ~

— “” MUSEUM

CAPE TOWN

INSTRUCTIONS TO AUTHORS

1. MATERIAL should be original and not published elsewhere, in whole or in part.

2. LAYOUT should be as follows:

(a) Centred masthead to consist of

Title: informative but concise, without abbreviations and not including the names of new genera or species

Author’s(s’) name(s) A

Address(es) of author(s) (institution where work was carried out)

Number of illustrations (figures, enumerated maps and tables, in this order)

(b) Abstract of not more than 200 words, intelligible to the reader without reference to the text

(d) Introduction f ; ;

(e) Subject-matter of the paper, divided into sections to correspond with those given in table of contents

(f) Summary, if paper is lengthy

(g) Acknowledgements

(h) References

(i) Abbreviations, where these are numerous

3. MANUSCRIPT, to be submitted in triplicate, should be typewritten and neat, double spaced

with 2,5 cm margins all round. First lines of paragraphs should be indented. Tables and a list of

legends for illustrations should be typed separately, their positions indicated in the text. All

pages should be numbered consecutively.

Major headings of the paper are centred capitals; first subheadings are shouldered small

capitals; second subheadings are shouldered italics; third subheadings are indented, shouldered

italics. Further subdivisions should be avoided, as also enumeration (never roman numerals)

of headings and abbreviations.

Footnotes should be avoided unless they are short and essential.

Only generic and specific names should be underlined to indicate italics; all other marking

up should be left to editor and publisher.

4. ILLUSTRATIONS should be reducible to a size not exceeding 12 x 18 cm (19 cm including

legend); the reduction or enlargement required should be indicated; originals larger than

35 x 47 cm should not be submitted; photographs should be rectangular in shape and final

size. A metric scale should appear with all illustrations, otherwise magnification or reduction

should be given in the legend; if the latter, then the final reduction or enlargement should be

taken into consideration.

All illustrations, whether line drawings or photographs, should be termed figures (plates

are not printed; half-tones will appear in their proper place in the text) and numbered in a

single series. Items of composite figures should be designated by capital letters; lettering of

figures is not set in type and should be in lower-case letters.

The number of the figure should be lightly marked in pencil on the back of each illustration.

5. REFERENCES cited in text and synonymies should all be included in the list at the end of

the paper, using the Harvard System (ibid., idem, loc. cit., op. cit. are not acceptable):

(a) Author’s name and year of publication given in text, e.g.:

‘Smith (1969) describes...’

‘Smith (1969: 36, fig. 16) describes...’

“As described (Smith 1969a, 19696; Jones 1971)’

‘As described (Haughton & Broom 1927)...’

“As described (Haughton et al. 1927)...’

Note: no comma separating name and year

Dagination indicated by colon, not p.

names of joint authors connected by ampersand

et al. in text for more than two joint authors, but names of all authors given in list of references.

(b) Full references at the end of the paper, arranged alphabetically by names, chronologically

within each name, with suffixes a, b, etc. to the year for more than one paper by the same

author in that year, e.g. Smith (1969a, 19695) and not Smith (1969, 1969a).

For books give title in italics, edition, volume number, place of publication, publisher.

For journal article give title of article, title of journal in italics (abbreviated according to the World list o,

scientific periodicals. 4th ed. London: Butterworths, 1963), series in parentheses, volume number, part

number (only if independently paged) in parentheses, pagination (first and last pages of article).

Examples (note capitalization and punctuation)

BULLOUGH, W. S. 1960. Practical invertebrate anatomy. 2nd ed. London: Macmillan.

FISCHER, P.—H. 1948. Données sur la résistance et de le vitalité des mollusques. J. 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. 19606. Spawning behaviour, egg masses and larval development in Conus from the Indian Ocean.

Bull. Bingham oceanogr. Coll. 17 (4): 1-51.

THIELE, J. 1910. Mollusca: B. Polyplacophora, Gastropoda marina, Bivalvia. In: SCHULTZE, L. Zoologische

und anthropologische Ergebnisse einer Forschungsreise im westlichen und zentralen Siid-Afrika 4: 269-270.

Jena: Fischer. Denkschr. med.-naturw. Ges. Jena 16: 269-270.

(continued inside back cover)

ANNALS OF THE SOUTH AFRICAN MUSEUM

ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

Volume 84 ~ Band

July 1981 Julie

Part 1 Deel

PALAEOECOLOGY OF THE LATE

TERTIARY FOSSIL OCCURRENCES IN

‘E>’ QUARRY, LANGEBAANWEG,

SOUTH AFRICA, AND A REINTERPRETATION

OF THEIR GEOLOGICAL CONTEXT

©” B. HENDE Y

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 8000

Die ANNALE VAN DIE SUID-AFRIKAANSE MUSEUM

word uitgegee in dele op ongereelde tye na gelang van die

beskikbaarheid van stof

Verkrygbaar van die Suid-Afrikaanse Museum, Posbus 61, Kaapstad 8000

OUT OF PRINT/UIT DRUK

Ik, AUIS S39), KIS As he al); HUIS, 5, 7-2).

61 tpi); 70—4),8, 9a—25, 7), lou 3)2

11(@=2, 5, 7, tp), 15(4>5), 242), 27, 31053), 329), 33; 450)

EDITOR/REDAKTRISE

Ione Rudner

Kopieregnavrae aan die Suid-Afrikaanse Museum

ISBN 0 86813 009 5

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

PALAEOECOLOGY OF THE LATE TERTIARY FOSSIL OCCURRENCES

IN ‘E’ QUARRY, LANGEBAANWEG, SOUTH AFRICA, AND A

REINTERPRETATION OF THEIR GEOLOGICAL CONTEXT

Q. B. HENDEY

South African Museum, Cape Town

(With 17 figures and 9 tables)

[MS. accepted 26 November 1980|

ABSTRACT

The late Tertiary succession in the vicinity of Langebaanweg, Cape Province, is correlated

with the global sea-level changes of that period. The principal fossiliferous elements are the

Quartzose Sand Member (QSM) and Pelletal Phosphorite Member (PPM) of the Varswater

Formation, which were laid down during the global early Pliocene transgression. The nature of

the succession was determined by the physical geography of the region, while the preservation

of an unusually large body of early Pliocene sediment in the Langebaanweg area was due to the

development at that time of an overlying coastal barrier complex, part of which survived

subsequent erosion.

All elements in the succession are fossiliferous in parts. Approximately 230 invertebrate

and vertebrate taxa are recorded from the Varswater Formation, in which a variety of marine,

freshwater and terrestrial depositional environments are represented. Species assemblages,

body part representation and condition of specimens vary according to the facies from which

they are derived. The fossils reflect the changing environmental conditions of the late Miocene

and early Pliocene, when local temperatures moderated from tropical to temperate, rainfall was

changing from a summer to a winter maximum, and woodland vegetation was giving way to

more open grasslands and fynbos.

CONTENTS

PAGE

RVR O CU CHORUS Pears ee Reece Rios wie de Renn gem ac, 8 Or a eon wee arc Z

(SASTINSIE | ie sk oe ats, sac aha pO Neer rr ORR SOME ed ge A,

PRCSEMeDIVITONMICM bettas is teat and Facade oes ae we Ota Bon Hex yteeas Bh a b)

(SiOOMCEAM sis, Bin 0 Pet ok POCO css SP RARER Aen RE eer aire Ser RO Grae 6

ithedbaneebaamvce successiOn.. 7.0 4e ee soe sabe Soe: 6

Tertiary sea-level movements on the west coast of South Africa ...... 8

Geological history and depositional environments of the Langebaan-

WC PEST CCC SCO Meee rie Bah Mics te Fatt ie aa aS Me oa el ets Bi le asl 18

ithe pne-WMiocene sca-levelichangess).-s4ean sor) sane eee as eee 18

iihelcanly: tomuddie Miocene transeression’ 22.57 -54 2257.62 42 19

ihemniddletovlateiNTOcene neeressiOniey ea: ae See Cee eae M1

itheveanly BhoOceme transenessiOMi sn jcu5-- atyaek yn ee ass 24

The late Pliocene regression and early Pleistocene transgression . . 39

|BHOMOPAT . o, os8 eek Ca aR cae eRe Se ARS a Oe cae eer ee, Re 42

ISIE DS 8 eg oo ook Bie Sk ets ea eee cee en Se 42

| EHEITUUOYE) i cccotene SS 3 chokes thane Olt Roe NOt AEE te ROT onan La ee ene gre eee oe ee 43

COMlPOsiWomere eres 4 ee ee RO en ee ree ne Canna eerie ss 43

glia ltonomyancrs® sesh: 2 Ae eaer Se et ae ee Cok ed ot) 56

Ann. S. Afr. Mus. 84(1), 1981: 1-104, 17 figs, 9 tables.

p} ANNALS OF THE SOUTH AFRICAN MUSEUM

PAGE

Palacoenvironment: s'ssee-csyl. i See oe tees eee Oe es ae ee ere 65

Evolutionany aspects ol the mamimalianitaunian eee oe ee ee eee 79

Dating of the Laneebaanwes succession® sige a eo ee ee ee ee 91

Summary and conclusions si. ..02 2 ew fs secon ae ere Rois oe eee ey eee 96

Acknowledeementstss. Joana Gide. soaee sa ciots catioee agate ogee es en eae 99

FRETEROM CEST oA nn 8 See heen eet Sr hPa OD eC HLT ce 99

INTRODUCTION

General

Langebaanweg (18° 9’E, 32° 58'S) is situated in the south-western Cape

Province approximately 110 km north-north-west of Cape town, and a little less

than 2° north of the most southerly tip of Africa (Cape Agulhas) (Figs 1-2). It

is about 13 km inland (north-east) from Saldanha Bay, and 21 km south of the

mouth of the (Great) Berg River at St Helena Bay. The settlement originally

consisted of little more than a railway siding, but during the Second World War

the commencement of phosphate mining operations and the construction of a

military airfield enlivened the area.

The phosphate deposits of the area have long been known, and although

commercial exploitation commenced in 1943, it was not until 1958 that the

occurrence of vertebrate fossils in these deposits was reported (Singer &

Hooijer 1958). Since then the area has become one of the most prolific sources

of late Tertiary fossils in Africa. Phosphate has been mined in two areas, and

both have yielded fossils. Initially mining was confined to an area immediately

west of the Langebaanweg railway station. This open-cast mine, known as

Baard’s Quarry, was relatively small, and although large quantities of fossils

were recovered, most are unidentifiable fragments (Hendey 1978a). About 2

km further west is Varswater, a subdivision of the farm Langeberg, where

mining also started on a small scale, with few fossils having been recovered.

During 1965 a start was made on the New Varswater Mine (‘E’ Quarry), a

comparatively large undertaking which has been the source of the vast majority

of fossils from the Langebaanweg area (see Hendey 1970a, 1974a, 1976a, and

other publications cited below).

This report deals largely with the fossils and deposits of that part of the ‘E’

Quarry sequence which comprises the late Miocene/early Pliocene Varswater

Formation. In order that this element in the succession be placed in perspec-

tive, an account is also given of other late Tertiary deposits in the vicinity. The

entire succession is correlated with others on the west coast of southern Africa,

and to the record of global sea-level changes.

The emphasis in the palaeontological sections is on the palaeoecology of

the Varswater Formation. ‘Palaeoecology’ was defined by Olson (1962: 134) as

referring “precisely to the ecology at some ancient time or times and thus,

strictly, is an interpretation of the once living biological-physical system’. The

primary requirement in any such study is the identification of the plants and

animals which lived at the time in question. To this must be added a determina-

PALAEOECOLOGY OF LANGEBAANWEG 3

tion of the prevailing environment, that is, the physical setting and the climate.

Only then can the relationships of the plant and animal communities to each

other, and to their surroundings, be assessed.

Although comparatively little is known of the vegetation in Varswater

Formation times, the geology and fauna of the deposits have already been

Ltideritz

Bogenfels

Kleinzee 2) SPE Mga

30°

Hondeklip Bay

ATLANTIC QO, .

“% ¢Vanrhynsdorp

OCEAN val =

2 32°

(=)

e &

\Lange- \%

baanweg =

19AIY

Cape Town “ 34°

False Bay_ Ne

0 100 200 km Gansbaai

Cape Agulhas

Fig. 1. The west coast of southern Africa.

4 ANNALS OF THE SOUTH AFRICAN MUSEUM

studied in some detail. Previous publications resulting froin these studies have

included palaeoecological interpretations, and these are either simply referred

to, or the information they contain is repeated here, usually in an abbreviated

form. In addition, some previously unpublished observations are included. This

synthesis is intended only as an interim report, since relevant research is

continuing.

St Helena Bay

VREDENBURG

\WER COURSE

a Pe =

o ING THE

ee cloce,

Ee’ QUARRYX&@ %q,

SALDANHA 4y, "0's quarry

33° © Sho

Ss \

NOES?

2 4 Karnberg

QO °g

Massenberg

<n, oO *ELANDSFONTEIN

°6.

(=) %

(-)

& Gr. Swartberg

ATLANTIC

OCEAN

4 Slangkop

0 10 20km

18°

Fig. 2. The Saldanha region of the south-western Cape Province.

18°30

PALAEOECOLOGY OF LANGEBAANWEG >)

While the ‘E’ Quarry fossils and deposits have already provided a wealth

of palaeoecological information, a great deal more is promised by further

identification of taxa and detailed analyses of the nature and composition of

fossil assemblages from particular horizons or areas. The combination of an

unusually rich fauna, some palaeobotanical information, and a complex variety

of depositional environments, dating from a period when significant environ-

mental change was taking place, makes ‘E’ Quarry an excellent subject for

palaeoecological studies.

Stemming from the accounts of the geology and biology of the ‘E’ Quarry

deposits is a statement on some evolutionary aspects of the recorded fossil

mammals, together with a reassessment of the age of the principal fossil

occurrences in the area.

Present environment

The south-western Cape Province has a Mediterranean type of climate (Cs

of Koppen), and Langebaanweg 1s situated in the more arid north-westerly part

of this region. Rainfall is brought by the westerly wind system of southern

mid-latitudes, which is usually far enough north to affect the area only during

winter. Coastal fogs sometimes penetrate to the area during cooler months.

The mean annual precipitation is about 250 mm. Winters tend to be cool rather

than cold, and very rarely are sub-zero temperatures experienced, and then

only at night. The summers are dominated by high-pressure systems in the

south Atlantic and are long and dry. Although temperatures may be high (up to

40 °C), they are often moderated by the prevailing south-easterly winds. These

winds are frequently moderate to strong, and, being dry, they aggravate the

summer aridity by desiccating both the soil and plants. The low temperatures of

the adjacent South Atlantic Ocean are caused by an upwelling of Southern

Ocean water in the Benguela Current System. This cold water is a major factor

contributing to the aridity of the region. A summary account of the climate of

the south-western Cape and adjacent regions was given by Fuggle & Ashton

(STS).

In the immediate vicinity of Langebaanweg there is no naturally occurring

surface fresh water, except for small ephemeral ponds during winter. Otherwise

surface fresh water is available locally only at isolated springs, such as that

associated with the Quaternary fossil occurrences on the farm Elandsfontein, 20

km to the south-east (Hendey 1974a), or in the Berg River, 15 km to the

north-east. The Berg River is perennial and is the largest river in the south-

western Cape, draining the major part of this region (Harrison & Elsworth

1958; Harrison 1964).

The sclerophyllous natural vegetation, which has been considerably dis-

turbed by human activity, is classified as ‘coastal fynbos’, which in this sandy

area 1S ‘predominantly ericoid and rather open, the bushes being somewhat

rounded and up to 1 m high’ (Taylor 1978: 204). Proteoids, restioids, grasses,

geophytes, and annuals are present, but there are no indigenous trees. A

6 ANNALS OF THE SOUTH AFRICAN MUSEUM

detailed account of the vegetation in the near-by area surrounding Langebaan

was given by Boucher & Jarman (1977).

The fauna of the area has been depleted during historic times (post-

seventeenth century). No larger mammals survive, although smaller ones are

not uncommon. They include a variety of imsectivores, bats and rodents

(including a porcupine), a hare, small carnivores (genets, mongooses, a wildcat,

foxes), and small antelope (steenbok, grysbok and probably grey duiker). Birds

are both diverse and abundant, nearly 100 species having been recorded in or

near ‘E’ Quarry (G. Benfield, pers. comm.). This is due largely to the

artificially well-watered environment created by the open-cast mining, which

has extended below the water-table in places. Lower vertebrates include frogs,

tortoises, lizards, and snakes. Invertebrates are common and include the land

snail, Trigonephrus, shells of which are frequently found in local terrestrial

deposits of late Cenozoic age.

GEOLOGY

THE LANGEBAANWEG SUCCESSION

The Cenozoic deposits in the vicinity of Langebaanweg have been the

subject of many published and unpublished studies, partly because of the

economically important phosphate occurrences, and partly because they are in

places so richly fossiliferous. At its maximum development the local succession

comprises over 100 m of clastic sediments resting on a pre-Mesozoic bedrock.

These deposits are largely unconsolidated, and the earliest evidently date back

to the early or middle Miocene, followed by a substantial element dating from

the Pliocene, while the succession is completed by a nearly ubiquitous covering

of Quaternary sands.

The most recent account of this succession, and the only one that deals

with it in its entirety, is that of Rogers (1980), whose lithostratigraphy (Table 1)

is basically similar to that recognized in earlier studies (e.g. Hendey 1973,

1974a, 1976a; 1980; Vankard 19744, 1975a, 19756; Dingle ci aly 1979) hie

earlier studies centred largely on the phosphate and fossiliferous deposits of the

Varswater Formation (sensu Tankard 1974a), which are best known from the

exposures in “E’” Quarry.

The *E’ Quarry succession was initially interpreted largely on palaeontolo-

gical grounds (Hendey 1973, 1974a), and this was followed by detailed studies

of the deposits themselves. Tankard’s (1974a) definition, description and dis-

cussion of the Varswater Formation in ‘E’ Quarry and adjacent areas is the

standard reference on the subject. In a subsequent study, Tankard (1975a)

dealt with those deposits underlying the exposures in ‘E’ Quarry and included

them in the ‘Saldanha Formation’. He gave a useful review of all the informa-

tion then available in an unpublished thesis (Tankard 1975b).

The ‘E’ Quarry succession has since been discussed by Hendey (1976a,

1980), Dingle et al. (1979), and Rogers (1980), while the latter two studies also

take into account underlying and overlying deposits.

PALAEOECOLOGY OF LANGEBAANWEG 7

TABLE |

The lithostratigraphy of Cenozoic deposits in the vicinity of Langebaanweg, Cape Province.

Rogers 1980 This Report

Formations Subdivisions Subdivisions Age

Bredasdorp Langebaan Limestone Surface sands Pleistocene and

Member (in part) (including duricrusts) Holocene

Baard’s Quarry

fluviatile deposits

Late Pliocene or

early Pleistocene

Anyskop Late Pliocene

terrestrial deposits

Bredasdorp . Langebaan Limestone

Member (in part)

Anyskop

marine deposits

‘phosphorite member’ Pelletal Phosphorite

Member (PPM)

(including beds 3aS and | (including beds 3aS and

3aN) 3aN)

Varswater Early Pliocene

‘quartzose sand Quartzose Sand

member’ Member (QSM)

(including ‘peat bed’ & | (including floodplain, peat

‘clay bed’) & tidal flat beds)

‘gravel member’ Gravel Member (GM) Late Miocene

Saldanha

— Pre-GM deposits Middle Miocene

Elandsfontyn

Changes in the nomenclature of the subdivisions of the Varswater Forma-

tion have been frequent enough to cause confusion, which is unfortunate since

the lithostratigraphy recognized in the various studies has remained virtually

unchanged (Rogers 1980, table 3.8). The most distinct of the recent termino-

logies is that of Dingle et al. (1979), and, although it is concise and comprehen-

sive, the largely informal terminology used in all post-1975 palaeontological

studies is retained here, with the addition of some completely informal terms

which refer to pre- and post-Varswater Formation (sensu stricto) deposits.

The latter deposits have hitherto received relatively little attention, and are

clearly in need of more detailed study. Only those which have a bearing on the

present interpretation of the area’s geological history are discussed below.

Since Langebaanweg is located close to the coast, the local succession has

been considerably influenced by past changes in sea-level. Although this

succession itself provides evidence of the nature and timing of such changes,

these can best be interpreted on the basis of regional and global evidence.

Consequently, before proceeding to a more detailed account of deposits in the

Langebaanweg area, the record of relevant sea-level movements is examined.

8 ANNALS OF THE SOUTH AFRICAN MUSEUM

TERTIARY SEA-LEVEL MOVEMENTS ON THE WEST COAST OF SOUTH AFRICA

The nature of the Langebaanweg succession is such that it provides some

unique data relevant to the interpretation of southern African sea-level

movements. However, since the local record is both complemented and sup-

plemented by evidence from further afield, account will be taken here of such

evidence from the approximately 600 km west coast of South Africa. This

stretch of coast can conveniently be divided into two regions. They are the

south-western Cape coast between False Bay and the Olifants River, and the

Namaqualand coast between the Olifants and Orange Rivers (Fig. 1). A coastal

plain of varying width fronts on a more or less continuous series of mountains,

which separates this plain from an inland plateau.

Shorelines of Cenozoic age are preserved along most of this coast, and

much attention has been focused on them, mainly because of associated

occurrences of phosphate and diamonds. Although particular reference 1s made

to the shoreline successions in the Langebaanweg area of the south-western

Cape and the Hondeklip Bay—Kleinzee section of the Namaqualand coast,

successions in other areas may be as well known and as relevant.

There are altimetric similarities between the recorded shorelines of the two

regions, and correlations between them, and shorelines elsewhere, have been

suggested (e.g. Hendey 1969; Carrington & Kensley 1969). The inter-regional

correlations have been modified (Tankard 1975b, 1976), while the broader

correlations have generally been discounted (e.g. Wolff et al. 1973). While the

criticism of the latter correlations was justified, more recent studies have

indicated that there is a sound basis for correlating the west coast shoreline

succession with global phenomena of the Cenozoic. Consequently, broader

correlations are feasible. For example, Tankard (19755, 1976) has correlated

the lower and more recent of the west coast shorelines with high sea-levels of

the last interglacial, and it follows that they are the local equivalents of last

interglacial shorelines recorded elsewhere in the world.

This aspect of Tankard’s studies is irrelevant here, since it is only the older

and higher of the late Cenozoic shorelines which are represented in the

immediate vicinity of Langebaanweg. According to Tankard (1975b, 1976) the

older late Cenozoic shorelines in the south-western Cape are at lower eleva-

tions than their counterparts in Namaqualand, this difference being ascribed to

differential epeirogenesis along the west coast. However, the difference is here

regarded as more apparent than real, and there is, in fact, a remarkable

altimetric similarity between all the late Cenozoic shorelines of the two regions.

The Namaqualand succession is both well developed and well documented

over a distance of about 400 km, and there has been little disagreement about

the elevations of the shorelines constituting this succession. The difference of

opinion over the inter-regional correlation centres on the elevation of the

higher shorelines in the south-western Cape, and it stems from the nature of

the record of these shorelines. The record in the two regions differs simply

because of distinctive characteristics in their physical geography.

PALAEOECOLOGY OF LANGEBAANWEG )

The Namaqualand coast has a more or less regular north-north-west to

south-south-east trend, and, in those areas where ancient shorelines are best

developed, it has a bedrock profile which rises at a generally moderate angle

from the present shoreline to beyond the limit of the highest late Cenozoic

shoreline (c. 100 m). The situation is more complex at the mouths of rivers,

where marine and river terraces merge, and where fluviatile and estuarine

deposits are represented. In general though, the ancient shorelines have a fairly

regular development, with associated deposits forming a readily identifiable

sequence between 0 and 100 m elevation. The shorelines themselves are

generally incised on bedrock. This situation can be largely ascribed to the

configuration of the bedrock, and the fact that it is uniformly comprised of

‘Namaqualand gneiss’ (see Haughton 1968).

By contrast, the coastal area of the south-western Cape is much less regular

due to the more complex solid geology (i.e. Basement Complex granites,

Malmesbury System metamorphics and intrusives, and Cape Supergroup

sedimentary rocks). The more resistant rocks, and those of the relatively young

and uplifted Cape Supergroup, form areas of high relief. The bedrock may thus

rise steeply from present sea-level to the hills and mountains of the region,

although there are also wide areas where bedrock is below present sea-level and

where it 1s overlain by largely unconsolidated late Cenozoic sediments.

The areas of high relief are particularly significant since at times of the

higher late Cenozoic sea-levels (c. 50 m and above) those in the east would

have formed an irregular and often steep-sided coast, whereas those in the west

would have formed two island complexes off the southern and central parts of

the region. The islands would also have had irregular and steep-sided coasts,

and would have provided some shelter to the adjacent mainland coast from the

open ocean. Mechanical erosion by wave action on this coast might, therefore,

have been reduced.

Shoreline environments in the south-western Cape, botn now and at times

of past high sea-levels, thus differ appreciably from those of the Namaqualand

coast. Consequently, their features and associated deposits are likely to differ

in some respects.

In Namaqualand wave-cut platforms on bedrock, on which gravels and

other sediments accumulate, often had their onshore limits marked by notches

and low cliffs. In other words, they exhibit many of the classic manifestations of

‘raised beaches’, and are usually easily identified and differentiated.

The situation in the south-western Cape is by no means as simple. In those

areas where bedrock is below present sea-level, wave-cut platforms were

developed on unconsolidated deposits. They were, therefore, insubstantial and

very susceptible to subsequent erosion. In addition, deposits on the platforms

were later to intermingle with essentially similar unconsolidated deposits accu-

mulated under very different circumstances (e.g. subaerially, by aeolian

action). Furthermore, gravels would not normally have been deposited on thse

platforms because of the absence of a source-rock, while notches and cliffs

10 ANNALS OF THE SOUTH AFRICAN MUSEUM

would also have been absent, with impermanent coastal dunes having been

present instead. As a result, evidence of past high sea-levels over wide areas in

the south-western Cape may be difficult to detect.

In those areas where bedrock rose steeply into hills and mountains, both

on the mainland and on the islands, older and higher notches, wave-cut

platforms and superimposed deposits would tend to be undercut and eroded

away during subsequent periods when sea-level was at lower elevations. In

addition, shoreline features developed on the mainland coast in the lee of the

two island complexes might have been relatively insubstantial because of

lower-energy waves. Consequently, some of the more obvious manifestations of

‘raised beaches’ may also be obscured in these areas of the south-western Cape.

It is, therefore, not surprising that surveys have found evidence of the older

and higher late Cenozoic shorelines in this region either absent or equivocal.

For example, Davies (1973: 722) believed that there was no direct proof of

high shorelines in that part of the south-western Cape which includes the

Langebaanweg area (Fig. 2), claiming that ‘the existence of pleistocene sea-

levels at 90 m and 60 m in [this area seems to be] inferred from these levels in

Namaqualand’. The existence of a c. 90 m marine platform in the vicinity of

Elandsfontein south-east of Langebaanweg was suggested by Mabbutt (1956), a

view subsequently substantiated by Rogers (1980), although he records marine

deposits up to an elevation of only about 80 m. This platform is on unconsoli-

dated sediments and is an example of one which lacks some of the classic

features of such platforms. The existence of a 50-55 m high sea-level in the

vicinity of Langebaanweg has been demonstrated by Tankard (1974a—see

p. 40), while Mabbutt (1956: 50) recorded a marine terrace at ‘45-60 m’ ‘inland

from Saldanha Bay’. This one also lacks the features of the kind that Davies

and others have sought.

Davies (1972: 270) also found ‘few indications of quaternary shorelines

above 30 m’ in the southern part of the south-western Cape (i.e. around False

Bay and along the southern coast to Gansbaai). However, he does record some

evidence of high sea-levels in this area at 45-60 m and 75-94 m. The fact that

these records are insubstantial is probably due to the generally precipitous

nature of the coast in this area and the destruction of higher shoreline features

by subsequent erosion. In addition, the eastern shore of False Bay was

sheltered by islands that now constitute the Cape Peninsula.

The fact that there is some evidence of these higher shorelines in the

south-western Cape is a certain indication that they exist. The evidence for

their presence cannot be dismissed because the shorelines are not obvious over

wide areas, or because they lack some of the characteristics of shorelines

elsewhere. Uncertainty about the actual elevation of the higher shorelines, or

discrepancies between recorded elevations in the south-western Cape and

Namaqualand do not nullify the correlation between the two regions. They are

simply explained by the fact that the south-western Cape record is poorer than

that in Namaqualand. In the latter region the stillstands in marine transgres-

PALAEOECOLOGY OF LANGEBAANWEG itt

sions and regressions are usually determinable, whereas over wide areas of the

south-western Cape this is not necessarily the case. In spite of the difference in

the nature of the records, elevations for the higher shorelines in the two regions

are in remarkably close agreement.

Other Namaqualand shorelines which are supposedly absent or poorly

developed in the south-western Cape are those which Carrington & Kensley

(1969) recorded at ‘29-34 m’ and ‘17-21 m’. There is, in fact, unequivocal

evidence for a shoreline at c. 30 m in ‘E” Quarry at Langebaanweg, while there

is an indication of a much later high sea-level at about 20 m in the Langebaan-

weg area (see below), which was probably the one responsible for the 15 m

marine terrace recorded in the same area by Mabbutt (1956).

It is worth noting in this connection that Tankard (19755, 1976) has

suggested that shorelines recorded at 10 m and 13 m in the south-western Cape

may be the equivalent of the 17-21 m shoreline in Namaqualand. He ascribed

the lower elevation of the south-western Cape beaches to differential epeiro-

genesis on the west coast. Since this factor is dismissed here, the discrepancy

must have another explanation. It is possible that the 10 and 13 m records in

the south-western Cape represent vestiges of a shoreline at a higher elevation

(? c. 20 m), since features associated with shorelines may extend over appreci-

able vertical and horizontal distances. Ideally only the maximum elevation of a

high sea-level should be recorded (i.e. the shoreline itself), but this is not

always possible.

It may be significant that, whereas the 17-21 m shoreline is present and

well developed at Hondeklip Bay, it is apparently either not represented at

Kleinzee, 80 km further north (Hallam 1964), or its deposits are mixed with,

and not distinguishable from, those of the older 30 m shoreline. This may be an

indication that the stillstand during which the 17-21 m shoreline was developed

was not prolonged, and that, consequently, it was only because of some

exceptional circumstance of local geography that sporadic records of it have

survived. This would be consistent with the nature of the record of the c. 20 m

shoreline in the Langebaanweg area (see below).

The situation on the west coast of South Africa may be summed up by the

statement that there is evidence for four relatively high late Cenozoic shore-

lines, which are for the sake of convenience referred to by their approximate

elevations of 90 m, 50 m, 30 m and 20 m. All represent ‘transgressive’ or

‘regressive complexes’ and deposits associated with them may therefore be

encountered at elevations above or below those given above.

It has generally been assumed that there is a direct correlation between the

age and the elevation of these shorelines, the oldest being the highest, and with

elevations decreasing with age. It is evident from the Langebaanweg record

that this is not the case. In this area the 30 m shoreline is the oldest, followed in

descending order of age by the 90 m, 50 m and 20 m shorelines. The

Langebaanweg succession is also important in providing evidence of the actual,

or likely, age of these shorelines.

12 ANNALS OF THE SOUTH AFRICAN MUSEUM

Up until a decade ago it was common practice to suggest correlations between

South African shoreline sequences with others elsewhere, notably the classic ones

of the Mediterranean Basin, which in turn were correlated with the glacial—

interglacial chronology of the Quaternary. Since then the nature and timing of

sea-level changes during the late Cenozoic have been much more thoroughly

documented, while the complexities imposed by tectonic instability have become

more widely appreciated, and the earlier correlations are no longer accepted.

A positive advance in long-range correlation of west coast shorelines came

with the studies of Tankard (19755, 1976), who suggested that most, if not all, of

the lower shorelines in the south-western Cape date back to the last interglacial.

The validity of this conclusion will not be examined here as none of the shorelines

is represented in the immediate vicinity of Langebaanweg. At issue though is the

age of the older and higher beaches, which have for the most part continued to be

regarded as being of Pleistocene age as well. Vertebrate fossils associated with the

higher shorelines in the Langebaanweg area indicate beyond all doubt that this

succession Is largely, or entirely, late Tertiary in age. In addition, the nature of the

Langebaanweg fossil assemblages allows for an interpretation of depositional

environments which is in good accord with the late Tertiary sea-level movements

around southern Africa recorded by Siesser & Dingle (1981). These are in turn in

close agreement with the global sea-level changes which were determined by Vail

et al. (1977). Siesser & Dingle (1981: 83) were concerned only with the ‘gross

movements of the seas around southern Africa during the Tertiary’, and details of

these movements during the late Tertiary determined on the basis of the

Langebaanweg record match those of the global changes (Vail et al. 1977; Vail &

Hardenbol 1979) remarkably well (Fig. 3).

Siesser & Dingle (1981: 83) summarized the sea-level movements relevant

to the Langebaanweg succession as follows: ‘The major Neogene transgression

began in the middle Miocene and probably reached its greatest extent in the

late Miocene or early Pliocene. The overall middle Miocene to early Pliocene

transgression was interrupted by a brief regressive pulse near the Miocene—

Pliocene boundary. Seas withdrew again in the late Pliocene.’

The interpretation of global sea-level changes of Vail & Hardenbol (1979)

differs slightly in that they have the major Neogene transgression commencing

during the early Miocene, reaching a climax early in the middle Miocene, and

followed by a middle to late Miocene regression, which took the form of three

major drops in sea-level, separated by periods when sea-level was more or less

static. The last of these major drops evidently coincides with the ‘brief regres-

sive pulse near the Miocene—Pliocene boundary’ recorded by Siesser & Dingle

(9S t-7183):

The sea-level movements which affected the Langebaanweg area are

referred to in the discussions which follow as:

|. ‘the early to middle Miocene transgression’,

2. ‘the middle to late Miocene regression’, with ‘the terminal Miocene

regression’ climaxing this event,

PALAEOECOLOGY OF LANGEBAANWEG

‘syisodap a]nerany Arend s.pieeg °/

‘syisodap yeigso1i9} doyskuy “9 ‘Samureqosuery jo jseo-yinos wuoyeid surrey “¢ ‘syisodap ourewu doyshtuy “‘p ‘Ndd ‘NSO ‘€ ‘WD ‘7

‘ssodap JIO-ud ‘[ ‘UoIssaoons SamuLegasuey] sy) Aq poayeoIpUl asoy] puK ‘sasuRYS [9Ad]-BIS UPOL}Y UTOYINOS pu [eqo]s Jo splosdy “¢€ ‘SIF

JIGGIW

INIIOIN

14A31-W4S INITLSVOI ~~

INIS3ud INIS4ud

—<— 9NI11vI QNISiy ——e (w) NOILWAaT4

(GL6L loquapieH B [len 1893;e) (1861 a1Gurq B 4aSsaig 12248)

dVINO I1ViSVOI NO G4SVa8 1]3A41-V4S NOISSZIINS JIMNVVEIINVI JHL SJONVWHI 13A341-V4S

40 SFIONVHI JAAILV19H JO S4I1DAN 1V98019 AG G4ILVIIGNI SAONVHD 13A51-W4S NVIIHIVY NYAHLNOS

= S

> >

e [=

(=<) m

ES) —

S4I1IAINIdNS

S419A9

14 ANNALS OF THE SOUTH AFRICAN MUSEUM

3. ‘the early Pliocene transgression ,

4. ‘the late Pliocene regression’ and ‘the early Pleistocene transgression’.

According to the present interpretation of the Langebaanweg succession,

deposits beneath the floor of “E’ Quarry (‘pre-GM’) are associated with the

early to middle Miocene transgression; the 30 m shoreline (GM) in ‘E” Quarry

dates from the middle to late Miocene regression; the 90 m shoreline represents

the climax of the early Pliocene transgression, with a large proportion of the

deposits in the immediate vicinity of “E’ Quarry having been laid down during

this transgression; while the 50 m shoreline represents a stillstand in the late

Pliocene regression. The 20 m shoreline is problematical, having been deve-

loped either during a second stillstand in the late Pliocene regression, or at the

climax of the subsequent early Pleistocene transgression. The reasons for this

interpretation will become evident from the discussions which follow in this,

and later, sections of this report.

On the basis of the altimetric correlation of south-western Cape and

Namaqualand shorelines, it follows that the 29-34 m shoreline in Namaqualand

must be of late Miocene age. Carrington & Kensley (1969) believed this

shoreline dated from the middle Pleistocene, but there 1s no firm foundation for

this suggestion. The 29-34 m shoreline is represented at Hondeklip Bay, where

it is ‘thin’ and ‘discontinuous’ (Carrington & Kensley 1969: 190), and at

Kleinzee (Hallam 1964), where it is also poorly developed (personal observa-

tion). This poor development sets it apart from other elements in the Nama-

qualand, succession, which suggests that it might have been subjected to

erosion during a subsequent period when sea-level was at a higher elevation. In

other words, it is likely to pre-date the 90 m and 50 m high sea-levels, rather

than post-date them as suggested by Carrington & Kensley (1969).

Again on altimetric evidence, the early Pliocene 90 m, late Pliocene 50 m,

and the late Pliocene/early Pleistocene 20 m shorelines in the south-western

Cape are correlated with the *75—90 m’, ‘45-50 m’ and ‘17-21 m° ‘transgressive

complexes’ in Namaqualand, which Carrington & Kensley (1969: 190-191)

dated as ‘Basal’, ‘Lower’ and ‘Middle Pleistocene’ respectively.

According to the above interpretation of west coast shorelines, there are

none of middle Pleistocene age. There undoubtedly were glacio-eustatic sea-

level fluctuations during this period, and it is likely that during the interglacial

phases sea-level was higher than at present, just as it was during the last

interglacial, and apparently also during the mid-Holocene. However, Shackle-

ton & Opdyke (1973) have suggested that middle Pleistocene interglacial

sea-levels would have been no higher than the highest of the last interglacial

levels. Consequently, in tectonically stable area all traces of their presence

might have been obscured or destroyed during the last interglacial.

Coastal deposits about which there is some uncertainty are those in the

Hondeklip Bay area which in mining terminology are referred to as ‘E stage’

(Tankard 1975b). Although they, too, relate to periods of higher sea-level, they

have been regarded as distinct from the deposits of the ‘transgressive com-

PALAEOECOLOGY OF LANGEBAANWEG iL)

plexes’, and both Carrington & Kensley (1969) and Tankard (19755) dated

them as late Tertiary. Since the higher of the ‘transgressive complexes’ are here

also regarded as late Tertiary, it is necessary to reconsider the age and

relationships of the ‘E stage’ deposits.

The ‘lower E stage’ deposits were correlated by Tankard (19755) with the

middle Miocene phosphatic rock horizon in ‘E’ Quarry at Langebaanweg. This

correlation is accepted here and the ‘lower E stage’ is regarded as remnant

evidence for the early to middle Miocene transgression on the Namaqualand

coast.

The ‘middle E stage’ deposits were correlated by Tankard (19755) with the

QSM of the Varswater Formation, and this is also accepted here, although the

correlation is extended to include the PPM of the Varswater Formation. The

QSM and PPM date from the same episode (i.e. the early Pliocene transgres-

sion), and there are no grounds for believing that the ‘middle E stage’ is

contemporary with the QSM rather than the PPM.

The age of the ‘middle E stage’ was partly inferred from remains of two

vertebrate taxa which were believed to be derived from this ‘stage’ and

redeposited in younger sediments. One of these taxa, Ceratotherium praecox, is

represented by a single rolled tooth (Hooijer 1972), which almost certainly is a

derived specimen (A. J. Carrington, pers. comm.). The second, ‘Prionodelphis’

(= Homiphoca) capensis, is represented by an isolated canine and a metatarsal.

They are unrolled, but Carrington believed them to be derived since they are

phosphatized specimens from non-phosphatic deposits. This is not a convincing

point of view since phosphatized fossils in situ in non-phosphatic deposits have

been recorded from ‘E’ Quarry. A new development concerning the Namaqua-

land ‘Homiphoca capensis’ is that the metatarsal has been found to be appreci-

ably longer than any of the now numerous specimens known from ‘E’ Quarry.

This suggests that the Namaqualand specimen represents a species which was

more advanced than the ‘E’ Quarry H. capensis. Consequently, it may well be

contemporaneous with the post-‘middle E stage’ deposits from which it and the

isolated canine were recovered.

Other deposits on the Namaqualand coast which were a likely equivalent

of the ‘middle E stage’ were those at Kleinzee from which was recovered the

small vertebrate fauna described by Stromer (193la, 1931b). This fauna is

evidently broadly contemporaneous with those from the QSM and PPM at

Langebaanweg (Hendey 1974a, 1978c, 1978d).

The ‘upper E stage’ deposits in the Hondeklip Bay area contain pelletal

phosphorite, which led Tankard (19755) to correlate it with the PPM of the

Varswater Formation. While this interpretation cannot be discounted, it is not

favoured here. According to one interpretation, the ‘E’ Quarry pelletal phos-

phorite is largely derived from a ‘mechanical erosion’ of the ‘Miocene bedded

authigenic phosphorite’ (Tankard 1975c: 375), and there is no reason to believe

that a similar deposit on the Namaqualand coast is necessarily contempora-

neous. While it, too, might have been derived from a Miocene phosphorite, the

16 ANNALS OF THE SOUTH AFRICAN MUSEUM

erosion process in Namaqualand could well have pre- or post-dated deposition

of the Varswater Formation PPM. The same applies if the ‘upper E stage’ and

‘E’ Quarry pelletal phosphorites are authigenic.

Tankard (19755) records that the ‘upper E stage’ sediments contain a rich

fossil fauna, the molluscs of which indicate warm-water conditions, and that the

sediments are recorded only up to an elevation of 30 m. Perhaps significantly,

the GM of the Varswater Formation is also comprised of a mechanically eroded

Miocene phosphorite, although in the form of a gravel rather than pellet-sized

particles, it is also associated with a warm-water mollusc fauna (see p. 68), and

it also occurs up to an elevation of 30 m. By contrast, the PPM contains an

element which suggests cold-water conditions, and is recorded at elevations

well above 30 m.

The ‘essentially bedrock-depression infilling’ ‘upper E stage’ (Tankard

1975b: 277) may, in fact, represent a part of the ‘thin, discontinuous’ ‘29-34 m

beach’ of Carrington & Kensley (1960: 190), and is recorded as such in Table 2.

The temperature of the adjacent ocean at the time that the various

elements in the west coast succession were laid down is of interest. The revised

interpretation of the Namaqualand succession means that all those deposits

from which warm-water mollusc faunas are recorded date back to the late

Tertiary, while those with cold-water faunas are Quaternary.

There are, however, apparent exceptions to this general rule. The 75-90 m

shoreline deposits in Namaqualand are unfossiliferous (Tankard 1975b), and

consequently provide no direct evidence of water temperatures prevailing at

that time. However, deposits laid down in ‘E’ Quarry during an early stage in

the 90 m transgression (i.e. the QSM and PPM) contain fossils that indicate

cold-water conditions (see p. 68). There is thus likely to have been a similar

cold-water fauna on the Namaqualand coast at that time. This period of cold is

correlated with the aftermath of the Antarctic glacial maximum during the

terminal Miocene (see p. 70). Since the 90 m early Pliocene transgression

post-dates this glacial maximum, it is possible that cold conditions prevailed

only during the early stages of the transgression, and that by the time it reached

its peak sea temperatures were higher.

Tankard (1975b) recorded that the 17-21 m shoreline in Namaqualand is

associated with a cold-water fauna. If this shoreline is late Pliocene in age, then

it would be an exception to the ‘late Tertiary/warm water’ rule. On the other

hand, this may be interpreted as evidence that the 17-21 m shoreline dates

from the early Pleistocene, which would be in keeping with evidence from the

Langebaanweg area (see p. 41).

Although the present study was confined to the shorelines on the west

coast of South Africa, it is obvious that a similar reinterpretation of coastal

sucessions elsewhere in southern Africa is possible. For example, according to

the present interpretation and contrary to earlier opinion, the ‘Upper Terraces

Group’ on the southern coast of the Namib Desert, with their warm-water

faunas (Hallam 1964), are likely to be late Tertiary in age.

PALAEOECOLOGY OF LANGEBAANWEG Ly

TABLE 2

Records of late Tertiary and early Pleistocene sea-level changes on the west coast of southern

Africa.

oe) —~

_ |

72) n ww =

m | & @ | 2 y| SOUTH-WESTERN| HONDEKLIP KLEINZEE SOUTHERN

2196 7 “ CAPE BAY NAMIB DESERT

Za as (2) (3) (4) (4), (S)

ae)

20 m

(Baard’s Quarry

fluviatile)

17-21 m ? in part

Pleistocene

= eee cane

a 50 m 40-64 m 15 m

(Anyskop (E Beach)

terrestrial)

@ 90 m 75-90 m 82-88 m 20-25 m

S| . |= |CE’ QSM, PPM;| (? including (F Beach)

= = Anyskop ‘middle E

a marine) stage)

= aha eee ee

g2 3

= 0 =

Es x at No deposits No deposits No deposits No deposits

ES 3

2 30 m 29-34 m 21-40 m 2 tanh

3 (‘E’ GM) (? including (D Beach)

ia ‘upper E stage’)

» 2 a Pre-GM ‘lower E stage’ 33 spe ee

37 oO & rock gravels’ an

ae ce on Arrisdrift fossil

= = site)

(1) Vail & Hardenbol (1979). (2) Langebaanweg area (this report); complementary data from

Davies (1972, 1973), Tankard (1974a) and Rogers (1980). (3) Carrington & Kensley (1969)

and Tankard (19755). (4) Hallam (1964) and Corvinus & Hendey (1978). (5) Downwarped

coast — terraces at lower elevations than those on South African coast.

18 ANNALS OF THE SOUTH AFRICAN MUSEUM

The poorly developed ‘D Beach’ (12 m) of the ‘Upper Terraces Group’ is

here tentatively correlated with the late Miocene 30 m shoreline on the South

African west coast, although it could instead be the equivalent of the late

Pliocene 20 m shoreline. The latter alternative is less likely because the 20 m

shoreline is apparently associated with a cold-water fauna (see above). In

addition, the poor development of the ‘D Beach’ may be an indication of

erosion during a subsequent high sea-level. which was suggested in the case of

the 30 m shoreline on the South African west coast.

The ‘F Beach’ (25 m) and ‘E Beach’ (15 m) are here correlated with the

South African west coast 90 m and 50 m shorelines.

The ‘false bedrock gravels’, which extend northwards from the mouth of

the Orange River, evidently pre-date the ‘Upper Terraces’ (Hallam 1964),

and could therefore date back to the middle Miocene. They are the likely

equivalent of the pre-GM deposits at Langebaanweg and the ‘lower E stage’

of the Namaqualand coast. In addition, they are probably the coastal equiva-

lent of the middle Miocene river terrace(s) at Arrisdrift. 30 km inland from

the mouth of the Orange River (Corvinus & Hendey 1978; Hendey 1978b).

It is worth noting in this connection that the bedrock elevation of the chan-

nel which contains the middle Miocene fossils at Arrisdrift is 40 m, while the

river level is 7 m above sea-level. It follows that their coastal equivalent

should be at 33 m, which is, in fact, the elevation of the base of the cliff

against which the ‘false bedrock gravels’ are banked (Hallam 1964). It may

also be significant that while there are four river terraces recognized at

Arrisdrift (Corvinus & Hendey 1978), there are also four marine terraces of

late: Tertiary age on: the coast (les the DD’, = and Fo Beaches andmile

‘false bedrock gravels’).

The Arrisdrift fossil occurrence is of particular significance in the interpre-

tation of the southern African west coast succession, since it provides the most

reliable evidence of the time at which the early to middle Miocene transgres-

sion began to result in deposition on the present coast. The Arrisdrift fossils

date back about 16 Ma (Hendey 19785), that is, to the early part of the middle

Miocene.

The correlation of the west coast shorelines suggested here is summarized

in Table 2.

GEOLOGICAL HISTORY AND DEPOSITIONAL ENVIRONMENTS OF THE LANGEBAANWEG

SUCCESSION.

As indicated earlier, the Langebaanweg succession is most conveniently

interpreted in relation to the southern African sea-level changes recorded by

Siesser & Dingle (1981).

The pre-Miocene sea-level changes

There is no evidence in the Langebaanweg area (Figs 2,4) for any deposits

dating back to the early Tertiary. Two marine transgressions are recorded

PALAEOECOLOGY OF LANGEBAANWEG 19

rerein PRESENT

St Helena Bay Berg River

ae Phosphate

mines

LBW Langebaanweg

station

AK Anyskop

‘© Quarry

E Quarry Baards Quarry

Contours in metres

above present sea-

level

Saldanha Bay

Fig. 4. The Langebaanweg area today.

during this period, one in the late Paleocene/early Eocene and the other in the

late Eocene (Siesser & Dingle 1981).

There are records of Eocene marine deposits at 70 and 163 m on the west

coast in the southern Namib Desert (Bogenfels and Buntfeldschuh—Siesser &

Dingle 1981), as well as possible Eocene deposits at about 140 m at Buffels

Bank (Kamaggas) west of Springbok (South African Museum records), and of

uncertain elevation at Quaggaskop near Vanrhynsdorp (Lamont 1947). If the

latter records are indeed Eocene, then the shorelines of this period must have

abutted the western escarpment in Namaqualand, whereas in the downwarped

southern Namib they approached the present coast near Bogenfels following

the east-west trend of the Klinghardt Mountains in this area (Fig. 1). The

implication is that during the Eocene transgressions much of the south-western

Cape, including the Langebaanweg area, was below sea-level, the coastline

being along higher ground many kilometres east of Langebaanweg.

There was then a major regression spanning the entire Oligocene, and much

or all of the early Miocene, when sea-level reached several hundred metres

below that of the present (Siesser & Dingle 1981). This is likely to have been the

period of major continental erosion during which all traces of early Tertiary

sediments in the Langebaanweg area, and elsewhere along the west coast, were

removed. It might have been during this period that the bedrock in the

Langebaanweg area was eroded down to its present elevation of about —40 m.

The early to middle Miocene transgression

The presence of early Miocene vertebrates in fluviatile and lacustrine

deposits near the coast of South West Africa between Bogenfels and Lideritz

20 ANNALS OF THE SOUTH AFRICAN MUSEUM

(Stromer 1926; Hopwood 1928; Hamilton & Van Couvering 1977; Hendey

1978b) is a possible indication that at least this section of the west coast was

then being influenced by the early to middle Miocene transgression. The

southern Namib region was certainly being influenced by this transgression

early in the middle Miocene, when the fossil assemblage at Arrisdrift was

deposited on a river terrace (Corvinus & Hendey 1978: Hendey 19785).

Apparently contemporaneously the ‘false bedrock gravels’ of the southern

Namib coast were being laid down, and these are in turn correlated with the

‘lower E stage’ and related deposits on the Namaqualand coast. and the

pre-GM deposits in the Langebaanweg area.

The pre-GM deposits are known only from boreholes, although small

exposures of their uppermost levels formerly exised in ‘C’ and *E’ Quarries.

About 70 m of these deposits are now known, ranging in elevation from about

—40 m to +30 m (Rogers 1980), although until recently only the upper 15-20

m had been placed on record (Tankard 19755). The sequence is apparently

comprised largely of deposits laid down under terrestrial environments, but at

least in the case of the upper levels studied by Tankard (19756: 34), ‘it is likely

that these deposits formed close to sea level’ during ‘a slow transgression .

Tankard found evidence in the sequence of a ‘marine incursion’, which was ‘a

minor event of very short duration. . . readily explained by breaching of a

barrier’. A second marine incursion ‘was accompanied by phosphate minerali-

zation to form a phosphatic sandstone’ (Tankard 19755: 34).

Peats in the succession have yielded abundant pollens which indicate a

forested environment, including subtropical elements such as palms (J. A.

Coetzee, pers. comm.), which was very different from the treeless sclerophyll

(fynbos) vegetation of the area today. Macroplant remains also include rem-

nants of trees. The plant fossils provide no direct evidence of age. but the

vegetation is consistent with a local one of Miocene age (Coetzee 1978).

The phosphatic sandstone mentioned above is recorded up to an elevation

of 30 m, and is a significant element in the succession. It contributed to

Tankard’s (1975a: 262) dating of at least this element as ‘middle Miocene

(Langhian)’, and provided some indication of the likely marine environment at

the time that it was developed (Tankard 1974a; Birch 1977). In addition. its

resistant nature served to protect the underlying deposits from erosion during

the marine regression that followed during the late Miocene.

The 30 m elevation of the phosphatic sandstone recorded in *E° Quarry

does not necessaily reflect the maximum elevation reached by the early to

middle Miocene transgression. If this phosphatic rock formed in an open shelf

environment as suggested by Birch (1977, fig. 4), then the ‘E’ Quarry area

might have been as much as 200 m below the sea-level of that time.

Another indication that this transgression reached an elevation substan-

tially higher than 30 m is the presence on the southern Namib Desert coast of

apparently contemporaneous deposits at 33 m (Table 2). It is evident from the

elevations of the late Tertiary shorelines post-dating the 33 m terrace that there

PALAEOECOLOGY OF LANGEBAANWEG ul

has been appreciable downwarping of this coast. For example, the suggested

equivalent of the 90 m early Pliocene shoreline on the South African west coast -

is at an elevation of only 20-25 m in the southern Namib Desert, and drops to

even lower elevations northwards from the Orange River (Table 2; Hallam

1964). The implication is that the 33 m terrace is now also much lower than its

as yet unrecorded counterpart on the South African west coast.

It is, therefore, unlikely that the 30 m record at Langebaanweg and the

33 m record in the southern Namib represent the maximum elevation of the

early to middle Miocene transgression. There is apparently no way of determin-

ing this elevation on the basis of available evidence, and this is reflected in the

Langebaanweg record represented in Figure 3. However, it is possible that the

?Eocene marine deposits at Buffels Bank and Quaggaskop in Namaqualand

(see p. 19) are, in fact, of middle Miocene age. The former are at an elevation

of about 140 m, and although the elevation of the Quaggaskop deposits is

uncertain, it is apparently about 120 m. The middle Miocene transgression

might therefore have reached an elevation in the order of 120-140 m.

The middle to late Miocene regression

The middle to late Miocene regression affected the Langebaanweg area by

causing erosion of deposits laid down during the preceding transgression. The

middle Miocene phosphatic sandstone exposed in ‘C’ and ‘E’ Quarries, and

encountered in boreholes to the south and south-west of these quarries, was

reduced by wave action to a gravel comprising pebble-, cobble- and boulder-

sized elements. The north-easterly limit of this gravel, which was briefly

exposed in trenches on the floor of ‘E’ Quarry during 1972, occurs at an

elevation of about 30 m. The gravel itself forms the basal unit of the Varswater

Formation (i.e. the Gravel Member—GM).

The GM evidently had a complex history. For example, Butzer (1973: 238)

recognized five stages in its development, excluding the formation of the

original phosphatic sandstone. They are:

1. ‘Reworking of [the primary middle Miocene] phosphatic rock.’

2. ‘Induration of the previous agglomeration with phosphate’ to form

what is here termed the secondary phosphatic rock of the GM (Fig. 5).

3. ‘Reworking of this new generation of [phosphatic rock] into cobbles

and blocks.’

4. ‘Partial carbonate cementation and iron-enrichment of the previous

agglomerate’.

5. “Reworking of all the previous products of induration into a matrix of

unconsolidated, white (1OYR) medium-to-coarse-grade sands.’

Assemblages of marine fossils indicative of sandy and rocky beach environ-

ments occur in the secondary phosphatic rock (stage 2), and in the unconsoli-

dated sands (stage 5). These were termed ‘Marine Faunal Units 1 and 2’

respectively by Hendey (1974a), and their invertebrate components were

DD ANNALS OF THE SOUTH AFRICAN MUSEUM

14 Ol

Alii

VLU

AAU

Fig. 5. A Gravel Member (GM) boulder of primary phosphatic rock (labelled

section), encrusted with fossiliferous secondary phosphatic rock.

described by Kensley (1972), with some additional taxa having been recorded

by Tankard (1974a, 1975b).

Evidently an appreciable period of time elapsed between stages 1 and 5 in

the development of the GM, and throughout this period the deposits must have

been at, or near, sea-level. Since the primary phosphatic rock might also have

been formed in a similar environment, it is possible that sea-level was at an

elevation of about 30 m for much of the middle to late Miocene.

However, this interpretation is not favoured and it seems much more likely

that the early to middle Miocene transgression, during which the primary

phosphatic rock was formed, reached a much higher elevation, as indicated

above.

Another possible interpretation of the history of the GM is that it developed

during a temporary stillstand in the middle to late Miocene regression. Accord-

ing to Vail & Hardenbol (1979) this regression was interrupted by two periods

when sea-level was nearly static, one between 13 and 10 Ma (cycle TM2.3 of

supercycle Te), and the other between 10 and 6,6 Ma (cycle TM3.1 of supercycle

Te). Sea-level during the former period was comparable to that during the

Langhian (Vail & Hardenbol 1979: fig. 8), the age when the primary phosphatic

rock was formed (Tankard 1975a), and this coincidence, coupled with the

coincident occurrence of the primary and secondary phosphatic rocks in ‘E’

Quarry, may indicate that the GM dates from cycle TM2.3. On the other hand,

the elevation of the GM relative to that of subsequent high sea-levels recorded in

the Langebaanweg area suggests that the GM was. developed during the cycle

TM3.1, that is, immediately prior to the terminal Miocene regression.

PALAEOECOLOGY OF LANGEBAANWEG 23

It is also possible that one or more of the later stages in the development

of the GM took place during the subsequent early Pliocene transgression.

While this possibility cannot be dismissed, it is not favoured here, because both

invertebrate assemblages of the GM include warm-water taxa, suggesting sea

temperatures consistent with the late Miocene, whereas at least during the early

stages of the early Pliocene transgression sea temperatures were lower (see

p. 16). In addition, the repeated reworking of previous products of deposition

observed by Butzer (1973) is more consistent with a regression than a transgres-

sion.

A considerable amount of attention has been focused on the origin and age

of phosphatic deposits, including those in South Africa (e.g. Tankard 1974b;

Birch 1977; Siesser 1978) and further investigation of the primary and second-

ary phosphatic rocks of the GM may provide additional evidence relevant to

the history of this horizon. Tankard’s (1975a) belief that the phosphatic rock

horizons on the west coast are all of middle Miocene age is now discounted

(Siesser 1978; Dingle et al. 1979). This probably is indeed the age of the

primary phosphatic rock represented in the GM, but the secondary phosphatic

rock may be appreciably younger. Siesser (1978) has concluded from evidence

on the South African continental shelf that phosphate deposition also occurred

during the late Miocene and into the Pliocene. Although there is at present no

direct evidence linking this period of phosphate deposition with the secondary

phosphatic rock, such a link would be consistent with the inferred late Miocene

age of the GM.

It was suggested elsewhere (Hendey 1976a) that the GM includes some

fossils derived from pre-existing deposits. This suggestion was based on the

belief that the GM was broadly contemporaneous with the overlying QSM and

PPM, and the presence in the GM of some teeth tentatively identified as

belonging to the late Miocene horse, Hipparion primigenium (Hendey 1976a),

which is distinct from the H. cf. baardi and H. ct. namaquense found in

overlying deposits (Hooijer 1976). However, now that the GM 1s interpreted as

a late Miocene deposit, the H. cf. primigenium teeth may well be contempora-

neous with its deposition. If so, then the GM can be no older than 12,5 Ma, the

earliest date at which Hipparion could have entered Africa (Churcher &

Richardson 1978).

The reworking and accumulation of deposits constituting the GM ceased

with the retreat of the sea from the Langebaanweg area, probably at the time

of the terminal Miocene regression. Tankard (1974a) suggested that this retreat

was caused by the development of a sandbar to the south of ‘E’ Quarry during

a temporary stillstand in the transgression which led to the deposition of the

Varswater Formation. Tankard has demonstrated that such a bar did, indeed,

exist during the period of deposition of the QSM and PPM, while it is certain

that these elements in the succession were deposited during a transgression.

However, the stillstand postulated by Tankard is here interpreted as a regres-

sion (i.e. the terminal Miocene one), and it is believed that the bar was built up

24 ANNALS OF THE SOUTH AFRICAN MUSEUM

only during the subsequent transgression (i.e. the early Pliocene one), when a

river began feeding sediment into the area.

The interpretation of the sea-level history at that time as ‘transgression—

regression—transgression’ over a prolonged period, rather than a shorter period

of ‘transgression-stillstand—renewed transgression’, was the key to the correla-

tion of the ‘E’ Quarry succession with certain global events of the late Miocene

and early Pliocene (see p. 70). This, in turn, suggested a correlation with the

record of sea-level movements around southern Africa as interpreted by Siesser

& Dingle (1981), and to the global sea-level changes recorded by Vail ef al.

(1977) and Vail & Hardenbol (1979).

Hitherto the deposits of ‘E’ Quarry have been interpreted as having been

laid down during a single transgression, probably that which resulted in

deposition of Dingle’s (1971, 1973) unit T, (e.g. Bishop 1980). The revised

interpretation of events indicates that while the OSM and PPM are referable to

T,, the underlying deposits correlate with unit T,. These units are respectively

correlated with sea-level supercycles Tf and Te of Vail & Hardenbol (1979).

The early Pliocene transgression

Introduction

Whatever the history of the GM and the period immediately following its

deposition, it is clear that thereafter there was a major change in the local

environment. This was brought about by another marine transgression, during

the early stages of which a river met the sea in the immediate vicinity of °*E”

Quarry and was responsible for discharging most of the sediment making up the

QSM and PPM of the Varswater Formation. The nature of these sediments was

dependent on the position of the river channel, which moved northwards as the

transgression progressed.

The QSM and PPM are by far the best documented deposits in the

Langebaanweg area (Table 3). The largest assemblages of fossils from this area

are from these deposits and they provide the evidence of age (c. 5,0 Ma) which

indicates their association with the early Pliocene transgression (see p. 94).

Judging from fossil evidence and the likely relationship of these deposits to

others in the vicinity, the OSM and PPM date from the earlier part of this

transgression. The earliest phase in this transgression is not represented by

deposits in the area, while towards its climax the area was largely, or entirely,

inundated by the sea. The only surviving record of local deposition during the

latter period are some of the deposits making up Anyskop, which overly the

PPM south of ‘E’ Quarry.

Three distinct positions have been recognized for the lower course of the

river during the period of deposition of the QSM and PPM. There were

intervals of undetermined duration between the successive positions of the river

channel. These intervals are reflected by evolutionary changes in taxa common

to more than one set of deposits associated with the channel positions (Hendey

1978d, 1980; Gentry 1980; De Muizon & Hendey 1980).

PALAEOECOLOGY OF LANGEBAANWEG 25

When the OSM was laid down, the river channel was to the south-east and

south of ‘E’ Quarry, its estuary being immediately north of the sand-bar

recorded by Tankard (1974a) (Figs 6-7). The configuration of the estuary and

the bar separating it from the sea was then probably similar to the analogous

area of the Berg River today, the principal difference being in the position of

the sea relative to the estuary. The sea is to the north of the present estuary,

which has a north-east to south-west trend, but in QSM times the bar had a

more or less east to west trend, with the sea to the south.

The second position of the channel was about 500 m further north, and it

cut diagonally across ‘E’ Quarry from north-east to south-west. The northward

movement of the channel was probably caused by the transgressing sea breach-

ing the sand-bar. The fossiliferous deposits of the second channel constitute bed

3aS of the PPM, while other largely unfossiliferous deposits make up part of

the undifferentiated element of the PPM. These deposits were laid down over

the OSM, filling the area north of the original bar. The extent of the bar was

thus considerably increased, both vertically and horizontally (Fig. 7). The

sea-floor beyond the river mouth was marked by a submarine channel (Fig. 6).

Later still the channel shifted northwards again, but this time the change was

slight. It might have affected only a relatively short section of the river, perhaps

as little as the last kilometre from its mouth, with the mouth itself probably

remaining in essentially the same position as in bed 3aS times. The new channel

still had a north-east to south-west trend, although once beyond the northern

perimeter of the quarry it swung southwards and truncated the south-westerly

parts of bed 3aS (Fig. 6). The deposits of the third channel. and associated

fossiliferous ones, constitute bed 3aN of the PPM, and once again contemporary

unfossiliferous deposits are included in the undifferentiated element of the PPM.

Thereafter, the river must have continued to discharge sediment (PPM

undifferentiated) into the area for a while, but as the transgression progressed,

the river mouth must have moved further north or east from the *“E*° Quarry

area, with a consequent diminution in the influence of the river on local

sedimentation. At the time of the climax of the transgression (90 m), the river

mouth must have been many kilometres away. The final extent of local

deposition at this time cannot be determined since the PPM in ‘E* Quarry. and

to the north and east. was truncated to a maximum of 50-55 m by erosion

during the subsequent late Pliocene regression. The only other recorded local

deposits dating from the early Pliocene transgression were some of those

comprising Anyskop (see below).

It has previously been assumed that the progressive northward shift in the

river channel during QSM and PPM times was caused by the rise in sea-level

during the transgression. A second factor involved may have been the build-up

in a northerly direction of the sand-bar separating the estuary from the sea.

Substantial sedimentation was then taking place, and this, together with the

rising sea, might have affected the lower course of the river. especially in the

case of the slight change from bed 3aS to bed 3aN.

ANNALS OF THE SOUTH AFRICAN MUSEUM

yueulWOpeid

SO}VIQSIOA =: [VIIJSOIIO} «= S pues

oneydsoyd Ul UOWUWOS ssoy

SUIWIODIQ Jnq ‘JURpUNQe S|ISSOJ

pues oneyd

-soyd ojur spiemdn suIpeis

sey jissoy} pue pues zjienb

PaPIAIp

-qns

jou

yueg IOAlI pue jouuvyo JdAII

quosoid Ajqeqoid

sud][od ynq ‘popso0deI $91¥1G9}10A

AjUO ‘]]J] JO YINos uUOUWUWOS

SSO] AJPAISSoISOId S]ISSOJ

suOzIjoy Avid pue

pues zjienb Jopun sowitjawWo0sS

‘AeJO pue pues snoasdeuogieo

puod pue ysiew

Aejo pue

JUCUIWIOPSId Sa}e1Q9}JOA [P11] | PUBS SNOSDeUOGIeS OJU! SUIPCIS | A] PU J] USOMJOq J}RIPOWIO}UI Ill

-S9119} PUB S]BIS ‘ UOWIUOD SjISsOJ | Sey jIssoJ} pue pues zjienb

UdddWNAAY ALIYOHdSOHdG 1WV14d114dd

yueuIWwOpeld Wdd

S3}VIQSJIOA [C1I}S9119} y901 a}eydsoyd jo 99] UI pue JQUUBYD J9ATI Il

pue sjeas ‘juepunqe sjissoj| “uo se] jISsoJ pue pues z}1enb

tay td

yUvUTWOpeid

SO}CIQO}IOA [CII}SOIIO} 4901 a3eydsoud uo pues Adkej9 yurg JOA I

pue sjeos ‘yuepunqe sjissoj

aie AI9A S3yeIq payenusoJayipun 1d

-OJIOA “S|ISSOJOIOIL JUITBU IWOS pues oeydsoyd j1O}}] QUITBUT ‘Wdd a-V

=®

SAONTNAUNIIO TIsso4 SLNAWIGIS OLLSINALOVUVH) SLNINNOWIANY TVNOILISOdIG }JOda1 SIU} Pu ORG] ABPUSH ~ oO

sas)

| :

“‘SOMULLQISUP |

‘AIIENG) .q, “UONPUNOT JO}eMSIVA 34} JO Wedd PUB WSO 94) UL S9dUdTINIDO |ISSOJ PUB SJUSWIPIS SIVSIIBJOVIeYS ‘SJUIUUOIIAUS |BUOI}ISOdaq

€ dav

PALAEOECOLOGY OF LANGEBAANWEG

"(¢ “8 “6161 7/2 /2 a[8ulq 92S)

syun AleUdUUIpPaIs JO Sade} 918 (AT O01 [ INSO Puke AT O1 | N&E Wedd ‘S}ug;eAinba sJ194y} pue) pa O} [| pue ¢q pu ZC SUOISIAIPqns su

UMOUy 30U

yueUIWOpald so}eIq

-SJIOAUI «~JoyeMYsos] pue sUIIe

-njso “oulIeU 2JUBpUNQe S|ISso}

UuOWWOS

$d} BIQIJIOA [BII{S91I9} ‘JURUTLUOp

-o1d suajjod ‘yuepunqe syjissoj

yUeUIWOpold

SOIBIQIJIOA [BIISOIIO}

>UOWWOS SdWI}IWIOS SjISSOJ

yueuIWwopaid

SOJCIQOIIOA [CII}S9II9}

:jUBpUNge SOUWIIJOWIOS s]ISso}

uMOUy JOU

ys Appnur

Aejo pue pues snosseuoqgies

pues z}1enb

(Ajzenb jo yynos

pue jsea-yyNOS s}sIxo Ajqeqoid

ynq pasodxd jOU) JOUUeYD JOATI

syey [ep!)

Ysieul

JOUUBYD JIAIJ 0} JOSO]O UTe]d

-pooy ¢ ‘| JO uorerseA o[qissod

(jouUvYS BSeUIeIP IOUIW

‘puod °3°d) S}UOWUOJIAUSOJOIUU

poyersosse = JIM

‘ulejdpooy

"10N

(Al) va

fe)

IIT S cd

v2)

= = o)

IT WSO y rae

iw)

™

(V)I cS

v2)

ANNALS OF THE SOUTH AFRICAN MUSEUM

(¢ “S14 “PPL6] preyurL wos

podepy) “UMop pir] 919M NPE Wdd Pur “Sve Wdd “(NSO) Joquiayy purs dsozyeNd ay) UdYM SAUIl] dy] Je JIUUBYS dUTIIeUIQNS B pUe sjauURYO

JAM BY} JO SUONISOd posiajur ayeoIpul soul] PaMOLIW “(WWdd) Joquiay asoydsoyg [e12@]][9g ay} JO aseq ay} yo dew imojuod aInjong “9 “SIF

AYHVND 3,

PALAEOECOLOGY OF LANGEBAANWEG 29

While the bed 3aS deposits overlie and are generally at a higher elevation

than those of the OSM, the situation in respect of bed 3aN relative to bed 3aS

is more equivocal. Some of the bed 3aN deposits are higher than any of bed

3aS, but nowhere was the former clearly superimposed on the latter. In fact,

there was a lateral (westerly) truncation of bed 3aS by bed 3aN. and both were

superimposed on the OSM (Fig. 6, Table 8).

The actual elevation of sea-level when the river was in its various positions

cannot be accurately determined. The fossiliferous deposits of the OSM and PPM

range between elevations of 30 m and 40 m, which is a possible indication of the

magnitude of the sea-level rise which took place during their deposition. However,

the actual rise might have been less, since much of the deposition associated with the

3aS and 3aN channels apparently took place during floods when river levels were

high. Not even the elevation of the OSM tidal mudflat bed (QSM I—Table 3) is of

assistance in resolving this issue, partly because this bed was truncated by overlying

deposits, and partly because nothing is known of the local tidal range at that time.

This range may have been large because of the estuarine situation of the mud-flat

bed. The complex nature of tides in estuaries is well known, with ranges being

markedly affected by the volume and strength of the river flow. In addition, it is

possible that at the time that the QSM was laid down, the sea connection between

the present Saldanha and St Helena Bays was incomplete, in which case the ancient

Berg River, with its seasonal floods, would have discharged into a large bay, a

combination of circumstances which can also affect tidal ranges.

The OSM

The first of the three positions of the river channel was not exposed in *E’

Quarry, although it is possible that some of the deposits in the south and

south-east of the quarry were laid down in an enlarged flood-season channel.

The proximity of the channel to this area is indicated by the nature of the

deposits in the south-central part of the quarry, and, to a lesser extent, by other

exposures of the QSM on the floor of the quarry. Three main depositional

environments (i.e. facies) of the QSM are recognized. All were truncated

during the initial stages of the deposition of overlying deposits and _ their

original depth and areal extent can no longer be determined. Each of the facies

is characterized by distinctive sediments and fossil assemblages (Table 3).

In the south-central part of ‘E’ Quarry there survives the vestiges of a

tidally inundated mud-flat bed (QSM III of this paper, or Layer E3 of Dingle et

al. 1979). The fossil assemblage of this muddy silt includes an extensive

invertebrate assemblage which was described and discussed by Kensley (1977).

Vertebrate remains are rare and are comprised mostly of small bone fragments

washed in from elsewhere. The most remarkable of the non-invertebrate fossils

are fragments of bird egg-shell, which, like some of the invertebrates, retain

traces of their original colour. The invertebrates include terrestrial, freshwater,

estuarine and marine species, and comprise an assemblage typical of a mud-flat

environment (Kensley 1977).

30 ANNALS OF THE SOUTH AFRICAN MUSEUM

EARLY PLIOCENE (A)

ee floodplain X fy

=| salt marsh X /j

Stidal Mack /4;

EARLY PLIOCENE (B) ,

Fig. 7. The Langebaanweg area during the early Pliocene (sea-level cycle TP1). A. Period of

deposition of the QSM (X-‘E’ Quarry exposures of the QSM). B. Period of deposition of the

PPM 3aN (Y-‘E’ Quarry exposures of the PPM 3aN). (See Fig. 4 for key.)

Immediately adjacent to the mud-flats bed, largely in a north-easterly

direction, is the more extensive ‘peat’ bed of the QSM (QSM II of this paper,

or Layer E2 of Dingle er al. 1979). These black, carbonaceous sands and clays

represent a marsh deposit. Their fossil content has yet to be analysed in detail,

but they are rich in both pollens and vertebrate fossils. They were briefly

discussed by Hendey (1976a: 225-226) and Rich (1980).

PALAEOECOLOGY OF LANGEBAANWEG Sil

Elsewhere to the north-west, north, north-east, and east are exposures of

the quartzose sands which are the main component of the QSM (QSM I of this

paper, or Layer E1 of Dingle et al. 1979). These deposits apparently accumulated

on the floodplain of the river, and were therefore partly subaerially and partly

subaqueously deposited (Hendey 1974a: 32-33, 35-36, 349-353: 1976a:

223-224). It is believed that the inundation of these deposits during flood periods

was gentle rather than torrential. In other words, the area was not affected by the

strong currents in, and immediately adjacent to, the main river channel, but was

inundated by slow-moving backwaters which overflowed from the main channel.

The representation of fossils in QSM I is variable both in terms of condition of

specimens and associations of taxa. Apart from indicating the subaerial and

subaqueous conditions of dry and wet seasons respectively, they suggest the

existence of such microenvironments as ponds and minor drainage channels

(Hendey 1976a: 223, 226). Terrestrial vertebrates predominate, and there is good

evidence that at least some, and perhaps even most, were subaerially accumulated

(Hendey 1974a: 351; 1976a: 224). On the other hand, some subaqueous deposition

also took place, together with subaqueous disturbance of previously deposited

materials. This would account for the occasional presence of aquatic species in

association with terrestrial ones, and the nature of certain of the occurrences.

Although a wide variety of taxa are represented in QSM I, the most

characteristic are a tortoise (Chersina sp.), a francolin (gen. and sp. not

determined), a rhinoceros (Ceratotherium praecox), a pig (Nyanzachoerus cf.

pattersoni (or kanamensis)), and a boselaphine antelope (Mesembriportax (or

Miotragocerus) acrae). The tortoise is the most commonly represented of these

taxa. Remains of this animal were sometimes present in astonishing quantity,

and they occasionally occurred where no other fossils were obvious (see p. 78).

While much of QSM I was probably deposited some distance from the main

channel of the river, this was not the case with the more southerly and

south-easterly exposures. Here there are areas of coarser-grained sands, as well

as clayey sands, while the associated fossils differ in both condition and species

representation. Amongst the species recorded from these deposits are a seal

(Homiphoca capensis) and giraffids (Giraffa sp., Sivatherium hendeyi*), which

are otherwise characteristic of QSM II and the channel deposits of the PPM. The

seal is known from elsewhere in OSM I, but there it is represented mainly by the

remains of very young individuals, whereas in the other deposits mentioned

above, sub-adults and adults predominate. The latter probably died in, or near

one of, their natural habitats (i.e. the river channel), whereas the very young

ones of QSM I might have been carried away from their nursery by terrestrial

carnivores.

To sum up, the QSM was laid down when the river channel was largely to

the south and south-east of ‘E’ Quarry, where it was separated from the sea by

a sand-bar (Fig. 7). Remnants of a tidal mud-flat deposit laid down on the

* Churcher (1978) suggested that S$. hendeyi is conspecific with S. maurusium, but it is here

regarded as sufficiently distinct to warrant separate specific status.

32 ANNALS OF THE SOUTH AFRICAN MUSEUM

northern margins of the estuary are preserved, as are remnants of an imme-

diately adjacent salt marsh. Elsewhere there are deposits which were inundated

by floods during the wet season. The picture that emerges is the almost classic

one of an estuarine environment and its associated sedimentary facies. The

present Berg River estuary appropriately provides an ideal modern analogue.

Perhaps the most significant characteristic that distinguishes the QSM from

the PPM is that the former are non-phosphatic. This is simply explained by the

fact that the development of authigenic phosphate did not occur in those facies

of the QSM exposed in ‘E’ Quarry (i.e. floodplain, salt marsh and tidal flat).

However, there evidently was phosphate deposition in other local environments

which existed at the time that the “E’ Quarry QSM deposits were laid down.

For example, the deposits comprising the sand-bar which separated the estuary

from the sea are phosphatic (Tankard 1974a), as are those which accumulated

in the marine littoral environment. The latter environment encroached on the

‘E’ Quarry exposures of the QSM as the transgression progressed, and resulted

in the deposition of the PPM in this area. There was a coincident north-easterly

migration of the estuarine facies in this depositional system (i.e. the QSM).

The QSM and PPM therefore represent lithostratigraphic units, and not

chronostratigrahic units (Fig. 8), and may be termed ‘magnafacies’, while any

chronostratigraphic units which may yet be defined would be termed ‘parva-

facies’ (see Krumbein & Sloss 1963: 320, fig. 9-6). It is fortuitous that none of

the PPM deposits laid down contemporaneously with the QSM have yet been

exposed in ‘E’ Quarry, while the QSM counterparts of the ‘E’ Quarry PPM

deposits are also not exposed, and may, in fact, no longer be preserved.

E’ QUARRY =

ne =

m= oa

= Se pe igs

i-5 ec —

<x So —

= =

oO Ls yv~o — 2

= i 2 & OF a)

= ° oo S oo g c

° ° fe} ° ° °

c Oey Seo oe? ke

9° ° °° fe} © fe} w

Ls ° ° ° o

n ° ° oo ° oo °

Oo ° fe} 9,0 °,0 ° =

° oo

fe}

2 eek PPM »50 x

on BO O69 859° 9 =

cre 5 ©9 6 89 5 GO

= — 959 9,59 0,0

oz =— oO o—|]-<0 S70 te Oo =O

°o? °,° 25° °

= 0 22 5 9° oo-<

= 02 969° Nry

C~¢ OO 6 OO & Sy

a FROMM OM AS NS ie eG aS

Os f i

= fo} Ye ~ =

S 1S Se as M Si

LNT oR SS

= sy RE 2ST “>

OSS ee WP UGS

= Se ean oe

=) Se

a aoe

> ae

Fig. 8. A diagrammatic representation of the QSM and PPM lithostratigraphic units (‘magna-

facies’), and their relationship to hypothetical chronostratigraphic units (‘parvafacies’).

PALAEOECOLOGY OF LANGEBAANWEG 33

The PPM

The channel deposits, and incorporated fossil assemblages, of beds 3aS and

3aN of the PPM are essentially similar. However, bed 3aS was in the form of a

relatively straight channel superimposed on an unconsolidated substratum,

whereas in the bed 3aN area an outcrop of phosphatic rock caused a southward

deflection in the course of the river, and, at least in flood times, it provided a

consolidated substratum for part of the channel (Hendey 1980: 58-62, fig. 26).

In addition, the seaward end of bed 3aS was truncated by the bed 3aN channel.

(Fig. 6), which resulted in the loss from bed 3aS of the marsh and pond facies

evident in bed 3aN (Table 3).

The bed 3aN deposits were recently discussed in some detail (Hendey

1980: 58-67), and were also mentioned earlier (Hendey 1976a: 228-230). The

latter account also deals with bed 3aS, while this bed (formerly 3a) was also

discussed by Hendey (1974a: 33, 36-37, 353).

While beds 3aS and 3aN have many species in common with the OSM, their

faunas are also in some respects distinct (see pp. 45-53), and the condition of

specimens and relative numbers of individuals in any given taxon may be very

different. The most characteristic of the beds 3aS and 3aN taxa are the seal

(Homiphoca capensis), giraffids (Giraffa sp., Sivatherium hendeyi), alcelaphine

antelope (Damalacra neanica and D. acalla), and a reduncine (Kobus subdolus).

There are, however, differences in the relative numbers of individuals repre-

sented in beds 3aS and 3aN, and to a lesser extent also in the taxa represented.

For example, the alcelaphines occur much more commonly than giraffids in bed

3aS, whereas in bed 3aN the reverse applies. In addition, species of Palaeotra-

gus, Tragelaphus, and Kobus are known only from bed 3aN.

Most of the fossils from beds 3aS and 3aN represent the remains of animals

washed into the area by the river, but at least some must represent animals that

were resident locally. The latter include aquatic species such as the seal and

fishes, while others might have lived on the river-banks. For example, hyaenas

might have roamed the banks scavenging carcasses stranded there.

The river might have had an important effect on the composition of the

bed 3aS assemblage. At the time that much of the latter was being deposited,

the river was flowing over the unconsolidated and often highly fossiliferous

deposits of the QSM. It is certain that at least some, and perhaps even a great

many, OSM fossils were picked up by the river and redeposited in bed 3aS.

Since such derived fossils might have suffered only minimal transport, they

would not necessarily show obvious signs of reworking. There is little likelihood

of a similarly serious complication with the bed 3aN assemblage (Hendey

1976a: 230).

Although the deposits and faunas of the QSM, bed 3aS, and bed 3aN

differ, and despite the evidence that intervals of time elapsed between their

deposition, there can be little doubt that they date from a relatively short

period and represent repetitive depositional phases of a single geological

episode. The period was long enough for evolutionary changes to take place in

34 ANNALS OF THE SOUTH AFRICAN MUSEUM

taxa common to more than one level (see p. 24), and it is possible that the

composition of the local fauna also changed during this time (see p. 86).

However, most of the faunal differences, and all of the geological ones, can be

ascribed to the exposure in various areas and levels of differing depositional

environments within a major depositional system (i.e. an estuary and associated

environments).

Fossils do occur in the PPM deposits overlying bed 3aN, but they are not

common and probably represent the remains of animals occasionally washed

into a marine littoral environment by currents emanating from the river mouth,

which was progressively further from the “E’ Quarry area as the early Pliocene

transgression advanced.

The upper levels of the PPM, and, indeed, all the undifferentiated parts of

this unit, include marine microfossils such as foraminifera, fish teeth, and

mollusc fragments (Tankard 1974a).

The record of deposition of the PPM in the Langebaanweg area is

probably incomplete, since there was evidently truncation of the higher levels

during the subsequent late Pliocene regression.

Mobilization of phosphate in the PPM and its redeposition into phosphatic

rock horizons might have continued after deposition of the PPM itself ceased.

There are several phosphatic rock horizons in the more northerly exposures of

the PPM in ‘E’ Quarry and it is evident that they did not form contempo-

raneously (Dingle et al. 1979: 89). The one that affected distribution of bed 3aN

was formed after the accumulation of the QSM and before deposition of bed

3aN, while others higher in the sequence were formed subsequently. A possibly

significant feature of these horizons is that they occur only in the more northerly

exposures of the PPM. The bulk of the PPM deposits, which are spread out south

and south-west from ‘E’ Quarry, have at most thin lenses of phosphatic rock

which are superficially very different in appearance. The more northerly

phosphatic rock horizons cannot have formed in a deep-water continental shelf

situation, which was suggested in the case of the middle Miocene phosphatic rock

(see p. 20). The fact that they occur only on what was the landward side of the

developing Varswater Formation, suggests the possibility that seasonal

movement of the watertable in the phosphate-rich sediments might have led to

precipitation of a phosphatic matrix in the zone of movement.

Deposition of pelletal phosphorite apparently ceased during a stage in the

transgression when the Langebaanweg area was so deeply covered by the sea

that conditions necessary for its formation were no longer being met (see Birch

1977, fig. 4).

The Anyskop marine deposits

It was previously assumed that all those deposits overlying the PPM in the

vicinity of ‘E’ Quarry were deposited subaerially during the Quaternary (e.g.

Hendey 1974a; Tankard 1974a; Bishop 1980). However, this apparently applies

to only part of the overlying succession.

PALAEOECOLOGY OF LANGEBAANWEG 35

Although pelletal phosphorite continued to accumulate at higher eleva-

tions south-east of Langebaanweg (e.g. Elandsfontein—Rogers 1980) during

the latter stages of the early Pliocene transgression (i.e. in the shallower water

close to the river mouth and shoreline), in the immediate vicinity of Lange-

baanweg only non-phosphatic deposits were being laid down. This conclusion is

based on a reassessment of the calcareous deposits making up the bulk of

Anyskop, the hill immediately south of ‘E’ Quarry. These deposits have been

included in the ‘Langebaan Limestone’ (Visser & Schoch 1973), which Tankard

(1976) defined as a member of the Bredasdorp Formation. The ‘Langebaan

Limestone’ 1s generally regarded as a composite ‘of limestone types of . . . aeo-

lian origin’ (Tankard 1976: 114). The continued recognition of this lithostrati-

graphic unit is open to question, since it is evident that limestones and calcretes

of different ages and origin are included in it. The Anyskop calcareous deposits

are a case in point. The reinterpretation of these deposits which is outlined

below is tentative, and, although it may well be modified in future studies, it is

obvious that previous interpretations are incorrect. |

Anyskop is a prominent feature in the Langebaanweg area, rising to an

elevation of over 70 m, while the mean elevation in the immediate vicinity is

probably about 30 m. Deposits on its northern slope have been exposed by

overburden stripping associated with the ‘E’ Quarry mining operation. Features

of these deposits, including a previously unknown fossil content, have been

revealed by three years (1977-1980) of subaerial weathering and erosion which

took place while mining operations were suspended. The hill itself is a curious

feature, since it may be the only one in a wide area around Langebaanweg

which does not have a core of Cape granite. In fact, the bedrock beneath the

hill is at about 35 m below sea-level (Rogers 1980), a situation which apparently

applies within a radius of several kilometres. The existence of the hill, there-

fore, has no direct connection with the solid geology of the area.

Also anomalous is the fossil content of the calcareous deposits comprising

much of Anyskop. According to Tankard (1974a: 280-281), foraminifera are

abundant, with Anomalina tests outnumbering those of Elphidium 5: 1, the

former being ‘essentially fresh’, while the latter are weathered. Tankard’s belief

that the forams were ‘blown inland subsequent to the regression of the sea

responsible for the Varswater Formation’ is inconsistent with the condition of

the Anomalina tests and the fact that this genus occurs most commonly in

continental shelf rather than beach situations (R. A. Martin, pers. comm.). It is

much more likely that the fresh Anomalina specimens were deposited in a

marine environment where no abrasion by wind or wave action was possible.

On the other hand, Elphidium is common in inshore situations and the abraded

conditions of the Anyskop specimens suggest that they might have been

exposed to the wind or waves. It is probably also significant that, whereas

Elphidium and other inshore forams are recorded from the underlying marine

littoral PPM, Anomalina is not.

One possible interpretation of the above observations is that the calcareous

36 ANNALS OF THE SOUTH AFRICAN MUSEUM

deposits of Anyskop were in part accumulated in a marine environment of

deeper water than the PPM, and also in part on a beach or adjacent terrestrial

environment. This is a combination of circumstances which exists on a sub-

merged barrier or barrier island. It is perhaps significant that the south-east to

north-west trend of Anyskop parallels at least a part of the sand-bar which

existed in QSM and PPM times, and it is conceivable that a submerged barrier

or barrier island was a vertical and north-eastward development from this

sand-bar as the Pliocene transgression progressed. Figure 9 illustrates the

hypothetical submerged barrier or barrier island in a ‘drumstick’ form of the

kind described for such features by Hayes (1976). The most substantial part of

this ‘drumstick’ survives as Anyskop today. The east-west trend of the barrier

(Fig. 9) is conjectural, and it might instead have extended southwards towards

Karnberg (Fig. 2; see also p. 39).

The development of this barrier was a very significant event, since the

deposits constituting it were subsequently to protect from erosion the underly-

ing and northward-extending Varswater Formation that survives today. The

preservation of so substantial a body of unconsolidated late Tertiary sediment

has hitherto been difficult to explain, just as there was no obvious reason why

Anyskop should have been developed by aeolian action in its present position

and form.

An apparent anomaly with this interpretation is the observation by Tan-

kard (1974a) that the most commonly occurring fossils in the calcareous sands

are shells of the land snail, Trigonephrus globulus. Also common are shells of

EARLY PLIOCENE

(C)

Fig. 9. The Langebaanweg area during the early Pliocene (sea-level cycle TP1). C. Period of

deposition of the Anyskop marine deposits. Sea-level at about 60 m above present level

(transgression maximum at 90 m). (See Fig. 4 for key.)

PALAEOECOLOGY OF LANGEBAANWEG 37)

another, unidentified land snail. The snail shells are now known to occur

together with fragmentary vertebrate remains, mainly of terrestrial species,

although marine and freshwater taxa are also recorded. Associated with these

fossils are ‘fossil roots’ (rhizoconcretions) and as yet unexplained circular

structures of calcrete with infillings of reddish sands (Fig. 10). The latter have

been suggested to represent remnants of tree stumps or clumps of sedge,

termitaria, or solution cavities.

Fig. 10. A. An exposure of the Anyskop terrestrial deposits showing one of the unidentified

features referred to in the text, together with snail shells and the horn-cores of an unidentified

antelope. B. A close-up view of a snail shell and the antelope horn-cores.

38 ANNALS OF THE SOUTH AFRICAN MUSEUM

There can be little doubt that these structures, the rhizoconcretions, and

the fossils mark an old land surface. They appear to be confined to a horizon

extending from high up the north slope of Anyskop down towards the ‘E’

Quarry exposures of the PPM, although they are not known to occur in situ

below an elevation of about 40-45 m. In areas on the north slope which have

been excavated well below the present surface, the structures and fossils are

absent. These observations suggest that the old land surface was more or less

coincident with the present northern contours of Anyskop, occurring over a

limited vertical extent of the calcareous deposits immediately beneath the

unconsolidated surface sands. In other words, the fossils were deposited on

Anyskop at a time when the north slope was more or less in its present form.

The vertebrate fossil assemblage is as yet unstudied, and it may be too

small and the fossils too fragmentary to give an accurate indication of age. The

assemblage is, however, definitely distinct from those of the underlying QSM

and PPM. It includes at least one species otherwise known only from the late

Pliocene/early Pleistocene Baard’s Quarry assemblage (i.e. Hipparion baardi,

an advanced member of the lineage which includes the ‘E’ Quarry H. cf.

baardi—Hooijer 1976; Hendey 1978a). On the other hand, an unidentified

antelope from the Anyskop fauna (Fig. 10) is clearly distinct from the ones

represented at Baard’s Quarry, and, indeed, all other antelopes hitherto

recorded from the south-western Cape.

The implications of the Anyskop ancient land surface will be discussed

below, but its significance here is that there appears to be no direct association

between its terrestrial fossils and the marine fauna (including forams) recorded

by Tankard (1974a). Consequently, a distinction is made between the ‘Anyskop

marine deposits’ and the ‘Anyskop terrestrial deposits’, with the former being

correlated with the early Pliocene transgression, and the latter post-dating this

event. There is an as yet undetermined relationship between the Anyskop

deposits and those overlying the Varswater Formation on the western side of

‘E’ Quarry (exposure 1 of Dingle et al. 1979, figs 2-3).

The early Pliocene transgression reached an elevation of 90 m, which is

about 20 m above the maximum height of Anyskop. While it is possible that

post-early Pliocene erosion reduced the height of Anyskop, it might never have

been as high as 90 m, which means that at least during the latter stages of the

transgression, Anyskop was probably in the form of a submerged barrier rather

than a barrier island.

Remnants of the 90 m marine platform exist to the south-east of Lange-

baanweg. Its presence is inferred from the topography (Mabbutt 1956), while

marine deposits up to an elevation of 80 m have been recorded on the farm

Elandsfontein (Rogers 1980). Evidently it was developed in the lee of a series

of hills formed by outcrops of Cape granite. The trend of these hills is from

north-west to south-east, roughly parallel to the present coast. The hills are

Karnberg (177 m), 12 km south-south-west of Langebaanweg, followed by

Massenberg (161 m), Groot Swartberg (287 m), and Slangkop (258 m). Subse-

PALAEOECOLOGY OF LANGEBAANWEG 39

quent to its formation the 90 m platform was incised by the valley of the Sout

River, a northward-flowing tributary of the Berg River (Fig. 2).

It was suggested above that the Anyskop barrier island may have extended

southwards in the direction of Karnberg, rather than eastwards as indicated in

Figure 9. If this was indeed the case, then Karnberg and the other hills in the

series must have been influencing the local topography at an earlier stage

during the early Pliocene transgression. The Anyskop barrier island might in

fact, have been a longshore spit extending northwards from Karnberg.

This, and other, uncertainties relating to the later history of the early

Pliocene transgression should be resolved by further investigation of relevant

deposits in the Langebaanweg area.

The early Pliocene transgression which resulted in deposition of the OSM,

PPM, the Anyskop marine deposits and other marine deposits south-east of

Langebaanweg, represents sea-level cycle TP1 of supercycle Tf (Vail & Har-

denbol 1979: fig. 8).

The late Pliocene regression and early Pleistocene transgression

The history of late Pliocene and early Pleistocene deposition in the

Langebaanweg area is as yet not well documented, and inference played a large

part in the interpretation outlined below, with the nature and likely age of the

Anyskop terrestrial deposits and Baard’s Quarry fluviatile deposits being of

particular significance in this interpretation.

It is unlikely that Anyskop was in its present form during the early stages

of the late Pliocene regression. Had this been the case then it would have been

in the form of an island a few kilometres off the emergent mainland. The fact

that a vertebrate fauna, which is correlated with an early stage of the regres-

sion, left traces of its presence on the north slope of Anyskop when the sea was

at an elevation of between 40 and 60 m, means that a connection with the

mainland must have existed. This connection probably followed the trend of

the pre-existing coastal barrier, which might have been north-south, rather

than east—west, as indicated in Figure 11.

Figure 11 also shows the ancient Berg River passing westwards along the

northern side of this connection and meeting the sea to the south-west of ‘E’

Quarry. There is, in fact, no direct evidence for the position of the river, while

its local presence is inferred simply because it is the most likely source of fresh

water required by species such as the hippopotamus in the Anyskop fauna. A

more tenuous indication that the river was present locally is the unequivocal

evidence for its presence both during the early Pliocene (see above) and

subsequently during the late Pliocene/early Pleistocene (see below).

A sea-level elevation of about 50 m at this time is deduced from the

presence of a marine element in the Anyskop fauna, and from the recorded

elevation of the Anyskop terrestrial fossils and associated features. In addition,

the PPM of the Varswater Formation has been eroded down to a maximum

elevation of 50-55 m in the vicinity of Anyskop. The maximum elevations for

40 ANNALS OF THE SOUTH AFRICAN MUSEUM

LATE PLIOCENE(A)

Groot Springfontein

i ( 50m shoreline)

4 km

LATE PLIOCENE /EARLY PLEISTOCENE(B)

eLBW

Fig. ll. The Langebaanweg area during the late Pliocene (sea-level cycle TP2) and the late

Pliocene or early Pleistocene (sea-level cycle TP3 or Ql). A. Period of deposition of the

Anyskop terrestrial deposits. B. Period of deposition of the Baard’s Quarry fluviatile deposits.

(See Fig. 4 for key.)

the PPM recorded by Tankard (1974a) are 53-54 m near ‘E’ Quarry, 54 m on

the farm Witteklip and at Paternoster, and 47-50 m at Duiker Eiland. Tankard

also records a marine horizon at 50 m on the farm Groot Springfontein,

south-east of Anyskop, the nature of the deposits and associated fossils indicat-

ing a shoreline situation.

PALAEOECOLOGY OF LANGEBAANWEG 4]

The various occurrences at about 50 m elevation are here interpreted as

representing a temporary stillstand in the Pliocene regression, probably the first

of two such stillstands during this event recorded by Vail & Hardenbol (1979)

in their synthesis of global sea-level changes (i.e. cycle TP2 of supercycle Tf).

Slender though the evidence may be, the vertebrate fauna of the Anyskop

terrestrial deposits suggests an age close to, but not necessarily contempora-

neous with, the fauna from the Baard’s Quarry fluviatile deposits (i.e. late

Pliocene/early Pleistocene). It is, therefore, not inconsistent with the inferred

Pliocene age of the 50 m shoreline.

As indicated above, there is unequivocal evidence for the presence of the

river during the late Pliocene/early Pleistocene. This is in the form of channel

deposits revealed west of Langebaanweg station by the prospecting for, and

mining of, phosphate (Tankard 1974a; Hendey 1978a). They are here referred

to as the ‘Baard’s Quarry fluviatile deposits’, and are correlated with the west

coast 20 m shoreline. This represents either the second of the stillstands in the

Pliocene regression recorded by Vail & Hardenbol (1979). (i.e.cycle TP3 of

supercycle Tf), or the climax of the early Pleistocene transgression (i.e. cycle

Q1 of supercycle Q).

According to Vail & Hardenbol (1979, fig. 8) cycle TP3 spanned the period

between 3,8 and 2,8 Ma, the latter being taken by them as the Pliocene—

Pleistocene boundary, while cycle Q1 is dated between 2,8 and 1,7 Ma. The

mammalian fauna of the Baard’s Quarry fluviatile deposits (= ‘lower levels’

fauna of Hendey 1978a) also suggested an age close to the Pliocene—Pleistocene

boundary, and it has recently become evident that at least one element in this

fauna is unlikely to be older than 1,9 Ma (see p. 95). Consequently, the Baard’s

Quarry fluviatile deposits are more likely to date from cycle Q1 than cycle TP3.

This is consistent with the record of the cold-water fauna associated with the

counterpart of the Baard’s Quarry fluviatile deposits on the Namaqualand coast

(Table 2; see also p. 16), and the suggestion that cold-water faunas are

characteristic of the Quaternary.

Figure 11 shows the river meeting the sea south-east of Anyskop, with the

pre-existing barrier island link to the mainland having been severed. If this link

did exist, then it is most likely to have been destroyed by river erosion. the

coincidence of this event with the 20 m stillstand is suggested by the fact that

the Baard’s Quarry fluviatile deposits are situated on a broad, flat plain (? an

ancient floodplain), which extends southwards through the area where the

hypothetical Anyskop-‘mainland’ connection existed (Fig. 11). It would have

been at this time that Anyskop took on its present form.

Subsequently, the river underwent a radical change in direction in its lower

course, shifting northwards to begin incising its present valley in the direction

of St Helena Bay. This process continued during the Pleistocene and at least by

the beginning of the late Pleistocene the Berg River must have been more or

less in its present position, since Tankard (1976) has identified last interglacial

marine deposits in the existing valley.

42 ANNALS OF THE SOUTH AFRICAN MUSEUM

Once the river and the sea were no longer in the vicinity of Langebaanweg,

sedimentation in the area was insignificant. The minor events of the Quaternary

included the development of the unconsolidated aeolian sands which blanket

the area, and duricrust (mainly calcrete) formation within them and in underly-

ing deposits. Early Stone Age artefacts, probably of middle Pleistocene age,

have been found cemented into the calcrete immediately underlying the surface

sands on Anyskop, and there is evidence for post-middle Pleistocene human

occupations in the area as well. There are recorded vertebrate fossil occur-

rences dating from the ?early Pleistocene (Skurwerug, near Saldanha), middle

Pleistocene (Elandsfontein), and late Pleistocene (Sea Harvest at Saldanha,

and others).

The late Cenozoic sequence in the Langebaanweg area is in many respects

unique, and although much attention has been focused on it, especially during

the last decade, most of the geological studies (e.g. Butzer 1973; Visser &

Schoch 1973; Tankard 1974a; Bishop 1980) have dealt only superficially with

the fossil occurrences and the nature of the various depositional environments

represented. Consequently, there is still a need for further study directed

specifically at the context of the recorded fossil assemblages and the history of

the deposits, especially those post-dating the PPM of the Varswater Formation.

Since large volumes of deposit have been mined away, it is no longer possible

to retrieve all the information on the succession which was once available.

Nevertheless, documentary records, sediment samples and exposures still exist,

and they offer a potential for further research.

BIOLOGY

FLORA

The flora of the Langebaanweg area during the period of deposition of the

late Cenozoic succession is not well known. Initially only indirect evidence was

available from which the nature of the local vegetation during QSM and PPM

times could be deduced. For example, large browsers amongst the herbivores,

particularly the long-necked giraffe (Giraffa sp.), have been cited on several

occasions as indicating the presence of trees (e.g. Hendey 1973, 1974a, 1976a,

1980). On the other hand, hypsodont grazing species such as the rhinoceros

(Ceratotherium praecox) indicate that grasslands were also present. The pos-

sible significance of the relative abundance of browsers and grazers at different

levels within the Varswater Formation will be discussed later (see p. 75).

Other indirect evidence for the presence of plants was that of leaf impres-

sions, probably of grass or reeds, on coprolites from the OSM (Hendey 1976a:

224). Attempts to recover pollens from coprolites have so far proved fruitless.

Physical remains of plants themselves are limited to pollens, algal nucules,

and fragments of petrified roots. Studies on pollens from the ‘peats’ of the

Varswater Formation and underlying deposits have yet to be published, but

passing references to these pollens have been made by Tankard (1975b),

Hendey (1976a), Tankard & Rogers (1978), and Dingle et al. (1979). In

PALAEOECOLOGY OF LANGEBAANWEG 43

addition, although not specifically mentioned, the pollens contributed to Coet-

zee’s (1978) study on vegetational changes in the south-western Cape during

the late Tertiary. These pollens indicate that the local vegetation during

pre-GM times (i.e. Miocene) was dominated by forests and woodlands,

whereas by QSM times (i.e. early Pliocene) sclerophyll (fynbos) vegetation was

becoming prominent. Details of the Langebaanweg pollen record will be

published elsewhere (J. A. Coetzee, in preparation).

The algal nucules (charophytes) were identified and discussed by Kensley

(1977), and they have also been studied by I. Soulié (Montpellier), although

her report on them is as yet unpublished. They were preserved in the tidal flat

bed of the QSM (i.e. QSM III), and were probably derived from freshwater

ponds on the OSM floodplain (1.e. QSM I). The existence of such ponds was

postulated by Hendey (1976a).

The fossil roots, which are in the form of slender fragments mostly from

the OSM, have not been studied. It is unlikely that they will reveal much about

the plants to which they belonged.

In conclusion, it should be mentioned that the rhizoconcretions and the

unidentified structures of the Anyskop terrestrial deposits may provide some

indication of the local vegetation later in the Pliocene.

FAUNA

Composition

This section deals only with the fauna of the Varswater Formation, from

which the assemblage of invertebrate and vertebrate fossils is exceptionally

large, both in terms of the numbers of specimens and in the variety of species

represented. However, it represents only a small fraction of the material which

was preserved in the deposits, much of which was lost during mining operations

and much of which remains unexcavated.

At least 230 distinct taxa have been recognized, ranging from protozoans to

mammals. Vertebrates predominate, and there are many phyla of lower animals

that are not recorded. Important groups such as the insects are also lacking, the

only tangible evidence of their presence being grooves on bovid horn-cores left

by keratophagous insect larvae (Fig. 12). Most specimens are well preserved and

individual species assemblages range from a few to many thousands of spe-

cimens. Most taxa are identifiable to species level, and incomplete identifications

usually indicate that taxonomic studies have yet to be undertaken.

The list of taxa which follows is given in systematic order and according to

the horizon from which material was derived. Material of doubtful provenance

(?) is listed under horizons where its presence was likely. In the case of the

equid, Hipparion cf. namaquense, it is still not known if the relevant material

came from the uppermost levels of the PPM, or the lowermost levels of the

overlying deposits (Hooijer 1976). Inferred habitat preferences are given, and

in instances where resting and feeding habitats differ (e.g. amongst many

birds), the latter are recorded.

44 ANNALS OF THE SOUTH AFRICAN MUSEUM

Fig 12. Marks left on a Mesembriportax acrae horn-core from the PPM 3aS

(SAM-—PQ-40071) by keratophagous insect larvae.

Although the lists for individual horizons broadly reflect the local fauna of

the period of deposition, they do not necessarily represent the fauna of a single

habitat. In the case of some levels it is possible that the assemblages include

specimens derived from pre-existing deposits. This applies particularly in the

case of the fluviatile deposits (beds 3aS and 3aN), which are, in addition,

comprised largely of taxa whose remains were washed into the area from

upstream habitats. Palaeoecological interpretations of the assemblages constitu-

ting the ‘E’ Quarry fauna must, therefore, take these factors into account.

An example of the palaeoecological studies which are possible is that by

Kensley (1977) on the invertebrate assemblage from QSM III (1.e. the tidal-flat

bed). In this instance the remains of the resident animal community were

supplemented by others washed downstream by the river and upstream by the

tides. This assemblage proved to be essentially similar to ones in comparable

situations today, and, interestingly, it proved a better indication of the nature

of the depositional environment than the sediments themselves.

This was a general rule with most ‘E’ Quarry sediments, and attempts to

interpret their history without taking into account the composition of individual

assemblages and the condition of specimens, resulted in mistaken conclusions.

For example, Butzer (1973) misinterpreted the depositional environments of

the QSM (= Bed 2) and bed 3aS (= bed 3a) through using only geological

evidence. While Bishop (1980) evidently appreciated the significance of the

fossils, he supported Butzer’s conclusions and ignored the palaeontological

evidence which became available after 1970.

PALAEOECOLOGY OF LANGEBAANWEG 45

PPM PPM PPM Inferred habitat

GM QSM 3aS 3aN_ undiff. preference

PHYLUM PROTOZOA

ORDER FORAMINIFERA

Ammonia beccarii — — — — x marine (littoral) and

Elphidium advenum _. — — — — x estuarine

Cibicides lobatulus a _- — — x marine (littoral;

Planorbulina EE CIOS —— — — — x attached to vegetation

Rosalina cf. bradyi : = = = = x etc.)

PHYLUM BRACHIOPODA (lamp shells)

Kraussina rubra . : oe OX — — = == marine (rocky shore)

PHYLUM ECHINODERMATA

CLASS ECHINOIDEA (sea urchins)

Parechinus angulosus . mene — — — — marine (rocky shore)

PHYLUM ARTHROPODA

CLASS CRUSTACEA

ORDER CIRRIPEDIA (acorn barnacles)

Gen. and sp. indet. . x — — — — marine (rocky shore)

? ORDER DECAPODA (lobsters etc. y.

Gen. and sp. indet. . RIBS — — = == marine, estuarine and

fresh water

SUBCLASS OSTRACODA

Family Cytheridae

Gomphocythere expansa — ‘< — = —

Family Cyprididae

Zonocypris cordata . oo x — — — fresh water

Family not det.

Gen. and at least 3 spp not det. — x — — —

PHYLUM MOLLUSCA

CLASS GASTROPODA

ORDER ARCHAEOGASTROPODA

Family Patellidae (limpets)

Cellana capensis . : hat — — — — marine (rocky shore,

warm water)

? Cellana sp. : : ae — — — — :

Patella granularis Ws - -— ~— — f manne noch Shore)

Family Fissurellidae (keyhole

limpets)

Diodora parviforata . 2 xX --— — — _. marine (rocky shore)

Family Haliotidae (perlemoens or

abalones)

* Haliotis saldanhae : vero — — — — :

Panonis 3B. . wes Z. ie ea ins } marine (rocky shore)

Family Trochidae (top shells

Oxystele tigrina . wR TEX — — — —

Oxystele variegata : — x — — — cok (rocky shore)

Gibbula benzi__.. — x. — — —

Family Turbinidae (turban shells

Turbo sarmaticus : x — —— — = marine (rocky shore,

warm. water)

Family Phasianellidae (pheasant

shells)

Tricolia neritina . : an aX x — — — marine (rocky shore)

Tricolia capensis . : — x — _ ey marine (rocky shore,

cold water)

ORDER MESOGASTROPODA

Family Littorinidae (periwinkles)

Littorina cf. knysnaensis _ ‘< — —

; ; marine (rock

? Littorina sp. (rocky shore)

46 ANNALS OF THE SOUTH AFRICAN MUSEUM

Family Hydrobiidae (snails)

Tomichia ventricosa . ——

Family Assimineidae (snails)

Assiminea sp... s oo

ORDER NEOGASTROPODA

Family Muricidae (rock shells or

whelks)

Ocenebra scrobiculata . aeX

Family Thaisidae (rock shells or

whelks)

Thais dubia : x

Family Columbellidae (dove shells)

Pyrene albuginosa : _

Family Nassariidae (plough shells,

dog whelks)

Bullia sp. . j : Sx

Bullia digitalis. : es

Bullia laevissima ; —

Bullia sp. nov. . A =

Nassarius cf. analogicus —

Nassarius sp.B . ‘ _

Nassarius sp. C . : ._

Family Marginellidae

Marginellasp. . : a

Family Turridae (screw shells)

‘Crassispira’ sp. . : 8 ay

‘Clavatula’ sp. . . a

*‘Turris’ sp. P : $x

ORDER ENTOMOTAENIATA

Family Pyramidellidae

Turbonilla kraussi ; wre

? Pyramidella sp. : yx

ORDER BASOMMATOPHORA

Family Siphonariidae (false limpets)

Siphonariasp. . AX

Family Ferrissiidae (snails)

Burnupia capensis : -_

Family Planorbidae (snails)

Ceratophallus natalensis

Bulinus ‘tropicus’ ; <a

ORDER STYLOMMATOPHORA

Family Succineidae (snails)

Succinea sp. ; : _

Family Endodontidae (snails)

Trachycystis cf. capensis oo

PPM PPM PPM

QSM_ 3aS

——$$<—_— ee

Inferred habitat

preference

fresh or brack water

mainly estuarine

marine (rocky shore)

marine (rocky shore,

warm water)

marine and/or

estuarine

marine (sandy shore)

and estuarine

marine and/or

estuarine

marine (sandy shore)

and estuarine

marine and/or

estuarine

marine and/or

estuarine

marine and/or

estuarine

marine

s marine

marine (rocky shore)

on rushes in fresh

water or estuaries

fresh water (sub-

tropical, tropical)

fresh water

terrestrial or semi-

aquatic (fresh water)

terrestrial

PALAEOECOLOGY OF LANGEBAANWEG 47

PPM PPM PPM Inferred habitat

GM QSM 3aS_ 3aN_undiff. preference

CLASS AMPHINEURA

Family Chitonidae (chitons)

Chiton nigrovirescens . — x — = = marine (rocky shore,

cold water)

CLASS PELECYPODA

Family Donacidae (sand mussels)

Donax serra ; : mesa x — = ' Pain

Donax sp(p) ¥ LAS s eee marine (sandy shore)

Family Carditidae (false cockles)

*Cuna aquaedulcensis . — x — — — marine

Family Arcidae (ark shells)

Barbatia obliquata , tess = — — — marine (rocky shore,

warm. water)

Family Mytilidae (mussels)

? Perna sp. eK a —_ == = marine (rocky shore)

Family Veneridae (venus shells)

Pitar sp. . : : a yox — — — — marine

MOLLUSCA INDET. ; : —— = — a

PHYLUM CHORDATA

CLASS CHONDRICHTHYES

ORDER SELACHII (sharks)

Family Hexanchidae

Notidanus serratissimus A pee x. — — — — marine

Family Carcharhinidae

G archarhinus melanopterus x S marine (warm water)

Carcharhinus limbatus os

Galaeorhinus sp. . : aX — _ -— a=

Prionace glauca . iS a

Negaprion or Hypoprionsp.. xX

Family Odontaspidae

Odontaspis accutissima ain OX — — —- —

Odontaspis sp. B ; Mat — — — —

Odontaspis sp. C : eX — = — —

Family Otodontidae marine

Megaselachus megalodon . x = — — —

Family Carcharodontidae

Carcharodon sp. . : Bes — — — —

Family Isuridae

Tsurus sp. . 4 f eee — — — —

Family Squalidae

Squalus sp. ‘ : sy os — — — —

Family Squatinidae

Squatina africana : a eX — — — _ marine (warm water)

Squatinasp.B . : ikex — — — — marine

SELACHII not det. 3 ‘ aa x x < —

ORDER BATOIDEA (skates, rays)

Family Rajidae (skates)

Rajasp. . oyoX — — = —=

Family Trygonidae (sting rays)

Gen. and sp. not det. x —_ — —~— —_ marine

Family Myliobatidae (eagle rays)

Myliobatis sp. . ‘ x — — = =

CLASS OSTEICHTHYES

ORDER PERCIFORMES

Family Sparidae (mussel-crackers)

Gen. and sp. not det. . eS —- — — — _ marine

48 ANNALS OF THE SOUTH AFRICAN MUSEUM

PPM PPM PPM Inferred habitat

GM QSM 3aS— 3aN_undiff. preference

ORDER SILURIFORMES

Family Tachysuridae (catfish)

? Tachysurus sp. ; — x x x = marine and estuarine

OSTEICHTHYES not det. : a OX x x ‘< x

CLASS AMPHIBIA

ORDER ANURA (frogs, toads)

Gen. and sp(p) not det. — x x x = terrestrial and fresh

water

CLASS REPTILIA

ORDER CHELONIA (tortoises, turtles)

Chersina sp. ; ; “ox x x x ass terrestrial

Gen. and sp. not det. . — x = as = marine or fresh water

Gen. and sp. not det. . — = = x == marine or fresh water

ORDER SQUAMATA

SUBORDER LACERTILIA

Family Chameleontidae (chameleons)

Gen. and sp(p) not det. — x x x — bushes and trees

Family Gekkonidae (geckos) :

Gen. and sp(p) not det. — x x x — terrestrial

Family Varanidae (leguans,

monitor lizards)

Varanus sp. : : oo x — —_ -- terrestrial or fresh

water

SUBORDER OPHIDIA (snakes)

Gen. and sp(p) not det. — x x x — terrestrial

SQUAMATA not det. . : — x x x —

CLASS AVES

ORDER STRUTHIONIFORMES

Family Struthionidae (ostriches)

Struthio sp. : : — == x x = terrestrial (open

country)

ORDER SPHENISCIFORMES

Family Spheniscidae (penguins)

**Inguza predemersus . — x Si. Y — | marine (mainly coastal,

**Dege hendeyi . Sancle x ? x — f cold water)

ORDER PODICIPEDIFORMES

Family Podicipedidae (grebes)

Gen. and sp. not det. . to x = x — ? fresh water

ORDER PROCELLARIIFORMES (petrels

etc.)

Gen. and at least 3 spp not

dete: : : —— x = x — marine (pelagic)

ORDER PELECANIFORMES (pelicans

etc.)

Fam., gen. and sp. not det.. — x —- — =

Family Phalacrocoracidae

(cormorants)

Gen. and at least 2 spp not

det. —; : . — x — x = marine and/or fresh

water

? Family Sulidae (gannets)

Gen. and sp. not det. . oe x — — —_ marine

ORDER CICONIIFORMES

Family Ciconidae (storks)

Gen. and sp. not det. . pee x a x — _ terrestrial (open

country and/or marsh)

PALAEOECOLOGY OF LANGEBAANWEG 49

PPM PPM PPM Inferred habitat

GM QSM 3aS_ 3aN_undiff. preference

Family Threskiornithidae (ibises,

spoonbills)

Gen. and sp. not det. .

ORDER ANSERIFORMES

Family Anatidae (ducks, geese)

Gen. and at least 4 spp not

det. : :

ORDER FALCONIFORMES

Family Falconidae (falcons)

Gen. and sp. not det. .

Family Accipitridae (hawks, eagles

etc.)

Gen. and at least 4 spp not

det ; ; c

? Gypaetinae (vultures) gen.

and sp. not det.

ORDER GALLIFORMES

Family Phasianidae (game birds)

Gen. and at least 2 spp

(francolin and quail) not

det.

ORDER GRUIFORMES

Family Gruidae (cranes)

Gen. and sp. not det. .

Family Rallidae (rails)

Gen. and sp. not det. .

Family Otidae (bustards)

ORDER CHARADRIIFORMES

(shorebirds)

Fam., gen. and at least

10 spp not det.

Family Pteroclidae (sandgrouse)

Gen. and sp. not det. .

ORDER COLUMBIFORMES

Family Columbidae (pigeons,

doves)

Gen. and at least 2 spp not

Gciamar : : ;

ORDER PSITTACIFORMES (parrots)

Fam., gen. and at least 2 spp

not det. ;

ORDER STRIGIFORMES

Family Strigidae (owls)

Gen. and at least 2 spp not

deel L : ;

terrestrial

(open country) and

fresh water

fresh water and

estuarine

aerial, arboreal and/or

terrestrial

aerial, arboreal and/or

terrestrial

terrestrial (open

country)

terrestrial (open

country with low cover)

terrestrial (open

country, marsh)

terrestrial (dense cover)

and fresh water

terrestrial (cpen,

mainly dry country)

fresh water, estuarine

and/or marine

terrestrial (near water)

terrestrial and arboreal

arboreal (forests,

woodlands)

arboreal and terrestrial

ANNALS OF THE SOUTH AFRICAN MUSEUM

PPM PPM PPM

GM QSM 3aS- 3aN_ undiff.

ORDER COLIIFORMES

Family Coliidae (colies)

Gen. and sp. not det. .

ORDER CORACIIFORMES (rollers etc.)

Fam., gen. and sp. not det. .

? Family Alcedinidae (kingfishers)

Gen. and at least 2 spp not

det.

ORDER PICIFORMES (woodpeckers

etc.)

Fam., gen. and at least 2 spp

not det.

ORDER APODIFORMES

Family Apodidae (swifts)

Gen. and sp. not det. .

ORDER PASSERIFORMES (songbirds)

Fam., gen. and at least 9 spp

not det.

AVES not det.

CLASS MAMMALIA

ORDER INSECTIVORA

Family Chrysochloridae (golden

moles)

Chrysochloris sp.

Family Soricidae (shrews)

Mysorex sp.

Suncus sp. ;

Soricidae cu and sp(p) not

deta

Family Macroscelididae (elephant

shrews)

Elephantulus sp. .

ORDER CHIROPTERA (bats)

Family Vespertilionidae

Eptesicus sp.

ORDER PRIMATES

Family Cercopithecidae (monkeys

ete»)

Gen. and sp. indet.

ORDER PHOLIDOTA (pangolins)

Phataginus sp.

ORDER TUBULIDENTATA Gani)

Gen. and sp. not det.

ORDER CARNIVORA

Family Canidae (foxes, jackals

etes)

Gen. and sp. not det. (? aff.

‘Canis’ brevirostris)

Vulpes sp. (fox) .

Family Ursidae (bears)

* Agriotherium africanum

Inferred habitat

preference

arboreal (bushland and

woodland)

arboreal and terrestrial

fresh water

arboreal and terrestrial

aerial

varied

fossorial

terrestrial

aerial

arboreal and terrestrial

terrestrial

(? woodlands)

PALAEOECOLOGY OF LANGEBAANWEG >|

PPM PPM PPM Inferred habitat

GM QSM 3aS- 3aN_undiff. preference

Family Mustelidae (weasels,

martens etc.)

Plesiogulo monspessulanus

(wolverine) . : —— x 2 — — terrestrial

(? woodlands)

*Mellivora benfieldi (honey

badger) : — = x x — terrestrial

Enhydriodon africanus fetien) = = x x = fresh water

Family Phocidae (seals)

** Homiphoca capensis. cf: x x S — marine and ? estuarine

Family Viverridae (mongooses ete, )

‘Viverra’ leakeyi (? aff.

Civettictis) (civet) — ‘ = ~ —

Viverrinae gen. and sp. not

det. (? aff. Pseudocivetta)

(civet) : ; — x x

Genetta sp. (genet) : —— x == =

Herpestes spp A, B

(mongooses) 2 — x x =

Herpestinae spp C, D, E

(mongooses) : — x —

Herpestinae not det. . — — ~ XX

Family Hyaenidae (hyaenas)

**Adcrocuta australis . —

*Ictitherium preforfex . —

*Hyaena abronia . : —

Ayaenictitherium RTE TBI —

Euryboas sp. ; : —

Hyaenidae sp. E : —

Hyaenidae not det. . ——

Family Felidae (cats)

‘Machairodus’ sp. (sabre-

tooth) g . = x =

Homotherium sp. (sabre

tooth) : —— x x cf.

Felis sp. (wildcat- like) — x — =

Felis aff. issiodorensis (lynx-

hike) ee : — x x

* Felis obscura (lynx- like) a — x =

Dinofelis diastemata (false

sabre-tooth) : oo x x

Felidae not det. . : ._ — x x

CARNIVORA not det.

Gen. and sp. not det.

(Canidae or Viverridae) — x = = =

Gen. and sp. not det.

(? Procyonidae) . oo x — — =

Gen. and sp. not det.

(? Lutrinae) : , = x — — —

ORDER PROBOSCIDEA (elephants and

kin)

Family Gomphotheriidae

Anancus sp. : : . x x — —

Family Elephantidae fers (woodlands)

Mammuthus subplanifrons . — x 2 x —

terrestrial

Sl Se Ses

| | xxx | x

Kec oe

terrestrial

)

Sy ANNALS OF THE SOUTH AFRICAN MUSEUM

PPM PPM PPM Inferred habitat

GM QSM 3aS 3aN_ undiff. preference

ORDER HYRACOIDEA (hyraxes or

dassies)

Family Procaviidae

Procavia cf. antiqua . Se x — ? — terrestrial or arboreal

ORDER PERISSODACTYLA

Family Equidae (horses)

Hipparion cf. primigenium . x — — — —

Hipparion cf. baardi . oo x x

Hipparion cf. namaquense . — — — — z terrestrial (grasslands)

Family Rhinocerotidae (rhinos)

Ceratotherium praecox a x x — —

ORDER ARTIODACTYLA

Family Tayassuidae (peccaries)

*Pecarichoerus? (or

Barberahyus) africanus . — — x x — terrestrial

Family Suidae (pigs)

Nyanzachoerus cf. pattersoni

(or kKanamensis) . — x = — — | terrestrial

Nyanzachoerus cf. jaegeri . — == x = — f (? woodlands)

Family Hippopotamidae (hippos)

Gen. and sp. not det. . ae —- x x — fresh water and

terrestrial

Family Giraffidae (giraffes)

*Sivatherium hendeyi . — x x x =

Palaeotragus cf. germaini . — — — x — Lone t (woodlands)

Giraffa sp. _— x x x —

Family Bovidae (buffaloes,

antelopes etc.)

Tragelaphus sp. A (nyala-like) — x x x =

Tragelaphus sp. B (nyala-like) — — — x —

** Mesembriportax (or Mio- terrestrial (woodlands)

tragocerus) acrae (kudu

like relative of nilgai) . — x x x —

*Simatherium demissum

(buffalo) . , — x x x — terrestrial

(? grasslands)

* Kobus subdolus (kob-like) . — == x x — \ terrestrial (woodlands

Kobus sp. B (kob-like) eS x — f near fresh water)

** Damalacra neanica

(hartebeest-like) . — = x x =

** Damalacra acalla (hartebeest- > terrestrial (grasslands)

like) ee : — x x x —

* Raphicerus paralius (steenbok) — x x x — terrestrial

Gazella sp. (gazelle) . —— x x x — f (? grasslands)

Ovibovini gen. and at least

2 spp not det. , — x x x = terrestrial

ORDER LAGOMORPHA

Family Leporidae (hares, pee

Pronalagussp. . : — x x x — terrestrial

ORDER RODENTIA

Family Bathyergidae (rodent moles)

Bathyergus sp. . : tos x x x — \ fossorial

Cryptomys sp. . = x ma x == ff

Family Hystricidae (porcupines)

Gen. and sp. not det. A oo x oa Ee ae Re ror

Gen. and sp. not det. B _— = x x = J

PALAEOECOLOGY OF LANGEBAANWEG 53

PPM PPM PPM Inferred habitat

GM QSM 3aS 3aN_ undiff. preference

Family Cricetidae (rats, mice,

gerbils etc.)

Mystromys sp. A ; _— x — — — |

Mystromys cf. darti . oo x — == — ference

Mystromys cf. hausleitneri . — x = ==

Gerbillus or Desmodillus sp. — x _— os =

Dendromus sp. . : — x — — — terrestrial

(? woodlands)

Steatomys or Malacothrix sp. — x = — —

Family Muridae (rats, mice)

Aethomys spp A,B. oo x — — — terrestrial

MussppA,B . : = x — — —

Rhabdomys sp. . : — x -= os =

** Furyotomys pelomyoide oo x — — — terrestrial (? near fresh

water)

Family Muscardinidae (dormice)

Graphiurus sp. . ; Sf x — — — terrestrial

(? woodlands)

RODENTIA not det. . ; a x x x —

ORDER CETACEA (whales, dolphins)

Gen. and spp not det. eX x x x — marine

* Sp. nov. ** sen. et Sp. NOV.

References:

Foraminifera—Tankard 19756.

Other invertebrates—Kensley 1972, 1977; Tankard 19755; unpublished.

Chondrichthyes— Hulley, in Hendey 1976a.

Osteichthyes — unpublished.

Amphibia/Reptilia — unpublished.

Aves (Spheniscidae) — Simpson 1971, 1975, 1979.

Aves (other)— Rich 1980; unpublished.

Mammalia (Insectivora, Chiroptera, Lagomorpha, Rodentia —excluding Hystricidae) — Pocock

1976.

Mammalia (Rodentia — Hystricidae)— unpublished.

Mammalia (Primates, Pholidota, Tubulidentata, Hyracoidea, Hippopotamidae, Cetacea)—

unpublished, but some details in Hendey 1976a.

Mammalia (Carnivora)—Hendey 1972a, 1974a, 1974b, 1977, 1978c, 1978d, 1980; Hendey &

Repenning 1972; Wolff et al. 1973; De Muizon & Hendey 1980.

Mammalia (Proboscidea)— Maglio & Hendey 1970; Maglio 1973; Coppens ef al. 1978.

Mammalia (Perissodactyla)—Hooijer 1972, 1976, 1978; Churcher & Richardson 1978;

Hendey 1978a.

Mammalia (Tayassuidae)— Hendey 19765; Cooke & Wilkinson 1978.

Mammalia (Suidae)—unpublished, but referred to by Cooke & Wilkinson 1978, and Harris

& White 1979.

Mammalia (Giraffidae)—Harris 1976; Churcher 1978; unpublished.

Mammalia (Bovidae)—Gentry 1974, 1978, 1980.

The observations that follow supplement the last summary statement on

the composition of the ‘E’ Quarry fauna (Hendey 1976a: 231-243).

Most of the identified invertebrates are conspecific with, or closely related

to, living forms and most have counterparts still living in the Langebaanweg

area or adjacent ocean. Two new species have been recorded amongst the

Mollusca. Marine forms predominate, but estuarine, freshwater and terrestrial

54 ANNALS OF THE SOUTH AFRICAN MUSEUM

species are also recorded. Palaeoenvironmentally significant species will be

discussed later (see p. 68).

Lower vertebrates (fish, amphibians and reptiles) are largely unstudied,

although the cartilaginous fish (sharks, skates and rays) and the terrestrial

tortoise (Chersina sp.) have received some expert attention. Other groups may

well be more diverse than the preceding list indicates but, although available

material is abundant, it is for the most part fragmentary and identification of

taxa may be difficult. With the exception of some sharks (see p. 68), the lower

vertebrates apparently represent forms which still occur in the region, or in the

adjacent ocean.

The remains of birds are also abundant, a preliminary study having

revealed that ‘E’ Quarry is the richest pre-Pleistocene fossil bird locality in the

world (Rich 1980). Although bird remains are less common than those of

mammals, and although fewer taxa are presently recognized, they may ulti-

mately prove to be the most diverse group represented in the ‘E’ Quarry

deposits. So far only the penguins have been studied in detail (Simpson 1971,

1975, 1979), and the fact that both identified taxa represent new genera

suggests that other birds may be equally distinctive.

The birds are a potentially valuable source of palaeo-environmental infor-

mation. For example, at least two species of parrots are represented, and

parrots no longer occur in the south-western Cape, being confined to wooded

tropical and sub-tropical regions further north. While this may indicate a warm,

wooded environment at Langebaanweg during the early Pliocene, the represen-

tation of marine birds, particularly the Procellariiformes, apparently indicates

temperate conditions in QSM and PPM times, which is in keeping with other

evidence (see p. 68).

However, a detailed analysis of available material is required before the

palaeoenvironmental implications of the avifauna can be assessed. Better

identification of the taxa represented is also necessary. For example, there are

several passerines that are characteristic of the south-western Cape fynbos

region today (McLachlan & Liversidge 1978), and the identification of such

forms in the ‘E’ Quarry avifauna would support the evidence which indicates

that the fynbos was being established locally during the early Pliocene (see

p. 43).

A curious feature of the avifauna is the under-representation of obvious

scavengers such as vultures and crows. Scavengers (hyaenas) are well repre-

sented amongst the mammalian carnivores, and the inferred presence of

abundant animal carcasses at the time that the deposits were laid down suggests

that avian scavengers should also have been common.

The potential importance of the birds as indicators of depositional environ-

ments was briefly discussed by Rich (1980).

Mammals are the best represented group in the ‘E’ Quarry fauna, both in

terms of the number of specimens and the number of species. Marine, fresh-

water, aerial and terrestrial species are recorded, the latter predominating. The

PALAEOECOLOGY OF LANGEBAANWEG 55)

mammals are also the most intensively studied group, although descriptive

accounts of about half the identified taxa have yet to be published. There are

apparently no extant species represented, and the described forms include four

new genera and sixteen new species.

A preliminary study of the small mammals (insectivores, bats, rodents and

hares) by Pocock (1976) revealed nothing obviously inconsistent with a late

Cenozoic fauna from this region. The most commonly represented of the small

mammals are fossorial forms (Chrysochloridae, Bathyergidae), which are today

still common in the vicinity of Langebaanweg, and elsewhere in the south-

western Cape. The small mammals are a potentially valuable source of

palaeoenvironmental information. For example, the only described rodent, a

new genus and species (Euryotomys pelomyoides) which is common in the

QSM, is a primitive otomyinine, a group which is characteristic of well-watered

environments. The presence of Dendromus and Graphiurus may be further

evidence for a wooded environment in the area at the time that the deposits

were laid down.

One primate only has been recorded from deposits in ‘E’ Quarry, and it is

known from fewer specimens than all other species. The under-representation

of this important order is discussed later (see p. 87).

The carnivores are still the most diverse of the mammalian orders repre-

sented in ‘E’ Quarry. They are known from a large number of often complete

and well-preserved specimens, a situation which is remarkable in view of the

generally poor representation of this group at other African localities of late

Miocene and Pliocene age. Since the last summary statement on them (Hendey

1976a: 234-240), further studies have been published on the bear (Hendey

1977, 1980), hyaenas (Hendey 1978c), mustelids (Hendey 1978d), and the seal

(De Muizon & Hendey 1980). However, only the accounts of the bear and

mustelids can be regarded as complete at this stage. Much material belonging

to other families is undescribed, and at least some of the recorded taxa require

reinterpretation.

For example, an incomplete study of the civet material has suggested that

two species may be represented, one being related to the living African civet

(Civettictis civetta), while the other may be related to the late Pliocene/early

Pleistocene Pseudocivetta (Fig. 17). Similarly, the material referred to the canid

(Vulpes sp.) may also represent two species, one being a generalized form

possibly related to later African Vulpes, while the second may be related to the

late Pliocene/early Pleistocene ‘Canis’ brevirostris. The difficulty in identifying

the ‘E’ Quarry civet and canid specimens is evidently due to the material

representing early stages of lineages which were only clearly differentiated later.

This applies in the cases of other ‘E’ Quarry mammals as well (see p. 82).

Although the identification of, and comments on, the ‘E’ Quarry probo-

scideans made by Coppens et al. (1978) are probably substantially correct, this

group requires further study. None of the Anancus specimens, which

apparently represent a new species, has been described, and the relationship

56 ANNALS OF THE SOUTH AFRICAN MUSEUM

between the ‘E’ and Baard’s Quarry species (Hendey 1978a) has yet to be

substantiated. An appreciable number of unstudied teeth of the elephant

(Mammuthus subplanifrons) are now available, and they may contribute

towards resolving the problems surrounding this unsatisfactory taxon.

The ‘E’ Quarry perissodactyls are palaeoenvironmentally significant since

both the rhinoceros and the Hipparion are hyposodont forms, indicating the

presence of grasslands, whereas most of the other large herbivores are wood-

land browsers. Although the study of the perissodactyls is largely complete, the

Hipparion specimens warrant further attention directed at their implications in

respect of the succession of strata in the vicinity of Langebaanweg. For

example, it is the Hipparion from the Anyskop terrestrial deposits which

suggests that they are broadly contemporaneous with deposits in Baard’s

Quarry (see p. 38). In addition, it 1s the Hipparion from the GM which

supports the suggestion that this horizon may predate the QSM and PPM by an

appreciable period (see p. 23).

The artiodactyls, which are predominantly browsers, are a_ well-

represented and diverse group, ranging in size from a very small peccary to the

giant Sivatherium. With the exception of the pigs, hippopotamus, and palaeo-

tragine, all have now been described, although additional material of most

species is available. The palaeotragine is of particular interest since, although

Africa is the home of the only living palaeotragine (Okapia johnstoni), the

post-Miocene history of this group is poorly known, and the ‘E’ Quarry species

is comparatively well represented by both cranial and postcranial material.

Other ‘E’ Quarry mammals which are unstudied are a pangolin, an

aardvark, two species of porcupine, and several cetaceans.

Taphonomy

Taphonomy ‘involves all aspects of the transference of organic remains

from the biosphere to the lithosphere, and includes both the biological and

physical factors and processes that are involved’ (Olson 1962: 134). It 1s,

therefore, a branch of palaeontology that is particularly relevant to palaeo-

ecological studies. Some information of a taphonomic nature, such as the mode

of accumulation of many of the ‘E’ Quarry fossils, has been given above, and

the summary account which follows deals with other matters.

A taphonomic study of the ‘E’ Quarry fossils such as the now classic one

on a late Tertiary vertebrate fauna from Nebraska by Voorhies (1969) has not

been undertaken. Nevertheless, almost unavoidably observations of a tapho-

nomic nature have been made in the course of fieldwork at Langebaanweg, and

subsequently also in the laboratory. Some of these observations have already

been recorded elsewhere (Hendey 1974a: 348-353; 1976a: 222-230; 1980:

53-67).

In any assessment of the taphonomy of individual fossil assemblages from

‘E” Quarry, it is important that allowance be made for collecting biases, recent

damage to bones and the post-depositional disassociation of body parts. Conse-

PALAEOECOLOGY OF LANGEBAANWEG >//

quently, the manner in which material was acquired needs to be known.

specimens were recovered from the ‘E’ Quarry sediments in the following ways:

Surface collecting.

Controlled excavation.

Uncontrolled excavation.

Screening of small sediment samples in the field or the laboratory using

mesh of 2 mm or less.

5. Screening of bulk sediment samples in the field using a double-bank of

sieves with mesh of 10 and 5 mm.

Pee

In the case of surface collecting, material had mostly been exposed by the

mining operations, in which case specimens were often out of context, damaged

and disassociated from other skeletal elements of the individual involved. A

large quantity of material was acquired in this way, with most specimens being of

medium to large size. Alternatively, specimens were exposed by wind or water

erosion, in which case their condition was often as in burial, and associations of

specimens were sometimes preserved. Once again much material was collected

in this way, and all except microscopic specimens were recovered.

Only a few controlled excavations were undertaken in ‘E’ Quarry, two in

1966 and one or two in the years 1969, 1970, 1975, and 1976. However, although

most of the fossiliferous horizons and areas of the QSM and PPM were sampled,

the size of the excavations was very variable, ranging from a few to hundreds of

cubic metres of deposit. The condition of specimens for the most part reflects

that in burial, and associations were recorded. Uncontrolled excavations differed

in that they were always on a small scale, usually being centred on a significant

fossil discovered on the surface (Fig. 13). Except for the recording methods, they

could be considered as controlled excavation when carried out by qualified

personnel, but otherwise they served only to build up sample sizes.

It was the latter factor, together with the need to recover small fossils, that

was the motivation in the screening of sediment samples. The screening of

small samples usually involved the use of hand-sieves washed in ponds on site,

or in containers of water in the laboratory. Since the deposits were mostly

unconsolidated fine- to medium-grade sands, little agitation of the sieves was

required and damage to specimens was minimal. However, associations of body

parts were usually lost.

The co-operation of the mining company made the screening of bulk

sediment samples possible. This was done when mining or other factors

threatened fossiliferous deposit, and involved the removal by mechanical means

of samples ranging in size from a few tons to hundreds of tons, and the

transportation of the sample by truck to an area where a stand with double-

banked two-man sieves was set up. Piped water was used to wash the ‘dumps’

through the sieves. This process allowed the recovery of some very large

samples of fossils, and was particularly useful in the recovery of rare taxa.

However, the process which variously involved mechanical excavators, bull-

58 ANNALS OF THE SOUTH AFRICAN MUSEUM

RP i

whiny

ie i, a

fe pe

we ee

ye,

pl Se

Fig. 13. The skull and vertebral column of a buffalo Simatherium demissum,

(SAM-PQ-L23400) in situ in the QSM.

dozers, front-end loaders, and tip-trucks, was very destructive. In addition,

there were instances where the removal process was not carefully controlled

and fossiliferous sediments of the QSM and bed 3aN were mixed.

The provenance of material was one of the recurrent problems encoun-

tered in the fossil-collecting programme. Material of doubtful provenance

includes much of that collected prior to 1969, but the problem continued even

when the nature of the succession was well known and exact sites of discovery

were recorded. Both beds 3aS and 3aN truncated the QSM, while bed 3aN

truncated bed 3aS, and it was in such situations that doubts about provenance

arose. Nevertheless, assemblages of known provenance do exist, and it is these

that will ultimately be used in the detailed analyses of fossils from particular

horizons. A problem which may be impossible to deal with is that of material

reworked from pre-existing deposits. It affects mainly bed 3aS, which includes

specimens derived from the QSM.

The depositional environments of the ‘E’ Quarry succession need not be

discussed again, but some comments on the agencies responsible for concen-

trating fossils in certain areas are necessary.

The principal agency involved was water action. In the case of the GM

fossils, concentration was by wave action, and in QSM III it was the combined

effects of the river and tides which led to the accumulation of fossils. Elsewhere

in the QSM and in beds 3aS and 3aN the river was largely responsible for

concentrating fossils deposited in a subaqueous environment, although a

subsidiary role for wave and tidal action cannot be ruled out.

PALAEOECOLOGY OF LANGEBAANWEG 59

One of the common field practices in taphonomic studies is the recording of

the orientation of in situ fossils, since the resultant data may indicate whether or

not the fossils were deposited by flowing water, and, if so, the current

direction(s). In one of the earlier controlled excavations in ‘E’ Quarry

(LBW/E/1969/1) some such recording was done, but the procedure was aban-

doned since it was found to be time-consuming and probably pointless. It seemed

obvious at that time that the deposits in question (bed 3aS) were channel-laid

and that the trend of the channel was north-east to south-west. All subsequent

observations on the bed 3aS deposits confirmed this early impression. Much the

same applied in the case of the bed 3aN deposits (see Hendey 1976a: 228-230;

1980: 57-63), although in this instance the nature and distribution of the fossils

were much more obviously indicative of a fluviatile environment.

Evidence for the subaerial accumulation of some QSM fossils has been

discussed elsewhere (Hendey 1974a: 349-353; 1976a: 223-224). This includes

the record of a vertebrate microfaunal concentration, apparently an owl pellet

accumulation, which was burnt prior to incorporation in the deposits. Microfau-

nal concentrations of this kind were not common in ‘E’ Quarry. The most

notable examples were two that occurred at the interface between the QSM

and bed 3aS. These assemblages, designated 1/1968 and 12/1968, were dis-

cussed by Hendey (1970a: 81, 86-88), and remain something of a mystery.

They are probably comprised largely of the residues of owl pellet accumulations

which were transported a short distance and redeposited by minor drainage

channels in QSM or bed 3aS times. This process would account for the

admixture of larger fossils which obviously did not feature in the diet of owls.

Two species of owls are recorded from the OSM.

Otherwise there is no evidence for animals having been ‘bone collectors’

during the period of deposition of the ‘E’ Quarry deposits. Elsewhere in the

south-western Cape concentrations of fossils in porcupine lairs and hyaena dens

have been recorded (Hendey 1974a; Klein 1975). Some of these are in rock

shelters, of which none existed in the ‘E’ Quarry area, while others are found in

what are presumed to be aardvark burrows. Such burrows would not be

expected in the largely subaqueously deposited PPM, and it is unlikely that any

existed on the floodplain of the river in QSM times. An aardvark is recorded

from the OSM, but it is an extremely rare element in the fauna. The aardvark

remains from beds 3aS and 3aN were probably washed into the area by the

river.

Large numbers of fossils were found in certain exposures of the QSM,

although not necessarily concentrated in restricted areas. For example, the

remains of small mammals and birds were abundant in the QSM I collecting

area named ‘East Stream’ (Hendey 1974a, fig. 3). The East Stream fossils

probably represent a microenvironment (floodplain grassland or thicket) in

which predation by small carnivores (mainly mongooses) occurred frequently.

Larger mammals that died there include an elephant (MVammuthus subplani-

frons) several pigs (Nyanzachoerus cf. pattersoni), a large cat (Dinofelis dias-

60 ANNALS OF THE SOUTH AFRICAN MUSEUM

temata), and hyaenas (Hyaena abronia). Some redistribution of animal remains

apparently took place when this area was inundated during floods (Hendey

19744: 351-3525 19 7]oaz 223-225).

An impression gained during field and laboratory work was that the body

part representation of vertebrates varied according to the source of material.

The only body part analyses undertaken to date are those of bird remains from

the QSM I and QSM II assemblages (Rich 1980), to which may be added the

detailed record of the number and condition of penguin bones from QSM I

given below. Many more such analyses are required before meaningful tapho-

nomic interpretations are obtained, but some general statements on body part

representation are already possible.

Complete skeletons were not found, although partial skeletons were recov-

ered from both the QSM and PPM. The most complete was that of the

holotype of Hyaena abronia from QSM I, which comprises the skull and about

eighty postcranial bones (Hendey 1974a: 103-115, 351-352). Otherwise

elements of individual skeletons tended to be completely disassociated, or

represented by only a few bones such as parts of a vertebral column or limb.

Even in the latter instances it was exceptional to find bones still articulated. An

example of a partial skeleton from the QSM in which some elements were in

articulation is that of the buffalo illustrated in Figure 13. It is one of several

medium to large mammals from the QSM which probably represent the

remains of animals killed in situ and subsequently dismembered by the primary

predator and by scavengers (Hendey 1974a: 351).

By contrast, the recently described partial skeleton of an Agriotherium

(SAM-PQ-L45062) from bed 3aN, in which all remaining skeletal elements

were disassociated (Hendey 1980: 62-63), was a more common type of occur-

rence. In this instance, the specimen was transported to its point of deposition

by the river and it might well have been complete when deposited, but suffered

subsequently from a variety of destructive post-depositional processes, includ-

ing disarticulation after loss of soft tissue, fire, pressure of overlying deposits,

and, more recently, the mining operation and collecting.

It is virtually certain that the skull of L45062 was intact at the time of

deposition and that its subsequent fragmentation was due largely to the pressure of

overlying deposits. Complete, or nearly complete, skulls are not common in the

‘E’ Quarry assemblage, and all the known specimens suffered some degree of

crushing. Restoration of these specimens was, however, facilitated by the fact that

fragments were easily recovered from their unconsolidated matrix and were

usually undistorted. Missing parts were probably lost during the recovery process

rather than earlier. One of the most complete skulls recovered is illustrated in

Figure 17, while others have been illustrated elsewhere (e.g. Hendey 1974a, figs

19-20; Gentry 1980, figs 28-29; De Muizon & Hendy 1980, figs 1-3, 5). One

instance is known where an exceptionally robust skull might have been intact in

the deposits but was damaged by the mining operation. This is the skull of the

Plesiogulo, SAM-PQ-L40042, described by Hendey (1978d: 330-336).

PALAEOECOLOGY OF LANGEBAANWEG 61

In general, the fossils from ‘E’ Quarry are well preserved, although

damage caused by post-mortem processes is not uncommon. These processes

include the mining operation and collecting methods which led to the recovery

of specimens, and they are obviously of no taphonomic significance. Neverthe-

less, it is important that they be recognized and not be confused with that

post-mortem damage which is relevant to taphonomic studies.

This damage is significant since it reveals some of the physical, chemical and

biological processes which affected animal remains from the time of death until

their discovery. From a palaeoecological point of view it is the factors operating

at the time of death and shortly thereafter that are important, since they may

reveal something of prevailing biological interactions and physical conditions.

For example, the activities of predators and scavengers may be revealed by

distinctive types of damage and thus provide evidence of trophic relationships,

while the condition of specimens may also indicate the nature of depositional

environments, which are a reflection of the prevailing physical environment.

The bed 3aN Agriotherium may again be cited as an example, since the

recently discussed post-mortem damage to the specimens (Hendey 1980: 63-66)

is typical of that encountered amongst the larger mammals from ‘E’ Quarry.

Similar and additional types of damage were discussed elsewhere (Hendey

1974a: 349-353). Table 4 lists various types of post-mortem damage evident in

specimens from ‘E’ Quarry. Only one of the more unusual types of damage is

discussed below, since the example cited has not previously been recorded.

Hendey (1974a: 353) mentioned a series of bones from the QSM which had

apparently been etched by the stomach acids of a hyaena. Such specimens are

not common, but one striking example involves the penguins from ‘E’ Quarry.

Two penguin species are recorded, namely, /nguza predemersus and Dege

hendeyi (Simpson 1971, 1975, 1979), and a high proportion of the bones in the

available assemblages have been etched, evidently by stomach acids (Fig. 14).

By contrast, this type of damage is either rare or not recorded in bones of other

birds from the same deposits.

The most likely explanation of this difference in condition is that the

penguins were preyed upon by an animal that ingested their carcasses and later

regurgitated the indigestible remnants, whereas other bird bones found in the

deposits were not ingested by predators. It is only the stouter and more durable

bones of the penguin skeleton which have been recorded (i.e. mainly limb

bones). While other bones of the skeleton are less diagnostic, and therefore less

likely to be recognized as belonging to penguins, their absence is probably due

_ to their having been destroyed by the stomach acids which damaged most of the

surviving bones. The most delicate of bones belonging to other birds are

preserved together with those of penguins, which indicates that the effects of

depositional processes can be eliminated as the cause of the condition and

representation of penguin bones.

Modern penguins are preyed upon by sharks and other large predaceous

fish, and seals (Simpson 1976), and these animals are all possible predators in

ANNALS OF THE SOUTH AFRICAN MUSEUM

“susuideds Juing Woy sqeysinsuNsip AjIpee. Ajjensn aie Kay} yng ‘y4oRIqQ Ose 2B NeE Peg ‘Wdd PUB WSO 24) JO ,spoq jead, oy) WoL sjIssoq,

(82 “SY “0861 *C ‘OLL Ssy

‘DpL6| AIPpusH) SoIOATUIeS

Aq sUIMeUS puUe sug

auo0g jJusdeipe

JO SSO] Je[NSe1IT YPM sow}

-9WOSs ‘soUOg UO pd}ej}USTIO

A[WOpue ‘SoAOO1S Ie[N3I11]

(ZI “314) (6c “8Y ‘0861 AepucH)

QLAIL] JOOSUI SNOSeYdo wy | s}UDPO! Aq SUIMeUD

$o109

-UJOY PIAOG UO poqeUdTIO S9UOg UO SdAOOI3

A}WOpues ‘SOAOOIS Je[NSoy | [o];esed ssoj 10 DIO

(€ “Sy

>DQL6| A2pucx{) uontIsodep

Jayye usuIdeds JOAO SUIMOY

1oyeM posieyo-pues ‘10d

AdpusH)

sung

(GLL “SY ‘PPLol

SOIOATUIES Kq

(t “3U ‘0861

AgpusyH =‘3'd—s]]NyS jew

-weuw Suljdoye Alpeisedso)

SUsOdep BUIA[IZAO JO

ainssoid [euOolIsodep-1sog

( “3Y ‘6L6l :S-p S8y

"¢L6| Uosduis) (p] “3!4)

SOIOAIUILS JO SPIOB YOVWOIS | (OE “SY °086] AOPUSH) B14

-SUBI] ADATI SUOIIOR SALMA

souoqg poysiq Povel]Ojxa puke PayxdeVid UdIJO| YI90}] puke soudg popriqy

pue ‘Asis s10/pue , 49e]q

Po2INO/ODSIP 41901 puke souog

soauoq peysnid

AgpuoH)

sug

(WLL “30 “PrLol ;

SOIOATUIBS Aq aGNV

souoq

suo]

(siisodap se] us °S°d) Sul

-1oy}eoM Snoonbegns jeuon

-Isodep-isod ‘fSulioyjeVoM

Jelseeqns yeuorlsodop-o1g

GNV

y}90} puke sau0q

poweyojxa pue payorig

SNOILVULSATIVT

SASNVD alavaoug

SNOILVULSONTIT

Sasnv_ ATEVdOud

ADVWAVG JO sadAL

“SIMULLQISUL]T ‘AIIENG JF, WO} Sjisso} Aq poviqiyxs asewep Wajow-}sod jo sajdwexy

bp ATV

PALAEOECOLOGY OF LANGEBAANWEG

Fig. 14. Undamaged (above) and acid-etched (below) penguin tibiotarsi, femora and

carpometacarpi from the QSM (East Stream sample).

the case of the ‘E’ Quarry penguins. In addition, penguins are preyed upon by

terrestrial animals, and ‘E’ Quarry is unique amongst recorded Tertiary pen-

guin localities in that the penguin bones come from non-marine deposits

(Simpson 1976) occurring in association with a wide variety of terrestrial

vertebrates, including predaceous forms. Consequently, it could also be signi-

ficant that ‘E’ Quarry is apparently the only known locality from which penguin

bones etched by stomach acids have been recorded. In fact, the ‘E’ Quarry

penguins are likely to have been preyed upon by several aquatic and terrestrial

predators, although only one is likely to have caused the acid-etched bones.

This was probably an animal that bolted its food, and subsequently regurgitated

indigestible residues. This behaviour applies in the case of sharks (P. A.

Hulley, pers. comm.), and of the predators recorded in association with the ‘E’

Quarry penguins, sharks are here favoured for the role.

Further clues concerning predation on the ‘E’ Quarry penguins may

emerge from more detailed examinations of the available material and compari-

sons with stomach contents of modern penguin predators. Even the most

superficial examination of the ‘E’ Quarry penguin assemblages reveals features

that may be of taphonomic significance. For example, etching is more pro-

nounced on leg bones than those of the flipper (Table 5). In addition, certain

TALULA LULL

ilo

PURE ALLAAH

112 11S 114

64 ANNALS OF THE SOUTH AFRICAN MUSEUM

TABLE 5

Limb bones of adult penguins (QSM I, East Stream sample) from ‘E’ Quarry, Langebaanweg.

Inguza predemersus Dege hendeyi

Specimens Specimens Specimens} Specimens Specimens

with little with with little with

or no pronounced Ap orno pronounced We

etching etching etched etching etching etched

Huneius ton a geese ee eG 67

eee a a 6 10 63 — 1 100

Radius ay ere 8 Si 1 — 0

Carpomencannes ey Ce ees bares 43 a — —

Tone i 19 27 59 1 1

Eo rs:

Tibiotarsus.

Tasounaintenons :

Total leg ae

TABLE 6

Etching on penguin leg bones (QSM I —East Stream sample) from ‘E’ Quarry, Langebaanweg.

Inguza predemersus Dege hendeyi

Little or Pronounced Yep Little or Pronounced Wi

no etching _ etching etched |noetching etching etched

Femur,

proximal end : 4 5 56 7 1

Femur,

distalend . : 2, i 78 72 6

Tibiotarsus,

proximal end : 1 9 90 1 1

Tibiotarsus,

distalend . . 4 8 67 l 3

Tarsometatarsus,

proximal end : — 13 100 — >

Tarsometatarsus,

distalend . 5 10 67 1 5)

PALAEOECOLOGY OF LANGEBAANWEG 65

parts of individual bones are more affected than others. For example, the

proximal ends of femora are less often etched than distal ends, while all the

proximal ends of the tarsometatarsi are etched, and, at least in the case of

Inguza predemersus, the proximal ends of the tibiotarsi are more affected than

distal ends (Table 6). Furthermore, although there is a general similarity in the

condition of the bones of the two penguin species, there are very marked

differences in the representation of individual bones. This may simply be due to

the smaller Dege hendeyi sample size, but it is striking that whereas there are

nearly as many femora of D. hendeyi (ten) as there are of J. predemersus

(twelve), other limb bones of the former are appreciably less common. This

applies particularly in the case of flipper bones (Table 7).

TABLE 7

Flipper and leg bones of penguins (QSM I—East Stream sample)

from ‘E’ Quarry, Langebaanweg.

Dege

hendeyi

Inguza

predemersus

Total flipper

Total leg

Flipper: leg

The above observations are not interpreted here, but they presumably

reflect events which followed shortly after the deaths of the birds concerned,

and thus indicate a potential value in analyses of this kind.

Certain of the ‘E’ Quarry fossils also reveal something about the condition of

animals during their lifetimes, and although this does not necessarily fall within the

scope of taphonomy, it is convenient to mention it here. The majority of

specimens show no signs of abnormality, but pathological conditions amongst

carnivore specimens have already been recorded (Hendey 1974a: 184-185; 1978d:

336; 1980: 53), while abnormalities observed amongst herbivores will be men-

tioned later (see p. 76). The latter are ascribed to dietary factors, but most of the

other recorded conditions are evidently due to old age or injury. Apart from their

intrinsic interest in the field of palaeopathology, dental and osteological abnor-

malities are significant since they may provide evidence of prevailing environmen-

tal conditions, the age structure of populations, cause of death, and behaviour

patterns. Consequently, a detailed study of relevant specimens is clearly desirable.

PALAEOENVIRONMENT

The nature of environments in the south-western Cape and adjacent

regions during the Palaeozoic, Mesozoic, and early Cenozoic are known only in

a very general way, but the record for the late Cenozoic is better (Deacon

1979). However, even for this period the record is only intermittently detailed,

66 ANNALS OF THE SOUTH AFRICAN MUSEUM

and for the late Tertiary it is the deposits in the vicinity of Langebaanweg that

provide the most substantial body of information.

The environment at Langebaanweg at the time that the Varswater Forma-

tion and other late Tertiary deposits were laid down was clearly very different

from that of the present. An account of the physical setting of the area and of

the geological history of these deposits was given earlier, and the present

section will deal with the climate and the environmental implications of the

Varswater Formation flora and fauna. It is important to note that information

presented below contributed to the revised interpretation of the geological

history of the Langebaanweg area. In other words, although these two aspects

of the present study are interrelated, the interpretation of the geological history

stemmed in part from the palaeoenvironmental study rather than the reverse,

which is implied by their order of appearance in this report.

The late Miocene and early Pliocene (i.e. between about 10 and 4 Ma) was

a period of great change over much of the earth’s surface. It was the time when

the generally moderate climatic conditions which characterized most of the

Tertiary were coming to an end, and when the most recent of the world’s ‘ice

ages’ was beginning. This ‘ice age’ has been a period of marked oscillations in

world climates, and these have had a profound effect on the biology of the

earth.

The Miocene—Pliocene boundary is defined on the basis of the marine

sedimentary succession in the Mediterranean Basin, and it is now generally

taken to date back about 5 Ma, with marine microfossils and palaeomagnetic

data having facilitated widespread correlations (e.g. Berggren & Van Couv-

ering 1974). Apart from changes in marine microfaunas, the terminal Miocene

was characterized by a world-wide marine regression, a phenomenon that was

caused by increased glaciation in Antarctica (see below), and which had

significant effects in coastal and continental shelf regions of the world.

In the Mediterranean this regression contributed to the so-called ‘Messinian

[or Mediterranean] salinity crisis’, which ‘was an interval of highly restricted

circulation reflected in extensive deposition of stagnant water and evaporative

sediments’ (Van Couvering et al. 1976: 263). This event, which lasted approxi-

mately 1,5 m.y., was concluded by ‘a return to former water depths coinciden-

tally with the beginning of the Pliocene . . . which is calibrated to t = 5 Ma

(Van Couvering et al. 1976: 263). Hsti et al. (1977: 402) have stated that the

Messinian salinity crisis is ‘such a geologically recent catastrophic event [that it]

had a great impact on the modern world, . . . on regional and global climates

and on the evolution and distribution of plants and animals’.

An example of the biological impact of this event was that it allowed the

late Miocene faunal interchange between Africa and Eurasia which will be

mentioned later (see p. 84), and hence the presence in the ‘E’ Quarry fauna of

certain ‘Eurasian’ taxa. However, the principal significance of the Messinian

event here is that it is well documented and thus illustrates the nature and

timing of certain of the global phenomena that occurred during the late

PALAEOECOLOGY OF LANGEBAANWEG 67

Miocene and early Pliocene. For example, the Messinian event was simply one

result of the increased late Miocene Antarctic glaciation and the consequent

global marine regression, which are events of far greater significance in the present

instance. Langebaanweg is situated in southern mid-latitudes, 7 500 km south of

the Mediterranean, whereas it is only 4 000 km from Antarctica, and separated

from it by an ocean that influences the climates of both regions. In addition, its

near-coastal location makes it subject to the effects of sea-level changes.

In a recent review of late Cenozoic palaeoenvironments on the west coast

of southern Africa, Tankard & Rogers (1978) related the aridity of this region

to the history of the Southern Ocean and the Antarctic ice-cap. Although there

are many conflicting statements on the timing of Antarctic ice-cap origin and

growth (see Mercer 1978), there is evidence that maximum growth was

achieved late in the Miocene (see McLachlan & McMillan 1979, fig. 2). Once

the Antarctic ice-cap had formed it ‘remained a semi-permanent feature exhi-

biting some changes of volume’, ‘the most important [being] during the latest

Miocene (t = 5 MY ago) when ice volumes increased beyond those of the

present day’ (Kennett 1978: 41). This was followed by ‘a regressive ice phase

with extensive melting and iceberg calving’ (Hayes et al. 1973: 24). The glacial

maximum led to the terminal Miocene global marine regression (cycles TM3.2,

TM3.3 of Vail & Hardenbol 1979), while the subsequent regressive ice phase

caused the early Pliocene global transgression (cycle TP1).

The fact that these events are correlated not only with the Messinian

Salinity crisis and its conclusion, but also with coastal strata elsewhere (e.g.

New Zealand—see Loutit & Kennett 1979), suggested that correlation with

Miocene-—Pliocene strata in South Africa is also feasible.

Relevant here is evidence that local sea temperatures declined during the

period of deposition of the Varswater Formation.

Present sea temperatures on the Cape west coast are low, and the area falls

within the Cold Temperate Province of the southern African coast (Brown &

Jarman 1978). However, the coast near Langebaanweg is close to the southern

boundary of this province and the western limit of the ‘overlap region’ between

the Cold and Warm Temperate Provinces (Brown & Jarman 1978: 1246). This

‘overlap region’ includes areas where the proportion of cold west coast species is

high, and others, such as False Bay, where ‘south coast species are much in

evidence’ (Brown & Jarman 1978: 1261). The northern shore of False Bay is at

present only about 130 km south of Langebaanweg, and during the past when

sea-levels were substantially higher, this bay was linked across the Cape Flats

with Table Bay, 20 km further north. Furthermore, Langebaanweg is today only

13 km inland from the ‘almost landlocked body of water comprising Langebaan

Lagoon and Saldanha Bay’, with its ‘biota of great richness, diversity and

productivity’ that includes species that ‘are characteristic of other provinces,

notably the south coast warm-temperate region’ (Brown & Jarman 1978: 1267).

Langebaanweg is thus situated in an area where the local marine fauna is

potentially subject to marked differences in composition depending upon

68 ANNALS OF THE SOUTH AFRICAN MUSEUM

prevailing climatic and environmental factors. Even today elements of two .

contrasting faunas (warm and cold temperate) are represented near by (Lange-

baan Lagoon, Saldanha Bay), or not far away (False Bay), while the local

occurrence of thermally anomalous molluscs during the late Pleistocene has

been documented by Tankard (1975d).

Molluscs of the GM are mostly of taxa found along the adjacent coast

today, but some ‘suggest water temperatures considerably warmer (about

3-5 °C) than today. These warm water molluscs include Cellana capensis, Turbo

sarmaticus, Barbatia obliquata, Ostrea atherstoni, and Striostrea margaritacea

(Tankard 1974a: 281). The oysters Ostrea and Striostrea are not known from ‘E’

Quarry exposures of the GM, but are from equivalent deposits on the near-by

farm of Sandheuwel (Tankard 1974a). Cellana capensis is today found north of

Port Alfred on the east coast (i.e. the Subtropical Province), while the other

four species occur eastwards from False Bay (i.e. the Warm Temperate and

Subtropical Provinces) (Kensley 1973; Day 1974; Brown & Jarman 1978).

In addition, although the sharks of the GM have yet to be studied in detail

(Hendey 1976a: 233), three of the species that have been recorded are today

characteristic of warmer oceanic regions. They are Carcharhinus melanopterus,

C. limbatus, and Squatina africana (Bass et al. 1973, 1975).

Indications are, therefore, that the GM was laid down during a warm

phase (i.e. warm temperate to subtropical), probably the period late in the

Miocene that preceded the onset of colder conditions towards the end of this

epoch.

Molluscs of the QSM are also mainly of taxa found along the adjacent

coast today, although one is a warmer water species. This is Pyrene albuginosa,

which presently ranges from False Bay to Natal. On the other hand, the QSM

assemblage, unlike that of the GM, also includes two essentially cold-water

species, namely, Chiton nigrovirescens and Tricolia capensis, which are largely

confined to the Cold Temperate Province of the west coast. In this instance, T.

capensis is particularly significant, since it is the most commonly occurring of

the QSM marine species (Kensley 1977).

It is, therefore, likely that local sea temperatures at the time of deposition

of the QSM were somewhat lower than those prevailing earlier when the GM

was laid down, and might have been little different from those of the present

(i.e. cold temperate). Consequently, the OSM is likely to date from the period

when the effects of Antarctic glaciation were being manifested in southern

mid-latitudes (i.e. terminal Miocene or later).

A similar, and probably contemporaneous, change from warm to cold

water is recorded on the west coast of South America. Mercer (1978: 80).

quoting W. J. Zinsmeister, noted that the ‘presence of distinctly warm-water

genera of molluscs indicates that during Middle and Late Miocene time the

coastal waters of southern Chile were warm subtropical’, and that these

‘warm-water faunas were replaced by distinctly cool temperate faunas in latest

Miocene or earliest Pliocene time’.

PALAEOECOLOGY OF LANGEBAANWEG 69

Although no identifiable molluscs are recorded from deposits overlying the

QSM (i.e. the PPM), remains of another marine species, the seal (Homiphoca

capensis), suggest that either there was a further temperature drop in the

period between deposition of beds 3aS and 3aN, or that temperatures remained

consistently cold during this period. The evidence for this concerns the greater

development of the maxillo-turbinals in the bed 3aN population of H. capensis

(De Muizon & Hendey 1980: 123). This adaptation to low sea and ambient air

temperatures may either reflect a lowering of such temperatures during the

period in question, or a delayed reaction to the prevailing cold conditions. In

the case of the latter alternative, the implication is that the cold adaptation in

H. capensis lagged behind the onset of the cold. This may be a more reasonable

supposition in the case of a large mammal than the one that would have the

adaptation immediately coincident with the temperature change.

It is worth noting in this connection that H. capensis was almost certainly

resident in the Langebaanweg area, and consequently that the cold adaptation

reflects local conditions. H. capensis was definitely not represented by occasio-

nal vagrant individuals as is the case today with certain Antarctic and sub-

Antarctic seals recorded on the Cape coast (e.g. Mirounga leonina, Lobodon

carcinophagus and Hydrurga leptonyx). The presence of the remains of very

young, probably neonate, individuals in the ‘E’ Quarry assemblage indicates

that breeding took place in the immediate vicinity. Otherwise the individuals

represented range from young to very old, and at least in beds 3aS and 3aN

they are present in large numbers, indicating large local populations at the time

of deposition.

In addition, it is unlikely that H. capensis was seasonally migratory. Its

dental and postcranial characteristics suggest it was a coastal species, unlike its

closest living relatives, the pelagic crabeater and leopard seals (Lobodon

carcinophagus and Hydrurga leptonyx) (De Muizon & Hendey 1980). These

two species are migratory, moving southwards during summer when the pack-

ice begins to break up (King 1964). Southward migration during summer was

probably characteristic of all the southern middle and high latitude migratory

seals that are now extinct, and since beds 3aS and 3aN were probably laid down

during summer (see p. 73), H. capensis evidently did not follow this practice.

Other evidence of prevailing temperatures may yet be obtained from a

study of the QSM and PPM avifaunas. There are indications that the avifaunas

also reflect cold conditions (see p. 54), but this has yet to be substantiated.

Available evidence therefore indicates a lowering of sea temperatures

during the period between deposition of the GM and QSM, and either a further

drop between deposition of beds 3aS and 3aN, or consistently low temperatures

at that time. Although there is as yet no positive evidence of what occurred

between deposition of the QSM and bed 3aS, it is unlikely that there was a

deviation from the generally cold conditions otherwise indicated, since the

QSM, bed 3aS and bed 3aN represent deposition during a single geological

episode (i.e. a marine transgression). Sea temperatures were low at this time

70 ANNALS OF THE SOUTH AFRICAN MUSEUM

apparently because it followed immediately after the terminal Miocene glacial

maximum in Antarctica. Temperatures might have risen again later in the early

Pliocene transgression (see p. 16).

Thus, while deposition of the GM probably took place before the onset of

the terminal Miocene glacial maximum, the QSM and PPM are likely to

post-date this event. This conclusion, taken in conjunction with others dis-

cussed in this report, contributed to the correlation of the ‘E’ Quarry deposits

with the late Miocene-early Pliocene events in Antarctica, the Mediterranean

Basin and elsewhere. This correlation is summarized in Table 8, which formed

the basis of the correlation of the entire late Tertiary succession in the

Langebaanweg area with the global sea-level changes recorded by Vail &

Hardenbol (1979) (Fig.3, Table 2).

To sum up, the GM is here interpreted as representing an event dating back

to a warm phase during the late Miocene, while the post-GM regression is

correlated with the world-wide lowering of sea-level which coincided with the

glacial maximum in Antarctica during the terminal Miocene (i.e. between 5,5

and 5 Ma—see Van Couvering et al.1976). Amongst other significant events, this

regression has been correlated with the climax of the Messinian salinity crisis,

and the formation of phosphate nodule beds in Australia and alsewhere in the

world (Carter 1978). The transgression during which the QSM and PPM were

laid down is in turn correlated with the regressive ice phase in Antarctica during

the early Pliocene, between 4,5 and 5 Ma (Van Couvering et al. 1976). This

transgression, which is recorded in countries as far apart as Spain and New

Zealand (Loutit & Kennett 1979), brought the Messinian salinity crisis to an end.

This interpretation of events indicates that the main exposures of the Varswater

Formation in ‘E’ Quarry (i.e. the QSM and PPM) are of the same age as the

early Pliocene deposits in the Mediterranean Basin (including those at Montpell-

ier in France), and at least part of the Opoitian Stage in New Zealand.

Correlation with other early Pliocene coastal deposits is clearly possible.

One possible inconsistency with this interpretation concerns the suggestion

that sea temperatures might have declined during the period of deposition of the

QSM and PPM. Since the early Pliocene transgression post-dates the late

Miocene glacial maximum in Antarctica, it would be expected that sea temper-

atures in southern mid-latitudes would be rising at this time. However, according

to Hayes et al. (1973: 24) ‘the waters around [Antarctica] would not necessarily

exhibit warming at that time but simply a slowdown in the rate of cooling’. At

issue here is the nature and timing of events in Antarctica during the late

Tertiary, and, according to at least some interpretations, Southern Ocean

surface temperatures did decline during the early Pliocene (Mercer 1978: 84-86,

fig.4). This is consistent with palynological evidence from the south-western

Cape that suggested to Coetzee (1978: fig.2) that the local climate changed from

‘Cool Wet’ during the late Miocene to ‘Colder Drier’ during the Pliocene.

There is abundant evidence from various parts of the world that the

Miocene and Pliocene were epochs when terrestrial environments were under-

(861) 992190) «+ (6L61) JOQUOPIEH WF TPA x

puejsseis pue (19]009

(YW 9 ‘9-8 ‘6) | S010} UOOSUOLU SUTUUI050q ) («FEIN 91949)

SUDDOIJY 93°] jeordosjqns jestdosqns 0} jesidosjqns 0} UOISSIISOI yoevog

x (QIed Ul) IAT =| o}eJOdUIN} We AA | 9} eJOdW9} WIAA | SULINP PUeISTINS

(ND)

PALAEOECOLOGY OF LANGEBAANWEG

(BW 7 “S—9 °9)

QUdDOII

[eUrUIS |,

tm 2s @ AUT

pe)

. oul

2 [eAIO}UI

a OUI TL

oo)

dOV

Spuvysseis

pue spur[poom

9) e13d Wd [|

4 ¢SOQUAJ

dy} jo

yusUIdooIANp

BUOI}S 1SIU

: Spuvysseis

pue spuevypoom

9} e10d Wd |,

ax(Jded Ul) HAT

NOILV.LAO4A

CNV

¢%S4INOZ NAT1Od

9} e19d Wd |

+ 4910 JOQ

uBy} JOLIp

pue Iaploo

‘oye1Odw9 |,

ALVAIIO

9)eJ0d UID}

plo Ajqeqoig

a}eJ0odw9}

PIOD

SAYOLV aad NAL

vds

(x€°€

‘TEL $1949)

UOISSOISOY

(41d. L 91949)

UOISSOISSUBI |,

THAAT-VAS

syisodop ON

[e11So119}

pue oulien}sy

o[yeIANy

a[eIAN]

[e10} I] OUTIL

SLNAUNNOYIANA

TIVNOILISOdaa

TVdIONId

‘soMuvegosur’y “AlIend) 4, ‘UONeWIO. In}eMSIeA IY

Q ATAV

UAGWNAWN THAVAD

(WSO)

daa NaN

GNVS dSOZLUVNO

(poyejue

-Joyipun Wdd)

ddan

HLIMOHdSOHd

TVLAT Td

SLINO.

JIHdVaADILVULs

NOILVWUOdT YALVMSUVA

72 ANNALS OF THE SOUTH AFRICAN MUSEUM

going significant change. For example, Webb (1977: 355) has recorded that

‘early in the Cenozoic Era North America was covered almost entirely by

forest’, and that ‘during the mid-Cenozoic . . . an increasing proportion of the

land opened up, forests giving way to woodland savanna, thorn forest, and

thorn scrub’, while ‘by the late Cenozoic forested areas had decreased still

further and much of the savanna was being replaced by grassland steppe and

even desert’. Webb (1977: 371) also states that ‘the last step in dismantling the

once continuous forests of North America came in the Pliocene, about five

million years ago’. He goes on to note that ‘a remarkably similar series of

changes affected the fauna of temperate South America during the same

40-million-year interval’ (Webb 1977: 355).

Kemp (1978) has recorded the environmental changes undergone in the

south-east Indian Ocean region during this period and, amongst other observa-

tions, she noted that the ‘latest Miocene was marked by an intense and sudden

chilling [which] must have caused marked precipitation decrease in much of

Australia’ (Kemp 1978: 170). These and subsequent climatic events had a

pronounced affect on the vegetation of the continent. For example, Axelrod &

Raven (1978: 112) recorded that ‘evidence in south-eastern Australia ...

[indicates] that the transition from a humid temperate rainforest to the present

dominant Eucalyptus-Acacia vegatation occurred at ~ 4.5 m.y. (Gill 1975)’.

There were comparable changes in Europe at this time. Delson (1975: 46)

has suggested that the deterioration in the European environment late in the

Miocene was determined largely by ‘relative decrease in moisture’, there being

‘a gradient from well-watered deciduous (or even evergreen) woodland in the

north, through parkland, scrub and into steppelike vegetation . . . with gallery

forests along watercourses’. The situation in southern Europe is relevant in the

present instance, since those areas bordering the Mediterranean now have a

climate similar to that of the south-western Cape, and changes undergone in

the two regions during the late Tertiary are likely to have been comparable.

The faunas of Langebaanweg and Montpellier in southern France are

broadly contemporaneous and similar in composition (see p. 89), and these

localities are similarly situated in near-coastal environments which today have a

Mediterranean type of climate. Consequently, it is possible that conclusions

reached concerning the palaeoenvironment of Montpellier early in the Pliocene

are relevant in the case of Langebaanweg.

Delson (1975: 47), quoting Lobreau-Callen & Suc (1972), states that

pollens from Montpellier indicate a ‘monsoon/dry season climate’, ‘closely

analogous to that of ... North Vietnam’. Schulze & McGee (1978, table 1),

following Képpen & Geiger (1936), have monsoon climate characterized by

‘mean temperatures above 18 °C for all months’, with ‘forest-vegetation despite

[a] dry season’.

There is evidence to suggest that these conditions might have prevailed in

the Langebaanweg area during the period of deposition of the middle to late

Miocene elements in the succession (i.e. the GM and the upper levels of the

PALAEOECOLOGY OF LANGEBAANWEG 73

pre-GM deposits), but that by the early Pliocene (i.e. QSM, PPM) they were

changing, with the present climate and vegetation patterns being in the process

of development.

The predominance of browsers amongst the QSM and PPM herbivores

indicates that forests or woodlands existed in the area during the early

Pliocene, probably having been present locally for an appreciable period. This

is confirmed by palynological evidence from the upper levels of the pre-GM

deposits, which also indicates that the forests of the time included tropical

elements such as palms (J.A. Coetzee, pers. comm.). The local vegetation

during at least a part of the middle to late Miocene might therefore have been

of a monsoon-forest type. The presence of grazers in the QSM and PPM faunas

indicates that by the early Pliocene grasslands had developed locally, while

there is palynological evidence for fynbos vegetation types making an appear-

ance (see p. 43).

At the time of deposition of the QSM and PPM, rainfall was strongly

seasonal, with the wet season probably being summer. However, the rainfall

regime was then apparently changing, and the summer-wet/winter-dry pattern

was probably more characteristic of the late Miocene than the early Pliocene.

This possiblity, together with the likely vegetation, suggests the monsoonal

combination of forest/dry season quoted above.

It has yet to be certainly established when the summer-dry/winter-wet

rainfall pattern was first established in the south-western Cape, but there can

be little doubt that the QSM and PPM date from a period when the transition

was taking place; or was about to take place. According to Taylor (1978: 75)

‘dry summers are of recent origin in southern Africa and probably only

appeared at the beginning of the Pleistocene’. Axelrod & Raven (1978: 112)

are more cautious in stating the ‘there is no evidence that the summer-dry

(mediterranean) climates in southern Africa existed before the formation of a

major ice sheet on Antarctica (~ 5 m.y.)’. Although the fynbos vegetation is

today centred on the south-western Cape winter-rainfall region, it evidently had

its origins under summer-rainfall conditions (Levyns 1964; Axelrod & Raven

1978; Taylor 1978). Consequently, the fact that fynbos was being established in

QSM times does not necessarily mean that the summer-dry/winter-wet pattern

already existed then.

The situation in respect of prevailing temperatures during the Miocene and

early Pliocene is less complex. It was suggested earlier that temperate condi-

tions prevailed in QSM and PPM times, and conditions then may have been

little different from those of today. The present mean annual temperature in

the south-western Cape is about 17 °C, with a mean annual range of about 8 °C

near the coast (Fuggle & Ashton 1979). The region is thus cooler than those

with monsoon climates, but the temperature difference is not great. During

warmer phases such as that during the Miocene when the GM and upper levels

of the pre-GM deposits were laid down, temperatures characteristic of mon-

soon climates might well have prevailed.

74 ANNALS OF THE SOUTH AFRICAN MUSEUM

Langebaanweg during the Miocene, like Montpellier during the early

Pliocene, might therefore have had a monsoon climate and vegetation. The

essential difference between these two localities was in the timing of a drop in

temperature and its consequent affect on rainfall and vegetation. The glaciation

of the Arctic lagged behind the glaciation of Antarctica, with the expansion of

the Arctic ice-cap and sea-level ice dated to 3 Ma (Berggren & Van Couvering

1974). Thus, while conditions at Montpellier during the early Pliocene were still

‘tropical’, at Langebaanweg they had already moderated to ‘temperate’.

In his account of the palaeoenvironments of the south-western Cape and

adjacent regions, Deacon (1979, table 1) recorded that the Miocene was

characterized by an ‘alternation between dominance by tropical palm and

temperate vegetation’, while during the Pliocene ‘ancestral fynbos communi-

ties’ were established. The evidence for these changes comes from fossil pollens

studied by Coetzee (1978), who found that there was a warm phase in the

region during the late Miocene when the vegetation was palm-dominated.

Coetzee dated this phase to 8 Ma, but since it immediately preceded the

terminal Miocene temperate phase associated with the Antarctic glacial maxi-

mum, it could have lasted to about 6 Ma. Palms and a casuarina apparently

persisted through the terminal Miocene cool phase (Coetzee’s Pollen Zone

Lvi), into the Pliocene (Pollen Zone Lvii), when species typical of the fynbos

make their first strong appearance. Consequently, it is possible that even

though temperatures were no longer favourable, vestiges of a pre-existing

monsoonal vegetation were still in evidence at Langebaanweg when the QSM

and PPM were being deposited.

Axelrod & Raven (1978: fig. 6C), following Greenway (1970), indicate

that during the latter half of the Miocene (i.e. c. 15-7 Ma) the south-western

Cape was an area of ‘subtropical forest’ bounded a little to the north of

Langebaanweg by ‘sclerophyll vegetation’. During the period of environmental

transition that followed, the forest was replaced by the sclerophyll vegetation.

The displacement of the forests was evidently a gradual process which was still

under way at the time that the QSM and PPM were being laid down.

It seems then that the main fossiliferous deposits of ‘E’ Quarry date from a

period of transition, when climate, vegetation and fauna were in the process of

change. These sometimes astonishingly rich fossiliferous deposits represent a

period when a previously hospitable environment was undergoing a marked

change for the worse. It was colder and drier, perhaps with rainfall tending

towards a winter rather than summer peak, but in any case with a pronounced

dry season; the vegetation was changing from forests or woodlands to more open

types (fynbos and grassland), while there was a corresponding change amongst

the terrestrial vertebrates, with open-country forms making a strong appearance.

The fynbos of the south-western Cape, in contrast to some other vegeta-

tion regions in Africa, is a less favourable habitat for larger mammals, and

although the situation was to vary in response to changing environmental

conditions during the Pleistocene (Hendey 1974a; Klein 1980), the environmen-

PALAEOECOLOGY OF LANGEBAANWEG TD

tal deterioration early in the Pliocene brought to an end the period when

mammals perhaps enjoyed their greatest success in this region. The Varswater

Formation provides the final and, indeed, only local evidence of this period.

The inferred habitat preferences of many of the ‘E’ Quarry taxa were given

earlier (see pp. 45-53). The habitats of invertebrates tend to be very specific, and

the ‘E’ Quarry species were invaluable in determining the depositional environ-

ments of the horizons in which they occur. As a general rule, vertebrates are less

useful in this repect, although they are usually indicative of the major habitat

types, namely, marine, fresh water and terrestrial. For example, representatives

of the marine environment include sharks, certain birds, the seal, and cetaceans;

freshwater taxa include the otter and the hippopotamus. The majority of the

vertebrates are, however, terrestrial forms, and it is this group that will be dealt

with here, since some provide an additional insight into the nature of the

environment at the time that the Varswater Formation was laid down.

Amongst the herbivorous mammals both woodland species (browsers) and

grassland species (grazers) occur. While the former predominate in terms of the

number of species, grazers were more commonly represented at certain levels

in the succession, or areas within the mine, evidently because of the tapho-

nomic factors in operation. For example the alcelaphines (Damalacra neanica

and D. acalla), which were undoubtedly grazers although relatively primitive

representatives of their group, were the most commonly occurring of the larger

mammals in the bed 3aS channel deposits of the PPM.

Before analysing the significance of faunal representation at any given

level or area, it would be desirable if minimum numbers of individuals were

determined. This has yet to be done because in many instances assemblages are

very large and the sorting of body parts and identification of taxa represented is

incomplete. Nevertheless, some patterns of representation are already obvious.

It was stated elsewhere (Hendey 1980: 56) that the remains of certain

woodland species were incorporated in the Varswater Formation with increas-

ing frequency as deposition progressed. The example cited was that of the

giraffids, Sivatherium, Palaeotragus, and Giraffa. Conversely, the grazing rhi-

noceros Ceratotherium is represented in decreasing numbers.

This was interpreted as indicating that either woodlands were an increas-

ingly widespread habitat in the vicinity, or that taphonomic factors were such

that the remains of large woodland species were incorporated in the deposits in

increasing numbers, while those of a large grazer correspondingly decreased.

The former alternative is less likely in view of the representation of other

grazers. These include the alcelaphines, Damalacra neanica and D. acalla,

which are common in the higher levels (beds 3aS and 3aN), but very rare lower

down (QSM). If grasslands were, indeed, giving way to woodlands, then the

alcelaphines would also have been represented in diminishing numbers.

It may, in fact, have been the reverse situation that led to the observed

representation of woodland-grassland species, that is, a diminution of wood-

lands in the area. The deteriorating local environment probably caused wood-

76 ANNALS OF THE SOUTH AFRICAN MUSEUM

lands to be increasingly confined to the immediate vicinity of the river, thus

restricting the habitat available to browsing species. Such animals would then

have been increasingly hard-pressed in exploiting the diminishing food resour-

ces, with an increase in mortality caused by starvation being likely. In addition,

it is possible that as the rainfall pattern was changing from a summer to winter

maximum, it became more irregular and perhaps even diminished. Droughts

might, therefore, have been an added hardship confronting the fauna.

There is evidence that at least some of the ‘E’ Quarry browsers were

ill-adapted to prevailing conditions by bed 3aN times. For example, teeth of

Sivatherium from this level sometimes exhibit hypoplasia of the enamel (fig.15)

iii PUA LLL LLLP LLL i

Fig. 15. A-C. Incisors of Sivatherium hendeyi from the PPM 3aN (unnumbered

Dump 10 specimens), with hypoplasia of the enamel. D. A healthy specimen.

This condition results from disturbances in the formation of the enamel matrix

during development of the teeth, and is a positive indication of ill-health in the

immature individual (Scott & Symons 1974). The aetiology in the case of the

bed 3aN Sivatherium is not known, but a nutritional inadequacy is most likely.

Such inadequacies could be caused by reduced food resources.

A second dental abnormality observed amongst bed 3aN browsers, is

irregular and excessive wear of teeth. Although this was observed in several of

the bovid species, it is most common in the teeth of the reduncines and/or

tragelaphines. There are problems in distinguishing the teeth of these animals

(see below), but it is virtually certain that the abnormal bed 3aN specimens

belong to the reduncines (mostly Kobus subdolus, but some Kobus sp. B), the

PALAEOECOLOGY OF LANGEBAANWEG Fig)

species most commonly represented by easily distinguishable horn-cores.

Abnormal wear on the teeth (Fig. 16) is reminiscent of that which is often

found in zoo animals that have been provided with inappropriate food. For

example, abrasive food such as hay, which is suitable only for hypsodont

grazers, causes excessive and irregular wear on the teeth of browsers. Living

reduncines are grazers, but the ‘E’ Quarry species have teeth resembling those

of the browsing tragelaphines (Gentry 1980: 255-256). The implication is that

the bed 3aN reduncines were browsers that were including in their diet an

abrasive food (probably grasses) to which they were ill-adapted.

Indications are, therefore, that by bed 3aN times at least some, and perhaps

all, the browsing herbivores in the fauna were under stress, with mortality

probably being higher than usual. Thus, it could have been the diminution rather

than an increase in woodland habitats that led to the increased number of

browsers being incorporated into the accumulating ‘E’ Quarry deposits.

It has previously been postulated that rainfall at that time was strongly

seasonal and that the fossiliferous channel deposits of beds 3aS and 3aN

represent flood-season accumulations (e.g. Hendey 1980: 60-62). Conversely,

the subaerially accumulated fossils of the QSM accumulated during the dry

season (Hendey 1976a: 223-225). These conclusions are based on the nature of

the deposits and the incorporated fossils.

For example, burnt bone is a not uncommon element of the QSM and

PPM assemblages, indicating that vegetation and/or peat fires were a feature of

the area at the time of their deposition (Hendey 1980: 66-67). Since hominid

activity can be excluded as a possible cause of fires, as can volcanic activity and

HHI

nhl

9\1

Zi1

—

Fig. 16. Reduncinae upper and lower molars from the PPM 3aN (unnumbered

Dump 10 specimens), showing abnormal and excessive wear.

78 ANNALS OF THE SOUTH AFRICAN MUSEUM

rock falls, it is most likely that fires were started by lightning. Such fires could

be expected only if a pronounced, and perhaps prolonged, dry season had left

the vegetation and peat deposits desiccated and inflammable. These fires must

have placed the fauna under additional stress. Firstly, animals trapped in the

fires would be injured or killed. This would apply especially in the case of

slow-moving forms, such as tortoises, and this could account for the large

numbers of tortoises found in the QSM, with its component of fossils accumu-

lated subaerially during the dry season (see above). Secondly, fires would have

further reduced the food supply available to herbivores, aggravating a starva-

tion problem caused by a deteriorating environment, and the drought which

preceded the first thunderstorms of the rainy season.

Lightning as the cause of fires is itself of palaeoenvironmental interest.

Under existing climatic conditions thunderstorms occur in the south-western

Cape only rarely (‘nearly five occasions per year—Schulze 1965: 313), and

although lightning is known to cause fires in this region, such fires are infrequent,

especially in the coastal areas (Kruger 1979). On the other hand, since

thunderstorms are a characteristic accompaniment of strong solar heating,

‘lightning is about ten times more frequent’ in summer-rainfall regions (Kruger

1979, quoting Kroninger 1978). This may be further evidence for summer rather

than winter rainfall in the Langebaanweg area during the early Pliocene.

Another possible effect of fires and the consequent reduction in vegetation

cover at this time was the exposure of the unconsolidated sediments of the

region to erosion at the onset of the wet season. This would account for the

large volumes of sediment carried by the river and deposited at its mouth, so

building up the Varswater Formation. In addition, there was probably an

increased rainfall run-off from the devegetated areas and, consequently, an

increase in the volume and duration of flooding. The devastating effects of the

floods would, therefore, have been intensified.

There is now some additional evidence to support the hypothesis of

seasonal deposition in the PPM during flood periods. A preliminary analysis of

a sample of Sivatherium teeth from bed 3aN suggests that it represents a

‘catastrophic’ rather than ‘atritional’ mortality (see Voorhies 1969: 46, pl. 13

(figs 1-2)). There is also a suggestion that the material in this sample was

accumulated during a restricted period of the year. Since it is abundantly clear

that the bed 3aN Sivatherium specimens were deposited in a river channel, it is

almost certain that they reflect a flood-season accumulation.

The situation that resulted in many of the fossils being incorporated in the

QSM and PPM is summed up below.

During the dry season terrestrial vertebrates would tend to concentrate in

the immediate vicinity of the river, including the now exposed QSM floodplain,

where fresh water in the generally sandy region was probably most readily

available. The area adjacent to the floodplain was probably the one to which

woodlands were largely confined, and was therefore the main habitat available

to browsing herbivores. Towards the end of each dry season the shortage of

PALAEOECOLOGY OF LANGEBAANWEG 79

food for herbivores might have become acute, with fires caused by electrical

storms further reducing available supplies. The fires probably killed many

animals, and might have caused others to drown when they sought refuge in the

river. Most affected would have been slow-moving non-swimmers such as

tortoises, which are abundantly represented in the QSM.

The first heavy rains of the wet season would cause flooding of the river.

If the rains fell inland and not locally, the flooding may have caught animals

concentrated on the floodplain unawares and swept them downstream to be

deposited at its mouth, which was then in the vicinity of the present ‘E’ Quarry.

Animals weakened by starvation or injured by fires were less likely to escape

the flooding than healthy animals. In addition to animals drowned in the flood,

the river would have carried with it remains of the animals that had previously

died on the floodplain. The arrival at the river mouth of large numbers of

carcasses of terrestrial vertebrates must have attracted aquatic carnivores such

as seals and sharks to the area, while carcasses stranded on the river-banks

would have attracted terrestrial scavengers. Those that died had their remains

added to the accumulating deposits.

The representation of animals in the QSM and PPM, the large numbers of

individuals involved, and the condition of their remains are thus indicative of

the deteriorating environment in the region during the early Pliocene. A

comparison between the mammals represented then and in modern times

indicates very clearly that the fauna of the region underwent a radical change in

the intervening period. Evidently this change had already been initiated by the

early Pliocene.

The period of deposition of the Varswater Formation was one of transi-

tion, with the local physiography, climate, vegetation and fauna all being in the

process of change. Patterns established earlier during the Miocene were still in

evidence, but those that were to characterize the Quaternary were already

being established.

EVOLUTIONARY ASPECTS OF THE MAMMALIAN FAUNA

Dating as it does from the period near the Miocene—Pliocene boundary, it

is to be expected that the fauna of the Varswater Formation would have

characteristics of both the time that Kurtén (1971: 152) has termed the ‘climax

of the Age of Mammals’ (i.e. the late Miocene), and the Quaternary, which is

characterized by similarly spectacular faunas only in some tropical regions, and

with progressively less diverse faunas in higher latitudes. The ‘E’ Quarry fauna

is, indeed, of a transitional nature and, at least in terms of its mammals, it

provides a record unparalleled in Africa of the final flowering of Tertiary life

before the climatic instability and extremes of the Quaternary so drastically

altered the character and composition of the continent’s fauna.

The ‘E’ Quarry carnivores serve admirably to illustrate the transitional

nature of the fauna. They include representatives of genera such as Agriother-

80 ANNALS OF THE SOUTH AFRICAN MUSEUM

ium, Plesiogulo and Enhydriodon, which were characteristic of the late Tertiary

over wide areas of Africa, Eurasia and North America. Similarly, the hyaenas

have previously been cited as an example of a group that had maintained a

pattern of representation established during the late Miocene, and which

contrasted with that during the Quaternary when the family was less diverse

and individual species were more highly specialized (Hendey 1978c). The larger

felids from ‘E’ Quarry are all sabre-toothed forms, and although such forms

had a long subsequent history, they were never again to predominate as they

did during the late Tertiary. By contrast, the mongooses of the early Pliocene

were then only in the early stages of their radiation, which was to reach a

climax during the Quaternary (Hendey 1974b).

Perhaps the best single example of a species intermediate between genera-

lized Miocene forms and highly specialized Quaternary ones was not a terres-

trial species but the seal, Homiphoca capensis (Hendey & Repenning 1972; De

Muizon & Hendey 1980). In this instance the intermediate character could be

fully documented, since H. capensis is very well represented, and there are

closely related forms still living which are both generalized (monk seals) and

specialized (Antarctic seals).

The ‘E’ Quarry carnivores also include early ancestors of species that were

to become characteristic of later carnivore faunas. For example, Hyaena

abronia and Mellivora benfieldi are likely ancestors of H. hyaena and M.

capensis respectively (Hendey 1978c, 1978d). It is, however, equally significant

that ancestors of several important elements of the Quaternary carnivore fauna

of Africa are not represented at ‘E’ Quarry. These elements are the lion

(Panthera leo), leopard (P. pardus), cheetah (Acinonyx jubatus), spotted

hyaena (Crocuta crocuta), hunting-dog (Lycaon pictus), and jackals (Canis

spp). They are largely open-country carnivores, with the cursorial cheetah

being a particularly good example of an animal evolved to meet the changing

environmental conditions of the late Tertiary. The cheetah has no counterparts

amongst the ‘E’ Quarry felids, although the fauna does include an early

‘hunting-hyaena’ (Euryboas), which was an ecological vicar of the cheetah

during the late Pliocene and early Pleistocene.

While the evolution of specialized carnivores such as the lion, leopard,

cheetah, spotted hyaena, and hunting-dog can be ascribed ultimately to the

changing environment of the late Tertiary, the reasons for the subsequent

success of generalized forms such as the Jackals are less obvious. The ecological

vicars of the jackals in the ‘E’ Quarry fauna were the civets (Hendey 1974a). At

least in terms of their size and dentitions there was little difference between

jackals and civets and it can be assumed that they had similar feeding-habits.

However, the early Pliocene civets had relatively small brains (Fig. 17), and it

is possible that in terms of sight, smell, and/or hearing they were inferior to

jackals and were thus unable to compete with them successfully. The living

African civet that survived this competition did so by becoming a nocturnal

omnivore and coincidentally increasing its brain size.

PALAEOECOLOGY OF LANGEBAANWEG 81

Wa a NUNN LN EL TTL ELAR VLE VELL ELE VALERA LEL LLL ELH LL LLL LL i ni

Fig. 17. Dorsal, lateral and ventral views of a civet skull (SAM-—PQ-L51590) from the

PPM 3aN.

82 ANNALS OF THE SOUTH AFRICAN MUSEUM

In the case of the marine vertebrates such as the seal, the changing

characteristics manifested in late Tertiary lineages can also be ascribed to

environmental factors, principally falling sea temperatures. The development of

the nutrient-rich cold upwelling in the Benguela Current System resulted not

only in an adaptation in Homiphoca to colder conditions, but also led to

adaptations in the postcranial skeleton, which permitted the more pelagic way

of life required to exploit the rich off-shore food resources, and those parts of

the skull connected with feeding in order to deal with the pelagic foods.

The fact that the ‘E’ Quarry mammals date from a period when the fauna

was adapting to changing conditions has contributed to the difficulties in

identifying certain taxa. In such instances the differentiation of new lineages

was in its early stages, as in the case of the civets and canids (see p. 55), and

consequently the distinguishing characteristics, which were later to become

unequivocal, were still only slhght. Apart from the examples cited above, forms

that resemble one another in many respects are included amongst the mon-

gooses (Hendey 1974a, 19746), hyaenas (Hendey 1978c), and bovids (Gentry

1980).

The diminution of forested or wooded environments was probably the

most significant factor in bringing about change amongst the terrestrial mam-

mals. This change was manifested by extinctions, by reducing the distribution

of species, and by the evolution of new taxa that were better adapted to the

changed environment. Large herbivorous species must have been most

affected, while the viability of carnivores which depended on them as a source

of food would have been placed in jeopardy. The large bear, Agriotherium,

falls into the latter category (Hendey 1980; 69-72). Ultimately though, the

entire fauna must have been affected to some extent.

A herbivore that became extinct without issue was the woodland bos-

elaphine, Mesembriportax (or Miotragocerus) acrae. During the late Miocene,

near relatives of this bovid had been amongst the most successful and wide-

spread of their family in Eurasia, and perhaps also Africa. Apart from the

living four-horned antelope (Tetracerus quadricornis) and nilgai (Boselaphus

tragocamelus) of India, the ‘E’ Quarry boselaphine is probably the last

recorded survivor of its group.

Much the same applies in the case of the okapi-like Palaeotragus cf.

germaini. Palaeotragines were an important element in the late Miocene faunas

of Eurasia and Africa, but the sole survivor of the group today is the rare

Okapia johnstoni, which is confined to forests in equatorial Africa (Churcher

1978). The Pliocene and Pleistocene history of the Palaeotraginae is poorly

documented, or perhaps misinterpreted, but ‘E’ Quarry is one relatively late

occurrence where its presence is indisputable. It is, however, a rare element in

the fauna, and by the early Pliocene the heyday of the subfamily was clearly

past.

Sivatherium hendeyi is the most commonly occurring of the ‘E’ Quarry

giraffids and, although the genus survived well into the Pleistocene in Africa, it

PALAEOECOLOGY OF LANGEBAANWEG 83

is nowhere as commonly represented as it is in ‘E’ Quarry. The group to which

it belonged flourished in Eurasia, and perhaps also Africa, during the late

Miocene, and ‘E” Quarry may provide a record of its final flowering.

There are several instances where descendants of taxa recorded from ‘E’

Quarry, or later related forms, had adapted to changing conditions through

altering their dietary preferences. Adaptations to grazing rather than browsing

were already evident in the ‘E’ Quarry rhinoceros (Ceratotherium praecox) and

alcelaphines (Damalacra acalla and D. neanica), and perhaps also the buffalo

(Simatherium demissum), gazelle (Gazella sp.), and neotragine (Raphicerus

paralius). Notochoerus, the warthog-like descendant of Nyanzachoerus (Cooke

& Wilkinson 1978; Harris & White 1979), was subsequently to follow suit. The

‘E’ Quarry grazers were still relatively primitive members of their groups, and

from the stocks that they represent were to arise the characteristic grazing

species which flourished in Africa during the Quaternary.

Although it was vegetational change that had the major affect on the

composition of the fauna, the related factors of rainfall and temperature must

also have played a role. There can be little doubt that the ‘E’ Quarry fauna had

its origins under tropical rather than temperate conditions, and, consequently,

that adaptation to cooler conditions in the terminal Miocene and early Pliocene

was required. Certain of the ‘E’ Quarry taxa, both mammals and non-

mammals, have modern counterparts only in the warmer parts of Africa, and

this suggests that ultimately they did not successfully adapt to the temperate

conditions that prevailed in the south-western Cape after the late Miocene.

Such taxa include a snail (Ceratophallus natalensis), the parrots (gen. and spp

not determined), the giraffe (Giraffa sp.), the ‘okapi (Palaeotragus cf.

germaini), and the reduncines (Kobus spp). It is possible that even by the early

Pliocene, falling temperatures had influenced the composition of the fauna,

with some tropical forms having already become extinct locally (see p. 89).

The effects of a change in rainfall are not obvious. If there was indeed a

change from summer to winter rainfall, this would not necessarily have had an

advserse effect on the mammals, although there might have been a reaction to

a lowering of rainfall and/or a lengthening of the dry season.

In assessing evolutionary aspects of the ‘E’ Quarry fauna, account must be

taken of events elsewhere, since this fauna was as much influenced by them as

it was by local environmental changes.

In the Old World the record of terrestrial life during the late Miocene is

well documented in parts of Europe and Asia, but in Africa the fossil record for

this period is poor. It is, however, safe to assume from the little that is known,

and from earlier and later records, that Africa during the late Miocene was as

richly endowed faunally as Eurasia. Its fauna was comprised of both endemic

lineages and ones that had their origins in Eurasia and North America.

After a long period early in the Tertiary when it was isolated from other

continents by seas and oceans, Africa—Arabia was joined to south-western Asia

early in the Miocene (Berggren & Van Couvering 1974). This event had a

84 ANNALS OF THE SOUTH AFRICAN MUSEUM

profound impact on the faunas of both Africa and Eurasia. Whereas earlier in

the Tertiary faunal interchange had been limited to ‘sweepstakes routes’ (sensu

Simpson 1967), or ‘Noah’s Arks’ (sensu McKenna 1973), the early Miocene saw

the opening of ‘corridor routes’ which allowed the first free faunal interchange

in perhaps 40 million years. African emigrants to Eurasia included primates and

proboscideans, while an array of Eurasian taxa, including carnivores and

perissodactyls, made their appearance in Africa (Maglio 1978).

Thereafter, the corridors between Eurasia and Africa were restricted or

closed intermittently, and Africa’s faunal history was punctuated by periods of

isolation which were interrupted by new periods of immigration. In spite of the

closer connection with Eurasia, immigrant lineages to Africa evolved taxa that

were usually in some respects distinct from their Eurasian counterparts, and

they, together with representatives of endemic lineages, gave the African fauna

a largely distinctive character throughout the later Cenozoic.

The ‘E’ Quarry fauna, and others of late Miocene and early Pliocene age

elsewhere in Africa, provide evidence for a period of immigration from Eurasia

during the late Miocene. Conversely, there is evidence from Eurasia of the

immigration of African mammals at this time. For example, Azzaroli (1975: 69)

concluded that it is ‘firmly established that interchange of terrestrial mammals

between southern and northern shores of the Mediterranean took place some

time around 6,5 m.y. ago or shortly later’.

Amongst those mammals from ‘E’ Quarry with a late Miocene Eurasian

connection is the bear, Agriotherium africanum (Hendey 1980). A slightly

earlier member of the lineage which includes A. africanum is recorded from c.

6 Ma deposits at Sahabi in Libya (Boaz et al. 1979), and it is likely that the

earliest African representative of the group entered the continent about that

time or slightly earlier, that is, during the c. 6,5 Ma faunal interchange

recorded by Azzaroli (1975). Other likely immigrants to Africa at this time are

the hyaenids, /[ctitherium, Palhyaena (=Hyaena), and Hyaenictitherium; the

wolverine, Plesiogulo; and the false sabre-toothed cat, Dinofelis, the latter

being a descendant of Eurasian Metailurus (see Hendey 1974a, 1978c, 1978d,

1980). The largest of the ‘E’ Quarry hyaenids, ‘Adcrocuta’ australis, was

previously thought to be included in this category, having been descended from

Eurasian Adcrocuta eximia, but it now seems likely that these two species had a

common ancestor in an earlier African percrocuta.

Immigrants amongst the herbivores are less obvious. One possiblility is the

boselaphine, Mesembriportax (or Miotragocerus) acrae. According to one inter-

pretation, this species was descended from the widespread Miotragocerus of the

Eurasian late Miocene, again with an intermediate form being recorded from

Sahabi (Thomas 1979). On the other hand, it may represent an endemic

African lineage that had its origins with the middle Miocene Protragocerus, and

which evolved in parallel with Eurasian Miotragocerus (Gentry 1974, 1980).

It is significant that descendants of late Miocene immigrants are most

obvious amongst the larger Carnivora. Such animals are less affected than

PALAEOECOLOGY OF LANGEBAANWEG 85

herbivores by environmental factors, and are thus capable of more rapid

dispersal, while individual species tend to have wider distributions. Conse-

quently, descendants of late Miocene immigrant carnivores to Africa might

have become established in the far south of the continent sooner than their

herbivorous counterparts. The more slowly dispersed herbivores are also likely

to undergo more rapid evolutionary changes as they adapted to new environ-

mental conditions, and their relationships to pre-existing forms may therefore

be more difficult to determine.

Not all the ancestors of Eurasian immigrants represented in the ‘E’ Quarry

fauna reached Africa late in the Miocene. One obvious exception is Hipparion,

whose appearance in Africa dates back about 12,5 Ma (Churcher & Richardson

1978). By the end of the Miocene endemic African species of Hipparion had

evolved (Churcher & Richardson 1978). The Langebaanweg Hipparion is of

particular interest, since it is recorded from all the main fossil-bearing horizons

of the Varswater Formation, as well as from younger deposits. At least one

lineage (H. cf. baardi-H. baardi) is represented which is distinct not only from

its counterparts in Eurasia, but also those elsewhere in Africa. The same

apparently applies in the case of the two alcelaphines from ‘E’ Quarry, one of

which has a descendant form in Baard’s Quarry, although in this instance the

origins of the group are still obscure (Gentry 1980).

Another of the ‘E’ Quarry species that might have been descended from an

Eurasian middle Miocene species is the peccary, Pecarichoerus? africanus.

Previously a connection with the Asian Pecarichoerus orientalis was suggested

(Hendey 19766), but a recently described peccary from the late Vindobonian of

Spain, Barberahyus castellensis (Golpe-Posse 1977), is remarkably similar to the

‘E” Quarry species and they may have a close phylogenetic relationship, with

descendants of the Spanish form having entered Africa at the same time as

Hipparion.

In interpreting the evolutionary history of Varswater Formation mammals

it is therefore necessary to take into account the following possibilities:

1. Species represent lineages endemic to Africa.

2. Species represent lineages of Eurasiatic origin, and with a long history

in Africa.

3. Species represent lineages of Eurasiatic origin, and with only a short

history in Africa.

4. Species represent lineages of African or Eurasian origin, but which are

endemic to the southerly parts of the continent.

In addition, account must be taken of the possibilities that certain lineages

represented elsewhere in Africa failed to make an appearance in the Lange-

baanweg area, only arrived there long after they were established further north,

and became extinct there earlier, or survived later, than they did further north.

The matter of locally endemic species, non-arrivals, late arrivals, early

extinctions, and late ‘survivors during any given period depends on local

86 ANNALS OF THE SOUTH AFRICAN MUSEUM

environmental conditions, as well as the degree to which the south-western

Cape, or southern Africa, was zoogeographically isolated from regions further

north. The potential for isolation of the south-western Cape, largely by a

barrier of aridity or semi-aridity, has already been established (Hendey 1974a;

14-19), although the degree of isolation at various times in the past is difficult

or impossible to determine. This applies particularly to pre-Quaternary times,

when African environments were very different from those of the present, and

when the fossil record is generally poor.

The fact that ‘E’ Quarry species such as Hipparion cf. baardi and the two

alcelaphines are, or may be, different from contemporary counterparts further

north in Africa, suggests that the south-western Cape, or an even larger area of

southern Africa, had been zoogeographically isolated some time during the late

Miocene. On the other hand, this isolation could not have been complete, or

had broken down very late in the Miocene, since there evidently had been

some immigration of taxa shortly before deposition of the QSM and PPM. Such

taxa include Agriotherium and the others referred to above.

Since the Varswater Formation was laid down over an appreciable period,

it is also possible that during this period some species arrived in the area as

immigrants, while others became extinct. However, it may not be possible to

determine that this actually happened. The simple fact that certain species are

not recorded low in the sequence does not necessarily mean that they were not

present in the area. In other words, the first appearance of a species in the

sequence was not necessarily coincident with its first appearance in the region.

Conversely, the presence of certain taxa low in the sequence, and their absence

higher up, is not necessarily indicative of their extinction. Both sampling

inadequacies and taphonomic factors are potential biases in the “E’ Quarry

record.

Sampling inadequacies is the less serious problem, since the sample sizes

from the QSM, bed 3aS, and bed 3aN are large. Nevertheless, past experiences

at this locality have shown that it is a factor that cannot be dismissed. For

example, it was once thought that porcupines had not been an element of the

fauna (Hendey 1974a), but remains of a single individual were subsequently

found in the QSM (Hendey 1976a), and since then fragmentary remains of a

second species have turned up on several occasions in beds 3aS and 3aN.

The problem with taphonomic factors is that the environments of deposi-

tion of the various elements in the succession vary, and it is obvious that taxa

represented in one (e.g. river floodplain) may differ from those in another (e.g.

river channel).

The QSM assemblage largely represents the remains of animals that lived

and died in the immediate vicinity, whereas the beds 3aS and 3aN assemblages

are largely comprised of animal remains washed in by the river from elsewhere.

Consequently, the presence of species in these two sets of deposits simply

reflects the opportunity for their remains to reach their points of deposition,

and may have nothing to do with their presence or absence in the region.

PALAEOECOLOGY OF LANGEBAANWEG 87

Individual taxa must, however, be assessed separately. For example, the

fact that hippos are not recorded from the QSM may be due to the lack of

exposures of those deposits where their remains were most likely to be

preserved, that is, those of the main river channel of that time. However, even

in the recorded channel deposits hippos are unequally represented. They were

extremely rare in bed 3aS, whereas they were much more common in bed 3aN.

It is thus possible that the representation of this animal in the ‘E’ Quarry

sequence is a true reflection of its initial absence (i.e. in QSM times), and its

subsequent presence in increasing numbers (i.e. in beds 3aS and 3aN).

The occurrence of the giant pig, Nyanzachoerus, is in curious contrast to

that of the hippo. Nyanzachoerus occurs quite commonly in the QSM, N. cf.

pattersoni being one of the characteristic species of this horizon. This species

may also be present in bed 3aS, although it is extremely rare. A second species,

N. cf. jaegeri, is definitely recorded from bed 3aS, although it, too, is un-

common. However, neither species, nor any other pig, is recorded from bed

3aN. This creates the impression that Nyanzachoerus was ‘replaced’ locally by

the hippo. There is, however, no evidence from elsewhere that hippos and

Nyanzachoerus were mutually exclusive taxa, although they might well have

competed for a common food source.

The situation in respect of these animals at Langebaanweg might have

differed from that in areas where they did co-exist. Since the local environment

was deteriorating during the early Pliocene, it is possible that competition

between hippos and Nyanzachoerus for diminishing food resources was critical

enough to cause the extinction of the latter. Thus the record of hippos in bed

3aS may, indeed, coincide with the first appearance of these animals in the

Langebaanweg area and the initiation of competition with Nyanzachoerus, the

latter process being concluded in the favour of the hippos by the time that bed

3aN was deposited. Alternatively, the two forms might have co-existed throu-

ghout the period of deposition of the QSM and bed 3aS, with the displacement

process being more gradual. The first of these alternatives is perhaps the more

likely in view of the observed relative abundance of the two forms.

One of the more intriguing features of the QSM and PPM faunas is the

extreme rarity of primates in the former and their complete absence in the

latter. Of the many thousands of mammalian teeth recovered to date, there are

only two specimens from the QSM identified as primate, while there are no

primate bones amongst the far more numerous identified postcranial remains

from the QSM and PPM. Primates are clearly grossly under-represented in the

‘FE’ Quarry fauna, given its location in Africa, the great number and variety of

mammals recorded, and the vast number of specimens already collected and

identified. In older and broadly contemporary faunas elsewhere in Africa, and

in southern Eurasia, both terrestrial and arboreal primates are represented,

sometimes in appreciable numbers and variety and often in assemblages far

smaller than that from ‘E’ Quarry.

It was suggested elsewhere that the location of Langebaanweg on the coast

88 ANNALS OF THE SOUTH AFRICAN MUSEUM

in the far south of the continent might have had a bearing on the under-

representation of primates (Hendey 1976a: 234). The coastal situation is

unlikely to have been significant, since primates are not necessarily averse to

this kind of environment, either now or in the past. For example, the broadly

contemporary fauna from Montpellier, which was accumulated in a similar

environment, includes primates (see below).

The situation of Langebaanweg at nearly 34°S may be a more significant factor.

Over 80 per cent of Africa’s non-hominid primates occur in low latitudes, between

15°N and 15°S (see distribution maps in Dorst & Dandelot 1970), and are thus most

characteristic of the hot tropical regions. In the warm-temperate south-western

Cape there is only one non-hominid primate recorded, which is about 2 per cent of

the total in the Ethiopian Region. This species is the chacma baboon, Papio ursinus.

A monkey (probably Cercopithecus) is tentatively recorded on the basis of early

historical records. The Pleistocene fossil record for the south-western Cape is

comparatively good, but once again only one non-hominid primate is recorded, 1.e.

the baboon, Theropithecus oswaldi, from the middle Pleistocene of Elandsfontein.

Clearly, the frequency in the occurrence of primates can be correlated with latitude,

and, consequently, also with temperature.

However, the lower temperatures of higher latitudes do not necessarily

directly limit the occurrence of primates. Instead, they have the effect of

limiting the food resources available to essentially herbivorous primates. Tropi-

cal regions remain productive in terms of such food resources throughout the

year, but in temperate regions they are readily available only during summer.

Consequently, although the vegetation at Langebaanweg during the early

Pliocene was such that the occurrence of both terrestrial and arboreal primates

might have been expected, winter temperatures were probably already too low

to have maintained adequate food supplies for most primates. This order may,

therefore, be included in the hypothetical group of animals that became extinct

in the south-western Cape earlier than elsewhere (see p. 83).

It is worth noting that primates were still present in contemporary faunas

of comparable latitudes in the Northern hemisphere. This is readily explained

by the fact that the existing zonality in world climates had yet to be established.

The glaciation of the Arctic lagged behind that of Antarctica, and it was not

until about 3 Ma that the situation in the two hemispheres was more or less

equalized.

The above theory implies that primates were still present in the far south

of the continent during the warmer period in the late Miocene and earlier.

However, although the Miocene mammal record for southern Africa is very

poor, it may again be significant that this record includes no primates. Only one

Miocene occurrence is presently known where the assemblage is sufficiently

large for primates to have been expected. It is from Arrisdrift in the southern

Namib desert, and in this instance it was suggested that zoogeographic barriers

between east and southern Africa might have impeded the southward dispersal

of primates (Hendey 1978b: 35).

PALAEOECOLOGY OF LANGEBAANWEG 89

Other notable absentees from the ‘E’ Quarry fauna are the Deinother-

lidae, Chalicotheriidae, and, amongst the non-mammals, the Crocodilia. Pro-

boscideans are a comparatively rare element in this fauna, and the absence of a

deinothere may therefore reflect a sampling deficiency. The same may apply in

the‘case of the chalicothere, an animal that was uncommon elsewhere in Afriica

as well. The situation in respect of the Crocodilia is likely to be different, since

wherever these animals are present they are likely to be well represented at

least by isolated teeth and scutes, if not by other skull parts and postcranial

bones. Crocodiles have not been recorded from any other south-western Cape

deposits either, and their absence is probably due to the relatively low temper-

atures in local aquatic environments in post-Miocene times.

Both deinothere and crocodile are recorded from the early middle

Miocene deposits at Arrisdrift, 500 km to the north, so they, like primates,

might have had their ranges restricted by falling temperatures in the far south

of the continent during the late Miocene and early Pliocene.

Although the ‘E’ Quarry fauna includes Eurasian immigrants, it was

dominated by taxa that were, and, in some cases, still are, characteristically

‘African’. These include groups such as the Macroscelididae, Chrysochloridae,

Tubulidentata, Proboscidea, Hyracoidea, and certain tribes of Bovidae.

The ‘African’ character of this fauna is clearly manifested by a comparison

with the broadly contemporary Montpellier fauna (Table 9). Apart from these

faunas being of comparable age, they are from similar coastal situations. In

their overall composition the faunas of ‘E’ Quarry and Montpellier are remark-

ably alike, comprising of a similar array of carnivores, proboscideans, perisso-

dactyls, artiodactyls, cetaceans, and some smaller mammals. The latter are

under-represented at Montpellier probably because of sampling deficiences,

while the under-representation of primates in ‘E’ Quarry is ascribed to the

climatic factors discussed above.

In spite of their overall similarities, the two faunas are taxonomically distinct,

the differences being mainly at generic and specific level. Conspecificity, or very

close relationship, is largely confined to the Carnivora, for reasons discussed

earlier (see p. 84), but even in this group there are obvious differences.

The most striking and unexpected of these differences is the absence of

hyaenas at Montpellier. Hyaenas are well represented in Europe at localities of

late Miocene and Pleistocene age, and they are also common in ‘E’ Quarry.

Their comparative rarity (or absence) at European localities of early Pliocene

age has yet to be satisfactorily explained. Smaller viverrids are also not

recorded at Montpellier, but otherwise differences from ‘E’ Quarry carnivores

are at generic or specific level.

For example, the badgers represent the genera that are still characteristic of

the continents concerned (i.e. Mellivora in Africa, Meles in Europe). Similarly,

the seals, both of which belong to the subfamily Monachinae, are most closely

related to taxa that are still found in geographical proximity (i.e. Homiphoca to

Antarctic Lobodontini, Pristiphoca to Mediterranean Monachus monachus).

TABLE 9

ANNALS OF THE SOUTH AFRICAN MUSEUM

The mammalian faunas of the Varswater Formation (QSM and PPM), Langebaanweg, and

the Pliocene deposits at Montpellier, France.

Insectivora and Chiroptera

Primates

Pholidota and Tubulidentata

Carnivora

Canidae

Ursidae

Mustelidae

Phocidae

Viverridae

Hyaenidae

Felidae

Proboscidea

LANGEBAANWEG

Present

1 species

Present

. 2 species

. Agriotherium africanum

(aff. insigne)

. Plesiogulo monspessulanus

Mellivora benfieldi

Enhydriodon africanus

. Homiphoca capensis

. ‘Viverra’ leakeyi

Viverrinae sp. B

Genetta and Herpestinae

. Several species

. Machairodontinae (2 spp)

Felis aff. issiodorensis

(? aff. christoli)

Other Felinae (3 spp)

. Mammuthus subplanifrons

MONTPELLIER*

Not recorded

3 species

Not recorded

Agriotherium insigne

Plesiogulo monspessulanus

Meles gennevauxi

Lutra affinis

Pristiphoca occitana

Viverra aff. pepratxi

Not recorded.

Not recorded

Machairodontinae (1 sp.)

Felis christoli

Not recorded

Mastodon arvernensis

Pliehyrax occidentalis

Hipparion crassum

Dicerorhinus megarhinus

Tapirus arvernensis

Not recorded

Sus arvernensis

Not recorded

3 species

2 or 3 species,

Gazella sp.

including a

Prolagus sp.

Anancus sp.

Hyracoidea . Procavia cf. antiqua

Perissodactyla

Equidae . Hipparion cf. baardi

Rhinocerotidae . Ceratotherium praecox

Tapiridae . Not recorded

Artiodactyla

Tayassuidae . . | species

Suidae . . Nyanzachoerus (2 spp)

Hippopotamidae . 1 species

Giraffidae . 3 species

Cervidae . Not recorded

Bovidae . 12 species, including a Gazella

sp.

Lagomorpha . Pronolagus sp.

Rodentia . Many species

Cetacea

*Anonymous 1975

. Several species

1 species

Several species

PALAEOECOLOGY OF LANGEBAANWEG 9]

Amongst the non-carnivores, giraffids and bovids are the predominant

ungulates in ‘E’ Quarry, whereas at Montpellier giraffids are absent (or not

recorded), and cervids are present in addition to bovids. Hipparion is recorded

at both localities. The combinations of Bovidae—Giraffidae—Equidae in ‘E’

Quarry and Cervidae—Bovidae—Equidae at Montpellier are still characteristic of

the continents concerned.

DATING OF THE LANGEBAANWEG SUCCESSION

One of the most persistent problems pertaining to the Varswater Forma-

tion is that of its age. As is frequently the case in southern Africa with deposits

of late Cenozoic age, only indirect dating methods can be used. The situation

is, however, exceptional in that in this instance faunal, botanical and geological

evidence is actually or potentially available for relative dating purposes. The

present study has clarified the dating problem.

Hitherto dating has depended largely on the fossil mammals of the Vars-

water Formation. Initially these fossils were thought to be of early Pleistocene

age (e.g. Singer 1961; Boné & Singer 1965; Hendey 1969, 1970a), but subse-

quently it was suggested that the formation included both Pliocene and Pleis-

tocene elements (Hendey 1970b). This is, indeed, the case with the fossil

assemblage from Baard’s Quarry (Hendey 1978a), but the Varswater Formation

in “E’ Quarry is entirely of late Tertiary age. This was established with the

realization that geological evidence had been misinterpreted, that some of the

mammalian taxa represented had been misidentified, and that the provenance of

an Equus tooth had been incorrectly recorded (Hendey 19726). Thereafter, the

most widely accepted age estimate was early Pliocene (c. 4-5 Ma) (e.g. Hendey

1973, 1974a). Instead of resolving the issue, more recent faunal studies have

provided contradictory dating evidence and have also indicated that deposition

of the Varswater Formation took place over a prolonged period (e.g. Hendey

1976a, 1978d, 1980). As a result, when referring to the age of this formation it

has been the practice to give only the outside age limits, which are 7 and 3,5 Ma.

These limits include the terminal Miocene as well as the early Pliocene.

The principal difficulty with faunal dating has been that broadly contempor-

ary faunas elsewhere in Africa, some of which are securely dated by absolute age

determinations, are not well known. Consequently, the basis for comparison is far

less than the lengthy ‘E’ Quarry faunal list would suggest. In addition, the location

of Langebaanweg near the southern continental extremity raises the problem of

regional peculiarities in the fauna of the kind discussed above (see pp. 85-89).

In that discussion the ‘E’ Quarry hippo was used as an example to illustrate

a point, and it can be used again to indicate the kind of difficulties that arise in

faunal dating. The fossil record of Hippopotamidae in east Africa is good, and

the evolutionary history of the family is well understood (Coryndon 1978).

Hippos are not the best mammals to use for relative dating purposes, since they

are conservative animals in an evolutionary sense, and are difficult to identify

92 ANNALS OF THE SOUTH AFRICAN MUSEUM

to species level on the basis of fragmentary material. Nevertheless, they may

serve as age indicators in the case of faunas such as the one from ‘E’ Quarry.

It was suggested earlier that hippos might have made their appearance in

the Langebaanweg area only during the period of deposition of bed 3aS (or

shortly before). Hippos are first recorded in east Africa as early as 10 Ma

(Coryndon 1978), but it would be unwise to conclude that their first appearance

there coincided with their appearance in the Varswater Formation. It is much

more likely that environmental factors retarded the southward dispersal of this

water-demanding animal, perhaps for an appreciable period. Hippos may thus

be an example where geography is a complicating factor when they are used for

relative dating purposes.

A contrasting example of a herbivore noted for its rapid dispersal and

utility in faunal dating is that of Hipparion, the three-toed horse that first made

its appearance in Africa at about 12,5 Ma, and which survived on this continent

well into the Pleistocene (Churcher & Richardson 1978). Again there is a

complicating factor with the Langebaanweg representatives of this genus. The

most commonly occurring of the ‘E’ Quarry species 1s Hipparion cf. baard1i,

which evidently belongs on a lineage endemic to the more southerly parts of

Africa, and comparisons with broadly contemporaneous Hipparion elsewhere

in Africa, therefore, provide little evidence of their likely relative age.

The Langebaanweg Hipparion has, however, provided useful relative age

information on deposits occurring in the immediate vicinity. For example, the

GM H. cf. primigenium is consistent with the late Miocene date for this horizon

deduced on other evidence (see p. 23). Another species, H. cf. namaquense, is

known only from the uppermost levels of the Varswater Formation, and may

even come from overlying deposits (Hendey 1976a; Hooijer 1976). Finally, the

Baard’s Quarry H. baardi, which is now also recorded from the Anyskop

terrestrial deposits, is clearly distinct from the ‘E’ Quarry H. cf. baardi, and

suggests a younger date for the deposits in which it occurs (Hendy 1978a; see

also p. 38).

Bearing in mind the difficulties in long-range correlations suggested by the

hippo and Hipparion, and the other potential regional peculiarities discussed

earlier, it is evident that in faunal dating of the Varswater Formation the

following two factors are of the greatest significance:

1. The occurrence at more securely dated localities of species conspecific

with ones occurring in “E’ Quarry.

2. The evolutionary state of ‘E’ Quarry taxa relative to those of related

taxa from dated localities elsewhere.

Initially it was two species from the QSM that were at the core of the

correlations indicating an early Pliocene age. They were the elephant, Mammu-

thus subplanifrons, and the pig, Nyanzachoerus cf. pattersoni (or kanamensis).

A 5 Ma maximum age for the OSM was based on the belief that counterparts of

PALAEOECOLOGY OF LANGEBAANWEG 93

these two species in the Lothagam 1 fauna from Kenya were aged between 5

and 6 Ma (Patterson et al. 1970; Maglio 1973). Subsequently Hooijer & Maglio

(1974: 4) concluded that this fauna ‘may be somewhat older than [6 Ma, but] it

is not likely to be much younger’. Behrensmeyer (1976: 167) noted that it is

‘obviously important to think of the Lothagam 1 fauna as representative of a

relatively long time span’ and that ‘the fauna from 1C may be somewhat

younger than 6,0 m.y.’ The base of the Lothagam sequence is underlain by

volcanics dated at 8,3 Ma (Behrensmeyer 1976), which is, therefore, the

absolute maximum age of the Lothagam fauna.

Although there is still uncertainty about the actual age of the Lothagam 1

fossils, there can be virtually no doubt that they pre-date those from the OSM,

and, consequently, also those from beds 3aS and 3aN. The Lothagam 1

elephants (Stegotetrabelodon orbus, Primelephas gomphotheroides) and the pig

(Nyanzachoerus tulotus) are more primitive than their counterparts from ‘E’

Quarry, as is an undescribed hyaenid from Lothagam 1.

On the other hand, the c. 4 Ma fauna from Kanapoi, 75 km south of

Lothagam (Behrensmeyer 1976), is apparently younger than that from ‘E’

Quarry, although the evidence for this is more slender. An undescribed

hyaenid from Kanapoi is certainly more advanced than its counterparts in the

‘E’ Quarry ‘Hyaena group’ (see Hendey 1978c). The presence of the pig,

Notochoerus, at Kanapoi and its absence in ‘E’ Quarry is also indicative of a

younger age for the former. In addition, the QSM Nyanzachoerus cf. pattersoni

is slightly more primitive than the N. pattersoni from Kanapoi in retaining P’.

The preceding evidence indicates that the age difference between the two

faunas is of no great magnitude, an opinion supported by the fact that their

elephants are in a comparable evolutionary state (Maglio 1973).

There is, however, a complication with the Nyanzachoerus cf. jaegeri from

bed 3aS in ‘E’ Quarry. White & Harris (1977) and Harris & White (1979) have

suggested that the relative stratigraphic positions of N. cf. pattersoni (or

kanamensis) and N. cf. jaegeri in ‘E’ Quarry are evidence in support of their

theory that the latter evolved from the former. According to their correlation

the upper levels of the Varswater Formation (i.e. beds 3aS and 3aN) actually

overlap with the Kanapoi Formation (Harris & White 1979, fig. 134). This

conclusion is in conflict with the evidence that suggests that all of the Varswater

Formation is older.*

The most positive indication of the earlier date for the Varswater Forma-

tion relative to the Kanapoi Formation comes from the hyaenids. The three ‘E’

Quarry species most closely related to the Kanapoi hyaenid are Ictitherium

preforfex, Hyaena abronia, and Hyaenictitherium namaquense, all of which

retain P; and Mp, teeth which the Kanapoi species has lost.

* A recent examination of the ‘E’ Quarry N. cf. jaegeri by T. D. White has led him to

believe that it represents a more primitive variety than the one from Kanapoi, and that the ‘E’

Quarry pigs are consistent with a 5 Ma date.

94 ANNALS OF THE SOUTH AFRICAN MUSEUM

Correlation of the ‘E’ Quarry fossils with those from east African localities

such as Lothagam and Kanapoi was recently further complicated by Gentry’s

(1980: 333) conclusion that ‘the bovids best indicate an age of about 6 m.y. for

the fauna’. The bovids thus suggest an age comparable to Lothagam 1, and

appreciably older than Kanapoi. However, the bovids do not necessarily rule out

a younger age, and at least in the case of the buffalo, Simatherium demissum, are

consistent with such an age. Gentry (1980: 332) found that ‘Parabos boodon,

which is at a comparable evolutionary level [to S$. demissum, but] on a different

lineage, comes from Perpignan, which is given an age of about 4,8 m.y.’

While the utility of groups such as the elephants, pigs, and bovids in faunal

dating is undeniable, preference is here given to that of the carnivores. The

environmental tolerance of these animals allows at least the larger species to

disperse rapidly when the opportunity arises, with little or no evolutionary

adaptation necessarily required, and with consequent wide distributions for

individual species. This facilitates long-distance correlations. The post-

Lothagam 1 and pre-Kanapoi age for the ‘E’ Quarry fauna indicated by the

hyaenids is, therefore, regarded here as more acceptable than the conflicting

evidence of the pigs and bovids, and the equivocal evidence of the elephants.

The outside age limits of 3,5 and 7 Ma are thus reduced to 4 and 6 Ma, with a

median estimate of 5 Ma applying.

The ‘E’ Quarry carnivores provide further evidence in support of the 5 Ma

estimate. The recent study of the Agriotherium from this locality suggested that

it is at a comparable evolutionary stage, or, in the case of the bed 3aN material,

perhaps even slightly more advanced than the Agriotherium from Montpellier

in France. Similarly, available materials of the Plesiogulo and a small Felis

from the two localities are virtually indistinguishable. Other, more general

similarities between the two faunas were referred to earlier (see p. 89, Table 9),

and faunal evidence thus points to broad contemporaneity between the ‘E’

Quarry deposits and those of Pliocene age at Montpellier.

The latter were laid down during the marine transgression that terminated

the Messinian salinity crisis and ushered in the Pliocene. It was this correlation

that first suggested that deposits in the Langebaanweg area could be related to

global phenomena of the late Tertiary, and which prompted the re-examination

of the local succession, a process that was to culminate with the correlation with

global sea-level changes (see pp. 12-18, Fig. 3, Table 2).

This correlation has placed the matter of the age of individual elements in

the Langebaanweg succession in an entirely new light. Whereas previously even

correlations with deposits and faunas elsewhere in Africa were generally

tentative, there is now the potential for secure correlations with successions

recorded from all the continents and oceans of the world. In addition, the

Langebaanweg area could become the key to interpreting other late Tertiary

strata in southern Africa, which have often proved difficult to date, and which

have had obscure histories and relationships.

While the correlation of individual elements in the Langebaanweg succes-

PALAEOECOLOGY OF LANGEBAANWEG 95

sion with subdivisions of the late Tertiary epochs appears secure, the absolute

or chronometric ages of these elements have yet to be precisely determined.

For example, although the QSM, PPM, and Anyskop marine deposits date

from the early Pliocene transgression, there is as yet no consensus on the timing

and duration of this event. On the basis of evidence from the Mediterranean

Basin, Van Couvering et al. (1976) and Van Couvering & Berggren (1977)

indicate that this transgression took place between 5,0 and 4,5 Ma, whereas

Vail & Hardenbol (1979) indicate it as a rapid event that took place at 5,2 Ma.

The essential difference between these interpretations concerns the dating of

the Miocene—Pliocene boundary, which is taken respectively at 5,0 and 5,2 Ma,

while there are also implications concerning the duration of the transgression.

Clearly, the dating of the Miocene—Pliocene boundary is a matter that

cannot be resolved on the basis of evidence from Langebaanweg. There is,

however, little doubt that a generally acceptable date, probably 5,2 Ma (see

Van Couvering 1978), will result from the many investigations relevant to this

question. The same applies in the case of the duration of. the early Pliocene

transgression. Judging from evolutionary changes manifested in certain mam-

malian taxa common to more than one level of the QSM and PPM (see p. 24),

the phase of the transgression represented by these deposits might have been in

the order of many millennia, while the transgression as a whole could conceiv-

ably have lasted several hundred thousand years.

Until these issues are resolved, it will be convenient to use the 5 Ma

estimate suggested above in reference to the deposits in the Langebaanweg

area that date from the early Pliocene transgression (i.e. the QSM, PPM and

Anyskop marine deposits).

There are more marked differences of opinion over the dating of the

Pliocene—Pleistocene boundary, and these have a bearing on the inferred

chronometric dates of the later elements in the Langebaanweg succession,

particularly the Baard’s Quarry fluviatile deposits. In recent years it has become

common practice to date the Pliocene—Pleistocene boundary at either 1,8 Ma

(e.g. Berggren & Van Couvering 1974), or 1,6 Ma (Haq et al. 1977). A

contrasting opinion is that of Vail & Hardenbol (1979), who fixed this boundary

at 2,8 Ma, which coincides with the termination of their sea-level cycle, TP3.

It was with an age range of 2-3 Ma in mind that the age of the Baard’s

Quarry fluviatile deposits fauna was suggested to be late Pliocene/early Pleis-

tocene. It now appears that at least a part of this fauna is younger than 2 Ma, a

conclusion that is based on the presence of Equus. According to Lindsay et al.

(1980: 135) ‘there were at least three major dispersal events of large mammals

during the Pliocene (at 1,9, 2,6 and 3,7 Myr)’, with Equus having dispersed into

Africa during the 1,9 Ma event. The Baard’s Quarry fluviatile deposits,

therefore, cannot be older than 1,9 Ma, and are more likely to date from

sea-level cycle QI than cycle TP3 (see p. 41).

This raises another problem concerning the fauna of these deposits. There

are other elements in the assemblage that would be more consistent with an age

96 ANNALS OF THE SOUTH AFRICAN MUSEUM

of more than 2 Ma. This was discussed by Hendey (1978a) and Gentry (1980),

and it now seems that the possible temporal heterogeneity of this fauna was

dismissed too lightly in the former study. The composition of this fauna can be

most readily explained by the hypothesis that while the deposits themselves

date from cycle Ql, they include both contemporary fossils and others

reworked from pre-existing Pliocene deposits.

There are thus several issues relating. to the age of the post-Varswater

Formation deposits in the Langebaanweg area that have yet to be settled, while

the chronometric ages of all elements in the succession have also to be

established. In addition, Vail & Hardenbol (1979: 79) concede that there ‘is no

question that considerably more research needs to be undertaken to document

the magnitudes and timing of eustatic sea-level changes’. Nevertheless, little

significance is attached to the present uncertainties concerning details of the

Laangebaanweg succession and global sea-level changes. The agreement

between them is sufficient to justify the conclusion that the former can for the

most part be regarded as securely dated in a geological sense. At the very least,

there is now a sound basis on which future research on dating can be based.

In conclusion, it should be mentioned that botanical evidence of age is

potentially available for certain elements in the succession, although for the

present and in the immediate future the faunal and geological dating of the

Langebaanweg sequence will provide data for palaeobotanical sudies. A tenta-

tive ‘Pollen Zone’ sequence for the south-western Cape has already been

established (Coetzee 1978), and there is apparently considerable potential for

refining this scheme (J. A. Coetzee, pers. comm.). The existing pollen sequence

is only tentatively correlated with the ‘E’ Quarry succession (Table 8).

SUMMARY AND CONCLUSIONS

Cenozoic deposits in the vicinity of Langebaanweg are comprised largely of a

late Tertiary succession of clastic sediments. They include economically important

phosphate deposits and some immensely rich fossil occurrences. The history of

this succession has been reinterpreted on the basis of geological and palaentologi-

cal evidence, and it has been correlated with the record of South African west

coast sea-level changes, for which successive shorelines at 30 m, 90 m, 50 m and 20

m above present sea-level are recognized. This interpretation is in accord with the

global sea-level changes recorded by Vail & Hardenbol (1979).

The succession is as follows:

1. Basal element (‘pre-GM deposits’)—early to middle Miocene trans-

gression (part of global sea-level cycles TM1.3, TM1.4, TM2.1 and

TM2.2); terrestrial and marine complex.

2. No local record—middle to late Miocene regression (cycle TM2.3).

3. Gravel Member (GM) of the Varswater Formation—late Miocene

regression (30 m shoreline—cycle TM3.1); marine beach complex.

4. No local record—terminal Miocene regression (cycles TM3.2, TM3.3).

PALAEOECOLOGY OF LANGEBAANWEG 97

5. Quartzose Sand Member (QSM) of the Varswater Formation—early

Pliocene transgression (cycle TP1); estuarine complex with floodplain

(QSM I), salt marsh (QSM II) and tidal flat (QSM III) facies.

6. Pelletal Phosphorite Member (PPM) of the Varswater Forma-

tion—early Pliocene transgression (cycle TP1); fluviatile (beds 3aS

and 3aN) and marine littoral (PPM undifferentiated) complex.

7. Anyskop marine deposits—early Pliocene transgression (cycle TP1);

coastal barrier complex.

8. Marine platform south-east of Langebaanweg—early Pliocene trans-

gression (90 m shoreline—cycle TP1).

9. Anyskop terrestrial deposits—late Pliocene regression (50 m shore-

line—cycle TP2).

10. Baard’s Quarry fluviatile deposits—late Pliocene regression, or, more

probably, early Pleistocene transgression (20 m shoreline—cycle TP3

on cycle @)):

The local succession is completed by Quaternary deposits, including a nearly

ubiquitous covering of aeolian sands. The area also has fossil occurrences and

hominid occupation sites dating from the middle Pleistocene, late Pleistocene,

and Holocene.

The nature of this succession was determined by the physical geography of

the region, while the preservation of an unusually large body of early Pliocene

sediment was due to the development of a coastal barrier complex, part of

which survived subsequent erosion, and which protected underlying deposits.

Palaeontological investigations have centred on the Varswater Formation

as it is exposed in an open-cast phosphate mine, ‘E’ Quarry. Some plant fossils

are represented in these and underlying deposits, while about 230 invertebrate

and vertebrate taxa, ranging from protozoans to mammals, have been

recorded, mainly from the QSM and PPM. The fossils were recovered from a

variety of marine, freshwater and terrestrial depositional environments. The

species and body part representation, and the condition of specimens, was

dependent on the source of the material.

Most of the collected invertebrates have been described, but the lower

vertebrates are still largely unstudied. The birds evidently represent the largest

pre-Pleistocene assemblage known anywhere, and they are also largely un-

studied. The mammals are the most intensively studied and best represented

group, although only about half of the eighty species recognized have been

described. They range from shrews and mice to elephants and whales, and

those already studied include four new genera and sixteen new species.

Detailed analyses of assemblages from the various depositional environ-

ments have yet to be undertaken. However, preliminary observations have

revealed evidence for both subaerial and subaqueous deposition of material

(e.g. in QSM 1, and in PPM, 3aN respectively). Individual specimens show

signs of damage by such processes as carnivore activity (e.g. toothmarks,

etching by stomach acids) and fires.

98 ANNALS OF THE SOUTH AFRICAN MUSEUM

Late Tertiary environments contrast sharply with that of the present, which

is semi-arid (summer dry/winter wet), with no naturally occurring surface

freshwater, and with a rather sparse sclerophyll (fynbos) vegetation without

indigenous trees. During the Miocene, tropical forests, possibly associated with

a monsoon climate, existed in the area, but by the early Pliocene the environ-

ment was deteriorating. Temperatures had moderated from tropical to temper-

ate, while precipitation had probably declined. Rainfall was then strongly

seasonal, with a summer-wet/winter-dry pattern being likely. Wooded areas

were probably restricted to the immediate vicinity of the river which was

responsible for feeding in most of the sediment comprising the QSM and PPM.

Grazing species amongst the mammalian herbivores indicate the presence of

grasslands, and there is palynological evidence for the presence of fynbos

vegetation types. However, woodland browsers still formed a_ significant

element in the mammalian fauna, although dental abnormalities indicate

stresses in these populations which might have been due to the diminution of

suitable habitats, and aggravated by factors such as droughts and fires.

The evidence for changing environments at Langebaanweg during the late

Tertiary is in keeping with the record elsewhere in the word. This contributed

to the correlation of the local succession with others elsewhere in the world

(e.g. the Mediterranean Basin), which were also under the influence of such

phenomena as changes in the volume of the Antarctic ice-cap.

The mammalian fauna of the QSM and PPM is in several respects

‘intermediate’ in character, which is in keeping with the environmental changes

of the period in question. Although it is essentially ‘African’ in character, it

includes descendants of several Miocene immigrants from Eurasia (e.g. the

three-toed horse, Hipparion, and the bear, Agriotherium). The composition of

certain groups (e.g. the hyaenas) still reflected a pattern more characteristic of

the Miocene than the Quaternary. However, early representatives of successful

Quaternary lineages are recorded (e.g. the white rhinoceros, Ceratotherium).

While individual taxa represent advances over their Miocene ancestors, many

were much less specialized than their Quaternary descendants (e.g. the seal,

Homiphoca capensis). Some ‘tropical’ elements were still present (e.g. parrots,

giraffes), but the moderating climate might already have caused some to

become extinct, or nearly extinct (e.g. crocodiles, primates).

Faunal evidence of age is in good accord with the geological evidence,

although absolute, or chronometric, dates for the various elements in the

succession are still lacking.

In a little more than a decade the Langebaanweg succession has become

one of the more intensively studied and best understood complexes of late

Tertiary deposits in southern Africa. It has also become the source of the

largest assemblage of late Tertiary vertebrates known anywhere in Africa. The

unique set of geological, botanical and faunal data already available have

combined to provide an unparalleled insight into the nature of local late

Tertiary environments and the changes that they underwent. In spite of all this,

PALAEOECOLOGY OF LANGEBAANWEG 99

there is still a considerable potential for further research, and the significance

of this area in southern African Cenozoic studies should increase still further.

ACKNOWLEDGEMENTS

I am indebted to the following persons who assisted in the preparation of

the manuscript of this paper: Mr F. Grine, Miss P. Haarhoff, Miss B. Mann,

Miss J. Nolte and Miss L. Scott (South African Museum); Dr J. A. Coetzee

(University of the Orange Free State); Prof. H. J. Deacon (University of

Stellenbosch); Prof. R. V. Dingle, Mrs R. Martin and Mr D. Salmon (Univers-

ity of Cape Town); Mr H. Jenner-Clarke (ASAM Minerals (Pty) Ltd);

Dr R. G. Klein (University of Chicago); Mr C. A. Repenning (United States

Geological Survey); Dr J. Rogers (South African Geological Survey);

Dr W. G. Siesser (Vanderbilt University); Dr C. G. Stocken (Consolidated

Diamond Mines of SWA (Pty) Ltd); and Dr A. J. Tankard (University of

Tennessee).

Much of the information presented here was derived from the research of

the many past and present collaborators in the Langebaanweg Research

Project, and my thanks go also to them. This research was made possible by the

co-operation of the management and employees of Chemfos Ltd, and the many

field and laboratory assistants involved in the collection and preparation of

material. Special thanks are due to Mr I. Tembo (Orapa, Botswana) for his

dedicated work over many years as senior field assistant. His efforts did much

to ensure the success of the collecting programme.

The Langebaanweg Research Project is supported by Chemfos Ltd, the

South African Council for Scientific and Industrial Research, and the

Wenner-Gren Foundation for Anthropological Research (Grant no. 2752-1834),

and the assistance of these organizations is gratefully acknowledged.

REFERENCES

Anonymous, 1975. Données sur quelques localités 4 mammiféres d’Age Néogéne supérieur du

Languedoc. Act. I Col. Intern. Sobre Biostrat. Cont. Neogeno Sup. Cuat. Inf. 4: 23-29.

AXELROD, D. I. & Raven, P. V. 1978. Late Cretaceous and Tertiary vegetation history of

Africa. In: WERGER, M. J. A. ed. Biogeography and ecology of southern Africa: 77-130.

The Hague: W. Junk.

AZZAROLI, A. 1975. Late Miocene interchange of terrestrial faunas across the Mediterranean.

Act. I Col. Intern. Sobre Biostrat. Cont. Neogeno Sup. Cuat. Inf. 4: 67-71.

Bass, A. J., D’AuBrey, J. D. & Kistnasamy, N. 1973. Sharks of the east coast of southern

Africa. I. The genus Carcharhinus (Carcharhinidae). Invstl Rep. oceanogr. Res. Inst.

Durban 33: 1-168.

Bass, A. J., D’AusBrey, J. D. & Kistnasamy, N. 1975. Sharks of the east coast of southern

Africa. V. The families Hexanchidae, Chlamydoselachidae, Heterodontidae, Pristiophori-

dae and Squatinidae. Invstl Rep. oceanogr. Res. Inst. Durban 43: 1-50.

BEHRENSMEYER, A. K. 1976. Lothagam, Kanapoi and Ekora: A general summary of stratigra-

phy and fauna. In: Coppens, Y. et al. eds. Earliest man and environments in the Lake

Rudolf Basin: 163-170. Chicago: University Press.

BERGGREN, W.A. & VAN COUVERING, J. A. 1974. The Late Neogene. Amsterdam: Elsevier.

100 ANNALS OF THE SOUTH AFRICAN MUSEUM

BircH, G. F. 1977. Phosphorites from the Saldanha Bay region. Trans. R. Soc. S. Afr. 42:

223-240.

BisHop, W. W. 1980. Palaeogeography and continental taphonomy. Jn: BEHRENSMEYER, A. K.

& Hin1, A. P. eds. Fossils in the making: 20-37. Chicago: University Press.

Boaz, N. T., Gaziry, A. W. & EL-ARNaAuTI, A. 1979. New fossil finds from the Libyan Upper

Neogene site of Sahabi. Nature, Lond. 280: 137-139.

Bone, E. L. & SINGER, R. 1965. Hipparion from Langebaanweg, Cape Province and a revision

of the genus in Africa. Ann. S. Afr, Mus. 48: 273-397.

BoucHErR, C. & JARMAN, M. L. 1977. The vegetation of the Langebaan area, South Africa.

Trans. R. Soc. S. Afr. 42: 241-272.

Brown, A. C. & JARMAN, N. 1978. Coastal marine habitats. Jn: WERGER, M. J. A. ed.

Biogeography and ecology of southern Africa: 1239-1277. The Hague: W. Junk.

ButTzer, K. W. 1973. On the sedimentary sequence from the ‘E’ Quarry, Langebaanweg, Cape

Province. Quaternaria 17: 237-243.

CARRINGTON, A. J. & KENSLEY, B. F. 1969. Pleistocene molluscs from the Namaqualand coast.

Ann. S. Afr. Mus. 52: 189-223.

CarTER, A. N. 1978. Phosphatic nodule beds in Victoria and the late Miocene—Pliocene

eustatic event. Nature, Lond. 276: 258-259.

CHURCHER, C. S. 1978. Giraffidae. In: Macuiio, V. J. & Cooxe, H. B. S. eds. Evolution of

African mammals: 509-535. Cambridge, Mass.: Harvard University Press.

CHURCHER, C. S. & RicHARDSON, M. L. 1978. Equidae. In: Mac.io, V. J. & Cooke, H. B. S.

eds. Evolution of African mammals: 379-422. Cambridge, Mass.: Harvard University

Press.

CorETzEE, J. A. 1978. Late Cainozoic palaeoenvironments of southern Africa. In: VAN

ZINDEREN BAKKER, E. M. ed. Antarctic glacial history and world palaeoenvironments:

115-127. Rotterdam: Balkema.

Cooke, H. B. S. & Wiikinson, A. F. 1978. Suidae and Tayassuidae. In: MAGLIO, V. J. &

Cooke, H. B. S. eds. Evolution of African mammals: 435-482. Cambridge, Mass.: Harvard

University Press.

Coprens, Y., Mactio, V. J., MADDEN, C. T. & BEDEN, M. 1978. Proboscidea. In: MaG.io, V.

J. & Cooke, H. B. S. eds. Evolution of African mammals: 336-367. Cambridge, Mass.:

Harvard University Press.

Corvinus, G. & HENDEy, Q. B. 1978. A new Miocene vertebrate locality at Arrisdrift in

Namibia (South West Africa). N. Jb. Geol. Palaont. Mh. 1978: 193-205.

CoryNpDon, S. C. 1978. Hippopotamidae. /n: MaG.io, V. J. & Cooke, H. B. S. eds. Evolution

of African mammals: 483-495. Cambridge, Mass.: Harvard University Press.

Davies, O. 1972. Pleistocene shorelines in the southern and south-eastern Cape Province (2).

Ann. Natal Mus, 21: 225-279.

Davies, O. 1973. Pleistocene shorelines in the western Cape and South West Africa. Ann.

Natal Mus. 21: 719-765.

Day, J. H. 1974. A guide to marine life on South African shores. Cape Town: Balkema.

DEAcon, H. J. 1979. Palaeocology. In: Day, J., SIEGFRIED, W. R., Louw, G. N. & JARMAN, M.

L. eds. Fynbos ecology: a preliminary synthesis: 58-69. Pretoria: S. Afr. natn. scient.

Progm. Rep. 40.

Detson, E. 1975. Paleoecology and zoogeography of the Old World monkeys. Jn: Tuttle, R.

H. ed. Primate functional morphology and evolution: 37-64. The Hague: Mouton.

DincLeE, R. V. 1971. Tertiary sedimentary history of the continental shelf off southern Cape

Province, South Africa. Trans. geol. Soc. S. Afr. 74: 173-186.

DINnGLE, R. V. 1973. The geology of the continental shelf between Ltideritz and Cape Town

(south-west Africa), with special reference to Tertiary strata. J. geol. Soc. Lond. 129:

337-363.

DINGLE, R. V., Lorp, A. R. & HENDEY, Q. B. 1979. New sections in the Varswater Formation

(Neogene) of Langebaan Road, south-western Cape, South Africa. Ann. S. Afr. Mus. 78:

81-92.

Dorst, J. & DANDELOoT, P. 1970. A field guide to the larger mammals of Africa. London:

Collins.

FUGGLE, R. F. & Asuton, E. R. 1979. Climate. In: Day, J., SIEGFRIED, W. R., Louw, G.N. &

JARMAN, M. L. eds. Fynbos ecology: a preliminary synthesis: 7-15. Pretoria: S. Afr. natn.

scient. Progm. Rep. 40.

PALAEOECOLOGY OF LANGEBAANWEG 101

Gentry, A. W. 1974. A new genus and species of Pliocene boselaphine (Bovidae, Mammalia)

from South Africa. Ann. S. Afr. Mus. 65: 145-188.

Gentry, A. W. 1978. Bovidae. In: MaG.io, V. J. & Cooke, H. B. S. eds. Evolution of African

mammals: 540-572. Cambridge, Mass.: Harvard University Press.

Gentry, A. W. 1980. Fossil Bovidae (Mammalia) of Langebaanweg, South Africa. Ann. S.

Afr. Mus. 79: 213-337.

GiLL, E.D. 1975. Evolution of Australia’s unique flora and fauna in relation to the plate

tectonics theory. Proc. R. Soc. Victoria 87: 215-234.

Go pE-PosseE, J.-M. 1977. Barberahyus castellensis n. g. n. sp., Tayasuido del Vindoboniense

terminal de Castell de Barbera (Cuenca del Vallés, Espana). Paleont. Evoluc. 12: 31-43.

GREENWAY, P. J. 1970. A classification of the vegetation of east Africa. Kirkia 9: 1-68.

Hatta, C. D. 1964. The geology of the coastal diamond deposits of southern Africa (1959).

In: Haucuton, S. H. ed. The geology of some ore deposits in southern Africa 2: 671-728.

Johannesburg: Geological Society of South Africa.

HamILTon, W. R. & VAN COUVERING, J. A. 1977. Lower Miocene mammals of South West

Africa. Namib Bull. (Suppl. 2, Transv. Mus. Bull.): 9-11.

Hag, B. U., BERGGREN, W. A. & VAN Couverinc J. A. 1977. Corrected age of the

Pliocene/Pleistocene boundary. Nature, Lond. 269: 483-488.

Harris, J. M. 1976. Pliocene Giraffoidea (Mammalia, Ariodactyla) from the Cape Province.

Ann. S. Afr. Mus. 69: 325-353.

Harris, J. M. & WuirtE, T. D. 1979. Evolution of the Plio-Pleistocene African Suidae. Trans.

Am. phil. Soc 69 (2): 1-128.

Harrison, A. D. 1964. An ecological survey of the Great Berg River. Jn: Davis, D. H. S. ed.

Ecological studies in southern Africa: 143-158. The Hague: W. Junk.

Harrison, A. D. & Etswortu, J. F. 1958. Hydrobiological studies on the Great Berg River,

Western Cape Province. Part 1. A general description, chemical studies on the water and

main features of the fauna and flora. Trans R. Soc. S. Afr. 35: 125-226.

HaucutTon, S. H. 1968. Geological history of southern Africa. Johannesburg: Geological

Society of South Africa.

Hayes, M. O. 1976 Beaches—barrier islands. In: Hayes, M. O. & Kana, T. W. eds. Terrigenous

clastic depositional environments: I 81-I 108. Columbia, SC: University of South Carolina.

Hayes, D. E., FRAKEs, L. A., BARRETT, P., BURNS, D. A., CHEN, P. H., Forp, A. S., KANEps,

A. G., Kemp, E. A., McCoLium, D. W., Piper, D. J. W., WALL, W. E. & WEsB, P. N.

1973. Leg 28 deep-sea drilling in the Southern Ocean. Geotimes 18 (6): 19-24.

HENDEY, Q. B. 1969. Quaternary vertebrate fossil sites in the south-western Cape Province. S.

Afr. archaeol. Bull. 24: 96-105.

HENDeEY, Q. B. 1970a. A review of the geology and palaeontology of the Plio/Pleistocene

deposits at Langebaanweg, Cape Province. Ann. S. Afr. Mus. 56: 75-117.

HENDEY, Q. B. 1970b. The age of the fossiliferous deposits at Langebaanweg, Cape Province.

Ann. S. Afr. Mus. 56: 119-131.

HENDEY, Q. B. 1972a. A Pliocene ursid from South Africa. Ann. S. Afr. Mus. 59: 115-132.

HENDEY, Q. B. 1972b. Further observations on the age of the mammalian fauna from

Langebaanweg, Cape Province. Palaeoecol. afr. 6: 172-175.

HENDEY, Q. B. 1973. Fossil occurrences at Langebaanweg, Cape Province. Nature, Lond. 244:

13-14.

HENDEY, Q. B. 1974a. The late Cenozoic Carnivora of the south-western Cape Province. Ann.

S. Afr. Mus. 63: 1-369.

HENpeY, Q. B. 1974b. Faunal dating of the late Cenozoic of southern Africa, with special

reference to the Carnivora. Quaternary Res. 4: 149-161.

HENDEY, Q. B. 1976a. The Pliocene fossil occurrences in ‘E’ Quarry, Langebaanweg, South

Africa. Ann. S. Afr. Mus. 69: 215-247.

HENDEY, Q. B. 1976b. Fossil peccary from the Pliocene of South Africa. Science 192: 787-789.

HENDEY, Q. B. 1977. Fossil bear from South Africa. S. Afr. J. Sci. 73: 112-116.

HENDEY, Q. B. 1978a. The age of the fossils from Baard’s Quarry, Langebaanweg, South

Africa. Ann. S. Afr. Mus. 73: 1=24.

HENpDEY, Q. B. 1978b. Preliminary report on the Miocene vertebrates from Arrisdrift, South

West Africa. Ann. S. Afr. Mus. 76: 1-41.

HENDEY, Q. B. 1978c. Late Tertiary Hyaenidae from Langebaanweg, South Africa, and their

relevance to the phylogeny of the family. Ann. S. Afr. Mus. 76: 265-297.

102 ANNALS OF THE SOUTH AFRICAN MUSEUM

HENDEY, Q. B. 1978d. Late Tertiary Mustelidae (Mammalia, Carnivora) from Langebaanweg,

South Africa. Ann. S. Afr. Mus. 76: 329-357.

HENpDEY, Q. B. 1980. Agriotherium (Mammalia, Ursidae) from Langebaanweg, South Africa,

and relationships of the genus. Ann. S. Afr. Mus. 81: 1-109.

HENDEY, Q. B. & REPENNING, C. A. 1972. A Pliocene phocid from South Africa. Ann. S. Afr.

Mus. 59: 71-98.

Hooyer, D. A. 1972. A late Pliocene rhinoceros from Langebaanweg, Cape Province. Ann. S.

Afr. Mus. 59: 151-191.

Hoover, D. A. 1976. The late Pliocene Equidae of Langebaanweg, Cape Province, South

Africa. Zool. Verh. 148: 1-39.

Hoover, D. A. 1978. Rhinocerotidae. In: MaGctio, V. J. & Cooke, H. B. S. eds. Evolution of

African mammals: 371-378. Cambridge, Mass.: Harvard University Press.

Hoover, D. A. & Macuio, V. J. 1974. Hipparions from the late Miocene and Pliocene of

northwestern Kenya. Zool. Verh. 134: 1-34.

Horwoop, A. T. 1928. New and little known mammals from the Miocene of Africa. Am. Mus.

Novit. 344: 1-9.

Hsu, K. J., Montapert, E., BERNOULET, D7, Cms, MB) ERICKSON; A“, GARRISONS Rem

Kipp, R. B., MELIERES, F., MULLER, C. & WriGuT, R. 1977. History of the Mediterranean

salinity crises. Nature, Lond. 267: 399-403.

Kemp, E. M. 1978. Tertiary climatic evolution and vegetation history in the southeast Indian

Ocean region. Palaeogeogr., Palaeoclimatol., Palaeoecol. 24: 169-208.

KENNETT, J. P. 1978. Cainozoic evolution of circumantarctic palaeoceanography. Jn: VAN

ZINDEREN BAKKER, E. M. ed. Antarctic glacial history and world palaeoenvironments:

41-56. Rotterdam: Balkema.

KENSLEY, B. 1972. Pliocene marine invertebrates from Langebaanweg, Cape Province. Ann. S.

Afr. Mus. 60: 173-190.

KENSLEY, B. 1973. Sea-shells of southern Africa. Gastropods. Cape Town: Maskew Miller.

KENSLEY, B. 1977. A second assemblage of Pliocene invertebrate fossils from Langebaanweg,

Cape. Ann. S. Afr. Mus. 72: 189-210.

Kine, J. E. 1964. Seals of the world. London: British Museaum (Natural History).

KLEIN, R. G. 1975. Paleoanthropological implications of the non-archaeological bone assembIl-

age from Swartklip 1, south-western Cape Province, South Africa. Quaternary Res. 5:

275-288.

KLEIN, R. G. 1980. Environmental and ecological implications of large mammals from Upper

Pleistocene and Holocene sites in southern Africa. Ann. S. Afr. Mus. 81: 223-283.

KOppPEN, W. & GEIGER, R. 1936. Handbuch der Klimatologie. Berlin: Borntraeger.

KRONINGER, H. 1978. National Lightning Recording Scheme. News/. natn. Elect. Engng Res.

Inst. (CSIR): 1-18.

KruGEr, F. J. 1979. Fire. In: Day, J., SIEGFRIED, W. R., Louw, G. N. & JARMAN, M. L. eds.

Fynbos ecology: a preliminary synthesis: 43-57. Pretoria. S. Afr. natn. scient. Progm. Rep. 40.

KRUMBEIN, W. C. & SLoss, L. L. 1963. Stratigraphy and sedimentation. San Francisco: W. H.

Freeman.

KurtTEN, B. 1971. The Age of Mammals. London: Weidenfeld & Nicolson.

Lamont, G. 1947. The geology of part of the Vanrhynsdorp Division, Cape Province.

University of Cape Town: Unpublished Ph. D. thesis.

Levyns, M. R. 1964. Migrations and origins of the Cape Flora. Trans. R. Soc. S. Afr. 37:

85-107.

Linpsay, E. H., OppykE, N. D. & JoHNnson, N. M. 1980. Pliocene dispersal of the horse Equus

and late Cenozoic mammalian dispersal events. Nature, Lond. 287: 135-138.

LoBREAU-CALLEN, D. & Suc, J.-P. 1972. Présence de pollens de Microtropus fallax (Celas-

traceae) dans le Pleistocéne inférieur de Celleneuve (Hérault). C.r. hebd. Séanc. Acad.

Sci. Paris (D) 275: 1351-1354.

Loutit, T. S. & KENNETT, J. P. 1979. Application of carbon isotope stratigraphy to Late

Miocene shallow marine sediments, New Zealand. Science 204: 1196-1199.

Massutt, J. A. 1956. The physiography and surface geology of the Hopefield fossil site. Trans.

R. Soc. S. Afr. 35: 21-58.

Maa uio, V. J. 1973. Origin and evolution of the Elephantidae. Trans. Am. phil. Soc. 63: 1-149.

Maac.io, V. J. 1978. Patterns of faunal evolution. In: MaGuio, V. J. & Cooke, H. B. S. eds.

Evolution of African mammls 603-619. Cambridge, Mass.: Harvard University Press.

PALAEOECOLOGY OF LANGEBAANWEG 103

Macuio, V. J. & HENDEy, Q. B. 1970. New evidence relating to the supposed stegolophodont

ancestry of the Elephantidae. S. Afr. archaeol. Bull. 25: 85-87.

McKenna, M. C. 1973. Sweepstakes, Filters, Corridors, Noah’s Arks, and Beached Viking

Funeral Ships in palaeogeography. Jn: TarLinc, D. H. & Runcorn, S. K. eds. Implications

of Continental Drift to the Earth Sciences 1: 295-308. London: Academic Press.

MCLACHLAN G.R. & LIVERSIDGE, R. 1978. Roberts birds of South Africa. Cape Town: The John

Voelcker Bird Book Fund.

McLacuian, I. R. & McMiLran, I. K. 1979. Microfaunal biostratigraphy, chronostratigraphy,

and history of Mesozoic and Cenozoic deposits on the coastal margin of South Africa.

Spec. Publs. Geol. Soc. S. Afr. 6: 161-182.

Mercer, J. H. 1978. Glacial development and temperature trends in the Antarctic and in South

America. In: VAN ZINDEREN BAKKER, E. M. ed. Antarctic glacial history and world

palaeoenvironments: 73-93. Rotterdam: Balkema.

Muizon, C. DE & HENDEY, Q. B. 1980. Late Tertiary seals of the South Atlantic Ocean. Ann.

S. Afr. Mus. 82: 91-128.

Otson, E. C. 1962. Late Permian terrestrial: vertebrates, USA and USSR. Trans. Am. phil.

Soc. 52 (2): 3-224.

PATTERSON, B., BEHRENSMEYER, A. K. & SILL, W. D. 1970. Geology and fauna of a new

Pliocene locality in north-western Kenya. Nature, Lond. 226: 918-921.

Pocock, T. N. 1976. Pliocene mammalian microfauna from Langebaanweg: A new fossil genus

linking the Otomyinae with the Murinae. S. Afr. J. Sci. 72: 58-60.

Ricu, P. V. 1980. Preliminary report on the fossil avian remains from Late Tertiary sediments

at Langebaanweg (Cape Province), South Africa. S. Afr. J. Sci. 76: 166-170.

Rocers, J. 1980. First report on the Cenozoic sediments between Cape Town and Elands Bay.

Rep. geol. Surv. S. Afr. 1980-165: 1-64. (Open file.)

SCHULZE, B. R. 1965. Climate of South Africa. 8. General Survey. Pretoria: Government

Printer.

SCHULZE, R. E. & McGee, O. S. 1978. Climatic indices and classifications in relation to

biogeography of southern Africa. Jn: WEeRGER, M. J. A. ed. Biogeography and ecology of

southern Africa: 19-52. The Hague: W. Junk.

Scott, J. H. & Symons, N. B. B. 1974. Introduction to dental anatomy. Edinburgh & London:

Churchill Livingstone.

SHACKLETON, N. J. & OppykeE, N. D. 1973. Oxygen isotope and palaeomagnetic stratigraphy of

Equatorial Pacific Core V28-238; Oxygen isotope temperatures and ice volumes on a 10°

and 10° year scale. Quaternary Res. 3: 39-55.

SIESSER, W. G. 1978. Age of phosphorites on the South African continental margin. Marine

Geol. 26: M17—M28.

SIESSER, W. G. & DINGLE, R. V. 1981. Tertiary sea-level movements around southern Africa.

J. Geol 89: 83-96.

Smmpson, G. G. 1967. The geography of evolution. New York: Capricorn Books.

Simpson, G. G. 1971. Fossil penguin from the late Cenozoic of South Africa. Science 171:

1144-1145.

Simpson, G. G. 1975. Notes on variation in penguins and on fossil penguins from the Pliocene

of Langebaanweg, Cape Province, South Africa. Ann. S. Afr. Mus. 69: 59-72.

Simpson, G. G. 1976. Penguins. Past and present, here and there. New Haven: Yale University

Press.

Simpson, G. G. 1979. A new genus of late Tertiary penguin from Langebaanweg, South Africa.

Ann. S. Afr. Mus. 78: 1-9.

SINGER, R. 1961. The new fossil sites at Langebaanweg (South Africa). Curr. Anthrop 2:

385-387.

SINGER, R. & Hoover, D.A. 1958. A Stegolophodon from South Africa. Nature, Lond. 182:

101-102.

STROMER, E. 1926. Reste land- und stisswasser bewohnender Wirbeltiere aus den Diamantenfel-

dern Deutsch-Stidwestafrikas. Jn: KaAIsER, E. ed. Die Diamantenwiiste Stidwestafrikas 2:

107-153. Berlin: D. Reimer.

STROMER, E. 193la. Reste stisswasser- und land bewohnender Wirbeltiere aus den Diamanten-

felden Klein-Namaqualandes (Stidwest-afrika). Sber. bayer Akad. Wiss. 1931: 17-47.

StroMER, E. 1931b. Palaeothentoides africanus, nov. gen., nov. spec., ein erstes Beuteltier aus

Afrika. Sber. bayer. Akad. Wiss. 1931: 177-190.

104 ANNALS OF THE SOUTH AFRICAN MUSEUM

TANKARD, A. J. 1974a. Varswater Formation of the Langebaanweg-—Saldanha area, Cape

Province. Trans. geol. Soc. S. Afr. 77: 265-283.

TANKARD, A. J. 1974b. Petrology and origin of the phosphate and aluminium phosphate rock of

the Langebaanweg—Saldanha area, south-western Cape Province. Ann. S. Afr. Mus. 65:

217-249.

TANKARD, A. J. 1975a. The marine Neogene Saldanha Formation. Trans. geol. Soc. S. Afr. 78:

257-264.

TANKARD, A. J. 1975b. The late Cenozoic history and palaeoenvironments of the coastal

margin of the south-western Cape Province, South Africa. Rhodes University: Unpu-

blished Ph. D. thesis.

TANKARD, A. J. 1975c. Varswater Formation of the Langebaanweg—Saldanha area, Cape

Province. (Discussion.) Trans. geol. Soc. S. Afr. 78: 373-375.

TANKARD, A. J. 1975d. Thermally anomalous Late Pleistocene molluscs from the south-western

Cape Province, South Africa. Ann. S. Afr. Mus. 69: 17-45.

TANKARD, A. J. 1976. Pleistocene history and coastal morphology of the Ysterfontein—Elands

Bay area, Cape Province. Ann. S. Afr. Mus. 69:73-119.

TANKARD, A. J. & RoGers, J. 1978. Late Cenozoic palaeoenvironments on the west coast of

southern Africa. J. Biogeogr. 5: 319-337.

TayLor, H. C. 1978. Capensis. In: WeRGER, M. J. A. ed. Biogeography and ecology of southern

Africa: 171-229. The Hague: W. Junk.

Tuomas, H. 1979. Miotragocerus cyrenaicus sp. nov. (Bovidae, Artiodactyla, Mammalia) du

Miocéne supérieur de Sahabi (Libye) et ses rapports avec les autres Miotragocerus.

Geobios 12: 267-281.

VaIL, P. R. & HARDENBOL, J. 1979. Sea-level changes during the Tertiary. Oceanus 22: 71-79.

VaIL, P. R. Mitcuum, R. M. & THompson, S. 1977. Seismic stratigraphy and global changes in

sea level. Part 4: Global cycles of relative changes of sea level. Mem. Am. Ass. Petrol.

Geol. 26: 83-97.

VAN COUVERING, J. A. 1972. Radiometric calibration of the European Neogene. /n: BisHop, W.

W. & MiLieR, J. A. eds. Calibration of hominoid evolution: 247-271. Edinburgh: Scottish

Academic Press.

VAN CoUuveRING J. A. 1978. Status of late Cenozoic boundaries. Geology 6: 169.

VAN COUVERING, J. A. & BERGGREN, W. A. 1977. Biostratigraphical basis of the Neogene time

scale. In: KAUFFMANN, E. G. & HazeEL, J. E. eds. Methods of biostratigraphy: 283-306.

Stroudsburg, Pa.: Dowden, Hutchinson & Ross.

VAN COUVERING, J. A. BERGGREN, W. A., DRAKE, R. E., AGuirrE, E. & Curtis, G. H. 1976.

The terminal Miocene event. Mar. Micropaleontol. 1: 163-286.

Visser, H. N. & ScHocn, A. E. 1973. The geology and mineral resources of the Saldanha Bay

area. Mem. geol. Surv. S. Afr. 63: 1-150.

VooruieEs, M. R. 1969. Taphonomy and population dynamics of an Early Pliocene vertebrate

fauna, Knox County, Nebraska. Univ. Wyoming, Contr. Geol., Spec. Paper 1: 1-69.

Wess, S. D. 1977. A history of savannah vertebrates in the New World. 1. North America. A.

Rev. Ecol. Syst. 8: 355-380.

Wuite, T. D. & Harris, J. M. 1977. Suid evolution and correlation of African hominid

localities. Science 198: 13-21.

Wo rr, R. G., SINGER, R. & BisHop, W. W. 1973. Fossil bear (Agriotherium Wagner, 1837)

from Langebaanweg, Cape Province, South Africa. Quaternaria 17: 209-236.

i x

‘ in

“

4 7 » ; ; a

’ : 5 Ri ii a 5 iy 4

%: 4 & at

Vb it =) i ;

» by L i ‘

ors *

y ~t t ur

Sai . a i

a oi ‘

{ C = ;

y ‘ i

vf é 3 =

your e ’

i i > fic a «

pane 5 i ‘

> . “i — ¢

\\ t . i

x a i} 7

i ‘ at at

4 i ) f

i — i 8

i) aA ; hy vi

( 5 ; By ‘

oe ; A is :

} ee, d

(ee

A | t

; vy o %,

sg” y 2 ‘3 i

Y ail ad J F }

i i < j ;

Die , =

i oo A i

£ ‘

‘ 5 .

K P t

Ea i i

. ¢

~ s

) ~ rk

Ve ry : s ‘

* =~ ‘

? i

Li ‘

2 -

: a 5 ‘

2 es i

; &) ene

5 i a 5

' és

' bt

\ fe i

a j a fi

r >

2 }

a .

bs iS

, > ~ rt eT,

9 ~

es YY er

aa \ St

re J

r ‘ 2

q t

| frat

¥ ‘ !

} ms Ny é J 1 i

} PB Ke i

{ {

e ? Po

Fi ct &

aw 1

B i. te

4 ’ t

ra t v

is ¢

I , =X)

a i % W ey -

ji E oF f 2

j :

7 A

by 1 Gi ‘] 2

{ i.

ro) # 4

6. SYSTEMATIC papers must conform to the International code of zoological nomenclature

(particularly Articles 22 and 51).

Names of new taxa, combinations, synonyms, etc., when used for the first time, must be

followed by the appropriate Latin (not English) abbreviation, e.g. gen. nov., sp. nov., comb.

nov., syn. nov., etc. .

An author’s name when cited must follow the name of the taxon without intervening

punctuation and not be abbreviated; if the year is added, a comma must separate author’s

name and year. The author’s name (and date, if cited) must be placed in parentheses if a

species or subspecies is transferred from its original genus. The name of a subsequent user of

a scientific name must be separated from the scientific name by a colon.

Synonymy arrangement should be according to chronology of names, i.e. all published

scientific names by which the species previously has been designated are listed in chronological

order, with all references to that name following in chronological order, e.g.:

Family Nuculanidae

Nuculana (Lembulus) bicuspidata (Gould, 1845)

Figs 14-15A

Nucula (Leda) bicuspidata Gould, 1845: 37.

Leda plicifera A. Adams, 1856: 50.

Laeda bicuspidata Hanley, 1859: 118, pl. 228 (fig. 73). Sowerby, 1871: pl. 2 (fig. 8a—b).

Nucula largillierti Philippi, 1861: 87.

Leda bicuspidata: Nicklés, 1950: 163, fig. 301; 1955: 110. Barnard, 1964: 234, figs 8-9.

Note punctuation in the above example:

- comma separates author’s name and year

semicolon separates more than one reference by the same author

full stop separates references by different authors

figures of plates are enclosed in parentheses to distinguish them from text-figures

dash, not comma, separates consecutive numbers

Synonymy arrangement according to chronology of bibliographic references, whereby

the year is placed in front of each entry, and the synonym repeated in full for each entry, is

~ not acceptable.

In describing new species, one specimen must be designated as the holotype; other speci-

mens mentioned in the original description are to be designated paratypes; additional material

not regarded as paratypes should be listed separately. The complete data (registration number,

depository, description of specimen, locality, collector, date) of the holotype and paratypes

must be recorded, e.g.:

Holotype

SAM-A13535 in the South African Museum, Cape Town. Adult female from mid-tide region, King’s Beach

Port Elizabeth (33°51’S 25°39’E), collected by A. Smith, 15 January 1973.

Note standard form of writing South African Museum registration numbers and date.

7. SPECIAL HOUSE RULES

Capital initial letters

(a) The Figures, Maps and Tables of the paper when referred to in the text

e.g. ‘.. . the Figure depicting C. namacolus...’; ‘. .. in C. namacolus (Fig. 10)...’

(b) The prefixes of prefixed surnames in all languages, when used in the text, if not preceded

by initials or full names

e.g. Du Toit but A.L.du Toit; Von Huene but F. von Huene

e.g. Therocephalia, but therocephalian

Punctuation should be loose, omitting all not strictly necessary

Reference to the author should be expressed in the third person

Roman numerals should be converted to arabic, except when forming part of the title of a

book or article, such as

‘Revision of the Crustacea. Part VIII. The Amphipoda.’ a

Specific name must not stand alone, but be preceded by the generic name or its abbreviation

to initial capital letter, provided the same generic name is used consecutively.

Name of new genus or species is not to be included in the title: it should be included in the

abstract, counter to Recommendation 23 of the Code, to meet the requirements of

Biological Abstracts.

Q. B. HENDEY

PALAEOECOLOGY OF THE LATE

TERTIARY FOSSIL OCCURRENCES IN

‘FE’ QUARRY, LANGEBAANWEG,

SOUTH AFRICA, AND A REINTERPRETATION

OF THEIR GEOLOGICAL CONTEXT

OF THE SOUTH AFRICAN

~~ MUSEUM

CAPE TOWN.

INSTRUCTIONS TO AUTHORS

1. MATERIAL should be original and not published elsewhere, in whole or in part.

2. LAYOUT should be as follows:

(a) Centred masthead to consist of

Title: informative but concise, without abbreviations and not including the names of new genera or species

Author’s(s’) name(s)

Address(es) of author(s) (institution where work was carried out)

Number of illustrations (figures, enumerated maps and tables, in this order)

(b) Abstract of not more than 200 words, intelligible to the reader without reference to the text

(d) Introduction

(e) Subject-matter of the paper, divided into sections to correspond with those given in table of contents

(f) Summary, if paper is lengthy

(g) Acknowledgements

(h) References

(i) Abbreviations, where these are numerous

3. MANUSCRIPT, to be submitted in triplicate, should be typewritten and neat, double spaced

with 2,5 cm margins all round. First lines of paragraphs should be indented. Tables and a list of

legends for illustrations should be typed separately, their positions indicated in the text. All

pages should be numbered consecutively.

Major headings of the paper are centred capitals; first subheadings are shouldered small

capitals; second subheadings are shouldered italics; third subheadings are indented, shouldered

italics. Further subdivisions should be avoided, as also enumeration (never roman numerals)

of headings and abbreviations.

Footnotes should be avoided unless they are short and essential.

Only generic and specific names should be underlined to indicate italics; all other marking

up should be left to editor and publisher.

4. ILLUSTRATIONS should be reducible to a size not exceeding 12 « 18 cm (19 cm including