o X»N PCX ~ O Xj*. PCX “ Q

ITHSONIAN^INSTITUTION NOIlfUUSNI "NVINOSHIUMS S3 I d Vd a I T L I B RAR I ES^SMITHSONIAN-'lNSTITU

VA CD

'/%/ ^ ffi

i > pf

^ v\

m xg»os*tx ^ x^vasv^ m ^ m HI m x^vosw

CD = CD ± w £ CD

inoshiiins saiavaan libraries Smithsonian institution NouniusNi nvinoshiums S3iav*

CD 2 CD 2 * CD Z CD

gfflj/ >' 2 s •' 5'- 5

ITHSONIAN INSTITUTION NOIlfUUSNI NVINOSHIIIMS^SB 1 a Vd 9 II^LI B RAR I ESWSMITHSONIAN~INSTITU

— * ID — V CO *=-

c n

z* -i

/ 5 to

o -vr = 5 v<^^/ 2-*' ' ' 5 v“

IN0SH1IWSZS3 I avaa n^LI BRAR 1 ES SMITHSONIAN INSTITUTION NOIiniUSNI^NVINOSHliWS^Sa I a VI
-5- r~ z r~ z f z

— O O Xvtsv/i

03
30
>

S3

m — <sgg^ m - n^jiiw rn

_ £D _ CD £ ' 03 _

ITHSONIAN INSTITUTION NOIlfUUSNI NVINOSHIUMS S3iaVaail LIBRARIES SMITHSONIAN INSTITL

Z v- CD Z __ _ W Z CD 2

< j£k^ S < 2 ..< jK\* s

z ^ ^ <£

I |

CD jF CD

_ vx S

5 ^ 2 3,, ^ '\ 35. x^utisgx s

CD V 2 CD Z CD V Z jg

INOSHIIINS S3 I ava a ll_ L! B RAR I ES SMITHSONIAN_INSTITUTIONOTNOIinillSNI_NVINOSHllWS S3 I BV

-s 2 5 S : 5 2 5 S

I ITHSONIAN""* INSTITUTION NOlifUllSNI^NVINOSHillNS S3 1 H VH 9 11 "*11 B R AR I ES^SMITHSONIAN^INSTITU

z r~ z r- z «“

inoshiiins saiavaan libraries Smithsonian institution NouniusNi nvinoshiums saiavi

CD Z CD Z * CD Z CD

2 < | z -■% 1 1 1 ^

s >cw 1 CJP 1 'w I.*5* I IOI !/#

Jr X^VAjt^X £5 > ' 2 Xfoofv<X^ > ■

ITHSONIAN INSTITUTION ^ NO liniliSN I NVINOSHIUMS^ S 3 I dVd 9 ll^LI B RAR I ES^SMITHSONIAN SNSTITU

CD 5 \ CD = CD

0 Xj'osi'jx' — tJ " .>3^ _ O

INOSHlIIMS^Saiavaan^LIBRARIES SMITHSONIAN^INSTITUTION NOIlfUUSNI NVINOSHIUMS S3IHV;

1 m ~ . . A-S a o

.« Of

w

q Xjjjy pc^x niti' o

ldillSNI^NVINOSHimS S3 ! HVH0 II ^LIBRAR ! ES^SMITHSONIAN^INSTITUTIOn'" NOlinillSNI^NVINOSH,
r- * Z r- z in z r-

m Xi^osv^x ^ m x^vosv^x ^ rn

o> ' — w 5 c/> _ 05

SRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIIASS S3iaVHan LIBRARIES SMITHSOI

O) Z , 05 Z O) Z •, 05

r >

x /?r
o

>'^PN > 5 ■ ' > ' 5 Xgyo fov>X >• 5

lninsNi nvinoshiiwswS3 s avaan librar i es^smithsqnian institution ^noiidiiisni^nvinosh.

■ - </> — gn —• o) ~

05

LU

/■" H m

0 5 5 2 xjvA?>gy O

3 RAR I ES^ SMITHSONIAN INSTlTUTlGN^NQIinillSNS^MWINOSHIIlAIS^ $3 !HWHanILIBRAR! kS^SMlTHSO

1 /V1’ I /«gg&v 5 ^5 2 ti££& 5 §

°° 03
^%U H

— 05 £ — w \ Z —

IlfUIlSNI NVIN0SH1IWS SBiaVaail LIBRARIES SMITHSONIAN "INSTITUTION NOlinillSNI NVINOSH

(/)

I ^

/ <yf p; /yj£V ' ' PX^\ -r ''/ti ///ti o — /6*/SiCfFKi?

(//r Z/ti 9 (>M f!'5l m /JW'.tiW m

a ^

\ 2j X-fr)mst£X ^ > ‘

3 RAR I ES1" SMITHSONIAN^ INSTITUTION^NOIiniUSNI NVINOSHIIMS^SS I HVH 8 SI L1 8 RAR 1 ES^SMITHSO

Of -g 05 . ~ . O) 2 ^

~ „ ,«W w .^V w w ~

^ 5 x^oc^ “ 5 ' ^ n

inillSNI^NVINOSHHIAiS S3IUVaai1LIBRARI ES^SMITHSONIAN^INSTITUTION ** NOlinillSNI “"nVINOSH.

ir -3>* — r? r-=* T?' r~

3RARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIIAIS S3iavaan LIBRARIES SMITHSOI

'« — - ^ 05

<

z

^ o

^ 05

f- X

CO

2 _

x

05 .. „

O W

s §v s

linillSNI^NWINOSHHIAjS^Sa I avaa n^LI BRAR I ES^SMITHSONIAN ^INSTITUTION NOIiniUSNI_NVINOSH.

— \ O) — OT — o) 2

^ ^ S I - A. 1

<».w = S W&. .S3W s «r«5.- 5 ¥W,, - ™

3 RAR I ES SMITHSONIAN^INSTITUTION^NOIinillSNI^NVINOSHllWS^Sa IHVaail^ LIBRARIES SMITHSO

^ 2 r~ z r= -z r~ Z

^spT^v o - O ^s5vsx - 4A-'.. O 3 XxIlvx o

c

ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

Ann. Cape Prov. Mus. (nat. Hist.)

VOLUME 13 • PART 1
8th MAY 1979

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

Printed by Cape & Transvaal Printers (Pty.) Ltd., Cape Town

BD8154

Notes on the elevation of Rana fasciata montana FitzSimons, 1946 to specific
rank, and on the identity of Rana fasciata sensu Burchell, 1824 (Anura:

Ranidae)

J. COMRIE GREIG

Cape Department of Nature and Environmental Conservation,

Jonkershoek Nature Conservation Station,

Stellenbosch, South Africa

R. C. BOYCOTT

Port Elizabeth Museum,

Port Elizabeth, South Africa

and

A. L. DE VILLIERS

Cape Department of Nature and Environmental Conservation,

Jonkershoek Nature Conservation Station,

Stellenbosch, South Africa

CONTENTS

Page

Abstract/Opsomming 2

Introduction 2

Differentiation between Rana fasciata and Rana montana stat. nov 2

Vocalisation 2

Morphological features 8

General 8

Length of snout 8

Extent of webbing 8

Length of foot 9

Length of first toe 10

Preorbital caruncle 10

Colour markings 11

Size and sexual dimorphism 13

Morphological separation from R. fasciata and R. grayii 13

Distribution 15

The type-localities 22

Life history notes 23

The name Rana montana 24

Taxonomic history of the name Rana fasciata 25

Acknowledgements 29

References 29

1

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 1, MAY 1979
OPSOMMING

Die Suid-Afrikaanse padda van die Attakwaberge wat in 1946 as Rana fasciata montana
beskryf is, word hier tot spesifieke rang verhef. Die kenmerkende roep en morfologiese verskille
onderskei Rana montana stat. nov. ongetwyfeld van Rana fasciata Smith. Simpatrie van die
twee spesies by drie lokaliteite ondersteun die verandering van rang van die spesies. Aandag
word gevestig op die feit dat die naam Rana fasciata Burchell, nou onderdruk, waarskynlik na
Rana fuscigula Dumeril en Bibron verwys en nie na Rana grayii soos voorheen voorgestel nie.

ABSTRACT

The South African frog described from the Attakwa Mountains in 1946 as Rana fasciata
montana is here raised to specific rank. Distinct call and morphological differences separate
Rana montana stat. nov. unequivocally from Rana fasciata Smith. Sympatry of the two species
at three localities confirms the species rank separation. It is pointed out that the name Rana
fasciata Burchell, now suppressed, is probably referable to Rana fuscigula Dumeril and Bibron
and not to Rana grayii as previously suggested.

INTRODUCTION

In 1946, the late V. F. M. FitzSimons described a new subspecies of Rana fasciata from the
southern Cape Province, South Africa. The types were collected on the “upper slopes’ ’of the
Attaqua Mountains and the Langeberge, and the subspecies was accordingly dubbed Rana
fasciata montana. After examining series of specimens from all over southern Africa, Poynton
(1964) assigned most south-west and southern Cape material to this subspecies, thus extending its
altitudinal and geographical limits from the mountain slopes of the Jonkersberg to sea-level near
Cape Town.

In June 1976 the authors made a joint collecting expedition to Kleinmond near Cape
Hangklip, an area of outstanding botanical and zoological interest, with the object of compiling a
checklist of local anurans. One of us (J.C.G.) was familiar with the melodious piping call of
Rana fasciata fasciata from the eastern Cape, and queried the identification by R.C.B. of a
lower-pitched creaking call as the same species. When a male ‘R. fasciata montana” emitting
this creaking call was eventually tracked down, it was apparent to both of us that eastern and
western Cape forms were separable at species level, and that Rana fasciata montana would have
to be elevated to full specific rank. This paper represents observations on both Rana fasciata and
Rana montana stat. nov.

DIFFERENTIATION BETWEEN RANA FASCIATA AND RANA MONTANA STAT. NOV.
VOCALISATION

Mating calls in the Anura are generally recognised as being species-specific, their principal
function being to guide gravid females to conspecific males (Blair 1958; Pengilley 1971). They
are regarded as sensitive taxonomic tools which can be used with confidence in the separation of
species, even the morphologically similar forms known as sibling species (Passmore 1977;
Passmore and Carruthers 1975).

The calls of the two species under discussion are totally dissimilar. Poynton (1964)
describes the call of Rana fasciata Smith as a “clear, high-pitched ‘pip’ ... a familiar sound
near streams on a frosty night.’’ Wager (1965) describes it as a “high-pitched, short, melodious
whistle like ‘whip’, usually single or double and at long intervals, but sometimes repeated rapidly

2

COMRIE GREIG, BOYCOTT & DE VILLIERS: ELEVATION OF RAN A FASCIATA MONTANA

a number of times, almost like a tinkle.” Stewart (1967) notes that the call of the subspecies
fuelleborni in Malawi is a “loud, musical ‘put, put, put’, developing into a melodic chorus
during the day or night during most of the year . . . The call may be single, or repeated several
times as a musical ripple”.

Unlike these authors, Walter Rose was based in the western Cape, the range of Rana
montana. In his popular account of southern African herpetology (1962), there is evidence of
confusion over the calls; he photographically illustrates both “Rana fasciata montana ” and R.
fasciata fasciata, and describes the call of the “species” as follows: “The mating call is a
high-pitched ‘Wheet, wheet’. At other times it now and again utters a slow ‘Chuck, chuck,
chuck’.” The first description is clearly that of the call of Rana fasciata', the second is equally
certainly of the call of Rana montana .

The “chuck, chuck, chuck” could well be used as a verbal description of the call of Rana
montana. It never develops into the aptly named “musical ripple” of the R. fasciata mating
chorus. It is medium-pitched, relatively infrequently produced, and could be described as a short
explosive creak. Onomatopoetic descriptions of frog calls, however, are notoriously subjective,
and a “short explosive creak” to one observer could well be rendered by another as a “staccato
belch”. The calls are therefore best graphically represented as oscillograms or sonograms
(occasionally spelt “sonagram”) for objective comparison between species.

The oscillograms and sonograms shown below were derived from recordings made in the
field on a Uher 4400 Report Stereo 1C portable tape-recorder at a tape speed of 19 cm/s, using an
AKG D900 directional microphone. The oscillograms were obtained using a Biomation Model
1015 Waveform Recorder and a Tektronix D12 Oscilloscope. The sonograms were obtained
after analysis on a sound spectrograph (Kay model 7029A spectrum analyser). The frequency
range was 80-8 000 Hz and a wide band filter (300 Hz) was used.

The mating call of Rana montana is illustrated in the oscillograms in Figs 1, 2 and 3. These
particular calls were recorded at Kleinmond (3419 AC Hermanus) on 25 July 1978. It can be
clearly seen from Figs 1 and 2 that the call consists of a series of closely-grouped pulses. The call
durations are respectively 80 ms and 60 ms (although we have found the duration may extend to
120 ms). Fig. 3 displays the detail of the first 2Vi pulses of the call shown in Fig. 2.

Fig. l

Rana montana

oscillogram of single mating call
(whole frame 200 ms)

3

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 1, MAY 1979

Rana montana calls are produced slowly and deliberately, often several seconds separating
each. From tape-recordings made at Kleinmond and Jonkersberg, we calculated the minimum call
interval at 1,3 seconds, although this would slow down to 3 to 10 seconds between faster bouts of
calling. One fast sequence was noted as 1,8; 2,3; 1,9; 1,6; 2,0; 2,0; 1,6; 1,3; 1,4; 1,6 seconds.

Rana fasciata, however, produces calls which are almost pure tones with no subdivision
into pulses. Fig. 4 shows a portion of a call sequence recorded on 26 July 1978 at Buffeljags-
rivier (3420 BA Suurbraak) on the same scale as the montana calls in Figs 1 and 2. Call
duration here is about 20 ms, the “pip”, “put”, or “whip” of earlier writers. When this

4

COMRIE GREIG, BOYCOTT & DE VILLIERS: ELEVATION OF RAN A FASCIATA MONTANA

particular “ripple” of five calls is compressed on to one oscillogram (Fig. 5), it can be seen that
the temporal separation of the calls is remarkably uniform, about 80-90 ms between each. A
single call expanded to show structural detail is illustrated in Fig. 6; this is directly comparable in
scale with Fig. 3 and shows the symmetrical nature of the single call.

The sonogram of a single call of Rana montana is shown in Fig. 7; its duration is 75 ms,
and its frequency range is 2,7 kFlz-3,25 kHz. This call was also recorded at Kleinmond (3419
AC Hermanus).

5

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT I, MAY 1979

Fig. 6

Rana fasciata

enlargement of one call from “rip-
ple” in Fig. 5 (whole frame 20 ms)

Fig. 8 is a sonogram of two successive calls of Rana fasciata from Port Elizabeth (3325 DC
Port Elizabeth); these calls are not part of a “ripple” sequence, and consequently have longer
durations of 30 ms and 20 ms respectively, with a frequency range of 3,25 kHz-4,0 kHz.

Eight calls from a “ripple” sequence of R. fasciata are shown in Fig. 9. They are more
staccato than the single calls, lasting only 15-17 ms each, with a slightly contracted frequency
range of 3,25 kHz-3,83 kHz. The gaps between the calls of the ripple vary between 90 and 115
ms.

Fig. 7

Rana montana

sonogram of single call using wide-
band (300 Hz) filter

6

COMRIE GREIG, BOYCOTT & DE VILLIERS: ELEVATION OF RAN A FASC1ATA MONTANA

Rana fascicita

sonogram of 8 calls in a “ripple” using wideband (300 Hz) filter

7

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 1, MAY 1979

The calls of the two species are clearly distinctive as shown by both oscillograms and
sonograms. Rana montana does not produce a rapid succession of calls, the frequency range is
lower than in R. fasciata, and each call is subdivided into pulses, giving rise to a creaking effect.
There can be no question but that the call difference alone can be used to separate the two
species.

MORPHOLOGICAL FEATURES
General

The main features used by FitzSimons (1946) to distinguish the subspecies montana from
fasciata were:

(1) the longer, more pointed snout of montana

(2) reduced webbing on feet of montana

(3) proportionately shorter feet of montana

(4) colour markings

We do not consider that the snout can be demonstrated unequivocally to be longer or more
pointed, nor that the foot can be shown to be proportionately shorter. We do, however, add two
other diagnostic characters to replace those two of FitzSimons.

(1) proportionately longer first toe of montana

(2) presence of a preorbital caruncle in montana

Length of snout

FitzSimons (op. cit.) described the snout of montana as “sharply pointed and strongly
projecting, its length at least twice diameter of eye (in typical fasciata , snout distinctly less than
twice diameter of eye)”; he also stated that the snout length is not more than 2,1 times into the
distance between snout and vent, as opposed to 2,2 to 2,3 times in fasciata.

We do not agree that the snout of montana is obviously longer or more sharply pointed. In
certain individuals it may be proportionately longer than in typical R. fasciata but in such cases
extra length is in our view not quantifiable. The length of snout/diameter of eye ratio is not a
valid point of difference, and we find it difficult to understand FitzSimons’ statement that the
snout- vent length/snout length ratio is 2,1 in montana and 2,3 in fasciata, when this ratio is
normally of the order of 5,4 to 6,1, and does not appear to us to present a clear-cut species
difference.

Extent of webbing

FitzSimons stated that the webbing on the feet does not extend as far as the basal
subarticular tubercles of the digits, while in R. fasciata the webbing extends as far as the basal
tubercles. Loveridge (1953), discussing the validity of R. fasciata fuelleborni , examined 119
specimens of various races of the species (including an unspecified, but presumably very small
number of "R. fasciata montana ”) and asserted that the extent of webbing on the toes was the
same in all races, viz. from the first to the fifth, the free phalanges were 2, 2, 3, 4, and 3. Inger
(1959) adduced this statement as additional evidence for the rejection of montana as a distinct
subspecies. Poynton (1964), however, states that in montana, four phalanges of the fourth toe are
free of web, compared with “3% to 4” in R. fasciata .

In all 80 Rana montana specimens examined, we have found that the webbing does not
reach the basal subarticular tubercle on the fourth toe; this means in effect that four phalanges of
the fourth toe are free of web (Fig. 10a). In Rana fasciata the webbing in all specimens
examined (from the S. and E. Cape) clearly extends beyond this tubercle, but not, however, as far

8

COMRIE GREIG, BOYCOTT & DE VILLIERS: ELEVATION OF RAN A FASCIATA MONTANA

5

Fig. 10b

Left foot of R. montana showing extent of webbing Left foot of R. fasciata showing extent of webbing

as half-way along the phalanx (Fig. 10b), thus perhaps contributing to the confusion concerning
the extent of the webbing, which is normally expressed in units of whole or half phalanges. In
some specimens the web just reaches the tubercle on the side adjacent to toe 3, but in these cases
the web on the side adjacent to toe 5 still extends beyond the tubercle. Reduction in webbing is
also apparent at the basal subarticular tubercle of toe 5; in montana the web reaches this tubercle,
while in fasciata it extends beyond it.

Another feature noticeable in fasciata but less so in montana is the continuation of the web
to the tip of each toe in the form of a narrow flange. We define the web as commencing when the
edge of the flange ceases to run parallel to the digit.

As FitzSimons (1946) claimed, therefore, there is a slight but significant and consistent
reduction in the extent of webbing on the feet of Rana montana in comparison with R. fasciata.

Length of foot

Again it is necessary to draw attention to an apparent contradiction in FitzSimons’
description. He states that montana has a proportionately shorter foot than fasciata , but goes on
to say that the length of the foot of montana goes less than 1,1 times into the snout/ vent length,
while in fasciata it is over 1,1 times. These statements are clearly contradictory.

After measuring a number of specimens of both species, we conclude that there is no
marked difference between the species in proportionate length of the feet.

9

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 1, MAY 1979

Poynton (1964) compares foot length (distance from the proximal end of the metatarsal
tubercle to the tip of the fourth toe) with the urostyle/tympanum or urostyle/eye length. Both R.
fasciata and R. montana are variable in this character, and in both, the length of the foot is equal
to the distance from the tip of the urostyle to the anterior margin of the tympanum or to the eye
(this character may be used to compare these species with R. grayii where the foot length is
equivalent to the distance between urostyle tip and the axilla, or rarely the posterior margin of
the tympanum).

Length of first toe

When the first toe of Rana montana is adpressed parallel to the second toe, the tip of the toe
normally reaches to the middle of the subarticular tubercle at the base of the second toe, and
frequently to the distal end of the tubercle. In our largely eastern and southern Cape sample of R.
fasciata , the tip of the first toe usually reaches only to the proximal end of the same tubercle, and
frequently falls short of it.

Preorbital caruncle

At the anterior comer of the eye of Rana montana is a small fleshy lobe, which is absent in
R. fasciata. Although the word “caruncle” is perhaps most commonly associated with hard or
solid structures, its correct application is to small fleshy excrescences, and we therefore use the
term “preorbital caruncle” to describe this character.

In Rana fasciata , the skin over the top of the orbit (the upper eyelid) meets the skin of the
snout at the anterior comer of the orbit in a fold or shallow groove. The fold presumably caters
for expansion and contraction of the skin in movements of the eye (Fig. lib). There is a small
notch in the fold next to the eye itself.

Rana montana, on the other hand, possesses two folds about 0,5 mm apart. The skin tissue
between the folds appears to be slightly swollen, hence the caruncle effect (Fig. 11a). The upper
fold, which is analogous to the single fold of R. fasciata , is more deeply incised next to the eye,
the incision being overlapped by the caruncle like a flap. The lower fold is also deeply incised
but the two sides of the fold are adpressed together and do not overlap. In only one specimen, out
of 80 examined, was there any doubt about the presence of the lower fold, but in that specimen
the other preorbital caruncle was clearly visible. (The caruncle is also visible in Fig. 12).

Fig. 11a

Preorbital caruncle of R. montana

10

Fig. lib

Preorbital groove of R. fasciata

COMR1E GREIG, BOYCOTT & DE VILLIERS: ELEVATION OF RAN A FASCIATA MONTANA

Colour markings

The colour markings of Rana montana have been described in some detail by FitzSimons
(1946). His description was based, however, on a small sample of six specimens and he did not
explain clearly how montana differs from fasciata. The description which follows is based on 80
specimens, the majority of which were collected by officers of the Cape Department of Nature
and Environmental Conservation. The occurrence of colour pattern polymorphism in R. montana
is here recorded for the first time. The colour markings of the dominant morph are illustrated in
Fig. 12.

A mesial vertebral stripe is always present, from behind the nostrils (rarely tip of snout or
eyes) to the tip of the urostyle, either one third of the width between the two sacral hump
projections, or rarely much broader (as wide as, or wider than the sacral hump), usually silver,
yellow or orange-brown but often silver suffused mesially with orange-red. The narrow stripe
morph has this stripe bounded by dark brown stripes of the same width, broken for the posterior
third of their length, and rarely along their entire length. A silver dorsolateral stripe from the
axilla to groin is bounded by shorter dark brown stripes above and below. The dorsal ground
colour is usually light brown or olive-grey. Occasionally a rufous stripe may be present between
the dark brown vertebral stripe and the upper dark brown dorsolateral stripe above the axilla, but
this does not extend as far as the sacral hump. A few raised white spots are scattered over the
flanks posteriorly. A rufous spot is occasionally present mesially to the anterior comer of the eye.
The upper surface of the upper arm is often tinged with rufous. A dark brown stripe runs from
the snout through the eye to the axilla, bounded below by a raised silver stripe which extends
from the axilla to a point mid-way between the nostrils and the snout, but often less boldly from
the anterior comer of the eye to the nostrils. Occasionally specimens may be encountered with all
light-coloured areas on the upper surfaces of the limbs and dorsum suffused irregularly with
blood-red pigment.

The belly is immaculate silvery gold. The throat is also immaculate, but in females there is a
deepening of the golden colour at the sides, while in males the gular area may be quite suffused
with light brown or black especially between the edge of the lower jaw and the longitudinal folds
at the sides of the vocal sac, and often between these folds. The latter are distinct and extend
from the pectoral region to the anterior edge of the lower jaw. The under surface of each limb is
dusky pink or brown. The upper surfaces of both femur and tibia are strongly banded transversely
with brown or dark brown, the femur with 4-6 bands, the tibia with 3-6 bands. The main tibial
and femoral bands match when the frog is at rest with folded limbs. The tarsus is also banded,
the banding extending the length of the fourth toe.

Rana montana is separable from R. fasciata on the leg banding alone, as pointed out by
FitzSimons (1946) and Poynton (1964). R. fasciata never possesses tibial cross-bands (Fig. 13).

Markings similar to cross-bands may occur in fasciata on the femur (which can, however,
possess no recognisable markings at all), but if present, they are attenuated and longitudinally
rather than laterally extended. Tibial markings never develop into transverse bands but are always
longitudinally oriented and sometimes extend in one stripe almost the whole length of the tibia;
such markings are usually weakly developed. The light coloured longitudinal vertebral stripe is
usually about half the width between the sacral projections. The throat of Rana fasciata is
immaculate and also suffused with gold. In males, however* the throat is a rich golden yellow.

Like many anurans, R. montana is capable of colour change in response to environmental
conditions (although Rose 1962 did not note “any appreciable power of colour change”). The
dark phase is in our experience the most frequently encountered, and here the entire dorsal
pattern is obscured except for the light vertebral and dorsolateral stripes and spots. Appropriate
heat and light treatment can induce change to the light phase within a half-hour.

The existence of colour pattern polymorphism in Rana montana is of considerable interest.
R. fasciata apparently does not vary significantly from the basic pattern as portrayed by Poynton

11

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 1, MAY 1979

Fig. 12. Male R. montana from Kleinmond (3419 AC Hermanus).

(1964), Wager (1965), Stewart (1967) and Fig. 13, but R. montana exhibits two basic pattern
morphs. The normal morph is that illustrated in Fig. 12, with a narrow light vertebral stripe, in
width approximately one third of the distance between the two projections of the sacral hump.
The other is the dorsoconcolor morph (Fig. 14a), so named by Lynch (1966) after a leptodactylid
frog from Central America; this pattern variant consists of a broad unicoloured dorsal area,
extending from the snout over the eyes, tapering to the tip of the urostyle.

Sixteen of our sample of 80 specimens (i.e. 20%) showed the dorsoconcolor pattern but this
figure is slightly biased. One collector of five specimens, three of which were dorsoconcolor
morphs, believes that he selected the more attractive frogs out of a total of twenty frogs handled,
these probably being dorsoconcolors (Herrington pers. comm.). If this sample is deleted from the
total, the relevant proportions are 13 out of 75 (17,3%), and if sight records are added the
percentage drops to 15%. Geographical distribution of the dorsoconcolor morph is not clear
because of inadequate sampling. However, two came from Betty’s Bay (3418 BD), two from
Kleinmond (3419 AC), eleven from the Muizenberg Mountains on the Cape Peninsula (3418
AB), and one from the Great Winterhoek Mountains (3319 AA). We have none as yet from the
eastern range of the species.

12

COMRIE GREIG, BOYCOTT & DE VILLIERS: ELEVATION OF RANA FASCIATA MONTANA

Fig. 13. Male R. fasciata from Theescombe, Port Elizabeth (3425 AB Uitenhage).

The same form of polymorphism occurs in Phrynobatrachus (Stewart 1974), and in Rana
grayii, which possesses “no stripe”, “thin stripe” and dorsoconcolor morphs. The dorsocon-
color patterns of R. montana and R. grayii are compared in Figs 14a and 14b. Polymorphic
pattern systems obviously play a role in protection from predators, especially in diurnal species;
there is some evidence that Rana montana is at least partially diurnal (see life history notes).

Size and sexual dimorphism

Poynton (1964) gives snout/ vent lengths (of the flattened animal) for Rana fasciata as 50
mm, and for R. montana as 47,5 mm. There is probably, however, no significant size difference
between the two species, as we have examined a large female R. montana of length 50 mm from
Swartboschkioof, Jonkershoek State Forest (coll. R. A. Haynes; in the Department of Forestry
collection).

Females of R. montana attain a larger size than males, averaging 40-50 mm and 31-40 mm
respectively. The males are recognisable when sexually mature by the presence of nuptial pads
on the thumbs and the longitudinal fold at each side of the vocal sac, as in Rana fasciata .
Morphological separation from R. fasciata and R. grayii

The three species, R. montana, R. grayii, and R. fasciata, resemble each other in general
conformation, are of approximately the same size, live in similar habitats, call in winter and are
long-toed jumping frogs with reduced webbing. Identification may therefore present problems,

13

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 1, MAY 1979

a. R. montana from Muizenberg Mtns. (3418 AB
S imonstad/S i monsto wn ) .

b. R. grayii from Muizenberg Mtns. (3418 AB
Simonstad/Simonstown).

Fig. 14. Dorsoconcolor pattern morphs

especially with dorsoconcolor variants of R. montana and R. grayii which are often sympatric.
Guidelines for separating them are provided below, but it should be noted that a hand-lens may
be necessary for some characters.

Rana montana

(1) possesses preorbital caruncle.

(2) webbing does not reach the basal subarticular tubercle on the fourth toe, and reaches
basal tubercle on the fifth toe.

(3) tip of first toe normally reaches to middle of basal subarticular tubercle of second toe or
beyond.

(4) femur and tibia distinctly cross-banded, the main bands corresponding when the femur
and tibia are adpressed. Even in the case of the dorsoconcolor morph, the cross-bands
are distinct and complete (see R. grayii).

(5) length of foot equal to distance from tip of urostyle to anterior margin of tympanum or
to the eye. (The length of the foot is taken to be the distance from the proximal end of
the metatarsal tubercle to the end of the fourth toe.)

Rana grayii

(1) possesses single preorbital groove.

(2) webbing does not extend beyond the basal subarticular tubercle on the fourth toe, but
reaches perceptibly beyond basal tubercle on the fifth toe.

(3) tip of first toe may fall short of basal subarticular tubercle of second toe or extends
beyond it. A variable character in this species.

(4) femur and tibia usually distinctly cross-banded, the bands corresponding when femur
and tibia are adpressed. Femur and tibia in certain individuals (especially in the
dorsoconcolor morph) may have cross-bands absent or incomplete, but never longitudi-
nally oriented.

(5) length of foot equal to distance from tip of urostyle to axilla (very rarely to posterior
margin of tympanum).

14

COMRIE GREIG, BOYCOTT & DE VILLIERS: ELEVATION OF RANA FASCIATA MONTANA

Rana fasciata

(1) possesses single preorbital groove.

(2) webbing (at least in E. and W. Cape specimens) reaches beyond basal subarticular
tubercle of fourth toe and fifth toe (but only about one quarter way along the phalanx).

(3) tip of first toe normally barely reaches the basal subarticular tubercle of the second toe,
and frequently falls short of it.

(4) femur and tibia not cross-banded. Femur may have attenuated patches of brown
corresponding to the cross-bands of the other two species; tibia has no cross-bands, but
may have longitudinally oriented streaks.

(5) length of foot equal to distance from tip of urostyle to anterior margin of tympanum or
to the eye.

DISTRIBUTION

Distribution localities given in this paper are cited according to the standard system for
biogeographical mapping proposed by Greig and Boshoff (in prep.). The grid blocks on the
accompanying maps are one quarter degree of latitude by one quarter degree of longitude
“square” and each is known as a “locus”.

Poynton (1964; Map 34) provides an accurate distribution map for Rana montana under the
name Rana fasciata montana, correctly attributing all western Cape “fasciata” material to this
form. It may be noted from his map that the “subspecies” montana and fasciata each overlap
the other’s range in the southern Cape.

This apparent overlap was one of the reasons why Inger (1959) suggested that Rana fasciata
montana was a dubious subspecies. In 1950/51, the Lund University Expedition collected four
specimens of Rana fasciata, from the Bloukrans River east of Plettenberg Bay (3323 DC
Nature's Valley), Grootkop, Knysna (3423 AA Knysna), 16 miles north of Matatiele (3028 BB
Lehlohonolo), and Barotta, 27 miles east of Louis Trichardt (2330 AB Levubu). In his
assessment of this material, none of which of course differed from typical R. fasciata, Inger
noted that FitzSimons’ localities for montana were interspersed with those he gave for fasciata
in the southern Cape Province, and that the supposed preference of montana for higher elevations
was shared by the Lund Expedition’s Matatiele fasciata specimen (given as 1905 m).

The localities in question given by FitzSimons were Jonkersberg and Grootvadersbos ( =
Grootvadersbosch) for montana , and Garcia and Still Bay for fasciata. If FitzSimons’ subspecies
had been valid, the existence of a population of fasciata at Garcia would be difficult to explain,
as this locality lies directly between Grootvadersbos and Jonkersberg, on the same mountain
range as Grootvadersbos.

The recognition of Rana montana as a species in its own right, however, resolves the
problems of altitudinal and geographic overlap. Indeed, the overlap is more complete than Inger
had believed; we re-visited each of FitzSimons’ southern Cape localities for Rana montana and
Rana fasciata, and while confirming the validity of his collections, we found that in fact
R. montana and R. fasciata exist more or less sympatrically at Grootvadersbos, Garcia and in the
Outeniqua Mountains, while clearly maintaining their specific identities. “True” sympatry, in
Mayr’s (1969) sense of the “existence of a population in breeding condition within the cruising
range of individuals of another population”, would be difficult to prove and may not exist;
nevertheless, at these three localities, calling males of both species could be heard simultaneously
by an observer although from different locations and differing habitats. This observation
constitutes the final proof, if such were required, of the specific distinction of the two forms.

It was chance which led FitzSimons to collect only R. montana at Grootvadersbos and
Jonkersberg, and only R. fasciata at Garcia; he could so easily have collected both species at
each collecting station.

15

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 1, MAY 1979

The question of altitude is not significant in itself. The determinant is habitat availability.
Further detailed study is of course required, but/?, montana distribution is apparently tied to the
“fynbos” vegetation type. It occurs from mountain top to the sea in the winter rainfall zone of
the western Cape, but in the southern Cape, where the climate is transitional to a summer rainfall
pattern, it is restricted to the mountains. In the mountains of the eastern Cape Province, and in
the Natal Drakensberg, R. fasciata finds suitably moist grassy habitat at high altitude, as the
Lund University Expedition found, and it extends down to sea-level all along the Natal and Cape
coasts as far west as Still Bay; it does not occur in the coastal fynbos of the winter rainfall region
of the S.W. Cape.

Figs 15 and 16 show the distribution ranges of Rana montana and Rana fasciata
respectively. They are based on Poynton’s map, with one or two corrections together with a
number of additional localities from a survey by the Cape Department of Nature and Environ-
mental Conservation from 1972 to 1978. The range of Rana fasciata is here shown to extend
nearly 110 km further to the south-west than previously recorded (Garcia; see FitzSimons 1946;

16

COMRIE GREIG, BOYCOTT & DE VILLIERS: ELEVATION OF RANA FASCIATA MONTANA

Fig. 16

Distribution of Rana fasciata Smith in southern Africa plotted on the latitude/longitude grid.

Poynton 1964), to Bonnievale (3320 CC Montagu). Rana montana is shown to have a wider
distribution range than previously noted; it occurs in the Cedarberg Mountains (3219 AC
Wuppertal) and eastwards to the Outeniqua Pass (3322 CD George).

Two nocturnal collecting trips have been made to Prince Alfred’ s Pass east of George (3323
CC Kruisvallei) where R. montana might also be expected to occur but it was not found,
perhaps, because of unsuitable weather conditions.

Rana montana is probably always found together with its relative R. grayii, but the latter is
a much more abundant and ubiquitous frog. We have never found Rana montana specimens easy
to acquire by night-collecting (see, however, life history notes), and it does not appear to form
numerically strong localised choruses like Rana fasciata. A table of spot localities for known
populations of R. montana is therefore provided below, which may prove useful to investigators
of this species’ life history (Table 1). The locus name and number correspond with the South
African 1:50 000 map series designation; these maps should be used for spot locality identifica-
tion.

17

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 1, MAY 1979

Fifty-four loci for Rana fasciata additional to those provided by Poynton {op. cit. ) are also
tabulated (Table 2). Of particular interest are those from the same localities as Rana montana, so
recorded in the column set aside for “notes”.

Table 1.

Spot localities for known Rana montana populations.

Locus

Locus Name

Spot Locality

Altitude

(m)

Notes

3219 AC

Wuppertal

32° 28'22"S; 19° 09'30"E

1 220

Coll. R. Haynes in Cedarberg State
Forest at Hoogvertoon. C.D.N.C. 4711

3319 AA

Groot-Winterhoek

33° 03'16"S; 19° 04'47"E
33° 04'28"S; 19° 06'02"E

640

680

Coll. R. Haynes at Driebosfontein.
C.D.N.C. 4706

Coll. D. Pepler at “De Tronk”, near
“Groot-Winterhoek”, on 12/xii/1976

3319 AC

Tulbagh

33° 24'45''S; 19° 07'19"E

450

Coll. R. Haynes at Kluitjieskraal

3319 AD

Ceres

33° 29' IT'S; 19° 19'13"E

1 670

Coll. C. Gow at U.C.T. Mountain Club
Hut, 1,5 km south of Waaihoek Peak,
25/iv/1965

3318 CD

Kaapstad/Cape Town

33° 58'22"S; 18° 23'58"E
33° 58'43"S; 18° 24'54"E

760

745

Poynton’ s record based on S.A.M.
specimen labelled “Table Mountain”.
Confirmed by specimen in S.A.M.
collected by C. Gow at the Mountain
Club Hut on Table Mountain on
2/i/ 1 965

Coll. E. van Jaarsveld. In S.A.M.
collection

3318 DC

Bell ville

Based on S.A.M. specimen labelled
“Cape Flats”. Doubtfully present in this
locus now, but specimen in any case
could have been collected in 3318 CD
Kaapstad/Cape Town

3318 DD

Stellenbosch

33° 59'36"S; 18° 57'1T'E

395

Coll. R. Haynes, Swartboschkloof,
Jonkershoek State Forest

3319 CC

Franschhoek

33° 52' 14"S; 19° 02'49"E
33° 48'18"S; 19° 03'26''E

33° 57'20"S; 19° 10'35"E

185

595

380

C.D.N.C. 4410. Wemmershoekvlei
Plateau to the north-west of
Wemmershoek Dam. R. A. Haynes
1 9/x/ 1 977

“Purgatory" outspan at east end of
Franschhoek Pass. C.D.N.C. 4719

3319 CD

Villiersdorp

33° 57'01"S; 19° 18' 13"E

550

In the Elandsrivier Valley (tributary
valley, the “Langkloof”) 4 km north of
Villiersdorp

3320 DD

Warmwaterberg

33° 59'29"S; 20° 51' 13"E

185

In a marsh at “Goede Hoop”,
Grootvadersbos. C.D.N.C. 4623

3321 CC

Muiskraal

33° 57'24"S; 21° 13'37"E
33° 57'49"S; 21° 13' 17"E

c. 455
c. 425

Near the Vetrivier, Garcia’s Pass
Near the Vetrivier, Garcia’s Pass

18

COMRIE GREIG, BOYCOTT & DE VILLIERS: ELEVATION OF RANA FASC1ATA MONTANA

Altitude

Locus

Locus Name

Spot Locality

(m)

Notes

3322 CC

Jonkersberg

33°

56'02"S; 22°

13'39"E

305

In the valley beside Jonkersberg Forest

Station

33°

55'43"S; 22°

13'34"E

410

C.D.N.C. 4612. Topotypical material

33°

55'44"S;

22°

1 3'41"E

410

C.D.N.C. 4613. Topotypical material

33°

51'44"S; 22°

02'39"E

610

North side of Robinson Pass, in the

Moerasrivier valley. Rana fasciata also
here

33°

51'52"S; 22°

01'49"E

745

North side of Robinson Pass

33°

55'09"S; 22°

01'27"E

455

South side of Robinson Pass

3322 CD

George

33°

54,37"S; 22°

24'39"E

505

In Outeniqua Pass north of George

33°

54'20"S;

22°

24'27"E

550

In Outeniqua Pass north of George

33°

53' 19"S; 22°

24'06"E

655

Just below the top of the Outeniqua Pass

on the south side

3418 AB

Simonstad/Simonstown

34°

05'29"S;

18°

25'20"E

290

Montane marshy “fynbos”. C.D.N.C.
4621; 4708. Steenberg Plateau

34°

13'26"S;

18°

24' 15"E

75

Coll. H. Langley at Modderdam, Cape
Point Nature Reserve. C.D.N.C. 4710

34°

1 4'46"S;

18°

23' 1 2"E

45

Coll. B. Rose and R. C. Boycott near
Olifantsbos, Cape Point Nature Reserve.
C.D.N.C. 4703

34°

06'26"S;

18°

26'56"E

410

Coll. D. Herrington at Nellie’s Pool on
Muizenberg Mtns. C.D.N.C. 4709

3418 BB

Somerset West

34°

09'06"S;

18°

56'23"E

c. 395

To the east of Sir Lowry’s Pass,
Hottentots Holland Mountains

3419 AA

Grabouw

34°

09'19"S;

19°

08'28"E

670

Lebanon State Forest. C.D.N.C. 4702

34°

09'22"S;

19°

08'31"E

670

Lebanon State Forest. C.D.N.C. 4704

34°

10'09"S;

19°

09'51"E

655

Lebanon State Forest. C.D.N.C. 4712

34°

09'28"S;

19°

09'27"E

825

Lebanon State Forest. C.D.N.C. 4707

34°

08'50"S;

19°

08'23"E

885

Coll. J. W. Esterhuysen, Lebanon State
Forest. 6/ii/ 1 974

34°

00'26"S;

19°

00'35"E

1 220

Coll. R. Haynes at Dwarsberg in
Jonkershoek State Forest. C.D.N.C.
4705

34°

05'26"S;

19°

03'09"E

490

2V4 km due south of Nuweberg Forest
Station in Viljoen’s Pass

34°

04' 19"S;

19°

03'59"E

490

0,5 km on north side of Viljoen’s Pass
10 km north of Grabouw

3419 AB

Caledon

34°

1 3' 1 9"S;

19°

28'59"E

335

Above the farm “Voorhoede” east of
Caledon, in the hills

3419 BA

Greyton

34°

13'24"S;

19°

3 1 ' 1 6"E

275

In the hills near the national road 8V2
km east of Caledon

3420 BB

Heidelberg (Kaap)

34°

00'06''S; 20°

51 '23"E

170

In a marsh on the farm ‘‘Goede Hoop”

3421 AA

Groot-Kragga

34°

01'27"S; 21°

13'50"E

150

Vetrivier valley below Garcia Forest

Station. Rana fasciata also here

34°

01 ' 1 5"S;

21°

13'54"E

150

In a marsh immediately to the west of
Garcia Forest Station

3418 AD

Simonstad/Simonstown

34°

18' 13"S;

18°

25' 19"E

80

Cape Point Nature Reserve, Klein
Blouberg

19

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 1, MAY 1979

Altitude

Locus

Locus Name

Spot Locality

(m)

Notes

3418 BD

Hangklip

34° 20'40"S; 18° 50'50"E

25

C.D.N.C. 4699; C.D.N.C. 4701

34° 16'38"S; 18° 59'52"E

110

Coll. B. Durand, Kogelberg State
Forest. 26/viii/ 1975

3419 AC

Hermanus

34° 20T1"S; 19° 03'30"E

15

Coll. C. Burgers. C.D.N.C. 4016

34° 19'18"S; 19° 09'38"E

40

Hills on east side of Botrivier

34° 20'00''S; 19° 03'10"E

15

C.D.N.C. 4594. Sonogram in this paper
is from this specimen

34° 19'42"S; 19° 05'26''E

30

East of Lamloch at Kleinmond

34° 19'03"S; 19° 07'15"E

45

In hills to the west of
Botrivier/Kleinmond road. Several
localities along this road

3419 AD

Stanford

34° 24'24"S; 19° 19' 13"E

10

Marsh next to sea at mouth of
Voelgatrivier, 2 km east of Hermanus.
C.D.N.C. 4700

34° 23'40"S; 19° 16'00"E

90

In Femkloof Nature Reserve, Hermanus

34° 19'00"S; 19° 24'37"E

245

South side of Shaw’s Mountain Pass, on
the farm “Hartebeesrivier”

34° 1 9'24"S; 19° 22'05''E

200

Below Shaw’s Mountain

Table 2.

Loci for Rana fasciata populations not recorded by Poynton (1964) with altitudes of sampled

populations

Locus

Locus Name

Altitude (m)

Notes

2530 BB

Sabie

1 190

2929 CC

Bushman’s Nek

3128 DC

Elliotdale

670

3128 DD

Mqanduli

670

3129 CC

Coffee Bay

90

3226 AD

Spring Valley

1 615

3226 BC

Hackney

1 660

3226 BD

Fairford

1 400

3226 CA

Bedford

790

3226 DB

Seymour

1 280

See text

3227 AB

St. Marks

1 065

3227 AC

Cathcart

1 415

3227 CA

Keiskammahoek

1 435

See text

3227 CB

Stutterheim

520

3227 CD

King William’s Town

560

20

COMR1E GREIG, BOYCOTT & DE VILLIERS: ELEVATION OF RAN A FASCIATA MONTANA

Locus

Locus Name

Altitude (m)

Notes

3227 DD

Cambridge

120

3228 BB

The Haven

455

3228 BD

Dwesa

15

3320 CC

Montagu

215

3320 DD

Warmwaterberg

200

Shared locality with R. montana

3322 CC

Jonkersberg

355, 610

Shared locality with R. montana

3322 DC

Wilderness

10

3322 DD

Karatara

215

3323 CD

The Crags

15, 230

3323 DC

Nature’s Valley

15

3324 CB

Akkerdal

305, 670

3325 BC

Coemey

105

3325 CB

Uitenhage (Noord)

120

3325 CD

Uitenhage

180

3325 DC

Port Elizabeth

90

3326 AA

Riebeek-Oos

655

3326 BA

Fort Brown

400

3326 BB

Breakfast Vlei

170

3326 AD

Salem

370

3326 BD

Trappe’s Valley

400

3326 DA

Boesmansriviermond

50, 270

3327 AC

Prudhoe

185

3327 CA

Great Fish Point

15

3420 AA

Stormsvlei

105

3420 AB

Swellendam

105, 120

3420 BA

Suurbraak

75, 105

3420 BB

Heidelberg (Kaap)

170, 365

3421 A A

Groot-Kragga

135, 170

Shared locality with R. montana

3421 AB

Riversdale

105, 425

Probable source of FitzSimons’ Garcia
material

3421 BA

Albertinia

200

3421 BB

Herbertsdale

170

3422 AA

Mosselbaai

15, 160

3422 AB

Pacaltsdorp

170

3422 BA

Wilderness

90

3422 BB

Sedgefield

10

3423 AB

Plettenbergbaai

215

3424 AB

Clarkson

215

3424 BA

Kruisfontein

170

3425 AB

Uitenhage

135

21

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 1, MAY 1979

It is probable that the locus given by Poynton (1964) for FitzSimons’ R. montana record
from Grootvadersbos, 3320 CD Scheepersrus, is incorrectly cited (Poynton, pers. comm.).
FitzSimons’ locality was “Langebergen, above Grootvadersbosch”, but Grootvadersbos is not
indicated on the 1:500 000 map of the area, and the word “Langeberge” is printed on the locus
3320 CD. However, both the farm “Grootvadersbos” and the forest of the same name actually
lie mainly in 3320 DD Warmwaterberg and 3420 BB Heidelberg (Kaap), and marginally in 3320
DC Barrydale and 3420 BA Suurbraak; no part of the estate lies in 3320 CD Scheepersrus. As
stated below in the section dealing with the type-localities, we suggest that this record was
obtained in 3320 DD Warmwaterberg.

Although the locality “Cape Flats” is cited for R. montana by Poynton (from British
Museum and South African Museum material), we have been unable to confirm its occurrence in
the locus cited, viz. 3318 DC Bellville. This locus has therefore been deleted from figure 6 until
such time as the species’ presence can be confirmed there. Old South African Museum
specimens of R. montana are catalogued from “Rondebosch”, “Table Mountain” and “Cape
Flats”. The latter specimen was collected in 1875 and accessioned only in 1896, which might
suggest that the selection of a locality could have been somewhat arbitrary. The Cape Town
suburb of Rondebosch lies partly on the Flats and partly on the lower slopes of Table Mountain;
it lies in the locus 3318 CD Kaapstad/Cape Town. The Cape Flats proper, which lie largely in the
loci 3318 DC Bellville and 3418 BA Strandfontein, have now been so altered by urbanisation,
agriculture, and the uncontrolled spread of exotic vegetation, that any formerly suitable habitat
for Rana montana has now all but disappeared.

All other Rana montana loci cited on Poynton’ s map have been confirmed by recent
collections.

It will be noted from Fig. 16 that the locality given by Poynton {op. cit .) for the FitzSimons
R. fasciata specimen from “Garcia” has been changed from 3321 CC Muiskraal to 3421 AB
Riversdale. FitzSimons (1946) acknowledged the assistance of Mr H. J. Vockins who was the
forester at Garcia in November 1940, and who now lives in retirement at George. We approached
the latter for information as to FitzSimons’ collecting sites, and in his recollection, the site where
the R. fasciata was collected was at the “Perdedam” in the Garcia Plantation. This locality was
accordingly visited, and despite an initial impression of unsuitability of habitat, we found Rana
fasciata calling in indigenous vegetation along a stream in a pine plantation at 10.30 a.m. at an
altitude of 425 m. It is suggested that this locality be regarded as the source of Rana fasciata
T.M. 20099. As a matter of interest, R. montana was heard calling in a marsh just below the
Garcia Forest Office that same morning (3 August 1978) at an altitude of only 150 m, and only
300 m distant from another Rana fasciata population calling at an artifical pond.

The locality given by Poynton for an Albany Museum R. fasciata specimen from
“Hogsback”, 3227 CA Keiskammahoek, should perhaps have been cited as 3226 DB Seymour,
as we interpret the label to mean the Hogsback holiday resort and township and not the mountain
of the same name. R. fasciata is abundant in the forest glades nearby. We have, however,
confirmed the species from both Hogsback township (3226 DB) and from below the Hogsback
Mountain (3227 CA).

THE TYPE-LOCALITIES

The type-locality of Rana montana is “the upper slopes of the Jonkersberg, Attaqua
Mountains” (FitzSimons 1946). FitzSimons used the Jonkersberg Forest Station as his base in
November 1940 and it is reasonable to suppose that the type specimen was collected within the
forest limits, above the forest offices. We have heard the species calling within 300 metres of the
office complex, and caught two calling males in natural vegetation within the pine plantations 75
m in altitude above the station and 650 m to the north-east. FitzSimons claimed that all his

22

COMRIE GREIG, BOYCOTT & DE V1LLIERS: ELEVATION OF RAN A FASCIATA MONTANA

specimens (two from Jonkersberg and four from Grootvaderbos) were “taken in grass-covered
marshy hollows on the higher mountain slopes, at an altitude of 3-4 000 ft.” (900-1 200 m). In
both places he could have collected them within 500 m of his camp-sites on the lower slopes, but
this would not conform to the stated altitudes. The type-locality may with confidence be
restricted to the mountain slopes in the Jonkersberg State Forest (holotype T.M. 20223; 3322 CC
Jonkersberg).

The paratypes (T.M. 19966-9 and T.M. 20222) were collected respectively at Groot-
vadersbos and Jonkersberg. As stated earlier, the Grootvadersbos estate lies in four loci, but
marginally in two of these. The main section lies in the locus 3320 DD Warmwaterberg including
the large section of indigenous forest and the State Forest plantations. A timber extraction track
which existed in 1940 runs up the mountain from the forest station to the natural forest

Boesmansbos, and passes through “grass-covered marshy hollows . . . at an altitude of 3-4 000

ft.”. We have in fact a record of/?. montana from Helderfontein above Boesmansbos at 3 900 ft
(1 190 m). It is almost certain that these paratypes were collected in 3320 DD Warmwaterberg.

Rana fasciata is interpreted by reference to the lectotype B.M. 58.11.25.127 with the
locality “Africa”, collected by Sir Andrew Smith. The type-locality may never be ascertained
with accuracy.

UFE HISTORY NOTES

Very little is known of the life history of Rana montana, although much of the detail

provided by Rose (1962, and in the 1950 edition of the same book) under the name of R.

fasciata, in fact refers to R. montana. We present the following incomplete notes with some
diffidence.

Both species are confined to areas of high rainfall (over 500 mm per annum), and are rarely
found far from permanent water. R. fasciata, however, appears to prefer long grass and reeds
adjacent to streams, ditches and dams with relatively deep water; relatively deep that is, when
compared with R. montana which is associated with marshy areas and seepage zones with
shallow trickles of water.

A normal R. fasciata chorus consists of many males closely spaced, often only half a metre
apart, while calling montana males are usually at least two or three metres apart. This is of
course attributable to the difference in breeding site; the “coastline” of a marsh is much longer
than that of a pond, and the montana males are not forced to jostle for a water’s edge territory.
Both species appear to be winter breeders.

It is our impression that Rana fasciata benefits by man’s agricultural activities. It is
commonly found around well-vegetated farm dams, in damp areas near irrigation canals, and in
town gardens; it is often found in exotic vegetation. Rana montana, on the other hand, is
probably intolerant of disturbance; we have never found it in anything other than natural veld,
and it does not seem to occur for example in Hermanus or Kleinmond, although these towns have
been built on what is presumed to have been formerly R montana habitat. This may be because
garden simulation of a fasciata breeding site is a simpler matter than that of a montana breeding
site.

Frog collections are often made at night, when calling males can be tracked down by ear and
captured. This is especially easy if the frogs concentrate around a circumscribed breeding site as
in the case of R. fasciata, but difficult in the case of R. montana whose calling males are
dispersed over a wide area. Consequently, specimens of R. montana are not numerous in
museum collections.

Specimens collected for our survey by colleagues and Forestry Department personnel were
largely obtained during the day, suggesting that/?, montana is at least partially diurnal. One of us
(A. L. de V.) collected eighteen specimens in just over one hour in an 80 m X 40 m area of damp
ericoid-restioid fynbos on the Steenberg Plateau (10/ix/1978), and twenty in 45 minutes in a 70 m

23

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 1, MAY 1979

x 35 m area near Nellie’s Pool a week later (both 3418 AB Simonstad/Simonstown), in the late
afternoon. A similar number was collected by D. Herrington under similar conditions in July
1976. The frogs were found jumping through short (V2 metre) vegetation on disturbance, in damp
seepage zones.

The positions adopted by calling males of R. fasciata and R. montana are similar,
depending upon the cover. If tall reeds or rushes are present, fasciata males will clamber up to
two metres above water level and call from a spread-eagled position clasping two or even three
separate stalks; if the water’s edge is surrounded by tussocks of grass, the males will call at water
level, crouching on the ground below grass cover, or in the open.

R. montana males have been found calling from the crouching position at water level below
an overhanging tussock of a short restioid sedge 8 cm high in a marsh at Grootvadersbos. In the
Jonkersberg State Forest they called from within a dense mat of interlaced ferns about IV2 metres
deep, overhanging a runnel of water below a road embankment. At Kleinmond they called from a
roadside ditch 10-20 cm above ground level spread-eagled between tall grass stems.

The eggs of Rana montana are probably laid, like those of grayii, not far from the water’s
edge, but not in water. At the Jonkersberg State Forest we found a batch of 39 eggs on a
waterlogged moss/earth substrate about 6 cm from a shallow runnel of water at the edge of a
level road below a bank overhung with ferns. These eggs were lying singly, or in groups of up to
six or seven attached in rows. The only frog found here was R. montana and the assumption is
that these were montana eggs. Rose (1962) also states for fasciata ( = montana ) that he found
some “capsule” eggs by the side of an upland trickle, which could never have reached to them;
from these eggs he reared fasciata ( = montana) froglets. The Jonkersberg egg capsules
averaged 7 mm in diameter (6, 0-7, 6 mm).

It is interesting to note that although Rana fasciata is generally supposed to lay its eggs in
water (Wager 1965), Stewart (1967) records that the subspecies fuelleborni lays its eggs out of
water on sedge blades above the water, or on moist earth an inch above water level. Stewart
erroneously attributes Rose’s remarks on ‘ Rana fasciata montana ” to Rana fasciata fasciata.

A single trombiculid mite was found on an adult male R. montana from Jonkersberg State
Forest (specimen C.D.N.C. 4613, see table 1). It was visible as a minute reddish speck on the
sole of the left foot in soft tissue immediately below the webbing between the second and third
toes. The frog was placed in formalin before the mite was noticed, rendering the latter unsuitable
for identification. Such mites should be preserved directly in 70% ethanol taking care not to
damage the mouthparts when removing them from the host.

THE NAME RANA MONTANA

The elevation of FitzSimons’ subspecies to species level is straightforward, "montana” is a
species-group name and was applied by FitzSimons with due regard to the dangers of synonymy.
In other words, there is no evidence of previous usage of the name montana in the genus Rana ,
and there is thus no impediment in the way of its employment despite the fact that the species is
not necessarily montane in its habitat preference. The full citation, if desired, should read Rana
montana FitzSimons, 1946.

Tschudi in 1838 proposed the genus Strongylopus for his concept of Boie’s Rana fasciata ,
as Strongylopus fasciatus . Steindachner (1867) employed the same genus for Rana grayii. Parker
and Ride (1962) of course showed that Boie’s "Rana fasciata” was probably what we now
know as R. grayii (see section dealing with the taxonomic history of the name Rana fasciata).

Although Poynton (1964) and the majority of his predecessors placed the species fasciata
and grayii in the genus Rana, Van Dijk (1966, 1971) revived Strongylopus as a genus and used
it for R. fasciata , R. grayii, R. wageri and R. hymenopus, without however, publishing the
diagnostic characters justifying the separation of the two genera. Parker and Ride (1962) in fact

24

COMRIE GREIG, BOYCOTT & DE VILLIERS: ELEVATION OF RANA FASCIATA MONTANA

suggested that Strongylopus might be found useful as a subgenus of Rana ; this suggestion
appears to us to have some merit, at least in respect of R. fasciata, R. grayii and R. montana.

Strongylopus as used by Tschudi is masculine in gender. If used as a generic name for the
species fasciata, montana and grayii, therefore, the former two names would become fasciatus
and montanus. If Strongylopus is used as a subgenus, however, no alteration in the specific name
is required, as the specific name agrees in gender with the generic name, e.g. Rana ( Stron-
gylopus) montana. The citation “Strongylopus (Rana)’’ as used by Van Dijk (1971; 1977) to
indicate equivalence of the two genera is incorrect. By convention, subgenera are cited in
parenthesis between the generic and specific names. It is best to indicate an alternative generic
name by the use of an “equals” sign, e.g. Strongylopus ( = Rana) grayii.

The choice of a vernacular name for Rana montana is not straightforward. If the specific
epithet is anglicised, the resulting name would be the “montane frog” or “mountain frog”
which would be misleading in view of its occurrence in coastal marshes. R. fasciata has been
anglicised in this way and has been called the “striped rana”, “striped frog”, “striped
grass-frog” and “long- toed grass frog”. Despite the fact that sedges of the family Restionaceae
rather than true grasses dominate the habitat of R. montana, we feel that “grass frog” fulfils a
useful role as a vernacular “generic” name for fasciata and montana, and we therefore
tentatively suggest that they could be called respectively the “striped grass frog” and the “Cape
grass frog”.

TAXONOMIC HISTORY OF THE NAME RANA FASCIATA

The history of the name Rana fasciata is marked by confusion over authorship and mistaken
identity. Authorship has been variously ascribed to Tschudi (FitzSimons 1946), Boie (Hewitt
1911), and Smith (Poynton 1964). Parker and Ride (1962), however, pointed out that the name
was first proposed by Burchell (1824). The background to this confused state of affairs was
succinctly presented by Parker and Ride (op. cit.) in an application to the International
Commission on Zoological Nomenclature. They claimed, apparently correctly, that the names
Rana fasciata and Rana grayii* were popularly assigned to the wrong frogs, and that if the law
of priority were to be strictly enforced, the species now known as R. grayii would have to
become R. fasciata, while the species now referred to as R. fasciata would require a new name
altogether. In the interests of stability in nomenclature, therefore, they proposed that the currently
used names should be retained by the exercise of the Commission’s plenary powers, and they
further requested the Commission’s blessing for a proposed neotype for R. fasciata and a
lectotype for R. grayii. This application was supported by Poynton (1963). The matter was
settled finally by Opinion 713 of the Commission (China 1964).

As Parker and Ride were of course ignorant of the existence of Rana montana , which bears
a superficial resemblance to R. fasciata sensu Smith, and as they based part of their submission
regarding the true identity of Burchell’s Rana fasciata on what appears to be faulty reasoning, a
brief summary of their 1962 application and the Commission’s subsequent findings is provided
here.

(1) The case concerned Rana fasciata Burchell, 1824, and Rana grayii Smith, 1849.

(2) Burchell (1824) described a new species of frog, Rana fasciata, from 30°Q5'S;

23°28'E. It was “a very pretty and new species of frog of a green colour, and marked

*Note: This species was described by Smith (1849), who used the spelling “grayii” . All subsequent authors except
FitzSimons have “corrected” this to “ grayi ”. The “correct original spelling”, however, is “ Rana grayii”
(Article 32, International Code of Zoological Nomenclature, 1964), which spelling should henceforth be used. It
should be noted too that the same stricture applies to four other southern African frogs. These are Tomopterna
( = Pyxicephalus) delalandii, Hyperolius horstockii, Kassina wealii and Leptopelis bocagii , where the specific
epithet variants delalandei, delalandi, horstocki, wealei and bocagei are no longer acceptable.

25

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 1, MAY 1979

by a longitudinal yellow stripe on its back, and by transverse stripes of brown on its
hind legs”. He also stated that it was a silent frog which croaked very seldom. It was
found on the 3rd March 1812. No type-specimen has been traced and no sketch or
figure exists (see below).

(3) Boie (1832) stated that he collected Rana fasciata Burchell near Muizenberg on the
Cape Peninsula.

(4) Tschudi (1838), unaware of Burchell’ s work, proposed a new genus Strongylopus , with
Rana fasciata as type-species, citing Boie incorrectly as author.

(5) Dumeril and Bibron (1841) described four varieties of Tschudi’s Strongylopus fas-
ciatus ( = Boie’ s Rana fasciata).

(6) Andrew Smith (1849) realised that Dumeril and Bibron had more than one species in
their four varieties, and named one of them Rana grayii. He also described and figured
“Rana fasciata Boie” corresponding to variety D of Dumeril and Bibron. Since 1849
the name Rana fasciata has been consistently applied to the from illustrated under that
name by Smith.

(7) Parker and Ride examined the specimens used by Dumeril and Bibron and confirmed
that Smith was correct in dividing them into two species. They also confirmed that
“variety D” corresponded with Smith’s figure of R. fasciata. Dumeril and Bibron
believed this variety to include the Leiden Museum specimens labelled R. fasciata by
Boie, and Boie of course believed these to belong to the species described by Burchell.
Parker and Ride therefore deemed it correct to use the name “ Rana fasciata Burchell”
for the species known popularly until 1962 as “ Rana fasciata Boie” or “ Rana
fasciata (Tschudi)”.

(8) Unfortunately Boie’s two surviving specimens at Leiden are both referable to Smith’s
R. grayii.

(9) Therefore the current concept of Rana fasciata is based on Smith’s acceptance of
variety D of Dumeril and Bibron as Rana fasciata; these authors in turn believed their
variety D corresponded to Boie’s Leiden specimens labelled R. fasciata, which Boie of
course believed to correspond to Burchell’ s Rana fasciata, but which correspond in
fact to Smith’s Rana grayii .

(10) This meant that if Boie correctly applied Burchell’s name, the species now known as
R. grayii would have to be known as R. fasciata, and the species now known as
R. fasciata would require a new name.

(11) Parker and Ride then selected a neotype for Rana fasciata Burchell ( sensu Smith) from
Smith's material in the British Museum (locality “Africa”), and a lectotype for Rana
grayii Smith also from Smith’s collection, and requested that the Commission preserve
stability of nomenclature by using its plenary powers to direct that Rana fasciata
Burchell, 1824 be interpreted by reference to that neotype, and that Rana grayii Smith,
1849 be interpreted by reference to that lectotype.

Poynton (1963) supported the application, pointing out that Burchell’ s description was
inadequate and that it could not be demonstrated unequivocally that R. fasciata Burchell is R.
grayii of later authors.

Professor Hobart Smith agreed with the need to preserve the name Rana fasciata in its
accustomed sense (i.e. sensu A. Smith), but proposed an alternative means of achieving this, viz.
by suppressing the name Rana fasciata Burchell, 1824, together with all other uses of the same
name prior to that by Andrew Smith in 1849, and that the proposed neotype for Burchell’s
species selected by Parker and Ride (one of Smith’s specimens) should be designated as a
lectotype of Rana fasciata A. Smith, 1849.

The Commissioners voted to suppress the name Rana fasciata Burchell, 1824, and all other

26

COMRIE GREIG, BOYCOTT & DE VILLIERS: ELEVATION OF RAN A FASCIATA MONTANA

uses of the specific name fasciata in the combination Rana fasciata prior to that by Smith, 1849.
They also voted to accept both the Rana grayii lectotype designated by Parker and Ride, and the
specimen designated by them as neotype of Rana fasciata as the lectotype of R. fasciata Smith,
1849 (China 1964).

As the lectotype of Rana fasciata Smith, 1849 in the British Museum bears no more precise
locality on its label than “Africa”, two pertinent questions arise from the melange of misconcep-
tions described above.

(a) does the recognition of Rana montana as a species in its own right in any way affect
the selection of the Rana fasciata lectotype by Parker and Ride?

(b) what was the species which Burchell named Rana fasciata ?

Firstly, let us take the case of the Rana fasciata lectotype whose type-locality, “Africa”,
could hardly be more vague. Fortunately, Smith published an illustration of his concept of Rana
fasciata, and Parker and Ride (1962) and Poynton (1963) believed that the selected lectotype
(B.M. 58.11.25.127) may well have been the one used in the preparation of this illustration
(Smith, 1849, pi. 78, fig. 1). This plate is reproduced here (Fig. 17). As has been pointed out
earlier (in section dealing with colour markings), Rana montana possesses well-marked trans-
verse bands on the tibia while R. fasciata has longitudinal tibia! markings; Smith’s plate clearly
shows a frog with the tibial markings of Rana fasciata . Thus the lectotype of Rana fasciata in
the British Museum was correctly selected, and the separation of Rana montana as a distinct
species causes no upset in existing taxonomic arrangements.

The second problem is that of the identity of Burchell’s Rana fasciata. This was collected at
30° 05' S; 23° 28 E, i.e. in the locus 3023 AB Good Hope, in the heart of the Karoo. Reference

Fig. 17. Smith's (1849) illustration of Rana fasciata on which the selection of the lectotype was based.

27

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 1, MAY 1979

to Poynton’s (1964) distribution maps, and to figures 6 and 7 in this paper, shows that this frog
could scarcely have been Rana fasciata Smith, 1849, which in the Cape is confined to the wetter
coastal regions. It is also very unlikely to have been Rana grayii, although Poynton did examine
a specimen of this species apparently collected at Vosburg (3022 DB Vosburg) approximately 80
km to the south-west in the collection of the Department of Zoology, University of the Orange
Free State. Van Dijk (1971) draws attention to the fact that Rana grayii has adapted well to
living in poplar plantations and points out that Karoo farms often possess such poplar groves
maintained by borehole water. In Burchell’s day this was of course not the case. Parker and Ride,
who were not familiar with South African conditions, enlisted the aid of the South African
herpetologist, the late Walter Rose of Cape Town, to find a solution. It may be presumed that
they gave Rose the alternatives as R. grayii and R. fasciata , together with Burchell’s descriptive
notes (i.e. a green frog, with a single longitudinal yellow dorsal stripe, transverse brown stripes
on its hind legs, a short small body and of a taciturn nature). Rose correctly replied that R.
fasciata Smith, 1849 is never green “though this is the prevalent colour of R. grayi” , that it has
the more powerful call of the two species, and a markedly long and slender body. Quite apart
from the fact that the Cape Town-based Rose was almost certainly describing R. montana ( vide
Rose 1962) rather than R. fasciata Smith, it is arguable that Rana grayii is not a noticeably
green frog, but is perhaps better described as pale olive as Wager (1965) states; green and yellow
are certainly not the first chromatic adjectives which spring to mind in regard to Rana grayii. It is
difficult to see why the loudness of the call should be regarded as significant by Parker and Ride,
as Burchell said only that “it was further distinguished by its silence, or at least by croaking very
seldom”. In any case a subjective decision on the relative power of the calls of R. montana, R.
grayii and R. fasciata is hardly relevant.

On the other hand it could be postulated that Rana fasciata Burchell was in fact Rana
fuscigula Dumeril and Bibron, 1841. This frog is often bright green, often bears a single yellow
vertebral stripe, has a short body, is not a particularly vocal species and might easily have been
silent on 2nd March 1812 at the end of a hot dry Karoo summer. The locus 3023 AB Good Hope
is clearly within the range of Rana fuscigula which has in fact been collected in the neighbouring
locus 3023 AA Sodium. If Burchell’s Rana fasciata was indeed Rana fuscigula of Dumeril and
Bibron, then Rana fuscigula would have become a junior synonym of Rana fasciata , and Rana
fasciata Smith would have required a new name. However, the action of the International
Commission in suppressing Rana fasciata Burchell, 1824 has fortunately obviated this situation.

Many of Burchell's sketches still survive. They were held in the library of the University of
the Witwatersrand, Johannesburg, but later transferred to the Africana Museum, Johannesburg
Public Library. The Johannesburg Africana Museum Catalogue of Pictures, vol. 6, contains
reproductions of these sketches but the only amphibian illustrated by Burchell in 1812 was
clearly a toad (B1941, number 17, sketch-book 3 of 1812). The compiler of the book implies in a
descriptive note to B1941 that the toad is Burchell’s Rana fasciata, but the date — 4 September
1812 — is too late for the collection date of R. fasciata , and the sequence of dates indicates that a
frog collected in March would have been in sketch-book 1 or 2 of 1812.

(Note: The abbreviations used are as follows: T.M. (Transvaal Museum); S.A.M. (South
African Museum); C.D.N.C. (Cape Department of Nature and Environmental Conservation.) The
majority of the Rana montana specimens on which this paper is based are to be lodged in the
collection of the Albany Museum, Grahamstown, South Africa. Others collected by the staff of
the Department of Forestry, are lodged in their collection at Jonkershoek Forest Station,
Stellenbosch, South Africa.)

28

COMRIE GRE1G, BOYCOTT & DE VILLIERS: ELEVATION OF RAN A FASCIATA MONTANA

ACKNOWLEDGEMENTS

We would like to acknowledge the assistance of the Department of Forestry and in particular
Mr P. G. Reyneke (D.F.O., Langeberge), Mr L. M. Francis (Regional Director, Southern Cape),
R. A. Flaynes and B. J. Durand of Jonkershoek, J. G. M. Rossouw of Garcia, H. N. van Rhyn of
Jonkersberg, and J. W. Esterhuysen of Lebanon. H. J. Vockins, formerly of Garcia Forest, was
most helpful, as also D. Pepler of the Faculty of Forestry, University of Stellenbosch. Mr and
Mrs J. P. Wilson of “Snelsetter”, Grootvadersbos, and Mr and Mrs A. L. Searle of Great Brak
River are thanked for their hospitality. Mr C. Jacot-Guillarmod, former Director of the Albany
Museum, provided advice on nomenclatorial matters. Dr A. Semmelink of the Central Acoustics
Laboratory, University of Cape Town, produced the oscillograms for us; we thank him for giving
us so much of his time. Dr N. I. Passmore of the Department of Zoology, University of the
Witwatersrand processed our sonograms and was also generous with advice. Dr J. C. Poynton of
the Department of Biological Sciences, University of Natal, Durban, is also thanked for critically
reading the manuscript, as are Dr C. J. Loedolff, Dr I. G. Gaigher, C. J. Burgers, P. M. Norton
and C. W. Heyl of the Cape Department of Nature and Environmental Conservation. Mr C. T.
Stuart also of the same department is especially thanked for his assistance. Lastly we wish to
thank the Director of Nature and Environmental Conservation of the Cape Province and the
Director of the Port Elizabeth Museum for permission to undertake this work.

REFERENCES

Blair, W. F. 1958. Mating call and speciation of anuran amphibians. Am. Nat. 92: 7-51.

Boie, H. 1832. Briefe von Heinrich Boie geschrieben aus Ostindien und auf der Reise dahin. Neues Staatsbiirgerliches
Magazin 1: 126-218.

Burchell, W. J. 1824. Travels in the interior of southern A frica . Vo) 2. London: Longman, Hurst, Rees, Orme and
Brown.

China, W. E. 1964. Opinion 713. Rana fasciata Smith, 1849 (Amphibia): added to the Official List with suppression of
Rana fasciata Burchell, 1824, under the plenary powers. Bull. zool. Nomencl. 21(5): 352-354.

Dumeril, A. M. C. and Bibron, G. 1841. Erpetologie generate ou histoire naturelle complete des reptiles. 8. Pans.

FitzSimons, V. F. M. 1946 An account of the reptiles and amphibians collected on an expedition to the Cape Province,
October to December, 1940. Ann. Tvl. Mus. 20(4): 351-377.

Greig, J. C. and Boshoff, A. F. Biological distribution mapping in southern Africa: the methodology of locality
designation using the latitude/longitude system, (in prep.).

Hewitt, J. 1911. A key to the species of the South African Batrachia together with some notes on the specific characters
and a synopsis of the known facts of their distribution. Rec. Albany Mus. 2: 189-228.

Inger, R. F. 1959. Amphibia. In: Hanstrom et at. eds, South African animal life, Vol. 6. Stockholm: Almqvist and
Wiksell, pp. 510-553.

Loveridge, A. 1953. Zoological results of a fifth expedition to East Africa. IV. Amphibians from Nyasaland and Tete.
Bull. Mus. comp. Zool. Harv. 1 10(4): 325-406.

Lynch, J. D. 1966. Multiple morphotypy and parallel polymorphism in some neotropical frogs. Syst. Zool. 15(1): 18-23.

Mayr, E. 1969. Principles of systematic zoology . New York: McGraw-Hill.

Parker, H. W. and Ride, W. D. L. 1962. Rana fasciata Burchell, 1824 (Amphibia); proposed designation of a neotype
under the plenary powers. Z.N.(S.) 1253. Bull. zool. Nomencl. 19(5): 290-292.

Passmore, N. I. 1977. Mating calls and other vocalisations of five species of Ptvchadena (Anura: Ramdae). S. Afr. J.
Sci. 73: 212-214.

Passmore, N. I. and Carruthers, V. C. 1975. A new species of Tomopterna (Anura: Ranidae) from the Kruger
National Park, with notes on related species. Koedoe 18: 31-50.

Pengilley, R. K. 1971. Calling and associated behaviour of some species of Pseudophryne (Anura: Leptodactylidae). J.
Zool , Lond. 163: 73-92.

Poynton, J. C. 1963. Letter to the editor. Bull. zool. Nomencl. 20: 255.

Poynton, J. C. 1964. The Amphibia of southern Africa: a faunal study. Ann. Natal Mus. 17: 1-334.

Rose, W. 1962. The reptiles and amphibians of southern Africa . Cape Town: Maskew Miller.

Smith, A. 1849. Illustrations of the zoology of South Africa; Reptilia. London: Smith, Elder & Co.

29

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 1, MAY 1979

Steindachner, F. 1867. Reise der Osterreichischen Fregatte Novara um die Erde in den Jahren 1857, 1858, 1859.

Zoologischer Theil. 1. Amphibien. Vienna: Staatsdruckerei.

Stewart, M. M. 1967. Amphibians of Malawi. New York: State University of New York Press.

Stewart, M. M. 1974. Parallel pattern polymorphism in the genus Phrvnobatrachus (Amphibia: Ranidae). Copeia
1974 (4): 823-832.

Tschudi, J. J. 1838. Classification der Batrachier, mit Beriicksichtigung der fossilen Thiere. Mem. Soc. Sci. nat.
Neuchatel 2.

Van Dijk, D. E. 1966. Systematic and field keys to the families, genera and described species of southern African anuran
tadpoles. Ann. Natal Mus. 18(2): 231-286.

Van Dijk, D. E. 1971. The zoocartographic approach to anuran ecology. Zool. afr. 6(1): 85-117.

Van Dijk, D. E. 1977. Habitats and dispersal of southern African Anura. Zool. afr. 12(1): 169-181.

Wager, V. A. 1965. The frogs of South Africa . Johannesburg: Purnell.

Manuscript accepted for publication 15 December 1978.

30

L yi '

/

fir 3 (3

ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

Ann. Cape Prov. Mus. (nat. Hist.)

8th MAY 1979

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

Printed by Cape & Transvaal Printers (Pty.) Ltd., Cape Town

BD8196

Note on the type-locality, distribution and juvenile coloration of Naja
nigricollis woodi (Serpentes: Elapidae) and an account of the colour-pattern
variation in intergrade populations

by

RICHARD C. BOYCOTT

(Port Elizabeth Museum)
and

W. D. HAACKE

(Transvaal Museum)

INTRODUCTION

The western black spitting cobra, Naja nigricollis woodi Pringle, 1955, occurs in southern
South West Africa (Namibia) from the vicinity of the Namib Desert Park east of Walvis Bay,
southwards through Great and Little Namaqualand to Porterville, and east to Prieska on the
southern bank of the Orange River.

The Port Elizabeth Snake Park recently acquired a juvenile specimen of this little-known
subspecies. As its coloration differs so markedly from that of fully-grown specimens, a
description is provided here, along with an assessment of past distribution records for the
subspecies and a redefinition of its type-locality. Naja nigricollis woodi integrades with Naja
nigricollis nigricincta Bogert in western and central South West Africa. The change in
colour-pattern from typical woodi to typical nigricincta within this region is also discussed.

THE TYPE-LOCALITY

The original description of Naja nigricollis woodi was based on three adult specimens
collected near Citrusdal in the western Cape Province, South Africa (Pringle, 1955). The
subspecies was named for its collector Mr John Wood of the South African Snake Farm, Fish
Hoek, near Cape Town. A precise locality was not given in the original description nor has one
been provided in any subsequent publication. However, Mr John Wood has confirmed that the
type-specimens were in fact collected on the farm “Keerom” which is situated 29 km SSE of
Citrusdal on the upper reaches of the Olifants River. It is suggested, therefore, that this farm
should be accepted as the type-locality. Following the locality citation system laid down by Greig
and Boshoff (in prep.) the locus (15' X 15' square) in which this farm falls is 3219 CC Keerom.
(It is coincidental that the farm name should be the same as the locus name.)

COLORATION NOTES AND DESCRIPTION

One of the diagnostic characters used by Pringle (1955) to distinguish this subspecies was
the black dorsal and ventral coloration in the specimens he examined. Attention has been drawn
to this uniform coloration by various authors (FitzSimons, 1962, 1970; Broadley, 1968;
Isemonger, 1968; and Visser and Chapman, 1978).

31

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 2, MAY 1979

Fig. 1. Naja nigricollis woodi Pringle — juvenile specimen from Aninauspas, Little Namaqualand (PEM 152120).

The description provided by Mertens (1971) was based on an adult specimen (SMF 66021)
from the farm Plateau and, although he stated that this adult female showed signs of grey
colouring on the ventral surface, his description does not differ markedly from that provided by
other authors. He questioned the locality data of SMF 21085 and commented on its atypical
characteristics. Consequently this record has been excluded from the present account. Broadley
(1974) drew attention to the difference in coloration between juveniles and adults, pointing out
that juveniles between 550 and 700 mm in total length are “grey with the head and neck black”.

The following more detailed description is based on a juvenile specimen (PEM 152120)
collected near Steinkopf (2917 BA KOSIES. 29° 12'30"S; 17° 37'39"E. Altitude 775 m) on the
road to Port Nolloth by Mr S. de Ridder of Springbok during April 1978 (Fig. 1). The vegetation
type is classed by Acocks (1975) as Namaqualand Broken Veld.

Dimensions : 486 (407 + 79) mm.

Sex and age : Juvenile female.

Scales : The dorsal scales number 21 rows at midbody and 23 in the neck region. Ventral
scales total 225 and subcaudals 65 using the method proposed by Dowling (1951). Anal scale
entire, the first 6 subcaudals entire the remainder divided.

Head shields : Intemasals 2, prefrontals 2, supraocular 1, parietals 2 with 9 bordering scales
(i.e. temporals and nuchals). Each nasal shield divided medially by nostril, preoculars 2,
postoculars 3, supralabials 6 the 3rd entering orbit, infralabials 8 the first three in contact with
anterior sublinguals.

Coloration : Each ventral scale has a narrow margin of grey on its posterior edge which is
absent in the first 30 ventrals. With this exception the entire ventral surface is black and the

32

BOYCOTT & HAACKE: LOCALITY, DISTRIBUTION AND COLORATION OF NAJA NIGRICOLLIS WOODI

throat is devoid of bands or stripes. The head, the sides and back of the neck are black; the rest of
the dorsal surface and the sides are grey. At the base of each grey dorsal and lateral scale there is
a margin of black; the overall effect is that of a fine reticulate pattern. This is further emphasized when
the skin is distended, as the black interstitial skin is thereby exposed. Proceeding posteriorly from
the neck region each body scale shows more grey, the black colouring diminishing until the
overall grey colouring becomes completely dominant (subjectively around 20 scales from the
parietals). A juvenile specimen from Porterville in the Transvaal Museum (TM 32784) conforms
to this description.

DISTRIBUTION

The range limits of this subspecies have confounded herpetologists for many years, partly
because of taxonomic uncertainty. Mertens (1955) mentions some specimens from the vicinity of
Aus and Narudas Slid, considering them to be melanistic individuals of N. nigricollis mossambica
Peters. FitzSimons (1962) believing it to represent a colour variety placed Naja mossambica

Fig. 2. • Naja nigricollis nigricollis Reinhardt, □ Naja nigricollis nigricincta Bogert, ■ Naja nigricollis woodi Pringle,

k Indicates nigricincta /woodi intergrade populations.

33

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 2, MAY 1979

Peters as a synonym under Naja nigricoilis Reinhardt and recognised nigricincta Bogert and
woodi as subspecies of the latter. Broadley (1968) separated N. mossambica and N. nigricoilis
and provisionally placed nigricincta and woodi as subspecies of N. mossambica to which
Mertens (1971) agreed. After examining a larger series of specimens from south-western Africa,
Broadley (1974) placed nigricincta and woodi as subspecies of N. nigricoilis. Amongst the 15
specimens of Naja nigricoilis woodi examined by him, he recorded that some showed signs of
intergradation with Naja nigricoilis nigricincta . Since Broadley’s publication the Transvaal
Museum has acquired additional specimens, some of which are intergrades. All in all sixteen
specimens, eleven in the Transvaal Museum (TM 30172, TM 30462, TM 36603, TM 36881, TM
37183, TM 39962, TM 48166, TM 49786, TM 50675, TM 52154, TM 52155), four in the
collection of the Namib Desert Research Station (NDRS R124, 125, 126, 127) and one in the
State Museum (SM R1640), which are representative of nine distribution records (Fig. 2) appear
to be Naja n. nigricincta /woodi intergrades. The intergrade zone between these two forms
appears to be in western and central South West Africa between the latitudes 22°S and 25°S
(Fig. 2). The distribution of Naja nigricoilis nigricincta extends from southern central South
West Africa northwards throughout the northern half of the country, excluding the Kungveld and
Kavango-Caprivi area, to south-western Angola where it intergrades with Naja nigricoilis
nigricoilis. The distribution records for Naja n. nigricincta (Fig. 2) have been taken from
FitzSimons (1962), Broadley (1974) and newer additions to the Transvaal Museum collection.
The records for Naja nigricoilis nigricoilis in south-western Angola and Zambia have been taken
from Broadley (1974). The specimen illustrated (Fig. 3) is an intergrade between Naja n.

Fig. 3. Naja nigricoilis nigricincta /woodi — juvenile specimen from Gobabeb,
Namib Desert Park (TM 36881).

34

BOYCOTT & HAACKE: LOCALITY, DISTRIBUTION AND COLORATION OF NAJA NIGRICOLLIS WOODI

nigricincta and Naja n. woodi, collected at Gobabeb in the Namib Desert Park (TM 36881) and
is typical for juveniles of that area.

Other juveniles of similar appearance have been collected, at the same locality (NORS
R127) and although without specific data, probably from the same general area (TM 52155). The
juvenile (TM 36603) from Hardap Dam, N. n. nigricincta according to Broadley (1974), is
actually considerably darker than the juveniles from the Kuiseb River and the dorsal, white
crossbands are even less distinct, for which reasons it must also be considered to be an intergrade.
Adults of intergrade populations show some consistency within a population. Along the Kuiseb
as far as the Desert Research Station, Gobabeb, they are dorsally more or less pitch black with a
black collar which usually covers the first 18 or 19 ventrals, followed by one or two salmon pink
to orange ventral scales (fading to white in preservative) and then irregularly alternating dark
grey and white ventrals with black flecks scattered over both background colours. The whitish
ventral markings in adults are best developed on the anterior third while the rear of the ventrum
tends to be more uniform grey with flecks. This pattern is also found on TM 49786 from
Kraaipoort, TM 48166 from Mooirivier and a specimen which was photographed in detail at
close range by the junior author between Homeb and Hudaob in the Kuiseb River. A second
specimen from Mooirivier (TM 39962) indeed gives the impression of a hybrid with Naja nivea
as described by Broadley (1974). A new specimen from Voigtskub (TM 52154), a flat skin, is
very close to a typical N. n. woodi, with a black dorsum and collar but the right tip of ventral 18
is white. This is equivalent to the salmon pink collar of the other dorsally black intergrades.

The adult specimens from Rostock (TM 30462), Onanesberg (NDRS R124) and Rossing
(TM 50675) have a ventral pattern similar to that described above but dorsally on a black
background indications of thin incomplete white transverse lines occur indicating their closer
position to N. n. nigricincta than the lower Kuiseb population. This fits in with the clinal
character of the intergrade zone, since the specimens from the Hakos Mountains (TM 24350, TM
34925) which are the closest true nigricincta, are rather dark specimens with comparatively
poorly developed white transverse bands. The nigricincta pattern is at its brightest in the
Kaokoveld and south western Angola. Unfortunately only one intergrade specimen (MBL 1965,
Quissange) has been described (Broadley, 1974) from the intergrade zone between the banded
N. n. nigricincta and typical nigricollis, which is uniform black above with a black throat and a
grey black speckled ventrum.

The change from one subspecies to another in Naja nigricollis through intergrade zones, is
apparently correllated with a clinal change in the number of ventrals (Broadley, pers. comm.).
Since further population analyses are pending by Broadley, this paper concentrated on clearing
up some of the distribution record errors of N. n. woodi, to add new records and to provide more
information on the change in pattern of populations of the intergrade zone between woodi and
nigricincta without presenting any scale count data.

A list of localities is provided in Table 1. Ten of the localities post-date those provided by
Broadley (1974). Locus citation follows Greig and Boshoff (in prep.). As the spelling of southern
African place names is notoriously variable, the locality names are spelt according to the latest
editions of the 1:50 000 Topographical series of South Africa and South West Africa.

GENERAL

The specimen figured by FitzSimons (1962) as Naja nigricollis nigricollis Reinhardt (TM
24653) from Que Que is in fact Naja mossambica Peters. In the field notes of Naja nigricollis
(referring to mossambica as well) FitzSimons (1962) stated that they will usually “spit” from
the reared position and also from ground level, without raising the head. N. mossambica is less
inclined to rear and spread its hood than other cobras (Visser and Chapman, 1978). From 30
observations in the Port Elizabeth Snake Park during which both species were approached, Naja
nigricollis woodi reared and elevated its head on all occasions. Naja mossambica raised its head

35

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 2, MAY 1979

on only 3 occasions, and then only when directly threatened. Thus a distinct behavioural
difference appears to exist between the two species. In captivity N. n. woodi has been known to
eat toads, Bufo sp. (Wood, pers. comm.); the Port Elizabeth Snake Park specimen accepted
laboratory mice.

MUSEUM ABBREVIATIONS

MBL

NDRS

NM

PEM

SAM

SM

SMF

TM

WNM

= Museum Bocage, Lisbon, Portugal.

= Namib Desert Research Station, Gobabeb, South West Africa (Namibia).
= Natal Museum, Pietermaritzburg, South Africa.

= Port Elizabeth Museum, Port Elizabeth, South Africa.

= South African Museum, Cape Town, South Africa.

= State Museum, Windhoek, South West Africa (Namibia).

= Senckenberg Museum, Frankfurt, West Germany.

= Transvaal Museum, Pretoria, South Africa.

= Wiesbaden Naturhistorisches Museum, Wiesbaden, West Germany.

ACKNOWLEDGEMENTS

The authors would like to express their gratitude to the following for providing valuable
information: Dr D. G. Broadley (Umtali Museum), Dr J. A. Pringle (Natal Museum), Dr G. R.
McLachlan (South African Museum), Miss R. M. Tietz (McGregor Museum), and Mr M. Penrith
(State Museum). Mr S. de Ridder of Springbok is thanked for supplying the juvenile specimen
and Mr John Wood of the South African Snake Farm, Fish Hoek, for furnishing details regarding
the type-locality. John Greig (Herpetologist, Cape Department of Nature and Environmental
Conservation) is thanked for supplying information concerning the localities. The Board of
Trustees of the Port Elizabeth Museum and the Director, Dr John Wallace are thanked for
permission to publish this paper. Dr J. A. Pringle, John Greig and Allan Batchelor read the
manuscript and provided useful suggestions.

REFERENCES

Acocks, J. P. H. 1975. Veld types of South Africa. Mem. bot. Surv. S. Afr. 40: 1-128.

Broadley, D. G. 1968. A review of the African cobras of the genus Naja (Serpentes: Elapinae). Arnoldia Rhod. 3(29):
1-14.

Broadley, D. G. 1974. A review of the cobras of the Naja nigricollis complex in south western Africa (Serpentes:
Elapidae). Cimbebasia (A). 2(14): 155-162.

Dowling, H. G. 1951. A proposed standard system of counting ventrals in snakes. Br. J. Herp. 1(5): 97-98.
FitzSimons, V. F. M. 1962. Snakes of Southern Africa. Cape Town: Purnell.

FitzSimons, V. F. M. 1970. A field guide to the snakes of Southern Africa. London: Collins.

Greig, J. C. and Boshoff, A. F. (in prep.). Biological distribution mapping in southern Africa: the methodology of
locality designation using the latitude/longitude grid system.

Isemonger, R. M. 1968. Snakes of Africa. Cape Town: Books of Africa.

Lampe, E. 1911. Erster Nachtrag zum Katalog der Reptilien- und Amphibien — Sammlung des Naturhistorischen
Museums der Stadt Wiesbaden. Jahrb. Nass. Ver. Naturk 64: 137-236.

Mertens, R. 1955. Die Amphibien und Reptilien Siidwestafrikas. Aus den Ergebnissen einer im Jahre 1952 ausgefiihrten
Reise. Abh. Senckb. Naturf. Ges. 490: 1-172.

Mertens, R. 1971. Die Herpetofauna Siidwest-Afrikas. Abh. Senckb. Naturf. Ges. 529: 1-110.

Pringle, J. A. 1955. A new subspecies of the spitting cobra Naja nigricollis Reinhardt from the Cape Province. Ann.
Natal Mus. 13(2): 253-254.

Visser, J. and Chapman, D. S. 1978. Snakes and Snakebite . Cape Town: Purnell.

Manuscript accepted for publication 29 January 1979.

36

Table 1.

Details of past and present distribution records for Naja nigricollis woodi Pringle and N. n. nigricincta /woodi intergrades

BOYCOTT & HAACKE: LOCALITY, DISTRIBUTION AND COLORATION OF NAJA NIGRICOLLIS WOODI

37

ANN. CAPE PROV. MUS. (NAT HIST.) VOL. 13 PT 2, MAY 1979

U

<

«<

-o

i>

c:

JS

; -o E
! -a “ c
c §
u 2

OB

E

o

£ 5 =

<u <-

— DO >
«"U-S
« x u-i 73

I <N — O >

I ,„vo2 «
JS o-^

xi ~ o

«n«2"

? \0 -*> 3 XJ

>> <L>

a: co

o ON ^

— h o

S w2 M

3 2 C O

>,Z
E Ji

• — • — 73

b <*> c5

<u

N O

2 £ cS

„ D

<y ut
On ^

C — y

<D 9- £
w 2 do

c S '5

Si T3
c«

Z &

ON Q>
w ^

CQ £

I 1

“ o

^ (J O
os P .s:
x ^ C
U ^

■ m ,c>0

O '-5 5

= £ si
o <d
U2^

T3

C«

6b

£r

On

CQ 'C

_Oo

<D T{ 03

a

hc<.e

6j

On q.

C 9

>,-4.

jv a
03 s:

2 *Q
CQ C

"O

03

6)

x ^ ^

a) s:

• *”■ T“

Q .
b

k^J c/5 .®C

^ ^ s

S ? c
O <L>

uz^

B cJ t>

73 p S

^ . .g

H uj ^

. C/5 C>C

U'-B S

S £ si
o <u
UZ^

>>

o

o

c*

§ d

<u s:

<| |
•^’O 5

d ^ si
o <u

CQ

U

(N NO
NO NO
On -<
On OC
cn rf

SS

H H

rnrninc S'°SvoSr'

SSSQ-Q2q2

HHhZrtZojZDi

a! 3
Q 23
Za:

<

£

D

<<

s*

CQ CQ

E

oj

Q

§■<

"O

Xw

■g <

o •>
Z on

s .

d oo

X) .

0) M

■s S

X CL,
O ^

o ^

<u

X X

>2
•- 03

x .

<L» X
cn &

'3 2
U Z

-§<

073

X

«« J
g<

o CQ

.s<
s>
:0 >,
CC I CQ

38

frit 3

d P

ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

Ann. Cape Prov. Mus. (nat. Hist.)

VOLUME 13 • PART 3
28th SEPTEMBER 1979

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

Printed by Cape & Transvaal Printers (Pty.) Ltd., Cape Town

BD 8373

Printed by Cape & Transvaal Printers (Pty.) Ltd., Cape Town

BD8373

A new species of Megerlina (Brachiopoda) from the Pleistocene of Zululand, South Africa

by

NORTON HILLER

Department of Geology, Rhodes University, Grahamstown 6140, South Africa

(with 3 figures)

ABSTRACT

A new fossil species of the Terebratulacean brachiopod, Megerlina, is described from late
Pleistocene limestones in the False Bay area of Lake St Lucia, Zululand. It is designated
Megerlina levis sp. nov.

CONTENTS

Page

Introduction 39

Systematic description 40

Acknowledgements 44

References 44

INTRODUCTION

The present specimens were collected by Dr D. K. Hobday who sent them to the author
for identification. They occurred, in local concentrations, within limestones of late Pleistocene
(Eemian) age exposed in small outcrops at Lister’s Point and Picnic Point on the western shores
of the False Bay area of the Lake St Lucia lagoonal complex on the Zululand coastal plain (Fig.
1). Associated with the brachiopods were a variety of corals, gastropods, bivalves and cirripeds
as well as echinoderm and sponge fragments and possible crab remains. Vertical tubes with
horizontal branches of the ichnogenera Thalassinoides and Ophiomorpha were also present. The
overall fauna and lithology of the enclosing sediments indicate an unrestricted, high energy,
shallow marine environment not far from shore, with water temperatures slightly warmer than at
present (Hobday, 1976: 103-105).

39

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 3, SEPTEMBER 1979

Fig. 1. Map showing location and geology of the Lake St Lucia lagoonal complex (After Hobday 1976).

SYSTEMATIC DESCRIPTION

Class ARTICULATA

Superfamily Terebratulacea King, 1850
Family Kraussinidae Dali, 1870
Genus Megerlina Eudes-Deslongchamps, 1884

Type-species: Kraussia lamarckiana Davidson, 1852: 80 by original designation.

Megerlina levis sp. nov. (Figs. 2, 3)

Type material

Holotype: AM-4778

Paratype 1: AM-4779

Paratype 2: AM -4780

Paratype 3: AM-4781

The specimens are housed in the Albany Museum, Grahamstown, South Africa.

40

NORTON HILLER: NEW SPECIES OF MEGERLINA (BRACHIOPODA)

Diagnosis

Subpentagonal Megerlina with carinate pedicle valve and sulcate brachial valve, and lacking
any radial ornamentation.

Material and horizon

Fourteen pairs of conjoined valves and fragments of several other pedicle and brachial
valves from late Pleistocene deposits on the Zululand coast.

Description

Small ventribiconvex shells of subpentagonal to transversely subcircular outline; maximum
width about midvalve; hinge-line width about three-quarters of maximum width; postero-lateral
margins (beak ridges) straight, subtending an angle of about 120° at the umbo. Pedicle valve over
nine-tenths as long as wide and just over one-third as deep as long. Lateral profile evenly
convex; anterior profile convex with narrow, round-topped median fold extending from umbo and
separating flat to very gently convex flanks; profile of younger forms more evenly convex, fold
not as pronounced. Umbo suberect; pedicle opening subinesothyridid. Small, but conspicuous,
triangular palintropes developed on either side of delthyrium; deltidial plates narrow.

A

Fig. 2. A. Diagram of the cardinalia of Megerlina levis sp. nov. b — buttress; c.p. — cardinal process; f.p. — fulcral plate;
s — socket; s.r. — socket ridge; About 5,5 times natural size. B. C. D. Posterior, ventral and lateral views of brachidium.

About 12 times natural size.

41

ANN. CAPE PROV. MUS. (NAT. HIST ) VOL. 13 PT 3, SEPTEMBER 1979

Brachial valve about four-fifths as long as wide and a little over one-fifth as deep as long.
Lateral profile gently convex with maximum convexity near umbo; anterior profile gently
convex but with narrow median sulcus which arises at 2 mm growth stage and expands and
deepens gradually towards anterior margin. Sulcus of young forms shallower than in adults.
Ornamentation of faint concentric growth lines and single prominent growth lamella near anterior
margin of some specimens; otherwise shells are smooth. Punctae well seen over whole of shell
surface, with density of 120-180 per sq. mm.

Ventral interior with stout hinge teeth without supporting dental plates; pedicle collar very
slightly excavate anteriorly. Dorsal interior largely obscured by infilling sediment but one
specimen shows weakly developed, elliptical, knob-like cardinal process between strong, widely
divergent socket ridges. These are buttressed on their inner sides by a pair of short plates which
converge on to valve floor to define notothyrial cavity (Fig. 2A). About mid-valve a low median
septum supports a pair of anteroventrally directed diverging plates, which are slightly concave
towards one another; each bears a blunt point at the distal end which reaches almost to the floor
of the opposite valve and a curved accessory ledge is present on the outside of each plate. From
about two-thirds to three-quarters the way up the posterior edge of each plate, a slightly curved
prong protrudes (Fig. 2B-D). These prongs have been interpreted as elements of a rudimentary
transverse band in the case of the closely related genus Kraussina (Elliot 1949: 541) but
Thomson (1927: 225), Jackson (1952: 30) and Cooper (1973: 22) think they represent the
anterior ends of the descending branches of a loop. However, the posterior ends of such branches
have not been described. Details of the musculature are obscured.

Dimensions

Holotype:

pair of conjoined valves

length
13,5 mm

width
15,3 mm

Paratype 1:

pair of conjoined valves

12,2 mm

13,2 mm

Paratype 2:

fragment of brachial valve

—

—

Paratype 3:

fragment of conjoined valves

—

—

Discussion

The present specimens

of Megerlina can immediately be

distinguished

from M.

lamarckiana (Davidson), from the shallow waters around the south-east coast of Australia, M.
davidsoni (Velain) from St Paul Island, and from M. capensis (Adams & Reeve) and M. striata
Jackson, both of which occur off the south-west coast of Cape Province. These modem forms are <
all distinctly ribbed, although this ribbing may develop upon a smooth early growth stage. Another
form which occurs off the South African coast is M. pisurn (Lamarck) which is described by
Jackson (1952: 30) as being nearly smooth but with faint traces of costation; the new species
shows no traces of costation and can be further distinguished by its narrower, deeper sulcus.
Jackson (1952: 33) ends his discussion of M. pisum by noting that the internal characters of the
type specimens from Mauritius were unknown to him and that his description was based on i
South African specimens which he identified by their outward appearance only. He states that if il
further specimens from Mauritius should prove to differ in their cardinalia and brachidium then
the South African shells should be known as M. natalensis. However, Cooper (1973: 22)
described a M. natalensis (Krauss) from the south-east coast of Cape Province, a locality also
quoted by Jackson for his M. pisum.

In synonymy with M. natalensis. Cooper places Terebratula natalensis Krauss 1844 and
Terebratula algoensis Sowerby 1847, two forms which are also included in the synonymy list
given by Jackson for M. pisum. There appears to be a certain amount of confusion over the
identity of these two forms but the shells described by Cooper are quite distinctly ribbed and
must be separate from M. pisum as well as the specimens described herein.

42

NORTON HILLER: NEW SPECIES OF MEGERLINA (BRACHIOPODA)

Smith (1901: 1 16) records Kraussina atkinsoni (Woods) from Algoa Bay but Jackson (1952:
32, 33) regards the specimens as juvenile forms of M. pisum, an identification he also gives to a
small, smooth sulcate shell with strong growth lamellae, that was dredged off Durban and
originally identified as K. atkinsoni by G. B. Sowerby. The superficial resemblance of these
juvenile forms to Megerlina levis sp. nov. tempts one to speculate that the Zululand species may
be ancestral to M. pisum.

Fig. 3. Megerlina levis sp. nov. A, B, C, D. Dorsal, lateral, anterior and ventral views of conjoined valves, Holotype
, (AM-4778). Respectively x 2,5, X 2,7, x 2,8, x 2,5. E. Dorsal view of conjoined valves, (AM-4779) x 2,7. F, G.
Ventral and lateral views of a broken brachial valve interior, (AM-4780). Respectively x 2,7, X 4,1. H. Posterolateral

view of brachidium, (AM-4781) x 2,2.

43

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 3, SEPTEMBER 1979

In discussing the generic position of K. atkinsoni, Jackson (1952: 32, 33) says that he does
not consider it a true Kraussina and that it should be removed from that genus. Although it
resembled M. lamarckiana, Jackson was doubtful whether the species should be included in
Megerlina without further investigation. However, Hatai (1965: H834) gives the time range for
Megerlina as Neogene (Tasmania)- Recent and the type specimens of K. atkinsoni are from
Miocene deposits in southern Tasmania. Thus it seems that K. atkinsoni has indeed been
transferred to Megerlina and, apart from the present specimens, is the only other fossil form of
the genus so far recorded. It is a ribbed form and is therefore distinct from M. levis sp. nov.

ACKNOWLEDGEMENTS

The author is indebted to Dr Hobday for providing the specimens for study, and to Mr W.
Holleman, of the Albany Museum, for his assistance with the photography.

REFERENCES

Cooper, G. A. 1973. Verna’s Brachiopoda (Recent). Smithson. Contr. Paleobiol. 17: 1—51.

Elliot, G. F. 1949. The brachial development of Kraussina (Brachiopoda). Ann. Mag. nat. Hist. (12) 2: 538-546.
Hatai, K. 1965. Kraussinidae. In: Williams, A. et al. in Treatise on Invertebrate Paleontology, ed. R. C. Moore, Part H, [
Vol. 2, Brachiopoda. University of Kansas Press, 927 pp.

Hobday, D. K. 1976. Quartemary sedimentation and development of the lagoonal complex. Lake St Lucia, Zululand.
Ann. S. Afr. Mus. 71: 93-113.

Jackson, J. W. 1952. A revision of some South African Brachiopoda, with descriptions of new species. Ann. S. Afr.
Mus. 41 (1): 1-40.

Smith, E. A. 1901. On South African marine shells, with descriptions of new species. J. Conch., Loud. 10: 104-116.
Thomson, J. A. 1927. Brachiopod morphology and genera ( Recent and Tertiary). Wellington, N.Z.: Dominion Museum.
(New Zealand Board of Science and Art. Manual 7), 338 pp.

Manuscript accepted for publication 28 May 1979.

/'i- (p /

A i-t

ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

Ann. Cape Prov. Mus. (nat. Hist.)

VOLUME 13 • PART 4
11th JANUARY 1980

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

Printed by Cape & Transvaal Printers (Pty.) Ltd., Cape Town

BD 8594

Ethological notes on Kohliella alaris Brauns (Hymenoptera: Sphecidae:
Larrinae) in the Eastern Cape Province of South Africa

by

F. W. GESS

(Albany Museum, Grahamstown)
and

S. K. GESS
INTRODUCTION

The genus Kohliella Brauns, 1910 is endemic to southern Africa and includes two species
poorly represented in collections.

Kohliella alaris Brauns, the type species, black and ranging in length from 6- 1 1 mm, was
described from Willowmore (Cape Province) (Brauns, 1910: 669) and was subsequently recorded
from Hex River (Cape Province) and Bulawayo (Zimbabwe) (Arnold, 1924: 43) and from two
localities on the Cape Peninsula (Beaumont, 1967: 510). In the Albany Museum collection it is
represented by specimens collected by the authors at New Year’s Dam, Alicedale and on Hilton,
Grahamstown (both Cape Province). K. stevensoni Arnold, the second species, was described
from Sawmills (Zimbabwe) (Arnold, 1924: 42).

It has been pointed out by Bohart and Menke (1976: 286) that, although Kohliella is similar
to Tachysphex in general facies, only a few features such as the form of the collar, the male
forefemoral notch, and the bare pygidial plate are common to both, and that Kohliella is probably
best regarded as a specialized relic.

Nothing has hitherto been published regarding the biology of the genus.

The present paper is the sixth of a series of publications dealing with the ethology of certain
solitary wasps occurring at Hilton, a farm situated 18 kilometres WNW. of Grahamstown (33°
19' S, 26° 32' E) in the Albany Division of the Eastern Cape Province of South Africa. An
account of the climate and vegetation of Hilton has previously been given (Gess and Gess, 1974:
191-192).

THE NESTING OF KOHLIELLA ALARIS BRAUNS

Description of the nesting sites

At Hilton Kohliella alaris nests in sand on the floor of a sandpit and in close proximity to
the sandpit in places where the vegetation has been removed leaving the sand bare or with very
sparse plant cover (Figs 1 and 2). The sand, light coloured and fine grained, is derived from the
weathering of Witteberg Quartzite and is of alluvial origin having been deposited upon its flood
plain by a seasonal tributary of the New Year’s River. K. alaris tends to nest in pseudo-colonies
making use of only small areas within nesting sites which appear to offer suitable conditions for
nesting throughout.

45

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 4, JANUARY 1980

Fig. 1. Hilton, 12.i. 1978. Portion of sand pit with figure marking a nesting site of Kohliella alaris.

Fig. 2. Hilton, 12. i. 1978. Entrance to a nest of Kohliella alaris in nesting site shown in Fig. 1.

46

GESS & GESS: NOTES ON KOHLIELLA ALARIS BRAUNS

Flight period

Males and females of K. alaris have been collected at Hilton from early December to early
March and nesting takes place throughout this period. During the summer of 1977-78, which
was a particularly successful season for this species, nesting was at its height in mid- January.

Whereas most of the sphecids nesting in the sandpit commence their daily nesting activities
before the hottest time of day and continue them into the early afternoon or on the hottest days
cease activity before noon, K. alaris never appears before noon and rather commences its
activities after noon, when the heat of the day is past its peak, and on the hottest days may delay
its nesting activities until as late as 4. 15 p.m., with the peak of nesting activity half an hour later.

Males are present in the nesting area and have been seen to attempt to mate with females in
close proximity to the nests.

Plants visited by adult wasps

There are no records of K. alaris visiting plants. However, trees and shrubs must be visited
by the females when hunting for it is in such situations that the prey occurs. It is probable that the
wasps visit a belt of shrubs and small trees which occurs along the banks of the seasonal tributary
near the sandpit. Although K. alaris was not observed in this situation, both males and females
were caught in a Malaise trap erected between shrubs on the edge of this belt.

Identification of the prey

Twenty-six prey were recovered from K. alaris or its nests and without exception were
nymphs of the Tree Cricket Oecanthus capensis Saussure (Gryllidae: Oecanthinae). Adult
O. capensis in the Albany Museum were collected at Brak Kloof, a farm adjoining Hilton, during
the months February to April.

Details of the prey and of their captors are given in Tables 1 and 2. It will be seen that the
prey is mutilated by the wasp in that the antennae are cut off short (Fig. 3). Mutilation must take
place immediately after prey capture as prey being transported to the nest are already in the
“dressed” condition.

Fig. 3. Contents of the cell of nest 1 1 (see Table 1) consisting of three tree crickets, Oecanthus capensis, the third from
the left bearing the egg of the nest builder shown on right, (x 2,8)

47

Table 1. Details pertaining to Kohliella alaris females, their prey and eggs associated in nests.

GESS & GESS: NOTES ON KOHLIELLA ALARIS BRAUNS

Table 2. Details pertaining to Kohliella alaris females and their prey associated during prey transport.

Date

Prey sex &
developmen-
tal stage

Prey

weight

(mg)

Prey

length

(mm)

Mutilations to
prey

Wasp

weight

(mg)

Wasp

length

(mm)

16.ii.76

female

nymph

43

12,5

Both antennae
shortened

18

9,3

16.ii.76

male

nymph

21

8,3

Both antennae
shortened

24

9,5

9.xii.77

male

nymph

—

11,0

Both antennae
shortened

—

—

20.xii.77

nymph

—

6,3

Both antennae
shortened

—

10,0

1 7. i.78

female

nymph

46

14,5

Both antennae
shortened

—

c. 10

1 7. i.78

male

nymph

26

11,0

Both antennae
shortened
Right hind leg
missing

25

10,0

Description of the nest (Figs 4a and b, 5a and b)

The nest of K. alaris consists of a short trench, approximately 10 mm long, leading to a
subcircular entrance hole, slightly wider than high, average width 4,3 mm (sample of 12), from
which a shallow entrance passage of similar diameter extends from 20-35 mm (sample of 9) and
reaches an average depth of 1 3 mm (sample of 7) below the surface of the ground. From this
shallow burrow one to several secondary branches of a steeper gradient lead off, some of which
end in a more or less horizontal single cell at an average depth of 35 mm (range 16-70 mm,
sample of 15). Some passages appear to remain unused and are free from sand whereas the
passages which end in a cell are filled with loose sand. A secondary passage may be excavated in
the same direction as the entrance passage or it may make a sharply acute to obtuse angle with it.
Similarly, a cell may have its long axis following the same direction as the secondary passage or
' it may be at an angle to it. Such secondary passages which terminate in a cell are 15-60 mm long
; (sample of 1 1) and 4-5 mm in diameter and the cells have an average length of 16 mm (sample
of 8) and an average diameter of 6,6 mm (sample of 9), therefore being of a slightly larger bore
than that of the secondary passages.

’ Method of construction of the nest, provisioning and oviposition

K. alaris is a sand raker. After the wasp has selected a nesting site, nest construction is
immediately initiated by her digging and raking away the soil. When the sand which is excavated
in the construction of the nest begins to accumulate forming a pile, the wasp spreads it from side
to side using a light dancing motion. Any particles which are too big to be extracted from the
excavation by raking are carried out by the wasp in her mandibles. At the start of nest excavation
the wasp is easily distracted by passing insects such as ants but makes no attempt to chase them.
However, as nest construction becomes more advanced she becomes more determined and
aggressive making attempts to drive away passing insects.

Hunting takes place after a cell has been excavated and the nest temporarily sealed with
sand. The wasp having located, captured, stung and mutilated her prey she flies with it to the nest
holding it beneath her with its head facing the direction of travel. She alights close to the

49

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 4, JANUARY 1980

100m m

50

GESS & GESS: NOTES ON KOHLIELLA ALARIS BRAUNS

Fig. 5a and b. Plans of nests 1 and 7 of Kohliella alaris .

51

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 4, JANUARY 1980

entrance to the nest, puts down the prey, rakes the sand out of the nest entrance, enters, turns
around within, and draws the prey in head first.

Fourteen nests were excavated from which the contents of ten provisioned or partially
provisioned cells were obtained (see Table 1) and were found to consist of twenty prey, seven of
which bore wasp eggs. In a cell several prey are positioned venter up facing the inner end of the
cell and parallel to each other but with every prey slightly in advance of that succeeding it. The
prey are incompletely paralysed and exhibit occasional trembling movements of the palps and
continuous pumping respiratory movements of the abdomen.

The egg of K. alaris is slightly curved, pearly white and 2,6 mm long (average of 5) and
0,5 mm wide at its mid length. It is attached by its anterior end to the underside of the prothorax of
one of the prey, posterior to either the right or left prothoracic coxa, and extends transversely
across the venter of the prey (Fig. 3). For six of the eggs it was possible to establish the order of
introduction into the cell of the prey to which they were attached. Three eggs were on the first
prey, two on the second prey and one was on the fourth prey. In the instance of the fourth prey
being selected for oviposition it was the only one of any considerable size weighing 37 mg
whereas the first three weighed only 5, 4,5 and 7,0 mg respectively. It is therefore likely that if
the first prey to be captured are very small oviposition is postponed until one of a suitable size is
obtained.

The secondary passage leading to a cell is usually filled with loose sand whereas the
entrance passage is left clear. When leaving the nest the wasp closes the entrance. She stands
inside the burrow and rakes sand in towards herself and through underneath to behind herself,
gradually advancing out of the nest. When only a small depression is left she rakes sand over it
from all directions until it is completely obscured. Three nests which had been seen to be closed
in this way were excavated and were found to have as yet incompletely provisioned cells. It is
therefore clear that the wasp closes her nest between visits.

Parasites

Five miltogrammine sarcophagids were reared from maggot infested prey recovered from a i
cell excavated on 1 8. i. 1978. The flies emerged from their puparia on 6.ii. 1978.

DISCUSSION

Kohliella with Parapiagetia, Holotachysphex and Prosopigastra may be grouped around lj
Tachysphex, one of the dominant sphecid genera and one of the most highly evolved members of (.
the tribe Larrini (Bohart and Menke, 1976: 269). As these five genera together form a distinct tj
evolutionary branch within the subtribe Tachytina and as nothing has hitherto been published I
concerning the biology of Kohliella it is of interest to compare the five genera with respect to )j
various aspects of nesting ethology.

A considerable number of papers on the biology of Tachysphex species have appeared,
important ones being listed by Bohart and Menke (1976: 270). A useful summary of the known i
biology of the Palaearctic species has been provided by Pulawski (1971: 16-20). In contrast little
has been published on the biology of Prosopigastra and almost nothing on that of Parapiagetia.
For both genera the known biology has been reviewed by Bohart and Menke (1976: 285 and 281 i
respectively) and for Prosopigastra additional data have been presented by Pulawski (1979) who
has also summarized the life history of the genus. An account of the nesting of a species of lj
Holotachysphex, H. turneri (Arnold) in trap-nests has been published by Gess (1978: 209-215).

In addition, since the publication of this account, natural nests of H. turneri have been found by
the authors at Hilton in old abandoned galleries of the carpenter bee, Xylocopa caffrariae
Enderlein in the hollow intemodes of dry culms of Phragmites australis, a reed fringing
permanent and semi-permanent water bodies.

52

GESS & GESS: NOTES ON KOHLIELLA ALARIS BRAUNS

The nests of Kohliella like those of Tachysphex and Prosopigastra are situated in the
ground, in friable soil; however, Holotachysphex by contrast nests up off the ground in hollow
plant stems. The nests of Parapiagetia have not as yet been located but the presence in the
female of fore-tarsal sand rakes and of a pygidial plate indicates that this genus like Kohliella is
ground nesting.

Kohliella like almost all species of Tachysphex excavates its nest itself whereas Pro-
sopigastra, Holotachysphex and the exceptional species of Tachysphex modify pre-existing
cavities, frequently the galleries of other aculeate Hymenoptera.

The nests of Kohliella that were investigated were all still under construction; eight
contained a single provisioned or partially provisioned cell and one contained two cells. It is
therefore not known whether the number of cells constructed ever exceeds two. Tachysphex nests
may be one- or two-celled or multicellular, those of Prosopigastra are multicellular and
Holotachysphex turneri nests have been found with one, two and three cells. Thus it appears that
all four genera show a tendency to construct more than one cell per nest.

Kohliella unlike Prosopigastra and Holotachysphex which leave the nest entrance open
during provisioning maintains a temporary closure during this period. Tachysphex displays
variability in its behaviour in this respect. As Kohliella maintains a temporary closure it may be
assumed that like Tachysphex, Prosopigastra and Holotachysphex it constructs a final closure.

Kohliella like Holotachysphex, Parapiagetia and some species of Tachysphex preys upon
Qrthoptera. Other species of Tachysphex, however, prey upon Dictyoptera and Prosopigastra
preys upon Hemiptera. In all genera nymphs are generally taken. The report that an unknown
Parapiagetia species was seen transporting a caterpillar should, it is considered, be treated with
great caution.

Kohliella like Tachysphex transports its prey held beneath its body and facing the direction
of travel. Kohliella, Tachysphex and Prosopigastra are known to carry their prey in flight and
judging from the small size of its prey and the situation of its nest Holotachysphex probably also
transports its prey in flight. When the prey is large and heavy relative to the wasp, Tachysphex is
known to transport it along the ground in short hopping flights.

Kohliella like some species of Tachysphex deposits the prey on the ground at the nest
entrance, opens and enters the nest, turns around within it and then draws in the prey unlike
Prosopigastra and some other species of Tachysphex which enter the nest directly.

Kohliella like Holotachysphex and Prosopigastra provisions each cell with several prey,
however, Tachysphex when large prey are taken may provision with a single prey. In all four
genera paralysis of the prey is incomplete. Kohliella like Holotachysphex and those Tachysphex
species which provision their nests with Acrididae positions its egg attached immediately
posterior to one of the prothoracic coxae and extending transversely across the venter of the prey.

From the above comparison of various aspects of the nesting ethology of the five genera it is
evident that the latter form a fairly close-knit group. Tachysphex, the species-rich dominant
genus around which the other genera are grouped shows a considerable ethological latitude or
plasticity whereas the ethology of the other genera appears more circumscribed. However, this is
likely to be merely an expression of the fact that the other genera have far fewer species and that
data concerning these are limited. Among those ethological aspects considered Kohliella does not
manifest any character which is unique to itself and which is not found also in the genus
Tachysphex seen as a whole (though of course if compared with individual species of Tachy-
sphex various differences do become apparent). On the other hand the ethology of Kohliella
differs in several aspects from that of both Holotachysphex and Prosopigastra. (Not enough is
known of the ethology of Parapiagetia to allow comparison.) Ethologically Kohliella therefore
has the greatest affinity with Tachysphex, the affinity being greater than that shown by either
Holotachysphex or Prosopigastra with that genus.

53

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 4, JANUARY 1980

SUMMARY

Some aspects of the ethology of Kohliella alaris Brauns (Hymenoptera: Sphecidae:
Larrinae) in the Eastern Cape Province of South Africa are described. In the study which was
based upon a series of nests excavated in sand attention is given to the description of the nesting
sites, flight period, plants visited by adult wasps, identification of the prey, description of the nest
and method of its construction, provisioning and oviposition. The nesting of Kohliella is
compared with that of the genera Parapiagetia, Holotachysphex, Prosopigastra and Tachysphex
and is found to be very similar to that of the last named.

ACKNOWLEDGEMENTS

The authors wish to thank Mr T. C. White of the farm Hilton for his much appreciated
kindness over the years in allowing them free access to his land. The senior author is grateful to
the C.S.I.R. for a running expenses grant for field-work during the course of which the present
observations were made.

REFERENCES

Arnold, G. 1924. The Sphegidae of South Africa Part 5. Ann. Transv. Mus. 11 (1): 1-73.

Beaumont, J. de 1967. Hymenoptera: Sphecidae. In Hanstrom et al. eds, South African animal life. Vol. 13. Stockholm:
Swedish Natural Science Research Council, pp. 502-512.

Bohart, R. M. and Menke, A. S. 1976. Sphecid wasps of the world: a generic revision. Berkeley: University of
California Press.

Brauns, H. 1910. Neue Sphegiden aus Siidafrika Deutsche Ent. Zeitschr. 1910: 666-670.

Gess, F. W. 1978. Ethological notes on Holotachysphex turneri (Arnold) (Hymenoptera: Sphecidae: Larrinae) in the
Eastern Cape Province of South Africa Ann. Cape Prov. Mus. (nat. Hist.) 11 (11): 209-215.

Gess, F. W. and Gess, S. K. 1974. An ethological study of Dichragenia pulchricoma (Arnold) (Hymenoptera:
Pompilidae), a southern African spider-hunting wasp which builds a turreted, subterranean nest. Ann. Cape
Prov. Mus. (nat. Hist.) 9 (11): 187-214.

Pulawski, W. 1971. Les Tachysphex de la region palearcticjue occidentale et centrale. Wroclaw: Zaklad Zool. SysL,
Polskiej Akad. Nauk.

Pulawski, W. 1979. A revision of the world Prosopigastra Costa (Hymenoptera Sphecidae). Polskie Pismo Ent. 49:
3-134.

Manuscript accepted for publication 21 September 1979.

54

<Yi+

fir&B
A) H

ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

Ann. Cape Prov. Mus. ( nat . Hist.)

VOLUME 13 ® PART 5
24th JUNE 1980

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

Printed by Cape & Transvaal Printers (Pty) Ltd, Cape Town

BD8876

A new species of Lemniscomys (Muridae) from Zambia

by

E. VAN DER STRAETEN

(Rijksuniversitair Centrum, Antwerp)

ABSTRACT

A new species of Lemniscomys , L. roseveari sp. nov. is described and the status of the
components of the L. griselda species-complex is examined by statistical analyses.

INTRODUCTION

During an overall revision of the species-group Lemniscomys griselda the author had the
opportunity of examining an important series of a “single-striped grass mouse” collected in
Balovale, Northern Rhodesia (now Zambia) by G. C. Shortridge and T. D. Carter in 1938.

The conclusion that this “grass mouse” belongs to an undescribed species was the result of
an extensive statistical analysis, using canonical analysis and discriminant analysis (see Van der
Straeten & Van der Straeten-Harrie, 1977, and Van der Straeten & Verheyen, 1978).

The author is very pleased to name the mouse Lemniscomys roseveari sp. nov. in honour of
D. R. Rosevear whose book The Rodents of West Africa was a great stimulus and help when he
started his studies on African Muridae.

A complete discussion of the Lemniscomys griselda species-group will be the subject of a
later publication.

DESCRIPTION OF LEMNISCOMYS ROSEVEARI SP. NOV.

Measurements in mm of adult Lemniscomys roseveari sp. nov. are given in Table 1.

General colour of the back brownish, anteriorly more greyish, posteriorly more yellow-
ochre, most like L. linulus from the Ivory Coast. Well defined black mid-dorsal stripe
commencing between the ears. Back with pale-tipped hairs (buff or ochraceous-buff) forming
small spots; the spots not forming a well defined pattern. More brownish than L. rosalia (cf.
sabulata) and L. griselda but a little paler than L. rosalia (cf. spinalis). Ears and feet with
yellow-ochre hairs. The hairs of the underside are pure white; underside edged with a buff line.
Tail longer than head and body, about 110% of the head and body length, dark above,
ochraceous-orange or yellow-ochre laterally and white below. Mammary formula 2(2-2). Skull
large with a broad M\ supra orbital ridges well developed. Upper incisors mostly with faint
grooves. Some specimens have a small Sm in My. The skull measurements of males and females
show only small differences. The skull measurements, PAL, UPDE, M1, BUL and BRCA are
highly significantly larger (= 1% significance level) than, DIN and ROB are significantly larger
(= 5% significance level) than, and PAF is highly significantly smaller than for the other
members of the East and South African Lemniscomys griselda species-complex (see discriminant
analysis). The hindfoot is significantly longer in males than in females.

55

Measurements in mm of adult Lemniscomys roseveari. Number of specimens measured, mean, range and standard deviation are given from left to right for each of the
measurements. For the exact description of the skull-measurements see Van der Straeten & Van der Straeten-Harrie (1977).

The external measurements are those shown on the museum labels.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13, PT 5, JUNE 1980

56

VAN DER STRAETEN: NEW SPECIES OF LEMNISCOMYS FROM ZAMBIA

Specimens examined

Holotype: adult 9, age-class 3, skin and skull, RG. No. 6396, from Balovale, Zambia, 1 015
metres, obtained 27 September 1938 by G. C. Shortridge and T. D. Carter during the
Vemay-Kaffrarian Museum Zambesi Expedition. Type specimen in the collections of the
Kaffrarian Museum, King William’s Town, Republic of South Africa.

Paratypes: 8 specimens. Kaffrarian Museum, RG. No. 6353, 9, age-class 6, 19/8/1938; RG.
No. 6355, (J, age-class 2, 20/8/1938; RG. No. 6356, 6, age-class 3, 21/8/1938; RG. No. 6360, 6,
age-class 3, 22/8/1938; RG. No. 6373, 9, age-class 5, 4/9/1938; RG. No. 6387, 6, age-class 3,
17/9/1938; R.G. No. 6391, 6, age-class 4, 25/9/1938; British Museum Natural History, RG. No.
49.354, 6, age-class 5, 1/9/1938.

All specimens from Balovale, Zambia, with skin and skull, collected by G. C. Shortridge and
T. D. Carter.

For details concerning age-classes see Van der Straeten (1979).

Other material examined: 28 specimens. Kaffrarian Museum, RG. Nos.: 6352, 6369, 6377,
6378, 6381, 6385, 6386, 6389, 6394, 6397, 6402, 6404, 6415, 6416, 6418, 6420 (all 66, all from
Balovale); 6358, 6361, 6362, 6370, 6374, 6383, 6392, 6395, 6412, 6417 (all 99, all from
Balovale); British Museum Natural History, RG. No. 1937.1.4.108, 9, from Balovale; RG. No.
55.1164, 9, from Solwezi.

DISTRIBUTION

Lemniscomys roseveari sp. nov. at the present time is known only from two localities in
Zambia: Balovale (13° 33’S, 23°07’E) and Solwezi (12° 10’S, 26° 24’E).

STATISTICAL ANALYSES

Canonical analysis.

For the canonical analysis 179 specimens from East and South Africa were used. These
were divided into eight groups as indicated in Table 2. The specific names used in Table 2 for the
different forms of the Lemniscomys griselda species-group are based upon a biometrical revision
of this species group.

Table 2.

Grouping of 179 specimens of Lemniscomys from East and South Africa for purposes of

canonical analysis.

Group

No.

Species

Number of
specimens

1

Lemniscomys roseveari

28

2

Lemniscomys rosalia (cf. spinalis-zuluensis-sabiensis ) 66

34

3

Lemniscomys rosalia (cf. spinalis-zuluensis-sabiensis ) 99

45

4

Lemniscomys rosalia (cf. calidior)

23

5

Lemniscomys griselda

18

6

Lemniscomys rosalia (cf. sabulata )

14

7

Lemniscomys rosalia (cf. rosalia )

8

8

Lemniscomys rosalia (cf. macculosus-phaeotis)

9

57

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13, PT 5, JUNE 1980

The analysis is based upon 18 skull measurements (see Table 3). There are seven canonical
variates, only five of which are significantly different from zero. In Table 4 are given the
eigenvalues of the canonical transformation and the importance of the different canonical
variates. In Table 3 are given the eigenvectors of the 18 variables for the first three canonical
variates. Using these eigenvectors a diagram was drawn. For each group the centre and the most
extreme values of each cluster of points are indicated. Fig. 1 shows the graphical representation
of the first and second canonical variates. The other canonical variates are not illustrated as they
give no further information concerning the problem.

Table 3.

Eigenvectors of 18 variables for the first three canonical variates.

Variable

code

1

2

3

GRLE

- 0,8632

0,5884

- 0,9461

PRCO

2,5536

- 1,9409

1,7076

HEBA

- 2,3949

1,7373

- 1,8240

HEPA

- 0,2616

- 0,0931

- 0,0034

PAF

- 2,3517

0,3968

0,8509

DIA1

- 1,6253

- 0,3303

2,6270

DIA2

2,9291

- 0,5666

- 0,6978

INT

0,9013

- 1,6658

0,2164

ZYG

0,0128

1,1459

1,7343

UPTE

- 1,6967

0,0369

1,1756

UPDE

0,7826

- 0,0420

- 2,9892

Ml

4,6652

10,2576

3,9417

ZYPL

0,5395

- 0,3966

- 0,0048

BNAS

1,1364

0,2550

- 0,9504

LNAS

0,3861

- 0,1074

0,4877

LOTE

- 0,1950

- 0,0873

- 0,2200

BUL

0,3856

0,3351

- 0,1951

BRCA

0,5189

0,0111

- 0,7968

Table 4.

Eigenvalues of the canonical transformation with test of significance.

No.

Eigenvalue

Relative (%)
importance

Chi-square

Degrees of
freedom

Probability

1

47,455

41,5

518,568

126

1,000

2

27,952

24,5

339,170

102

1,000

3

19,147

16,8

212,575

80

1,000

4

11,091

9,7

116,624

60

0,999

5

4,039

3,5

54,643

42

0,910

6

3,203

2,8

29,348

26

0,705

7

1,366

1,2

8,987

12

0,295

58

VAN DER STRAETEN: NEW SPECIES OF LEMNISCOMYS FROM ZAMBIA

©

Fig. I. Canonical analysis: canonical means (solid circles) and extreme limit of each cluster of points; canonical variates:
1 (abscis) and 2 (ordinate); 1, A: Lemniscomys roseveari ; 2, B: L. rosalia (cf. Spinalis ) 6 <3; 3, C: L. rosalia (cf. spinalis )
9?; 4, D: L. rosalia (cf. calidior)-, 5, E; L. griselda', 6, F: L. rosalia (cf. sabulata ); 7, G: L. rosalia (cf. rosalia ); 8, H: L.
rosalia (cf. macculosus); black star: type of L roseveari.

Considering the first canonical variate there are distinguishable three groups within the East
and South African Lemniscomys griselda species-complex: 1. L. griselda, 2. L. rosalia (with
different subspecies separated by the second and third canonical variates) and 3. L. roseveari.
The overlap between the three groups is very small. Each of these groups is considered here as
corresponding with a valid species.

Discriminant analysis and dendrogram.

The same groups and specimens were used as for the canonical analysis with the addition,
however, of a group of 113 specimens of Lemniscomys linulus (see Van der Straeten, 1979). Out
of the 18 measurements used for the canonical analysis 11 were selected for the present analysis:
GRLE, PRCO, HEBA, PAF, DIA2, INT, ZYG, UPTE, UPDE, M1 and BUL. All groups were
combined in a two by two groups discriminant analysis in order to determine the generalized
distances of Mahalanobis (D^) and the probability of misclassification. The results are sum-
marized in Table 5.

To illustrate the biometrical affinities a dendrogram was drawn up (Fig. 2) based on the
matrix of values and generated by the U.P.G.M.A. clustering method (see Sneath & Sokal,
1973).

59

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13, PT 5, JUNE 1980

The results of this dendrogram and those of the canonical analysis are almost the same. The
West African Lemniscomys linulus is quite different from the East and South African species of
the Lemniscomys griselda species-group. There is a very close relationship between Lemnis-
comys rosalia (cf. spinalis ) (=spin in Fig. 2), L. rosalia (cf. subulata) (= sab), L. rosalia (cf.
calidior) (= cal), L. rosalia (cf. rosalia ) (= rosa) and L. rosalia (cf. macculosus) (= phae)
which are considered to belong to the same species namely Lemniscomys rosalia . Further study
of more material is needed, however. The biometrical distance between Lemniscomys rosalia, L.
griselda and L. roseveari is important enough to consider these three forms as valid species.

Table 5.

Matrix of generalized distances of Mahalanobis (= D^) (lower triangle) and probability of misclassification (%) using a
1 1 variate discriminant function (upper triangle) (abbreviations see Fig. 2).

Species

No. Specimens

lin

113

ros

28

spin

79

cal

23

gris

18

sab

14

rosa

8

phae

9

lin

—

0,00

0,01

0,19

0,00

0,00

0,08

0,00

ros

90,0530

—

6,08

2,38

0,21

4,78

0,40

0,17

spin

53,7478

9,5844

—

18,42

5,13

30,25

2,71

13,82

cal

33,4167

15,6836

3,2364

—

1,25

8,37

14,74

5,97

gris

85,6822

32,6681

10.6644

20,0785

—

2,68

0,19

0,37

sab

85,5641

11,1154

1,0702

7,6221

14,9091

—

3,43

1,30

rosa

40,1762

28,2083

14,8163

4,3912

33,3432

13,2701

—

0,89

phae

56,6743

34,1637

4,7392

9,6963

28,6237

19,8033

22,4351

—

In order to eliminate errors as much as possible, two discriminant functions using skull
measurements were elaborated. The first discriminant function makes it possible to differentiate
between Lemniscomys roseveari and the species of the L. griselda species-group from East and
South Africa.

K = - 3,454 x PAF + 3,543 x DIA 2 - 5,417 x UPTE + 5,950 x UPDE + 20,819 x M1 -
26,96

If K>0 then the specimen is a Lemniscomys roseveari ; if K<0 it is a species of the L. griselda ■<
species-group. The chance of an erroneous determination is 6,8%. In the material examined two (>
Lemniscomys roseveari and eight L. rosalia (from the Transvaal and Namibia) were misidentified ij
with this function. If the eighteen available factors are used for a discriminant function, this
percentage drops to 4,9.

The second discriminant function, using three factors, makes it possible to differentiate
Lemniscomys roseveari from the adjacent L. griselda. It gives a 1,9% chance of an erroneous
determination.

K = 14,074 x PRCO - 14,581 x HEBA - 10,880 x PAF + 21,85

Specimens of Lemniscomys griselda give negative values for K whereas specimens of L.
roseveari give positive ones.

VAN DER STRAETEN: NEW SPECIES OF LEMNISCOMYS FROM ZAMBIA

SPIN SAB CAL ROSA PHAE GRIS

ROS

LIN

Fig. 2. Dendrogram, based on a matrix of generalized distances of Mahalanobis, generated by the U.P.G.M.A. clustering
methods; spin: Lemniscomys rosalia (cf. spinalis), sab: L. rosalia (cf. sabidata), cal: L. rosalia (cf. calidior ), rosa: L.
rosalia (cf. rosalia), phae: L. rosalia (cf. macculosus ), gris: L. griselda, ros: L. roseveari , lin: L. Hindus.

61

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13, PT 5, JUNE 1980

ACKNOWLEDGEMENTS

The author wishes to thank his colleague Mr. P. Swanepoel for allowing him to study the
material in the collections of the Kaffrarian Museum and for the loan of some specimens.

REFERENCES

Sneath, P. H. A. and Sokal, R. R. 1973. Numerical taxonomy. San Francisco. Freeman & Co.

Van der Straeten, E. 1979. Etude biometrique de Lemniscomys linulus (Afrique occidentale, Mammalia, Muridae). In
press.

Van der Straeten, E. and Van der Straeten-Harrie, B. 1977. Etude de la biometrie cranienne et de la repartition
d'Apodemus sylvaticus (Linnaeus, 1758) et d 'Apodemus flavicollis (Melchior, 1834) en Belgique. ActaZool. &
Pathol. Antverp. 69: 169-182.

Van der Straeten, E. and Verheyen, W. N. 1978. Taxonomical notes on the West-African Myomys with the
description of Myomys derooi (Mammalia, Muridae). Z. Saugetierk. 43: 31-41.

Manuscript accepted for publication 20 February 1980.

62

I

frto/3

n 14 •

ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

Ann. Cape Prov. Mas. (nat. Hist.)

VOLUME 13 • PART 6
24th JUNE 1980

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

Printed by Cape & Transvaal Printers (Pty) Ltd, Cape Town

BD8908

Ethological studies of Jugurtia confusa Richards, Ceramius capicola
Brauns, C. linearis Klug and C. lichtensteinii (Klug) (Hymenoptera:
Masaridae) in the Eastern Cape Province of South Africa

by

F. W. GESS and S. K. GESS
(Albany Museum, Grahamstown)

CONTENTS

Page

Abstract 63

Introduction 63

Description of the nesting sites 64

Flight seasons, daily flight period, location of mating and of sheltering 64

Plants visited by adult wasps and composition of provision 65

Description of the nest 66

Method of construction of the nest, oviposition and provisioning 67

Life history 74

Parasites and other associated insects 76

Discussion 78

Acknowledgements 82

References 82

ABSTRACT

Some aspects of the ethology of Jugurtia confusa Richards, Ceramius capicola Brauns, C.
linearis Klug and C. lichtensteinii (Klug) (Hymenoptera: Masaridae) in the Eastern Cape
Province of South Africa are described. The account of the ethology of J. confusa is the first
published for Jugurtia. Those of the three Ceramius species clarify some uncertainties and
misconceptions in the literature and add to the overall picture of the ethology of the genus.

INTRODUCTION

The two Old World genera Ceramius Latreille and Jugurtia Saussure each encompass about
twenty-five described species the greatest concentration of which is found in southern Africa,
particularly in the Cape Province. The remaining species are scattered over Africa north of the
Sahara, southern Europe, Turkey, Armenia, and in the case of Jugurtia North-West Africa
(Northern Nigeria, Upper Volta, Gambia and Senegal), Yemen and Persia.

Nothing has hitherto been published concerning the ethology of any Jugurtia species. With
respect to Ceramius species a general outline of the ethology may be assembled from the
publications of Fonscolombe (1835), Giraud (1871), Ferton (1901), Brauns (1910) and Gess
(1965, 1968 and 1973). However, despite the number of publications dealing with the nesting of
Ceramius species certain aspects have never been described. On the other hand incorrect
interpretation of observed facts has led to the presence in the literature of uncertainties or
misconceptions, chiefly connected with the form of cell provisioning practised.

63

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13, PT 6, JUNE 1980

The present paper is based upon field work carried out over a period of years at Hilton, a j
farm situated 18 kilometres WNW. of Grahamstown (33° 19'S., 26° 32'E.) in the Albany
Division of the Eastern Cape Province of South Africa and as such is the seventh of a series of
publications dealing with the ethology of certain solitary wasps occurring at that locality. The
account is augmented as indicated in the text by a few observations made at Clifton, a farm
situated 7 kilometres NNE. of Hilton. An account of the climate and vegetation of Hilton and
Clifton has previously been given (Gess and Gess, 1974: 191-192).

DESCRIPTION OF THE NESTING SITES

The sites chosen for nesting by Jugurtia confusa Richards, Ceramius capicola Brauns, C.
linearis Klug and C. lichtensteinii (Klug) on Hilton lie within the area utilized by Dichragenia
pulchricoma (Arnold), Bembecinus cinguliger (Smith) and Parachilus insignis (Saussure) and
described for these species (Gess and Gess, 1974, 1975 and 1976) as localized patches of bare
clay earth occurring in low-lying areas, sparsely covered by small low-growing shrubs, largely
Pentzia incana (Compositae), and situated close to temporary sources of water. Pentzia incana
only is mentioned as being the dominant plant. For the purpose of this study it must, however,
also be mentioned that the second most dominant plant is Chrysocoma tenuifolia (Compositae)
and that there is in addition a lesser scattering of low-growing Mesembryanthemaceae including
Drosanthemum parvifolium, Drosanthemum floribundum and Ruschia species. The nesting of the i
four species of masarids appears to be very localized as, although an extensive search was made,
they were found to make use of only a few of the available apparently similar bare areas and to
utilize the same sites each year. The chosen sites lie on both sides of a water furrow mentioned in
Gess and Gess (1974, 1975 and 1976) and on a flat area between the furrow and a small shallow
earthen dam situated below the former. Water is provided in shallow rainwater pools in the !
furrow in which water only flows for short periods after rain.

J. confusa and C. capicola nest in pseudo-colonies on level bare patches caused by sheet
erosion or by the presence of a road or path and it seems probable from the situation of the single
nest of C. linearis discovered that this species selects similar areas. J. confusa and C. capicola i
have not been found nesting in close proximity either to each other or to any of the other :j
ground-nesting wasps with the exception of Bembecinus cinguliger although this was more in the r
nature of a fringe overlap in time and space. A single nest of J. confusa was found also in a
pocket of clayey soil on a horizontal ledge of a raised river bank at some distance from the main i1
nesting area. A second species of Jugurtia, J. braunsiella (von Schulthess) has occasionally been
collected at Hilton but its nests have not as yet been located. C. lichtensteinii differs from the ij
other three species of masarids under consideration in that it favours raised bare ground for
nesting purposes, the nests located having been either on raised earth around a shrublet in an i
eroded area or else on artificially constructed banks of a furrow and of small drainage channels.
Similarly, on the farm Clifton, C. lichtensteinii nests on raised banks of bare clayey soil on the
sides of a donga (erosion gulley). At Clifton where C. lichtensteinii is more common than at d
Hilton, it has been found to nest in populous pseudo-colonies, thirty nests having been counted
within an area of about a square metre.

FLIGHT SEASONS, DAILY FLIGHT PERIOD, LOCATION OF MATING
AND OF SHELTERING

i

The flight periods for the masarids under consideration, like those of all clay-nesting wasps,
are variable according to the climatic conditions prevailing in a particular year. Certain
generalisations may, however, be made. The species sighted earliest are J. confusa and J.
braunsiella which have been collected at Hilton as early as late September. The most abundant
nesting observed for J. confusa was in the summer of 1976-77 when the greatest activity was

64

GESS AND GESS: ETHOLOGICAL STUDIES OF JUGURTIA CONFUSA RICHARDS, ETC. (HYMENOPTERA)

from early November to mid-December after which there was no activity till mid-February to
early April when there was a second but insignificant flush. In other years when the rain has been
late there has been a shift in activity to January-February but this was less successful than in the
years when nesting was early in the summer.

C. capicola and C. lichtensteinii are normally reaching their peak after the high-point of
activity of J. confusa is over. Again the intensity of activity and the extent in time is governed by
the climatic conditions in any particular year, however, early December to early January seems to
be the most favourable time for these two species.

A nesting date for C. linearis is only known from one nest found on 10. xii. 1974, however,
from collecting dates for this species in the area it seems to be flying most commonly in
November-December and therefore its greatest activity is probably somewhere between that of J.
confusa on the one hand and C. capicola and C. lichtensteinii on the other.

The males appear in numbers shortly before the females and are present for most of the
flight period, becoming scarce as the season advances. The hottest part of the day shows the
greatest activity and this is the time favoured for nesting activities. J. confusa males were
observed to fly low, 5-8 cm above the ground, particularly skirting bushes at the periphery of the
nesting site and also alighting on the ground within the nesting area where they sun themselves
and rise up to chase the females and each other. They are seen to descend rapidly upon the
females and although mating was not observed it seems likely that it takes place within the
vicinity of the nesting area. The females were always seen alone at the water where they stand on
the mud at the edge of the water to fill their crops. Two females of J. braunsiella were collected
whilst similarly engaged. In the three species of Ceramius on the other hand both males and
females were observed together at the water and the females filled their crops whilst standing on
the surface of the water. In C. capicola, the only species present in large numbers at Hilton, both
sexes were frequently observed flying up and down the length of a puddle. They flew over the
water 5-8 cm above the surface. The females frequently alighted on the water surface, the legs
being held wide-spread. While thus resting on the water surface and gently drifting, the males
were seen to alight on top of the females, when both would take off and fly away together, the
male above the female grasping her. This was presumably followed by mating. The behaviour of
C. linearis and C. lichtensteinii at the water and preparatory to mating was observed at Hilton
and more particularly at Clifton to be identical to that of C. capicola as described above.

When the sun was obscured by cloud or when a breeze got up females of J. confusa,
C. capicola and C. lichtensteinii were found to be sheltering in their nests. Those of J. confusa
were seen backing down their burrows into the sheltering position, head uppermost about 5 mm
below the opening of the shaft.

PLANTS VISITED BY ADULT WASPS AND COMPOSITION OF PROVISION

Records of Jugurtia species visiting flowers are few in number: at Hilton J. confusa was
found on Drosanthemum parvifolium (Mesembryanthemaceae) (1 male, 8. xii. 1976) and on
Acacia karroo (Leguminosae) (1 male, lO.ii. 1977) and J. braunsiella was found on Lasiosper-
mum bipinnatum (Compositae) (1 male, 12.x. 1977); at Clifton J. braunsiella was found on
Pteronia paniculum (Compositae) (1 female, 27.x. 1972). A female J. confusa caught in flight
was found to be carrying an abundance of pollen on the fore-tibiae and tarsi which are covered
with dense hairs, on the swollen antennae, and on the labrum and front of the clypeus.

I Flower-visiting records pertaining to Ceramius species at Hilton are likewise few in
number: C. linearis on Drosanthemum floribundum (Mesembryanthemaceae) (1 male,
29.xi. 1976) and C. lichtensteinii on a ‘purple mesem’ (Mesembryanthemaceae) (1 male,
26.x. 1977) and on Senecio pterophorus (Compositae) (1 female and 2 males, 29.xi. 1979; 1
female, 1. xii. 1979; 2 males, 2. xii. 1979). However, both males and females of C. capicola, C.

65

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13, PT 6, JUNE 1980

linearis and C. lichtensteinii have previously been recorded in the Grahamstown district on
various species of Mesembryanthemaceae (Gess, 1973: 115-116).

The provision supplied by J. confusa, C. capicola, C. linearis and C. lichtensteinii for the
nourishment of their young is pollen bound together with nectar to form a loaf which partially
fills the cell. This pollen bread varies in consistancy from dry and firm in C. linearis and C.
lichtensteinii to moist and sticky in J. confusa and C. capicola.

A sample of pollen loaves was examined microscopically for each species, the pollen of the
provision being compared with that of flowers found in the vicinity of the nesting area, including
species of Mesembryanthemaceae and Compositae. In all cases it was found that the pollen used
was that of Mesembryanthemaceae. With respect to the species of Mesembryanthemaceae present
in the vicinity of the nesting areas an attempt was made to match the pollen by size with that
used in the formation of the pollen loaves. In the case of J. confusa five pollen loaves were
examined and one pollen size was found to be common to all and was matched with that of
Drosanthemum parvifolium. However, some loaves contained in addition one or two other pollen
sizes but these were not successfully matched. C. capicola loaves contained only one size of
pollen which was matched with that of Drosanthemum floribundum. The C. linearis loaf on the
other hand contained pollen of three sizes one of which matched with that of Drosanthemum
floribundum. C. lichtensteinii was found to be using pollen of the same size as that of several I
species of Ruschia.

DESCRIPTION OF THE NEST

The nests of J. confusa, C. capicola, C. linearis and C. lichtensteinii all consist of a
subterranean burrow surmounted by a cylindrical mud turret constructed from mud pellets
cemented together; in J. confusa loosely so that in some cases the turret has a lacy appearance; in
C. capicola, C. linearis and C. lichtensteinii closely and smoothed on the inside so that open
interstices are extremely rare except near the free end in turrets of C. lichtensteinii . In all four i
species the turret is initially vertical to sub-vertical but, if of any considerable length, curves over
becoming horizontal as in C. lichtensteinii or curves downwards and then, sometimes, continues
in a horizontal plane close to the ground but always free from it as in C. capicola. In J. confusa
more than a short vertical turret is extremely rare and may be considered aberrant.

The subterranean burrow of a newly-constructed nest consists of a vertical shaft which is
either of constant diameter along its length as in J. confusa or has in the upper third a short
bulbous portion as in C. capicola and C. lichtensteinii and from which at its lower end there t
branches a subhorizontal secondary shaft terminating in an excavated-cell within which is a
constructed mud-cell in J. confusa and C. lichtensteinii but not in C. capicola and C. linearis.

In nests at a more advanced stage of construction further secondary shafts each terminating i
in a cell are present. In J. confusa and C. lichtensteinii a secondary shaft including its cell is
barely longer than the cell itself whereas in C. capicola and C. linearis it is considerably longer
than the cell. C. linearis differs from the other three species in that, in the single nest excavated,
the cells all lay at a similar depth below the lower end of the main shaft whereas of those in the
other three species only the deepest cell lies below the lower end of the main shaft, the other cells li
being at varying depths above it. In C. lichtensteinii the cells are arranged in loose whorls, one or
two whorls being constructed in a single year. Additional whorls are constructed in succeeding
years. These nests are therefore perennial in nature as probably are those of J. confusa.

All completed cells are sealed with a mud-plug which, in constructed mud-cells, takes the
form of a stopper fitting into the neck of a cell like a cork into a bottle. The section of a
secondary shaft between the sealed cell and the main shaft is filled with tightly packed earth and
its opening to the main shaft sealed with a mud-plug.

66

GESS AND GESS: ETHOLOGICAL STUDIES OF JUGURTIA CONFUSA RICHARDS, ETC. (HYMENOPTERA)

Figs 1-3. Vertical plans of the turrets and underground workings of the nests of three species of Masaridae at Hilton.

Lettering of cells as in key.

Figs la and b. Two nests of Jugurtia confusa Richards, investigated on 8.xi. 1976 and 27. ix. 1978 respectively.

Figs 2a, b and c. Three nests of Ceramius capicola Brauns, investigated on 7.xii. 1976, 9.xii.l976 and 1 1 . i. 1 977

respectively.

Fig. 3. Single nest of Ceramius linearis Klug, investigated on lO.xii. 1974.

METHOD OF CONSTRUCTION OF THE NEST, OVIPOS1TION AND PROVISIONING

The four species studied differ considerably in their size and as body size has a profound
effect upon the dimensions of the nest built by each species, total body length and greatest body
width (measured across the mesothorax) of average females of each species are given here. J.

• confusa is the smallest species, 10 mm long and 2,7 mm wide; the Ceramius species range from
C. capicola, 12 mm long and 2,9 mm wide, through C. linearis, 15 mm long and 3,4 mm wide
to C. lichtensteinii , 20 mm long and 5,6 mm wide. Males are generally slightly smaller than the
females.

67

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13, PT 6, JUNE 1980

Figs 4a, b and c. Vertical plans of the turrets and underground workings of three nests of Ceramius lichtensteinii (Klug)
at Hilton, investigated on 29.xii. 1976, 31.xii. 1973 and 3. i. 1977 respectively. Lettering of cells as in key.

GESS AND GESS: ETHOLOGICAL STUDIES OF JUGURTIA CONFUSA RICHARDS, ETC. (HYMENOPTERA)

Figs 5a and b. Vertical plans of the turrets and underground workings of two nests of Ceramius lichtensteinii (Klug) at
Hilton, both investigated on 3. i. 1977. Lettering of cells as in key.

69

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13, PT 6, JUNE 1980

Water is required for nest construction and is collected by a female from a nearby puddle.
Whilst filling her crop with water J. confusa stands on the mud at the edge of the water whereas
the three species of Ceramius alight on the water surface.

Nest excavation is initiated by the female’s regurgitating water from her crop onto the
ground. Using her mandibles she works this water into the earth to form mud from which she
forms a pellet. A number of pellets are formed in this way from each crop-full of water. The first
pellets excavated from the shaft-initial may be discarded. The shaft-initial is circular in cross
section due to the female’s rotating evenly, not altering the direction of rotation without first
completing a circle. At the commencement of turret construction, the pellets, instead of being
discarded, are laid down in a circle around the shaft-initial in such a way that the inner diameter
of the turret will be the same as that of the shaft; in J. confusa 3 mm (average of sample of 13
measurements), in C. capicola 3,2 mm (average of sample of 10 measurements), in C. linearis 1
3,5 mm (one measurement) and in C. lichtensteinii 6,2 mm (average of sample of 7 measure-
ments). In the construction of a vertical cylindrical turret pellets are added regularly whereas in a
sloping or curved turret more pellets are added to what will be the upper side than to what will be
the lower side. It is noticeable that in turrets of C. capicola the upper edge of the turret opening
projects further than the lower edge.

The method of placement of pellets by C. capicola was observed most clearly. The wasp
backs up the shaft with a pellet held in her mandibles and reaching the turret opening holds the
sides of the turret with her legs whilst placing the pellet in position and smoothing it on the inner
surface with her mouthparts and supporting it on the outer surface with the tip of the ventral
surface of her abdomen which is curved around for this purpose (Fig. 8). As many as twelve
pellets may be added to the turret per water load. If the turret is destroyed by rain or mechanical
means, the wasp will build a new one of similar design and dimensions to the original one. In C.
lichtensteinii and probably also in J. confusa , if a newly emerged female instead of initiating a
new nest expands a maternal nest, a turret is still constructed at the commencement of nesting. In
this case the mud required is obtained from the bottom of the main shaft.

Whilst turret construction is in progress the shaft increases in depth. The diameter of the
shaft of J. confusa is maintained constant whereas the diameters of the shafts of C. capicola and lj
C. lichtensteinii after the shafts have reached average depths of 35 mm (sample of 7) and 85 mm
(sample of 5) respectively increase to 5,4 mm (average of 7 measurements) and 11,5 mm
(average of 5 measurements) after which they decrease to their original measurements. The
resultant bulbous portions of the shafts are 10 mm and 24-30 mm long. As the only nest of C.
linearis excavated was in difficult ground it is not known whether or not this species constructs a
‘bulb’.

After completion of the turret the wasp continues to excavate the shaft but the pellets then
extracted are discarded. J. confusa has no clearly defined pellet-dropping area, however, the
wasp does confine her arrivals and departures from the nest to a set quarter segment. Like J.
confusa, C. lichtensteinii has no set pellet-dropping area but discards pellets in bushes at the edge i1
of the clearing in which her nest is situated with the result that the distance from the nest is
variable; 60-90 cm in one instance and about 300 cm in another. However, C. capicola has a
clearly defined pellet-dropping area a few centimetres from and to one side of the turret. When i
discarding a pellet, a female C. capicola backs out from her nest until her head is free from her
turret, flies sideways and slightly forwards just above the surface of the ground to the 1 1
pellet-dropping area, drops the pellet and still orientated parallel to the turret flies in reverse
sideways motion back to the nest entrance which she is then facing and enters (Fig. 6). In this
way the pellet-dropping operation takes up the minimum of time and exertion and differs from
that of most mud-excavating wasps including J. confusa and C. lichtensteinii which fly up in a
wide circle when dropping pellets.

At this stage in nest construction C. capicola and C. lichtensteinii females when leaving the

GESS AND GESS: ETHOLOGICAL STUDIES OF JUGURTIA CONFUSA RICHARDS, ETC. (HYMENOPTERA)

O

o

o

Backs out of turret onto
ground in front of turret
opening.

O o

o

Flies up, moves to pellet
dropping area, drops
pellet.

U

°°e00o

o o
o o

Moves back to in front
of turret opening,
alights.

Walks f or w a rd s, ste ps up
into turret and disappears
down its length.

Fig. 6. Sequence showing, in diagrammatic form, the method of pellet deposition after turret construction by Ceramius

capicola Brauns.

71

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13, PT 6, JUNE 1980

nest to fetch water emerge from the turret head first whereas those of J. confusa still continue to
emerge backwards. It is probable that the C. capicola and C. lichtensteinii females turn around
in the newly constructed bulbous portion of the nest shaft.

Cycles of water carriage and pellet extraction were timed for J. confusa, C. capicola and C.
lichtensteinii and were found to be performed rapidly and without interruption during active nest
excavation. Water carriage took 1-2 minutes depending upon the distance from the nest to the
water source and pellet extraction took 2-4 minutes depending on the dryness and hardness of the
soil. Similarly, the number of pellets extracted per water load varied according to the dryness of
the soil. Thus at the outset of nest excavation after rain when the surface layers of the soil only
were damp wasps were observed to carry out a far larger number of pellets and more rapidly than
they did once they had reached a lower level in the soil where the rain had not penetrated.

The shaft is sunk vertically to depths of 37-52 mm (average 43 mm, sample of 18) in J.
confusa, 75-100 mm (average 89 mm, sample of 10) in C. capicola, 135-170 mm (average 150
mm, sample of 7) in C. lichtensteinii and 90 mm in C. linearis (single nest). From the bottom of
the shaft a secondary shaft of the same diameter as the main shaft is constructed in a
subhorizontal plane so that the distal end lies deeper than the bottom of the main shaft and is
enlarged to form a cell. The average lengths of secondary shafts and of cells of the four species
are: J. confusa 17 mm (sample of 7) and 15,6 mm (sample of 5); C. capicola 24 mm (sample of
10, range 10-45 mm) and 11 mm (sample of 10, range 10—1 2 mm); C. lichtensteinii 33 mm and
29 mm (sample of 10) and C. linearis 30 mm and approximately 12 mm (sample of one). The
average diameters of cells are: J. confusa 5 mm (sample of 5); C. capicola 6 mm (sample of 10);
C. lichtensteinii 1 1 mm (sample of 7); and C. linearis approximately 6 mm.

After cell excavation has been completed the cells of C. capicola and C. linearis are ready
for oviposition whereas in J. confusa and C. lichtensteinii a mud-cell is first constructed within
each excavated-cell. Mud for the construction of these cells must be quarried within the nest as
these wasps do not fetch mud from elsewhere. In nests of J. confusa in which a mud-cell has
been constructed there is an enlarged ‘heel’ at the bottom of the shaft. It is thought probable that
at least part of the mud used in constructing the mud-cell is excavated from this source and that
the mud used by C. lichtensteinii is similarly obtained by a deepening of the lower end of the
main shaft. The mud-cells are easily separable from the walls of the excavated-cells. They have a
rough outer surface on which the separate applications of mud are discemable. The inner surface,
however, is carefully smoothed. The average thickness of the walls is 0,7 mm in J. confusa and
0,9 mm in C. lichtensteinii.

Oviposition takes place before the commencement of provisioning. The eggs are strongly
curved, white in J. confusa and C. capicola, whitish- yellow to pale yellow in C. lichtensteinii,
and are of average dimensions 2,51 mm by 0,76 mm (sample of one), 3,55 mm by 0,89 mm
(sample of 8) and 6,24 mm by 1,42 mm (sample of 7) respectively.

The cell is then rapidly provisioned and sealed before the egg hatches. Exceptionally, open
cells of C. lichtensteinii were found containing a young larva and little or no pollen bread. Thus,
at Hilton during the period 12-3 l.xii. 1973, of nine open cells in which oviposition had taken
place seven contained eggs only and no provision, one contained a newly hatched larva (5 mm in
length) and no provision, and one contained a small larva (6 mm in length) and a small amount
of provision. However, of ten newly sealed cells, one contained an as yet unhatched egg and a
complete pollen loaf and nine contained larvae of various sizes ranging from newly hatched (5,5
and 6 mm in length) to large with varying amounts of as yet unconsumed provision in inverse
proportion to larval size. It therefore appears that under favourable conditions the cell is fully
provisioned and sealed before the egg hatches but that, exceptionally, under unfavourable
conditions such as bad weather, when ‘mesem’ flowers stay closed and wasps do not readily fly,
or such as a scarcity of forage flowers the beginning of provisioning may be preceded by the
hatching of the egg.

72

GESS AND GESS: ETHOLOGICAL STUDIES OF JUGURTIA CONFUSA RICHARDS, ETC. (HYMENOPTERA)

Fig. 7. Hilton, lO.xi. 1976. Sequence showing shaft sinking and turret building by Jugurtia confitsa Richards. (x2,4)

73

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13, PT 6, JUNE 1980

After a cell has been sealed the remaining section of the secondary shaft is filled with earth
and sealed off from the vertical shaft with mud which is smoothed so that the entrance to a
secondary shaft, once it is sealed, is no longer visible on the surface of the main shaft. C.
capicola and C. linearis do not introduce earth into the nest from elsewhere and so the earth for
filling a secondary shaft must be obtained within the nest and can only come from the excavation
of another secondary shaft. Succeeding cells are constructed in the same manner, the number
probably being dependent on the availability of water for nest construction and pollen and nectar
for cell provisioning.

Twenty newly-constructed nests of J. confusa were excavated in the summer of 1976-77
and of these eleven contained one cell each (Fig. la) and nine were still at the stage of shaft
excavation. Due to the drying out of the water supply nesting in the pseudo-colony to which
these nests pertained ceased at this stage. However, the single nest found on the river bank and
excavated in September, 1978, a year after it had first been noted contained eleven cells (Fig.
lb). This nest had a shaft 81 mm in depth with the cells radiating out from the shaft at depths of
from 35-79 mm. Eight cells were open, empty and parchment-lined and were therefore cells
from which the occupants had already emerged. Three cells were sealed, one containing an adult
female ready to emerge and the other two each a prepupa. In addition one adult female was
coming into and out of the nest and one male was sunning itself on the ground next to the nest. It
seems highly probable that this nest may have been a perennial one.

Sixteen nests of C. capicola were excavated in the summer of 1976-77 and of these seven
were one-celled (Fig. 2a), three two-celled (Fig. 2b), one four-celled (Fig. 2c) and the remaining
five nests had not reached the stage of cell excavation.

The single nest of C. linearis excavated in the summer of 1974 contained four cells (Fig. 3).

Nineteen nests of C. lichtensteinii were excavated in the summer of 1973-74 and of these
six were newly constructed (Fig. 4b) and thirteen had been initiated one or two years previously.
A further five nests were excavated in the summer of 1976-77. Of these one nest was newly
constructed (Fig. 4a) and the remaining four nests had been initiated one or two years previously.
These four nests contained eleven (Fig. 4c), fifteen, twenty (Fig. 5a) and twenty-one (Fig. 5b)
cells arranged in whorls of three to seven or more cells. As there seems to have been a low
percentage of emergence of C. lichtensteinii at Hilton during the years when observations were
made the infrequency of occurrence of newly constructed nests is not surprising.

Little information was gathered concerning the immature stages of the four masarid species.
Hatching of the larva of C. lichtensteinii appears to take place about three days after oviposition
for of four eggs obtained from unprovisioned open cells excavated on 3 1 . xii. 1973, one hatched
on 2.i. 1974 and three hatched on 3. i. 1974.

After consuming the stored provision the then mature larva spins its cocoon which in C.
lichtensteinii is parchment-like, brittle, brown in colour and in intimate contact with the inner end
and the walls of the mud-cell from which its detachment is difficult. The posterior end of the
cocoon — that is the end towards the inner end of the cell is rounded, whereas the anterior end of
the cocoon — that is the end towards the mud-plug closing the mud-cell is truncate and is in the i{
form of a flat circular plate separated from the mud-plug by an empty space. Cocoons of C.
lichtensteinii measured from 19,0-21,5 mm in length, from 7, 3-9,0 mm in width at their middle
(where widest) and from 6,3-7, 0 mm in width at their truncate end. The empty space between
the truncate end of the cocoon and the inner surface of the mud-plug was from 4-8 mm long.

Having spun the parchment cocoon the larva changes into a prepupa characterised by its
hunch-backed and flaccid appearance. Diapause takes place in the prepupal stage and may last
until the following spring or may be extended over longer periods. Development after pupation in
the spring appears to be rapid.

LIFE HISTORY

74

GESS AND GESS: ETHOLOGICAL STUDIES OF JUGURTIA CONFUSA RICHARDS, ETC. (HYMENOPTERA)

Fig. 8. Hilton, 6. xii. 1976. Sequence showing turret building by Ceramius capicola Brauns. (x2,5)

75

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13, PT 6, JUNE 1980

PARASITES AND OTHER ASSOCIATED INSECTS

Parasites of Masaridae recorded at Hilton consist of species of Chrysididae, Mutillidae and
Meloidae.

The genus Allocoelia (Chrysididae), the species of which as far as is known are all parasitic
in the cells of Masaridae, was found to be represented at Hilton by A. bidens Edney. During the
period 8-19.xi. 1976 seven females of this small species were caught resting on the ground in the
nesting area of J. confusa which without doubt constitutes its host. The appearance of the
chrysidid coincided with the beginning of the nesting period of its host.

Though not yet collected at Hilton there seems to be no reason why other species of
Allocoelia, namely A. latinota Edney and A. capensis Smith should not also occur there as both
species have been collected in nearby vicinities. Thus ,4. latinota has been found associated with
nesting C. capicola at Strowan and at Grahamstown (Cradock Dam), with Ceramius species
(capicola, lichtensteinii and linearis) at Alicedale and with C. linearis at Kenton-on-Sea (Gess,
1973: 118-119). A. capensis was found associated with an extensive pseudo-colony of C.
lichtensteinii on the farm Clifton on 7. i. 1979 (D. W. Gess) and was recorded by Brauns (1910;
446) as parasitic in the cells of this wasp at Willowmore. The relative smallness of the nesting
aggregations of Ceramius species at Hilton compared with the large pseudo-colonies of C.
capicola and C. lichtensteinii located at Strowan and at Clifton respectively is the probable
reason why the Allocoelia species parasitic in the nests of these wasps have not yet been
collected at the former locality.

Three C. lichtensteinii cells that showed clear evidence of parasitism by Mutillidae were
found in a nest excavated on 25. i. 1973. In each instance there was an emergence hole sited in or
near the mud-plug sealing the mud-cell and another in the truncate, disc-like anterior end of the
parchment-like C. lichtensteinii cocoon within. The mutillid emergence hole in the truncate
cocoon end was situated in the centre of the disc and left the edges of the latter attached to the
sides of the cocoon. By contrast, C. lichtensteinii itself when emerging from its cocoon removes
the disc entirely. Within each of the three C. lichtensteinii cocoons was found the smaller, very
tough and leathery cocoon of the parasite, the space between the two cocoons being almost I
completely filled with very extensive, tough, brown silken spinnings. Also found in the space
between the two cocoons was in one case the dry shrivelled remains of the mature larva or i
prepupa of the host and in the other two cases the remains of the host pupae. The mutillid
cocoon, orientated the same way as the host’s cocoon, had at its hind end an extensive deposit of I
meconium. In the same nest was found a live female Dasylabroides caffra (Kohl). The fine state
of the vestiture and the completely unworn mandibles indicated that this individual had just
emerged, doubtless from one of the three parasitized cells.

Subsequently six adults of D. caffra were reared from C. lichtensteinii cells excavated at
Hilton and kept in Petri dishes in the laboratory. Three cells excavated on 5.xii. 1973 yielded a
male (on 17. i. 1974) and two females (on 28. ii. 1974 and 31. i. 1975) and three cells excavated on
27. xii. 1973 yielded three females (on 17.iii. 1974, during January, 1975, and on 3 l.i. 1975
respectively). Examination of the mud-cells and cocoons vacated by the parasites confirmed the
earlier findings. It is probable, however, that D. caffra is not species-specific as regards its host
and that C. lichtensteinii is therefore only one of a range of species parasitized.

Two adult specimens of a species of Meloidae, Ceroctis groendali (Billb.) were found in the
nests of C. lichtensteinii . In addition, six meloid larvae in various stages of development were
found associated with the cells of this wasp. Though none of these larvae was reared through to
the adult stage and proof of the larvae’s identity is thus not available, it is nevertheless believed
that they were conspecific with the adults. On this assumption it may be stated that C. groendali
is a fairly common cleptoparasite or predator in the nests of C. lichtensteinii where in its larval
stage it feeds either upon the stored pollen loaf after first destroying the egg or newly hatched
larva of the wasp or upon the fully grown larva itself. The eight individuals of the meloid found

GESS AND GESS: ETHOLOGICAL STUDIES OF JUGURTIA CONFUSA RICHARDS, ETC. (HYMENOPTERA)

Figs 9a, b and c. Hilton, 3.i. 1977. Nesi turrets of Ceramius lichtensteinii (Klug)
Fig. 9a (x0,4); Figs 9b and c (xl, 85)

77

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13, PT 6, JUNE 1980

in association with C. lichtensteinii nests are dealt with in an order determined by their
developmental stage when found.

A (riungulin larva with darkly pigmented body and with well developed long legs and long
slender cerci was found in a sealed cell on 1 2. xii. 1973, situated on the pollen loaf next to the C.
lichtensteinii egg the hatching of which appeared to be imminent. Kept in a gelatin capsule this
triungulin larva moulted to give rise to the second (caraboid) instar before dying.

A small larva, probably early third (first scarabaeoid) instar was found on 31. xii. 1973 in a
sealed cell containing a pollen loaf and a small, dead C. lichtensteinii larva. On the same date
another sealed cell was found to contain a somewhat larger larva and a pollen loaf considerably
reduced in size. This larva, kept in a gelatin capsule, continued to feed for some time on the
pollen loaf but died as a large fourth (second scarabaeoid) instar larva.

A small larva, probably early third (first scarabaeoid) instar was found on 3. i. 1977 in a
sealed cell containing a large pre-spinning C. lichtensteinii larva and no pollen. The two larvae
were transferred to a gelatin capsule and when next examined on 6. i. 1977 the beetle larva was
found to have grown considerably and to have eaten a hole into the wasp larva which it had
thereby killed. The beetle larva moulted on 7. i. 1977 and again between this date and 12. i. 1977
when it was fully grown, having eaten all of the C. lichtensteinii larva except its skin.
Transferred onto damp sand in a Petri-dish it moulted to the coarctate resting larval stage on the
surface of the sand after a period of restless tunnelling through the sand. At the time of writing
(March, 1980) the larva was still in a state of diapause.

A fully developed fourth (second scarabaeoid) instar larva (circa 10 mm long) was found
associated with a C. lichtensteinii cell on 27. xii. 1973 as was a coarctate resting larva (8 mm
long).

The two adult beetles found in the C. lichtensteinii nests were both dead. The first was
found on 25. i. 1973, the second on 5. xii. 1973.

C. groendali appears to be widespread in its distribution, Peringuey (1909: 218) recording it
from “Cape Colony (Port Elizabeth, Grahamstown, East London), the whole of Orange River
Colony, Natal, and the Transvaal”. As this distribution is far more extensive than that of C.
lichtensteinii (see Richards, 1962: 102; Gess, 1965: 224, 1968: 10 and 1973: 113) it is clear that
C. groendali cannot be restricted to this wasp but must have other hosts as well, possibly j
ground-nesting solitary bees.

It is of interest that Brauns (1910: 446) recorded finding the coarctate larvae or the pupae ij
(“Puppen”) of a meloid in the cells of C. lichtensteinii at Willowmore. Specific identification ij
was impossible as he was unable to rear them through to the adult stage. It seems possible,
however, that the same species, C. groendali, was involved.

It is probable that nests of C. lichtensteinii may on occasion be utilized for nesting by
Megachile (Eutricharaea) stellarum Cockerell. A female of this species was on 3. i. 1977 seen r
entering and afterwards leaving a turreted nest of the wasp and on at least one occasion the
remains of an old megachilid leaf nest were found in a turreted nest. Nesting of M. stellarum has
previously been recorded in abandoned and incomplete burrows of Dichragenia pulchricoma
(Arnold) (Gess and Gess, 1974: 204-206) and the females of this bee have also been observed
leaving old burrows of Parachilus insignis (Saussure) (Gess and Gess, 1976: 98). As M.
stellarum nests in pre-existing cavities in the ground it is likely that it will make use of any
burrows of suitable size regardless of the identity of their excavators.

DISCUSSION

The earliest published accounts of the nesting of Ceramius species concerned two Palaearc-
tic species observed in southern France: C. fonscolombei Latreille at Aix (Fonscolombe, 1835:
426-427) and C. tuberculifer Saussure (cited as C. lusitanicus Klug — see Richards, 1962: 29

78

GESS AND GESS: ETHOLOGICAL STUDIES OF JUGURTIA CONFUSA RICHARDS, ETC. (HYMENOPTERA)

and 112-115) in the area of Hautes-Alpes and in the valley of the Vallouise near Brian^on
(Giraud, 1871: 375-379) and at Montlouis and Cerdagne (Ferton, 1901: 137-139).

More recent published accounts bearing upon the nesting of Ceramius have all been
concerned with southern African species observed in the Cape Province: C. beyeri Brauns, C.
bicolor (Thunberg) (cited as C. karrooensis Brauns), C. capicola Brauns, C. cerceriformis
Saussure (cited as C. schulthessi Brauns), C. linearis Klug (cited as C.fumipennis Brauns) and
C. lichtensteinii (Klug) near Willowmore (Brauns, 1910; 387, 445-446), C. bicolor (Thunberg)
on the banks of the Olifants River between Klawer and Clanwilliam (Gess, 1968: 13), and C.
capicola Brauns and C. lichtensteinii (Klug) on Strowan near Grahamstown (Gess, 1973:
117-119).

Some details at least have therefore been published concerning the nesting of a total of eight
Ceramius species, three of which, C. capicola Brauns, C. lichtensteinii (Klug) and C. linearis
Klug, are the subjects also of the present studies. In order to obtain an overall picture of the
ethology of Ceramius it is the intention in the present discussion to draw together and compare
the various published accounts augmented by the present studies which latter furthermore enable
some uncertainties and misconceptions present in the literature to be examined and corrected. At
the same time the nesting of i. confusa Richards, the first Jugurtia species to be examined with
respect to its ethology, is compared with that of species of Ceramius.

All the species of Ceramius enumerated above as also J. confusa have been recorded as
nesting in the ground, the nests being surmounted by variously shaped mud turrets. The nature of
the ground, when recorded, has been given as clayey and hard, bare or at most covered by sparse
vegetation (Giraud, 1871; Ferton, 1901; Gess, 1968 and 1973; present paper). Nesting on slightly
raised banks has been recorded for C. fonscolombei (Fonscolombe, 1835) and for C.
lichtensteinii (present paper) whereas nesting on level ground has been recorded for C. capicola,
C. linearis and J. confusa (present paper).

The formation of pseudo-colonies appears to be common and has been recorded for C.
tuberculifer (Giraud, 1871; Ferton, 1901), C. bicolor (Brauns, 1910), C. lichtensteinii (Brauns,
1910; present paper), C. capicola (Gess, 1973; present paper) as well as fori, confusa (present
paper).

The nearness of nesting sites to water is either stated or is implied by all authors and all the
species are recorded as visiting this water. In the present paper it has been shown that with
respect to the four species studied the purpose of visiting pools is to fill the crop with water
which when regurgitated upon the clayey nesting substrate makes the latter more easily worked
and thus makes nest construction possible. For C. capicola, C. linearis, C. lichtensteinii and J.
confusa no evidence was ever obtained by the present authors either during the course of the
present study itself or on any of the many occasions on which the species were collected at pools
that the wasps were collecting anything but water. Similarly, Ferton (1901) with respect to the
pool-visiting of C. tuberculifer made it abundantly clear that what the wasp collects is water, not
mud, and stated that when captured on her way from a pool a female disgorged into his net her
stored liquid (“son liquide”).

However, other authors have claimed that some species at least collect not water but mud.
Fonscolombe (1835) stated that C. fonscolombei went to ponds to collect sodden earth (“terre
delayee”) but later in his account appears to have been uncertain for he stated that the turret was
constructed of pellets derived from the excavation of the nest (which would indicate the
collection of water, not mud) or of pellets carried to the nest from without (which would support
his earlier contention).

Similarly, Brauns (1910) stated that whereas C. beyeri, C. lichtensteinii and C. linearis
settle on the water surface at the middle of the pool and collect water, C. cerceriformis, C.
bicolor and C. capicola alight at the edge of the pool and collect mud in little pellets which he
maintained are used by them for the construction of their cells and turrets.

79

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13, PT 6, JUNE 1980

With respect to C. capicola at least, the present study has shown Brauns to have been
mistaken. C. capicola differs from C. lichtensteinii and C. linearis neither with respect to the
part of the pool alighted upon nor with respect to what is collected at the pool: water. It seems
hightly likely that Brauns was mistaken also with respect to the other two alleged mud-collectors,
C. cerceriformis and C. bicolor.

It is of significance to note that alighting at the edge of the pool does not prove that mud,
not water, is being collected there. As noted in the present study, J. confusa stands on the mud at
the edge of the water but fills her crop with water like the three species of Ceramius which alight
on the actual water surface.

The construction of a mud-cell within the excavated-cell has been reported for C.
tuber culifer (Giraud, 1871; Ferton, 1901) and for C. lichtensteinii (Brauns, 1910). In the present
study the construction of such a mud-cell by C. lichtensteinii has been confirmed and has been
established also for J. confusa. However, no mud-cells could be demonstrated for either C.
capicola or C. linearis and these species therefore differ from C. tuberculifer and C.
lichtensteinii in what appears to be an important behavioural character. In the tentative grouping
of the species of Ceramius given by Richards (1962: 83) C. capicola and C. linearis are placed
in the same species-group together with C. bicolor and C. socius Turner which may therefore be
expected also to omit the construction of a mud-cell.

The mud-cell constructed by C. lichtensteinii within the excavated-cell bears a close
resemblance to the aerial cell constructed by Pseudomasaris edwardsii (Cresson) as described
and illustrated by Torchio (1970). It is possible therefore to regard the construction of a mud-cell
within an excavated-cell as in C. tuberculifer, C. lichtensteinii and J. confusa as being
behaviourally intermediate between the excavation only of a cell as in C. capicola and C.
linearis and the presumably more advanced construction of an aerial mud cell as in the genera
Gay el la, Masaris, Pseudomasaris and Celonites.

With respect to the four species of the present study it is interesting that a correlation exists
between the type of cell and the length of the secondary shaft. Thus in C. capicola and C.
linearis the simple excavated-cells terminate long secondary shafts whereas in C. lichtensteinii
and J. confusa the excavated-cells containing the constructed-cells terminate very short secon-
dary shafts.

Oviposition into the empty cell follows the pattern common to all Vespoidea and has been
recorded for C. tuberculifer (Giraud, 1871; Ferton, 1901), C. lichtensteinii (Brauns, 1910;
present paper), C. capicola and J. confusa (present paper). However, with respect to the latter
three species no evidence was found during the present study that the egg was in any way
attached to the cell wall being always found lying loose on the floor of the cell at its inner end.
This is in contrast to the suspension of the egg by a filament from the top or side of the cell as in
many Eumenidae or the glueing of the egg to a cell wall as in the Vespidae. The egg of
Pseudomasaris edwardsii (Cresson) (Masaridae) is reported to be normally anchored by its
narrowed posterior tip to the cell wall near its base (Torchio, 1970: 7) but it is not stated how this
attachment is effected.

According to Ferton (1901) the egg of C. tuberculifer is deposited only provisionally at the
bottom of the cell and after the cell has been provisioned with a firm pollen loaf of characteristic
retort-like shape (see Ferton, 1901: Plate 1, Fig. 10) the mother moves the egg onto the neck of
the “retort”, in which position the little larva is alleged to begin feeding. Much as been made of
Ferton’ s assertions by Malyshev (1968: 263) who, in his chapter on the genesis of bees, has
based his “Secondary Bee Phase of Vespoid Type” upon them.

The present authors cannot accept Ferton’ s assertions concerning the transfer of the egg by
the female wasp from a provisional site of deposition to a pollen loaf specially shaped for its
reception. In the cells of C. lichtensteinii of which a large number were examined over the years,
the egg was left where first deposited and the larva upon hatching found its own way onto the

80

GESS AND GESS: ETHOLOGICAL STUDIES OF JUGURTIA CONFUSA RICHARDS, ETC. (HYMENOPTERA)

nearby pollen loaf. There is no reason to suppose C. tuberculifer to be different in this respect.
Moreover, it is difficult to visualise how it would be physically possible for the female wasp to
reposition her egg onto the pollen loaf as the latter would be situated between her and the egg. It
is believed that the retort-shape of the pollen loaf described by Ferton was not the result of any
moulding by the female wasp but was simply an artifact of the larva’s feeding, that Ferton drew
the wrong conclusions, and that Malyshev’s hypothesis is therefore based upon false premises.

Some confusion exists in the literature as to whether Ceramius species practise mass
provisioning or progressive provisioning.

Giraud (1871) recorded open C. tuberculifer cells each with a larva that either had no
provision at all, had varying amounts of provision, or were almost full. The impression gained
from the above is that mass provisioning is probably practised. However, somewhat later in his
paper Giraud made a statement which may easily be taken ambiguously, namely that it appeared
to him evident that the female wasp continues to carry in nourishment after the hatching of the
larva, as was demonstrated by the insufficiency of the provision deposited in the cell of the young
larva and by the presence of the female in the galleries leading to the cells.

Ferton (1901) with respect to the same species, and as already discussed above, described
and figured a pollen loaf which by its completeness dearly indicates mass provisioning rather
than progressive provisioning.

Brauns (1910; 445) with respect to C. lichtensteinii claimed that he had never found stored
pollen and nectar masses in the cells with the larvae, even when the latter were still small and
stated that it was certain that the wasp feeds its larvae until such time that they are full-grown,
the provision being nectar (“Sicher ist, dass die Wespe ihre Larven solange fiittert, bis dieselben
erwachsen sind, und zwar mit Blumenhonig.”).

Brauns’ contention that C. lichtensteinii practises progressive provisioning has been men-
tioned by Richards (1962: 29) who did not comment other than to state that this was not recorded
for the European species studied. Torchio (1970: 31), presumably on the strength of Brauns’
assertion, has listed the genus Ceramius as practising progressive provisioning in contrast to the
genera Euparagia, Gayella, Paragia, Pseudomasaris and Celonites which he lists as not
provisioning progressively.

Malyshev (1968: 259) not only accepted Brauns’ statement but elaborated upon it, writing
that: “This method of progressive feeding of the larvae on honey (sic !), provided when it is
needed and only given directly into the larva’s mouth, is bound to reflect the moment . . . when
the instincts of the wasp were transformed into those of the bee”. Malyshev’s hypothesis once
again is based on false premises for in C. lichtensteinii provisioning is not progressive.

As was established in the present study and as has been dealt with in some detail earlier in
the present publication C. lichtensteinii (as also C. capicola, C. linearis and/, confusa) practises
mass provisioning, and, under optimal conditions of favourable weather and an abundance of
forage flowers, provisioning and sealing of the cell is completed by the female before the egg
hatches. Under less favourable conditions the rate of provisioning is slowed down leading to the
finding of unsealed cells containing larvae and varying amounts of provision as recorded by
Giraud, and under really unfavourable conditions the situation as reported by Brauns results.
There is certainly no shred of evidence that Ceramius differs in any way in its method of
provisioning from the other genera for which this facit of behaviour is known.

Brauns (1910: 445) stated with respect to C. lichtensteinii that several females appeared to
work in a single nest as he frequently saw several females disappearing one after the other down
the same turret. The present authors found no evidence that more than a single female nests in a
nest at any one time either with respect to C. lichtensteinii or with respect to C. capicola, C.
linearis and J. confusa. However, as already stated earlier in the present publication, C.
lichtensteinii (as also J. confusa ) has perennial nests resulting from the reuse of the nest in
successive nesting seasons by females produced in the nest. The present authors believe Brauns’

81

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13, PT 6, JUNE 1980

misconception may readily be explained by what appears to be the likely sequence of events in a
perennial nest at the beginning of the flight season. Thus, at that time it seems probable that the
several wasps, both males and females that have emerged from their cells in the nest should
initially remain associated with the nest for shelter, returning to it at night or when the weather is
unfavourable. It seems likely also that this co-existence in the nest should persist until such time
when one of the females, possibly the first-emerged and therefore the most advanced, has reached
the stage of maturity when her nesting urges begin, when she takes over the nest and drives out
from it the males and the other females which latter have to emigrate and initiate new nests. That
both sexes of species of Ceramius do for the greater part retire to the nests for the night is
recorded by Brauns (1910: 446). Similarly, in the present study, females of C. lichtensteinii, C.
capicola and J. confusa were found to shelter in their nests during periods of unfavourable
weather. It is therefore believed that Brauns’ observation of several females entering a single nest
was made during the post-emergence — pre-nesting period at the beginning of the flight season
and that the females concerned were sheltering rather than nesting. Brauns’ further observation ij
that sometimes both sexes may be found at night perched upon plants (1910: 446) most probably
pertains to individuals which have been evicted from the nest in which they developed, the
females amongst them not yet having initiated new nests.

Summing up, the picture of the ethology of Ceramius and Jugurtia which emerges from the
studies to date is as follows:

1. Nesting is in the ground.

2. There is a tendency towards the formation of pseudo-colonies.

3. Water is required for the excavation of the nest and is collected from a pool by the
female, in Ceramius standing on the surface of the water and in Jugurtia at the edge of
the water.

4. The nest consists of a main shaft and secondary shafts each terminating in an
excavated-cell and is surmounted by a turret constructed of mud pellets.

5. C. lichtensteinii , C. tuberculifer and J. confusa in addition construct a mud-cell within
each excavated-cell but such constructed mud-cells are absent in nests of C. capicola i
and C. linearis.

6. The mud for constructing turrets and mud-cells is obtained entirely from within nest
excavations.

7. The egg is laid free within the empty cell.

8. Mass provisioning is practised and the provision is in the form of a pollen loaf.

9. Each nest is worked upon by a single female in any one season.

10. In C. lichtensteinii and J. confusa there is reuse of nests in successive years.

11. Ceramius and Jugurtia are in their ethology very similar.

ACKNOWLEDGEMENTS

The authors wish to thank Mr T. C. White of the farm Hilton for his much appreciated
kindness over the years in allowing them free access to his land. Similarly, thanks are due also to
the late Mr H. H. Norton of the farm Clifton for permission to investigate insects there. Dr D. J.
Brothers of Natal University is thanked for determining the Mutillidae found associated with the
nests of C. lichtensteinii . The senior author is grateful to the C.S.I.R. for a running expenses it:
grant for field work during the course of which the present observations were made.

REFERENCES

Brauns, H. 1910. Biologisches iiber sudafrikanische Hymenopteren. Z. wiss. Insekt Biol. 6: 384-387, 445-447.

Ferton, C. 1901. Notes detachees sur l’instinct des Hymenopteres melliferes et ravisseurs avec la description de
quelques especes. Ann. Soc. ent. France 70: 83-148.

Fonscolombe, H. Boyer de 1835. Description du Ceramius fonscolombii (Latr.). Ann. Soc. ent. France 4: 421-427.

82

GESS AND GESS: ETHOLOGICAL STUDIES OF JUGURTIA CONFUSA RICHARDS, ETC. (HYMENOPTERA)

Gess, F. W. 1965. Contribution to the knowledge of the South African species of the genus Ceramius Latreille
(Hymenoptera: Masaridae). Ann. S. Afr. Mus. 48 (11): 219-231.

Gess, F. W. 1968. Further contribution to the knowledge of the South African species of the genus Ceramius Latreille
(Hymenoptera: Masaridae). Novos Taxa ent. 57: 1-14.

Gess, F. W. 1973. Third contribution to the knowledge of the South African species of the genus Ceramius Latreille
(Hymenoptera: Masaridae). Ann. Cape Prov. Mus. (nat. Hist.) 9 (7): 109-122.

Gess, F. W. and Gess, S. K. 1974. An ethological study of Dichragenia pulchricoma (Arnold) (Hymenoptera:
Pompilidae), a southern African spider-hunting wasp which builds a turreted, subterranean nest. Ann. Cape
Prov. Mus. (nat. Hist.) 9 (11): 187-214.

Gess, F. W. and Gess, S. K. 1975. Ethological studies of Bembecinus cinguliger (Smith) and B. oxydorcus (Handl.)

(Hymenoptera: Sphecidae), two southern African turret-building wasps. Ann. Cape Prov. Mus. (nat. Hist.) 11
(2): 21-46.

Gess, F. W. and Gess, S. K. 1976. An ethological study of Parachilus insignis (Saussure) (Hymenoptera: Eumenidae) in
the Eastern Cape Province of South Africa. Ann. Cape Prov. Mus. (nat. Hist.) 11 (5): 83-102.

Giraud, J. 1871. Miscellanees hymenopterologiques. Ann. Soc. ent. France (5) 1: 375-419.

Malyshev, S. I. 1968. Genesis of the Hymenoptera and the phases of their evolution. (Translated from the Russian; edited
by O. W. Richards and B. Uvarov.) London. Methuen.

Peringuey, L. 1909. Descriptive catalogue of the Coleoptera of South Africa: Family Meloidae. Trans. R. Soc. S. Afr. 1
(1): 165-297.

Richards, O. W. 1962. A revisional study of the Masarid wasps ( Hymenoptera , Vespoidea). London. British Museum
(Natural History).

Torchio, P. F. 1970. The ethology of the wasp, Pseudomasaris edwardsii (Cresson), and a description of its immature
forms (Hymenoptera: Vespoidea, Masaridae). Los Angeles County Museum Contr. in Science 202: 1-32.

Manuscript accepted for publication 18 March 1980.

KEY TO LETTERING OF CELLS IN FIGS 1 -5.

A. Cell open, containing old cocoon from which adult wasp has emerged.

B. Cell closed, containing pupa in cocoon.

C. Cell dosed, containing prepupa in cocoon.

D. Cell closed, containing mature larva spinning cocoon.

E. Cell dosed, containing mature larva prior to cocoon spinning.

F. Cell either open or closed, containing still feeding immature larva.

G. Ceil either open or closed, containing egg with provision.

H. Cell open, containing egg without provision.

I. Cell open, empty.

X. Cell either open or closed, development of young aborted.

83

ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

Ann . Cape Prov. Mus. (nat. Hist.)

VOLUME 13 • PART 7
9th JULY 1980

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

Prey and nesting sites of some sympatrk species of Cerceris (Hymenoptera:
Sphecidae) with a review and discussion of the prey diversity of the genus

by

F. W. GESS

(Albany Museum, Grahamstown)

CONTENTS

Page

Abstract 85

Introduction 85

The Cerceris species present at Hilton and the relative frequency of their occurrence .... 86

Identification of the nesting sites 86

Identification of the prey 87

Provenance of the prey 89

Discussion of the nesting sites and the prey in relation to ecological displacement 90

Review and discussion of the prey diversity of the genus Cerceris 91

Acknowledgements 93

References 93

ABSTRACT

Nineteen sympatric Cerceris species are examined in relation to frequency of occurrence,
nature of nesting sites and of prey. Their nesting sites and prey are discussed in relation to their
ecological displacement. There follows a review and discussion of the prey diversity of the genus
Cerceris which puts forward the hypothesis that Cerceris is an Old World genus and that its
original prey was hymenopterous. Prey records of a total of twenty Afrotropical species are
included of which twelve are new associations and three confirm previously published records.

INTRODUCTION

The genus Cerceris Latreille is cosmopolitan in its distribution and with over 850 known
species is the largest genus of the Sphecidae. All species are ground-nesting and provision their
young with insect prey.

The present paper is based upon observations of sympatric Cerceris species made over a
period of seven years at Hilton, a farm situated 18 kilometres WNW. of Grahamstown (33°19’S.,
26°32’E. ) in the Albany Division of the Eastern Cape Province of South Africa.

In view of the large overall number of species and the frequency with which considerable
numbers of species often occur together, both spatially and temporally, the ecological displace-
ment of the species — that is the different ways in which the species exploit their habitat — is of
great interest. Clearly of importance in such ecological displacement is specificity in such basic
ethological factors as the choice of nesting site, and more particularly the choice of prey. It is

85

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 7, JULY 1980

these aspects of the ethology of the Cerceris species at Hilton which are set forth in the present
account.

This is the eighth of a series of publications covering some aspects of the ethology of solitary
wasps occurring at Hilton, the climate and vegetation of which have been previously described
(Gess and Gess, 1974: 191-192).

Prey records, both published and unpublished, from localities other than Hilton have been
included in the account in an attempt to draw together information on the chosen prey of as many
Afrotropical Cerceris species as possible, little having hitherto been published on the subject.

THE CERCERIS SPECIES PRESENT AT HILTON AND THE RELATIVE FREQUENCY OF

THEIR OCCURRENCE

Nineteen species of Cerceris have been recorded from Hilton. Arranged in alphabetical order
these are: C. amakosa Brauns, C. armaticeps cajfrariae Empey, C. diodonta diodonta Schlet-
terer, C. discrepans discrepans Brauns, C. dominicana Brauns, C. erythrosoma Schletterer, C.
holconota holconota Cameron, C. hypocritica Brauns, C. languida languida Cameron, C.
latifrons latifrons Bingham, C. / uni g era Dahlbom, C. nasi dens obscura Schletterer, C. nigri-
frons nigrifrons Smith, C. oraniensis Brauns, C. pearstonensis pearstonensis Cameron, C.
pictifacies Brauns, C. ruficauda ruficauda Cameron, C. rufocincta polychroma Gribodo and C.
spinicaudata spinicaudata Cameron.

The species varied greatly in the frequency of occurrence, as reflected by catches and
sightings. The most commonly met with species, in decreasing order, were C. latifrons, C.
languida and C. rufocincta polychroma, C. holconota, C. spinicaudata and C. pearstonensis.
Less commonly met with species were C. nigrifrons and C. ruficauda, C. lunigera, C. nasidens |
obscura and C. oraniensis. Rare were C. hypocritica, C. amakosa, C. diodonta and C.
pictifacies, C. dominicana and C. erythrosoma, C. discrepans and C. armaticeps.

The flight periods of all the Cerceris species at Hilton fall between mid-October and
mid- April.

IDENTIFICATION OF THE NESTING SITES

Nests belonging to ten species were located at Hilton. All were constructed in level or nearly
level bare ground which was sufficiently friable to allow the females to excavate their burrows
using their mandibles and legs but no water. The ground was firm though in sandy situations the
firm underlying sand was sometimes overlain by a layer of loose material.

Six species were found to nest in sandy soil, two sites being particularly favoured: a sandpit
(see Gess and Gess, 1980: Fig. 1) and the area adjacent to it, and a very gently sloping bank
margining a car track where the latter crosses the bed of a seasonal tributary of the New Year’s
River. The sand, light coloured and fine grained, is derived from the weathering of Witteberg
Quartzite and is of alluvial origin having been deposited upon its flood plain by the above
seasonal water course.

The species found nesting in these sites were: C. holconota (seven nests; both in the sandpit
and on the gently sloping bank where it nested in company with Bembix albofasciata)', C.
languida (many nests; in the sandpit in very fine sand where it nested in company with
Bembecinus braunsii and B. haemorrhoidalis)', C. latifrons (many nests; as for C. holconota)', C.
oraniensis (4 nests; on level ground immediately above and below the lip of the sandpit); C.
rufocincta (several nests; in the sandpit where it sometimes nested in company with Bembecinus
braunsii and B. haemorrhoidalis)', and C. spinicaudata (three nests; as for C. holconota).

Four species were found to nest in disturbed clayey soil immediately adjacent to a water
furrow (see Gess and Gess, 1976: Plate 2). The clayey soil, reddish-brown in colour, is derived

86

PREY AND NESTING SITES OF SOME SPECIES OF CERCERIS

from the weathering of Upper Witteberg Shales previously known as Lower Dwyka Shales and
referred to as such by Gess and Gess (1974: 192 and 1975: 24). Though less compacted than
undisturbed clayey soil the substrate nested in by the four species was nevertheless less friable
than the soil in sandy areas. The species concerned were: C. lunigera (1 nest); C. pearstonensis
(1 nest); C. ruficauda (1 nest) and Cerceris sp. A (1 nest). The last listed species is unidentified
owing to the unfortunate escape during the opening up of the nest of the female nest builder.
Whereas it is very probable that the species was one of the nineteen recorded from Hilton, it is
clear, knowing its chosen nesting substrate and prey, that it was not one of the named species
recorded above as having been found nesting.

IDENTIFICATION OF THE PREY

The identity of the prey of thirteen of the nineteen species of Cerceris recorded from Hilton is
known. Prey identifications for nine of these species were made at Hilton, seven of the
associations having been previously unknown and the other two confirming previously recorded
associations. For four of the ten species for which prey identifications were not made at Hilton
information as to the nature of the prey taken is available from other sources. The Cerceris
species are dealt with individually below.

Cerceris sp. A.

In a nest excavated on 6.xii. 1974 and containing forty-one uneaten prey this species was found
to have provisioned its cells solely with a 3, 5-4,0 mm long metallic-green species of
PTEROMALIDAE (Hymenoptera: Chalcidoidea).

Cerceris erythrosoma Schletterer

This species was not associated with its prey at Hilton but is known to provision with beetles
of the family CURCULION1DAE. Brauns (1911: 239 and 1926: 278) recorded the prey as a
species of Tanymecus. Similarly, prey associated with females collected by Jacot Guillarmod at
Mamathes in Lesotho and now in the collections of the Albany Museum consist in one instance
of a 10 mm long specimen of Tanymecus makkaliensis Fhs. (det. G. A. K. Marshall) and in five
instances of 6, 8-7, 6 mm long specimens of Protostrophus sp. near sceleratus H. v. S. (det. R.
T. Thompson).

Cerceris holconota holconota Cameron

Six females seen transporting prey were captured, four after they had revealed their nests
which were subsequently excavated to recover the prey stored in the caches. The thirteen prey
obtained were representative of six families of Hymenoptera as detailed below: BRACONIOAE,
?Genus & sp. (1 female, 13. i. 1975); BETHYLIDAE, ?Genus & sp. (1 female, 20.xii. 1974);
TIPHIIDAE, Anthobosca sp. (1 male, 13. i. 1975), Braunsomeria sp. (4 males, 13. i. 1975 and 1
male, 16.xi. 1977), Mesa incisa (Cameron) (2 females, 20.xii. 1974); MUTILLIDAE, Dasylabris
eunyce (Peringuey) (1 male, 13. i. 1975); FORMICIDAE, ICamponotus sp. (1 winged male,
ll.xii. 1975); HALICTIDAE, Lasioglosswn sp. (1 female, 20.xii. 1974).

Cerceris languida languida Cameron

A female transporting prey was captured on each of the following three dates: 24.xi. 1977,
lO.i. 1978 and 23. ii. 1978. The first female was allowed to enter her nest which was subsequently
excavated and yielded 32 prey from the cache. All 34 prey obtained from the three females were
beetles of the family PHALACRIBAE and represented a single, 1,7 mm long, black Olibrus sp.

Cerceris latifrons latifrons Bingham

Females transporting prey were captured on 19.xi. 1973 (1), 20.xii.1974 (1), lO.i. 1975 (1),

87

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 7, JULY 1980

9.xii. 1975 (4), 2. xii. 1 977 (3) and 2. i. 1978 (1). All eleven prey obtained were beetles of the tribe
Hopliini, SCARABAEIDAE: MELOLONTHINAE. Two 4, 4-4,7 mm long species were rep-
resented. The prey taken at Hilton agrees with that recorded by Arnold (1940: 106) for the
subspecies C. latifrons sedula Arnold from Rhodesia, namely “an unidentified species of beetle
(Hoplinae)”.

Cerceris lunigera Dahlbom

One female seen transporting prey on 6. xii. 1974 was captured after revealing her nest. The two
prey obtained from the cache in the excavated nest were bees of the family HALICTIDAE,
namely Halictus sp. (female) and Lasioglossum sp. (female).

Cerceris nasidens obscura Schletterer

This species was not associated with its prey at Hilton but according to E. McC. Callan (pers.
comm.) it has been found provisioning nests in Grahamstown with “small melolonthid beetles”,
SCARABAEIDAE: MELOLONTHINAE. The tribe to which the beetles belong is apparently not
Hopliini.

Cerceris nigrifrons nigrifrons Smith

This species was not associated with its prey at Hilton. However, Brauns (1926: 320) reported
the prey to be beetles of the family BUPRESTIDAE, for instance Sphenoptera sp. in Rhodesia.

Cerceris oraniensis Brauns

Four females, each carrying a single prey, were captured at or near their nest entrances on
16.ii. 1978. A fifth prey was obtained from the cache of a nest excavated on the same day. All the
prey were beetles of the family CURCULIONEDAE and represented a single, 7, 0-7, 4 mm long
Protostrophus species.

Cerceris pearstonensis pearstonensis Cameron

This species was not associated with its prey at Hilton but was found at other localities to
provision with beetles of the family CURCULIONIDAE. At Strowan, near Grahamstown, on
30. xi. 1970, a female was caught when flying with a 4,6 mm long Protostrophus sp. Prey ji
associated with two females collected by Jacot Guillarmod at Mamathes, Lesotho, and now in the
collections of the Albany Museum are likewise small weevils. The larger of the two, 4,4 mm
long, is a Protostrophus sp., the smaller, 4,2 mm long, belongs to some other genus.

Cerceris ruficauda ruficauda Cameron

A nest of this species excavated on 10. xii. 1974 was found to contain twenty-six uneaten prey,
beetles of the family CHRYSOMELIDAE: CRIOCERINAE. A single 3,4 mm long blackish-
bronze species was represented.

Cerceris rufocincta polychroma Gribodo

Eight females seen transporting prey were captured, one after it had revealed its nest which
was subsequently excavated to recover the prey stored in the cache. The ten prey obtained were
representative of two families of Hymenoptera as detailed below: TIPHIIDAE, Anthobosca
rufithorax (Cameron) (1 female, 2. i. 1978), Tiphia sp. (1 male, 13. xii. 1977 and 1 male, 2. i. 1978);
MUTILLIDAE, Chrestomutilla sp. (1 male, 24. i. 1978), Dasylabroides caffra (Kohl) (1 female
!!, 13.xii. 1977; 1 male, 17.i. 1978; 2 males, 24.i. 1978), Psammotherma flabellata (F.) (1 male,
20. xii. 1974; 1 male, 13.U975).

Cerceris spinicaudata spinicaudata Cameron

Three females seen transporting prey were captured after revealing their nests which were
subsequently excavated to recover the prey stored in the caches. The nine prey obtained were all

88

PREY AND NESTING SITES OF SOME SPECIES OF CERCERIS

small bees of the family HALICTIDAE. Species of two genera were represented: Lasioglossum
spp. (4 females representative of 3 species, 20.xii. 1974; 1 female, lO.i. 1975) an&Nomioides sp.
(probably N. halictoides Bliithgen) (4 females, 20.xii. 1974).

Prey associated with two females collected by Jacot Guillarmod, one at Strowan, near
Grahamstown (27. ii. 1972) and the other at Mamathes, Lesotho, and now in the collections of the
Albany Museum are likewise females of species of Lasioglossum. The present records confirm
the association given by Brauns (1926: 337) who wrote: “I found it carrying Halictus sp. for
larval food”.

It should be noted that, though the genus Halictus occurs in southern Africa, most species at
one time allocated to it are now included in the closely related genus Lasioglossum. Brauns’
record may thus apply to either Halictus or Lasioglossum.

PROVENANCE OF THE PREY

For some species at least it was established where the prey taken by them was to be found,
therefore allowing the situation to be identified in which the wasps hunt and respond to prey of
certain size and behaviour. For other species it is possible to speculate concerning the identity of
the situation by using circumstantial evidence.

Olibrus sp. (Phalacridae) the sole prey of C. languida was found to be very common in the
flowers of Lasiospermum bipinnatum (Compositae), an annual herb growing in the sandpit and
elsewhere. Recorded from this plant in October and November during the first weeks of the
wasp’s flight period, the beetle was undoubtedly present both then and later in the summer in the
flowers of other herbaceous composites as well. The flowers of species of Senecio, a genus of
common occurrence at Hilton, were found by H. Andreae (pers . comm. ) to be frequented by
three Olibrus species at the Cape. The beetle is associated throughout its life with its host plant,
the egg being laid in the flower and the larva feeding in the capitulum until full-grown when it
bores down the stem to pupate in the ground. In European species there may be six generations in
the year (Imms, 1957: 801). From the above it is apparent that C. languida must hunt for its prey
in the flowers of Compositae, probably all small annuals likeL. bipinnatum . The commonness of
such plants and the large number of generations of Olibrus probably developing upon them
makes possible the extended flight and nesting period — continuously from the first half of
October to the beginning of March — established at Hilton for C. languida.

The Hopliini (Scarabaeidae) which constitute the prey of C. latifrons are as a tribe characteris-
tically found in the flowers of Compositae so that this wasp may be expected to hunt in a similar
though not necessarily the same situation as that pertaining to C. languida.

The Protostrophus sp. (Curculionidae) found to be the prey of C. oraniensis was common on
the foliage of the annual herbaceous weed Conyza bonariensis (Compositae) during the only
period, mid-February, during which this wasp was observed nesting. The beetle infested plants,
growing in the sandpit, were in close proximity to the nests.

Protostrophus, recently monographed by Van Schalkwyk (1968) is a very commonly met with
genus, the majority of the 136 recognized species being found in South Africa. The species are
inconspicuous, mostly dull coloured and are unable to fly. In body length they range from
3-9mm. The beetles are apparently not restricted to any particular plants, the eggs being dropped
to the ground where ever the females are feeding. The larvae lead a subterranean life, probably
feeding upon decaying vegetable matter. Mass emergences of thousands of adults have been
recorded when great damage may be done by their feeding on the foliage of many different plants
including seedling trees. From the observed presence at Hilton of Protostrophus sp. on the
foliage of Conyza and the known biology of the weevil genus it would appear that C. oraniensis
hunts for its prey amongst foliage at no great height above the ground. It is probable that the

89

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 7, JULY 1980

other two Cerceris species known to prey at least partly upon Protostrophus and occurring at
Hilton, C. erythrosoma and C. pearstonensis hunt in similar situations.

The species of Criocerinae (Chrysomelidae), the sole recorded prey of C. ruficaudci may be
expected to occur on foliage as is characteristic for the subfamily and it is there that the wasp
undoubtedly does its hunting.

For the species of Cerceris which hunt Hymenoptera two possible hunting situations im-
mediately present themselves: at flowers and young growth visited by the prey for purposes of
foraging for nectar or glandular exudates, or at the nests of the prey.

It was noted at the time when the nest of C. lunigera and those of C. spinicaudata were
excavated that they were situated in close proximity to the nests of various small bees some of
which at least were those of species represented amongst the prey of the wasps. This was
particularly striking with respect to C. spinicaudata and Nomioides sp. (probably N. halictoides)
which nested next to one another in the sand. It therefore seems possible that the hunting of these
two Cerceris species may take place at or near the nest entrances of their prey and in close
proximity to the entrances of their own nests. The fact that only female bees are recorded as prey
would lend support to this hypothesis. The latter would be in accord with the findings of Marchal
(1887) as reported by Hamm and Richards (1930: 106) on the behaviour of C. rybyensis (L.)

( = C. ornata Fabr.) in France. That wasp nested in a garden path in close proximity to
innumerable nests of Andrena and Hal ictus on which species it was preying. The Cerceris was
reported to circle round the nests, every now and then dropping to the ground, or even entering a
burrow. Only bees returning home laden with pollen were attacked, and these were knocked
down and stung as they hovered over their nests before entering.

With respect to C. holconota and C. rufocincta polychroma on the other hand, the wide range
of unrelated prey of both sexes and of very diverse habits and behaviour makes it likely that the
only situation in which these Cerceris species in hunting would meet all the prey species would
be at flowers and young growth to which the latter would go for foraging purposes. In the areas
in which the two Cerceris species nested — in and near the sandpit — the flowering plants known n
to have been visited by a mixed company of Braconidae, Tiphiidae, Mutillidae and others were i,
Selago corymbosa (Selaginaceae) and various low-growing Compositae including Helichrysum
and Lasiospermum. It therefore appears that C. holconota and C. rufocincta polychroma hunt for
their prey on and around flowers of low-growing herbaceous plants.

DISCUSSION OF THE NESTING SITES AND THE PREY IN RELATION TO ECOLOGICAL

DISPLACEMENT.

The existence at Hilton of two basically very different soil types, one clayey and the other
sandy, and the specificity shown by the wasps in their choice of nesting substrate has resulted in ij
the Cerceris species of that locality being divided for the purposes of nesting into two distinct
non-competitive groups, indicating that an area with more than one type of friable soil can |i
support a larger number of Cerceris species than one with a uniform soil type.

Other factors, for example the depth of the friable soil, will be limiting. The species nesting in
sand place their caches at depths ranging from 60 ± mm in C. rufocincta to 500 + mm in C.
latifrons. Therefore, whereas C. latifrons can only nest in relatively deep sand C. rufocincta and
other shallow-nesters are less restricted.

Hunting by Cerceris species at Hilton appears to take place at no great distance from the nest.
This being so, competition for prey between species such as for example the clay-nesting C.
lunigera on the one hand and the sand-nesting C. spinicaudata on the other is avoided as,
although both hunt halictine bees, their hunting areas are distinct. This is especially true for these
species which are believed to prey upon bees nesting in close proximity to their own nests.

All the species are highly prey specific with the exception of C. rufocincta and more

90

PREY AND NESTING SITES OF SOME SPECIES OF CERCERIS

particularly C. holconota which, with their recorded prey representing respectively two and six
families of Hymenoptera, appear to be catholic in their choice. With the high prey specificity
generally shown competition for prey is effectively avoided despite the fact that a number of
species nesting in the same soil type may hunt in very similar situations. The most commonly
taken prey recorded or known for Cerceris species occurring at Hilton are species of Halictidae
and Curculionidae, notably Protostrophus species. With respect to the three species hunting
Protostrophus weevils, differences in preferred size range of prey either alone or in conjunction
with differences in substrate-determined nest situation are likely factors of importance in the
avoidance of competition.

REVIEW AND DISCUSSION OF THE PREY DIVERSITY OF THE GENUS CERCERIS

The most interesting aspect of the behaviour of Cerceris concerns the diversity of insects taken
as prey by the genus and the specificity shown by individual species.

Provisioning is with adult Coleoptera or Hymenoptera, depending upon the Cerceris species.
From data presented by Bohart and Menke (1976: 576) and by Iwata (1976: 150-151) it appears
that eleven families of Coleoptera have so far been recorded as prey: Anthribidae, Bruchidae,
Buprestidae, Cerambycidae, Chrysomelidae, Coccinellidae, Curculionidae, Nitidulidae, Phalacri-
dae, Scarabaeidae and Tenebrionidae. Published records of Hymenoptera taken as prey mostly
concern Apoidea.

Coleoptera appear to be the more commonly taken order and have been recorded as prey of
Cerceris from many parts of the world. North American species of Cerceris provision exclu-
sively with Coleoptera and the few records pertaining to South America concern prey of the same
order. The few prey records for Australian Cerceris similarly concern Coleoptera. In the
! Palaearctic, Oriental and Afrotropical Regions of the Old World, however, there are, in addition
I to species provisioning with Coleoptera, a minority of species that utilize Hymenoptera. Thus the
Palaearctic C. rybyensis (L. ), C. hortivaga Kohl and C. sabulosa (Panzer) and the Oriental C.
pictiventris Dahlbom provision with Apoidea, especially Halictidae and some in addition with
Colletidae and Andrenidae.

Though Apoidea appear to be the most commonly taken non-coleopterous prey, four records of
the use of Hymenoptera other than bees have been published (see Bohart and Menke, 1976: 576).
In the Palaearctic, C. stratiotes Schletterer is believed to prey only on the chalcid, Stilbula
cynipiformis (Rossi), a Cerceris species provisioning with halictid and andrenid bees was found
to have also in its brood cells a species of Psen (Sphecidae), and a female C. pekingensis
Tsuneki was reported to have taken a species of Pison (Sphecidae) into her burrow. The fourth
record concerns the Oriental C. langkasukae Pagden which was reported carrying a species of
Hingstoniola (Sphecidae) although her regular prey were buprestids.

The prey of eight Afrotropical Cerceris species has been recorded in print. On the African
mainland, Curculionidae are taken by C. chirindensis Arnold (Arnold, 1932: 13), C. emeryana
varilineata Cameron (Brauns, 1926: 322, as C. varilineata Cam.) and C. erythrosoma Schletterer
(Brauns, 1911: 239 and 1926: 278); Buprestidae are taken by C. nigrifrons Smith (Brauns, 1926:
320); and Scarabaeidae are taken by C. latifrons sedula Arnold (Arnold, 1940: 106, as C. sedula
Arnold). On Malagasy, Curculionidae are taken by C. clypearis Saussure; Chrysomelidae and
Buprestidae by C. albotegula Arnold; and nomiine bees (Halictidae) by C. nenitra Saussure (all
Arnold, 1945: 23, 41 and 42 respectively).

Additional records (all C. F. Jacot Guillarmod’s and unpublished) obtained from the Albany
Museum collections concern C. bothavillensis Brauns and C. emeryana multicolor Arnold (both
from Mamathes, Lesotho) and C. multipicta fuscifacies Empey (from the Transvaal Lowveld) all
of which took Curculionidae, in the case of the first named species, Leurops sublineata Marshall.

With the addition of the records from Hilton, the prey of a total of 20 (i.e. 10,6%) of the 189

91

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 7, JULY 1980

Afrotropical species of Cerceris recognized by Empey (1969) is known. The corresponding
figures for America north of Mexico are 21 (i.e. 27%) of 78 species (see Scullen, 1965 and
Evans, 1971: 509). Of the twenty Afrotropical species concerned, fourteen species prey upon \
Coleoptera and six upon Hymenoptera whereas, as previously noted, all North American prey
recorded comprised Coleoptera. With respect to both regions the most commonly taken prey are
species of Curculionidae, 40% and 48% of the species for which prey is known taking this family
in the Afrotropical Region and in America north of Mexico respectively.

Cerceris is most unusual though not unique (see Bembix) in the Sphecidae in provisioning
with prey belonging to more than one order, with, as has been shown, a majority of species
utilizing Coleoptera and a minority utilizing Hymenoptera. It is therefore of interest to examine
which of these two orders constitutes the original prey taken by the genus.

Suggested relationships within the subfamily Philanthinae based entirely upon consideration
and evaluation of morphological characters of presumed phylogenetic value have been discussed
by Bohart and Menke (1976: 557-558) and shown by them in a dendrogram (Fig. 183). On the
basis of the number of advanced characters appearing in each genus, Cerceris is seen as a highly
advanced genus, surpassed only by the North American Eucerceris.

If on Bohart and Menke’s dendrogram are entered details of the prey for each of the included !
genera (unknown for Eremiasphecium, Odontosphex and Philanthinus) it is seen that the species
of Philanthinae prey pre-eminently upon Hymenoptera and that only Cerceris (partly) and
Eucerceris (wholly) prey upon Coleoptera.

It therefore appears that within the Philanthinae the use of Hymenoptera as prey is primitive or
unspecialized and that the use of Coleoptera is advanced or derived. In this case Cerceris in its
prey selection represents the transitional stage between the less advanced condition and that
shown by Eucerceris. In this connection Pagden’s record (see Bohart and Menke, 1976: 576) of
C. langkasukae carrying a hymenopteran although her regular prey were Buprestidae may
represent atavistic behaviour in time of shortage of the coleopterous prey.

The change from Hymenoptera to Coleoptera by Cerceris can be shown to have been
advantageous for two reasons. Firstly, it enabled that fraction which made the change to avoid If!
competition with related genera (e.g. Philanthus ) and possibly others (e.g. Palarus) for i1
hymenopterous prey, and secondly, it made available to Cerceris as potential prey vast numbers
of species of many families belonging to an order which at least in its adult stage was not being
exploited by any other wasps.

The vast possibilities opened up by the adoption of the new prey may be considered to have
led to an outburst of speciation in the fraction of Cerceris concerned, leading to the pre-
dominance of Coleoptera-preying species over Hymenoptera-preying species and to the overall
magnitude of the genus Cerceris, which with a total of over 850 known species (Bohart and
Menke, 1976: 575) is the largest genus of the Sphecidae.

Within the range of Coleoptera preyed upon, the most commonly taken by Cerceris appear to
be species of Curculionidae and species of this family constitute the sole prey of Eucerceris. The
preference for weevils may be attributable to the overwhelmingly greater number of species in i
the Curculionidae than in other families.

With respect of those Philanthinae which utilize Hymenoptera as prey, some interesting trends
relevant to the prey taken by the various genera are apparent. Thus the sole recorded prey for
Pseudoscolia and the most characteristic prey of Trachypus, Philanthus and the Hymenoptera-
preying species of Cerceris are species of Halictidae though species of all three latter genera may
also utilize other families of bees as well as wasps of several aculeate and non-aculeate families.
On the other hand prey taken by three of the four genera of the Aphilanthopsini (prey is not
known for the fourth genus) consists in each case of a single genus of Formicidae.

It is clear that the genera of the Aphilanthopsini are greatly specialized in their choice of prey
(as is indicated also by the presence in Clypeadon and Listropygia of an “ant clamp” formed of

92

PREY AND NESTING SITES OF SOME SPECIES OF CERCERIS

the pygidial plate and the hypopygium) whereas the other genera represent the more un-
specialized or primitive condition. In those species which utilize a wide spectrum of prey, the
prey taken at any locality or time is probably dependant upon its abundance in the area and upon
its size relative to the provisioning wasp. The frequency with which Halictidae are taken by
species of Trachypus, Philanthus and the Hymenoptera-preying Cerceris species is undoubtedly
due to the commonness of species and individuals of this bee family. There does, however,
appear to be a tendency towards restriction in the range of prey taken. This may be seen in those
of the Hymenoptera-preying Cerceris occurring at Hilton for which a number of prey records is
available. Thus, in contrast to C. holconota for which six families including Halictidae are
recorded as prey, C. spinicaudata appears to be restricted to Halictidae and C. rufocincta
polychroma appears to specialize in Tiphiidae and Mutillidae.

Finally, from the fact that both Hymenoptera- and Coleoptera-preying species of Cerceris
occur in the Old World but only Coleoptera-preying forms occur in the New World it may be
speculated that Cerceris was originally an Old World Genus.

ACKNOWLEDGEMENTS

The author wishes to thank Mr T. C. White of the farm Hilton for his much appreciated
kindness over the years in allowing him free access to his land. Gratitude is expressed to the
C.S.I.R. for a running expenses grant for field-work during the course of which the present
observations were made.

REFERENCES

Arnold, G. 1932. New species of Ethiopian Sphegidae. Occ. Pap. Rhod. Mus. 1(1): 1-31.

Arnold, G. 1940. New species of African Hymenoptera. No. 4. Ann. Transv. Mus. 20(2): 101-143.

Arnold, G. 1945. The Sphecidae of Madagascar. Bulawayo. National Museum of Southern Rhodesia.

Bohart, R. M. and Menke, A. S. 1976. Sphecid wasps of the world: a generic revision. Berkeley: University of California
Press.

Brauns, H. 1911. Biologisches liber siidafrikanische Hymenopteren. Z. wiss. Insekt Biol. 7: 238-240.

Brauns, H. 1926. The Ethiopian Cerceris species. Ann. Transv. Mus. 11(4): 268-337.

Empey, H. N. 1969. Revision of the Ethiopian species of Cerceris Latreille, 1802 (Hymenoptera: Sphecidae). 1. A
synonymical checklist of the described species. J. ent. Soc. sth. Afr. 32(2): 297-331.

Evans, H. E. 1971. Observations on the nesting behavior of wasps of the tribe Cercerini (Hymenoptera: Sphecidae). J.
Kans. ent. Soc. 44: 500-523.

Gess, F. W. and Gess, S. K. 1974. An ethological study of Dichragenia pulchricoma (Arnold) (Hymenoptera:
Pompilidae), a southern African spider-hunting wasp which builds a turreted, subterranean nest. Ann. Cape
Prov. Mus. (nat. Hist.) 9(11): 187-214.

Gess, F. W. and Gess, S. K. 1975. Ethological studies of Bembecinus cinguliger (Smith) and B. oxydorcus (Handl.)

(Hymenoptera: Sphecidae), two southern African turret-building wasps. Ann. Cape Prov. Mus. (nat. Hist.)
11(2): 2 1 — 46.

Gess. F. W. and Gess. S. K. 1976. An ethological study of Parachilus insignis (Saussure) (Hymenoptera: Eumenidae) in
the Eastern Cape Province of South Africa. Ann. Cape Prov. Mus. (nat. Hist.) 11(5): 83-102.

Gess. F. W. and Gess, S. K. 1980. Ethological notes on Kohliella alaris Brauns (Hymenoptera: Sphecidae: Larrinae) in
the Eastern Cape Province of South Africa. Ann. Cape Prov. Mus. (nat. Hist.) 13(4): 45-54.

Hamm, A. H. and Richards, O. W. 1930. The biology of the British fossorial wasps of the families Mellinidae, Gorytidae,
Philanthidae, Oxybelidae, and Trypoxylonidae. Trans. R. ent. Soc. Lond. 78(1): 95-132.

Imms, A. D. 1957. A general textbook of entomology. (9th ed. revised by O. W. Richards and R. G. Davies). London:
Methuen.

Iwata, K. 1976. Evolution of instinct : comparative ethology of Hymenoptera. New Delhi: Amerind Publishing Co. for
Smithsonian Institution and National Science Foundation, Washington, D.C.

Marchal, P. 1887. Etude sur l’instinct du Cerceris ornata. Archs Zool. exp. gen. (2)5: 27-60.

Scullen, H. A. 1965. Review of the genus Cerceris in America north of Mexico (Hymenoptera: Sphecidae). Proc. U. S.
natn.Mus. 116: 333-548.

Van Schalkwyk, H. A. D. 1968. A monograph of the genus Protostrophus Marshall (Col: Curcuiionidae, Brachyderinae).

Entomology Mem. Dep. agric. tech. sen*. Repub. S. Afr. 15: 1-173.

Manuscript accepted for publication 10 April 1980.

93

Printed by Cape & Transvaal Printers (Pty) Ltd, Cape Town
BD8933

46/J

f\l i-i

ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

Ann. Cape Prov. Mus. (nat. Hist.)

VOLUME 13 • PART 8
18th JULY 1980

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

Some aspects of the ethology of Dasyproctus westermanni {Dahlbom)
(Hymenoptera: Sphecidae: Crabroninae) in the Eastern Cape Province of

South Africa

by

F. W. GESS

(Albany Museum, Grahamstown)

CONTENTS

Page

Abstract 95

Introduction 95

Description of nesting sites 96

Flight period 96

Flowers and young growth visited by adult wasps 96

Identification of the prey 97

Description of the nest 98

Method of construction of the nest, oviposition and provisioning 99

Life history 101

Parasites and other associated insects; mould 102

Review and discussion of the ethology of the genus Dasyproctus 103

Acknowledgements 105

References 106

ABSTRACT

Some aspects of the ethology of Dasyproctus westennanni (Dahlbom) are described. Of
particular note is the orientation of the pupae which appears to be governed by gravity rather than
the position of the nest entrance. Foraging records are given for D. bipunctatus Lep. & Brulle,
D. dubiosus (Arnold), D. immitis (Saussure), D. ruficaudis (Arnold) and D. westermanni
(Dahlbom). The ethology of the genus Dasyproctus is reviewed and discussed.

INTRODUCTION

Sixty-seven species of the genus Dasyproctus Lepeletier and Brulle (Sphecidae: Crab-
roninae) are listed by Bohart and Menke (1976:419) of which just over half occur in the
Afrotropical Region. The remainder are known from the Oriental and Australasian Regions.
Published biological accounts varying from fragmentary to fairly extensive pertain to six species:

95

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 8, JULY 1980

D. agilis (F. Smith) and D. buddha (Cameron) from the Oriental Region and D. barkeri
(Arnold), D. bipunctatus Lep. & Brulle, D. kibonotensis Cameron and D. stevensoni (Arnold)
from the Afrotropical Region.

The present paper is an account of some aspects of the ethology of a seventh species,
D. westermanni (Dahlbom) and is the ninth in a series of publications dealing with the ethology of
certain solitary wasps occurring at Hilton, a farm situated 18 kilometres WNW. of Grahamstown
(33° 19'S., 26° 32'E.) in the Albany Division of the Eastern Cape Province of South Africa.

In all, five species of Dasyproctus have been recorded from Hilton. These are:
D. bipunctatus Lep. & Brulle, D. dubiosus (Arnold), D. immitis (Saussure), D. ruficaudis (Arnold)
and D. westermanni (Dahlbom). With respect to D. bipunctatus three colour forms are present
which by some authors (e.g. Leclercq, 1958) are accorded subspecific rank: D. b. bipunctatus
Lep. & Brulle, D. b. lugubris (Arnold) and D. b. simillimus (Smith). D. b. simillimus (Smith) is
listed as a good species by Bohart and Menke (1976). Foraging records are given for all five
species.

DESCRIPTION OF NESTING SITES

Of the five species only D. westermanni was observed engaged in nesting activities. Nests
were situated up off the ground within galleries hollowed out by the wasp in green subvertical i
pithy inflorescence stems of Urginea altissima (Liliaceae) growing on rising clayey ground
immediately to the east of the New Year’s River in dwarf karroo scrub characterised by Pentzia
incana (Compositae) (Fig. 1).

Evidence of nesting by Dasyproctus was also found in the inflorescence stems of two
species of Gasteria (Liliaceae) and in the stems of Berkheya decurrens (Compositae), all I
growing on clayey soil in various situations within thorn scrub. As the nests examined in these
plants were all old ones from which the wasps had emerged it was not possible to establish the
identity of the builders. It is possible, however, that these were the nests of one or more of the
other species of Dasyproctus .

FLIGHT PERIOD

The flight periods of all five species at Hilton and elsewhere in the vicinity of Grahamstown
fall largely between the beginning of October and the end of March though isolated individuals
have been found both earlier and later in the summer. D. westermanni is known from November
to mid March and was found nesting in Urginea at Hilton during December and January.

FLOWERS AND YOUNG GROWTH VISITED BY ADULT WASPS

At Hilton two species of Dasyproctus were collected on the yellow flowers of Acacia
karroo (Leguminosae): D. b. bipunctatus (6. i. 1 977, 1 female and 13. i. 1977, 1 male) and
D. dubiosus (29.xii. 1976, 2 males and 6. i. 1 977, 1 male). The young foliage of A. karroo was i!
visited by D. westermanni (6.xii. 1976, 1 female).

At Strowan, a farm lying between Hilton and Grahamstown, the yellow flowers of Berkheya
heterophylla (Compositae) were visited by D. b. bipunctatus (12.x. 1972, 4 females, 16.x. 1972, 4
females, and 25.x. 1972, 7 females) and by D. b. simillimus (16.x. 1972, 1 female); at Belmont
Valley, lying 25 kilometres SE. of Hilton, the yellow flowers of Foeniculum vulgar e (Umbel-
liferae) were visited by D. b. bipunctatus (26. i. 1970, 1 female), by D. b. simillimus (26. i. 1970, 1
female, 5.ii. 1970, 1 female, 28.iv.1970, 1 female), by D. dubiosus (20.L1970, 1 male), by
D. immitis (25. i. 1970, 1 female) and by D. ruficaudis (20.i. 1970, 1 female); at the Koonap River

96

ASPECTS OF ETHOLOGY OF DASYPROCTUS WESTERMANNI

| Fig. 1. Hilton, 4.i. 1978. Inflorescences of Urgenea altissima (Liliaceae) in dwarf karroo scrub characterised by Pentzia

incana .

near Adelaide, 70 kilometres N. of Hilton, the greenish-yellow flowers of Zizyphus mucronata
(Rhamnaceae) were visited by D. b. bipunctatus (1 female) and by D. immitis (1 male) (both
20-22. xii. 1972).

IDENTIFICATION OF THE PREY

Prey was obtained only from D. westermanni and as in all species of Dasyproctus consisted
of small flies. It appears that D. westermanni is an opportunist with respect to the flies it utilizes
as prey and may take any suitably sized fly which it finds in its hunting area. All cells in which
| prey flies were in a condition allowing identification were found to have been provisioned with
several species of flies though in some the preponderance of one or other species indicated that
| the wasp may, upon finding a ready supply of that species, have concentrated upon it or upon its
source.

97

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 8, JULY 1980

Seven dipterous families were represented amongst the prey found in nests examined during
January, 1978. Details are listed below.

Simuliidae

Simulium sp.

2,0 mm long

13 females, 4 males

Stratiomyidae

1 sp.

5,2 mm long

10 specimens

Bombyliidae

Bombylius delicatus Wied.

4,0 mm long

1 female, 1 male

1 sp. of Cyrtosiinae

1,4 mm long

3 specimens

1 sp. other

3-4 mm long

1 female, 1 male

Empididae

1 sp.

3,0 mm long

1 male

Syrphidae

1 sp.

3,6 mm long

1 specimen

Otitidae

species A

2,2 mm long

3 females, 10 males

species B

2,4 mm long

1 male

Chamaemyiidae

1 sp.

3,4 mm long

3 specimens

The presence amongst the prey of flies such as the Simuliidae and Stratiomyidae and
possibly also the Otitidae may indicate that the chief hunting area of D. westermanni was in
fairly close proximity to the nests, namely in the riverine vegetation fringing the New Year’s
River. Other flies such as the less commonly taken Bombyliidae were probably captured in the i
tract of open dwarf scrub nearer the nests and between the latter and the riverine vegetation.

DESCRIPTION OF THE NEST

The completed nest of D. westermanni consists of a circular entrance hole of 4 mm diameter
bitten through the side of the green inflorescence stem of Urginea altissima (Fig. 2) and giving ;
access to an ascending and a descending gallery of 4-4,5 mm bore hollowed out of the pithy
centre of the stem. Both galleries are divided serially into a number of cells (Figs 3a, b, c and d).

A total of twenty-two nests of D. westermanni within nineteen subvertical stems was
examined. Seventeen of the utilized stems each contained a single nest, one stem contained two
nests and another contained three nests. The entrance hole to the nest was situated from
500-1 070 mm (average 770 mm) above the ground and in all but one instance was sited below
the level of the lowermost elements of the cylindrical raceme which occupies the terminal
two-fifths of the inflorescence stem. Heights above the ground of the bottom and the top of the
raceme ranged from 820-1 360 mm (average 1 040) and from 1 550-2 100 mm (average 1 800
mm) respectively.

Of the twenty-two nests, twenty had completed ascending galleries and of this latter number
nine had in addition completed descending galleries. The ascending galleries ranged in length
from 42-114 mm (average 86 mm) and the descending galleries ranged in length from 78-122
mm (average 91 mm).

In length the serially arranged cells including the pithy plug sealing each ranged from 8-14
mm (average of 71: 11,2 mm), the thickness of the plugs or cell partitions being 1, 5-2,0 mm.
The number of cells in fully utilized ascending galleries ranged from five to eight; the only fully
utilized descending gallery found contained seven cells. The maximum number of cells found in
any one nest (the only one which had both galleries fully utilized and intact) was fourteen.

98

ASPECTS OF ETHOLOGY OF DASYPROCTUS WESTERMANNI

Fig. 2. Hilton, 4. i. 1978. Portion of inflorescence stem of Urgenea altissima (Liliaceae) showing entrance hole of nest of

Dasyproctus westermanni (x circa 0,5).

Neither gallery is filled with cells right up to the level of the entrance hole, the plug of the
outermost cell in each case being some distance removed from it. In the above fourteen-celled
completely utilized nest a vestibular space, 16 mm in length, was left between the two outermost
cells. The nest entrance opening into this vestibule was not sealed.

METHOD OF CONSTRUCTION OF THE NEST, OVIPOSITION AND PROVISIONING

A nest entrance hole having been bitten through the wall of the inflorescence stem and the
pith centre having been reached, the wasp hollows out its galleries by biting off and removing the
pith. Some at least of this pithy material appears, however, to be retained within the nest for the
purpose of constructing the cell partitions.

99

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 8, JULY 1980

The ascending gallery is invariably the first to be hollowed out and only after the full

number of cells destined for this gallery has been constructed within it is a start made with the
hollowing out of the descending gallery (Figs 3a, b, c and d).

Each cell is provisioned with a relatively large number of prey which are closely packed and

are orientated to face the inner (i.e. blind) end of the cell. The prey is alive but partially

paralysed. Two fully provisioned and sealed cells, the contents of which had neither been
consumed nor had become mouldy, contained fourteen and twenty-six flies respectively.

It is not clear on which prey, in order of provisioning, oviposition takes place but it appears
not to be on the first. A D. westermanni egg was found in each of the above two cells — in the
first it was attached to a Bombylius delicatus Wied. (4 mm long male), one of the first three prey

VJ

If

TZZZZ2

50mm

^2222=

Pigs 3a, b, c, and d. Plans of nests of Dasyproctus westermanni in inflorescence stems of Urginea altissima in

longitudinal section.

100

ASPECTS OF ETHOLOGY OF DASYPROCTUS WESTERMANNI

to be introduced into the cell; in the second it was attached to an Otitid (2,4 mm long male)
situated about in the middle of the cell. Two partially provisioned and still open cells containing
three and ten prey respectively were found in which no egg had yet been laid.

The egg of D. westermanni is white, very strongly curved and 1,9 mm in length. It is glued
to the underside of the prothorax of the fly anterior to the legs — that is more or less in the neck
region, the fly’s head being pushed forwards to accommodate the anterior end of the egg in the
space between the prothorax and the underside of the head. The egg extends laterally or
postero-laterally across the ventral surface of the fly to either left or right. The newly hatched
larva is orientated in the same way (Fig. 4).

Fig. 4. Diagrammatic representation of bombyliid fly showing position of newly hatched feeding larva of Dasyproctus

westermanni .

LIFE HISTORY

Due to the fact that in the nests examined all but two of the completed cells contained either
provision and wasp young that had gone mouldy or cocoons containing spinning larvae or pupae,
little information concerning the life history of the species could be obtained. The two eggs found
on prey on 4. i. 1978 and lO.i. 1978 hatched on 5. i. 1978 and 14. i. 1978 respectively. Neither larva
survived.

All twenty-six male and thirteen female wasps reared from the examined nests emerged
during the period 10-27. i. 1978. Development from egg to adult appears to be very rapid in the
generation reared in Urginea stems, probably owing to the short time during which these stems
are suitable for the nesting of Dasyproctus . It is probable that there is at least one other annual
generation of the species but its nesting site has not been identified.

The silken cocoon of D. westermanni is brown and papery, 9 mm long and with a maximum
width of 3,4 mm. Its anterior end is rounded whilst the posterior end has incorporated in it the

101

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 8, JULY 1980

dark brown meconium voided by the post-spinning larva. Adhering to the meconium at the hind
end of the cocoon are the prey remains — wings, legs and other uneaten fragments which the
mature larva on cleaning the cell prior to the start of cocoon spinning has concentrated behind it
at one end of the cell.

The orientation of the cocoons and of the pupae within them is unusual and of interest in
D. westermanni in that it appears to be governed by gravity rather than by the position of the nest
opening as indicated to the spinning larva by the curvature of the cell partitions. Thus in all cells,
whether constructed in ascending or descending galleries, the anterior end of each cocoon is
orientated upwards, facing away from the ground. Thus in the first-constructed ascending gallery
the anterior end of each cocoon is directed away from the nest entrance and the meconium and
prey remains are between each wasp and the outer partition of its cell. In the later-constructed
descending gallery, however, the orientation normally practised by wasps nesting serially in tubes
is present — the anterior end of each cocoon is directed towards the nest entrance and the
meconium and prey remains are between each wasp and the inner or blind end of its cell.

PARASITES AND OTHER ASSOCIATED INSECTS; MOULD

Three distinct categories of associated insects occur in D. westermanni nests: a parasitoid of
D. westermanni itself, a species cleptoparasitic with respect to the stored provision, and at least
one species possibly cleptoparasitic with respect to the nest galleries.

A 4,2 mm long dark metallic green and reddish bronze Perilampus sp. (Hymenoptera:
Chalcidoidea: Perilampidae) was found to be a parasitoid in nine cells distributed over five nests
studied during January, 1978. Previously, during the summer of 1974 two specimens of the
parasite were reared from the same host. The D. westermanni larva is attacked after it has spun
its cocoon in the normal manner but before it has changed into a pupa. Each parasitized
D. westermanni cocoon yields a single perilampid. The emergence of the adult parasitoid from
the host cocoon is timed to coincide within a few days with the emergence of adult
D. westermanni from adjacent cells in the nest.

A 2,6 mm long yellowish-brown species of Phoridae (Diptera) was found to be a
cleptoparasite in one nest. When opened on 18. i. 1978 it was found that in both the ascending and
descending galleries all the cell partitions, bar those sealing the two outermost cells (i.e. those
closest to the nest entrance), had been broken down and that the provision stored in the breached
cells had been almost completely devoured. In the ascending (older) gallery the cleptoparasites
had already pupated, thirteen puparia being cemented in a group to the gallery wall at the end
nearer the nest entrance. In the descending (newer) gallery were twenty-four as yet unpupated
maggots. These were transferred to a glass vial where they pupated on 20. i. 1978. Flies from both li
galleries emerged from their puparia towards the end of February.

A species of Trypoxylon (Sphecidae) which utilizes pre-existing cavities for its nesting was
found in three instances to have taken over the galleries excavated by D. westermanni. Though it i|
cannot with certainty be stated that actual competition for the excavated galleries took place
between Trypoxylon sp. and D. westermanni it does appear that this may have been the case and ill
that the galleries taken over by the former were ones being used by the builder, not abandoned
ones. Thus all three D. westermanni galleries (which were ascending ones) were either in the
process of being hollowed out or of being provisioned. One of the galleries had not yet reached
its final length as evidenced by the fact that its blind end had not yet been rounded off as is the
case with completed galleries; the second gallery had been completed but nesting by
D. westermanni had not yet commenced; the third gallery had had three cells completed by
D. westermanni , the point of succession being during the provisioning of the fourth cell.

Other insects found in D. westermanni galleries were females of Heriades ?spiniscutis
(Cameron) (Megachilidae) (three instances) and a female of Allodapula variegata (Smith)
(Anthophoridae) (one instance). The D. westermanni galleries used by these bees were clearly

ASPECTS OF ETHOLOGY OF DASYPROCTUS WESTERMANNI

abandoned ones, no recently provisioned cells being present. It is possible that the bees were
merely sheltering in the galleries but it cannot be ruled out that they might have commenced
nesting had they been left undisturbed. If they had, they would clearly not have been competitors
for the galleries as it appears that Trypoxylon sp. may have been.

The previously mentioned old nesting galleries of Dasyproctus sp. in Gasteria inflorescence
stems were found to have been utilized for nesting by Trypoxylon sp., Heriades ?spiniscutis
(Cameron) and Allodapula variegata (Smith) whereas the galleries in Berkheya decurrens stems
had been utilized by Heriades ?spiniscutis (Cameron) and A llodape sp. ( rufogastra Lep. & Serv.
or exoloma Strand) (Anthophoridae).

Mould attacking the stored provision is of common occurrence in the cells of
D. westermanni and effects a larger number of cells than do parasitoids and predators together.
Details are given in Table 1 of the nesting success and mortality rate due to various causes in one
hundred completed cells examined.

Table 1.

Nesting success and mortality rate due to various causes in one hundred completed cells of D. westermanni examined.

Total
No. of
completed
cells*

No. of
cells
attacked
by mould

No. of
cells
attacked
by Peri-
lampidae

No. of
cells
attacked
by

Phoridae

No. of
cells
attacked
by other
insects

No. of
cells in
which
pupae died
for un-
known
reasons

No. of cells
from which
D. westermanni
adults
emerged

100

30

9

13

1

8

39

♦Excluding two cells containing newly hatched larvae which did not survive examination.

REVIEW AND DISCUSSION OF THE ETHOLOGY OF THE GENUS DASYPROCTUS

As already noted, published biological accounts varying from fragmentary to fairly exten-
sive pertain to six species: D. agilis (F. Smith) and D. buddha (Cameron) from the Oriental
Region and D. barkeri (Arnold), D. bipunctatus Lep. & Brulle, D. kibonotensis Cameron and
D. stevenson i (Arnold) from the Afrotropical Region.

D. agilis and D. buddha were both recorded nesting in stems of Gramineae (sorghum and
solid dead stems of Coelorrachis respectively) in which their cells were provisioned with
Otitidae (mainly), Muscidae and Syrphidae and with Chloropidae and Otitidae respectively (see
Bohart and Menke, 1976: 419).

Of the African species, D. bipunctatus Lep & Brulle (including its colour forms simillimus
(Smith) and lichtenburgensis (Arnold)) is the best known. The species has been found by several
observers to be restricted in its nesting to the inflorescence stems of the monocotyledonous
families Amaryllidaceae, Iridaeceae and Liliaceae. The most comprehensive account of the
nesting of this species ( D . bipunctatus sensu stricto) is that by Bowden (1964: 425-437) who
found it nesting near Kampala (Uganda) in the flowering stems of Kniphofia and Aloe spp.
(Liliaceae), Gladiolus sp. (Iridaceae) and Hippeastrum (an American genus) (Amaryllidaceae).
Six families of Diptera were represented amongst the prey, namely Chloropidae, Lonchaeidae,
Muscidae, Simuliidae, Sphaeroceridae and Trypetidae. Most prey were Atherigona species
(Muscidae: Coenosiinae). On the basis of the prey taken Bowden theorized that D. bipunctatus
forages primarily over grass. D. bipunctatus was also found nesting in gladiolus stems near

103

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 8, JULY 1980

Salisbury (Rhodesia, now Zimbabwe) by Cuthbertson (1937: 28-31). That author listed the prey
as belonging to the Anthomyiidae, Sarcophagidae, Sapromyzidae, Syrphidae and Tachinidae.

D. bipunctatus simillimus, in the Western Cape Province at least, appears to nest by
preference in the green flowering stems of Watsonici sp. (Iridaceae). The present author found it
nesting commonly in such stems at Kirstenbosch (Cape Peninsula) on 22. xi. 1964 and A. J. Hesse
of the South African Museum (unpublished notes) recorded its prey found in Watsonia stems at
an unnamed locality as Otitidae.

It seems that the “watsonia wasp” described by Skaife (1953: 338) is referrable to
D. bipunctatus simillimus (Smith) and that the name, Dasyproctus capensis Skaife, given it in the
above publication may be considered a hitherto overlooked and therefore new synonym of the
former. In the new (1979) edition of Skaife’s book the revisers have come to the same conclusion
concerning the identity of the wasp in question and have substituted the name Dasyproctus
bipunctatus for Dasyproctus capensis.

D. bipunctatus lichtenburgensis at Tanga on the East African coast was found to provision
its cells in stems (unspecified) almost entirely with Otitidae (= Ortalidae) and to be subject to
parasitization by a species of Mutillidae (Carpenter, 1942: 48).

In contrast with the above considered species with respect to the stems utilized for nesting ii
are D. barkeri, D. kibonotensis and D. stevensoni all of which utilize dicotyledonous plants.

D. barkeri, described from Durban, Umgeni and Malvern in Natal, was recorded by Arnold ll
(1927: 127) as “nesting in dry stems of the Kaffir-boom” ( Erythrina sp.) (Leguminosae).

D. kibonotensis, studied as was D. bipunctatus at Kampala (Uganda), was found by
Bowden (1964: 425-437) to be restricted to the stems of Rubus sp. (Rosaceae) for the excavation 1
of its nests. The prey encompassed five families, namely Lonchaeidae, Muscidae, Platystomidae,
Tachinidae and Trypetidae. Most prey were Trirhithrum coffeae Bezzi (Trypetidae), a dominant j
of the dipterous fauna of robusta coffee. On the basis of the prey taken Bowden theorized that
this species forages in understory shrubs. Recorded parasites were Miltogramminae.

D. stevensoni in West Cameroon was found nesting in a stem of Conyza bonariensis j
(Compositae) in which the cells were provisioned with Milichiidae (Michener, 1971: 407).

As far as can be established from a consideration of so relatively few species, the nesting of
Dasyproctus appears to present great uniformity in its basic features and the nesting of
D. westermanni at Hilton here described appears, with the possible exception of pupa orientation tl
to be dealt with below, to be very similar to the general pattern.

All the species nest up off the ground in galleries which they themselves excavate in the pith li
of usually live green plant stems. The serially arranged cells are separated by pith partitions and ij:
are provisioned with numerous small adult Diptera.

In D. bipunctatus , D. kibonotensis and D. westermanni the nest entrance is on the side of
the stem and the nest galleries proceed in both directions through the stem, the gallery directed i|
towards the stem apex (that is the ascending gallery) being completed first in all three species.
Details of the nest plans of the two Oriental species and of D. barkeri are not available. With
respect to D. stevensoni it is stated that the only nest examined was damaged in its collection but
that apparently the entrance was at the broken end of the stem, rather than in the side of the stem.
Confirmation of this apparently unusual nest form is required.

It is in the orientation, within the galleries, of the cocoons and of the pupae within them that
D. westermanni differs from the other two species for which the orientation has been recorded.
Thus Bowden (1964: 429 and 430) recorded both D. bipunctatus and D. kibonotensis larvae as
pupating with their heads towards the entrance hole of the nest. In D. westermanni , as already
described, all pupae are orientated with their heads away from the ground and towards the apex
of the flower stalk, so that only those larvae in the descending gallery have their heads towards
the nest entrance.

Using the trends shown by the reviewed species some speculation is possible concerning the

104

ASPECTS OF ETHOLOGY OF DASYPROCTUS WESTERMANNI

identity of the builders of the old Dasyproctus nests found at Hilton in the inflorescence stems of
Gasteria spp. (Liliaceae) and in the stems of Berkheya decurrens (Compositae). All the nests
concerned had side entrances and both ascending and descending galleries. In the case of the
nests in Gasteria it is not impossible that these were the work of D. westermanni but the fact that
these plants were in a vegetation type different from that in which occurred the Urginea
frequently utilized by D. westermanni may indicate the involvement of a different species. The
nature of the inflorescence stalk utilized may indicate that D. bipunctatus might have been the
species concerned. In the case of the nests in Berkheya decurrens it is probably correct to
consider that they were the work of neither D. westermanni nor D. bipunctatus as all records
indicate that these two species are restricted to monocotyledonous inflorescence stalks and that no
species nests in both monocotyledonous and dicotyledonous plants. The nests in Berkheya are
thus probably those of one of the other three Dasyproctus species recorded from Hilton, namely
D. dubiosus, D. immitis and D. ruficaudis.

It is clearly of survival value to all species of Dasyproctus if the herbaceous plant stems in
which they make their nests are protected in some manner or other from being eaten by large
herbivores. With respect to those plants found to be utilized by Dasyproctus at Hilton it can be
shown that all are so protected at least from the herbivores now common there, namely cattle,
sheep and goats. Thus Urginea altissima, which grows freely exposed in the dwarf karroo scrub
and which is very conspicuous when flowering on account of the long inflorescence stalks, is
avoided by the above animals. This is undoubtedly due to the plant’s toxicity to stock resulting
from the presence in all parts of the plant of a glucoside having a digitalis action (see Watt &
Breyer-Brandwijk, 1962: 717-718). It has furthermore been observed by the author that Urginea
altissima is also left untouched under the more natural conditions of the Addo National Elephant
j Park where the herbivores in the portion of the park examined are buffalo, eland, hartebeest and a
variety of smaller buck.

The Gasteria spp. at Hilton appear to be innocuous to herbivores, no indication to the
contrary being found in Watt & Breyer-Brandwijk. However, these plants are found at Hilton
only in the shelter of thorny shrubs such as Acacia karroo where they are thus physically
protected from stock. Berkheya decurrens on the other hand grows in the open, often on
disturbed ground, but is well protected from herbivores by the spinose projections of the leaves
and involucral bracts.

Stems utilized elsewhere by Dasyproctus species may at least in some instances also be
shown to be immune from destruction by herbivores. Thus the trailing stems of Rubus sp. (used
by D. kibonotensis) are undoubtedly protected by the sharp prickles with which they are armed,
j With respect to Watsonia spp. (used by D. bipunctatus ) it has been observed near Grahamstown
that buck readily eat the actual flowers but not the inflorescence stalks (Jacot-Guillarmod, pers.
i com. ).

It is apparent therefore that there is no consistency in the nature of the protection offered by
j the plants in which Dasyproctus nests as it may be due to their physiological or morphological
nature or to their protected habitat.

ACKNOWLEDGEMENTS

The author wishes to thank Mr T. C. White of the farm Hilton for his much appreciated
kindness over the years in allowing him free access to his land. Thanks are due also to Dr G. L.
Prinsloo of the Plant Protection Research Institute, Pretoria for his help with respect to the
identity of the Perilampidae. Gratitude is expressed to the C.S.I.R. for a running expenses grant
for field work during the course of which the present observations were made.

105

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 8, JULY 1980

REFERENCES

Arnold, G. 1927. The Sphegidae of South Africa. Part 8. Ann. Transv. Mus. 12 (2): 55-131.

Bohart, R. M. and Menke, A. S. 1976. Sphecid wasps of the world: a generic revision. Berkeley: University of California
Press.

Bowden, J. 1964. Notes on the biology of two species of Dasyproctus Lep. and Br. in Uganda (Hymenoptera:
Sphegidae). J. ent. Soc. sth. Afr. 26 (2): 425-435.

Carpenter, G. D. H. 1942. Note on the bionomics of the sphegid wasp Dasyproctus bipunctatus Lepeletier (Hym.). Proc.
R. ent. Soc. Lond. (A) 17 (4—6): 48.

Cuthbertson, A. 1937. Biological notes on some diptera in Southern Rhodesia. Proc. Trans. Rhod. scient. Ass. 35 (1):
16-34.

Leclercq, J. 1958. Hymenoptera Sphecoidea (Sphecidae II. Subfam. Crabroninae). Explor. Parc. natn. Upemba Miss.
G. F. de Witte Fascicule 45: 1-114.

Michener, C. D. 1971. Notes on crabronine wasp nests. J. Kans. ent. Soc. 44 (3): 405-407.

Skaife, S. H. 1953. African insect life. Cape Town: Longmans Green & Co.

Skaife, S. H. 1979. African insect life. (2nd ed. revised by J. Ledger and A. Bannister.) Cape Town: Struik.

Watt, J. M. and Breyer-Brandwijk, M. G. 1962. The medicinal and poisonous plants of southern and eastern Africa.
Edinburgh: Livingstone.

Manuscript accepted for publication 18 April 1980.

Printed by Cape & Transvaal Printers (Pty) Ltd, Cape Town

BD8966

/? & / 3

NH

ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

Ann. Cape Prov. Mus . (nat. Hist.)

VOLUME 13 • PART 9
8th OCTOBER 1980

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

Printed by Cape & Transvaal Printers (Pty) Ltd, Cape Town

BD9014

The distribution and status of some eagles in the Cape Province

by

A. F. BOSHOFF

(The Lakes Nature Conservation Station, Private Bag X6546, George, 6530)

and

C. J. VERNON

(East London Museum, 319 Oxford Street, East London, 5201)

CONTENTS

Page

Uittreksel 107

Abstract 108

Introduction 108

Survey methods 108

Evaluation of the information obtained from the survey 110

Distribution and abundance of the eagles

Black Eagle 112

Crowned Eagle 117

Martial Eagle 120

Fish Eagle 123

Tawny Eagle 127

Bateleur 129

Long-crested Eagle 129

Black-breasted Snake Eagle and Bearded Vulture 130

Discussion 130

Conclusion 131

Acknowledgements 131

References 132

UITTREKSEL

’n Studie van die verspreiding en digtheid van verskeie arend spesies in die Kaapprovinsie,
gebaseer op ’n posopname gevolg deur ’n veldwaarheidsopname (“ground-truth survey”), word
beskryf. Ondanks sekere tekortkominge verbonde aan so ’n opriame, en die groot gebied wat
gedek is, was dit moontlik om ’n geldige prent van die verspreiding van die betrokke spesies, so
wel as realistiese beramings van hulle digtheid, te verkry. Dit was nie moontlik om verklarings te
kon maak nie oor enige neigings wat miskien teenwoordig is, maar die opname het getoon dat,
van die nege arend spesies wat in die Kaapprovinsie broei, is daar vyf wat blykbaar lewensvat-

107

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 9, OCTOBER 1980

bare, alhoewel klein, bevolkings het terwyl ’n sesde spesies moontlik besig is om tot uitroeiing te
daal.

ABSTRACT

A study of the distribution and abundance of some eagles in the Cape Province, based on a
postal survey followed by a ground-truth survey, is described. Despite certain disadvantages
attached to such a survey, and the vast area covered, it was possible to obtain a valid picture of
the distribution of the species involved, as well as realistic estimates of their abundance. It was
not possible to make statements on any trends that may exist but the survey indicated that, of the
nine species of eagles that are known to breed in the Cape Province, five appear to have viable,
albeit small, populations while a sixth species may be declining to extirpation.

INTRODUCTION

Recently raptors have become a source of international concern as they are at the top of
various food chains and consequently are important environmental indicators. The status of
eagles in the Cape Province is given in general terms in Skead (1967), Winterbottom (1968) and
McLachlan & Liversidge (1978). The factual basis for the statements by these authors is meagre,
primarily because of the paucity of observers in a vast province. Virtually nothing is known of
the status of eagles in the Cape Province and most of the work concerning eagles has been in the
form of random roadside counts (Rudebeck 1963; Rowan 1964; Siegfried 1966, 1968; Cade
1969; Maclean 1970; Dean 1975a).

Despite the lack of precise information it is common knowledge that some eagles have
declined in numbers and range and five species occurring in the Cape Province are considered to
be vulnerable to extirpation. The decline has been attributed to decreasing food supplies, loss of
habitat due to agricultural and urban development and to direct persecution (Siegfried 1963,
1968; Skead 1967; Heyl 1972; Winterbottom 1972; Siegfried et al. 1976; Vernon 1978; Boshoff
& Vernon in prep.). For many years farmers have persecuted eagles for taking their stock. At one
period, bounties were paid on eagles in the Cape Province, and there are records of nearly 5 000 I
bounties paid between 1930 and 1955 (Boshoff & Vernon in prep.). For example, over a seven
year period, 1947 to 1953, an average of 74 bounties was paid per year in the Bedford district. In
the Laingsburg district 150 Black Eagles were killed in one year (Siegfried 1963). The effects of r
this persecution are not known but the present authors suspect that in some areas it has caused the
extirpation of the eagles.

In 1976 a survey was initiated by the Cape Department of Nature and Environmental
Conservation (CDNEC) to determine the distribution and abundance of some of the eagles in the
Cape Province. Eight species were included in the survey namely Black Eagle Aquila verreauxi,
Crowned Eagle Spizaetus coronatus. Martial Eagle Polemaetus bellicosus, Tawny Eagle Aquila
rapax. Long-crested Eagle Lophaetus occipitalis. Fish Eagle Haliaeetus vocifer, Bateleur
Terathopius ecaudatus and Black-breasted Snake Eagle Circaetus pectoralis. Three species, two
of which are rare and are not known to breed in the Cape Province, were omitted.

An additional aim of the 1976 Survey was to obtain information on alleged stock predation by
eagles (Boshoff & Vernon in prep.).

SURVEY METHODS

In 1976 a questionnaire was sent by the Department of Nature and Environmental Conserva-
tion of the Cape Province (CDNEC) to all farmers and landowners in the province. The
questionnaire requested information about the presence, breeding and feeding of certain eagles
and vultures. It was accompanied by a coloured illustration depicting those eagles and vultures,

108

BOSHOFF & VERNON: DISTRIBUTION AND STATUS OF SOME EAGLES

together with a covering letter explaining the aims of the survey, and a stamped and addressed
reply envelope. The information gleaned from the returned questionnaires was placed on
computer and printouts of this information were made available to members of the CDNEC for
the ensuing ground-truth survey. This survey was considered necessary in order to evaluate the
results of the postal survey and farmers/landowners throughout the province were visited. The
detailed objectives were to establish whether the species that the farmers had reported actually
occurred or bred in the recorded area. Due to logistical problems it was not possible to conduct
the ground-truth survey on a totally systematic basis and farms to be visited were chosen on a
random basis according to proximity to the route followed. Attempts were made to include all the
major physiographic regions of the Cape Province. For convenience of discussion the province
was divided into six zones namely South-Western Cape (S.W. Cape), Southern Cape (S. Cape),
Karoo, North-Western Cape (N.W. Cape), Northern Cape (N. Cape) and Eastern Cape (E. Cape).
The boundaries of these zones are shown in Fig. 1.

Fig. 1. Regions of the Cape Province and number of questionnaires returned per locus in the CDNEC 1976 Postal Survey.

109

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 9, OCTOBER 1980

The records of presence, which include breeding reports, of the various species of eagles and
vultures were plotted according to the total number of returns per query per locus (locus = one
quarter-degree square; approximately 630 sq. km). The information was arranged in four ways:
presence and breeding reports or breeding reports alone, with records uncorrected or corrected. In
order to compensate for misidentifications it was arbitrarily decided to reject all records of only
one report per locus; this represents the “corrected” state. The authors decided that a record was
most likely to be incorrect if only one farmer in a locus reported the species, taking into
consideration that this may have resulted in a bias against particularly astute observers. This
factor may also have favoured loci from which were received a large number of replies to the
1976 questionnaire as opposed to those loci from which there were few replies. Thus all maps
reflecting the results of the 1976 Survey represent only the corrected data. The results of this
survey were compared with all records that could be gathered from other sources. These include
the observations of the members of the ground-truth survey and all published and unpublished
records from 1970 onwards.

The abundance of four species of eagles was estimated from the breeding reports of the 1976
Survey. The actual number of reports was multiplied by the proportion of those reports that were
investigated by the ground-truth survey and found to be correct. This figure was then multiplied li
by the reciprocal of the percentage poll to give an estimate of the total number of breeding pairs
of each species in the Cape. The confidence limits, at the 95% level, based on the proportion of
correct reports, were taken from tables in Diem & Lentner (1970).

Comparative estimates of abundance of the four eagle species were made from the ground-
truth survey. These were based upon the tabulation of nests seen during the 1976 Survey, or
found independently of the farmers, together with estimates of the numbers of potential pairs in
the areas of the Cape Province not visited by members of the ground-truth survey.

The above methods of assessing abundance were made for the Cape as a whole and for each i
degree square (16 loci — approximately 10 000 sq. km). In the E. Cape, where several degree
squares were thoroughly covered by the ground-truth survey, it was possible to make detailed
comparisons between the results of the 1976 Survey and the ground-truth survey.

j

EVALUATION OF THE INFORMATION OBTAINED FROM THE SURVEY

Distribution of farmers

A total of 30 487 farmers/landowners was sent questionnaires of which 14 761 were com-
pleted and returned to give a poll of 48,4%. The distribution of farmers replying to the 1976
Survey is shown in Fig. 1. This map reflects the actual distribution of the farmers as the blank
areas represent either non-farming areas, regions where the density of farmers per locus is low, or
African states/homelands which were not covered by the survey. The average number of farmers
replying to the 1976 Survey per locus was about 12, but in some loci there were none and in the
intensively farmed areas there were frequently more than 50. The density was lowest in the
Karoo and N.W. Cape and greatest in the S.W. Cape and S. Cape.

Identification of the eagles

Although the farmers, not unexpectedly, had problems with the identification of the various
eagles and vultures, the authors consider that, for persons not specifically interested in bird
watching, they were very competent. Problems arose through the variety of vernacular names, the
difficulties of identifying species and the pictures accompanying the questionnaires. The vernacu-
lar names caused considerable confusion, especially those linking the Black Eagle and the
Bateleur, the Black Eagle and the Bearded Vulture Gypaetus barbatus, and the Martial Eagle and
the Bearded Vulture. This problem is discussed in Boshoff (1979).

110

BOSHOFF & VERNON: DISTRIBUTION AND STATUS OF SOME EAGLES

The differences between the adult and immature plumages of the eagles also provided a source
of confusion. Only adult birds were illustrated in the questionnaire, and some farmers identified
immatures of commoner eagles such as the Black Eagle and Martial Eagle as the Bearded
Vulture, Black-Breasted Snake-Eagle or Tawny Eagle. As a result in this paper the Bearded
Vulture data have been included with that for the eagles. The inclusion in the illustrations of the
Bateleur but the omission of the Jackal Buzzard Buteo rufofuscus meant that farmers reported the
Bateleur instead of the more common Jackal Buzzard. This was compounded by the fact that in
some areas the buzzard is called the Bateleur.

There were consistent patterns of misidentification. The common species were more frequently
correctly identified than the rarer species. The farmers’ identifications and the likely species they
were reporting are shown in Table 1. There was no significant difference between the rate of
correct identification of the Black, Crowned, Martial and Fish Eagles, and a common correction
factor is used for these species.

Table 1

Identification of the eagles and the Bearded Vulture in the CDNEC 1976 Survey

Species identification by the farmers

Likely identification*

BE.

C.E.

T.E.

L.E.

F.E.

BAT.

ME.

S.E.

B.V.

Black Eagle (B.E.)

107

26

22

Crowned Eagle (C.E.)
Tawny Eagle (T.E.)
Long-crested Eagle (L.E. )

1

27

14

3

1

1

Fish Eagle (F.E.)
Bateleur (BAT.)

22

6

Martial Eagle (M.E.)

2

2

7

2

47

2

34

Snake Eagle (S.E.)
Bearded Vulture (B.V.)

2

Jackal Buzzard

1

2

Unknown

25

3

19

2

4

18

21

7

12

Number of nest reports
investigated

135

32

41

5

26

55

69

9

70

% correct

79

84

33

60

85

11

67

0

3

Upper c.l. (95%)

86

95

50

95

96

22

78

34

12

Lower c.l. (95%)

71

67

20

15

65

4

55

0

1

♦According to the ground-truth survey.

Reports of Presence and Breeding

The ground-truth survey proved that it was not possible to devise a method of assessing the
accuracy of reports of the presence of any eagle species on individual farms. Thus while
distribution maps based on these reports are given, no proper estimate can be made of their
accuracy, other than the “percentage correct” data given in Table 1. The authors consider that
both the presence reports and the breeding reports suffered from the above inaccuracies, but that
there was a greater degree of inaccuracy in the presence reports than in the breeding reports.
However, the results of the 1976 Survey pertaining to the Cape Vulture Gyps copr other es
indicated that the farmers reported the presence of the vultures if they had seen them on the farm
at some time during their lifetime (Boshoff & Vemon 1980). Thus any distribution map based on
presence records may reflect the pattern over the last generation; say twenty-five years.

In contrast to the presence reports, the breeding reports are considered to reflect the present

111

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 9, OCTOBER 1980

day situation. In addition the members of the ground-truth survey were able to assess the
accuracy of breeding reports as a nest is a static object. Three classes of accuracy were
established: correct, possibly correct and incorrect. The “possibly correct” records are those
where the farmer could not locate the nest reported but the member of the ground-truth survey
considered that there could easily have been a nest on the farm. Only the “correct” reports are
used in the estimates of the abundance of individual species of eagles.

i

DISTRIBUTION AND ABUNDANCE OF THE EAGLES

Black Eagle

Distribution

The 1976 Survey indicated that the Black Eagle is widespread in the Cape Province. The
distribution of farmers reporting the species on their farms is shown in Fig. 2 and indicates that

Fig. 2. The distribution of farmers reporting the presence of Black Eagles on their farms in the CDNEC 1976 Survey.

Circled digits represent totals per locus.

112

BOSHOFF & VERNON: DISTRIBUTION AND STATUS OF SOME EAGLES

the Black Eagle is widespread throughout the S.W., S., and E. Cape but is localised in the other
three regions. The pattern of distribution of farmers reporting that Black Eagles bred on their
farms is shown in Fig. 3. This map shows the same pattern as Fig. 2 but intensifies the trends to
suggest that even in the southern regions the species may be absent from some areas. Fig. 3 tends
to reflect the distribution of mountainous and hilly terrain, which constitutes the known habitat
preference of this eagle, and compares favourably with the distribution of Black Eagle sightings
obtained from other sources for the period 1970 to 1979 (Fig. 4). It should be mentioned that the
concentration of reports from portions of degree squares 28 22, 29 22 and 29 23 (Fig. 3) was not
investigated during the ground-truth survey ( vide Fig. 4).

Fig. 3. The distribution of farmers reporting breeding by Black Eagles on their farms in the CDNEC 1976 Survey. Circled

digits represent totals per locus.

113

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 9, OCTOBER 1980

114

BOSHOFF & VERNON: DISTRIBUTION AND STATUS OF SOME EAGLES

The authors consider that the 1976 Survey provides an accurate picture of the distribution of
the Black Eagle in the Cape Province.

Abundance

From the 1976 Survey it would appear that the Black Eagle is the most abundant eagle species
in the Cape Province. There are an estimated 487 breeding pairs, and the potential range of this
estimate is from 280 to 730 pairs (Table 2). The estimate of 487 pairs is considered to be
reasonable as the ground-truth survey gave an estimate of a minimum of 400 pairs. These figures
suggest that in the Cape Province as a whole the Black Eagle is about seven times more abundant
than any other eagle.

Table 2

Estimates of the numbers of breeding pairs of four species of eagles in the Cape Province, based
on the results of the CDNEC 1976 Survey and the ground-truth survey

Eagle species

Black

Crowned

Martial

Fish

Total Farmers — 30 487

% replying — 48,4

Number reporting eagles breeding

1 766

83

351

236

Number of breeding reports investigated . .
Number of occupied nests found (excluding

135 (7,6%)

32(38,6%)

70(19,9%)

26(1 1,0%)

“possibly occupied”)

18

10

5

3

% investigated that were correct

95% confidence limits

13

31

7

12

— upper

20

50

16

30

— lower

8

16

2

2

Estimated number of breeding pairs*

95% c.l. on estimate

487

54

52

56

— upper

730

86

116

146

— lower

280

28

17

12

Estimated number of breeding pairs extra-
polated from ground-truth survey data ....

400-500

50

?

45

*Calculation of estimated number of pairs based on:

e.g. Black Eagle x

KM v
48,4

1 766
1

The breeding density of the Black Eagle varies throughout the Cape Province (Fig. 5) and the
birds are most abundant in parts of the N.W. Cape and the E. Cape. The greatest density is found
in the degree square 32 26, which is the area between Adelaide, Bedford, Cradock and Tarkastad.
Within this area V. L. Pringle reported 14 eyries of the Black Eagle in the area of one locus
(circa 630 sq. km).

115

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 9, OCTOBER 1980

!

116

BOSHOFF & VERNON: DISTRIBUTION AND STATUS OF SOME EAGLES

Crowned Eagle

Distribution

The 1976 Survey indicated that the Crowned Eagle is mainly found in the S. Cape and the
E. Cape (Fig. 6) although there were isolated reports from all other regions. These records do not
appear in the map of the distribution of breeding Crowned Eagles (Fig. 7). This map compares
favourably with that of Crowned Eagle sightings made by observers other than farmers (Fig. 8)
and the authors consider that the 1976 Survey provides a valid assessment of the distribution of
the Crowned Eagle.

117

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 9, OCTOBER 1980

118

BOSHOFF & VERNON: DISTRIBUTION AND STATUS OF SOME EAGLES

Fig. 8. The distribution of the Crowned Eagle according to published and unpublished records (ex CDNEC 1976 Postal

Survey data) between 1970 and 1979.

Abundance

Estimates from the 1976 Survey suggest that there are 54 pairs of Crowned Eagles in the Cape
Province. The range of this estimate is between 28 and 86 pairs (Table 2). The ground-truth
survey tabulated 50 eyries of the Crowned Eagle but not all of these were checked during the
1970s to see if they were still occupied. It is considered that the number of these that have been
abandoned may be compensated for by eyries that have not been located, and that 50 breeding
pairs is a reasonable estimate.

The density of Crowned Eagles is variable within the E. Cape and the 1976 Survey correctly
identifies the areas of greatest density. Crowned Eagles are known from the ground-truth survey
to be most numerous in the degree squares 32 26, 32 27 and 33 27 indicated in Fig. 7; these are
the Bedford, Komga and Grahamstown districts.

119

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 9, OCTOBER 1980

Martial Eagle

Distribution

The 1976 Survey indicated that the Martial Eagle is widespread in the Cape Province (Fig. 9).
This eagle is found throughout the E. Cape but is more localised in other regions. The extent of
the localisation is emphasised in the map of the breeding distribution of the Martial Eagle (Fig.
10). These results compare favourably with the distribution of Martial Eagle records made by
other observers (Fig. 11).

Fig. 9. The distribution of farmers reporting the presence of Martial Eagles on their farms in the CDNEC 1976 Survey.

Circled digits represent totals per locus.

120

BOSHOFF & VERNON. DISTRIBUTION AND STATUS OF SOME EAGLES

121

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 9, OCTOBER 1980

Fig. 11. The distribution of the Martial Eagle according to published and unpublished records (ex CDNEC 1976 Postal

Survey data) between 1970 and 1979.

A bundance

The Martial Eagle is represented in the Cape Province by an estimated 52 pairs. This is an
average of less than one pair per degree square. The range of the estimate is between 17 and 1 16
pairs (Table 2) and therefore even at the upper estimate there is less than one pair of Martial
Eagles per 5 000 sq. km. If a pair of Martial Eagles has a home range of 400 sq. km., which is
double the estimate for birds in the Kruger National Park (Snelling 1969), there must be
considerable areas of the Cape where there are no resident pairs. It is not possible to provide a
comparative estimate of the abundance of the Martial Eagle in the Cape Province. There were 30
eyries tabulated by the ground-truth survey.

The density of Martial Eagles is greatest in the E. Cape and even within that region they are
localised. The 1976 Survey indicates that the degree squares 32 26 and 32 27 contain a greater
density than other squares; this is the area between Bedford and Komga. The ground-truth survey
also found the species to be most numerous in that area.

122

BOSHOFF & VERNON: DISTRIBUTION AND STATUS OF SOME EAGLES

The Abundance of the Black, Crowned and Martial Eagles, estimated from the 1976 Survey,
can be compared with estimates from the ground-truth survey for the E. Cape south of 32° S. The
totals for all three species for seven degree squares are given in Table 3. It is concluded that, for
this zone at least, the 1976 Survey provided a reasonable estimate of the numbers of breeding
pairs of these three eagles.

Table 3

Estimates of the numbers of breeding pairs of three species of eagle in part of the Eastern Cape,
based on the results of the CDNEC 1976 Survey and the ground-truth survey

Degree

Square

Black Eagle

Crowned Eagle

Martial Eagle

Total

1976*

G-T**

1976

G-T

1976

G-T

1976

G-T

32 25

27

12

1

2

3

4

31

18

32 26

19

17

5

13

8

6

i 32

36

32 27

12

7

3

8

6

6

21

21

32 28

0

2

0

4

0

1

0

7

33 25

10

7

1

2

3

4

14

13

33 26

7

4

5

17

3

4

15

26

33 27

1

0

0

1

0

0

1

1

*1976 Survey.
**Ground-truth survey.

Fish Eagle

Distribution

The 1976 Survey indicated that the Fish Eagle is found along the coast and along the Orange
River and that it avoids the arid western part of the Cape Province (Fig. 12). The breeding
distribution of the Fish Eagle is even more restricted (Fig. 13). The results of the 1976 Survey are
similar to the distribution pattern obtained from other sources (Fig. 14).

The Fish Eagle was the species most frequently correctly identified by the farmers. Accord-
ingly the authors consider that many of the reports of the presence of the Fish Eagle are valid.
Thus Fig. 12 is interpreted to mean that the species wanders widely from its permanent breeding
grounds to areas offering a suitable though temporary habitat.

Abundance

There are an estimated 56 pairs of Fish Eagles in the Cape. The confidence limits of this
estimate range between 12 and 146 pairs (Table 2). A comparable estimate of 45 pairs was made
by the ground-truth survey through tabulation of known eyries and extrapolation to areas not
covered by the ground-truth survey.

123

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 9, OCTOBER 1980

124

BOSHOFF & VERNON: DISTRIBUTION AND STATUS OF SOME EAGLES

Fig. 13. The distribution of farmers reporting breeding by Fish Eagles on their farms in the CDNEC 1976 Survey. Circled

digits represent totals per locus.

125

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 9, OCTOBER 1980

126

BOSHOFF & VERNON: DISTRIBUTION AND STATUS OF SOME EAGLES

Tawny Eagle

Distribution

The reports of the presence of the Tawny Eagle in the Cape Province are mapped in Fig. 15.
This map has little reliability because of the high frequency with which the Tawny Eagle was
misidentified (Table 1). The map of the breeding distribution of the Tawny Eagle (Fig. 16) shows
that it occurs mainly in the N. Cape. Most, if not all of the few reports from the E. Cape are
considered to be misidentifications. The distribution data obtained from other sources (Fig. 17)
confirm the pattern of distribution seen in Fig. 16. The concentration of records from the
Colesberg-Venterstad districts (portions of degree squares 30 24 and 30 25) is due to the

Fig. 15. The distribution of farmers reporting the presence of Tawny Eagles on their farms in the CDNEC 1976 Survey.

Circled digits represent totals per locus.

127

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 9, OCTOBER 1980

Fig. 16. The distribution of farmers reporting breeding by Tawny Eagles on their farms in the CDNEC 1976 Survey.

Circled digits represent totals per locus.

concerted efforts of R. C. Rous to locate eagle nests in these areas. Within this limited zone there
are six breeding records for the past three years. Consequently it is possible that this species
occurs more frequently in the nothem parts of the E. Cape and the north-western Karoo than was
thought. During the past nine years there have been a further six sight records from the E. Cape.

Abundance

No estimate could be made from the 1976 Survey. All the correct identifications of this eagle
came from the N. Cape and the Tawny Eagle may be as numerous there as the Black Eagle or the
Martial Eagle.

128

BOSHOFF & VERNON: DISTRIBUTION AND STATUS OF SOME EAGLES

Fig. 17. The distribution of the Tawny Eagle according to published and unpublished records (ex CDNEC 1976 Postal

Survey data) between 1970 and 1979.

Bateleur

The confusion of this eagle with other species in the 1976 Survey prevents any reasonable
conclusions from being made about the distribution and abundance of the Bateleur. The six
correct identifications confirmed by the ground-truth survey were all made by farmers in the
N. Cape. This was also the only region where Bateleurs were seen by the ground-truth survey.
There are a few isolated sight records of the Bateleur from the Karoo and E. Cape, made during
the 1970s by other observers. The findings of the 1976 Survey are restricted to the statement that
the Bateleur may still breed in the N. Cape but elsewhere in the Cape it is absent or a rare
vagrant.

Long-Crested Eagle

Little of value was detected by the 1976 Survey about the distribution and abundance of the
Long-crested Eagle. The breeding records (not illustrated) are confined to four, two from the

129

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 9, OCTOBER 1980

E. Cape and two from the S.W. Cape. The Long-crested Eagle was not observed by the
ground-truth survey and very few sight records were made during the 1970s. The paucity of
information resulting from the 1976 Survey is a valid indication of the rarity of this species in the
Cape Province.

Black-Breasted Snake Eagle and Bearded Vulture

The 1976 Survey did not reveal any worthwhile information about these two species. Both
species are rare in the Cape Province and most, if not all, reports by the farmers were based upon
misidentifications. The present distribution of the Bearded Vulture is confined to the north-
eastern Cape (Boshoff et al. 1978).

DISCUSSION

The 1976 Survey yielded a valid picture of the status of eagles in the Cape Province. It showed
that there are only four species which are present in any numbers. These are the Black Eagle,
Crowned Eagle, Martial Eagle and Fish Eagle. The inclusion of the Fish Eagle negates the
possibility that the farmers were only reporting the species that affected their stock. As the
farmers have a favourable attitude towards the Fish Eagle they were also reporting in general
about the eagles of which they were aware.

This suggests that the remaining species of eagles that occur in the province are scarce,
unobtrusive or localised. The Tawny Eagle and Long-crested Eagle represent localised species, !
the Booted Eagle Hieraaetus pennatus is unobtrusive and the Bateleur, Black-breasted Snake
Eagle and Long-crested Eagle are rare. The Booted Eagle was not included in the questionnaire, ;
and the ground-truth survey indicated that the farmers were unaware of its presence or identity. |
This is surprising as farmers in general were aware of the small raptors, which were not included
in the questionnaire, such as the Jackal Buzzard, Black-shouldered Kite Elanus caeruleus and
Lanner Falcon Falco biarmicus.

It is unlikely that the same knowledge of the status of eagles in the entire province could have
been obtained in any other way for the same expenditure of time, manpower and finance. A
hidden advantage of the survey was that all farmers were reached and that contact was made with
many genuinely interested in eagles and conservation. Over such a vast area it might otherwise
have been difficult to reach these potential opinion leaders.

The survey presents a picture of the status of eagles in the Cape Province in the 1970s. It is a
static sample and does not reveal any trends. Thus no conclusions can be made as to whether the
eagles are increasing or decreasing in numbers, or holding their own. Temporal trends could only
be established if the survey were to be repeated sometime in the future. As farmers’ reports of
raptors were often drawn from their lifetimes’ experiences, it will be necessary to re-phrase the
questionnaires before repeating the survey.

Although the 1976 Survey does not show trends in the status of eagles, deductions can be
made if the information is considered in conjunction with all past and present records of eagles
(Boshoff et al. in prep.) and an evaluation of the relationship between stock farmers and eagles
(Boshoff & Vernon in prep.).

The authors consider that the Black Eagle and the Crowned Eagle still have viable populations.
Though persecuted, several factors allow them to persist. The nature of their habitats, mountains
and forests respectively, make these eagles relatively inaccessible. Their natural prey, dassies and
monkeys, are still relatively numerous permitting pairs of breeding eagles to survive without
encroaching on domestic stock. Birds that take stock tend to be juveniles which have dispersed
from their natal areas and consequently the number of these eagles removed by farmers appears
to be less than the breeding increment. There is no reason why such a situation cannot persist
indefinitely.

130

BOSHOFF & VERNON: DISTRIBUTION AND STATUS OF SOME EAGLES

Little comment can be made about the status and conservation of the Fish Eagle in the Cape
Province. There is no indication that farming has affected the population of this species in any
adverse way. The species may, however, be declining due to human disturbance (Siegfried et al.
1976). The main causes of the decline are probably pollution and urban development around
estuaries. However, the many artificial water impoundments created in recent times appear to
have benefited the population by the provision of additional permanent food sources as well as
breeding sites such as partially drowned trees.

The Martial Eagle population may be in jeopardy and certainly gives cause for concern. The
species is rightly listed as vulnerable in the South African Red Data Book — Aves (Siegfried et al.
1976). The population is small and widely dispersed. Even if there were 100 pairs, each with a
territory of 10 km. radius, there would be no Martial Eagles breeding in nine-tenths of the Cape
Province. The species is not liked by farmers in general, even those who are favourably disposed
towards other eagles. It is testimony to the resilience of the Martial Eagle that it still exists in the
province. As with the Fish Eagle the Martial Eagle has capitalized to a large extent on man’s
activities in that it readily nests on power pylons (Dean 1975b; A. Daneel pers. comm.', pers.
rec.) and in exotic trees, such as pines, eucalypts and poplars, often introduced to farms (R.
Martin pers. comm. ; pers. rec. ). In the Karoo, where suitable nest trees are scarce, two nests of
this species on cliffs are known (M. H. Currie, pers. comm.). Similarly the Tawny Eagle is also
known to nest in exotic trees in the Colesberg district (R. C. Rous pers. comm. ).

As biologists the authors wonder how a species such as the Martial Eagle, which occurs at low
density and under such pressure, is able to persist. As conservationists they are alarmed at the
level of persecution. Other than the Kalahari Gemsbok National Park, there are no large reserves
1 in the Cape Province, where a viable population of these eagles might persist. Consequently the
responsibility for the conservation of eagles in the province depends upon the farmers. There
appear to be several socio-economic factors associated with eagles and the stock problem which
need further elucidation (Boshoff & Vernon in prep.).

CONCLUSION

Useful and valid information was obtained about the distribution and abundance of four
species of eagles in the Cape Province. The findings are applicable to the entire province. Despite
certain disadvantages, the method employed seems to be the only way of obtaining such
complete coverage. Although the results do not reveal trends, certain deductions can be made if
these data are linked with ancillary information. There are nine species of eagles known to breed
\ in the province and five — Black Eagle, Tawny Eagle, Crowned Eagle, Fish Eagle and Booted
Eagle — appear to have viable, if small, populations. A sixth species, the Martial Eagle may be
declining to extirpation. The 1976 Survey was valuable in that it presented a broad review of the
status of eagles and gave pointers for more intensive investigations.

ACKNOWLEDGEMENTS

The authors wish to express their sincere gratitude to those farmers/landowners who completed
and returned their questionnaires and, in many cases, provided additional information and
hospitality during the field survey. They are also indebted to those members of the CDNEC who
, assisted with the survey, both in the office and in the field, and to Mrs H. Glover and Mr A.
Bastiaans of the Cape Provincial Administration Computer Section who assisted with the
databanking and map production. The authors also thank Mr R. K. Brooke, operating on a CSIR
grant at the FitzPatrick Institute, University of Cape Town, and Drs R. C. Rous and M. J. F.
Jarvis for providing information, as did numerous other observers. The Director of the CDNEC is
thanked for permitting the authors to undertake and publish the results of the 1976 Survey.

131

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 9, OCTOBER 1980

REFERENCES

Boshoff, A. F. 1979. Variation in the Afrikaans vernacular for some eagles in the Cape Province. Bokmakierie 31 (1):
22-23.

Boshoff, A. F., Brooke, R. K. and Crowe, T. M. 1978. A computerised distribution mapping scheme for vertebrates in
southern Africa, illustrated by a range decrease in the Bearded Vulture Gypaetus barbatus (Linn.). SAfr. J.
Wild!. Res. 8 (4): 145-149.

Boshoff, A. F. and Vernon, C. J. 1980. The past and present distribution and status of the Cape Vulture in the Cape
Province. Ostrich 51 (In press).

Brown, L. H. 1970. African birds of prey. London: Collins.

Brown, L. H. 1976. Eagles of the world. Cape Town: Purnell.

Cade, T. J. 1969. The status of the Peregrine and other falconiforms in Africa. In Hickey, J. J. ed. , Peregrine falcon
populations: their biology and decline. London: Univ. of Wisconsin Press, pp. 289-381.

Dean, W. R. J. 1975j*. Dry season roadside raptor counts in the N. Cape, South West Africa and Angola. J. sth. Afr.
Wildl. Mgmt Ass. 5 (2): 99-102.

Dean, W. R. J. 1975b. Martial Eagles ( Polemaetus bellicosus) nesting on high tension pylons. Ostrich 46 (1): 1 16-1 17.
Diem, K. and Lentner, C. 1970. (Eds.). Scientific tables. Basle: Ciba-Geigy.

Heyl, C. W. 1972. Ondersoek van roofvoel probleme in die Kuruman-distrik, Noord-kaap. Ongepubl. versl.. Dept. Nat.
Omg. Bew. Kaap.

Maclean, G. L. 1970. An analysis of the avifauna of the southern Kalahari Gemsbok National Park. Zool. afr. 5 (2):
249-273.

McLachlan, G. R. and Liversidge, R. 1978. Roberts birds of South Africa. Johannesburg: Trustees of the John
Voelcker Bird Book Fund.

Murphy, J. R. 1976. Eagles and livestock — some management considerations. In Chancellor, R. D. ed. , World
Conference on Birds of Prey, Vienna, 1975 . London: ICBP, Report on proceedings, pp. 307-313.

Noble, J. 1886. Official handbook of the Cape of Good Hope . Cape Town: Solomon.

Rowan, M. K. 1964. Relative abundance of raptorial birds in the Cape Province. Ostrich 35 (3): 224-227.

Rudebeck, G. 1963. Studies on some palaearctic and arctic birds in their winter quarters in South Africa: 4. Birds of prey
(Falconiformes). In Hanstrom et al. eds. South African animal life. Vol. 9. Stockholm: Swedish Natural Science
Research Council, pp. 418-516.

Siegfried, W. R. 1963. A preliminary report on Black and Martial Eagles in the Laingsburg and Philipstown divisions.
Cape Dept. Nat. Cons. Invest. Rep. No. 5.

Siegfried, W. R. 1966. Relative abundance of raptorial birds in the south-western Cape Province. Ostrich 37 (1): 42-44. l!
Siegfried, W. R. 1968. Relative abundance of birds of prey in the Cape Province. Ostrich 39 (4): 253-258.

Siegfried, W. R., Frost, P. G. H., Cooper, J. and Kemp, A. C. 1976. South African Red Data Book: Aves. (S. Afr. r
National Scientific Programmes, Rep. No. 7). Pretoria: CSIR.

Skead, C. J. 1967. Ecology of birds in the Eastern Cape Province. Ostrich Suppi. No. 7.

Snelling, J. C. 1969. A raptor study in the Kruger National Park. Bokmakierie 21 (3). Suppi.: vii-xi.

Vernon, C. J. [1978], A review of the status of raptors in the Eastern Cape. In A symposium on African predatory birds. ii
Pretoria: Northern Transvaal Ornithological Society, pp. 87-95.

Wiley, R. W. and Bolen, E. G. 1971. Eagle-livestock relationships: livestock carcass census and wound characteristics.
Southwestern Nat. 16: 151-169.

Winterbottom, J. M. 1968. A checklist of the land and fresh-water birds of the Western Cape Province. Ann. S. Afr.
Mus. 53: 1-276.

Winterbottom, J. M. 1972. The ecological distribution of birds in southern Africa. Percy FitzPatrick Inst. Afr. Ornith.
Monogr. No. 1.

Manuscript accepted for publication 26 May 1 980.

Q )“F

**"* ANNALS.

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

Ann. Cape Prov. Mus. (nat. Hist.)

VOLUME 13 • PART 10
30th APRIL 1981

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

Printed by Cape & Transvaal Printers (Pty) Ltd, Cape Town

BD9446

Early names and records of two small Hieraaetus eagles (Aves: Accipitridae)
in the Cape Province of South Africa

by

R, K. BROOKE

(FitzPatrick Institute, University of Cape Town)
and

C. J. VERNON
(East London Museum)

ABSTRACT

Names and records of Ayres’s Hawkeagle and the Booted Eagle, chiefly in the Cape
Province of South Africa are discussed. Morphinus dubius Smith was said by Roberts to be the
first name applied to Ayres’s Hawkeagle and as a result the species is universally called either
Hieraaetus dubius or Aquila dubia. Smith’s type may not have survived but his description
lacks reference to any character diagnostic of Ayres’s Hawkeagle whereas it is a straightfor-
ward description of a pale phase Booted Eagle H. pennatus. Thus M. dubius is a junior
synonym of H. pennatus and the first available name for Ayres’s Hawkeagle is Spizaetus ayresii
Gurney. The type locality of M. dubius is in the Vredendal District of the western Cape
hundreds of kilometres from the woodlands and forests which H. ayresii frequents. There is no
satisfactory evidence for the occurrence of H. ayresii in the western Cape and all records are
rejected. In the eastern Cape some records are acceptable but all are of wandering immature
birds. The most southerly record of an adult is from the northern Transkei. Most misidentifica-
tions are of H. pennatus which is now known to be relatively common in the Cape Province. A
Zambezi valley record of H. ayresii is validated.

WHAT IS MORPHINUS DUBIUS?

The name Morphinus dubius Smith (1830a) (full citations of names will be found in
Appendix 1) lay unused for over 100 years because it was believed to be indeterminate.
Roberts (1936: 321) argued that M. dubius was the first description of Ayres’s Hawkeagle
Spizaetus ayresii Gurney (1862). There has been no further discussion of Roberts’s view but it
has been universally accepted and for over 40 years Ayres’s Hawkeagle has been referred to as
Hieraaetus dubius or Aquila dubia according to taxonomic taste.

Smith (1830a) used the generic name Morphinus and attributed it (p. 115) to Cuvier. In
Smith (1834) he corrected the name to Morphnus (p. 276) still attributing it to Cuvier. As
Friedmann (1950: 421) makes clear, Morphinus is an unjustified emendation by Fleming in
1822 of Dumont’s, 1816, Morphnus which Dumont attributed to Cuvier, presumably in MS.

133

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 10, APRIL 1981

In the course of C.J.V.’s investigations into the past and present distribution of eagles in
the Cape Province it has emerged that Ayres’s Hawkeagle is a rare vagrant and that several of
the claimed records are unacceptable. A critical reading of Smith (1830a) does not permit
agreement with Roberts’s (1936) findings on M. dubius. There are no characters in the descrip-
tion which apply it unequivocally to Ayres's Hawkeagle: it lacks any reference to an occipital
crest, an evenly barred tail or ginger coloured underparts, the three plumage characters that
separate juvenile Ayres’s Hawkeagle from the similarly sized Booted Eagle H. pennatus
(Gmelin) 1788.

Smith’s (1830a) text is a good description of a pale phase H. pennatus “plumage above
. . . the shafts of the feathers and a portion of the vanes adjoining being always darkest; . . .
chin, throat, and centre of breat chiefly white, with some brown shades in the form of longitu-
dinal streaks that include the shafts of the feathers; belly white, here and there variegated with
clouds of light reddish-brown or with dark brown narrow streaks, produced by the shafts being
that colour. Several of the long feathers on the flanks more or less distinctly crossed with
reddish-brown, and those on the outside of the thighs also slightly marked in the same way,
only with a lighter shade, the rest of the feathers of the legs pure white . . . insides of shoul-
ders white , with a few blackish-brown spots or blotches. Tail nearly even, the feathers blackish-
brown, variegated with narrow waved irregular bands of a darker tint,” (italics added).

The type of M. dubius may not have survived and this is true of several of the types of
animals described by Andrew Smith while still in South Africa. However, Smith (1834)
synonymized his M. dubius with his newly proposed Butaetes lessonii of which the type sur-
vives in the British Museum (Natural History) under reg. no. 43.2.28.64. The register, but not
the label, states that the specimen is Buteo (sic) lessonii and the entry lies among a batch of
Andrew Smith’s types registered at the same time (C. W. Benson in litt. 1978 to R.K.B.). The
specimen is, as the description would lead one to expect and did lead Hartert (1914) to find, an
undoubted pale phase H. pennatus. This view of what is meant by M. dubius is supported by
Smith’s (1834) revision coupled with the identity of the surviving second type.

The view taken of the relationship between M. dubius and B. lessonii in the preceding i
paragraph is not the only possible one in the light of known taxonomic practice in the early
XIX century. It is quite possible, indeed very likely, that Smith regarded B. lessonii as a
redescription of M. dubius in a more appropriate genus and with a less inappropriate specific
epithet. This is borne out by the fact that the description of B. lessonii is a shortened version of
that of M. dubius and does not differ from it positively in any respect. If this suggestion is truly
the case, then specimen no. BM(NH) 43.2.28.64 is the type not only of B. lessonii as argued
above but also of M. dubius. This suggestion could be disproved by critically comparing the
specimen with the description of M. dubius.

CONSEQUENCES OF FINDING

What follows from this finding? In the first place, as set out in Appendix 1, the name
dubius of Smith cannot be applied to Ayres’s Hawkeagle and workers must revert to calling
the species ayresii of Gurney in whatever generic combination seems warranted: preferably
Hieraaetus ayresii, cf. Brooke (1974). Smith’s M. dubius must be placed in and Butaetes lesso-
nii must be restored to the synonymy of H. pennatus (Gmelin) as Hartert (1914) held and
contra Brooke (1974). If the South African breeding population of H. pennatus proves taxono-
mically separable from west Palaearctic breeding populations (this does not seem likely to be
the case), M. dubius is the first name applied to South African birds and would form their
subspecific epithet. Heerenlogement, the type locality of M. dubius, must be restored to the
list of localities in the Cape Province at which H. pennatus has been collected since its rejec-
tion by Brooke (1974) has been shown to be unwarranted. Likewise, the sight and breeding
records at Franschhoek and east of Klaver made in 1917 by Roberts (1936: 322) attributed to

EARLY NAMES AND RECORDS OF TWO SMALL HIERAAETUS EAGLES

H. ayresii must be transferred to H. pennatus which is now known to be a widespread breeding
species in the Cape Province south of the Orange River (Brooke et al. 1980).

OTHER RECORDS REVIEWED

Morphinus albescens Smith (1830a) was transferred by Brooke (1974) to the synonymy of
Spizaetus coronatus (L.), 1766, from that of H. pennatus where Hartert (1914) had provisional-
ly placed it. In his 1834 revision Smith admitted S. coronatus sub nom. Aquila coronata to the
South African avifauna (p. 273) but did not realise that his M. albescens was only its juvenile.
Roberts (1936: 294) seems to have made a reciprocal error in dealing with Smith’s MS entry
under M. ornatus. Roberts did not publish the lengthy descriptions that Smith wrote so it is not
practicable to check what Smith was describing: one may accept Roberts’s view that they apply
to S. coronatus. Nonetheless, the notes that Roberts (1936) did publish, including the refer-
ence to a specimen shot at the Heerenlogementsberg in January 1829, only make sense when
applied to H. pennatus. Smith’s bird had a “Bill towards cere bluish . . . cere and feet a sort
of greenish yellow” whereas S. coronatus has a black bill with dark cere. The habitat is an arid
scrubcovered mountain rising out of a semidesert plain which is appropriate for H. pennatus
(Brooke et al. 1980) but not for H. ayresii and S. coronatus which are eagles of forest and
woodland. And Smith (Roberts 1936) said his species was “very common”.

Roberts (1936: 322) mentions a specimen of H. ayresii which he examined in the South
African Museum, Cape Town. R.K.B. has recently re-examined it and it is an undoubted
H. ayresii as Roberts said. It is a small adult male with hardly any markings on the underparts.
The label, written in the last century, says “Cape District — purchased” and there is no in-
formation given on who collected it or sold it to the museum. The unsatisfactory wording of
1 the label is not proof that H. ayresii has occurred in the western Cape (west of 25°E and south
of 29°S) in recorded history in the absence of any other records which might be called analo-
gously in its support.

In addition to the adult male specimen of uncertain provenance just discussed there are
three or four specimen records of H. ayresii from the eastern Cape which are acceptable. The
earliest was collected c. 1820 by Mundt and Le Maire, two Prussian botanists (ffolliott &
| Liversidge 1971), somewhere on the Sundays River: it is a juvenile female in the Humboldt
Museum, Berlin (Stresemann 1924). There is an unsexed juvenile from Grahamstown taken in
I 1910 in the Albany Museum, Grahamstown, which is also mentioned by Finch-Davies (1919)
S and another unsexed juvenile without date in the East London Museum merely labelled “east-
, ern Cape Province”. The label, at least, must have been written in or after 1910 when the
Cape Colony became the Cape Province. This is probably not the juvenile female from near
Port Elizabeth mentioned by Finch-Davies (1919).

Layard (1867) placed the Buse Gantee of Levaillant (1796) in the synonymy of H. penna-
tus. This is clearly incorrect. Smith (1830b) pointed out that European workers had placed it as
a synonym of the Roughlegged Buzzard Buteo lagopus (Pontoppidan, 1763, Danske Atlas p.
i 616: Denmark) which is clearly the species of the picture and much of the description. But the
species does not occur in the Cape Province: the most southerly recorded localities in Vaurie
! (1965) are Cyprus and Malta. What was the specific identity of the Outeniqua forests speci-
men, shot by Levaillant, which was so like B. lagopus that he failed to distinguish it? After
discussion between R.K.B. and Dr L. H. Brown it seems nearly certain that the specimen was
a subadult H. ayresii. Living solitarily in forested country is appropriate and so are the heavily
. blotched tarsal feathers reaching to the toes with their long thin, “plus effilees”, claws. R.
Martin (pers. comm, to R.K.B.) who has a good knowledge of the forested country of George
and Knysna, the Outeniqua forests of early writers, has never seen H. pennatus there despite
search. And it would seem less easy to muddle the European H. pennatus with B. lagopus than

135

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 10, APRIL 1981

the then unknown H. ayresii in immature plumage. Thus Levaillant’s Buse Gantee is almost
certainly the first record of a vagrant immature H. ayresii in the eastern Cape.

Skead (1967) gives three other records of H. ayresii from the eastern Cape. Those from
Golden Valley and King Williams Town in 1966 have been re-examined and they prove to be
specimens of H. pennatus. The 1962 record from King Williams Town is a MS sight record by

G. Ranger and C. J. Skead at Kei Road on 7 April 1962 of a juvenile H. ayresii described as
having an “unmistakable crest”. This record seems acceptable, not least because of the lengthy
field experience of both observers. It will be noted that all indubitable records from the Cape
Province (not the extralimital Transkei) are of juvenile or immature birds. The only adults are
the South African Museum specimen whose provenance is uncertain and one which was shot in
Pondoland, Transkei, in February 1909 but which was not preserved (Finch-Davies 1919).

H. ayresii was recorded by Winterbottom (1971: 60) from Goodhouse on the Orange
River where it forms the frontier between the Cape Province and Namibia but the source of
this record is not known (Dr J. M. Winterbottom pers. comm, to R.K.B.). In the absence of
supporting details this record of a forest and woodland eagle in the desert is unacceptable and

H. ayresii should not be admitted to the Namibian list on the strength of this record.

Sclater (1905) and Stoehr & Sclater (1906: 109) recorded the collection of H. spilogaster
on 25 February 1904 from George’s Ferry at 15° 40'S, 30° 18'E on the Zambian bank of the
Zambezi (Benson et al. 1970: 55) a few kilometres upstream from Feira at the junction of the
Luangwa and Zambezi Rivers. Finch-Davies (1919) redetermined this specimen as H. ayresii
and even provided a coloured portrait of it to illustrate his paper. This redetermination was
overlooked by Benson et al. (1971) who stated that H. ayresii had not yet been recorded in the
Zambezi and Luangwa valleys. The specimen has been demounted but is still preserved in the
South African Museum where R.K.B. has recently examined it and found it to be, as Finch-
Davies (1919) said, an adult female H. ayresii.

As a result of the foregoing findings Map 101 for H. ayresii in Snow (1978) needs to be
amended by the deletion of the four symbols in the western Cape and by the addition of one at
the Luangwa/Zambezi confluence.

Brooke (1974) was written before he had seen the South African Museum material of
H. pennatus. Table 1 may be regarded as a continuation of Table 1 in Brooke (1974) in that it
provides analogous data on the South African Museum specimens. The Wynberg specimen
which was formerly mounted was collected by Dr H. Exton, cf. Winterbottom (1955).

Table 1

Data on specimens of Hieraaetus pennatus pennatus examined in the South African Museum, i1

Cape Town. All measurements in mm.

Date

Place

Sex

Age

Wing

length

Culmen

length

Hind

Claw

length

Phase

c 1870

Wynberg, C.P.

cf

ad.

350

19,8

20,7

dark

c. 1880

—

Cf

ad.

355

21,5

22,5

pale

c. 1900

—

?

imm.

385

22,3

26,2

pale

4.10.1910

Robertson, C.P.

9

imm.

370

21 +

26,2

pale

24.11.1910

Lynedoch, C.P.

cf

ad.

360

20,5

22,0

pale

136

EARLY NAMES AND RECORDS OF TWO SMALL HIERAAETUS EAGLES

ACKNOWLEDGEMENTS

This work was carried out while R.K.B. was employed under the South African National
Programme for Environmental Sciences (C.S.I.R.). R.K.B. is obliged to Dr T. H. Barry,
Director of the South African Museum, Cape Town, for study facilities.

REFERENCES

Benson, C. W., Brooke, R. K., Dowsett, R. J. and Irwin, M. P. S. 1970. Notes on the birds of Zambia: Part V. Arnoldia
( Rhodesia ) 4 (40): 1-59.

Benson, C. W., Brooke, R. K., Dowsett, R. J. and Irwin, M. P. S. 1971. The birds of Zambia. London: Collins.
Brooke, R. K. 1974. On the material evidence of Hieraaetus pennaius in southern Africa. Bull. Br. Orn. Club 94:
152-158.

Brooke, R. K., Martin, R., Martin, J. and Martin, E.1980. The Booted Eagle, Hieraaetus pennatus , as a breeding
species in South Africa. Gerfaut 70: 297-304.
i Ffolliott, P. and Liversidge, R. 1971. Ludwig Krebs. Cape Town: Balkema.

Finch-Davies, C. G. 1919. Some notes on Hieraaetus ayresii Gurney Sen. (Lophotriorchis lucani Sharpe et auctorum).
Ibis (11) 1: 167-179.

' Friedmann, H. 1950. The birds of north and middle America. Bull, U.S. natn. Mus. 50 (XI).

Gurney, J. H. 1862. A fourth additional list of birds received from Natal. Ibis 4: 149-158.

Hartert, E. 1914. Die Vogel der paldarktischen Fauna 2. Berlin: Fncdlandcr.

Layard, E. L. 1867. The birds oj South Africa. Cape Town: Juta.

Levaillant, F. 1796. Histoire naturelle des oiseaux d'Afrique 1. Paris: Jansen.

Roberts, A. 1936. Some unpublished field notes made by Dr (Sir) Andrew Smith. Ann. Transv. Mus. 18: 271-323.

Sclater, P. L. 1905. An exhibition of a specimen of Dubus’ Hawk-Eagle. Bull. Br. Orn. Club 15: 67-68.

Skead, C. J. 1967. Ecology of birds in the eastern Cape Province. Ostrich suppl. 7.

Smith, A. 1830a. A description of the birds inhabiting the south of Africa. S. Afr. Q. Jl. 2: 105-120.

Smith, A. 1830b. A description of the birds inhabiting the south of Africa. S. Afr. Q. JL 4: 380-392.

: Smith, A. 1834. African zoology. S. Afr. Q. Jl. (2) 3: 273-288.

Snow, D. W. 1978. An atlas of speciation in African non-passerine birds. London: Trustees Brit. Mus. (Nat. Hist.).
Stoehr, F. E. (sic) and Sclater, W. L. 1906. Notes on a collection of birds made in north-east Rhodesia by Dr F. E.
Stoehr. Jl. S. Afr. Orn. Un. 2: 83-114.

1 Stresemann, E. 1924. Uber Hieraaetus ayresii und Spizaetus africanus. Novit. zool. 31: 214-216.

Vaurie, C. 1965. The birds of the palaearctic fauna 2. London: Witherby.

Winterbottom, J. M. 1955. On the range of the Booted Eagle in South Africa. Ostrich 26: 136.

Winterbottom, J. M. 1971. A preliminary checklist of the birds of South West Africa. Windhoek: S.W. Afr. Sci. Soc.

Manuscript accepted for publication 30 May 1980.

137

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 10, APRIL 1981
APPENDIX 1

Citations and type localities of names accepted as valid or synonymous in this paper. The

synonymies for H. pennatus and H. ayresii are believed to be complete.

"Indicates original reference examined by R.K.B.

HIERAAETUS Kaup

Hieraaetus Kaup, 1844, Class. Saeug. Voegel p. 120. Genotype Falco pennatus Gmelin by
original designation.

Synonyms

Eutolmaetus Blyth, 1845, J. Asiat. Soc. Beng. 14: 174. Genotype Aquila fasciata Vieillot by
original designation.

* Anomalaetus Roberts, 1922, Ann Transv. Mas. 8: 208. Genotype Spizaetus ayresii Gurney by
original designation and monotypy.

HIERAAETUS PENNATUS (Gmelin)

* Falco pennatus Gmelin, 1788, Syst. Nat. 1 (1): 272, no locality but restricted to France by
Swann, 1922, Synopsis Accipitres p. 113.

HIERAAETUS PENNATUS PENNATUS (Gmelin)

Synonyms

Aquila minuta Brehm, 1820, Beitr. Vogelkd. 1: 68, Orla River, western Poland.

* Morphinus dubius Smith, 1830, S. Afr. Q. Jl. 1 (2): 177, Heerenlogement at 31° 58'S, 18°
34'E in the Vredendal District, western Cape Province, South Africa.

*Butaetes lessonii Smith, 1834, S. Afr. Q. Jl. (2) 3 (3): 287, Heerenlogement.

Aquila brehmii von Mueller, 1851, Naumannia 1 (4): 24, Melpes, Kordofan Province, Sudan.

Aquila nudipes von Mueller, 1851, Naumannia 1 (4): 25, South Africa.

Aquila paradoxa Brehm, 1853, J. Orn ., Lpz. 1: 204, northeastern Africa at 15°N.

Aquila longicaudata Heuglin, 1855, Sber. Akad. Wiss. Wien 19: 258, Shereik, Northern Pro-
vince, Sudan.

Aquila gymnopus Heuglin, 1855, Sber. Akad. Wiss. Wien 19: 258, nom. nov. pro A. nudipes
von Mueller.

Aquila maculatirostris Brehm, 1856, Allg. dt. naturh. Ztg. p. 46, Sudan.

Aquila pygmaea A. E. Brehm & C. L. Brehm, 1861, Ber. dt. Orn. Ges. p. 99, Madrid, Spain.

Hieraetos (sic) pennatus major & medius A. E. Brehm, 1866, Verz. Samml. p. 2 are nom.
nud.

Aquila albipectus Severtsov. 1873, Turkest. Zhivotn. p. 63, Alma Ata, Kazakh S.S.R.,
U.S.S.R.

HIERAAETUS PENNATUS MILVOIDES (Jerdon)

Spizaetus milvoides Jerdon, 1839, Madras J. Lit. Sci. 10: 75, Tiruchchirappalli, Madras Pro-
vince, India.

Synonym

Hieraaetus pennatus harterti Stegmann, 1935, Orn. Mber. 43: 151, Kyakhta, southwestern
Transbaicalia, U.S.S.R.

HIERAAETUS AYRESII (Gurney)

*Spizaetus ayresii Gurney, 1862, Ibis 4: 149 and pi. 4, near the coast of Natal.

Synonyms

*Buse Gantee Levaillant, 1796, Histoire naturelle des oiseaux d'Afrique I: 79 and pi. 18, Out-
eniqua. Cape Province, South Africa (partim).

Lophotriorchis lucani Sharpe & Bouvier, 1877, Bull. Soc. zool. Fr. p. 471, Landana, Cabin
da, Angola.

138

EARLY NAMES AND RECORDS OF TWO SMALL HIERAAETUS EAGLES

HIERAAETUS SPILOGASTER (Bonaparte)

Aquila spilogaster Bonaparte, 1850, Rev. Mag. Zool. p. 487, Ethiopia.

SPIZAETUS Vieillot

*Spizaetus Vieillot, 1816, Anal. Nouv. Orn. Elem. p. 24. Genotype Falco ornatus Daudin by
subsequent designation by Gray, 1840, List Gen. Birds p. 2.

Synonym

* Stephanoaetus W. L. Sclater, 1922, Bull. Br. Orn. Club 42: 75. Genotype by original designa-
tion and monotypy Faleo coronatus Linnaeus.

SPIZAETUS CORONATUS (Linnaeus)

Falco coronatus Linnaeus, 1766, Systema Naturae, ed. 12, vol. 1: 124, Guinea coast of west
Africa.

Synonyms

Falco albescens Daudin, 1800, Traite Orn. 2: 45, Outeniqua, Cape Province, South Africa.

* Morphinus albescens Smith, 1830, S. Afr. Q. Jl. 1 (2): 155, Outeniqua, Cape Province, South
Africa.

139

QU

ft(el3
V t-f- '

ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

Ann. Cape Prov. Mus. (nat. Hist.)

VOLUME 13 • PART 11
30th APRIL 1981

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

Printed by Cape & Transvaal Printers (Pty) Ltd, Cape Town

BD9447

An annotated checklist of the lizards of the Cape Province,

South Africa

by

W. R. BRANCH

(Port Elizabeth Museum, P.O. Box 13147, Humewood 6013, South Africa)

CONTENTS

Page

Abstract 141

Introduction 141

Sauria

Infraorder: Gekkota 142

Family: Gekkonidae 143

Infraorder: Anguimorpha 149

Family: Varanidae 149

Infraorder: Iguania 149

Family: Chamaeleonidae 149

Family: Agamidae 151

Infraorder: Scincomorpha 151

Family: Lacertidae 151

Family: Scincidae 154

Family: Cordylidae 159

Doubtful Records 164

Acknowledgements 165

References 165

ABSTRACT

A checklist of the lizards of the Cape Province, South Africa is presented. A total of 143
species and subspecies is recognized. The taxonomic status of no less than 84 (59,1%) of these
has changed since the last review of the Cape lizards by FitzSimons (1943). Comments on the
distribution and taxonomic status of problematic groups are included.

INTRODUCTION

Recent years have seen a burgeoning interest in the small vertebrate fauna of the Cape
Province. This is coupled with the awareness of the uniqueness and zoogeographic importance
of the Fynbos Biome and adjacent regions (Kruger, 1978). Studies on the herpetofauna of the
Fynbos Biome are minimal, and hampered by poor taxonomy. This is particularly acute for

141

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 11. APRIL 1981

lizards, which present the greatest diversity among the reptiles. The suborder was last re-
viewed by Dr V. F. M. FitzSimons (1943) in his monograph on the “Lizards of South Africa.”
However, during the intervening 37 years numerous taxonomic changes at the familial, generic
and specific level have occurred.

This review originated as a checklist of the lizards inhabiting the Fynbos Biome, incorpor-
ating all the diverse taxonomic changes that have occurred subsequent to FitzSimons’ mono-
graph (1943). However, accurate distribution maps exist for very few lizards inhabiting the
Cape Province and this makes it impossible to identify all those species that may inhabit the
Fynbos Biome. The scope of the review has thus been expanded to cover the total lizard fauna
of the Province. The eastern border of the Cape Province remains in a state of political flux.
As currently recognized, it consists of a wavy line running from East London to Lesotho. The
only lizard affected by this is Afroedura pondolia , that reaches its western limit at Kentani
(3228 AD) and thus no longer qualifies for Cape residence.

The taxonomy of a number of genera and species is now relatively well stabilized, but that
of many remains confused. Some of these problem groups are currently under investigation,
and the results of these studies will in due course affect this review. In addition the distribu-
tions of most species are only poorly known and will alter as the region becomes better col-
lected. Subject to these limitations, 143 species and subspecies of lizards are known to occur in
the Cape Province (Table 1). No less than 84 (59,1%) of these have been affected by some
form of taxonomic change since FitzSimons’ monograph (1943). However, only one new spe-
cies and three new subspecies of lizard have been described from the Cape in this time, and it
is evident that the alpha-phase of the taxonomy of South African lizards, in which new species
are discovered, has all but passed. Future studies will, for the most part, refine our knowledge
of the relationships of the known forms, and few new species can be expected.

CLASS: REPTILIA
ORDER: SQUAMATA
SUBORDER: SAURIA

There is still discussion concerning the ordinal classification of squamate reptiles. The
extensive studies of Gans (see review 1978) have convincingly demonstrated that burrowing :j
“worm lizards” differ from true lizards in many characteristics, and should be elevated to
equivalent rank within the Squamata. The suborder Amphisbaenia is represented in southern
Africa by only four genera; Monopeltis, Dalophia Chirindia and Zygaspis. The species
Monopeltis capensis capensis, M. leonhardi , M. sphenorhynchus mauricei , Dalophia pistillum,
and Zygaspis quadrifrons all range into the northern Cape Province in the vicinity of the
Kalahari Gemsbok National Park and Kimberley. The taxonomy of all these forms has recent-
ly been reviewed (Broadley, et al ., 1976; Saiff, 1970) but refinements can still be expected.
Chromosome morphology indicates that Monopeltis capensis rhodesianus, at least, warrants
specific recognition (Branch, unpublished observations).

The osteological studies of McDowell and Bogert (1954) indicate that the Gerrhosauridae
and Cordylidae are closely related, and best treated as subfamilies of the Cordylidae. The
suborder Sauria is thus represented in southern Africa by only seven families out of the 17
families of Recent lizards currently recognized (Dowling and Duellman, 1978).

INFRAORDER: GEKKOTA

The Infraorder Gekkota contains two living families, Gekkonidae and Pygopodidae. The
pygopodids are derived from geckos through elongation of the body and loss of limbs, and are
restricted to the New Guinea — Australian region.

142

CHECKLIST OF CAPE LIZARDS

Family: Gekkonklae

An ancient and widespread family. Underwood (1954) proposed a division of gekkonid
lizards into three families, based mainly on the shape of the pupil. Within the Gekkonidae, he
placed the southern African genera Chondrodactylus , Colopus, Palmatogecko , Ptenopus,
Rhotropella and Rhotropus into the subfamily Diplodactylinae, whilst all remaining South
African geckos remained in the Gekkoninae. Kluge (1967) demonstrated that divisions based
on pupil shape were unnatural and recognized only a single family with four subfamilies. He
referred all South African geckos to the Gekkoninae, the Diplodactylinae being restricted to
the Australian subregion. Haacke (1975, 1976a, 1976b, 1976c, 1976d) in a detailed study of the
burrowing geckos of southern Africa (i.e. the first four genera listed above, plus the subse-
quently described Kaokogecko , Steyn and Haacke, 1966), supported this arrangement. Russell
(1976), moreover, has demonstrated that all these genera, with the exception of Ptenopus , are
closely related and are derivatives of the Pachydactylus radiation in the subregion.

Chondrodactylus Peters

Haacke (1976c) has recently reviewed the taxonomy and biology of this large, nocturnal,
terrestrial gecko. He recognizes a single species with two races.

Chondrodactylus angulifer angulifer Peters

Distributed through the northern, north-western, western and central Cape Province. The
region coincides closely with the 300 mm average rainfall zone.

Chondrodactylus angulifer namibensis Haacke X. C. a angulifer Peters.

Basically restricted to Namibia in the Namib and Pro-Namib, intergrades between the two
races are known from Brandkaross and Lekkersing in the extreme north-western Cape Pro-
vince.

Palmatogecko rangei Anderson

The range of this well-known Namib gecko is known to extend into the Richtersveld
(Brandkaross, Bloeddrif and Sendelingsdrif). The lower Orange River may occasionally run
dry and thus is no effective barrier to psammophilous forms. Haacke (1976a) has reviewed the
monotypic genus.

Colopus Peters

Haacke (1976b) has reviewed the taxonomy and biology of this monotypic genus of small,
terrestrial, nocturnal geckos. He recognizes two races, both of which occur in the northern
Cape Province.

Colopus wahlbergii wahlbergii Peters

The nominate race extends into the northern Cape Province at Vanzylsrus and Gemsbok,
in association with the Kalahari.

Colopus wahlbergii furcifer Haacke

The type locality of this new subspecies is Twee Rivieren on the Nossob River, Kalahari
Gemsbok National Park, northern Cape Province. It is found in association with the more arid
“dune areas” of the south-western Kalahari. It is characterized by a different colour pattern
and a more pointed, subtriangular head.

Phelsuma ocellata (Boulenger)

Originally described as a species of Rhotropus this diurnal gecko, which is restricted to
Little Namaqualand, was transferred to the monotypic genus Rhotropella by Hewitt (1937). It

143

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 11, APRIL 1981

has generally been assumed, and the new generic name implied, that ocellata was closely re-
lated to Rhotropiis. However, Russell (1977) has demonstrated that Schmidt (1934) was correct
in placing ocellata in Phelsuma, even though this generates an apparent zoogeographic anomaly.

Phelsuma is prevelant on Madagascar and the Indian Ocean Islands, although two species
are also known from the East African coast. However, the south-west arid region of Madagas-
car shows a number of faunal and floristic similarities to the Karoo. The expansion and con-
traction of the Kalahari Desert has fragmented the ranges of a number of other genera and
species of lizards, most notably that of Platysaurus (the species capensis is separated by 900 km
from its nearest congener) and that of Cordylus vittifer (the subspecies machadoi Laurent from
southern Angola is separated by nearly 1 000 km from the nominate race in the Transvaal). It
may, in part, be responsible for the hiatus in the distribution of Phelsuma , that now results
from the transfer of ocellata to the genus.

Pachydactylus Wiegmann

This is by far the largest genus of southern African lizards, and it has undergone an |
extensive radiation in the arid western regions. Many of the smaller genera of the region, i.e.
Palmatogecko, Colopus, Chondrodactylus, Kaokogecko and Rhotropiis are obviously derived
from Pachydactylus- like ancestors, and illustrate the diverse habitats into which the group has
radiated. The genus is undergoing active investigation by Dr G. McLachlan (South African
Museum) and Mr W. Haacke (Transvaal Museum), and it will be some time before its tax-
onomy is stabilized. There have been numerous important taxonomic changes in the Cape
representatives since FitzSimons’ (1943) monograph, and these are summarized in Table 2.

Pachydactylus austeni Hewitt

A small, terrestrial gecko, the habits of which are very similar to those of Palmatogecko
rangei (Haacke, 1976a). It is restricted to the coastal areas of Little Namaqualand, reaching as i
far south as Darling.

Pachydactylus mariquensis mariquensis A. Smith

A medium-sized, terrestrial gecko that lives in holes in the ground in karroid situations.
Extensively distributed in the inland areas of the Cape Province, from Uitenhage to Worces-
ter, and north to the Orange Free State.

Pachydactylus mariquensis latirostris Hewitt

Treated as a race of mariquensis by Loveridge (1947) and occurring in the north western
Cape Province from Victoria West to Pofadder.

Pachydactylus mariquensis macrolepis FitzSimons

Known from the northern parts of Little Namaqualand (type locality: Springbokfontein).
Loveridge (1947) suggested it may be synonomous with P. amoenus , which is known from the
same region. The status of macrolepis itself is doubtful (McLachlan, in litt.)

Pachydactylus amoenus Werner

Known only from the type collected at Kamaggas, Little Namaqualand. This area is now
relatively well-collected, and the status of this form has still to be resolved, particularly its
relationship to P. mariquensis macrolepis (see above).

Pachydactylus geitje (Sparrman)

A small, attractive, terrestrial gecko, previously referred to as P. ocellatus. Distributed
through the southern and western Cape, from Graaff-Reinet and Perdepoort in the Klein
Winterhoekberg to the Cape Peninsula, and north to the Bokkeveld Mountains. It is also
found on Robben Island.

144

CHECKLIST OF CAPE LIZARDS

Pachydactylus monticolus FitzSimons

A poorly known species, described from the Fransch Hoek mountains. Loveridge (1947)
though it may be synonomous with P. maculatus, but this is considered doubtful by McLach-
lan (in litt.), who has collected rugose P. geitje to which P. monticolus is very similar. Whether
the taxon deserves even subspecific recognition only further collecting will resolve.

Pachydactylus maculatus maculatus Gray

This widely-distributed terrestrial gecko has a confused taxonomy which is in need of
revision. Loveridge (1947) synonomized P. microlepis , P. maculatus albomarginatus and P.
capensis oculatus with P. maculatus. De Waal (1978) has agreed that oculatus is referable to
maculatus and not capensis, but recognizes it as a northern race (albeit tentatively, pending a
full revision of the species). The nominate race is distributed mainly in the eastern regions of
the Cape but extends as far west as Matjesfontein (Visser, 1979) and Matroosberg (McLach-
lan, in litt.)

Pachydactylus maculatus oculatus Flewitt

A north-eastern race that occurs from Tarkastad District to Colesberg and Norvalspont
and into the south-western Orange Free State (de Waal, 1978). P. m. albomarginatus is a
synonym of this race.

Pachydactylus serval onscepensis Hewitt

McLachlan and Spence (1966) reviewed the Pachydactylus weberi and P. serval groups.
They placed montanus in the synonomy of serval, and treated onscepensis and purcelli as sub-
species of serval. P. s. onscepensis is found only in the far northwest of the Cape Province
along the lower Orange River from Onseepkans to the Richtersveld.

Pachydactylus serval purcelli Boulenger

A southern race of serval, that reaches as far south as Touwsrivier and Matjesfontein.
Pachydactylus weberi weberi Roux

Three races of these small, rupicolous geckos are recognized. Loveridge (1947) treated
P. werneri (from northern Namibia) as a subspecies of weberi, but this was not followed by either
Mertens (1955) or McLachlan and Spence (1966). However, McLachlan (in litt.) now agrees
with this arrangement and also considers the southern race P. w. gariesensis Hewitt to be
untenable as enormous variation exists in the number of tubercules in southern specimens. The
nominate race thus extends into the northern Cape as far as the Calvinia-Clanwilliam districts
but is replaced in southern Namibia by P. w. acuminatus FitzSimons.

Pachydactylus punctatus amoenoides Hewitt

Now known to enter the Cape in the region of the Richtersveld (De Hoop, Brandkarros,
etc.) and the southern regions of the Kalahari Gemsbok National Park (Pianka, 1971). Mer-
tens (1955) doubts the validity of this race, although Laurent (1964) considers amoenoides to
be a full species. The group needs revision.

Pachydactylus rugosus rugosus A. Smith

The taxonomy of this species has recently been reviewed by Loveridge (1947), Mertens
(1955) and McLachlan (1979b). Loveridge (1947) resurrected the Namibian race frater but this
has not been accepted by either Mertens (1955) or McLachlan (1979b). The latter author has
transferred two subspecies of P. capensis, i.e. P. c. formosus and P. c. barnardi, to rugosus on
the basis of similar colouration and the structure of the dorsal tubercules. The nominate race
extends into the northern parts of Little Namaqualand in the vicinity of Onseepkans, Kakamas
and Rietfontein.

145

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 11, APRIL 1981

Pachydactylus rugosus formosus A. Smith

A southern race found in the western Cape Province from Ceres District to the southern parts
of Little Namaqualand. McLachlan (1979b) has restricted the type locality to Vredendal.

Pachydactylus rugosus barnardi FitzSimons

A central race of rugosus that is distributed throughout Little Namaquland. It is sympat-
ric with P. capensis labialis at Kamieskroon and Kleinsee. Broadley (1977b) has suggested that
barnardi is a full species as its caudal scalation differs from that of capensis and formosus. He
did not, however, compare its caudal scalation with typical rugosus.

Pachydactylus capensis capensis A. Smith

FitzSimons (1943) regarded P. capensis as polytypic with nine subspecies occurring in
southern Africa, of which five entered the Cape Province. The studies of Loveridge (1947) and
McLachlan (1979b), however, have reduced the number of Cape races to two. Broadley
(1977b) has further reduced the number of races in the north-eastern range of P. capensis. A 1
revision of the southern forms is needed. The nominate race extends north from the central
Cape Province through Botswana and the Orange Free State to southern Angola and the
northern Transvaal.

Pachydactylus capensis labialis FitzSimons

This western race is recorded from Little Namaqualand and may deserve specific recogni-
tion (McLachlan, in litt.)

Pachydactylus mentalis Hewitt

Known from the south-eastern Cape Province, from Graaff-Reinet to Tarkastad, this spe-
cies was treated as a subspecies of P. capensis by Loveridge (1947). However, de Waal (1978)
noted that the snout scales of mentalis differ from those of capensis and that its status requires
investigation.

Pachydactylus namaquensis (Sclater)

A large rupicolous gecko that is closely related to the following two species. It is found in i
Little Namaqualand and southwards into the western Karoo. A specimen (now in the Transvaal i
Museum) recently collected from the Langeberg Mountains (3321 DC) is a considerable southern i
extension (R. Boycott, pers. comm.).

Pachydactylus bibroni A. Smith

A wide-ranging large gecko, the range of which extends from the northern and western
Cape Province to Angola and Tanzania. Loveridge (1947) recognized three subspecies, bibro-
ni, turneri and pulitzerae, but these have not been accepted by Broadley (1966) or de Waal
(1978). The problem requires further investigation.

Pachydactylus laevigatus laevigatus Fischer

A large gecko closely related to and often confused with the previous species. The nomi-
nate race extends into the northern Cape reaching Marydale in the east and the vicinity of
Prieska and Springbok in the south. It extends north through Namibia to southern Angola,
being replaced by the race P. 1. fitzsimonsi (Loveridge, 1947) in Damaraland.

Lygodactylus Gray

An ancient genus that may be one of the few African lizards to have been directly affected
by the upheavals of continental drift. It is well represented on Madagascar, whilst two species
have also been recently described from South America. Although referred to a separate genus

146

CHECKLIST OF CAPE LIZARDS

(Vanzoia; Smith et al., 1977) these Neotropical representatives seem only doubtfully distinct
from Lygodactylus (Broadley, in litt.). One species (L. capensis) has even been found embed-
ded in amber (Kluge, 1967). The taxonomy of these small, diurnal geckos, notoriously dif-
ficult, has been recently reviewed by Pasteur (1964).

Lygodactylus bradfieldi Hewitt

This species, described by Hewitt in 1932, was treated as a western race of L. capensis by
FitzSimons (1943) but fared worse under Loveridge (1947) who placed it in the synonymy of
the latter. Pasteur (1964) re-elevated it to a full species but difficulty is still experienced in
separating the two forms. It is found at various localities in the northern and north-western
Cape, from Kimberley to the Richtersveld.

A specimen of L. capensis in the collection of the Addo Elephant National Park presum-
ably represents a translocation, possibly in a game crate, from the Kruger National Park.

Ptenopus Gray

Previously considered to be monotypic, containing the single species garrulus (FitzSimons,
1943). Two further species (carpi and kochi) have subsequently been described from Namibia.
The two subspecies of garrulus were synonomized by Brain (1962) but, more recently, Haacke
(1975) has resurrected maculatus. Both races occur in the Cape Province.

Ptenopus garrulus garrulus (A. Smith)

This small, terrestrial, pad-less gecko is widely distributed in those areas of southern Afri-
ca that are covered in Kalahari sand and receive less than 500 mm average rainfall. The range
extends into the northern Cape Province, reaching as far as the Orange Free State border near
Lemoenboord (de Waal, 1978).

Ptenopus garrulus maculatus Gray

A western form that extends through Namibia and the north-western Cape Province.
Somewhat surprisingly, specimens were recently collected near Calitzdorp in the Little Karoo
(Oelofsen and Kruger, 1976). This gecko may be more widespread in the central Cape Pro-
vince than realized.

Phyllodactylus Gray

A wide ranging genus found in Tropical America, Mediterranean Europe, Africa and
Australia. The taxonomy of the southern African forms remains confused.

Phyllodactylus peringueyi Boulenger

The status of this species remains problematic. The types, which remain the only known
specimens, are labelled as from Little Namaqualand and Port Elizabeth. That labelled as from
Little Namaqualand was accessioned with a Crotaphopeltis hotamboeia which was listed as
coming from the same area. As the herald snake is not known from Little Namaqualand this
locality must be in doubt (McLachlan, pers. comm.). However, the remaining locality (Chelsea
Point, near Port Elizabeth) is in a coastal dune region with little emergent rock. It is now
extensively invaded by exotic Australian Acacia and searches for this lizard at this spot by both
McLachlan (pers. comm.) and the author have proved fruitless. Hewitt (1937) has suggested
that the specimens are American. Their true identity, if not South African, remains unknown.

Phyllodactylus porphyreus (Daudin)

Known from the south and south-western Cape Province and extending east as far as
Cape St Francis and north to Little Namaqualand. FitzSimons (1943) recognized three races.
However, Loveridge (1947) synonomized the south coast form (cronwrighti — known from

147

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 11, APRIL 1981

Knysna to Cape St Francis) with typical porphyreus. The status of the west coast race nama-
quensis , which is known only from Little Namaqualand, is also in need of study. McLachlan
(in lift.) doubts the validity of the Bitterfontein locality (from which the type of namaquensis
comes). If correct, then Nieuwoudtville is the northern limit for the species.

Phyllodactylus lineatus lineatus Gray

A small, rupicolous gecko known from the south and central (Laingsburg) Cape Province
and north along the west coast to Little Namaqualand.

Phyllodactylus lineatus rupicolous FitzSimons

The validity of this northern form is doubtful as it is described from some localities listed
for typical lineatus (i.e. Springbok and Kamaggas). Haacke (1965) could also find no differences
between Oranjemund specimens and the typical race.

Phyllodactylus lineatus essexi Hewitt

This eastern form seems well-defined. Its range is now known to extend further west s
(Suurberg) than the type locality at Hounslow near Grahamstown.

i

Phyllodactylus microlepidotus FitzSimons

This large, nocturnal gecko has, since its original description in 1939, only recently been
rediscovered (McLachlan, 1979c; Visser, 1979). All known specimens come from the Cedar-
berg, mostly at Pakhuis Pass.

Afroedura Loveridge

This genus was described by Loveridge (1944a) to accommodate African geckos previous-
ly referred to the genus Oedura by FitzSimons (1943). Oedura is now restricted to Australia, ij
Loveridge (1947) synonomized a number of the species listed in FitzSimons (1943). However,
the taxonomy of the genus is still not stabilized and a much needed revision is presently in
progress (Onderstall, 1975).

Afroedura africana namaquensis (FitzSimons)

Known only from Little Namaqualand. Treated by FitzSimons (1943) as a full species, this il
form has been reduced to a subspecies of africana by Haacke (1965) who described a new sub-
species, tirasensis , from Great Namaqualand, intermediate in characteristics between typical
africana and namaquensis .

Afroedura nivaria (Boulenger)

Loveridge (1947) synonomized Oedura amatolica with Afroedura nivaria. The characters
used by Hewitt to differentiate amatolica are quite minor. The distribution of A. nivaria thus
closely parallels that of Pseudocordylus melanotus suhviridis being found in the Eastern Cape
Province in the vicinity of the Katberg and Amatola Mountains and thence northwards along
the Drakensberg escarpment.

Afroedura tembulica (Hewitt)

Known from the mountains of the south-eastern Cape Province, from Tembuland (from
which the species derives its name) to the Queenstown District. The relationship of this species
to nivaria and karroica requires investigation.

Afroedura karroica karroica (Hewitt)

Typical karroica was known to FitzSimons (1943) by only a poorly preserved type of
vague locality (Albany District). Hewitt’s subspecies A. k. wilmoti was recognized by Fitz-

148

CHECKLIST OF CAPE LIZARDS

Simons (1943) but was only dubiously separable from the nominate race due to the latter’s
vague provenance. Loveridge (1947) synonomized both forms, and the nominate subspecies is
now considered to range through the inner regions of the Eastern Cape Province, from Graaff-
Reinet to Tarkastad.

Afroedura karroica halli (Hewitt)

Loveridge (1947) treated halli as a race of karroica although FitzSimons (1943), who
treated the form as a full species in his monograph, considered it might only be a subspecies of
nivaria. Like tembulica it is restricted to the mountainous regions of the north-east-ern Cape
Province in the vicinity of Herschel.

INFRAORDER: ANGUIMORPHA
SUPERFAMILY: PLATYNOTA

Family: Varamidae

With the exception of the anguid genus Ophisaurus which ranges into North Africa, three
species of Varanus are the only anguimorph lizards found on the African continent. Varanus
griseus is restricted to Saharan Africa, whilst the remaining two species occur almost through-
out the continent.

Varanus niloticus niloticus (Linnaeus)

The race ornatus replaces typical niloticus in the West African rain forests. Along the
south-east coastal region the most westerly record of niloticus in the Cape Province appears to
be Patensie in the Gamtoos River Valley. In the north it extends along the Orange River to
Vioolsdrif ( pers . obser.) and possibly along its whole length, however, it is absent from the
central and southwest Cape.

Varanus exanthematicus albigularis (Daudin)

Entering the Cape Province in the northern, central and eastern regions, its southern and
western limits remain unknown. J. Greig (pers. comm.) has recorded the species from
Beaufort West.

INFRAORDER: IGUANIA

A group of generalized, primarily arboreal lizards, some of which have become secondarily
terrestrial or even semi-aquatic. None have become serpentiform, with elongate body or
reduced limbs. Three living families are recognized. The Iguanidae are the most primitive and
appear to have been the ancestoral stock for the Agamidae. The chamaeleons in turn may
have been derived from the latter family. Iguanids are not now known from mainland Africa,
although two genera occur on Madagascar. They have been displaced by their descendants,
the agamids.

Family: Chamaeleonidae

The generic status of the dwarf viviparous chamaeleons of the south and east coast of
southern Africa remains in a state of flux. Hillenius (1959) considered all the dwarf chamae-
leons placed by FitzSimons (1943) in the genus Microsaura (including the Cape forms taenio-
broncha, caffer, pumila, damarana , gutturalis and ventralis ) to be but subspecies of the wide-
spread species Chamaeleo pumilus. However, Raw (1976, 1978) has re-elevated these forms to
specific rank and revived the earlier generic name Bradypodion Fitzinger. This name has not
been used in over 130 years and, although it possesses priority over Microsaura Gray, it could
have been suppressed in favour of the more recently used and more widely known junior

149

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 11, APRIL 1981

synonym. These chamaeleons are usually found in allopatric populations and species limits are
difficult to define. Raw’s (1976) proposal to elevate all forms to specific rank is followed here
for convenience. However, it is probable that a number of these forms will prove to be but
local races, or even synonyms, as their biology and distribution become more well known.

Bradypodion taeniobronchum (A. Smith)

Apparently confined to the area of Algoa Bay, but very rare. Searches at Schoenmaker’s
Kop and Van Stadens River for further specimens have proved fruitless.

Bradypodion damaranum (Boulenger)

Perversely named as the species is only known from the coastal region around Knysna,
Tsitsikama and George.

Bradypodion pumilum (Daudin)

Its distribution is centred in the Western Cape Province, from the Cape Peninsula north-
wards to Clanwilliam and eastwards to Bredasdorp.

Bradypodion occidental (Hewitt)

Visser (1979) records this species from Melkbosstrand, 25 km north of Cape Town, a
considerable southerly extension from its previous record in Little Namaqualand.

Bradypodion ventrale (Gray)

Recorded from the Eastern Cape Province from Port Elizabeth to Port Alfred and inland
as far as Queenstown and the Katberg (FitzSimons, 1943). Visser’s (1979) record of this form
from Fort Beaufort (an area from which FitzSimons records karroicum ) may validate the spe-
cific status of the latter taxon, but this requires confirmation.

Bradypodion karroicum (Methuen and Hewitt)

As its name implies, extends through the karroid areas of the Cape Province from Fort
Beaufort (see note under B. ventrale) to Grootvadersbosch and inland as far as Nieuwoudtville
and Victoria West (Visser, 1979).

Bradypodion gutturale (A. Smith)

Poorly known and apparently confined to the Worcester and Robertson Districts.
Chamaeleo Laurenti

The genus is represented in the Cape Province by two species. Chamaeleo laevigularis
Muller was described from East London, Cape Province, but FitzSimons (1943) could not refer
it to any southern African forms and treated it as a “ species auctorum Hillenius (1963) has
subsequently compared the type to C. johnstoni from the Congo and found them to be conspe-
cific.

Chamaeleo namaquensis A. Smith

A large terrestrial chamaeleon that ranges into Little Namaqualand. The sexes are
dichromatic (Robertson,, 1978).

Chamaeleo dilepis dilepis Leach

The flap-necked chamaeleon enters the Cape Province in the Kalahari Gemsbok National
Park. It is known from a specimen collected at Nossob (Haacke pers. comm.)- De Waal (1978)
records a specimen from Bloemhof Dam, on the Orange Free State border, and it may enter
the Cape in this vicinity as well. Basically, though, its distribution extends north of the region
under discussion through the savannah regions of southern and eastern Africa. Broadley (1966)

150

CHECKLIST OF CAPE LIZARDS

and Hillenius (1959) have discussed the taxonomy of the many described races of C. dilepis.
All are of doubtful validity, with the exception of C. d. ruspolii Boettger from Somalia. Tri-
nomials are thus retained.

Family: Agamidae

Represented in Africa by only two genera, Uromastix and Agama. Uromastix is terrestrial
and restricted to Saharan Africa, whilst Agama has arboreal, rupicolous and terrestrial forms
and ranges throughout most of the savannah regions of the continent.

Agama Daudin

Represented in the Cape Province by three species, although the taxonomy of one species, A.
hispida, is chaotic and it may be polyphyletic.

Agama atra atra Daudin

One of the very few lizards the distribution of which encompasses the whole of the Cape
Province. Trinomials are necessary as Mertens (1955) has transferred knobeli (known from
Great Namaqualand) from anchietae to atra.

Agama anchietae Bocage

Similar in habits to atra, and extending from the north-western Cape Province into Nami-
bia.

Agama hispida (Linnaeus)

FitzSimons (1943) recognized six subspecies of hispida, but many of these forms occur at
the same localities. Broadley (1966) rejected all subspecies pending a full revision of the spe-
cies. This is presently being undertaken by Dr G. McLachlan (South African Museum). It
appears that aculeata is a full species and differs from hispida in having an enlarged ear hole.
The relationship of the other forms to these two species requires further investigation
(McLachlan, per. comm.).

INFRAORDER: SCINCOMORPHA

These lizards are almost cosmopolitan in distribution and diurnal in habit. The activity
patterns of many fossorial forms, however, require investigation. Six families are included of
which three families have representatives in southern Africa, i.e. the Lacertidae, Scincidae and
Cordylidae.

Family: Lacertidae

A large family that is common in the Palaearctic region and is well represented in the arid
and savannah areas of the Afro-tropical region. Six genera occur in the Cape Province.

“Lacerta” australis Hewitt

The rarest lizard in the Cape Province, known only from the type collected on the Mat-
roosberg near Ceres. Arnold (1973) has suggested that this species and “ Lacerta ” rupicola
FitzSimons from Lake Fundusi in the northern Transvaal are not congeneric with Palaearctic
Lacerta sensu stricta. It is probable that a new genus will have to be erected to accommodate
them. Their relationship to Tropidosaura is presently being investigated.

Tropidosaura Fitzinger

A small genus of terrestrial, montane lacertids that are endemic to southern Africa. Four
species are currently recognized, two of which occur in the Cape Province.

151

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 11, APRIL 1981

Tropidosaura gularis Hewitt

A montane form that is now known to extend along the southern Cape Fold Mountains
from Table Mountain to the Port Elizabeth District.

Tropidosaura montana montana (Gray)

A wide-ranging, but poorly-known lizard, with three recognized races. The typical race is
distributed in the western and south-western Cape from Table Mountain north to the Cold
Bokkeveld and east to Heidelberg.

Tropidosaura montana rangeri Hewitt

Known from the tops of the Katberg and Amatola Mountains down to the sea cliffs at
Hamburg in the Eastern Cape Province. The large distribution gaps between this race and the
typical form in the west and T. m. natalensis (described by FitzSimons from the Natal
Drakensberg in 1947) in the east are probably due to poor collecting and zones of intergrada-
tion (if any) have yet to be established.

Nucras Gray

Very secretive lacertids the taxonomy of which is confused due to the paucity of speci-
mens and the great overlap in “diagnostic” characters in the genus. Broadley (1972) has re-
viewed the tesselata group and placed great emphasis on colour pattern.

Nucras lalandi (Milne-Edwards)

A large Nucras with a thick tail. Widely distributed in the eastern regions from Knysna to
Natal and the Transvaal. Incorrectly called N. delalandii by FitzSimons (1943).

Nucras taeniolata taeniolata (A. Smith)

Erected to full specific status by Broadley (1972). Restricted to a small isolated, possibly
relic population centred around Grahamstown in the Eastern Cape Province.

Nucras taeniolata ornata Gray

A wide-ranging race that shows considerable regional variation. In the Cape Province it
enters the Kimberley region and reaches as far south as Burgersdorp. Nucras intertexta holubi
and N. i. damarana are synonyms of ornata (Broadley, 1972).

Nucras intertexta (A. Smith)

A Kalahari species that extends into the northern Cape Province in the Kimberley region.
Nucras tessellata tessellata (A. Smith)

A western Cape form that extends from Namibia to Worcester. A rufous-grey colour
variety (var elegans ) occurs along the eastern border of this subspecies (i.e. Caledon, Mary-
dale, Calvinia, etc.). A further colour variety (var. “T”) shows affinities in its dorsal colour
pattern to taeniolata taeniolata of the eastern Cape. As it occurs in sympatry with typical tessel-
lata at Klipfontein and Steinkopf it cannot be treated as a distinct subspecies. Broadley (1972)
has treated this form as a variety of typical tessellata, although he notes it may be a distinct
species or a western subspecies of taeniolata. The problem remains unsolved.

Nucras tessellata livida (A. Smith)

Inhabiting the Karoo areas of the Cape Province, from Matjesfontein to Port Elizabeth. It
separates the typical races of tessellata and taeniolata and obscures the relationship of tessellata
var “T” (see above).

CHECKLIST OF CAPE LIZARDS

Eremias Wiegmann

A very wide ranging genus previously considered to extend throughout Africa into south-
eastern Europe and through western and central Asia to China and Korea. Boulenger (1921)
in his Monograph of the Lacertidae divided Eremias s.l. into 5 “Sections” (Subgenera). Those
relevant to southern Africa are Lampreremias (including lugubris ) and Mesalina (including
namaquensis, undata , lineoocellata, laticeps and burchelli). Eremias s.s. is restricted to south-
east Europe and Asia. Scherbach (1975) treats Boulenger’s “Sections” as full genera, belong-
ing to the subfamily Eremiainae. Loveridge (1957) pointed out that the earliest available
generic name for southern African “ Eremias ” is Heliobolus Fitzinger 1843 (type by original
designation: Lacerta lugubris ). At present it seems best to recognize Heliobolus and Mesalina
as subgenra of Eremias pending an investigation of their relationships (Broadley, pers.
comm.).

Eremias (Heliobolus) lugubris (A. Smith)

Distributed throughout the Kalahari and neighbouring regions and recently recorded in
the Cape Province from near Vryburg and Tweerivieren.

Eremias (Mesalina) namaquensis Dumeril and Bibron

Extending from Little Namaqualand across the northern regions of the Cape Province to
Graaff-Reinet and Cradock in the eastern Cape.

Eremias (Mesalina) laticeps A. Smith

Extending across the karroid regions, from Graaff-Reinet to Little Namaqualand, and
across the Orange River into Great Namaqualand. Mertens (1955) revived the name for Lacer-
ta capensis (= Eremias capensis) as the name is preoccupied by Lacerta capensis Sparrmann
1783 (= Varanus n. niloticus).

Eremias (Mesalina) undata undata (A. Smith)

Closely related to E. namaquensis and mainly restricted to Namibia, although it ranges
into Little Namaqualand and the northwest Cape Province. A colour variety inornata was
recognized by FitzSimons (1943) in the southern part of the range of undata. However, it
occurs in a scattered pattern through the range of undata and cannot be considered as a valid
subspecies. Trinomials are required, nonetheless, as Mertens (1955) has described two subspe-
cies ( rubens and gaerdesi) from the Waterberg and north-west regions of Namibia respectively.

Eremias (Mesalina) burchelli Dumeril and Bibron

A montane form that extends across the southern Cape from Cape Agulhas and the
Cedarberg to Queenstown and on to the Natal Drakensberg.

Eremias (Mesalina) lineoocellata lineoocellata Dumeril and Bibron

A wide-ranging subspecies that enters the Cape Province in the northern regions around
Kimberley.

Eremias (Mesalina) lineoocellata pulchella Gray

The main subspecies in the Cape Province, extending in a wide belt from Little Namaqua-
land and the Western Cape Province, through the central and southern karroid areas to the
Eastern Cape Province. Replaced in Great Namaqualand by the spot-less subspecies inocellata
(Mertens, 1955).

Ichnotropis Peters

A small genus of rough-scaled, terrestrial lacertids, represented in southern Africa by
three species.

153

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 11, APRIL 1981
Ichnotropis squamulosa Peters

Broadley (1976a) has shown that these lizards are very short-lived, hatching and growing
to sexual maturity in six months and then dying soon after reproduction. I. capensis has a
similar abbreviated life history, but the cycles of the two species are phased such that their
eggs hatch at different times. This avoids competition for the same resources. Mainly distri-
buted north of the Cape Province, it has recently been recorded from the Kalahari Gemsbok
National Park (Pianka, 1971).

Meroles Gray

Mertens (1955) demonstrated that Scaptira Wiegmann, to which FitzSimons (1943) refer-
red the following species, is restricted to Central Asia. The next available generic name is
that of Meroles Gray 1938. Six monotypic species are recognized, of which four range into the
Cape Province.

Meroles suborbitalis (Peters)

Extending from the central Karoo west and north through Little Namaqualand and Gor-
donia.

Meroles knoxii (Milne-Edwards)

A common lizard throughout its range which extends from the western Province north-
wards through Little Namaqualand to Namibia (Liideritz). Haacke (1965) has demonstrated
that the northern subspecies pequensis (Hewitt) is invalid.

Meroles ctenodactylus (A. Smith)

A larger, slower-moving species that extends from Little Namaqualand and the adjacent
Karoo to just north of the Orange River. It reaches as far south as the Zout Rivier on the west
coast (McLachlan, in litt. ).

Meroles cuneirostris (Stauch)

A psammophilous species well adapted to life in the Namib Desert, the range of which
just extends into Little Namaqualand (Brandkaros, Holgat River Mouth, etc.).

Family: Scincidae

Greer (1970a) has revised the family on the basis of osteological characters. Four subfami-
lies are recognized: Scincinae (with approximately 28+ genera, 182 species), Feylininae (2
genera, 4 species), Acontinae (3 genera, 16 species) and Lygosominae (40+ genera, 600+ I
species). The Scincinae are considered to have independently given rise to all three other
subfamilies. The Acontinae and Feylininae are both viviparous, burrowing groups, whilst the
Lygosominae is composed of the most numerous, diverse and advanced skinks. All subfami-
lies, with the exception of the Feylininae (which is restricted to west and central Africa), are
found in the Cape Province.

Acontinae

Distributed throughout southern Africa, with an isolated population in south-east Kenya.
Malagasy Acontias (holomelas and hildebrandti) were referred to the new scincine genus Mala-
contias by Greer (1970a). Three genera are recognized in southern Africa; Acontias (7 spe-
cies), Typhlosaurus (8 species) and the monotypic Acontophiops.

Acontias Cuvier

The genus was revised by Broadley and Greer (1969), who recognized seven species, all of
which are found in the Cape Province.

154

CHECKLIST OF CAPE LIZARDS

Acontias plumbeus Bianconi

A very large burrowing skink that inhabits the tropical lowlands of south-eastern Africa.
An isolated, possibly relic, population occurs in the eastern Cape, around East London.

Acontias breviceps Essex

Previously treated by FitzSimons as a subspecies of A. plumbeus. It has a discontinuous
distribution at high altitudes along the eastern escarpment and occurs in the eastern Cape
around the Hogsback and Katberg.

Acontias gracilicauda gracilicauda Essex

Treated as a subspecies of A. plumbeus by FitzSimons, but differs from it -in having a
small post-temporal fenestra and a strongly tapering tail. Found in the eastern Cape as far
south as Port Elizabeth, extending northwards into higher altitudes on the high veld of the
Orange Free State and Transvaal and into the northern Cape as far as Kuruman.

Acontias gracilicauda namaquensis Hewitt

A poorly defined western race found in Little Namaqualand. It was also previously treated
as a subspecies of A. plumbeus by FitzSimons (1943).

Acontias percivali tasmani Hewitt

Although previously treated as a race of A. plumbeus by FitzSimons (1943), tasmani dif-
fers from this species in having no post-temporal fenestra. The species has a remarkable dis-
continuous distribution and typical percivali is found in the vicinity of Voi in south-eastern
Kenya. The race tasmani has a very restricted range in the Eastern Cape Province from Port
Elizabeth to King William’s Town.

Acontias meleagris meleagris Hewitt

A very variable form that ranges along the south-west Cape coast from Cape Town to
Knysna and inland as far as Fraserburg in the Karoo.

Acontias meleagris orientalis Hewitt

An eastern race that extends from Cape St Francis to the Transkei. FitzSimons recognized
a separate race lineicauda from the Port Elizabeth — Grahamstown region. Broadley and Greer
(1969) conservatively treated this form as a localized “morph” of orientalis but noted that it
was smaller and more slender and inhabits the dry habitats of the Sundays River and Great
Fish River valleys. Its status has still to be resolved.

Acontias lineatus lineatus Peters

Inhabiting the interior of the Western Cape Province from Matjesfontein north to Kaka-
mas in the west and Hopetown in the east, then across the Orange River as far north as
Liideritz and Keetmanshoop. The race orangensis Hewitt is a synonym of typical lineatus
(Broadley and Greer, 1969).

Acontias lineatus tristis Werner

Restricted to Little Namaqualand. Treated as only a variety of lineatus by FitzSimons but
elevated to subspecific level by Broadley and Greer (1969).

Acontias lineatus grayi Boulenger

A very localized race (known only from the type locality — Graafwater, between Clanwil-
liam and Lambert’s Bay, south-western Cape), the status of which is still problematic.

155

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 11, APRIL 1981
Acontias litoralis Broadley and Greer

A small species related to A. lineatus and restricted to the western Cape coastal strip from
Daberas to Elandsbaai.

I

Typhlosaurus Wiegmann

Highly specialized, limbless, fossorial skinks, of which FitzSimons (1943) recognized eight
species. Haacke (1964) described a greatly attenuated species ( braini ) from the Namib Desert,
whilst Broadley (1968) in revising the whole genus, placed plowesi in the synonymy of meyeri,
and thus returned the total number of species in the genus to eight. The genus ranges across
the northern regions of southern Africa, with five species occurring in the western Cape re-
gion.

Typhlosaurus lineatus lineatus Boulenger

A Kalahari species that extends into the northern Cape Province in the Kalahari Gemsbok
National Park and thence south to Witdraai and Vanzylsrus.

Typhlosaurus gariepensis FitzSimons

Related to lineatus and found in the northern Cape Province from Upington to the Kala-
hari Gemsbok National Park.

Typhosaurus meyeri Boettger

An attenuated species that inhabits the coastal regions of southern Namibia, from Liider- i
itz south to the Richtersveld and Little Namaqualand. Broadley (1968) synonomized
T. plowesi FitzSimons with this species.

Typhlosaurus caecus (Cuvier)

Another attenuated species, the one that extends the furthest south. Distributed in the
coastal areas of the Western Cape Province from Edgemead, Cape Town (Visser, 1979),
northwards, possibly as far as Port Nolloth (where it is known from a single specimen), but
certainly as far as Lambert’s Bay.

Typhlosaurus vermis Boulenger

Closely related to, and possibly conspecific with T. caecus. Extending from Alexander Bay
south to the Spoegrivier and inland as far as Putsonderwater. Broadley’s (1968) reference to
Lambert’s Bay is questioned by McLachlan (in lift.).

!

Scincinae

Basically Old World in distribution, with six genera occurring in sub-Saharan Africa. Only |
Scelotes extends into the Cape Province.

Scelotes Fitzinger

A large group of semi-fossorial to fossorial small skinks. The genus shows an interesting
progression from species with small but well-developed limbs to limbless species. It has under-
gone extensive partitioning (Greer, 1970b) since FitzSimons’ monograph (1943). However, all
the new genera are extralimital to this discussion. There has been no alteration in the tax-
onomy of the Cape species since FitzSimons’ monograph (1943) and the genus remains in
urgent need of revision.

Scelotes anguina (Boulenger)

Coastal areas of the Eastern Cape Province from Port Elizabeth to Peddie.

156

CHECKLIST OF CAPE LIZARDS

Scelotes capensis (A. Smith)

Known from Little and Great Namaqualand.

Scelotes caffer (Peters)

Apparently common in the Fish River Valley of the Eastern Cape Province and extending
through the Karoo to Little Namaqualand.

Scelotes kasneri FitzSimons

Known only from the type locality, i.e. Lambert’s Bay, Clanwilliam District, Western
Cape Province.

Scelotes bipes bipes (Linnaeus)

Restricted to the south-western Cape Province from Mossel Bay to the Cape Peninsula.
Scelotes bipes sexlineatus (Harlan)

Recorded from Port Nolloth and Kleinsee in Little Namaqualand and south to Clanwil-
liam and Calvinia.

Scelotes gronovii (Daudin)

Inhabiting the coastal regions of the south-western Cape Province from Dassen Island to
Lambert’s Bay.

Scelotes bicolor (A. Smith)

Although the type locality of this species is given as Little Namaqualand, no further speci-
mens have been collected. FitzSimons (1943) questioned the validity of the species as no other
legless Scelotes are known from the western Cape. It is very closely related to, and possibly
conspecific with, S. arenicola which is known from Zululand and Mozambique.

Lygosominae

An extremely widespread group that is particularly well-represented in, and probably ori-
ginated within, the Australasian region. In southern Africa it is represented by the genera
Mabuya, Panaspis (= Ablepharus Fitzinger and Afroablephams Greer), Lygosoma (= Riopa)
and Cryptoblepharus . Only the first occurs in the Cape Province.

Mabuya Fitzinger

A wide-ranging genus distributed throughout Africa and Madagascar and found also in
south-eastern Asia, Central and South America and the Caribbean. The most common skinks
in the Cape Province, with no less than nine recognized species, some of which have local
races.

Mabuya capensis (Gray)

A large, thick-bodied skink that is distributed throughout the Cape Province, although it
is rarer in the arid, western regions. It is now known to extend north of the Limpopo as far as
Zambia.

Mabuya occidentalis (Peters)

Closely related to the previous species, but of generally lighter build. In the Cape it is
restricted to karroid regions, from Cradock and Beaufort West, extending north-west through
Little Namaqualand and Namibia to southern Angola.

157

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 11, APRIL 1981

Mabuya homalocephala homalocephala (Wiegmann)

The nominate race of this medium-sized, terrestrial species extends along the coastal re-
gions and Cape Fold Mountains of the western and south-western Cape, from Little Namaqua-
land to Knysna.

Mabuya homalocephala peringueyi Boulenger

A northern race found in Little Namaqualand. The record from Namibia is incorrect
(McLachlan, in litt.).

Mabuya homalocephala smithii (Gray)

Extending along the eastern Cape coastal regions from Garcia Pass, Langeberg, to the
Transkei. It is now known to extend inland as far as Rubidge Kloof, north of Graaff-Reinet.
There appears to be a fairly extensive overlap with the nominate race from Knysna to Swellen-
dam that requires investigation.

Mabuya striata complex

The taxonomy of this problematic group has recently been stabilized by Broadley’s revi-
sion of the complex (1977a). Mertens (1955) split the previously monotypic striata (FitzSimons,
1943) into three races, i.e. nominate striata, sparsa and spilogaster. Sympatry between spilogas-
ter and sparsa in the Kalahari Gemsbok National Park was noted by Broadley (1969) who in
his subsequent revision of the complex (1977a) recognized two further races of striata , i.e.
wahlbergi and punctatissimus . Three of these taxa occur in the Cape Province.

Mabuya striata punctatissimus (A. Smith)

Extending just into the northern Cape in the vicinity of Mafeking and Aliwal North.
Intergrades between this and the following race occur in the vicinity of Vanzylrus.

Mabuya striata sparsa Mertens

A robust, dark subspecies common in the Kalahari Gemsbok National Park and extending
into neighbouring Namibia and Botswana.

Mabuya spilogaster (Peters)

A robust, speckled species that is commonly found on Acacia trees and is thus ecologically
separated from sympatric striata. It extends from southern Angola through Namibia and west-
ern Botswana into the northern Cape Province, reaching as far east as Kimberley.

Mabuya sulcata sulcata (Peters)

Considered monotypic by FitzSimons (1943). Mertens (1955) recognized a new race, nigra
from Liideritz in Namibia, and resurrected Boulenger’s ansorgii as a separate race for Angolan
material. Only the nominate race occurs in the Cape and is found throughout the Karoo and
adjoining regions, from Little Namaqualand to Perdepoort in the Klein Winterhoekberg.

Mabuya variegata variegata (Peters)

The confused taxonomy of these small skinks is too lengthy to be dealt with here and,
moreover, has been admirably reviewed by Broadley (1975). In summary the species dama-
rana and varia have long been misidentified. Two forms occur and Mertens (1955) resurrected
the name longiloba after demonstrating that the type of damarana was a synonym of varia.
Subsequently Broadley (1975) has shown that longiloba is a synonym of Euprepes variegatus,
and the name must change again. He also recognized a northern subspecies. The nominate
race extends along the west coast of southern Africa south of the Cunene River to as far south
as the Clanwilliam District in the western Cape. It also occurs eastwards through the Karoo to
Kimberley and the Albany District.

158

CHECKLIST OF CAPE LIZARDS

Mabuya variegata punctulata (Bocage)

A northern and eastern race that extends into the northern Cape Province in the region of
Mafeking and the Kalahari Gemsbok National Park. In addition de Waal (1978) refers to this
race specimens from Rietfontein on the border of the Orange Free State and the north-eastern
Cape Province even though typical variegata occurs at Zastron. It is possible that punctulata
and variegata are not conspecific.

Mabuya varia (Peters)

Due to confusion with damarana (= variegata , see above) this species now has a much
more restricted range in southern Africa than that given in FitzSimons (1943). The species
ranges southwards from Sudan and Somalia but only extends into the south-eastern corner of
the Cape Province in the vicinity of Grahamstown and Port Elizabeth.

Mabuya acutilabris (Peters)

A small, well-defined western species that extends into the Cape Province only in the
northern tip of Little Namaqualand.

Family: Cordylidae

The family now includes the plated lizards and related forms which were treated by Fitz-
Simons (1943) as a separate family but should now be treated as a subfamily of the Cordylidae
(McDowell and Bogert, 1954). The family is closely related to the Lacertidae but there is
uncertainty as to whether it is ancestoral to, or derived from these lizards. The family is basi-
cally distributed throughout sub-Saharan Africa, although a number of Gerrhosaurines occur
on Madagascar.

Cordylinae

Four genera are recognized, of which all occur in the Cape Province.

Chamaesaura Schneider

A group of elongate, serpentiform cordylids that show varying degrees of limb-reduction
and which are amazingly conservative in morphology. Three species are recognized, of which
two occur in the Cape Province.

Chamaesaura aenea (Wiegmann)

A montane form that reaches its southern limit in the vicinity of the Amatola Mountains
near Cathcart and Tarkastad in the north-eastern Cape Province.

Chamaesaura anguina anguina (Linnaeus)

The most widespread Chamaesaura in the Cape Province, extending in a wide belt across
the more humid regions of the southern Cape from Kei Mouth to Cape Town. It is separated
by nearly a 1 000 km from the northern race C. a. tenuior (which was treated as a separate
species by FitzSimons, 1943) found in East Africa.

Platysaurus A. Smith

These greatly flattened, highly specialized corylids have undergone rapid speciation in the
rocky areas of south-eastern Africa, and have been extensively studied by Broadley (1978). A
single, relic species occurs along the lower Orange River, 900 km from its nearest neighbour,
Platysaurus g. guttatus , in the Limpopo River basin.

159

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 11, APRIL 1981
Platysaurus capensis A. Smith

Found along the lower Orange River, west of Aughrabies Falls, extending into Namibia
along the Fish River Canyon and southwards to Garies in the Cape Province. Broadley (1978)
has discussed the possibility that the population at Aughrabies Falls may deserve subspecies
recognition.

Pseudocordylus A. Smith

There have been numerous alterations to the concept of Pseudocordylus since FitzSimons’
(1943) monograph and the status of the genus and its constituent species has still to be re-
solved. Pseudocordylus is obviously closely related to Cordylus but the boundaries between
the two genera are vague and they may indeed be congeneric. In addition, the species present-
ly placed in Pseudocordylus may in fact be polyphyletic and their similarities due to conver-
gence. A revision is long overdue but has been inhibited by the loss of Andrew Smith’s import-
ant types and the lack of specimens from key areas. Field studies of the Cape forms have
begun but it will be some time before sufficient material is collected for a thorough revision.
FitzSimons (1943) recognized only two species, with a total of five subspecies. However, cur-
rent concepts of the genus would include as many as six species, with up to 10-11 subspecies.

Pseudocordylus microlepidotus microlepidotus (Cuvier)

Although Loveridge (1944b) listed the range of this large lizard as being restricted to the
Cape Peninsula mountains, FitzSimons (1943) knew it to extend as far as Prince Alfred’s Pass
in the Outeniqua Mountains.

Pseudocordylus microlepidotus namaquensis Hewitt

Recorded by FitzSimons (1943) from the widely separated localities of Little Namaqua- 1
land and Beaufort West. The status of this subspecies needs investigation.

Pseudocordylus microlepidotus fasciatus (A. Smith)

Considered by Smith to be a full species, this race is restricted to the Eastern Cape Pro-
vince from Graaff-Reinet north to Colesberg and east to Tsomo in the Transkei. Specimens i
collected recently from the Port Elizabeth-Suurberg region may be referable to P. algoensis i
A. Smith which has usually been considered a synonym of fasciatus (FitzSimons, 1943; Loveridge,
1944b).

Pseudocordylus melanotus subviridis (A. Smith)

Previously treated as P. s. subviridis by FitzSimons (1943), and known from the Katberg
and Amatola Mountains in the Eastern Cape Province. The status of Pseudocordylus with
elongate temporal scales has caused tremendous problems, and FitzSimons (1943), Loveridge
(1944b), Broadley (1964) and de Waal (1978) have all held different views. The problem derives
from the loss of Smith’s types of melanotus and subviridis, and differing interpretations of
Smith’s illustrations of these types. As melanotus has paragraph preference over subviridis, de
Waal (1978) has elevated melanotus to a full species and treated subviridis as a subspecies of it.
Northern Cape specimens from near Herschel (Broadley, 1964) may be referable to the
nominate race, P. m. melanotus. Loveridge (1944b) referred Pseudocordylus from the Amatola
Mountains to his new species P. langi , but this species is restricted to the Natal Drakensberg at
Mont-aux-Sources and Cathedral Peak (Broadley, 1964).

Pseudocordylus capensis capensis (A. Smith)

This species was transferred to Pseudocordylus from Cordylus by Loveridge (1944b).
Although this species and the closely related robertsi (see below) share some characters with
Pseudocordylus (i.e. the neck is covered with granules instead of scales and the dorsal scales

160

CHECKLIST OF CAPE LIZARDS

lack osteoderms), these may be due to convergence. It is known from the Hottentots Holland
Mountains in the western Cape eastwards to Robinson’s Pass in the Attaqua Mountains.

Specimens collected recently in the Kammanassieberg share features with this and the
following subspecies.

Pseudocordylus capensis robertsi (van Dam)

Although FitzSimons (1943) treated this form as a subspecies of Cordylus capensis, Love-
ridge (1944b) re-elevated it to specific status, whilst transferring it to Pseudocordylus. It is not
found in sympatry with capensis and was previously known only from the Van Rhynsdorp-
Klawer District. Specimens from Pakhuis Pass in the Cedarberg are intermediate between
capensis and robertsi and the latter is here treated as only a race of capensis, pending a full
revision of the genus.

Cordylus Laurenti

Perhaps the most typical lizard of the Cape Province, although in number of species it is
surpassed by other genera, notably Pachydactylus and Mabuya. The genus is basically re-
stricted to southern Africa, although one species reaches as far north as Ethiopia. The tax-
onomy is chaotic and sorely in need of revision. Six, possibly seven, species occur in the Cape
Province.

Cordylus giganteus A. Smith

This species is listed more in hope than in fact. Boulenger (1910) recorded a specimen
from Colesberg. However, de Waal (1978) failed to find this species in the southern half of the
Orange Free State and its presence in the northern Cape is therefore doubtful. If it did pre-
viously exist in the northern Cape and is now extinct, it would be a worthy candidate for
re introduction, as is now being done with the Python (Python sebae) in the eastern Cape.

Cordylus cataphractus Boie

A well-defined species that inhabits the Western Cape Province from Little Namaqualand
to Matjesfontein.

Cordylus coeruleopunctatus (Methuen and Hewitt)

Another well defined species. It is the morphological antithesis of cataphractus, being
elongate and finely-scaled. It has many resemblances to Pseudocordylus capensis and may be
close to the ancestral form that gave rise to the latter radiation. Recorded from the southern
slopes of the Attaqua and Outeniqua Mountains, from Mossel Bay to Witelsbos Forestry Sta-
tion.

Cordylus macropholis (Boulenger)

A small species inhabiting the western Cape coastal belt from Little Namaqualand to near
Ysterfontein (McLachlan, 1979a).

Cordylus peersi (Hewitt)

A melanistic form known only from the Little Namaqualand region.

Cordylus cordylus cordylus (Linnaeus)

This species, more than any other in southern Africa, is a taxonomist’s nightmare. Fitz-
Simons (1943) recognized five subspecies, whilst Loveridge (1944b) increased this number to
nine. The most obvious and pressing problem remains the status of the small, melanistic,
rugose niger, that occurs on the Cape Peninsula but also at Saldanha Bay and the adjacent
Islands. Loveridge (1944b) considered it to be a separate subspecies. However FitzSimons

161

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 11, APRIL 1981

(1943) treated it as a variety as isolated melanistic individuals occurred throughout the range of
cordylus. Conservatively, FitzSimons’ (1943) views are followed here, although the problem is
not settled. Visser (1971) has suggested that niger may be a subspecies of C. peersi. Such a
relationship may be possible, although due to priority peersi would become a subspecies of
niger. The identity of specimens from Van Rhynsdorp, considered by FitzSimons (1943) to be
typical cordylus , but out of the ranges of any of Loveridge’s (1944) races of cordylus , may give
some insight into the problem. With these provisos, the range of the nominate race is currently
considered to extend in a wide belt across the Cape from Robben Island to Kokstad in East
Pondoland.

Cordylus cordylus minor FitzSimons

As its name implies, a dwarf form, presently known only from the type locality at Matjes-
fontein. Its relationship to the small rupicolous cordylids of Little and Great Namaqualand
(i.e. lawrenci, pustulatus , namaquensis and campbelli) and to other similar cordylids recently
collected from the Nieuweveldberg and Sneeuberg remains to be resolved.

Cordylus lawrenci (FitzSimons)

A small, rupicolous form that has been treated by Loveridge (1944b) as a race of cordylus
but which shows affinities to minor (see previous discussion). Known only from the type local-
ity (Lekkersing, Little Namaqualand) and from near Port Nolloth (Visser, 1979).

Cordylus tasmani (Power)

Although treated by both FitzSimons (1943) and Loveridge (1944b) as a subspecies of I
cordylus , this form probably deserves specific recognition. Known from scattered localities in i
the eastern Cape around Uitenhage, its full range has still to be determined.

Cordylus polyzonus polyzonus (Parker)

A wide ranging form found in a wide variety of colours. The Namibian form, jordani , that l[
was treated as a full species by FitzSimons (1943) has been reduced to a subspecies of polyzo-
nus by Loveridge (1944b). The non nate race extends in a wide belt across the west, north and
central Cape Province from Aliwal orth to Port Nolloth. It does not occur south of the Cape
Fold Mountains.

Gerrhosaurinae

Represented by four genera in Africa and two on Madagascar. Three genera occur in the
Cape Province.

Gerrhosaurus Wiegmann

The genus is very ancient and is the possible ancestor of most of the other cordylids. A
fossil lizard from the Lower Miocene of Mfanganu Island, Lake Victoria, Kenya, has been
referred to the extant species G. major (Estes, 1962). Two species occur in the Cape Province.

Gerrhosaurus flavigularis Wiegmann

Recorded from scattered localities throughout the Cape with the exception of the north-
western region. The species is monotypic as Broadley (1966) has demonstrated that G. f.
fitzsimonsi Loveridge from East Africa is not valid.

Gerrhosaurus typicus (A. Smith)

A rare species restricted to the western Cape from Ceres and Beaufort West (3122 CD) to
Little Namaqualand.

162

CHECKLIST OF CAPE LIZARDS

Table 1

Synopsis of the Lizards of the Cape Province

Family

Genera

Species

Species and
Subspecies

Taxonomic*

changes

Gekkonidae

9

28

42

26

Varanidae

1

2

2

2

Chamaeleonidae

2

9

9

7

Agamidae

1

3

3

3

Lacertidae

6

18

22

13

Scincidae

4

28

38

18

Cordylidae

7

20

27

15

Total

30

108

143

84

*Taxa affected by taxonomic revisions subsequent to FitzSimons (1943)

Table 2

Changes in the Nomenclature of Cape Province Pachydactylus

FitzSimons, 1943

Loveridge, 1947

Current status

purcelli

purcelli

serval purcelli

microlepis

maculatus

maculatus

austeni

austeni

austeni

latirostris

mariquensis latirostris

mariquensis latirostris

mariquensis mariquensis

mariquensis mariquensis

mariquensis mariquensis

mariquensis macrolepis

mariquensis macrolepis

mariquensis macrolepis ( = amoenu

ocellatus

geitje

geitje

amoenus

amoenus

amoenus (?)

monticolus

monticolus

geitje (?)

maculatus maculatus

maculatus

maculatus maculatus

maculatus albomarginatus

maculatus

maculatus oculatus

montanus onscepensis

serval

serval onscepensis

weberi weberi

weberi weberi

weberi weberi

weberi gariesensis

weberi gariesensis

weberi weberi

punctatus amoenoides

punctatus amoenoides

punctatus amoenoides

rugosus

rugosus rugosus

rugosus rugosus

mentalis

capensis capensis

mentalis ?

capensis capensis

capensis capensis

capensis capensis

capensis labialis

capensis labialis

labialis ?

capensis formosus

capensis formosus

rugosus formosus

capensis barnardi

capensis barnardi

rugosus barnardi

capensis oculatus

maculatus

maculatus oculatus

namaquensis

namaquensis

namaeen s

bibroni

bibroni bibroni

bibroni

laevigatus

laevigatus laevigatus

laevigatus laevigatus

163

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 11, APRIL 1981
Tetradactylus Merrem

Considered by FitzSimons (1943) to be endemic to southern Africa but now known to
reach as far north as Katanga, western and northern Zambia, and south-eastern Tanzania.
Three species and five subspecies occur in the Cape Province.

Tetradactylus seps seps (Linnaeus)

Occurring in the south-western Cape and extending eastwards as far as Cape St Francis.
Tetradactylus seps laevicauda Hewitt

Known from Stutterheim and the Katberg in the north-eastern Cape and extending east-
wards through the south-eastern Orange Free State to Natal.

Tetradactylus tetradactylus tetradactylus (Lacepede)

Restricted to the montane regions of the south-western Cape from Table Mountain to the
Cedarberg and Swartberg.

Tetradactylus tetradactylus bilineatus Hewitt

A rare eastern race known only from the type locality at Burghersdorp in the eastern
Cape.

Tetradactylus tetradactylus fitzsimonsi Hewitt

A similarly poorly known form apparently restricted to the southern Cape Province from
George to Port Elizabeth.

Cordylosaurus Gray

A monotypic genus of small, beautifully coloured gerrhosaurids, restricted to the western i
regions of southern Africa. Two species, one with two races, were recognized by FitzSimons
(1943), although he noted that the only available type of subtesselatus was possibly synono-
mous with trivittatus. More importantly it had priority. He retained the two species, however,
even though he was aware that Loveridge (1942) had synonomized them. Even more surpri-
singly, he still recognized the poorly-defined southern race C. trivittatus australis Hewitt (dis-
tinguished by a fused interparietal and parietal, and narrow dorsal light bands), although ear- j
lier Parker (1936) had shown these characters to be unreliable. Loveridge (1942) recognized
only a single, variable species, and this has been subsequently accepted (Mertens, 1955, 1971).

Cordylosaurus subtessellatus (A. Smith)

A small (usually less than 150 mm) gerrhosaurid that has a bright blue posterior region
(although the type lacks this). Ranging from southern Angola throughout the western regions
of Namibia and Little Namaqualand and reaching its southern limit at Matjesfontein in the i
Cape Province.

DOUBTFUL RECORDS

In 1913 Hewitt and Power reported on a collection of lizards from the McGregor
Museum, Kimberley. Two species listed by them for the Cape are of doubtful occurrence.

Lygosoma sundevalli sundevalli (A. Smith)

This small, semi-fossorial skink (for long known by the name Riopa sundevalli) has recent-
ly been transferred to the large Afro-Asian genus Lygosoma (Greer, 1977). It extends in a
wide belt across the northern regions of southern Africa, being replaced along the Mozambi-
que flood plain by the large, speckled species, L. afer (Broadley, 1966b). The Kyky record
given by Hewitt and Power (1913) is based on a dried scrap purchased from a Botswana

CHECKLIST OF CAPE LIZARDS

bushman (who considered it a much-prized antidote for snake bite). The specimen may thus
have been obtained further north. No confirmed records for the Cape exist, although Broadley
(1967b) collected specimens only 20 miles north of the tip of the Kalahari Gemsbok National
Park.

Cordylus tropidosternum jonesi (Boulenger)

Listed by Hewitt and Power (1913) as occurring at Kimberley, but having been introduced
during the diamond-rush with timber brought to the mining camps. No recent records confirm
its continued existence. Broadley (1966a) treated jonesi as a southern race of tropidosternum.

ACKNOWLEDGEMENTS

The author is grateful to Dr D. G. Broadley (Umtali Museum), Dr G. McLachlan (South
African Museum) and Mr W. Haacke (Transvaal Museum) for very helpful discussions con-
cerning the taxonomic status and distribution of certain Cape lizards. J. Greig and R. Boycott
(Jonkershoek Research Station) both supplied new locality records.

REFERENCES

Arnold, E. N. 1973. Relationships of the Palaearctic lizards assigned to the genera Lacerta, Algyroides and Psammo-
dromus (Reptilia: Lacertidae). Bull. Br. Mus. nat. Hist. Zoology 25 (8): 291-366.

Boulenger, G. A. 1921. Monograph of the Lacertidae. Vol. 2. London: British Museum (Natural History).

Brain, C. K. 1962. A review of the gecko genus Ptenopus with a description of a new species. Cimbebasia 1: 1-18.
Broadley, D. G. 1964. A review of the crag lizards (Genus Pseudocordylus) of Natal. Ann. Natal Mus. 16: 99-110.
Broadley, D. G. 1966a. The herpetology of South East Africa. Ph.D. Thesis. Univ. Natal, Pietermaritzburg.

Broadley, D. G. 1966b. A review of the Riopa sundevalli group (Sauria: Scincidae) in southern Africa. Amoldia, Rhod.
2 (34): 1-7.

Broadley, D. G. 1967a. The life cycles of two sympatric species of Ichnotropis (Sauria: Lacertidae). Zool. Afr. 3 (1):
1-2.

Broadley, D. G. 1967b. An expedition to the south-western Kalahari, January-March 1967. J. Herpetol. Assoc. Afr. 2:
19-30.

Broadley, D. G. 1968. A revision of the African genus Typhlosaurus Wiegmann (Sauria: Scincidae). Arnoldia. Rhod. 3
(36): 1-20.

Broadley, D. G. 1969. Two sympatric species of the Mabuya striata complex (Sauria: Scincidae) in the southern
Kalahari. Koedoe 12: 11-14.

Broadley, D. G. 1972. A review of the Nucras tessellata group. (Sauria: Lacertidae). Amoldia. Rhod. 5 (20): 1-36.
Broadley, D. G. 1975. A review of the Mabuva lacertiformis complex in southern Africa (Sauria: Scincidae). Arnoldia.
Rhod. 7 (18): 1-16.

Broadley, D. G. 1977a. A review of the north eastern forms of the Pachydactylus capensis complex (Sauria: Gekkoni-
dae). Arnoldia, Rhod. 8 (18): 1-20.

Broadley, D. G. 1977b. A review of the Mabuya striata complex in South East Africa. (Sauria: Scincidae). Occ. Pap.
natn. Mus. Rhod. Ser. B. 6 (2): 44-79.

Broadley, D. G. 1978. A revision of the genus Platvsaurus A. Smith (Sauria: Cordylidae). Occ. Pap. natn. Mus.
Rhod. Ser. B. 6 (4): 130-185.

Broadley, D. G. and Greer, A. E. 1969. A revision of the genus Acontias Cuvier (Sauria: Scincidae). Amoldia, Rhod.
4 (26): 1-29.

Broadley, D. G., Gans, C. and Visser, J. 1975. Studies on Amphisbaenians (Amphisbaenia: Reptilia) 6. The genera
Monopeltis and Dalophia in southern Africa. Bull. Am. Mus. nat. Hist. 157 (5): 341-485.

De Waal, S. W. P. 1978. The Squamata (Reptilia) of the Orange Free State, South Africa. Mem. Nat. Mus. Bloemfon-
tain. 1: 1-160.

Dowling, H. G. and Duellman, W. E. 1978. Systematic Herpetology: A synopsis of families and higher categories. New
York: HISS Publications.

Estes, R. 1962. A fossil Gerrhosaur from the Miocene of Kenya (Reptilia: Cordylidae). Brevoria 58: 1-10.

FitzSimons, V. F. M. 1943. The Lizards of South Africa. Transv. Mus. Mem. 1: 1-528.

FitzSimons, V. F. M. 1947. Description of new species and subspecies of reptiles and amphibians from Natal, together
with notes on some other little-known species. Ann. Natal Mus. 11: 111-137.

Gans, C. 1978. The characteristics and affinities of the Amphisbaenia. Trans, zool. Soc. Lond. 34 (3): 346-416.

Greer, A. E. 1970a. A subfamilial classification of scincid lizards. Bull. Mus. comp. Zool. Harv. 139: 151-183.

Greer, A. E. 1970b. The systematics and evolution of sub-Saharan Africa, Seychelles and Mauritius scincine scincid
lizards. Bull. Mus. comp. Zool. Harv. 140: 1-23.

165

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 11, APRIL 1981

Greer, A. E. 1977. The systematics and evolutionary relationships of the scincid lizard genus Lygosoma. Jnl. nat. Hist.
11: 515-540.

Haacke, W. D. 1964. Descriptions of two new species of lizards and notes on Fitzsimonsia brevipes (FitzSimons) from
the Central Namib Desert. Scient. Pap. Namib Des. Res. Stn. 25: 1-15.

Haacke, W. D. 1965. Additional notes on the herpetology of South West Africa with descriptions of two new sub-
species of geckos. Cimbebasia 11: 1-40.

Haacke, W. D. 1975. The burrowing geckos of southern Africa, 1 (Reptilia: Gekkonidae). Ann. Transv. Mus. 29 (12):
197-247.

Haacke, W. D. 1976a. The burrowing geckos of southern Africa, 2 (Reptilia: Gekkonidae). Ann. Transv. Mas. 30 (2):
13-28.

Haacke, W. D. 1976b. The burrowing geckos of southern Africa, 3 Reptilia: Gekkonidae). Ann. Transv. Mus. 30 (3):
29-39.

Haacke, W. D. 1976c. The burrowing geckos of southern Africa, 4 (Reptilia: Gekkonidae). Ann. Transv. Mus. 30 (5):
52-70.

Haacke, W. D. 1976d. The burrowing geckos of southern Africa, 5 (Reptilia: Gekkonidae). Ann. Transv. Mus. 30 (6):
71-89.

Hewitt, J. 1937. Descriptions of South African lizards. Ann. Natal Mus. 8 (2): 199-210.

Hewitt, J. and Power, J. H. 1913. A list of South African Lacertilia, Ophidia, and Batrachia in the McGregor
Museum, Kimberley; with field notes on various species. Trans. R. Soc. S. Afr. 3: 147-176.

Hillenius, D. 1959. The differentiation within the genus Chamaeleo Laurenti 1768. Beaufortia 8 (89): 1-92.

Hillenius, D. 1963. Notes on Chameleons II. Chamaeleo laevigularis (L. Muller, 1926), a synonym of Chamaeleo
johnstoni (Boulenger, 1901). Beaufortia 10 (117): 44-47.

Kluge, A. G. 1967. Higher taxonomic categories of Gekkonid lizards and their evolution. Bull. Am. Mus. nat. Hist. 135
(1): 1-59.

Kruger, F. 1978. A description of the Fynbos Biome Project. S. Afr. Nat. Sci. Program. Rept. No. 28: 1-23.

Laurent, R. F. 1964. Reptiles et Amphibiens de 1’ Angola. Publgoes cult. Co. Diam. Angola 67: 11-165.

Loveridge, A. 1942. Revision of the African lizards of the family Gerrhosauride. Bull. Mus. comp. Zool. Harv. 89 (11):
486-543.

Loveridge, A. 1944a. New geckos of the genera Afroedura, new Genus, and Pachydactylus from Angola. Am. Mus.
Novit. No. 1254: 1-4.

Loveridge, A. 1944b. Revision of the African lizards of the family Cordylidae. Bull Mus. comp. Zool. Harv. 95: 1-118.

Loveridge, A. 1947. Revision of the African lizards of the family Gekkonidae. Bull. Mus. comp. Zool. Harv. 98 (1):
1-469.

Loveridge, A. 1957. Checklist of the reptiles and amphibians of East Africa (Uganda, Kenya, Tanganyika, Zanzibar).
Bull. Mus. comp. Zool. Harv. 117 (2): 153-362, i-xxxvi.

McDowell, S. B.. and Bogert, C. M. 1954. The systematic position of Lanthanotus and the affinities of Anguiomor-
phan lizards. Bull. Am. Mus. nat. Hist. 105 (1): 1-142.

McLachlan, G. R. 1979a. Distribution of Cordylus macropholis. J. Herpetol. Assoc. Afr. 20: 5.

McLachlan, G. R. 1979b. The taxonomy of Pachydactylus rugosus. J. Herpetol. Assoc. Afr. 21: 4-8.

McLachlan, G. R. 1979c. A second specimen of Phyllodactylus microlepidotus. J. Herpetol. Assoc. Afr. 21: 9.

McLachlan, G. R. and Spence, J. M. 1966. The genus Pachydactylus (Part 1). Ann. Cape Prov. Mus. 5: 149-156.

Mertens, R. 1955. Die Amphibien und Reptilien Siidwest-afrikas. Abh. senckenb. naturforsch. Ges. 490: 1-172.

Mertens, R. 1971. Die Herpetofauna Siidwest-afrikas. Abh. senckenb. naturforsch. Ges. 529: 1-110.

Onderstall, D. 1975. Some taxonomic and zoogeographic aspects of the gecko genus Afroedura. Lecture presented at
the Symposium on Herpetology and Ichthyology, Kruger National Park, September 1975.

Oelofsen, B. and Vorster, W. 1976. A new record for Ptenopus (Reptilia: Gekkonidae) from Calitzdorp. Zool. Afr.
11 (1): 225.

Parker, H. W. 1936. Dr Karl Jordan's Expedition to South West Africa and Angola: Herpetological collections. Novit.
zool. 40: 115-146.

Pasteur, G. 1964. Recherches sur 1’evolution des lygodactyles, lezards Afro-Malgaches actuels. Trav. Inst, scient. cherif.
Ser. zool. 29: 1-132.

Pianka, E. R. 1971. Lizard species density in the Kalahari Desert. Ecology 52 (6): 1024-1029.

Raw, L. R. G. 1976. A survey of the dwarf chameleons of Natal, South Africa, with descriptions of three new species
(Sauria: Chameleonidae). Durban Mus. Novit. 11 (7): 139-161.

Raw, L. R. G. 1978. A further new dwarf chameleon from Natal, South Africa (Sauria: Chameleonidae). Durban Mus.
Novit. 11 (15): 265-269.

Robertson, M. D. 1978. Sexual dichromatism in the Namaqua chamaeleon, Chamaeleo namaquensis. Madoqua 11 (1):
81-83.

Russell, A. P. 1976. Some comments concerning inter-relationships amongst gekkonine geckos. In Bellairs and Cox eds,
Morphology and biology of reptiles. London: Academic Press, pp. 217-244.

Russell, A. P. 1977. The genera Rhotropus and Phelsuma (Reptilia: Gekkonidae) in Southern Africa: A case of
convergence and a reconsideration of the biogeography of Phelsuma. Zool. Afr. 12 (2): 393-408.

166

CHECKLIST OF CAPE LIZARDS

Saiff, E. I. 1970. Georgraphical variation in the genus Zygaspis (Amphisbaenia: Reptilia). Herpetologica 26: 86-119.

Scherbach, N. N. 1975. Catalogue of African Desert Lizards. Acad. Sci. Kiev, USSR.

Schmidt, K. P. 1934. The reptiles of the Pulitzer Angola Expedition. Ann. Carneg. Mus. 22: 1-15.

Smith, H. M., Martin, R. L. and Swain, T. A. 1977. A new genus and two new species of South American geckos
(Reptilia: Lacertilia). Papeis Dep. Zool. S. Paulo 30 (14): 195-213.

Steyn, W. and Haacke, W. D. 1966. A new web-footed Gekko (Kaokogecko vanzyli gen. et. sp. nov.) from the
North-western South West Africa. Cimbebasia 18: 1-23.

Underwood, G. 1954. On the classification and evolution of geckos. Proc. zool. Soc. Lond. 124: 469-472.

Visser, J. 1971. Cordylus cordylus and its races in the Cape Province — A general discussion. J. Herpetol. Assoc. Afr. 7:
18-20.

Visser, J. 1979. New and re-confirmed records for the Cape Province with notes on some “rare" species. J. Herpetol.
Assoc. Afr. 21: 40-48.

Manuscript accepted for publication 12 September 1980.

167

\

!\ (o 13
N \A

ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

Ann. Cape Prov. Mus. (nat. Hist.)

VOLUME 13 • PART 12
30th APRIL 1981

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTGWN

SOUTH AFRICA

Printed by Cape & Transvaal Printers (Pty) Ltd, Cape Town

BD9451

Collections of Recent mammals in southern Africa

by

LLOYD R. WINGATE

Department of Zoology, University of Transkei, Private Bag X5092, Umtata

and

PIERRE SWANEPOEL

Kaffrarian Museum, King William’s Town 5600

INTRODUCTION

“Systematics collections are complex information acquisition, storage and retrieval sys-
tems comprising specimen samples of organisms and associated information about them”
(Irwin et al., 1973).

“Systematics collections of plants and animals are the only permanent record of the
earth’s biota, and the specialized libraries attached to these collections are the written record
of the earth’s natural history” (National Research Council Committee on research in the life
sciences, 1970).

Although the need for general collections from all parts of the world has diminished
through the advanced development of traditional mammalogy, the need for improving regional
representation in museums continues. The main reason for this is the value of collections as
the origin of ideas. Although it is often possible to assemble enough specimens from various
museums to address a given problem, it is also true that the problem might not have been
obvious in the first place without sufficient material of a taxon in any one museum to indicate
that such a problem exists (Zusi, 1969).

Few southern African institutions have Recent mammal collections of sufficient size to
carry out an adequate systematic study of any one taxon. This is probably true for most
museums the world over in all biological disciplines. Therefore, a very important function of a
collection is to make study material available to any bona fide research worker.

Collections are scattered worldwide and with the proliferation and growth of collections,
locating material which one wishes to study has become a problem. The various mammal
collections spread across southern Africa are obviously of great importance to workers on
African, and especially on southern African mammals. If such workers do not know of the
existence or whereabouts of all the available material of a taxon under review they may be
unable to produce conclusive results. This would also be an indirect waste of time and money
by both the systematist involved and the institutes which collected the unused material. An
example in hand is the revision of the endemic African family Macroscelididae “based on the
entire collection of the British Museum, amounting to about a thousand specimens, along with

169

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 12, APRIL 1981

smaller numbers received on loan or examined in other institutions” (Corbet and Hanks, |
1968). By these authors’ own admission most species could not be treated adequately at the
subspecific level because “distribution and variation” were not better known. Fewer than 80
specimens were examined from collections in Africa, only 11 of these specimens being housed
in southern African collections. At the time more than a thousand additional specimens of the
family Macroscelididae were available for study in southern African mammal collections.

The authors believe that the present paper will help workers on African mammals by
focusing their attention on material which is available in collections from this area.

No previous survey of southern African collections of Recent mammals and their uses has
been conducted. Such surveys have regularly been undertaken in North America (Howell,
1923; Doutt et al., 1945; Anderson et al., 1963; Choate and Genoways, 1975; Genoways et al.,
1976). At present a survey concerning mammal collections of the world is being conducted by
Hugh H. Genoways and Duane A. Schlitter. Their survey, however, is not planned to be as
extensive as those recently carried out for North America. (Choate and Genoways, 1975;
Genoways et al., 1976), or the present study for southern Africa.

The general method of conducting the present survey closely follows that of Choate and
Genoways (1975). The authors’ objectives were to assess: (1) the size, scope, and location of
the systematic resources of mammalogy in southern Africa; (2) the curatorial status of these
collections; (3) the actual use of the collection by the systematics community and others; and
(4) how to obtain the maximum use out of these collections. They hope the information
obtained and presented here will permit the assessment of problems within the discipline, and
generate some thought and discussion on how to correct these.

For the purpose of this survey the authors considered southern Africa as the area south of
the Cunene — Zambezi rivers, including the following countries: Bophuthatswana, Botswana
(formerly Bechuanaland), Lesotho (formerly Basutoland), Mozambique, South Africa, South
West Africa/Namibia, Swaziland, Transkei, Venda and Zimbabwe (formerly Rhodesia, South-
ern Rhodesia).

A questionnaire (Appendix 1) based on that of Choate and Genoways (1975) was drawn
up requesting details about the status of collections of Recent mammals as on 1 June 1978.
Questionnaires were mailed to: (a) museums; (b) other institutions and private individuals
known to have or suspected of having collections of Recent mammals; (c) conservation and
wildlife departments; (d) national parks; (e) forestry departments; (f) agricultural depart-
ments; and (g) universities. A mailing list was included with each questionnaire and all reci-
pients were requested to inform the authors if they were aware of any additional collections in
southern Africa. Notices to this effect were also placed in the Bulletin of the Southern African
Museums Association (SAMAB). Further questionnaires were sent in response to answered
notices and to notifications of institutions or persons suspected of holding collections.

Subsequently, letters were sent to the curators of the 15 collections reported to have 1 000 or
more specimens, requesting the following information: When was the collection started? What
was the size of the collection on 1 June 1918, 1 June 1938, 1 June 1958, 1 June 1968, 1 June
1973 and 1 June 1976?

In response to the 121 questionnaires mailed, 91 (75%) replies were received, 41 of which
reported collections of over 20 specimens (Table 1). Collections of fewer than 20 specimens
were ignored for the purposes of this survey. Based on these replies, the 41 collections of
Recent mammals known to exist in southern Africa, held a total of 174 073 specimens on 1
June 1978.

No collections of Recent mammals were reported from the following countries:
Bophuthatswana, Botswana (formerly Bechuanaland), Lesotho (formerly Basutoland), Swazi-
land, Transkei, or Venda. Table 1 summarizes the numbers of collections and specimens
housed in the remaining four countries, viz. , Mozambique, South Africa, South West Africa/

170

COLLECTIONS OF RECENT MAMMALS IN SOUTHERN AFRICA

Table 1. — Numbers of Recent mammal specimens and of holotypes and other type specimens
in southern African collections (Numbers in brackets refer to collection numbers used in text).

Institution

Total number
of specimens

Number of
holotypes and
other types

National Museum of Rhodesia (37)

67 000

2

Transvaal Museum (30)

28 000

439

Kaffrarian Museum (10)

25 400

30

South African Museum (3)

9 200

13

Queen Victoria Museum (40)

8 030

0

State Museum (36)

7 800

0

S.W.A. Division of Nature Conservation and Tourism —
Etosha Ecological Institute (34)

3 689

0

Albany Museum (7)

2 952

4

Natal Museum (22)

2 800

0

University of Rhodesia — Museum of Zoology (41)

2 136

0

Blair Research Laboratory (39)

2 000

0

Cape Provincial Department of Nature & Environmental
Conservation — Vrolijkheid (13)

1 800

0

National Museum, Bloemfontein (25)

1 300

0

University of Natal, Pietermaritzburg (24)

1 216

0

Port Elizabeth Museum (12)

1 150

0

Cape Provincial Department of Nature and Environmen-
tal Conservation — Jonkershoek (14)

880

0

Universidade Eduardo Mondlane (2)

836

0

University of the Witwatersrand — Bernard Price Institute
(28)

700

0

University of Stellenbosch — John R. Ellerman Museum
(16)

668

0

Durban Museum (19)

667

0

University of Cape Town (4)

650

0

Orange Free State Division of Nature Conservation (26)

600

0

N. G. Palmer collection (5)

600

0

S.W.A. Division of Nature Conservation and Tourism —
Windhoek (35)

593

0

University of Pretoria (31 )

478

0

South African Institute for Medical Research, Depart-
ment of Epidemiology (27)

350

0

South African National Parks Board — Kruger National
Park (32)

350

0

University of Stellenbosch — Department of Forestry (15)

300

0

L. R. Wingate collection (11)

284

0

McGregor Museum (9)

250

0

Natal Parks, Game and Fish Preservation Board —
Hluhluwe (21)

241

0

Museu de Historia Natural (1)

208

0

171

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 12, APRIL 1981

Institution

Total number
of specimens

Number of
holotypes and
other types

Natal Parks, Game and Fish Preservation Board — Pieter-
maritzburg (23)

200

0

P. B. Fourie collection (29)

180

0

University of Natal, Durban (20)

173

0

Cape Provincial Department of Nature & Environmental
Conservation — Rolfontein (18)

132

0

South African Department of Forestry — Saasveld
Museum (6)

85

0

W. R. J. Dean collection (33)

60

0

Rhodes University (8)

45

0

Sable Park Museum (38)

40

0

South African National Parks Board — Tsitsikama
National Park (17)

30

0

Namibia and Zimbabwe (formerly Rhodesia, Southern Rhodesia). Collections in Mopambique
house 0,6 per cent (1 044 specimens) of the known material present in southern Africa, South
Africa 47,0 per cent (81 741 specimens), South West Africa/Namibia 6,9 per cent (12 082
specimens), and Zimbabwe 45,5 per cent (79 206 specimens). The size increase since 1918 of
10 of the biggest collections is shown in Table 2.

Ninety-five per cent (164 473 specimens) of this material is housed in 15 collections of
1 000 specimens or more, 69 per cent in three collections of over 20 000 specimens, and 39 per
cent in the largest collection (National Museum of Rhodesia). Of the 15 collections holding
over 1 000 specimens, four are situated in Zimbabwe, two in South West Africa/Namibia, and
the rest in South Africa. Three of these 15 collections (Etosha Ecological Institute, University
of Rhodesia, University of Natal at Pietermaritzburg) are essentially teaching collections, and
one (Blair Research Laboratory) houses mainly cranial material.

Five collections in southern Africa house all the holotypes and other types (488) known to
be present in southern Africa. The majority (90%) of these are held by the Transvaal
Museum. Only one publication listing holotypes and other types of mammals has been pub-
lished in southern Africa (South African Museums’ Association, 1958). Holotypes and other
types of Recent mammals present in only two southern African museums (Kaffrarian Museum,
Transvaal Museum) were listed in this publication. For the Kaffrarian Museum, at least, many
mistakes are present, and at the moment a revised and annotated list of holotypes and other
types is being drawn up for the Kaffrarian Museum collection.

The Umtali Museum in Zimbabwe, which houses a large herpetological collection, keeps
only a small reference collection of mammals, not reported herein, and passes on all other mam-
mal material to the National Museum in Bulawayo. Generally only a relatively small collection
of mammals is held at Queen Victoria Museum, apart from material the curator may be study-
ing. The bulk of the Zimbabwean mammal material is deposited in the National Museum.

The East London Museum mammal collection (approximately 1 000 specimens) is now
catalogued and housed on permanent loan in the Kaffrarian Museum.

172

Table 2. — Growth of the major collections of Recent mammals in southern Africa over the last 60 years.

COLLECTIONS OF RECENT MAMMALS IN SOUTHERN AFRICA

1978

9 200
1 150
25 400

1 800
2 800
1 300
28 000

7 800
67 000
2 000

O 04 O

0 3-00

»n o o

NO

r-

O O O

co o m

ON o O

i/l O' 3

r- o <n

of O 04

on

oo m

04 r-H NO

\o r-

r-H

C4

r4

NO

O ON o

0 3-0

ON o o

co

r-

O -H o

IT) -t

-f o

o oo oc

NO 04 j

O

on

r- on

04 1

3- tn

T— 1

in

o co no

O 04 O

O m O O

-H o o

00

no

ON O

r-H O

O O "t

CO ON

in o

On

NO ON

04 on

m

t—H

7—H

rf

o m oo

O <N O O

o o o

00

in

O ON 00

ON O

o

O 3 't

04 NO

o

Os

in oo

n4

in

r- H

r-H

^H

o r-

0 0-00

o o o

00

CO

o r-

VO NO

o

° 04 1

t-h no

in

On

*

t-H 00

0(N0

O OO O O
ON NO

o o o

O -h

m

O 04

00 CO

04

On

co

04

s I
3 §

3 3
22
3 -5

.y jo
£ «

3 t:
o o
c/n CL

o c
o

EU

<D —

E 2

tn

Q. E -

« £ y

|Q 2.t

<d -3 > C3

«.2 c o

3 o rjJ

s.s^>

C o 3

to 2 03 __
r cl o

3

CO

I

3 £

i U 3 -

•K 0.2 >

c3 cd cd w

Z &

c

'5

c

£

6

0)

o

5

S

c 0>

5 C/3

3 G

CL) ^

3 „
5 3
^ 3

3-2

rt CO

z z

£

<

e 3

E o

3 co
<u ..
to —

3 E

S 3

OJ

, C/)

3 3
G ^

2 S

H co

cd

<u

T3

O £
J3 2
OC 3

^ ti

<*- O

O „Q

e 3

3 J
3

<U -3

3

s

73

c

o

3 22

uZCQ

CO On O co

Cj vO O v«o
04 04 CO CO

r- on
co ro

173

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 12, APRIL 1981

Part of the material on which the “Check list and atlas of the mammals of Mozambique”
(Smithers and Lobao Tello, 1976) was based, is thought to be in Mozambique and housed at
either Gorongoza National Park or Maputo (formerly Lourengo Marques). However, the
authors were unable to determine the exact whereabouts or size of this collection. Another
institution suspected of housing a collection but from which no reply was received is the De-
partment of Zoology, University of the Witwatersrand.

The bulk of the old Medical Ecology Centre collection was deposited in the Transvaal
Museum collection in 1971. Only a small reference collection was kept and is now under the
auspices of the Department of Epidemiology, South African Institute for Medical Research,
Johannesburg.

According to the persons in charge, the following collections will be deposited in reput-
able museum collections upon completion of the research being done on them: the Cape Pro-
vincial Department of Nature and Environmental Conservation collection at Vrolijkheid
Nature Conservation Station, and the South West Africa Division of Nature Conservation
and Tourism collection in Windhoek.

Between the effective date of this survey (1 June 1978) and the date of publication, four of
the smaller collections have been incorporated into larger collections. More than half of the
Natal Parks, Game and Fish Preservation Board collection at Pietermaritzburg ( c . 120
specimens) was deposited in the Transvaal Museum; the L. R. Wingate collection at King
William’s Town (284 specimens), the Cape Provincial Department of Nature and Environmen-
tal Conservation collection at Jonkershoek, Stellenbosch (880 specimens), and the cranial
material from the P. B. Fourie collection at Pretoria (180 specimens) were deposited in the
Kaffrarian Museum (total 1 344 specimens).

Electronic data processing is receiving immediate attention in only two collections (John
R. Ellerman Museum of Zoology, McGregor Museum).

Two museums, the National Museum of Rhodesia and the Natal Museum have the same
acronym (NM). This is an unacceptable situation. The function of an “unique” acronym for a
collection is to aid in the identification of specimens belonging to that collection in future
studies, particularly with material on loan. It will also be particularly useful in case of a com-
puterized system of information retrieval. The authors suggest that these museums adopt com-
pletely new acronyms for their collections, possibly: NMB — National Museum of Rhodesia,
and NATM — Natal Museum.

Most of the major collections have their curatorial practices and documentation pro-
cedures standardized and many also report having these available in written form. However,
few have all procedures available in one document to assist potential collectors, visiting scien-
tists and also the staff in managing the collection in a standardized way. The Kaffrarian
Museum has compiled such a document for its collection.

The past and present use of mammal collections and recommendations concerning their
maximum utilization will be discussed elsewhere.

SURVEY OF COLLECTIONS

In the accounts of collections the information is arranged as follows: name (including
acronym) and address of collection. Number of specimens; number of holotypes and other
types; primary goal of collection (present status of collection and proposed new plans); sys-
tematic groups best represented; geographic areas best represented; special preparations;
availability of field notes; special files; what percentage of cranial and other skeletal material
is cleaned and adequately stored; how specimens are catalogued; significant other collections in-
corporated into this collection; plans for active development of electronic data processing;
professional scientists at the institution (and their main research interests); person directly
responsible for the collection; person who completed the questionnaire.

174

COLLECTIONS OF RECENT MAMMALS IN SOUTHERN AFRICA

Negative answers are omitted except where these are of particular importance.

The following accounts of reported collections are arranged alphabetically, first by coun-
try, then by province (where appropriate), and finally by town or city where the institution or
owner is placed.

MOZAMBIQUE

1. MUSEU DE HISTORIA NATURAL (MHN), Praga da Travessia do Zambeze, Maputo
(formerly Lourenyo Marques), Mozambique. 208; primary goal is research (the study of mam-
mals only started recently with a Section of Mammals being established within the newly cre-
ated Department of Natural Sciences); Mozambique; special preparations include 20 elephant
foetuses in alcohol, one hippopotamus skeleton, 1 dugong skeleton, 186 mounted specimens;
no field notes; special files include 100 colour transparencies of specimens and 30 of habitats,
one tape recording; specimens are catalogued by geographical areas, and by collectors; re-
ported by the person responsible — D. J. A. Travassos Santos Dias, Professor of Veterinary
Medicine.

2. UNIVERSIDADE EDUARDO MONDLANE (UEM), Faculdade de Biologia, Maputo
(formerly Lourenzo Marques), Mozambique. 836; primary goal is teaching (segregated
teaching collection of 23 specimens), also used for research; Rodentia; Gorongoza, Maputo
Reserve, Mozambique in general; special preparations include 101 specimens in formalin, 187
skeletons, 493 mounted specimens; field notes; 80 per cent of skulls and other skeletal material
are cleaned and adequately stored; specimens are catalogued by systematic groups, by geo-
graphic areas, and by collectors; person responsible — Hans Feijen; reported by Augusto
Cabral.

SOUTH AFRICA
CAPE PROVINCE

3. SOUTH AFRICAN MUSEUM (ZM), P.O. Box 61, Cape Town 8000, Republic of South
Africa. 9 200; 13 holotypes and other types; primary goal is research (85%), also used for
exhibition, and teaching; Carnivora, Pinnipedia, Bovidae; Cape Province; special preparations
include specimens in alcohol, specimens in formalin, skeletons, mounted specimens, specimens
of known age, specimens of domestic animals; no field notes; 95 per cent of skulls and other
skeletal material are cleaned and adequately stored; specimens are catalogued numerically;
scientists at the institution using the collection are Q. B. Hendey (palaeontology), P. Hirschon
(archaeology); person responsible — P. Hirschon, Research Assistant; reported by Q. B.
Hendey, Curator of Palaeontology, and P. Hirschon.

4. UNIVERSITY OF CAPE TOWN, DEPARTMENT OF ZOOLOGY (UCTZ), Rondebosch,
Cape Town 7700, Republic of South Africa. 650; primary goal is teaching (80%), also used for
research; Rodentia, Carnivora; Cape Flats, De Hoop Nature Reserve, Bontebok National
Park, Rooiberg Mountains (Cape Province), Namib Desert (South West Africa/Namibia); spe-
cial preparations include 20 skeletons, microscope preparations, 60 Rhabdomys skulls of
known age, more than 400 stomach contents of Rhabdomys, skulls and skeletons of domestic
animals; field notes; special files include a few colour transparencies of specimens and habitats;
95 per cent of skulls and other skeletal material are cleaned and adequately stored; specimens
are not catalogued but recorded by geographic areas; scientists at the institution using the
collection are Jennifer U. M. Jarvis (small mammal ecology and physiology), J. H. M. David
(rodent ecology); reported by the person responsible — Jennifer U. M. Jarvis, Senior Lecturer.

5. N. G. PALMER COLLECTION (NGP), Private Bag X546, George 6530, Republic of
South Africa. 600; primary goal is research and reference (at the moment the collection is not

175

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 12, APRIL 1981

used much since the owner is now involved in estuarine research); Rodentia, Carnivora, Insec-
tivora; Cape Peninsula and vicinity, southern Cape coast, Oviston Nature Reserve, Rolfontein
Nature Reserve; special preparations include 450 skulls only; field notes; 90 per cent of skulls
and other skeletal material are cleaned and adequately stored; specimens are catalogued
numerically; reported by N. G. Palmer, owner.

6. SOUTH AFRICAN DEPARTMENT OF FORESTRY— SAASVELD MUSEUM (DFSM),
Saasveld Forest Research Station, Department of Forestry, Private Bag X531, George 6530,
Republic of South Africa. 85; primary goal is reference collection for ecological studies, used
for professional research (90%), and teaching; Rodentia; southern Cape Province; special pre-
parations include 15 specimens in alcohol; field notes; special files include three colour trans-
parencies of specimens, and 50 of habitats; specimens are kept on open shelves; 90 per cent of
the skulls are cleaned and adequately stored; specimens are catalogued numerically; scientist
at the institution using the collection is G. J. Breytenbach (ecology); reported by the person
responsible — G. J. Breytenbach, Professional Officer.

7. ALBANY MUSEUM (AMM), Somerset Street, Grahamstown 6140, Republic of South
Africa. 2 952; two holotypes and two other types; primary goal is research (but at present used
for research, 20%, exhibition, 10% and teaching, 70% — there is a small teaching collection —
the vast majority of the mounted specimens and virtually all the records were destroyed in a
fire in 1943, no mammal work has been conducted since c. 1940, specimens are at present
inadequately stored in boxes, there are plans to reorganize and improve the curatorial condi-
tion of the collection in 1980, after which an assessment can be made of the status of the
collection and the few remaining records); Chrysochloridae, Viverridae, Muridae; eastern
Cape Province; special preparations include 150 specimens in alcohol, 10 skeletons, 50
mounted specimens, many skulls only; few field notes remain after the fire; unknown percent-
age of skulls and other skeletal material is cleaned and adequately stored; specimens are not
catalogued; reported by the person responsible — W. Holleman, Deputy Director.

8. RHODES UNIVERSITY, DEPARTMENT OF ZOOLOGY (RUZ), Grahamstown 6140,
Republic of South Africa. 45; primary goal is undergraduate teaching (segregated teaching
collection), and research; Rodentia; Albany district. Cape Province; special preparations in-
clude 20 specimens in formalin, five skeletons; no field notes; most skulls are cleaned, none
adequately stored; specimens are catalogued numerically; scientists at the institution using the
collection are M. R. Perrin (rodent population biology), T. S. Allen-Rowlandson (kudu bio-
logy); reported by the person responsible — M. R. Perrin, Lecturer.

9. McGREGOR MUSEUM (MMK), P.O. Box 316, Kimberley 8300, Republic of South Afri-
ca. 250; primary goal is research (to date the collection has mainly been used for teaching —
segregated teaching collection of 10 specimens — and exhibition, a research collection of mam-
mals representing the northern Cape Province is planned for the immediate future); Carnivora,
Rodentia, Artiodactyla; northern Cape Province; special preparations include specimens in
alcohol (46 embryos, mostly springbok), some skeletons, 98 mounted specimens; no field
notes; six per cent of skulls and other skeletal material are cleaned and adequately stored;
specimens are catalogued numerically; immediate plans for active development of electronic
data processing; scientists at the institution using the collection are R. Liversidge (springbok
ecology), P. Richardson (scavengers); reported by the person responsible — P. Richardson,
Curator of Mammals.

10. KAFFRARIAN MUSEUM (KM), 3 Albert Road, King William’s Town 5600, Republic of
South Africa. 25 400; 30 holotypes and other types; primary goal is research (90%), also used
for teaching (segregated teaching collection of 50 specimens), public education, and exhibition;
Rodentia, Insectivora, Chiroptera, Lagomorpha, Carnivora; South West Africa/Namibia,

176

COLLECTIONS OF RECENT MAMMALS IN SOUTHERN AFRICA

Zambia (formerly Northern Rhodesia), north-western and eastern Cape Province, Malawi
(formerly Nyasaland); special preparations include 268 specimens in alcohol, 40 in formalin,
9 skeletons, 30 karyotypic preparations, approximately 300 mounted specimens, a collection of
approximately 2 500 skulls, reproductive organs, stomach contents, kidneys, adrenals, eye
balls mainly of Tatera leucogaster and Aethomys chrysophilus preserved in formalin from one
locality in the northern Transvaal, South Africa; field notes (in part after Hall, 1962, since
January 1978); 70 per cent of skulls and other skeletal material cleaned and adequately stored;
specimens are catalogued numerically, and on cards by systematic groups; in 1974 the Kaffra-
rian Museum bird collection was exchanged for the East London Museum mammal collection
(c. 1 000 specimens) which was then incorporated into this collection; scientists at the institu-
tion using the collection are Pierre Swanepoel (small mammal systematics), Lloyd R. Wingate
(Chiropteran ecology); person responsible — Pierre Swanepoel, Curator of Mammals; reported
by Lloyd R. Wingate, Collection Manager, and Pierre Swanepoel.

11. L. R. WINGATE COLLECTION (LRW), 16 Peters Road, King William’s Town 5600,
Republic of South Africa. 284; primary goal is research (these specimens were preserved as
voucher specimens for ecological work); Chiroptera; Natal; special preparations include 70
skulls only, 150 specimens in alcohol, 15 specimens of known age; held notes; special hies
include 100 black and white photographs of specimens, eight of habitats, 80 colour transparen-
cies of specimens, 20 of habitats, 15 minutes of tape recordings; two per cent of cranial and
other skeletal material is cleaned and adequately stored; specimens are catalogued numeri-
cally; reported by Lloyd R. Wingate, owner.

12. PORT ELIZABETH MUSEUM (PEM), P.O. Box 13147, Humewood, Port Elizabeth
6013, Republic of South Africa. 1 150; primary goal is research (95%), also used for teaching
and exhibition; Cetacea, Pinnipedia, Rodentia; south-east coast of southern Africa, south-
western Indian Ocean, eastern Cape Province; special preparations include specimens in alco-
hol (e.g. stomach contents), specimens in formalin (e.g. testes), skeletons, microscope prep-
arations; held notes for Cetacea and Pinnipedia; special hies include hfteen black and white
photographs of specimens, 200 colour transparencies of specimens; all skulls and other skeletal
material of marine mammals are cleaned and adequately stored, one per cent of these are
degreased, a small proportion of terrestrial mammal skulls and other skeletal material is
cleaned and adequately stored; marine mammals are catalogued numerically, cataloguing of
terrestrial mammals is pending; scientist at the institution using the collection is G. J. B. Ross
(small cetacean biology and systematics); reported by the person responsible — G. J. B. Ross,
Curator of Marine Biology.

13. CAPE PROVINCIAL DEPARTMENT OF NATURE AND ENVIRONMENTAL CON-
SERVATION—VROLIJKHEID NATURE CONSERVATION STATION (CPA), Private Bag
614, Robertson 6705, Republic of South Africa. 1 800; primary goal is research (95%), also
used for teaching (when work on the collection has been completed, material collected in the
different regions of the Cape Province will be deposited in the appropriate regional Cape
Provincial Museums, in the meantime the material cannot be obtained on loan); Carnivora;
western Cape Province; special preparations include specimens in alcohol, specimens in forma-
lin, specimens of known age (several Carnivora species); held notes; special hies include black
and white photographs of specimens and habitats, colour transparencies of specimens and
habitats; all skulls and other skeletal material are cleaned and adequately stored; specimens
are catalogued numerically, and on cards by systematic groups; scientist at the institution using
the collection is C. T. Stuart (Carnivora); reported by the person responsible — C. T. Stuart,
Research Assistant.

177

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 12, APRIL 1981

14. CAPE PROVINCIAL DEPARTMENT OF NATURE AND ENVIRONMENTAL CON-
SERVATION—JONKERSHOEK NATURE CONSERVATION STATION (JNCS), Private Bag
5014, Stellenbosch 7600, Republic of South Africa. 880; primary goal is research; Chiroptera;
south-western Cape Province; special preparations include all specimens in alcohol; field notes;
special files include black and white photographs of specimens, colour transparencies of speci-
mens; some skulls have been removed from fluid-preserved specimens, cleaned and adequately
stored; specimens are catalogued numerically; scientist at the institution using the collection is
J. C. Herselman (chiropteran ecology and systematics); reported by the person responsible — J.
C. Herselman, Research Officer.

15. UNIVERSITY OF STELLENBOSCH. FACULTY OF FORESTRY (USF), Stellenbosch

7600, Republic of South Africa. 300; primary goal is teaching; Rodentia, Insectivora; south-
western Cape Province; special preparations include 40 specimens in formalin, 10 injected
anatomical preparations, some specimens of known age, most specimens skulls only; no field
notes; special files include 15 colour transparencies of specimens, 40 of habitats, 10 tape re-
cordings; 40 per cent of skulls and other skeletal material are cleaned and adequately stored;
specimens are catalogued numerically; reported by the person responsible — D. Pepler, Tech-
nical Assistant.

16. UNIVERSITY OF STELLENBOSCH— JOHN R. ELLERMAN MUSEUM OF ZO-
OLOGY (JE), Stellenbosch 7600, Republic of South Africa. 668; primary goal is research
(75%), also used for teaching — segregated teaching collection of 50 specimens (the collection
was recently established — 1977 — and is intended to encourage student research); Rodentia,
Insectivora, Carnivora; south-western Cape Province; special preparations include 50 speci-
mens in alcohol, 20 skeletons, 20 injected anatomical preparations, 20 microscope pre-
parations, reproductive organs, embryos, three mounted specimens, 25 specimens of domestic
animals; no field notes; special files include 10 black and white photographs of specimens, five
of habitats, three colour transparencies of specimens, five of habitats, 15 tape recordings; all
skulls and other skeletal material are cleaned and adequately stored; specimens are catalogued
numerically, and on cards by systematic groups; electronic data processing utilizing the
SELGEM programme package has been initiated; scientists at the institution using the collec-
tion are Alan Channing (systematics), staff of the Department of Archaeology (palaeo-ecolo-
gy); reported by the person responsible — Alan Channing, Curator.

17. SOUTH AFRICAN NATIONAL PARKS BOARD— TSITSIKAMA FOREST AND COAST-
AL NATIONAL PARK (TNP), P.O. Storms River 6308, Republic of South Africa. 30; pri-
mary goal is research (90%), also used for exhibition; Tsitsikama (southern Cape Province);
special preparations include specimens in alcohol, in formalin, skeletons, mounted specimens;
no field notes; special files include a few black and white photographs of specimens, a few
colour transparencies of specimens and habitats, one tape recording of otter sounds, two mo-
tion pictures; specimens are catalogued numerically; reported by the person responsible —
G. A. Robinson, Park Warden.

18. CAPE PROVINCIAL DEPARTMENT OF NATURE AND ENVIRONMENTAL CON-
SERVATION—ROLFONTEIN NATURE RESERVE (RNR), P.O. Box 23, Vanderkloofdam

8771, Republic of South Africa. 132; primary goal is reference and research; Artiodactyla, Car-
nivora; Rolfontein Nature Reserve (north-eastern Cape Province); special preparations in-
clude 13 specimens in formalin, 19 stomach contents in formalin; field notes; skulls and other
skeletal material cleaned and adequately stored; specimens are catalogued numerically, and by
systematic groups; reported by the person responsible — Ken Coetzee, Senior Nature Con-
servation Officer.

178

COLLECTIONS OF RECENT MAMMALS IN SOUTHERN AFRICA

NATAL

19. DURBAN MUSEUM AND ART GALLERY (DM), P.O. Box 4085, Durban 4000, Repub-
lic of South Africa. 667; the primary goal is research, teaching, and exhibition (but at present
used for professional research, 5%, teaching, 60%, public service, 30%, and exhibition, 5%);
Chiroptera, Rodentia, Carnivora ( Genetta ); Natal, Cape Province; special preparations include
a few bats in alcohol, some mounted specimens; no field notes; skulls and other skeletal mate-
rial are cleaned and adequately stored; specimens are catalogued numerically; reported by the
person responsible — C. Quickelberge, Scientific Officer.

20. UNIVERSITY OF NATAL — DURBAN, DEPARTMENT OF BIOLOGICAL SCIENCES

(UND), King George V Avenue. Durban 4001, Republic of South Africa. 173; primary goal is
teaching, and exhibition; Natal; special preparations include 20 specimens in alcohol, 32 in
formalin, 120 skulls only; no field notes; five per cent of skulls and other skeletal material are
cleaned and adequately stored; specimens are catalogued by systematic groups; reported by
the person responsible — Margaret I. Keogh, Graduate Assistant.

21. NATAL PARKS, GAME AND FISH PRESERVATION BOARD— HLUHLUWE GAME
RESERVE (HGR), P.O. Box 25, Mtubatuba 3935, Republic of South Africa. 241; primary
goal is research (compilation of checklists); Rodentia; Zululand (Natal); special preparations
include 187 fluid-preserved specimens, 14 known-age specimens; no field notes; 50 per cent of
skulls and other skeletal material are cleaned and adequately stored; specimens are catalogued
numerically; scientists at the institution using the collection are P. M. Brooks (general mam-
malogy), J. L. Anderson (predators); person responsible — P. M. Brooks, Research Officer;
reported by P. J. Birkenstock, Research Assistant.

22 NATAL MUSEUM (NM), Loop Street, Pietermaritzburg 3201, Republic of South Africa.
2 800; primary goal is research; Rodentia, Insectivora; Natal; field notes; 80 per cent of skulls
and other skeletal material are cleaned and adequately stored; specimens catalogued numer-
ically, and on cards by systematic groups; scientist at the institution using the collection is J. A.
Pringle (mammals of Natal); reported by the person responsible — J. A. Pringle, Director
Emeritus.

23. NATAL PARKS, GAME AND FISH PRESERVATION BOARD — PIETERMARITZ-
BURG (DR), P.O. Box 662, Pietermaritzburg 3200, Republic of South Africa. 200; primary
goal is reference; Carnivora, Artiodactyla; Natal Midlands, Natal Drakensberg; special prepar-
ations include 35 Artiodactyla skulls only; field notes; scientist at the institution using the
collection is D. Rowe-Rowe (small carnivore ecology); reported by the person responsible —
D. Rowe-Rowe, Research Officer.

24. UNIVERSITY OF NATAL — PIETERMARITZBURG, DEPARTMENT OF ZOOLOGY

(UNP), P.O. Box 375, Pietermaritzburg 3200, Republic of South Africa. 1 216; primary goal is
teaching (there is a good deal of research material, including karyotypic preparations and
voucher specimens in the care of the department from time to time, but these are all eventually
deposited in the Transvaal Museum upon completion of research, this temporary material is
not included in this account); Artiodactyla, Carnivora, Primates, Rodentia; Natal, southern
Africa; special preparations include 10 specimens in alcohol, 30 in formalin, 108 skeletons, five
injected anatomical preparations, 30 mounted heads, eight complete skeletons of domestic
animals, 1 000 skulls only; no field notes; special files include colour transparencies of speci-
mens and habitats; most skulls and other skeletal material are cleaned but few are adequately
stored; specimens are catalogued numerically; reported by the person responsible — J. Meester,
Head of the Department of Zoology.

179

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 12, APRIL 1981

ORANGE FREE STATE

25. NATIONAL MUSEUM (NMB), P.O. Box 266, Bloemfontein 9300, Republic of South
Africa. 1 300; primary goal is research (93%), also used for teaching and exhibition; Rodentia,
Carnivora; Orange Free State; special preparations include 30 specimens in alcohol, five speci-
mens in formalin, 260 skeletons, 1 300 microscope preparations, 20 mounted specimens, five
specimens of known age, 15 specimens of domestic animals; field notes; special files include 40
black and white photographs of specimens, 40 colour transparencies of specimens; 90 per cent
of skulls and five per cent of skeletons are cleaned and adequately stored; specimens are
catalogued numerically, by systematic groups, and by geographic areas; scientist at the institu-
tion using the collection is C. D. Lynch (zoogeography, reproduction); reported by the person
responsible — C. D. Lynch, Curator of Mammals.

26. ORANGE FREE STATE PROVINCIAL DIVISION OF NATURE CONSERVATION

(ORANVRY), P.O. Box 517. Bloemfontein 9300, Republic of South Africa. 600; primary
goal is research, also used for exhibition; Lepus capensis, Papio ursinus ; Willem Pretorius
Game Reserve, and Winburg, Bloemfontein, Brandfort and Theunissen districts (Orange Free
State); all specimens are skulls only except for a few used for display; no field notes; skulls and
other skeletal material are cleaned but not adequately stored; specimens not catalogued; re-
ported by the person responsible — N. A. Ferreira, Senior Professional Officer.

TRANSVAAL

27. SOUTH AFRICAN INSTITUTE FOR MEDICAL RESEARCH, DEPARTMENT OF

EPIDEMIOLOGY (MEC), P.O. Box 1038, Johannesburg 2000, Republic of South Africa. 350;
primary goals are research, teaching (segregated teaching collection of 20 specimens), and
exhibition (in 1971 the bulk of the Medical Ecology Centre, MEC, collection of over 6 000
specimens was donated to the Transvaal Museum and only a reference collection and some
extralimital material was kept to assist in disease surveillance, teaching, and exhibition — the
Medical Ecology Centre grew from the Plague Research Station of the Department of Flealth
under D. H. S. Davis into the Plague Research Laboratory which was established in 1946 as a
joint enterprise with the South African Institute for Medical Research, SA1MR, and still
headed by D. H. S. Davis, this became the Medical Ecology Centre in 1957, incorporating all
aspects of research into zoonoses with their insect vectors and mammal hosts — from the earl-
iest years D. H. S. Davis built up a comprehensive and valuable collection of small mammals
until he retired in 1971 — in the late 1960s the State Health Laboratory Sciences took over the
Medical Ecology Centre which then became the State Medical Ecology Laboratory and has
been stationed in Rietfontein, Pretoria since 1977, although the reference collection has re-
mained at the SA1MR); Rodentia, Carnivora, Primates; southern Africa; special preparations
include 50 skins only, 75 specimens formerly from the D. H. S. Davis extralimital rodent
collection; field notes; all skulls and other skeletal material are cleaned and adequately stored;
specimens are catalogued numerically, by systematic groups, and by geographic areas; scien-
tists at the institution using the collection are Hillary J. Keogh (mammalogy) and some of the
entomologists at the SAIMR; reported by the person responsible — Hillary J. Keogh, Mamma-
logist and Research Officer.

\

J

28. UNIVERSITY OF THE WITWATERSRAND— BERNARD PRICE INSTITUTE FOR
PALAEONTOLOGICAL RESEARCH (BPI-Pal), 1 Jan Smuts Avenue, Johannesburg 2001,
Republic of South Africa. 700; primary goal is research (60%), also used in teaching and
public education; Bovidae, Suidae, Hyracoidea, Carnivora, Hystricidae; Karoo (Cape Pro-
vince), Transvaal, western Kalahari (Cape Province); special preparations include mostly
skeletal material, many skulls only, 16 mounted specimens, several known-age specimens,

COLLECTIONS OF RECENT MAMMALS IN SOUTHERN AFRICA

several domestic animal specimens, in addition there are 700 bone specimens from cleared
porcupine lairs which are catalogued separately from the skeletons; no field notes; special files
include black and white photographs of Hystricidae, other specimens, and habitats; all skulls
and skeletal material are cleaned and adequately stored; specimens are catalogued numerical-
ly, and by systematic groups; scientists at the institution using the collection are Judy M.
Maguire (fossil mammals), J. W. Kitching (fossil mammals, reptiles, amphibians), C. E. Gow
(early mammals, dinosaurs); reported by the person responsible — Judy M. Maguire, Research
Officer.

29. P. B. FOURIE COLLECTION (PBF), Tuberculosis Research Institute, Private Bag
X385, Pretoria 0001, Republic of South Africa. 180; primary goal is research; Procavia capen-
sis\ north-eastern Orange Free State; all specimens are skulls only; no field notes; special files
include tape recordings (2,5 hours edited), sonograms of 22 different Procavia capensis vocal-
izations; all tape recordings are properly stored in dust free containers; tapes catalogued by
date and geographic area, but additionally listing information on systematic group, sex of
animal, age, associated behavioural patterns and tape speed (this information exists on the
tape itself on the leader section), cranial material not catalogued; reported by P. B. Fourie,
owner.

30. TRANSVAAL MUSEUM (TM), P.O. Box 413, Pretoria 0001, Republic of South Africa.
28 000; 439 holotypes and other types; primary goal is research (95%), used also for teaching
and public education; Rodentia, Insectivora, Chiroptera; South Africa, especially Transvaal
Province; special preparations include 250 specimens in alcohol, 200 skeletons 300 karyotypic
preparations, 500 mounted specimens, 70 specimens of known age; field notes (since 1970,
previously only on occasion); special files include 400 colour transparencies of specimens, 100
of habitats; all skulls and other skeletal material are cleaned and adequately stored; specimens
are catalogued numerically, and on cards by systematic groups; the bulk of the South African
Department of Health’s Medical Ecology Centre collection (MEC) was incorporated into the
Transvaal Museum collection in 1971; scientists at the institution using the collection are I. L.
Rautenbach and N. J. Dippenaar (mammal systematics), C. K. Brain and E. Vrba (palaeont-
ology), E. Voigt, and 1. Plug (archaeology); reported by the person responsible — I. L.
Rautenbach, Curator of Mammals.

31 UNIVERSITY OF PRETORIA, DEPARTMENT OF ZOOLOGY (UPZ), Pretoria 0002,
Republic of South Africa. 478; primary goal is teaching (100%); Rodentia; Transvaal; special
preparations include 50 fluid preserved specimens, 130 skeletons, three injected anatomical
preparations, several karyotypic preparations, many microscope preparations, three mounted
specimens; no field notes; special files include colour transparencies of habitats; five per cent
of skulls and other skeletal material are cleaned and adequately stored; specimens are not
catalogued; reported by the person reponsible — J. A. J. Nel, Professor of Zoology.

32. SOUTH AFRICAN NATIONAL PARKS BOARD— KRUGER NATIONAL PARK (KNP),
Private Bag X404, Skukuza 1350, Republic of South Africa. 350; primary goal is research
(85%), also used for teaching, and exhibition; Chiroptera, Insectivora, Lagomorpha, Roden-
tia; Kruger National Park (Transvaal), special preparations include specimens in alcohol, in
formalin, skeletons, and mounted specimens; no field notes; 50 per cent of skulls and other
skeletal material are cleaned and adequately stored; specimens are catalogued by systematic
groups; scientists at the institution using the collection are S. C. J. Joubert, U. de V. Pienaar,
and G. L. Smuts (zoology), V. de Vos (veterinary ecology); reported by person responsible —
S. C. J. Joubert, Chief Research Officer.

33. W. R. J. DEAN COLLECTION (WRJD), P.O. Box 1397, Tzaneen 0850, Republic of
South Africa. 60; primary goal is as a reference collection to identify prey animals of predatory

181

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 12, APRIL 1981

birds; Macroscelididae, Soricidae, Chiroptera, Rodentia; northern and western Transvaal; spe-
cial preparations include 30 skulls only; two skeletons; field notes; all skulls and other skeletal
material are cleaned and adequately stored; specimens are catalogued by systematic groups;
reported by W. R. J. Dean, owner.

SOUTH WEST AFRICA/NAMIBIA

34. SOUTH WEST AFRICA DIVISION OF NATURE CONSERVATION AND TOURISM—
ETOSHA ECOLOGICAL INSTITUTE (EEI), Etosha National Park, Okaukuejo, South West
Africa/Namibia. 3 689; primary goals are research (60%), training of staff, and education of
visitors (the purpose of holding study material is to build up a reference collection for local
reference and training programmes, additional material is deposited with the State Museum in
Windhoek); Insectivora, Chiroptera, Carnivora, Bovidae, Rodentia; Etosha National Park on-
ly; special preparations include 25 specimens in alcohol, 25 in formalin, 25 mounted speci-
mens, 3 569 skulls only; no field notes; all skulls and skeletal material are cleaned and ade-
quately stored; specimens are catalogued numerically; scientists at the institution using the
collection are John E. W. Dixon (taxonomy, management), H. H. Berry (ecology, behaviour),
P. de Villiers (management); person responsible — Senior Conservator (Research); reported by
John E. W. Dixon, Principal Nature Conservator.

35. SOUTH WEST AFRICA DIVISION OF NATURE CONSERVATION AND TOURISM—

WINDHOEK (NCTW), Private Bag 13186, Windhoek, South West Africa/Namibia. 593; pri-
mary goal is research (this collection will be deposited in the State Museum, Windhoek on
completion of research, only then will it be available on loan); Chiroptera, Gerbillinae
(Rodentia); central Namib Desert; special preparations include 160 specimens in alcohol, 240
in formalin, 12 skeletons, one specimen of known age; field notes; special files include 120
colour transparencies of habitats; 12 per cent of skulls and other skeletal material are cleaned
and adequately stored; specimens not catalogued; scientist at the institution using the collec-
tion is Michael Griffin (small mammal taxonomy, distribution); reported by the person re-
sponsible— Michael Griffin.

36. STATE MUSEUM (SM), P.O. Box 1203, Windhoek, South West Africa/Namibia. 7 800;
primary goal is research (95%), also used for teaching (teaching collection, a few specimens),
and public service; Insectivora, Lagomorpha, Rodentia, Carnivora, Hyracoidea; South West
Africa/Namibia; special preparations include specimens in alcohol, 100 skeletons, 30 karyo-
typic preparations, a few mounted specimens, one specimen of known age ; field notes; more than 90
per cent of skulls and other skeletal material are cleaned and adequately stored; specimens are
catalogued numerically, and on cards by systematic groups; scientist at the institution using the
collection is C. G. Coetzee (taxonomy, zoogeography); reported by the person responsible — C. G.
Coetzee, Director.

ZIMBABWE ( FORMERLY RHODESIA , SOUTHERN RHODESIA )

37. NATIONAL MUSEUM OF RHODESIA (NM), P.O. Box 240, Bulawayo, Zimbabwe.
67 000; 2 paratypes; primary goal is research (60%), also used for teaching, public service, and
exhibition (the main body of mammal collections in the National Museum of Rhodesia is kept
here); Rodentia, Insectivora, Carnivora, Lagomorpha, Artiodactyla, Perissodactyla, Hyra-
coidea; Zimbabwe (formerly Rhodesia, Southern Rhodesia), Zambia (formerly Northern
Rhodesia), Botswana (formerly Bechuanaland); special preparations include specimens in
alcohol, in formalin, mounted specimens, specimens of known age, a small foetus collection;
field notes; 99 per cent of skulls and other skeletal material are cleaned and adequately stored;
specimens are catalogued numerically, by systematic groups, and by geographic areas; scien-

182

COLLECTIONS OF RECENT MAMMALS IN SOUTHERN AFRICA

tists at the institution using the collection are K. M. Adams (small mammals, primarily
Muroidea), R. J. Peek; reported by person responsible — Curator of Mammals (K. M. Adams
till August 1978 and R. J. Peek subsequently).

38. SABLE PARK MUSEUM (SPM), P.O. Box 315, Que Que, Zimbabwe. 40; primary goal
is to provide an interpretive service (the study centre has only recently been established, it is
hoped that research will form the major part of its use, followed by teaching and exhibition);
Rodentia; Sable Park; held notes; all skulls and other skeletal material are cleaned and ade-
quately stored; specimens are catalogued by systematic groups; reported by the person re-
sponsible— J. M. Stephens, Member of Management Committee.

39. BLAIR RESEARCH LABORATORY (BRL), Ministry of Health. P.O. Box 8105, Cause-
way, Salisbury, Zimbabwe. 2 000; primary goal is research (all skulls and eye-lenses are re-
tained but only definitive collections of study skins are kept); Rodentia; Zimbabwe (formerly
Rhodesia, Southern Rhodesia); special preparations include 300 karyotypic preparations, eye-
lenses; held notes; special hies include hve black and white photographs of habitats, colour
transparencies of specimens and habitats, tape recordings of ultrasonic sounds, karyotype
photomicrographs; all skulls and other skeletal material are cleaned and adequately stored;
specimens are catalogued numerically, and on punch cards which can discriminate species,
locality, sex, age, ectoparasites; scientists at the institution using the collection are D. H.
Gordon and P. Taylor (genetics, ecology, epidemiology, systematics); reported by the person
responsible — D. H. Gordon.

40. QUEEN VICTORIA MUSEUM (QVM), P.O. Box 8006, Causeway, Salisbury, Zim-
babwe. 8 030; primary goal is research (the main body of the mammal collections in Zimbabwe
is held at the National Museum, Bulawayo, while only a small collection is held at Queen
Victoria Museum in addition to temporary mammal material that is of particular interest to the
curator) segregated teaching collection of 30 specimens; Muridae, Carnivora; Zimbabwe
(formerly Rhodesia, Southern Rhodesia), Botswana (formerly Bechuanaland), Mozambique;
special preparations include 500 specimens in propanol, 100 in formalin, 150 skeletons, 50
injected anatomical preparations, 100 mounted specimens, 15 specimens of known age; field
notes; special files include 200 black and white photographs of habitats, 50 colour transparen-
cies of live specimens, 200 of habitats; all skulls and other skeletal material are cleaned and
adequately stored; specimens are catalogued numerically, by systematic groups, by geographic
areas, and by collectors; small historical collection from C. J. Anderson, 1859; scientist at the
institution using the collection is Reay H. N. Smithers (mammalogy); reported by the person
responsible — Reay H. N. Smithers, Associate Curator of Mammals (till November 1978).

41. UNIVERSITY OF RHODESIA— MUSEUM OF ZOOLOGY (URMZ), Box MP 167, Mt.
Pleasant, Salisbury, Zimbabwe. 2 136; primary goal is teaching (95%). also used for exhibition
and student research (only 66 per cent of the skin and skull material have their proper skulls,
much material has incomplete data, collection has little research value); Rodentia, Microchir-
optera; Salisbury district, Kariba Lake area; special preparations include 1 600 specimens in
formalin, 42 skeletons, one injected anatomical preparation, microscope preparations, four
mounted specimens, five specimens of known age, 36 domestic animals, 236 skulls only; no
field notes; special files include electron-microscope photographs of hairs; not all specimens
are adequately stored; 75 per cent of skulls and other skeletal material are cleared and ade-
quately stored; specimens are catalogued numerically, and by systematic groups; scientist at
the institution using the collection is T. Choate (mammalogy, wildlife management, behaviour,
resource ecology); reported by the person responsible — T. Choate, Senior Lecturer.

183

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 12, APRIL 1981

ACKNOWLEDGEMENTS

The authors wish to thank all the persons who supplied information concerning this survey.
They are grateful to C. G. Coetzee, N. J. Dippenaar, Hugh H. Genoways, I. L. Rautenbach,

J. Meester and Duane A. Schlitter for reviewing earlier drafts of the manuscript and to Wanda

Maritz for clerical assistance.

REFERENCES

Anderson, S., Doutt, J. K. and Findley, J. S. 1963. Collections of mammals in North America. J. Mammal. 44:
471-500.

Choate, J. R. and Genoways, H. H. 1975. Collections of Recent mammals in North America. J. Mammal. 56: 452-
502.

Corbet, G. B. and Hanks, J. 1968. A revision of the elephant-shrews, family Macroscelididae. Bull. Br. Mus. nat.
Hist., Zoology 16: 47-111.

Doutt, J. K., Howell, A. B. and Davis, W. B. 1945. The mammal collections of North America. J. Mammal. 26:
231-272.

Genoways, H. H., Choate, J. R., Pembleton, E. F., Greenbaum, I. F. and Bickham, J. W. 1976. Systematists, other
users, and uses of North American collections of Recent mammals. Museology 3: 1-87.

Howell, A. B. 1923. The mammal collections of North America. J. Mammal. 4: 113—120.

Irwin, H. S., Payne, W. W., Bates, D. M. and Humphrey, P. S. eds. 1973. America’s systematics collections: a
national plan. Association of Systematics Collections, Lawrence, Kansas, xiii + 63 pp.

National Research Council Committee on Research in the Life Sciences. 1970. The life sciences. National Academy
of Science, Washington, D C., 526 pp.

Smithers, R. H. N. and LoBao Tello, J. L. P. 1976. Check list and atlas of the mammals of Mojambique. Mus. Mem.
Nat. Mus. Rhodesia 8: viii + 1-184.

South African Museums' Association. 1958. A list of zoological and botanical types preserved in collections in southern
and East Africa. Vol. 1 — Zoology. Part I. South African Museums’ Association, Pretoria, South Africa, v +
147 pp.

Zusi, R. L. 1969. The role of museum collections in ornithological research Pp. 651-661, In Cohen, D. M. and
Cressey, P. F., eds. Natural history collections past-present-future. Proc. biol. Soc. Wash. 82: 559-762.

Manuscript accepted for publication 15 September 1980.

APPENDIX I

Questionnaire on collections of Recent mammals in southern Africa.

1. What is the formal name (if any), standard abbreviation (if any) and address, of your
private or institutional collection?

2. What is the name, address and telephone number of the person directly responsible for
the collection?

3. Approximately how many specimens of Recent (not fossil) mammals were in the collec-
tion as of 1 June 1978? Approximately how many are not yet catalogued? As of what
date?

4. How many holotypes are in the collection? How many other type specimens (syntypes,
paratypes, etc.)? As of what date? Has a catalogue or list of these types been published?
If so, please give reference.

5. What geographic areas are best represented in the collection?

6. What systematic groups are best represented in the collection?

7. What preparations are represented? Give, if possible, approximate numbers of specimens.
Skin and skull? Specimens in alcohol/propanol? Specimens in formalin? Skeletons? In-
jected anatomical preparations? Karyotype preparations? Microscope preparations?
Mounted specimens? Segregated teaching collections? Specimens of known age? Domes-
tic animals? Others (specify)?

184

COLLECTIONS OF RECENT MAMMALS IN SOUTHERN AFRICA

8. Are original field notes and field catalogues routinely conserved with the collection?

9. Are specimens in the collection stored in containers adequate to protect them from dam-
age?

10. Approximately what percentage of skulls and other skeletal material are cleaned and indi-
vidually enclosed in appropriately labelled containers?

11. What special files are maintained with the collection (if possible, give approximate num-
bers)? Black and white photographs of specimens? Black and white photographs of habi-
tats? Colour transparencies of specimens? Colour transparencies of habitats? Tape record-
ings? Motion pictures? Others (specify)?

12. Are specimens recorded in a catalogue arranged: numerically? by systematic groups? by
geographic areas? by collectors? others (specify)?

13. Does the collection include specimens referred to in the literature as being in a different
collection (for example, a formerly private collection now deposited in an institutional
collection? If so, what collections previously referred to have been combined with yours?

14. Are curatorial practices and documentation procedures standardized for the collection?
Are they available in written form?

15. Do you have immediate plans for active development of electronic data processing as a
means of assisting collection management and information retrieval?

16. Can specimens be borrowed for study purposes? This presumes that loans are usually
made to institutions rather than individuals and that adequate insurance will be carried
during shipment; the loan of holotypes is not expected.

17. Approximately how many such loans have been made in the past two years? How many
specimens were involved in these loans?

18. Is the collection stored in such a manner as to permit use by visiting scientists? Is someone
usually available to aid a scientist who wishes to use the collection? Who should be con-
tacted for such arrangements? Should the arrangement be made in advance?

19. Is a special systematic library maintained near the collection? Do you regard the library as
poor? fair? good? excellent? Is use of the library available to persons who might visit the
collection?

20. Approximate number of visitors who have utilized the collection in the past two years:
professional mammalogists? post-graduate students? other professional scientists? provin-
cial or government employees? others (specify)?

21. In the past two years, how many visiting students have used the collection directly in their
thesis projects? How many for other projects? How many students have used the collec-
tion indirectly (such as for deposition of voucher specimens) in their thesis projects? How
many for other projects?

22. How is the collection utilized, according to your estimation? Give answer in percentage of
total utilization. Professional research? student research? teaching? public service? exhibi-
tion? other uses (specify)?

23. What do you regard as the primary goal of the collection (e.g. research, teaching, exhibi-
tion or others)?

24. Who are the professional scientists at your institution who use the collection, and what
are their main research interests?

25. Does your institution produce a periodic or annual report in which information on the
collection is routinely reported? If so, give the most recent reference.

26. This questionnaire was completed by: Name, Title, Date.

27. Please add below or on separate sheets, any additional remarks or statistics that may be of
interest.

185

m

/

13

rlH

ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

Ann. Cape Prov. Mas. ( nat . Hist.)

VOLUME 13 • PART 13
30th APRIL 1981

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTQWN

SOUTH AFRICA

Printed by Cape & Transvaal Printers (Pty) Ltd, Cape Town

BD9460

A new species of the whale-louse Syncyamus (Crustacea: Amphipoda:
Cyamidae) ectoparasitic on dolphins from South Africa

by

ROGER J. LINCOLN

Department of Zoology, British Museum (Natural History), Cromwell Road,

London SW7 5BD

and

DESMOND E. HURLEY

New Zealand Oceanographic Institute, P.O. Box 12-346, Wellington, New Zealand

ABSTRACT

A new species of whale-louse, Syncyamus aequus sp. nov., ectoparasitic on the Common
dolphin, Delphinus delphis Linnaeus, the Blue-white dolphin, Stenella coeruleoalba (Meyen),
and on Tursiops aduncus (Ehrenberg), is described and figured from material collected on the
South African coast. The cyamids were taken from the blowhole, snout, mouth, and eye of the
hosts, and have the smallest adult body size of any species known to date. The status of
Syncyamus is reviewed. A table summarizes the sparse literature on cyamid/dolphin associ-
ations.

INTRODUCTION

In an earlier paper (Lincoln & Hurley 1974a), the authors described a new genus and
species of cyamid, Scutocyamus parvus , from the common North Atlantic White-beaked dol-
phin, Lagenorhynchus albirostris Gray. With a maximum recorded body size (ovigerous
female) of 3,1 mm this was the smallest whale-louse species known at that time. Just com-
pleted (Lincoln & Hurley, in press) is an account of a second species of Scutocyamus , this time
from New Zealand, which has a still smaller body size than S. parvus , but in this instance the
measurement comes from a non-ovigerous female which may understate the true maximum
body size of the species. In the authors’ experience the ovigerous specimens are usually slightly
larger than the non-ovigerous ones. Scutocyamus is especially interesting as one of the few
examples of a cyamid that parasitizes the smaller toothed cetaceans, the porpoises and the
dolphins (Leung 1967).

A recent collection from the Cape coast of South Africa has now furnished further evi-
dence of cyamids infesting dolphins. The cyamid taken from three different hosts, Delphinus
delphis, Stenella coeruleoalba, and Tursiops aduncus, also has a body size smaller than Scu-
tocyamus parvus. The largest of the ovigerous females in the collection is only 2,8 mm in
length — so small that had it not been for eggs in the brood pouches the authors might have
overlooked the specimens as juveniles of an indeterminable species. The cyamids can be refer-
red to the genus Syncyamus Bowman, 1955, but represent a hitherto undescribed species. The

187

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 13, APRIL 1981

name Syncyamus aequus sp. nov. is proposed for this species — the specific epithet, taken from
the Latin meaning impartial, alluding to the non host-specific habit of this species.

The material was provided by Dr Graham J. B. Ross of the Port Elizabeth Museum and
reached the authors in two parts. The first collection came via Professor J. L. Mohr, recently
retired from the Department of Zoology at the University of California, Los Angeles. It had
been sent to the University by Dr Ross, in the first instance to Dr Yuk-Maan Leung, the
foremost of recent workers on cyamids, whose premature death in 1976 was a sad loss from the
ranks of amphipodologists. The remainder of the material came to the authors directly from
Dr Ross, when the manuscript of the present paper was nearing completion. The authors are
grateful to Dr Ross and Professor Mohr for the opportunity to examine this material.

SYSTEMATICS

Genus Syncyamus Bowman, 1955

DIAGNOSIS

Cyamidae with pereopod 2 larger than 1; antenna 1, 4-articulate; antenna 2, 2-articulate;
maxilliped reduced to a simple flap. Unguis not distinct on pereopod 1. Gills simple. Pereon
segments 6 and 7 fused. Type species Syncyamus pseudorcae Bowman, 1955.

The fusion of pereon segments 6 and 7, a key characteristic of this genus, caused some
initial confusion. Although apparently fused, there was sufficient indication of a possible su-
ture line to cast doubt on this character until closer examination of prepared specimens using
interference phase contrast removed any doubts. The impression of a weak suture line, which
is seen with a dissecting microscope under reflected light, may be produced by some subsurface
division or the topography of the cuticle.

Syncyamus aequus sp. nov.

(Figs 1 a-d, 2 a-c, 3 a-c)

DESCRIPTION

Length of body from front margin of head to posterior end of pereon in ovigerous female
2, 3-2, 8 mm, non-ovigerous female 1,6-2, 5 mm, male 1,7-2, 2 mm; maximum body width on
pereon segment 5. No body pigmentation present — all specimens preserved in alcohol. Body
narrowly oval in outline with head strongly immersed in pereon segment 2 (pereon segment 1
is fused with the head); anterolateral margins of pereon segment 2 asymmetrically bilobed
anteriorly (Fig. lb). Pereon segments 3-4 subequal in width, little shorter but wider than
segment 2 in female, very much shorter and narrower in male; pereon segments 5 and 6 sub-
equal in length and width, pereon segment 7 subtriangular, fused with segment 6, posterior
margin weakly sinuous. Ventral pereon surface with 1 pair of short spines on segment 7, 2
pairs on segment 6, and additionally in male only, 1 pair on segment 5. Pleon very small,
bilobed. Head sub-rounded; eyes small, oval.

Antenna 1 small, 4-articulate, with articles 2 and 4 much shorter than 1 and 3, apex
bearing small group of stout setae. Antenna 2 extremely small, 2-articulate; article 2 much
longer than 1, with about 8 stout distal setae. Upper lip fleshy, tending to bilobed (the mouth-
parts were studied in situ using interference phase contrast — the appendages were too fragile
to be satisfactorily removed without destroying the specimen); distal surface of upper lip
densely fringed with fine setae. Mandibles not easily distinguished, left incisor apparently with 7
teeth, right with 6 teeth. Lower lip inner lobes fused into slender setose plate, outer lobes
broad and rounded, setose distally. Maxilla 1 palp short, 1-articulate, about 6 well developed

188

A NEW SPECIES OF THE WHALE-LOUSE SYNCYAMUS

Fig. 1 Syncyamus aequus sp. nov. female holotype a, entire dorsal; b, anterolateral angle of first free pereon segment
c, pereopod 1 ventral; d, pereopod 2 ventral. Bar scale 1,0 mm.

189

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 13, APRIL 1981

Fig. 2 Syncyamus aequus sp. nov. female holotype a, entire ventral; b, pereopod 5 ventral; c, pereopod 7 ventral. Bar
scale 1,0 mm.

190

A NEW SPECIES OF THE WHALE-LOUSE SYNCYAMUS

Fig. 3 Syncyamus aequiis sp. nov. paratypes. a, entire dorsal, male; b, entire ventral, male (right pereopods omitted);
bar scale 1,0 mm; c, mouthparts in situ , female (high magnification interference phase contrast).

191

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 13, APRIL 1981

apical teeth on outer plate. Maxilla 2 a single plate, slightly tapering distally with 4-5 stout
apical setae; left and right maxillae coalesced along proximal three-quarters of mid-line, only
divergent distally. Maxilliped vestigial, formed from small triangular lobes fused into cleft flap,
apex of each lobe with 2-3 stout setae.

Pereopod 1 (Fig. lc) smaller than 2, 5-articulate, propodus subrectangular, palmar margin
straight, transverse, delimited by small tooth. Pereopod 2 (Fig. Id) robust, 4-articulate, pro-
podus slightly longer than wide, anterior and posterior margins strongly convex, palm trans-
verse, delimited by small tooth with accessory tooth on posterior margin close to apex of
dactylus. Pereopods 5-7 (Fig. 2b, c) robust, 5-articulate, basal article with large ventral spine
on anteroproximal angle, article 3 with very large spine on ventral surface and rounded lobe
on inner dorsal surface, propodus oval, palm simple, convex, dactylus strongly curved. Gills
short, bluntly rounded apically, directed anteromedially, little shorter in male than female; no
accessory gills in male. Brood plates triangular, tapering distally, adjacent margins setose.
Holotype female with 10 eggs in brood pouch. Genital valves quadrate.

MATERIAL EXAMINED

4 $$, 1 d\ 2 juv; from Delphinus delphis , 1,86 m, male (PEM N 364), 3 miles 110° off
East London, South Africa, 26 May 1978, floating dead at sea. Collected from mouth 3 $9
(ovig.), body lengths, 2,7 mm (holotype), 2,8 mm, 2,4 mm; 1 9 (non ovig.), 2,2 mm: from
axilla 1 d\ 2,2 mm: from eye 1 juvenile, 1,1 rnm: from blowhole 1 juvenile, 1,2 mm. Registra-
tion nos. PEM K2a (holotype), PEM K2 b-g (paratypes).

3 9?’ 3 cfcf, 3 juv: from snout and blowhole of Stenella coeruleoalba, 2,28 m, female
(PEM N 264), Humewood Beach, Port Elizabeth, 26 December 1975. Body lengths; 9 (ovig.),
2,4 mm; 2 99 (non ovig.), 2,1 mm, 2,0 mm; 3 cfcf, 1,7 mm, 1,8 mm, 1,9 mm. Registration
no. PEM K7.

1 9’ 1 Cf, 2 juv: from blowhole of Stenella coeruleoalba , 1,96 m, male (PEM N 229),
Swartkops River mouth, Algoa Bay, 18 February 1975. Body lengths; 9’ 1,6 mm; cf, 1,7 mm.
Registration no. PEM K8.

1 9* 1 Cf: from snout of Stenella coeruleoalba , estimated length 2,15 m, male (PEM N
443), Swartkops River mouth, Algoa Bay, 12 March 1980. Body lengths; 9 (with brood), 2,3
mm; cf, 1,7 mm. Registration no. PEM Kl.

1 cf, from blowhole of Tursiops aduncus, 2,34 m, male (PEM N 358), Salt Rock, Natal,
17 April 1978. Body length 1,9 mm. Registration no. PEM K3.

1 cf, from blowhole of Tursiops aduncus , 1,72 m, immature male (PEM N 331), Natal
coast, October 1977. Body length 2,0 mm. Registration no. PEM K4.

2 99’ from blowhole of Tursiops aduncus , 2,5 m, male (PEM N 333), Natal shark nets,
October 1977. Body lengths, 2,5 mm, 1,8 mm. Registration no. PEM K5.

1 cf, from blowhole of Delphinus delphis, immature (PEM N 320), 1 mile west Cape
Receife, Algoa Bay, 10 September 1977. Body length 1,8 mm. Registration no. PEM K6.

The holotype and all paratypes are registered in the collections of the Port Elizabeth
Museum, South Africa. One paratype, PEM K2g, has been transferred to the British Museum
(Natural History), London.

DISCUSSION

This is only the second species of Syncyamus named to date, the other being S. pseudor-
cae Bowman found on the False killer whale, Pseudorca crassidens (Owen), in the Gulf of
Mexico (Bowman 1955). The new species Syncyamus aequus is very close to S. pseudorcae, but
can be distinguished by the detailed morphology of pereopods 1 and 2, by the asymmetry of
the anterior lobes of pereon segment 2, and by body size.

192

A NEW SPECIES OF THE WHALE-LOUSE SYNCYAMUS

In a subsequent paper, Bowman (1958) identified as Syncyamus a collection of small
cyamids taken from a dolphin in Panama Bay, although he was uncertain whether they repre-
sented a species distinct from S. pseudorcae. The largest of these Panama amphipods was an
ovigerous female of 3,0 mm body length, considerably smaller than the 4,8 mm female holo-
type of S. pseudorcae , and according to Bowman pereon segments 6 and 7 were not fused. The
presence of a suture between segments 6 and 7 should strictly place the material outside the
diagnosis of Syncyamus , but in the authors’ experience this is a rather subjective character as
already mentioned for the present material. Where segments are ‘relatively fused’ and the
specimens very small, it can be extremely difficult to resolve precise surface structure using a
light microscope. Bowman’s Panama material may belong to the species described in this
paper — the brief notes available, especially the body size and the reference to the inner an-
terior lobe of pereon segment 2 being larger than the outer, are consistent with the present
description. The dolphin from which the Panama cyamids were collected was tentatively iden-
tified for Bowman from a colour photograph as “the long-snouted dolphin, Stenella graffmani
(Lonnberg )”.

Table 1

Cy amid! dolphin associations

HOST

PARASITE

Pseudorca crassidens (Owen)

Syncyamus pseudorcae Bowman

False killer whale

Isocyamus delphini (Guerin-Meneville)

Phocoenoides truei Andrews

Neocyamus physeteris (Pouchet)

True’s dolphin

Stenella graffmani (Lonnberg)

Syncyamus sp.

Gulf of Panama spotted dolphin

Stenella longirostris (Gray)

Syncyamus sp.

Long-beaked dolphin

Stenella coeruleoalba (Meyen)

Syncyamus aequus sp. nov.

Blue-white dolphin

Syncyamus sp.

Delphinus delphis Linnaeus

Isocyamus delphini (Guerin-Meneville)

Common dolphin

Syncyamus aequus sp. nov.

Grampus griseus (Cuvier)

Syncyamus pseudorcae Bowman
Isocyamus delphini (Guerin-Meneville)

Risso’s dolphin or Grampus

Tursiops nuuanu Andrews

Syncyamus sp.

Pacific or Little Bottle-nosed dolphin

Tursiops aduncus (Ehrenberg)

Syncyamus aequus sp. nov.

Indian Ocean Bottle-nosed dolphin

Steno bredanensis (Lesson)

Isocyamus delphini (Guerin-Meneville)

Rough-toothed dolphin

Lagenorhynchus albirostris Gray

Scutocyamus parvus Lincoln & Hurley

North Atlantic White-beaked dolphin

Cephalorhynchus hectori (Van Beneden)

Scutocyamus sp.

Pied Hector’s dolphin

193

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 13, APRIL 1981

The few additional references to Syncyamus in recent literature suggest that their associ-
ation with dolphins deserves further attention. Leung (1967, 1970) gives 4 different hosts and
localities for species listed simply as Syncyamus sp.; the Long-beaked dolphin, Stenella longi-
rostris (Gray), from the Gulf of California, the Common dolphin, Delphinus delphis Linnaeus,
from Gibraltar, the Pacific Bottle-nosed dolphin, Tursiops nuuanu Andrews, from Mexico,
and the Blue-white dolphin, Stenella coeruleoalba (Meyen), from an area between Hawaii and
the Marshall Islands in the Pacific. The latter record is from the same host as that of Syn-
cyamus aequus described in this paper. Leung noted a distinct suture line between pereon
segments 6 and 7 in the Gibraltar specimen mentioned above. The authors have had the
opportunity to examine this specimen and conclude that the posterior segments are in fact
fused, or at least relatively so — there is no suture on the cuticle surface, although a line is
apparent at low magnification under a dissecting microscope. The concept of fused or
coalesced in these small cyamids is open to interpretation and may cause problems in future
work on Syncyamus. The authors measured the body length of the Gibraltar specimen as 4,1
mm, not 3,8 mm as given by Leung, and in their view this cyamid belongs to Bowman’s
species, Syncyamus pseudorcae.

A potential third species, or a senior synonym of one of the two existing species, was
published by Costa (1866) as Cyamus chelipes , from an unidentified dolphin collected in the
Bay of Naples. Unfortunately, Costa’s description and figures are totally inadequate for com-
parison with S. pseudorcae and S. aequus , except to say that the species probably belongs to
the genus Syncyamus (Bowman 1958). It may well have been the same as Leung’s Gibraltar
species which the authors have determined as S. pseudorcae.

In view of Leung’s comments on the lack of host specificity in dolphins it is worthwhile
bringing the list of cyamid/dolphin associations up-to-date (Table 1). Source references for this
tabulation are Bowman (1955, 1958), Leung (1967, 1970), Lincoln & Hurley (1974a, b, present
paper, and in press), Morzer Bruyns (1971).

REFERENCES

Bowman, T. E. 1955. A new genus and species of whale-louse (Amphipoda: Cyamidae) from the False killer whale.
Bull. mar. Sci. Gulf Caribb. 5 (4): 315-320.

Bowman, T. E. 1958. First Pacific record of the whale-louse genus Syncyamus (Amphipoda: Cyamidae). Pacif. Sci. 12 (2):
181.

Costa, A. 1866. Descrizione di una specie di Cyamus parassita de’delfini: C. chelipes. Annuar. R. Mus. zool. R. Univ.
Napoli 3 (1863): 82-83.

Leung, Y.-M. 1967. An illustrated key to the species of whale-lice (Amphipoda, Cyamidae), ectoparasites of Cetacea,
with a guide to the literature. Crustaceana 12 (3): 279-291.

Leung, Y.-M. 1970. First record of the whale-louse genus Syncyamus (Cyamidae: Amphipoda) from the western
Mediterranean, with notes on the biology of odontocete cyamids. In, Pilleri, G. (ed.) Investigations on Cetacea
Vol. 2: 243-247.

Lincoln, R. J. & Hurley, D. E. 1974a. Scutocyamus parvus , a new genus and species of whale-louse (Amphipoda:
Cyamidae) ectoparasitic on the North Atlantic White-beaked dolphin. Bull. Br. Mus. nat. Hist. (Zool.) 27 (2):
59-64.

Lincoln, R. J. & Hurley, D. E. 1974b. Catalogue of the whale-lice (Crustacea: Amphipoda: Cyamidae) in the collec-
tions of the British Museum (Natural History). Bull. Br. Mus. nat. Hist. (Zool.) 27 (2): 65-72.

Lincoln, R. J. & Hurley, D. E. A new species of Scutocyamus (Amphipoda: Cyamidae) on the Pied Hector’s dolphin,
Cephalorhynchus hectori (Van Beneden) from New Zealand. N.Z. II. mar. freshw. Res. (in press).

Morzer Bruyns, W. F. J. 1971. Field guide of whales and dolphins. Amsterdam, C. A. Mees. 258 pp.

Manuscript accepted for publication 24 November 1980.

194

Q 14

i '

f\ 6? 1 3
f\l H

ANNALS

O F T H E

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

Printed by Cape & Transvaal Printers (Pty) Ltd

BD9649

A review of the populations of Kinixys (Testudinidae) occurring in

south-eastern Africa

by

DONALD G. BROADLEY

(National Museum, P.O, Box 240, Bulawayo, Zimbabwe)

CONTENTS

Page

Abstract 195

Introduction 195

Materials and methods 196

Character analysis 197

Systematic account 203

Key to the savanna species of Kinixys 203

Kinixys natalensis Hewitt 204

K. belliana belliana Gray 208

K. belliana spekii Gray 211

Acknowledgements 215

References 215

ABSTRACT

The populations of hinged tortoises found in Africa south of Latitude 25 °S. and east of
Longitude 30 °E. are reviewed. Kinixys natalensis Hewitt is reinstated as a full species. The
southern material of K. belliana Gray appears to represent two subspecies. The large, brightly
patterned tortoises with thick-boned convex carapaces are assigned to the typical form
(synonym: K.b. zulnensis Hewitt), which in this region is restricted to the coastal plain. The
smaller, relatively weakly patterned tortoises with thin depressed carapaces are assigned to
K.b. spekii Gray (synonym: K. australis Hewitt), which inhabits the escarpment and plateau
areas, but descends onto the coastal plain in southern Mozambique.

INTRODUCTION

In the last comprehensive revision of the genus Kinixys , Loveridge & Williams (1957)
recognized a single savanna species with two races: K. belliana nogueyi ocean ing from Came-
roun west to Senegal, and the typical form inhabiting a wide range in savannas of the rest of
sub-Saharan Africa. They were unaware of the description of K.b. mertensi from north-eastern
Zaire by Laurent (1956).

In 1974 Senor J. L. P. L. Tello donated a herpetological collection which included seven
Kinixys from near Ressano Garcia in southern Mozambique. Five of these specimens were
adults of the depressed spekii form of K. belliana , but another, an adult female, with a worn
shell, differed remarkably in proportions and in its relatively small size. Some time later, while

195

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 14, JULY 1981

studying Hewitt’s 1935 paper, the author was struck by the similarity of the “aberrant”
Mozambique tortoise to the old male syntype of K. natalensis Hewitt illustrated in PI. XXXV,
Fig. 4. When the author was able to examine the type series of K. natalensis, it became appar-
ent that this taxon represents a valid species, sympatric with K. belliana. Additional material
was examined from Natal, Zululand, Swaziland and southern Mozambique and it was obvious
that in this region specimens of K. belliana could readily be separated into two subspecies on
the basis of carapace height and colour pattern. A relatively clear-cut division between these
two taxa can be traced northwards at least to Malawi, so the name spekii Gray is revived for
the weakly patterned depressed form.

When the author examined the alleged type specimen of Kinixys belliana Gray in the
British Museum (BM 65.5.3.68: RR 1947.3.5.73), he was surprised to find that it has only four
claws on the forefeet and therefore agrees with the West African race nogueyi. As Gray’s
original description states that there are five claws on the feet, it seems clear that this specimen
is not the type. It is also smaller (carapace length 170 mm) than the specimen described by
Gray in 1831. Further investigation produced another mounted specimen (BM 1979.919) which
had recently been removed from display. Scrawled on the plastron in the legend “bought at
Humphrey’s” and with it is a printed display label recording the fact that the tortoise was
presented by Dr J. E. Gray in 1831. It seems clear that this large female (carapace length 191
mm) with five claws on the forefeet is the true type specimen of Kinixys b. belliana. It has a
shell length/height ratio of 2,10 but the worn shell lacks any pattern.

MATERIALS AND METHODS

The present study is restricted to Kinixys natalensis and sympatric or parapatric popula-
tions of K. belliana. The region covered consists therefore of that part of the generic range
which lies south of Latitude 25 °S. and east of Longitude 30 °E. The range of Kinixys in south-
ern Africa is mapped in Greig & Burdett (1976: Fig. 14).

The most useful diagnostic characters proved to be shape of beak, ratio of shell height/
shell length, and carapace pattern. Several ratios used by Laurent (1956; 1962) in his diagnosis
of K. belliana mertensi were calculated for the southern forms. The measurements were taken
with dial calipers. Gular length and width are maximum measurements, but measurements of
mid-ventral sulci are average figures in the numerous cases when shield margins are not pre-
cisely juxtaposed.

Due to great variability, the only ratio which showed any merit as a diagnostic character
was pectoral sulcus/abdominal sulcus. In K. belliana there is great variation in the relative
lengths of the femoral and anal sulci and this was measured by calculating the ratio femoral
sulcus/abdominal sulcus. The shallow gulars of K. natalensis were expressed in terms of the
ratio gular length/width.

Specimens less than 75 mm in shell length have atypical proportions and were not used in
the statistical analyses.

Shell length to the nearest millimetre between blocks (with carapacial hinge closed) was
taken as an indication of size, because plastron length is dependent on the very variable de-
velopment of the gulars. Shell height was also measured between blocks.

The nomenclature of bones and shields follows Zangerl (1969).

Seventy-five specimens were examined from the following institutions (catalogue number
prefixes in parentheses): Albany Museum, Grahamstown (AM); British Museum (Natural His-
tory), London (BM); Natal Museum, Pietermaritzburg (NM); Transvaal Museum, Pretoria
(TM) and Umtali Museum (UM). The collection from the Umtali Museum has been transferred
to the National Museum, Bulawayo.

A figure in parentheses after a catalogue number indicates number of specimens in a
series.

196

BROADLEY: REVIEW OF POPULATIONS OF KINIXYS (TESTUDINIDAE)

CHARACTER ANALYSIS

Beak

K. natalensis differs from all other species in the genus in having a strongly tricuspid beak
which is similar to that found in Homopus areolatus. The beak is unicuspid in K. belliana
(rarely bicuspid in K.b. nogueyi : Loveridge & Williams, 1957).

Ratio shell length! height (Fig. 1 & Table 1).

The southern material of K. belliana can in most cases be allocated to subspecies on the
basis of this ratio, i.e. less than 2,3 in K.b. belliana. more than 2,3 in K.b. spekii (Coefficient

Fig. 1. Scatter diagram showing ratio of shell height: shell length.

197

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 14, JULY 1981

of difference = 1,54). Loveridge (1936) found a range of 2,3 to 2,5 in adult spekii from north-
ern Tanzania and S.E. Kenya. K. natalensis is intermediate in this respect but closer to typical
K. belliana than to K.b. spekii.

Table 1.

Kinixys Populations South of Latitude 25 °S.

N

Range

Mean

S.D.

C.V.

C.D.’s

Shell Length /Height Ratio

natalensis

21

1,97—2,41

2,17

0,11

5,0

N:B = 0,33

b. belliana

26

1,82—2,34

2,09

0,13

6,2

N:S = 1,26

b. spekii

19

2,29—2,65

2,46

0,12

4,9

B:S = 1,54

Gular Length/Width %

natalensis

21

30,65

43,4

9,8

22,6

N:B = 1,36

b. belliana

26

45—81

67,6

8,0

11,8

N:S = 1,07

b. spekii

18

45—72

62,3

7,9

12,7

B:S = 0,33

Gular + Humeral Sulcus/ Plastron

Width at Humero-Pectoral Sulcus %

natalensis

21

55,5—81,1

68,1

7,3

10,7

N:B = 0,31

b. belliana

26

60,0—84,9

72,4

6,7

9,3

N:S - 0,18

b. spekii

18

58,9—90,6

70,8

8,1

11,3

B:S = 0,11

Pectoral Sulcus/Gular + Humeral

Sulci %

natalensis

21

22,5 — 40,0

31,1

5,2

16,7

N:B = 0,90

b. belliana

26

28,6—64,7

45,2

10,4

23,0

N:S = 0,87

b. spekii

18

31,8—59,1

42,8

8,3

19,0

B:S = 0,13

Pectoral Sulcus/ Abdominal Sulcus %

natalensis

21

18,7—38,2

26,9

5,0

18,6

N:B = 1,21

b. belliana

26

31,2—62,9

43,2

8,5

19,7

N:S = 1,07

b. spekii

20

25,8—52,0

39,8

7,1

17,9

B:S = 0,22

Femoral Sulcus/ Abdominal Sulcus %

natalensis

21

17,5—48,9

32,4

9,7

29,9

N:B = 0,53

b. belliana

26

27,8—68,2

42,4

9,0

21,2

N:S = 0,44

b. spekii

20

17,4—34,1

26,1

4,6

17,6

B:S = 1,20

Thickness of shell.

The southern populations of K.b. belliana have a heavy thick-boned shell with deep sulci
whereas adjacent populations of K.b. spekii have a moderate to thin shell often with a 11-
shaped fontanelle between the nuchal and the first neural (Fig. 11). A similar fontanelle was
noted in the carapace of a K. belliana nogueyi from northern Nigeria. The only available
skeletal shell of K. natalensis is of moderate thickness with no translucent areas.

198

BROADLEY: REVIEW OF POPULATIONS OF KINIXYS (TESTUDINIDAE)

Colour pattern.

In K. natalensis the carapacial pattern is basically zonary, with the areolae brown, encir-
cled by zones of orange/yellow and blackish brown, the latter having a ragged or weakly radi-
ate outer margin which tends to break up the peripheral zone of orange/yellow (Fig. 8, left).
The shields of the plastron are yellow mesially and along the sulci, these areas being separated
by zones of black which are variable in width (Fig. 8, right). The head, limbs and tail are
uniformly yellow. There appears to be no sexual dimorphism in colour pattern in this species.

The carapace of K. belliana belliana usually has a bold black radial pattern on a yellow
ground (Fig. 10). The pattern tends to be more ill defined and broken up in males. The plas-
tron usually has a broad uniformly yellow median zone flanked by an ill-defined radial pattern
(Fig. 10, lower right), which may be vestigial or even absent (especially in males).

Males of K. belliana spekii are uniformly yellow brown or with darker areolae ( spekii
pattern) or with darker areolae separated by a light zone from a broad dark zone (Fig. 12,
upper left). Females are usually more strongly marked than males, having two or three zones
of dark weakly radiate markings which may be convergent (Fig. 12, upper right). Plastron
markings may be distinct in females (Fig. 12, lower right) but are usually vestigial in males
(Fig. 12, lower left).

NATALENSIS

dV

B. BELLIANA

B. SPEKIS

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 192021

Fig. 2. Histogram to show distribution of specimens by sex in each shell length size class for south-eastern populations of
each taxon. The horizontal scale is in centimetres, e.g. class 12 = 116 to 125 mm. The smallest vertical division indicates

one specimen.

199

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 14, JULY 1981
Size and sexual dimorphism (Fig. 2).

Shell lengths for the samples of the three taxa are indicated in Fig. 2. K. natalensis is a
smaller species than K. belliana and the typical form of the latter species attains a larger size
than K.b. spekii. It should be noted that the seven largest specimens of K. natalensis are all
females, indicating that there may be sexual dimorphism in size in this species, a phenomenon
which is not apparent in K. belliana.

Gular lengthlwidth ratio (Fig. 3 & Table 1).

The scatter diagram shows that most K. natalensis have the gulars (together) more than
twice as wide as long whereas in K. belliana they average approximately one and a half times
as wide as long. The Coefficient of Difference between K. natalensis and K.b. belliana is
significant at 1,36 but the C.D. between K. natalensis and K.b. spekii is only 1,07.

Fig. 3. Scatter diagram showing ratio of gular length: width (of pair).

Plastron proportions (Table 1).

Three of the ratios used by Laurent (1956; 1962) in his diagnosis of K. belliana mertensi
were calculated for the present samples and the data appear in Table 1. These ratios provide

200

BROADLEY: REVIEW OF POPULATIONS OF KINIXYS (TESTUDINIDAE)

no clear diagnostic characters for the south-eastern populations but K. natulensis does have a
relatively short pectoral sulcus in relation to the abdominal sulcus (Fig. 4), a ratio used by
Wermuth & Mertens (1961) to key out K.b. mertensi.

The great variation in the relative lengths of the median anal and femoral sulci in a Beira
series of K.b. belliana has been mentioned by Boulenger (1907) but a scatter diagram for the
south-eastern populations shows a strong tendency for typical K. belliana to have the femoral
sulcus longer than the anal sulcus whereas in most K.b. spekii the situation is reversed (Fig. 5).
The variation in K. natalensis is nearly as great as in the other two taxa combined.

The carapacial hinge.

In K. natalensis the hinge is well developed between peripherals seven and eight but even
in the largest adults there is hardly any intervention of cartilage between the fourth and fifth
costal bones.

In adults of K.b. belliana the hinge is very well developed, with the anterior edge of costal
five smooth for more than half its length and cartilage often extending round the whole

201

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 14, JULY 1981

Fig. 5. Scatter diagram showing ratio of femoral sulcus: anal sulcus.

periphery of neural four, although the main hinge runs between neurals four and five. The
hinge between vertebral shields three and four lies in a deep gutter which links the upper ends
of the fifth costals across the anterior half of neural five.

In K.b. spekii the hinge is moderately developed, with cartilage rarely extending more
than half-way up the anterior edge of costal five.

Profile of posterior lobe of carapace.

The three taxa show marked differences in the profile of the carapace posterior to the
hinge. K. natalensis often has well developed protuberences beneath the fourth and fifth ver-
tebrals, the shell sloping at about 45° to the peak of the “knob” of the fifth vertebral and then
falling away almost perpendicularly to the lower edge of the marginals (Fig. 7), in this respect
showing some resemblance to K. homeana.

202

BROADLEY: REVIEW OF POPULATIONS OF KINIXYS (TESTUDINIDAE)

In K.b. belliana the carapace curves gradually from the anterior end of vertebral four
(Fig. 9) whereas in K.b. spekii it slopes very gently to the end of the fourth vertebral and then
falls away abruptly from there (Fig. 11).

Number of marginals.

K. belliana usually has 23 marginals, with 22 or 24 as rare variations due to fusion or
division of individual shields. In K. natalensis the number of marginals bordering the posterior
lobe of the plastron is increased by division of the “supracaudal” (M12) and/or the insertion of
extra marginals each side of it. Only five specimens have 23 marginals, fourteen have 24, three
have 25 and three have 26.

Shape of marginals in juveniles.

In juveniles of K. natalensis the marginal shields are slightly serrated but are not compara-
ble with the multiple sharp spines found on the marginals of the forest species K. erosa and K.
homeana. The marginals of K. belliana are not serrate.

Bones of the carapace.

The usual neural/suprapygal arrangement in K. belliana is 8 neurals + 2 suprapygals, the
only variations in seven K.b. belliana and four K.b. spekii checked for this character are in TM
14877 with the eighth neural fused with the first suprapygal, the type of zulucnsis with an extra
half suprapygal wedged between the standard ones, and a male paratype of K. australis which
has 9 neurals, 3 suprapygals and 9 costals. Six specimens of K. natalensis were checked and
two of these have 9 neurals; two have 3 suprapygals and one has a single suprapygal; two
(including the lectotype female) have 9 pairs of costals, the eighth pair being subdivided.

K. natalensis may consistently differ from K. belliana in having the sulcus between the first
and second vertebrals crossing the second costal bone rather than the first. Inis appears to be
due partly to larger size of the first vertebral shield and partly to the shape of the first costal
(Fig. 7). The same condition occurs in some K. belliana with nine costals.

A single shell of K. natalensis was disarticulated to allow examination of individual bones.
The gular lip of this female specimen (AM 77) is narrow, only 14,2% of the plastron length.
This ratio in K. belliana ranges from 14 to 25%.

The shortening of the plastron in K. natalensis is mainly due to a reduction in the length of
the hyoplastra.

Postorbital bar of skull.

One skull was prepared for each of the south-eastern taxa. Skulls of K. natalensis
(AM 77 — Greytown) and K.b. belliana (AM — St Lucia Bay) lack a postorbital bar whereas a
slender postorbital bar is present in the K.b. spekii skull (AM 78D — White River, E. Tvl.).
This character seems to be very variable in K. belliana.

SYSTEMATIC ACCOUNT
Key to the savanna species of Kinixys

la. Beak tricuspid; gulars (together) usually at least twice as wide as long; marginals usual-
ly 24 or more natalensis Hewitt

lb. Beak unicuspid (rarely bicuspid); gulars (together) usually less than twice as wide as

long; marginals usually 23 2

2a. Forelimb always with 4 claws; range northern Cameroun west to Senegal

belliana nogueyi Lataste

2b. Forelimb usually with 5 claws (occasional specimens have 4) 3

3a. Median pectoral sulcus short, 0 to 31% of both gular + humeral sulci and abdominal
sulcus; range northern Zaire and Uganda belliana mertensi Laurent

203

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 14, JULY 1981

3b. Median pectoral sulcus long, 26 to 69% of gular + humeral sulci and 21-60% of abdo-
minal sulcus 4

4a. Carapace distinctly convex, shell height usually included in length less than 2,3 times;
carapace pattern radial, black on yellow; range eastern Africa from (?) Somalia south

to Zululand belliana belliana Gray

4b. Carapace distinctly depressed, shell height usually included in length more than 2,3
times; carapace pattern basically zonary, males often unpatterned; range central
plateau areas of central and southern Africa belliana spekii Gray

Kinixys natalensis Hewitt
Kinixys belliana zuluensis (part) Hewitt, 1931: 475, PI. xxxviii. Figs 18-20.

Kinixys natalensis Hewitt, 1935: 353, PI. xxxv, Figs 3-4. Type locality: Jameson Drift, Tugela River, Natal. Archer.
1967: 62 & 1968: 13.

Kinixys belliana belliana (part) Loveridge & Williams, 1957: 384; Wermuth & Mertens, 1961: 180 & 1977: 70.

Diagnosis.

Beak tricuspid; gulars (together) usually at least twice as wide as long; marginals usually
24 or more, the posterior “supracaudal” usually divided.

Description.

Beak tricuspid; prefrontal longitudinally divided; frontal entire; forelimb covered with
more or less imbricate scales, some enlarged and pointed, forming on anterior edge a longitu-
dinal series of 7-8 from elbow to outer of five claws; hind foot with four claws.

Carapace moderately convex (shell length/height ratio 1,97 to 2,41), often with flat-topped
protuberences on vertebrals four and five, with end of carapace more or less vertical from top
of posterior knob; anterior margin not or but slightly expanded; posterior marginals slightly
reverted, forming a distinct “gutter”, and feebly serrated in adults (serrated in juveniles);
dorsal shields often raised, and with well marked concentric growth annuli; nuchal elongate in
adults, often minute externally; vertebrals 5, rarely 6 (four specimens); pleurals 4, rarely 5
(one hatchling; on one side of two specimens); marginals 23-26, usually 24, with the posterior
“supracaudal” usually partially or completely divided.

Front lobe of plastron truncate anteriorly, not or but slightly projecting beyond anterior
border of carapace; gulars paired, wider than long; pectorals with a rather narrow median
sulcus; axillaries small (approx. 3); inguinal large (absent only in AM 6975 G), in contact with
or separated from sixth marginal, in contact with femoral; hind lobe short and truncate, with a
very shallow posterior notch. Adult males without a distinctly concave plastron.

Colouration.

Carapace with a concentric pattern on each shield, the areolae light to dark brown, fol-
lowed by a broad orange yellow zone, then a blackish zone that may be broken up into short
rays, it may extend almost to the margin of the shield or may be followed by another orange/
yellow zone. The shields of the plastron have extensive more or less symmetrical blackish
figures, but the shields have yellow centres and margins. Two black rings on the abdominal
shields are particularly prominent (Fig. 8, right). The colour patterns become broken up and ill
defined in large adults.

The old female from near Ressano Garcia (UM 30453) appears to have been uniform
brown above and uniform yellow below throughout life.

204

BROADLEY: REVIEW OF POPULATIONS OF KINIXYS (TESTUDINIDAE)

Fig. 6. Kinixys: distribution of three taxa south of Latitude 25 °S. and east of Longitude 30 °E. Open symbols denote
literature records or specimens identified only from photographs.

205

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 14, JULY 1981

Fig. 7. Kinixys natalensis. AM — , Lectotype 2 from Dimane stream, near Jameson Drift, Tugela Valley, Natal. Dorsal,
lateral, ventral, anterior and posterior views. Heavy lines indicate sulci, thin lines indicate sutures between carapacial

bones.

206

BROADLEY: REVIEW OF POPULATIONS OF KINIXYS (TESTUDINIDAE)

Fig. 8. Kinixys natalensis $. AM 6975 F — Jameson Drift, Tugela Valley, Natal: dorsal and ventral views.

Size.

Largest male (lectoallotype, AM 6975A) 125 mm long, 94 mm wide, 56 mm high; largest
female (AM 6975B) 155 mm long, 1 13 mm wide, 70 mm high.

Remarks.

Hewitt (1935) selected and labelled as types an adult male from Jameson Drift (AM
6975A) and an adult female from nearby Dimane Stream (no catalogue number); his type
series also included 16 paratypes. The female syntype is here nominated lectotype because it is
in better condition and its loose dorsal shields made it possible to remove these and illustrate
the arrangement of neural and costal bones in Fig. 7. The lectoallotype male is atypical in
having only 23 marginal shields.

Habitat.

The type locality was described as “ Euphorbia tirucalli country, on the rather bare and
dry slopes of the Tugela River (altitude 300 m)”. The author has seen a photograph of a
female taken in a dry thicket near the summit of the Lebombo Range (550 m) in Swaziland,
also photographs of a pair at Itala Nature Reserve in Northern Tall Grassveld/Lowveld eco-
tone at ca 900 m altitude. A hatchling from Weenen Nature Reserve was in Valley Bushveld at
900 to 1 000 m altitude.

Distribution.

The Lebombo Range on the Mogambique-Swaziland border, extending south through
western Zululand to Greytown in Natal.

Localities.

MOZAMBIQUE: 10 km SSE of Ressano Garcia UM 30453-4. SWAZILAND: Groen-
pan Farm (photos: J. Culverwell); Ndzinda Nature Reserve (photo: J. Culverwell). ZULU-

207

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 14, JULY 1981

LAND: no precise locality (Hewitt, 1931) NM 1129; Manaba (Hewitt, 1935); Ntambanana
(Hewitt, 1931); Otobotini TM 19346. NATAL: Dimane Stream (Hewitt, 1935) AM — (4);
Greytown (Hewitt, 1931) AM 77, NM 1130; Impanza (Hewitt, 1931) NM 1143 (2); Itala Na-
ture Reserve (photos: O. Bourquin); Jameson Drift (Hewitt, 1935), AM 6975 (12) BM
1934.10.21.1; Weenen Nature Reserve TM 50682.

Kinixys bellianci belliana Gray

Kinixys Belliana Gray, 1831. Synopsis Reptilium: 69. No locality. Rowe-Rowe, Murray & Daniel, 1968: 53; Bourquin,
Vincent & Hitchins, 1971: 21.

Cinixys belliana Boulenger, 1905: 252 & 1907: 482.

Kinixys belliana zuluensis Hewitt. 1931: 471. PI. xxxviii. Figs 21-23. Type locality: Richard's Bay, Zululand. Hewitt,
1935: 350, PI. xxxv. Figs 1-2.

Kinixys belliana belliana Loveridge & Williams, 1957: 384 (part). Wermuth & Mertens, 1961: 180 & 1977: 70 (part);

Pooley, 1965: 54, PI. xi.

Kinixys zuluenisis Archer. 1967: 62 & 1968: 13.

Diagnosis.

Beak unicuspid; carapace distinctly convex, shell height usually included in length less
than 2,3 times; gulars (together) less than twice as wide as long; marginals usually 23, the
posterior ‘supracaudal’ undivided; colour pattern of carapace radial.

Description.

Beak unicuspid; prefrontal entire (but usually with a median cleft anteriorly) or longitudi-
nally divided; frontal large, rarely longitudinally divided; forelimb covered with more or less
imbricate scales, some enlarged and pointed, which on anterior edge form a longitudinal series
of 8-10 from elbow to outer of five claws (6 claws on right foot of TM 29400); hind foot with
four claws.

Carapace moderately convex (shell length/height ratio 1,82 to 2,34), highest point usually
on vertebral 3 or 4; vertebrals and pleurals flat or convex with deep sulci; anterior margin
often somewhat expanded, not reverted; posterior marginals not or but slightly reverted, not
serrated; dorsal shields with well marked growth annuli and deep sulci; nuchal elongate in
adults; vertebrals 5; pleurals 4 (5 on left side of type of K.b. zuluensis Hewitt); marginals 23
(24 in four specimens, only one due to division of 'supracaudal’ M12).

Front lobe of plastron truncate anteriorly, usually projecting beyond anterior border of
carapace, especially in males; gulars paired, usually less than twice as broad as long; pectorals
with a moderate median sulcus; axillaries 2-3, often very narrow; inguinal large, usually in
contact with sixth marginal, in broad contact with femoral; hind lobe short and truncate, with a
very shallow posterior notch. Plastron distinctly concave in adult males.

Colouration.

Carapace yellow, with a black radial pattern (usually 4-6 rays) on each vertebral and
pleural shield. In females the black rays may be largely confluent as in TM 14877 (Hewitt,
1935, PI. xxxv. Fig. 1), but at the other extreme is TM 13550, with only a few narrow black
rays on the vertebrals and pleurals (Hewitt, 1931, PI. xxxviii. Fig. 22). In males the black
pattern is usually less well developed than in females and it is obscure in senile individuals.

The plastral pattern is very variable. In juveniles the plastron is largely black except for a
broad median band of yellow extending along the median sulci of the pectorals, abdominals
and femorals. Thereafter the black is broken up by yellow intrusions until large adults retain
only vestigial black markings, usually in a vaguely radial pattern.

208

BROADLEY: REVIEW OF POPULATIONS OF KINIXYS (TESTUDINIDAE)

Fig. 9. Kinixys belliana belliana. TM 34680, 9 from Ndumu Game Reserve, Zululand, Dorsal, lateral, ventral, anterior

and posterior views. Conventions as in Fig. 8.

209

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 14, JULY 1981

Fig. 10. Kinixys belliana belliana. Left O' type of K.b. zuluensis Hewitt, NM 1203 — Richard’s Bay, Zululand: dorsal and
ventral views. Right, $ AM — St Lucia Bay, Zululand: dorsal and ventral views.

210

BROADLEY: REVIEW OF POPULATIONS OF KINIXYS (TESTUDINIDAE)

The head is yellow, sometimes mottled with grey. The forelimbs are usually mottled yel-
low and black, the enlarged scales yellow; the hindlimbs and tail are yellow.

Size.

Largest male (TM 51378 — Candover to Mkuze, Zuluiand) 211 mm long, 137 mm wide, 90
mm high; largest female (TM 43074 — 20 km NNW of Hluhluwe, Zuluiand) 205 mm long, 142
mm wide, 98 mm high.

Remarks.

The male type of K.b. zuluensis (NM 1203), not illustrated by Hewitt (1931), is shown in
Fig. 10 left. It is aberrant in having an extra pleural shield anteriorly on the left side, a super-
numerary marginal anteriorly on the right side and the first marginal on the right side is longer
than broad, as in K.b. mertensi and K.b. nogueyi.

Habitat.

Hewitt (1935: 355) states that “zuluensis seems to belong to the low country of palms,
swamps and sand dunes” and in the southern part of its range K.b. belliana does appear to be
restricted to the coastal plain (classed as Coastal Forest and Thornveld). Haacke (in litt.) has a
sight record of specimens in coastal dune forest at Mapelane, south of Lake St Lucia. TM
51378 was found on a road through savanna thornveld west of the Lebombo Range.

Ecology.

At Lake Sibayi, the faeces of two tortoises contained remains of pill millipedes (Glomer-
idae), shells of small snails and stones/skins of the fruit Syzygium cordatum , the latter fruit also
being found in faeces of a tortoise from Kosi Bay (Rowe-Rowe et al., 1968). In St Lucia Park
Kinixys were seen eating mushrooms with avidity (Rowe-Rowe et al., 1968).

Distribution.

East Africa from? Somalia south to Zuluiand.

Localities.

MOZAMBIQUE: Chimonzo TM 29399-400; Maputo Elephant Reserve TM 41762.
ZULULAND: no precise locality AM — (2); Candover to Mkuze TM 51378; Hluhluwe Game
Reserve (Bourquin et al., 1971); 20 km NNW of Hluhluwe TM 43074-5; Kosi Bay (Rowe-
Rowe et al., 1968); Lake St Lucia (Archer, 1967, 1968; Rowe-Rowe et al., 1968) TM 45828-9;
Lake Sibayi (Rowe-Rowe et al., 1968) TM 48265; Lalanek TM 47844; Manaba (Hewitt, 1935)
TM 14877; Manzangwenya TM 48729-30; Mtubatuba (Hewitt, 1931, 1935; Archer 1967, 1968)
TM 13544, 1549-50; Mkuzi Game Reserve (Pooley, 1965); Ndumu Game Reserve (Pooley,
1965) TM 34678-81; Otobotini TM 19363; Richard’s Bay (Hewitt, 1931) NM 1203, type of
K.b. zuluensis ; St Lucia Bay AM — (2); Umfolozi Game Reserve (Archer, 1967, 1968; Bour-
quin et al., 1971); Umfolozi Station (Boulenger, 1905; Hewitt, 1931, 1935) BM 1905.3.7.66 &
68.

Kinixys belliana spekii Gray

Kinixys spekii Gray, 1863, Ann. Mag. nat. Hist. (3) 12: 381 Central Africa.

Kinixys australis Hewitt, 1931: 477, PI. xxxvi. Figs 4-6. Type locality: White River, Eastern Transvaal.

Kinixys belliana belliana Loveridge & Williams, 1957: 384 (part); Wermuth & Mertens, 1961: 180 (part); Pienaar, 1966:
127 ; Switak, 1971: 9; Wermuth & Mertens, 1977: 70 (part); Pienaar, 1978: 213.

Kinixys zombensis (not Hewitt) Archer, 1967: 64 & 1968: 35.

211

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 14, JULY 1981

Diagnosis.

Beak unicuspid; carapace depressed, shell height usually included in length more than 2,3
times; gulars (together) usually less than twice as wide as long; marginals usually 23, the post-
erior ‘supracaudal’ undivided; colour pattern of carapace (if present) zonary.

Description.

Beak unicuspid; prefrontal usually longitudinally divided; frontal entire; forelimb covered
with more or less imbricate scales, some enlarged and pointed, which on anterior edge form a
longitudinal series of 7-10 from elbow to outer of 5 claws; hind foot with 4 claws.

Carapace depressed (shell length/height ratio 2,29 to 2,65), highest point frequently on
vertebral 1 or 2; dorsal shields relatively flat, areolae of vertebrals usually with a weak longitu-
dinal keel; anterior margin not or but feebly expanded, not reverted; posterior marginals
weakly reverted, not serrated; dorsal shields with well marked growth annuli except in old
males, which usually have very worn shells; nuchal moderate to elongate in adults; vertebrals 5
(6 in AM 78D only); pleurals 4 (5 in AM 78D only); marginals 23 (22 in 2; 24 in 2).

Front lobe of plastron truncate anteriorly, projecting well beyond the anterior border of
the carapace in both sexes; gulars paired, usually less than twice as broad as long; pectorals
with a moderate median sulcus; axillaries two, of moderate size; inguinal large, usually in
contact with sixth marginal, in broad contact with femoral; hind lobe short and truncate, with a
very shallow posterior notch. Plastron usually concave in adult males.

Colouration.

Carapace with a zonary pattern in juveniles and subadults. This may persist in adult
males, but they often become uniformly olive-brown or yellow-brown. In females the dark
zones tend to break up into ragged radiations. Fig. 12, upper right shows an unusually strongly
marked female.

A typical female or juvenile plastral pattern is shown in Fig. 12, lower right. In adult
males the plastron is uniformly yellow or may show vestiges of this pattern.

The head, limbs and tail are usually uniformly yellow, but UM 30451 (the strongly marked
female shown in Fig. 12, right) has black speckling on the forelimbs.

Size.

Largest male (TM 42735 — north of Manzini, Swaziland) 181 mm long, 125 mm wide, 77
mm high; largest female (AM 78C — White River, E. Transvaal, lectotype of K. australis ) 183
mm long, 115 mm wide, 70 mm high.

Remarks.

None of the photographs of the type specimens of K. australis (Hewitt, 1931: PI. xxxvi,
Figs 4-6) fits any of the White River specimens now in the Albany Museum collection and the
plastron of the female does not bear the collector’s number “LXXVIII” which appears on all
except the smallest of the White River specimens. As Hewitt does not indicate how many
specimens he had from White River, it is not known if some were destroyed in the fire at the
Albany Museum in 1941, when the early catalogues were destroyed. However, only the large
female bears a label inscribed “Type Kinixys belliana australis Hwtt” in Hewitt’s handwriting
and it almost fits the measurements provided by Hewitt, so it is here nominated lectotype. The
paratype series consists of two adult males, two small females and a 60 mm juvenile with a
separate catalogue number (AM 1295).

Habitat.

The area occupied by this taxon south of Latitude 25 °S. is mostly classified as Tropical
Lowveld, but the southernmost locality is Weenen in the (Karroid) Valley Bushveld.

212

BROADLEY: REVIEW OF POPULATIONS OF KINIXYS (TESTUDINIDAE)

Fig. 11. Kinixys belliana spekii. AM 78C, lectotype J of K. australis Hewitt from White River, E. Transvaal. Dorsal,
lateral, ventral, anterior and posterior views. Conventions as in Fig. 8. f = fontanelle.

213

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 14, JULY 1981

Fig. 12. Kinixys belliana spekii. Left, cf UM 30450; Right, 9 UM 30451; both from 10 km SSE of Ressano Garcia,

Mozambique, dorsal and ventral views.

214

BROADLEY: REVIEW OF POPULATIONS OF KINIXYS (TESTUDINIDAE)

Distribution.

Savannas of much of southern and central Africa excluding the eastern coastal strip inhab-
ited by typical K. belliana. Range extends north at least to Angola, southern Zaire, western
Tanzania and Kenya (Loveridge, 1936).

Localities.

TRANSVAAL: Farm De Hoop, Nelspruit TM 42393; Komatipoort (Archer, 1967, 1968);
Kruger National Park (Pienaar, 1966, 1978; Switak, 1971); White River (Hewitt, 1931 AM 77
(5), 1295 MOZAMBIQUE: Maputo Elephant Reserve TM 41761 (this specimen may be a
waif from further west, perhaps swept into a tributary of the Maputo River by a flash flood);
Moamba TM 29516-7; 10 km SSE of Ressano Garcia UM 30450-1, 30455, 33010; 15 km SSE
of Ressano Garcia UM 30452. SWAZILAND: near Manzini TM 24205; north of Manzini TM
42735; 5 km NW of Siphofaneni TM 51023; Tshaneni UM 33418. ZULULAND: Ndumu
Game Reserve TM 37895. NATAL; Weenen Nature Reserve TM 52219.

ACKNOWLEDGEMENTS

This study is based almost entirely on borrowed material. The author is extremely grateful
to Dr P. H. Skelton and Mr J. C. Greig for permitting him to examine the entire Albany
Museum collection of Kinixys, Mr W. D. Haacke for sending him the relevant material from
the Transvaal Museum collection and to Dr B. R. Stuckenberg for providing material from the
Natal Museum.

Mrs F. M. Clift typed the manuscript, Mr M. L. Angell took the photographs and Mr &
Mrs J. Akester printed them.

REFERENCES

Archer, W. H. 1967. The tortoise with a difference: Kinixys species or the hingebacks. Afr. wild Life 21 (1): 59-66.

Archer, W. H. 1968. (Reprint of 1967 article). J. Int. Turtle & Tortoise Soc. 2 (4): 11-13, 35-36.

Boulenger, G. A. 1905. On a collection of batrachians and reptiles made in South Africa by Mr C. H. B. Grant, and
presented to the British Museum by Mr C. D. Rudd. Proc. zool. Soc. Lond. 2: 248-255.

Boulenger, G. A. 1907. Second report on the batrachians and reptiles collected in South Africa by Mr C. H. B. Grant,
and presented to the British Museum by Mr C. D. Rudd. Proc. zool. Soc. Lond. 2: 478-487, Pis xxi-xxii.

Bourquin, O., Vincent, J. & FIitchins, P. M. 1971. The vertebrates of the Hluhluwe Game Reserve — Corridor (State-
land) — Umfolozi Game Reserve Complex. Lammergeyer (14): 5-58, Figs 1-61.

Greig, J. C. & Burdett, P. D. 1976. Patterns in the distribution of southern African terrestrial tortoises (Cryptodira:
Testudinidae). Zool. Afr. 11 (2): 249-273.

Hewitt, J. 1931. Descriptions of some African tortoises. Ann. Natal Mus. 6 (3): 461-506, Pis xxxvi-xxxviii.

Hewitt, J. 1935. Some new forms of batrachians and reptiles from South Africa. Rec. Albany Mus. 4: 283-357,
Pis xxvii-xxxvi .

Laurent, R. F. 1956. Contribution a l’Herpetologie de la Region des Grands Lacs de l’Afrique Centrale. Annls Mas. r.
Congo beige Ser. 8vo, 48: 1-390, Pis i-xxxi.

Laurent, R. F. 1962. On the races of Kinixys belliana Gray. Breviora 176: 1-6.

Loveridge, A. 1936. Scientific results of an expedition to rain forest regions in eastern Africa. V. Reptiles. Bull. Mus.
comp. Zool. Harv. 79 (5): 207-539, Pis i-iv.

Loveridge, A. & Williams, E. E. 1957. Revision of the African tortoises and turtles of the suborder Cryptodira. Bull.
Mus. comp. Zool. Harv. 115 (6): 161-557, Pis i-xviii.

Pienaar, U. de V. 1966. The reptiles of the Kruger National Park. 223 pp. Pretoria: National Parks Board.

Pienaar, U. de V. 1978. The reptile fauna of the Kruger National Park. 222 pp. Pretoria: National Parks Board of
Trustees.

Pooley, A. C. 1965. A preliminary check-list of the reptiles found within the Mkuzi and Ndumu Game Reserves in
northern Zululand. Lammergeyer 3 (2): 41-55.

215

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 14, JULY 1981

Rowe-Rowe, D. T., Murray, R. M. & Daniel, P. J. 1968. Miscellaneous notes of natural history interest. 29. Tortoise
diet. Lammergeyer (9): 53.

Switak, K. H. 1971. In Kruger National Park. J. Int. Turtle & Tortoise Soc. 5 (1): 7-9, 27.

Wermuth, H. & Mertens, R. 1961. Schildkroten, Krokodile, Briickenechsen. xxvi+422 pp. Jena: G. Fischer.
Wermuth, M. & Mertens, R. 1977. Liste der Rezenten Amphibien und Reptilien. Testudines, Crocodylia, Rhyn-
chocephalia. Tierreich 100: i-xxvii, 1-174.

Zangerl, R. 1969. The turtle shell, pp. 311-339. In Biology of the Reptilia 1. Morphology A. London and New York:
Academic Press.

Manuscript accepted for publication 24 November 1981.

VX

\

'A icV3

/ n w

A N N A L S

O F T H E

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

Ann. Cape Prov. Mus. (nat. Hist.)

VOLUME 13 • PART 15
10th AUGUST 1981

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

Printed by Cape & Transvaal Printers (Pty) Ltd, Cape Town

BD9672

The description and osteology of a new species of Gephyroglanis
(Siluriformes, Bagridae) from the Olifants River, South West Cape, South

Africa

by

P. H. SKELTON

(Albany Museum, Grahamstown, 6140, R.S.A.)

CONTENTS

Page

Abstract 218

Introduction 218

Methods 218

Description of Gephyroglanis barnardi sp. nov 220

Material examined 220

Diagnosis 220

Comparison with Gephyroglanis gilli and G. sclateri 225

Distribution, habitat preferences and status 225

Etymology 227

Osteology 227

Neurocranium 227

Infraorbitals and nasal bones 232

Suspensorium, palatine-maxilla and preopercular 232

Opercular Series 235

Lower jaw 235

Hyoid arch 235

Branchial skeleton 237

Pectoral girdle 239

Weberian apparatus 239

Precaudal vertebrae 242

Caudal vertebrae 242

Caudal skeleton 242

Dorsal fin skeleton 242

Pelvic fin skeleton 243

Anal fin skeleton 243

Discussion 245

Acknowledgements 248

References 249

217

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 15, AUGUST 1981

ABSTRACT

I

A new small bagrid catfish referred to Gephyroglanis is described from the Olifants River,
western Cape Province, South Africa. The osteology of the new species is described and
figured. Comparisons with other South African Gephyroglanis species are drawn and queries
concerning the generic placement of these species are raised.

INTRODUCTION

The Olifants River system in the south-western Cape Province (Fig. 7) has a particularly
high proportion of endemic fish species. One of these, Gephyroglanis gilli Barnard, 1938, is a
relatively small bagrid catfish. A second Gephyroglanis species, G. sclateri Boulenger, 1901,
occurs in the Orange River system to the north. Other Gephyroglanis species, found in the
Zaire River and in rivers of West Africa (Boulenger, 1911; Jayaram, 1966), are remote from
the southern African species.

Barnard (1943) commented on the intraspecific variation of both southern African
Gephyroglanis species pointing out that two rather distinct “fin forms” are found: one with a
high ragged dorsal fin and the other with a low dorsal fin. Barnard op. cit. (p. 229) remarked
that if the extremes of each form were taken from different river systems they could “almost
be regarded as constituting a specific difference”. He considered, however, that the differences
were phenotypic in nature, the high fin form coming from rocky, flowing waters and the low fin
form from placid, muddy waters.

During routine identification of fishes from a survey of the Olifants River system con-
ducted by the Department of Nature and Environmental Conservation, the author was im-
pressed by the remarkable difference of fin form between specimens of Gephyroglanis. Further j
investigations revealed that the differences of fin form were correlated with other characters
which, contrary to Barnard’s finding, indicated specific distinction. Re-examination of existing
collections revealed that several, including the syntype series of G. gilli, contained two clearly
different species of which one is referable to G. gilli and the other is here described as new.

METHODS

Measurements were taken to the nearest 0,1 mm and are expressed as percentages of
standard length.

Linear measurements (Fig. 1) were made according to Hubbs & Lagler (1958) except for
the following;

Pectoral to pelvic fin length is the median measurement between the posterior margins of
the bases of the pectoral and pelvic fins.

Pelvic to anal fin length is the median measurement between the posterior margin of the
base of the pelvic fin and the base of the first anal fin ray.

Predorsal length is measured from the symphysis of the premaxillae to the base of the first
dorsal fin spine.

Humeral process length is measured from the base of the pectoral spine to the tip of the
humeral process.

Vertebral counts were determined from radiographs. Counts were taken according to
Skelton (1976).

Anal fin rays were counted as far as possible by dissection. The number of anterior un-
branched rays is nevertheless difficult to determine accurately because these rays are fine and
small. Completely reliable counts are available from six cleared and stained specimens.

Specimens were cleared and stained for bones and cartilage using Taylor and Van Dyke’s
(1979) modified trypsin method (Alcian blue for cartilage, alizarin red for bone).

218

SKELTON: DESCRIPTION AND OSTEOLOGY OF GEPHYROGLANIS

SL

Fig. 1. Measurements as made in this study on Gephyroglanis.

Abbreviations: A — anal fin base; aMB— anterior mandibular barbel; BD — body depth; BW — body width; CPD — caudal
peduncle depth; CPL — caudal peduncle length; D — dorsal fin length; HL — head length; HP — humeral process length;
HW — head width; IO — inter orbit; MxB — maxillary barbel; O — orbit diameter; P — pectoral fin length; P-A — pelvic to
anal fin; P-P — pectoral to pelvic fin; pMB — posterior mandibular barbel; Pre D — predorsal length; S — snout length;

SL — Standard length.

Abbreviations: AM/P-Albany Museum fish collection, Grahamstown; BMNH-British
Museum (Natural History), London; MNHN-Museum National d’Histoire Naturelle, Paris;
MRAC-Musee Royal de l’Afrique Centrale, Tervuren; USNM-United States National
Museum (specimens in National Museum of Natural History, Smithsonian Institution),
Washington; RUSI-J. L. B. Smith Institute of Ichthyology, Grahamstown; SAM-South Afri-
can Museum, Cape Town.

219

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 15, AUGUST 1981

Fig. 2. Holotype of Gephyroglanis barnardi sp. nov., male of 68,5 mm SL. AM/P 7647 (a)

DESCRIPTION

Gephyroglanis barnardi sp. nov.

Barnard’s rock catlet (Fig. 2)

Material examined

Holotype: AM/P 7647 (a) cf 68,5 mm SL. Noordhoeks River at roadbridge, tributary of
Olifants River, Cape Province, S.A. (32° 43' 15"S; 19° 03'59"E). Electrofished by K. Hamman
& S. Thorne, 3 September 1979.

Paratypes: 3, AM/P 7647 (b) 71 mm SL; 57,4 mm SL, 51,5 mm SL. Collected together
with holotype. 29, (ex) AM/P 1369 33,5-67,2 mm SL. Collected at same site as holotype by
F. L. Farquharson, 7 April 1967. (5 specimens sent to BMNH, 5 to USNM, 5 to
MNHN). 14, AM/P 1879 (4 cleared and stained). Collected at same site as holotype by K.
van Rensburg, 17 March 1965. (5 specimens to RUSI). 7, AM/P 893 (2 cleared and stained).
Collected at same site as holotype by K. van Rensburg, 17 March 1965. 7, SAM 29232. Col-
lected by K. H. Barnard and C. W. Thorne, Noordhoeks River and Olifants River mainstream
at Keerom, February 1939, (ex-type series of G. gilli).

Other Material: 1, AM/P 8202, ripe 74,5 mm SL. Noordhoeks River at roadbridge. Elec-
trofished by P. H. Skelton and S. Thorne, 6 January 1981. 1, AM/P 8206, ripe 5 mm SL,

Thee River (32° 47' 49"S, 19° 05' 31"E). Electrofished by P. H. Skelton and S. Thorne, 7
January 1981.

Diagnosis

G. barnardi is relatively small, the longest specimen examined is 75 mm SL. Mor-
phometric proportions recorded in Table 1. Head depressed and flat with squarish profile from
above, 6,6 to 8 times in SL, broadest at level of the opercle, dorsal surface entirely covered by
skin. Angle of head shallow (20-30°) rising straight or with a very shallow arch to the nape and
dorsal fin. Eyes dorso-lateral, small, directed dorsally and widely spaced; mouth straight, sub-
terminal with moderately papillose and fleshy lips.

Nares widely separate, located antero-medial to orbits; anterior short and tubular, post-
erior elongated. Each posterior naris bordered in front with a short compressed barbel (Fig. 4)
the base of which extends posteriad as a low narrow ridge on either side of the naris. Snout
bluntly curved.

220

SKELTON: DESCRIPTION AND OSTEOLOGY OF GEPHYROGLANIS

I

Fig. 3. Gephyrogtanis gilli Barnard, male of 97,4 mm SL. AM/P 7646 collected in same locality as holotype of

G. barnardi.

ANTERIOR MANDIBULAR BARBEL POSTERIOR MANDIBULAR BARBEL

5mm

Fig. 5. Ventral view of head of holotype of G. barnardi sp. nov. to show mouth
form and position of mandibular barbels.

Fig. 4. Left lateral view
of nares and nasal barbel
of holotype of G. barnar-
di sp. nov. Scale bar = 1
mm.

221

Table 1.

Comparative measurements (%SL) of southern African Gephyroglanis species: G. barnardi holotype and paratypes; G. gilli — AMIP 1369; G. sclateri

AM! P 3921 (3), 1477 (8), 7840 (1), 1383 (7). Measurements of holotype in mm.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 15, AUGUST 1981

222

SKELTON: DESCRIPTION AND OSTEOLOGY OF GEPHYROGLANIS

Gill openings continuous ventrally and ascending on either side to the dorso-lateral sur-
face of head. Branchiostegal membranes free. Branchiostegal rays 6-6 (N 6).

Four pairs of barbels present including the short nasal pair described above, a pair of
simple maxillary barbels and two pairs of simple mandibular barbels. Maxillary barbels extend-
ing lateral from upper jaw, not reaching beyond posterior opercular groove. Anterior man-
dibular barbels placed ventrally just posterior to lower lip (Fig. 5). Posterior mandibular bar-
bels slightly longer than anterior pair, located behind and lateral to the anterior pair opposite
the gular node (Fig. 5).

Body broadest anteriorly becoming increasingly compressed caudad. Caudal peduncle
laterally compressed, twice as long as deep. Anus located posterior to base of pelvic fins and
covered by fan of pelvics. Skin smooth.

Dorsal fin short, with straight posterior edge, basal third fleshy. Dorsal fin spine without
serrations. Pectoral fins rounded. Anterior ray gently curved and spinous, serrated on post-
ero-medial edge with 6 (juveniles) to 12 or 13 (usually 10 or 11 in adults) basally directed
serrations. Pectoral fin formula I (N 54), 6 (N 4), 7 (N 50).

Pelvic fins rounded overlapping ventrally and short, not reaching anal fin. Pelvic fin for-
mula I, 5 (N 54).

Anal fin rounded, fleshy over basal third, formula in Table 2.

Last anal ray divided or undivided to base. Caudal fin irregularly truncate, extending
anteriorly on the dorsal and ventral edges of caudal peduncle. Principal caudal rays usually 8 +
9 (N 40), less frequently 9 + 8 (N 7); 9 + 9 (N 2); 7 + 9 (N 1); 7 + 8 (N 2); 8 + 8 (N 2)
(holotype has 9 + 8). Procurrent rays as determined from cleared specimens: dorsal 16 (N 4),
17 (N 1), 18 (N 1); ventral 15 (N 1), 17 (N 4), 18 (N 1) (Table 4).

Adipose fin long and well developed, originating behind pelvic fin base, posteriorly de-
tached.

Cephalic lateral line system unfragmented, course of canals shown in Fig. 9.

Lateral line straight, simple, extending mid-laterally from head to posterior end of caudal
peduncle.

Alimentary canal (Fig. 6) short, with stomach expanded, bulbose and leading to the intes-
tine on the left. Intestine crosses anteriorly to right before passing posteriad ventro-laterally
and tucking behind the stomach. Thereafter it proceeds directly to the anus.

Gonads paired, ovaries simple, testes villiform.

Vertebral counts are recorded in Table 3.

Table 2.

Distribution frequency of anal fin rays (A) G. barnardi N=54 ( B ) G. gilli N=20 (C) G.

sclateri N=20

Branched rays

A

10 11 12 13

B

10 11 12 13

C

10 11 12 13

iii

Unbranched iv
Rays v

vi

1 2 1

6 16 7 1

4 9 3 1

1 2

6 6 11

3 1 1

13 3 1

3 8

223

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 15, AUGUST 1981

Table 3.

Comparative distribution frequencies of vertebral counts of South African Gephyroglanis species

A. VERTEBRAE

N

41

42

43

44

45

46 47

G. barnardi

(32)

2

19

10

1

G. gilli

(20)

1

9

10

G. sclateri

(5)

1 4

B. PRECAUDAL VERTEBRAE

N

11

12

13

14

15 16 17

G. barnardi

(32)

15

10

6

1

G. gilli
G. sclateri

(20)

(5)

3

7

9

1

1 3 1

C. CAUDAL VERTEBRAE

N

28

29

30

31

32

G. barnardi

(32)

5

10

12

5

G. gilli

(20)

1

4

10

4

1

G. sclateri

(5)

2

2

1

D. PREANAL FIN VERTEBRAE

N

19

20

21

22 23 24

G. barnardi

(32)

3

21

8

G. gilli

(20)

2

8

8

2

G. sclateri

(5)

2 1 2

Table 4.

The number of caudal procurrent rays in six cleared and stained paratypes of Gyphyroglanis

barnardi

Dorsal Procurrent Rays

16 17 18

Ventral

15

Procurrent

16

Rays

17

18

224

SKELTON: DESCRIPTION AND OSTEOLOGY OF GEPHYROGLANIS

Fig. 6. Alimentary canal within the body cavity of Gephyroglanis barnardi
sp nov. (AM/P 1879 No. 2). Abbreviations: a — anterior; p — posterior;
i — intestine; s — stomach.

Comparison with Gephyroglanis gilli and G. sclateri

Gephyroglanis barnardi may be distinguished from the sympatric G. gilli by the following
characteristics; it is a smaller species (maximum recorded SL = 71 mm v. 127 mm of G. gilli):
G. barnardi has dark brown spots and blotches, G. gilli is invariably without blotches or mark-
ings; G. barnardi has more slender curved dorsal and pectoral fin spines and these fins are
shorter and more rounded than those of G. gilli which has well developed strong dorsal and
pectoral spines and longer, straight, angulate fins (Fig. 3); the posterior edge of the caudal fin
is more or less truncate in G. barnardi , in G. gilli this is emarginate (Figs 2 & 3); G. barnardi
has fewer dorsal branched rays (six) than G. gilli (seven).

G. barnardi differs in several morphometric and osteological characters from G. gilli,
particularly in having a smaller eye, shorter barbels, shallower head and body depth, shorter
humeral process, nuchal plates separate from the supraoccipital process (in G. gilli the nuchal
plates of the dorsal fin interlocks with the supraoccipital process), the caudal fin skeleton re-
latively consolidated (i.e. fused) compared with G. gilli, more vertebrae than G. gilli (Table
3).

G. barnardi is distinguished from G. sclateri , an Orange River endemic species, by its
smaller maximum size, smaller eye, shorter snout, shallower body depth, longer caudal pedun-
cle, shorter dorsal fin, number of branched dorsal fin rays, shorter humeral process and in the
position of the posterior mandibular barbels. In addition the cephalic lateral line pores of
G. barnardi are simple, those of G. sclateri are dendritic. The caudal fin of G. barnardi is
truncate whereas that of G. sclateri is moderately forked. The caudal skeleton is more consoli-
dated in G. barnardi. The supraoccipital process is joined with the first nuchal plate in
G. sclateri but not in G. barnardi. There are fewer vertebrae in G. barnardi than in G. sclateri
(Table 3).

Distribution, habitat preferences and status

G. barnardi is known only from the Noordhoeks and Thee rivers, two adjacent tributaries
of the Olifants River system of the western Cape Province (Fig. 7). Both these tributaries are
perennial mountain streams with gradients of approximately 1:30 (Noordhoeks) and 1:45
(Thee) over the accessible reaches where the species has been collected. The streams vary in
width from three to six or seven metres wide during normal flow and are bedded with loose
round rocks, pebbles and coarse clean sand. Vegetation is generally restricted to marginal

225

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 15, AUGUST 1981

Fig. 7. The distribution of Gephyroglanis barnardi sp. nov. (closed circles) and G. gilli (open circles) in the
Olifants River system, western Cape Province. Records as given in list of type material (G. barnardi) and in
Appendix 1 (G. gilli). Half closed circle refers to SAM 19359/SAM 29232 from which either or both species

may have been collected.

226

SKELTON: DESCRIPTION AND OSTEOLOGY OF GEPHYROGLANIS

patches, frequently of Cyperaceae. Riffles and shallow runs alternate with deeper pools and
the water is clear. G. barnardi is found together with G. gilli and the redfin minnow, Barbus
phlegethon , in shallow (less than 1 m deep) flowing riffle stretches.

The confinement of G. barnardi to these two tributaries is surprising. Barnard may have
collected this species from the mainstream at Keerom (cf SAM 29232) but it is possible that his
specimens of G. barnardi may have been derived only from Noordhoeks as the evidence is
equivocal. The habitat of the mainstream at Keerom is very similar to that in the Noordhoeks
and Thee except that it is deeper, up to about two metres. Extensive surveys by the author and
officials of the Department of Nature and Environmental Conservation throughout the Oli-
fants system indicate that the species is probably restricted to these two streams and possibly to
a few other adjacent streams in inaccessible localities. This restriction may be due in part to
specific habitat preferences — which appear similar to those of B. phlegethon. This minnow
species occurs almost exclusively in such shallow flowing, loose rock and pebble bottomed
habitats of the clear water western Cedarberg tributaries of the mainstream Olifants River.
The two strongest populations of B. phlegethon are those in the Noordhoeks and Thee rivers.

It is also likely that ecological changes in the Olifants River system resulting from human
interference have acted to restrict G. barnardi and other indigenous species including
B. phlegethon (Gaigher, 1978; Skelton, 1977). Agricultural practices and the introduction of
exotic predatory fish species, notably smallmouth bass ( Micropterus dolomieu) are undoubted-
ly the main factors involved (pers. obs.). From available records it would appear that
G. barnardi is probably the rarest species in the Olifants and is therefore clearly an endangered
species.

Etymology

The species is named for the late Dr Keppel Harcourt Barnard, former Director of the
South African Museum. This name is particularly fitting because both of the other southern
African Gephroglanis species are named after Dr Barnard’s predecessors at the South African
Museum and furthermore Dr Barnard himself described G. gilli with which G. barnardi has
been confused in the past. Finally it is clear from his writings that Dr Barnard had a particular
interest in the fish fauna of the Clanwilliam Olifants River.

OSTEOLOGY

The following description of the osteology of G. barnardi is provided in order to form a
base line study to which other Gephyroglanis species can be compared. The osteology of
G. sclateri has been described by Petrick (1976) but that description is not generally available.
Bone names are those in general usage (mainly after Weitzman, 1962) except in the following
cases: Gosline (1975) for suspensorium; branchial and hyoid arches from Nelson (1969); epioc-
cipital follows Paterson (1975); supracleithrum and posttemporal after Lundberg (1975);
Weberian vertebrae follow the homology suggested by Chardon (1968). The term ethmoid is
used here for the bone referred to as the mesethmoid or supraethmoid of various authors.
Patterson (1975) has shown this bone to be derived from a supraethmoid and rostrodermeth-
moid.

Neurocranium (Figs 8a, b, c, d)

The ethmoid is a T-shaped bone, sutured postero-ventrally with the lateral ethmoids and
the vomer and interdigitating postero-dorsally with the frontal and lateral ethmoids. The
ethmoid is visible on the ventral side only as a narrow neck linking the broad premaxillae with
the neurocranium.

The premaxillae are broad rectangular plates firmly attached to the ventral side of the
T-bar of the ethmoid. Their ventral surface is almost entirely covered with slender, curved,
sharp conical (macroscopically villiform) teeth.

227

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 15, AUGUST 1981

©

Fig. 8. Neurocranium of Gephyroglanis barnardi sp. no
left nasal bone removed; B. ventral viev?

228

SKELTON: DESCRIPTION AND OSTEOLOGY OF GEPHYROGLANIS

SUPRAOCCIPITAL

d/P 893). Scale bar = 2 mm. A. dorsal view,
left lateral view; D. posterior view.

229

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 15, AUGUST 1981

The lateral ethmoids are prominent bones forming broad synchondral sutures posteriorly
with the orbitosphenoids, dorsally with the frontals and anteriorly with the ethmoid. Antero-
medially the lateral ethmoids are partially overlapped by thin lateral lamellae of the vomer.
Their anterior and antero-ventral margin is encased in cartilage to form an articulation condyle
for each palatine. The vomer is a blade-like bone articulating on the postero-ventral surface of
the ethmoid and the antero-medial surface of each lateral ethmoid.

The canal-bearing frontals are elongated rectangular bones sutured anteriorly with the
lateral ethmoids and supraethmoid and posteriorly with the supraoccipital and sphenotic
bones. Each frontal counteroverlaps the other along a sinuous medial suture, except where the
bones are separated to form elongated anterior and posterior fontanelles.

The sphenotic, pterotic, supratemporal and supraoccipital bones form the posterior roofing
bones of the neurocranium. The sphenotics lie postero-laterally to the frontals, anteriorly to
the pterotics and usually latero-anteriorly to the supraoccipitals. The neurocranium broadens
at the level of the pterotics and these bones form the postero-lateral portion of the braincase.
Each pterotic provides passage for a portion of the posttemporal canal of the cephalic lateral
line system, as well as internally for the horizontal semicircular canal. A shallow hyomandibu-
lar facet is formed jointly by the pterotic and sphenotic bones. The large pan-shaped supraoc-
cipital extends over the postero-medial portion of the neurocranium. Anteriorly a deep medial
notch in the supraoccipital forms the posterior part of the posterior fontanel. Posteriorly a
slender medial process extends towards but does not reach the first nuchal plate of the dorsal
fin skeleton.

The parasphenoid forms a narrow strut anteriorly where it is partially overlapped by the
vomer. On the dorsal side it is sutured in turn to the lateral ethmoids, orbitosphenoids and
prootics. Posteriorly the parasphenoid interdigitates with the basioccipital.

Anteriorly the orbitosphenoid forms a dual synchondral suture divided by a groove and
foramen with the lateral ethmoids. Dorsally each orbitosphenoid meets the respective frontal
along a simple suture. Posteriorly there are synchondral sutures with the pterosphenoids. The
orbitosphenoids are closely applied to the parasphenoid without an inter-orbital septum being
formed.

The pterosphenoids meet the frontals on the antero-dorsal side, the sphenotics postero-
dorsally and, over a short length, the parasphenoids ventrally. Anteriorly the pterosphenoids
are synchondrally united to the orbitosphenoids. Posteriorly and postero-ventrally an irregular
junction with the prootics allows for the formation of the large trigeminal foramen. A smaller
facial foramen also provides a break in the suture between each pterosphenoid and prootic.

The large prootics form hemispherical bulges ( bulla acustica lagenaris) on the ventral sur-
face of the neurocranium. Anteriorly the prootics suture with the pterosphenoids, dorsally
with the sphenotics and the pterotics, ventrally with the parasphenoid, postero-ventrally with
the basioccipital and posteriorly with the exoccipitals.

The basioccipital forms an interdigitating suture anteriorly with the parasphenoid, and is
synchondrally sutured to either prootic antero-laterally and the exoccipitals laterally. The ven-
tral limb of the posttemporal is attached to the postero-lateral sides of the basioccipital by
means of a ligament. The basioccipital houses the major portion of each sagitta and, in part,
each asteriscus.

The exoccipitals suture with the basioccipitals laterally and dorsally. They include both the
glossopharyngeal (IX) and vagus (X) foramina. The foramen magnum is bounded by the ex-
occipitals (latero-dorsally) with a small portion provided by the basioccipitals on the ventral
side. (Lamellae of the exoccipitals meet in the midline to form the posterior floor of the brain
case.)

The epioccipitals are sutured to the exoccipitals, supraoccipitals, and pterotics and are over-
lain by the supratemporals and, in part, the posttemporals. The dorsal section of the posterior

230

SKELTON: DESCRIPTION AND OSTEOLOGY OF GEPHYROGLANIS

Fig. 9. A. Diagram of dorsal view of head of Gephyroglanis barnardi sp. nov. to illustrate in situ position of nasals,
infraorbital bones, lachrymal and the course of the cephalic lateral line system. Note preoperculomandibular lateral line
canal extends to dentary (cf. Fig. 12). B. Infraorbital bones and lachrymal of Gephyroglanis barnardi.

231

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 15, AUGUST 1981

vertical semi-circular canal passes obliquely through each epioccipital providing passage from
the exoccipitals, which continues dorsally into the supraoccipital.

The posttemporals are small plates overlying the epioccipitals and articulating with the
extremity of the dorsal limb of the supracleithrum. A short canal through the posttemporal
links the supracleithral portion with the pterotic portion of the cephalic lateral line canal.

The supracleithra are large V-shaped bones each with a dorsal limb articulating in a
groove between the epioccipital and the supratemporal. A channel through this limb provides
passage for the cephalic lateralis lateral line canal. The ventral limb is attached from its medial
extremity to the side of the basioccipital. At the lateral base of the ventral limb a large round
groove or cavity, closed posteriorly by tough connective tissue, accommodates the dorsal ex-
tremity of the cleithrum.

Infraorbitals and nasal bones (Figs 8a, 9a and b)

The nasal bones are elongate tubular units lying lateral to the stem of the ethmoid. They
have a short branch leading to an opening about midway along their length and their rostral
end turns laterad. The infraorbital series are tubular bones comprising a triradiate lachrymal, a
short first infraorbital and three longer, gently curved second, third and fourth infraorbitals.
The lateral line canal branches to a pore at each interconnection.

Suspensorium , palatine-maxilla and preopercular (Figs 10a and b, 11, 12).

The hyomandibular, preopercular, quadrate and metapterygoid are welded together to
form a large plate-like unit. The hyomandibular articulates in a facet formed by the sphenotic
and pterotic and also has a posterior spherical facet with which the opercle articulates. Anter-
iorly it sutures by means of an interdigitation with the metapterygoid. Ventrally it is sutured
with the preopercle. There is an arcuate symplectic-cartilage between the hyomandibular and
the quadrate on the antero-ventral edge of the former bone. A canal through the hyomandibu-
lar provides passage for the facial nerve which passes medio-dorsally — latero-ventrally and
emerges laterally from a notch on the preopercular. There is a narrow lateral ledge on the
hyomandibular providing attachment for the levator arcus palatini muscle.

The preopercle is firmly sutured dorsally to the hyomandibular and the quadrate. Antero-
ventrally a ledge of the quadrate lies lateral to the preopercle, on the posterior side a flange of
the preopercle lies lateral to the quadrate. The flange also covers portions of the symplectic
cartilage and the hyomandibular. The lateral line canal passes through the length of the
preopercle with two intermediate branches to pores terminating on the ventral edge of the
bone.

The quadrate is stout and squarish, forming an arcuate synchondral junction with the
hyomandibular dorsally, an interdigitating suture with the metapterygoid anteriorly and is su-
tured posteriorly with the preopercle.

The metapterygoid is the smallest bone in the unit and forms interdigitating sutures ven-
trally and posteriorly with the quadrate and hyomandibular respectively. The symplectic cartil-
age joins the bone postero-ventrally.

The mesopterygoid, ectopterygoid and palatine (Fig. 11) are loosely connected to the meta-
pterygoid by means of ligaments. The mesopterygoid is flat and square-shaped with a promin-
ent projection from the postero-medial corner. This bone is closely sutured with the smaller
flat ectopterygoid to form a single unit which is closely attached (ligamentously) to the pos-
terior portion of the palatine. The palatine is rod-like, the anterior portion thick and stout and
the posterior section narrow and cylindrical. Midway along its length on the dorsal side there is
a prominent saddle-like articulation facet for the lateral ethmoid. A cartilage pad extends from
each extremity, with the large anterior pad articulating directly with the maxilla.

232

SKELTON: DESCRIPTION AND OSTEOLOGY OF GEPHYROGLANIS

(1

Fig. 10. A. Lateral view of left suspensorium of Gephyroglanis barnardi sp. nov. B. Medial view of interopercle. Scale

bar = 2 mm.

Fig. 11. Medial view of mesopterygoid, ectopterygoid and palatine of Gephyroglanis
barnardi sp. nov. Scale bar = 1 mm.

233

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 15, AUGUST 1981

ANCLIDD- ARTICULAR

ANGUID- ARTICULAR

DENTARY

Fig. 13. Left lower jaw of Gephyroglanis barnardi sp. nov. A. medial view; B. lateral view. Scale bar = 1 mm.

234

SKELTON: DESCRIPTION AND OSTEOLOGY OF GEPHYROGLANIS

The maxilla (Fig. 12) is a stout compressed rod articulating proximally with the palatine
cartilage. A constriction precedes the dual ridged articulation head. A cartilage rod supporting
the maxillary barbel extends from a groove on the antero-ventral side.

Opercular series (Fig. 10a)

As in all catfish there are no subopercles. The opercle is a slender triangular bone arti-
culating, at the apex, with the hyomandibular by means of a ball and socket articulation. The
paddle-shaped interopercle is hollow medially with a transverse division forming a facet for the
abutment of the postero-lateral edge of the posterohyal. In front of the facet, an elongated
groove accommodates the large ligament joining the posterohyal to the angulo-articular.

Lower jaw (Fig. 13)

This consists of three bones, the angulo-articular , the dentary and the coronomecklian as
well as Meckel’s cartilage. The triangular dentary is gently curved and bears slender, curved
teeth in a broad band from two to seven or eight teeth wide. The mandibular lateral line canal
passes along the ventral side of the dentary and has four intermediate pores. A hollow on the
postero-medial aspect of the dentary accommodates the angulo-articular and Meckel’s cartil-
age.

Posteriorly the angulo-articular is stout and large supporting a large saddle-like articula-
tion facet. Anteriorly the bone is compressed and projects dorsally to form the posterior sec-
tion of the ascending process of the jaw. A rounded projection (boss) on the postero-lateral
edge of the dorsal limb provides an attachment point for the adductor mandibulae muscle.
There is a large medio-dorsal excavation in front of the articulation facet. The lateral line canal
passes obliquely through the postero-lateral part of the angulo-articular with a single in-
termediate pore. Meckel’s cartilage extends anteriorly to meet the dentary and also forms a
dorsal branch which lies in a medial groove on the dorsal limb of the angulo-articular.

The coronomeckelian lies medial to the dentary and angulo-articular and wraps around
the dorsal limb of Meckel’s cartilage.

Hyoid arch (Figs 14 and 15)

A small irregularly spherical interhyal is present above the triangular posterohyal.

The postero-lateral extremities of the posterohyal forms a rounded condyle which articu-
lates in ball and socket fashion within the facet on the interopercle (similar to that described
for G. gilli and G. sclateri by Petrick, 1973). A large ligament attaches to the anterior surface
of this condyle and joins the posterior edge of the angulo-articular.

Posteriorly the anterohyal is deep and laterally compressed and sutures to the posterohyal.
Its anterior end is triangular in transverse section where it is sutured with the relatively large
ventrohyal and the smaller dorsohyal.

The ventrohyal has a stout triangular lateral portion which sutures to the anterohyal.
Medially the ventrohyal is depressed and forms a condyle which articulates both the urohyal
and also its opposite fellow. Posteriorly the ventrohyal sutures with the dorsohyal. The smaller
dorsohyal is a depressed rectangular bone joined laterally with the anterohyal and anteriorly
with the ventrohyal (absent from the illustrated individual).

There are usually six branchiostegal rays on each arch, five articulating with the anterohyal
and one with the posterohyal. The first three or four curved acinaciform rays increase progres-
sively in size posteriorly. Each has a Y-shaped head which articulates on the ventral edge of
the anterohyal. The fifth and sixth branchoistegals are spathiform, the fifth articulates on the
postero-ventral angle of the anterohyal and the sixth on the antero-ventral angle of the post-
erohyal.

235

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 15, AUGUST 1981

INTERHYAL

Fig. 14. Medial view of left hyoid arch of Gephyroglanis barnardi sp. nov. (dorsohyal
absent). Scale bar = 2 mm.

Fig. 15. Lateral view (above) and dorsal
view (below) of the urohyal of Gephyrogla-
nis barnardi sp. nov. Scale bar = 1 mm.

236

SKELTON: DESCRIPTION AND OSTEOLOGY OF GEPHYROGLANIS

The urohyal (Fig. 15) is a short triangular bone truncated posteriorly with a median dorsal
flange. The flange divides anteriorly on the dorsal side to form an articulation bracket for the
leading basihyal cartilage. The bracket is flanked anteriorly on each side by an attachment
surface for a ligament connecting to the ventral side of either ventrohyal. A median canal
passes ventrad behind the attachment surface.

Branchial Skeleton (Figs 16a, b, c)

As the interpretation and homology of the siluroid branchial skeleton does not appear to
be satisfactorily determined, nomenclature and interpretation of Nelson (1969) will be fol-
lowed.

There are three infrapharyngobranchials on each side. The first is a small slender slivver
of bone, sometimes absent, lying anterior to the first epibranchial. The second is entirely carti-
lagenous and lies at the head of the first epibranchial between the latter and the third in-
frapharyngobranchial. The third infrapharyngobranchial is elongate and club-like with the
broader end posteriorly abutting the fourth upper toothplate.

There are two upper toothplates — here termed the fourth and fifth after Nelson (1969) —
the fourth is a short club-like bone, edentulous, suspending on its ventral side the large oval
fifth upper toothplate. The ventral side of the fifth toothplate is entirely toothed with long
slender caniniform teeth (Fig. 16c).

There are five epibranchials. The first epibranchial is the largest and has a rod-like medial
section extending laterally into a longer dorsoventrally compressed portion. There are three or
four long gill rakers along the outer anterior edge. There is a short cartilage pad at each end of
the bone. The second epibranchial is similar though slightly smaller than the first. The medial
extremities of these two epibranchials lie close together and abut the cartilagenous second
infrapharyngobranchial. The third epibranchial is a shorter more slender bone with a promin-
ent dorsal process on the mid-posterior side. The epibranchial joins betwen the third in-
frapharyngobranchial and the fourth upper toothplate. The fourth epibranchial is shorter than
the preceding three but is less rod-like and forms a low triangular posterior flange. The fifth
epibranchial is represented by a small but distinct cartilage rod extending off the posterior
edge of the fourth ceratobranchial.

There are five ceratobranchials the anterior three of which are relatively simple arched
struts, grooved to form a trough on the ventral side and having short cartilagenous extremities.
Each successive bone is progressively shorter than the previous one. The fourth ceratobranch-
ial differs in that the medial extremity turns anterad and forms an enlarged cartilage pad. The
fifth ceratobranchials are expanded posteriorly to form a dentigerous surface. Their extremities
consist of short cartilage pads. The teeth on the fifth ceratobranchials are slender, curved and
caniniform.

There are three pairs of hypobranchials of which the anterior two are ossified and the third
entirely cartilagenous. The bony hypobranchials are dorso-ventrally compressed, broader later-
ally and more slender medially. The extremities are in all cases cartilagenous. The third hypo-
branchials are stout cartilage bars against which on the posterior edge the third and fourth
ceratobranchials as well as the cartilagenous posterior basibranchial copula articulate.

The basibranchials are represented by an anterior series of closely connected cartilage and
bony units (representing basibranchials 1-3) and an independent cartilage copula posteriorly.
The anterior series consists of a cartilage pad backed with a bony section, a second cartilage pad
and bone which passes ventral to the third hypobranchial cartilages and itself ends in the form of
a small cartilage. The posterior hypobranchial forms a large rectangular cartilage pad articulat-
ing anteriorly with the third hypobranchials, posteriorly with the fifth ceratobranchials and
laterally with the medial cartilages of the fourth ceratobranchials.

237

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 15, AUGUST 1981

Fig. 16. A. Dorsal view of the branchial skeleton of Gephyroglanis barnardi sp. nov. with right
epibranchials, infra-pharyngobranchials and upper tooth plates removed. Left gill rakers only de-
picted on 1st cerato-branchial. Scale bar = 2 mm.

B. Dorsal view of right epibranchials, infrapharyngobranchials and upper tooth plates. Scale

bar = 1 mm.

C. Ventral view of right upper tooth plate, same scale as B.

238

SKELTON: DESCRIPTION AND OSTEOLOGY OF GEPHYROGLANIS

There are six to nine simple gill rakers along the anterior edges of all ceratobranchials and
three to four similar rakers along the anterior edges of the epibranchials (except the fifth). There
are no gill rakers along the posterior edges of the first, second and fifth ceratobranchials or the
first, and fifth epibranchials. A single small gill raker is irregularly present on the posterior edge
of the second epibranchial.

Pectoral girdle (Figs 17a, b, c)

The pectoral girdle on each side consists of the supracleithrum, cleithrum, mesocoracoid,
coracoid and radials. The supracleithum has already been described (see above). The cleithrum
is the largest single element in the girdle and unites medially in a symphysis with its opposite
member. The dorsal limb of the cleithrum terminates in a stout peg which articulates within the
socket of the supracleithrum. There is a short and relatively broad posterior humeral process at
the base of the dorsal limb. The ventral limb is directed medially and forms a broad flat plate.
The cleithrum folds over latero-ventrally to form a thick lip which partly encloses a deep ventral
groove in which the base of the pectoral spine articulates.

The coracoid is a large complex bone sutured postero-medially to the cleithrum and articu-
lating by means of a well developed interdigitating junction with its opposite member. On the
ventral side of the girdle the coracoid forms a bridge over the cleithral groove and sutures on the
thick overfolded lip. An articulation condyle for the pectoral spine is formed at the posterior
base of the bridge. On the antero-ventral side of the coracoid forms a narrow ridge or flange
which decreases in altitude to disappear medially.

The mesocoracoid is represented by a thin strut which runs from the posterior end of the
ventral coracoid ridge to the posterior edge of the coracoid. A notch on the posterior side of the
base of the strut forms an articulation facet for the median radial. A foramen on the coracoid
above the base of the mesocoracoid allows for the movement of the head of the pectoral spine.

There are three proximal radials and two smaller distal ones. The first proximal radial is a
large trapezoidal cartilage located posterior to the base of the pectoral spine. This assists in the
articulation of the pectoral spine as well as of the following two rays. The second and third
proximal radials are rod-like bones connecting the posterior four rays with the coracoid. The two
small secondary or distal cartilage radials occur between the first and second proximal radials
and the fin rays.

Weberian apparatus (Figs 18a, b, c, d)

The complexities of the siluroid Weberian apparatus have been described by several au-
thors, most notably by Chardon (1968). The fusions and modifications are such that the indi-
vidual components are not readily recognizable and the present description aims solely to
provide a comparative base for G. barnardi.

The first vertebra is represented by the centrum only, and this is compressed to a thin
plate. The second to fourth vertebrae are indistinguishably fused and joined by an elaborately
interdigitating suture to the fifth vertebra to form a single complex. A complex shield is de-
veloped above the centra of these vertebrae and has on each side a large bony strut (the
anterior branch of the parapophyses of the fourth vertebra) which extends antero-laterally and
is firmly bound to the ventral limb of the posttemporal. The posterior branch of the fourth
parapophyses form slender projections from the shield. On the dorsal side a strong V-shaped
strut (fourth neural spine) emerges, forms the neural arch and sutures with the third proximal
pterygiophore of the dorsal fin. On the ventral side of the complex there is a well developed
aortal groove which extends posteriad over the fifth and subsequent precaudal vertebrae.

The parapophyses of the fifth vertebra form prominent projections dorso-laterally. The
enlarged proximal pterygiophore of the second dorsal spine lodges within a pit on the dorsal
side of the fifth vertebra. The sixth vertebra is the first to carry pleural ribs. Like the fifth

239

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 15, AUGUST 1981

Fig. 17. A. Ventral view and B. posterior view of left pectoral girdle of Gephyroglanis
barnardi sp. nov. C. Lateral profile of cleithrum of Gephyroglanis barnardi sp. nov. Scale

bar = 2 mm.

240

SKELTON: DESCRIPTION AND OSTEOLOGY OF GEPHYROGLANIS

DORSAL SPINE 2

DORSAL SPINE

NUCHAL PLATE 3.

NUCHAL PLATE 2

NUCHAL PLATE 1

DORSAL SPINE 2V
DISTAL RADIAL

DORSAL SPINE lx
NUCHAL PLATE 3 ,
NUCHAL PLATE 2 •

NUCHAL PLATE 1

NEURAL SPINE 4 -
PARAPOPHYSIS 4

PROXIMAL RADIAL 3

CLAUSTRUM

Fig. 18. A. Lateral view and B. dorsal view of Weberian vertebrae and dorsal fin skeleton of Gephyroglanis barnardi sp.
nov. Scale bar = 2 mm. C. Dorsal view of right tripus and intercalarium and D. lateral view of right scaphium and
claustrum of Gephyroglanis barnardi. sp. nov. C & D Scale bar = 1 mm.

241

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 15, AUGUST 1981

vertebra the parapophyses of the sixth are well developed as dorso-lateral projections. The
pleural ribs are connected to the antero-ventral edge of the projections. The neural arches of
this vertebra are short and directed posteriad, uniting behind the fifth proximal pterygiophore.
The fifth pterygiophore is thus held within a receptacle bounded anteriorly by the neural
arches of the fifth vertebra and laterally and posteriorly by the neural arches of the sixth
vertebra.

The tripus is attached -to a bony flange beneath the dorsal shield of the anterior vertebral
complex. The tripus itself is elongate, and thin with a flat postero-medial projection and a
tapering lunate transformator process. Anteriorly the tripus is linked to the slender rod-like
intercalarium.

The scaphium lodged within a notch in the exoccipital is in turn connected by means of a
short ligament to the intercalarium. It is an L-shaped bone with the horizontal limb broader
and larger than the vertical limb. The simple leaf-like claustrum lies within the angle of the
scaphium.

Precaudal vertebrae

The neural arches of the seventh and eighth vertebrae form a broad groove which
accommodates the dorsal fin pterygiophores. The neural spines of the ninth and subsequent
vertebrae become progressively longer and reach a maximum at about the fifteenth or six-
teenth vertebra. The parapophyses of the precaudal vertebrae likewise progressively decrease
in size. The development of a closed haemal arch at about the twelfth vertebra precedes the
disappearance of distinct parapophyses and the formation of a haemal spine at about the
fifteenth vertebra. In consequence there are several vertebrae with a closed haemal arch, para-
pophyses and attached pleural ribs.

Caudal vertebrae (see Fig. 19)

The caudal vertebrae are, apart from the first few with non-united parapophyses and short
haemal spines, reasonably uniform in character. They all have well developed neural and
haemal spines, pre- and post-neurapophyses and the majority have pre- and post-
haemapophyses. Three or four preural vertebrae lack post-haemapophyses and their pre- and
post-neurapophyses merge with the neural arch.

Caudal skeleton (Fig. 19)

The caudal skeleton is characterized by the near complete fusion of the hypurals into two
hypural plates. The compound ural centrum (PU, + L^) is fused with the parhypural and
hypurals 1 and 2 as well as the uroneurals 1 and 2 to form a single unit. Weak suture lines
between the hypurals 1 and 2 and the parhypural are still evident. Hypurals 3 to 6 are also
fused into a single plate although here only a single suture line is still evident. A well de-
veloped hypurapophysis on the base of the parhypural is joined to a secondary hypurapophysis
on the base of hypural 1 (i.e. type A hypurapophysis of Lundberg and Baskin, 1969).

A small cartilagenous third uroneural is present between the first dorsal principal caudal
ray and the last dorsal procurrent ray. There is a single epural. Procurrent rays vary between
15 and 18 (Table 4) with three of the six cleared specimens having 16 dorsal and 17 ventral
procurrent rays.

Dorsal fin skeleton (Figs 18a and b, 20)

The first nuchal plate is small and triangular and is fused to a median plate or flange
derived, according to Alexander (1966), from a supraneural bone. This compound unit is su-
tured posteriorly to a second slightly larger but essentially similar nuchal plate. This in turn is
fused to a third larger Y-shaped nuchal plate and the proximal radial of the first pterygiophore.
This radial has bilateral transverse flanges which suture with the dorsal V-shaped projection of

242

SKELTON: DESCRIPTION AND OSTEOLOGY OF GEPHYROGLANIS

dcrsal procurrent rays

Fig. 19. Lateral view of caudal fin skeleton of Gephyroglanis barnardi sp. nov. Scale bar = 2 mm.

the fourth neural spine. The postero-dorsal portion of the first radial is thickened and knob-
like and is straddled anteriorly by the first spine. A large proximal radial of the second ptery-
giophore lodges within a pit on the dorsal side of the fifth vertebra and sutures anteriorly with
the first proximal radial. The dorsal surface is roughened and expanded to form a friction pad
for the base of the second dorsal fin spine. A median dorsal projection forms a ring of bone
which passes through a foramen on the large (second) dorsal spine (Fig. 20).

The proximal radial of the third pterygiophore is reduced in size relative to the first and
second and forms a narrow strut supported basally by the sixth vertebra. Its cartilagenous
dorsal extremity articulates with the cartilage distal radial of the second branched ray. There
are four subsequent pterygiophores; each is progressively reduced in size and shares the sup-
port of two branched rays. The proximal radial of the last pterygiophore is biramous and
supports the last branched ray. A cartilage distal radial is present between the divided base of
each branched ray and the cartilagenous head of each proximal radial.

Pelvic fin skeleton (Fig. 21)

Each pelvic bone is forked antero-laterally and has a broad medial plate which meets and
forms a cartilagenous symphysis with the opposite unit. A slender cartilagenous ischial process
is present. The fin rays articulate against a cartilage-capped posterior ridge. A short lateral
splint, similar to that described by Gosline (1961) for many lower teleostean fishes, lies adja-
cent to the first ray.

Anal fin skeleton

Apart from the first, the proximal radials of the anal fin are all simple and slender. The
first proximal radial is a short heart-shaped plate providing an attachment surface for the bi-

243

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 15, AUGUST 1981

Fig. 20. Antero-dorso-lateral view of portion of dorsal fin skeleton of Gephyroglanis barnardi sp.
nov. A ligament between dorsal spine 1 and 2 has been cut and the spines are separated more than

is natural. Scale bar = 1 mm.

Fig 21. Dorsal view of left pelvic fin
skeleton of Gephyroglanis barnardi
sp. nov. Scale bar = 1 mm.

244

SKELTON: DESCRIPTION AND OSTEOLOGY OF GEPHYROGLANIS

@

(D

Fig. 22. Ventral view of head of A. Gephyroglanis gilli (AM/P 7646, 61 mm SL) and B. Gephyroglanis sclateri (AM/P

7840, 62 mm SL). Scale bar = 5 mm.

lateral blocks of the infracarinalis medius muscle running from the pelvic bones. The articula-
tion end of each proximal pterygiophore terminates in a pad of cartilage. These pads articulate
with cartilage spheres (the distal radials) located between the base of each fin ray. The unbran-
ched fin rays are slender and flexible; only the small anterior-most two are unsegmented and
none can be described as spinous. On account of the minute size of the first ray, accurate
counts are not possible without clearing and staining. The branched rays are quadrifurcate
except the anterior and posterior rays which are trifurcate.

DISCUSSION

There is little doubt on the morphological evidence presented above that there are two
distinct Gephyroglanis species in the Olifants River system. A surprising outcome of the inves-
tigation is that Barnard, whilst fully aware of the two forms within his Olifants samples, did
not perceive the real extent of the differences between them. Barnard’s entire attention fo-
cused on the form of the fins, structures which are well known to be phenotypically influenced
by the environment (Hubbs, 1941).

In addition to the array of anatomical characters on which the existence of the new spe-
cies, G. barnardi , is recognized, the suggestion by Barnard (1943) that the two forms have
different habitat preferences was tested in the field. Barnard (1943) indicated that the low fin
form ( = G. barnardi ) preferred a sluggish muddy environment and the high fin form (G. gilli)
preferred a stony swiftly flowing habitat. On a visit to the Noordhoeks River in January 1981
the author and Mr S. Thorne electrofished both species from the same stony bottomed, swiftly
flowing stretch of river. Both species were taken from a site in the Thee River which, whilst
rocky bottomed, had been obstructed by means of a rocky wall so that the current was sluggish
and the rocks were covered by a layer of silt. In a third tributary, the Oudste River, at a site
where the current was sluggish and the bottom was a mosaic of bedrock, scattered loose rocks,
sand and mud and where the banks were undercut in a fine grained soil only G. gilli was

245

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 15, AUGUST 1981

encountered. Similarly in the Boskloof River, a tributary of the Boontjies River, where the
rocky bottom was covered by silt, only G. gilli was collected. In both the Boskloof and the
Oudste localities G. gilli was taken both from amongst the rocks and from under the muddy
banks.

It appears, therefore, that the correlation of fin form to habitat reported by Barnard
(1943) is not upheld. In this regard it is interesting that seven specimens taken from an irriga-
tion channel off the Jan Diessels River (SAM 22467) and reported by Barnard (1943) as being
of the low fin form are in fact specimens of the high fin form recognized as G. gilli. The only
low fin forms (G. barnardi) within the type series of G. gilli are seven specimens in SAM 19359
collected in February 1939 from the Noordhoeks River and the Olifants River at Keerom.
These facts contradict Barnard (1943: 229) and it appears that either the samples were mixed
up after collecting or that Barnard was confused in correlating specimens and localities. In the
light of more recent collecting and observations as reported in the previous paragraph the
latter explanation is the more likely. Furthermore, the author collected only G. gilli in the Jan
Diessels River in January 1973.

A lectotype for G. gilli (SAM 29231) has been selected from SAM 22467 which is part of
the original syntype series (of this species). The remaining specimens of the sample become
paralectotypes as does the remainder of SAM 19359 after the removal of seven G. barnardi
specimens. The seven G. barnardi specimens (SAM 29232) have been included as paratypes of
the species. All the material of G. gilli and G. sclateri examined is recorded in Appendix 1.

The three southern African Gephyroglanis species (G. barnardi, G. gilli and G. sclateri)
are geographically remotely isolated from other species of this genus. The key character on
which Boulenger (1901) originally placed G. sclateri within Gephyroglanis is the absence of
teeth on the palate (vomer). As pointed out by Roberts (1975) a loss character such as this is
not necessarily synamorph and could easily (and possibly more likely) be a convergent or
parallel development in such geographically remote species. Petrie (1973; 1976) suggested that
the manner in which the posterohyal articulated with the interopercle and the concomitant loss
of the interhyal in G. gilli and G. sclateri was an unusual feature of these South African
species. If this is so (the phenomenon does not appear to have been reported for other bagrid
species) it would strongly suggest a synapomorphy and indicate at least that the southern spe-
cies are monophyletic. Personal observations in the osteology of G. gilli and G. sclateri in the
course of the present investigation have revealed that the interhyal is frequently present in
both species, albeit in a vestigial form. The synapomorphy per se is not therefore negated
which suggests that the generic identity of the South African Gephyroglanis species is open to
question. The type series of this genus is G. congicus Boulenger, 1899, which inhabits the
Zaire River.

In recording the above it should also be noted that the South African Gephyroglanis do
not agree with all of Jayaram’s (1966) points of definition of this genus. Firstly that author
describes the snout as acuminate (narrow and pointed), however, the snouts of G. gilli and
G. barnardi are rather obtusely blunt and rounded. Furthermore, although the snout of
G. sclateri , a species examined by Jayaram, is more slender than that of G. gilli and G. barnar-
di it cannot be described as acuminate. Secondly the lips are recorded as thin and plain,
however, they are papillose in G. barnardi and G. gilli and rugose in G. sclateri. The number
of branchiostegal rays is smaller in the South African species than recorded for the genus by
Jayaram ( viz eight to ten); G. sclateri and G. gilli usually have seven on either side and
G. barnardi six. Further research on the generic status of the South African species is in progress,
and no change is made at this stage pending the outcome of the investigation.

The absence of gill rakers on the posterior border of the first and second gill arches is a
difficult character to evaluate without more in- and out- group comparisons. The only refer-
ence to this character which has been traced in the literature is Harry (1935). In describing the

246

SKELTON: DESCRIPTION AND OSTEOLOGY OF GEPHYROGLANIS

osteology of Amp hilius longirostris (Boulenger, 1901) Harry (op. cit 216) recorded that there
were gill rakers on the first four gill arches (ceratobranchials) and “In addition the third and
fourth ceratobranchials bear a single row of eight stout spinous rakers on the inner face”.
Petrick (1976) did not record the gill-rakers of G. sclateri in his account of the osteology of the
species. In both G. sclateri and in G. gilli the situation with regard to the absence of rakers on
the posterior edge of the first two arches is, however, the same as that described above for
G. barnardi. In addition to some amphiliids [Harry, 1953; pers. obs. on Amphilius platychir
(Gunther, 1864), A. natalensis Boulenger, 1917 and A. lampei Pietschmann, 1913], the mocho-
kids Synodontis leopardinus Pellegrin, 1914 and Chiloglanis anoterus Crass, 1960, also do not
have posterior edge rakers on the first two arches (pers. obs.). The situation in other bagrids is
evidently variable as Chrysichthys hildae Bell-Cross, 1973 and Auchenoglanis ngamensis
Boulenger, 1911 nave stout rakers on the posterior edge of the first four gill arches. Gephyro-
glanis longipinnis Boulenger, 1899 (MRAC 123287-288) has rows of knob-like rakers on the
posterior edges of all the gill arches. Gephyroglanis congicus (MRAC 96639) has an equally
elaborate series of branched rakers on the posterior edges of the gill arches. A complex dis-
tribution of the character is foreseen within the siluriformes which is likely to be closely corre-
lated with functional demands.

A few preliminary comments on the inter-relationships of the three South African species
can be made at this stage. Generally G. barnardi differs more from both G. gilli and G. sclateri
than these two species do from each other, which is not in itself any indication of the phy-
logenetic relationships between them. Many of the distinguishing features of G. barnardi are
apparently autapomorph: for example the relatively weak dorsal and pectoral fin spines; the
small humeral process; the reduction in size of the nuchal shield and its connection with the
supraoccipital.

The caudal fin skeleton possibly provides one synapomorphy for G. barnardi and G. gilli
in that these two species have a greater degree of hypural consolidation than does G. sclateri.
In G. sclateri the hypurals are generally free except for hypurals three and four which are fused
basally. In G. barnardi and G. gilli hypurals one and two are fused or separate only in juve-
niles. In G. barnardi hypurals three, four, five and six are also fused. The degree of fusion
shown by G. barnardi is greater than that shown by any bagrid species reported on by Lund-
berg and Baskin (1969). A few bagrids including Gephyroglanis longipinnis were found by
these authors to have hypural three and four fused but only one species (Chrysichthys ornatus
Boulenger, 1902) to have, in addition, hypurals one and two fused. The skeletal differences
relating to the caudal fins of the three South African species are (coincidentally) correlated
with the external forms of these fins: in G. barnardi the fin is truncate, in G. gilli emarginate
and in G. sclateri forked. Lundberg and Baskin (1969) found that the majority of bagrid spe-
cies they examined had B type hypurapophyses (the hypurapophysis and secondary hypur-
apophysis on hypural 1). The South African Gephyroglanis species appear to have an A type
hypurapophysis. All specimens of G. sclateri and G. barnardi examined had an A type and of
the three available specimens of G. gilli two had an A type, however, one had a B type. (A
type: hypurapophysis on parhypural arch, secondary hypurapophysis on base of hypural 1).

A feature of the caudql fin skeleton of several catfish examined to date which appears to
have been overlooked by Lundberg and Baskin (1969) is the presence of a third uroneural.
This element is clearly present in G. barnardi albeit as cartilage. Its cartilaginous nature poss-
ibly explains why it was not found in other catfish by Lundberg and Baskin (op. cit.).

Monod (1968) illustrated the caudal skeleton of a Gephyroglanis species and included a
cartilage “opistural” which is here interpreted as the third uroneural. This element is also
clearly present in cartilage form in the schilbeids, Schilbe mystus and Eutropius depressirostris
(AM/P 3121, 3256 and AM/P 742 respectively). It is difficult to detect in only alizarin red
stained specimens, but has been found in both dry and alizarin preparations of G. sclateri and
G. gilli (pers. obs.).

247

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 15, AUGUST 1981

One character which contradicts the caudal fin synapomorphy of G. bamardi and G. gilli
is the dorsal fin ray count. G. bamardi has usually six branched dorsal fin rays the number
which is most common among other Gephyroglanis species (Boulenger, 1911; Jayaram, 1966)
and other genera of the Chrysichthyinae (Jayaram, 1966). Based on an out-group comparison
this appears to be the plesiomorph condition. G. gilli and G. sclateri both have seven dorsal
branched rays, arguably synapomorph in the South African context. On the other hand it has
been pointed out (Heemstra, pers. comm.) that the configuration of the last dorsal proximal
radial (Fig. 18a) in G. bamardi suggests that the last ray has been lost. If this is so, then the
possession of the seven branched dorsal rays of G. gilli and G. sclateri is the plesiomorph
condition and is not indicative of a closer relationship to each other than either may have to
G. bamardi.

The placement of the mandibular barbels is similar in G. barnard and G. gilli but differs in
G. sclateri (Figs 5;22). In the former two species the posterior mandibular barbels are lateral to
the gular node and postero-lateral to the anterior mandibular barbels. In G. sclateri the pos-
terior mandibular barbels are distinctly anterior to the gular node and lie more or less posterior
to the anterior pair. The vertebral counts of the South African Gephyroglanis species are
compared in Table 3. There is a broad overlap in counts between G. gilli and G. bamardi ,
however, G. sclateri has a larger count. This finding agrees with Lindsey’s (1975) pleomeristic
rule as G. sclateri attains twice the size of the other two species.

Jayaram (1966) separated G. sclateri and G. gilli chiefly on the basis of anal fin rays
(G. gilli 11-13 v. 16-17 in G. sclateri). These figures were based on Boulenger (1901, 1911) and
Barnard (1943) with the latter author dividing the rays into spines, simple rays and branched
rays. Barnard’s (1943) count for G. sclateri was 2-4, 2-3, 10-12, usually 4, 2, 12, with the total
range 16-19. The count for anal fin rays given by Barnard for G. gilli was 2-3, 2, 10-11, most
frequently 3, 2, 10, with a total range of 14-16 not 11-16 as given by Jayaram (1966). A
comparison of anal fin rays of the three southern African species is given in Table 2. The
formula for G. barnardi is III— VI, 10-13 (spines and simple rays not separately distinguished),
most frequently IV, 11 with a total range of 14-18. G. sclateri is therefore generally character-
ized by more anal fin rays than either Clanwilliam Olifants species although this is not neces-
sarily a diagnostic feature of every specimen. The South African Gephyroglanis species have a
greater number of anal fin rays than other species of this genus (Jayaram, 1966), a fact which is
noteworthy in terms of the general distinction already made of these three remote southern
representatives.

The geographical distribution of Gephyroglanis species in the Orange and Olifants River
systems, together with other freshwater fish species in these two systems, offers substantial
support for the theory that the two systems were linked at some stage in the past (Jubb &
Farquharson, 1965; Skelton, 1980). Fryer (1977) also suggested such a link on the evidence of
the distribution of related species of the ectoparasitic crustacean genus Chonopeltis. Two fac-
tors suggest that this link is likely to have been a relatively ancient one, probably pre-
Pleistocene. These are the high degree of endemic speciation of the fish species concerned in
both the Orange and the Olifants, and secondly that there is little direct present day geological
indication of such a link.

ACKNOWLEDGEMENTS

The research was funded by the Department of Nature and Environmental Conservation
and was stimulated by collections received from two officials of that department, Mr K. Ham-
man and Mr S. Thorne. Assistance was given by Mr S. Thorne and Mr A. Welman. The
manuscript was kindly read and criticized by Mr W. Holleman and Dr P. C. Heemstra.

248

SKELTON: DESCRIPTION AND OSTEOLOGY OF GEPHYROGLANIS

REFERENCES

Alexander, R. McN. 1966. Structure and function in the catfish. J. Zool., Loud. 148 (1): 88-152.

Barnard, K. H. 1943. Revision of the indigenous freshwater fishes of the S.W. Cape region. Ann. S. Afr. Mus. 36 (2):
101-262.

Boulenger, G. A. 1901. Description of a new Silurid fish of the genus Gephyroglanis , from South Africa. Ann. S. Afr.
Mus. 2 (8): 227-228.

Boulenger, G. A. 1911. Catalogue of the Fresh-water fishes of Africa in the British Museum ( Natural History ). Voi. 2.
London. British Museum (Natural History).

Chardon, M. 1968. Anatomie comparee de l’appareil de Weber et des structures connexes chez les siluriformes. Annls
Mus. r. Afr. cent. Ser. 8vo., Sci. zool. 169: 1-277 .

Fryer, G. 1977. On some species of Chonopeltis (Crustacea: Branchiura) from the rivers of the extreme South West
Cape region of Africa. J. Zool., Lond. 182: 441-445.

Gaiger, I. G. 1978. The distribution, status and factors affecting the survival of indigenous freshwater fishes in the Cape
Province. Research report. Cape Provincial Department of Nature and Environmental Conservation, pp.
74-128.

Gosline, W. A. 1961. Some osteological features of modern lower teleostean fishes. Smithson, misc. Colins 142 (3):
1-42.

Harry, R. R. 1953. A contribution to the classification of the african catfishes of the family Amphiliidae , with descrip-
tion of collections from Cameroon. Revue Zool. Bot. afr., 47 (1-2): 177-232.

Hubbs, C. L. 1941. The relation of hydrological conditions to speciation in fishes. In: A symposium on Hydrobiology.
Madison. Univ. Wisconsin Press, pp. 182-195.

Hubbs, C. L. and Lagler, K. F. 1958. Fishes of the Great Lakes Region. Bull. Cranbrook Inst. Sci. 26: 213 pp.
Jayaram, K. C. 1966. Contributions to the study of the fishes of the family Bagridae 2. A systematic account of the
African genera with a new classification of the family. Bull. I.F.A.N. Ser. A 28 (3): 1064-1139.

Jubb, R. A. 1967. The freshwater fishes of southern Africa. Cape Town. Balkema.

Jubb, R. A. and Farquharson, F. L. 1965. The freshwater fishes of the Orange River drainage basin. 5. Afr. J. Sci. 65:
118-125.

Lindsey, C. C. 1975. Pleomerism, the widespread tendency among related fish species for vertebral number to be
correlated with maximum body length. J. Fish. Res. Bd Can. 30: 409-434.

Lundberg, J. G. 1975. Homologies of the upper shoulder girdle and temporal region bones in catfishes (Order Siluri-
formes), with comments on the skull of the Hologeneidae. Copeia 1975 (1): 66-74.

Lunderg, J. G. and Baskin, J. H. 1969. The caudal skeleton of the catfishes, order Siluriformes. Am. Mus. Novit.
(2398): 1-49.

Monod, T. L. 1968. Le complexe urophore des poissons teleosteens. Mem. d'lnst. Fond. d'Afr. Noire (81) 1-705.
Nelson, G. J. 1969. Gill arches and the phylogeny of fishes with notes on the classification of vertebrates. Bull. Am.
Mus. nat. Hist. 141: 479-552.

Patterson, C. 1975. The braincase of pholidophorid and leptolepid fishes, with a review of the actinopterygian brain-
case. Phil. Trans. R. Soc. Ser. B. 269: 275-579.

Petrick, F. O. 1973. An anomaly in the hyoid arch of Gephyroglanis sclateri Boulenger and Gephyroglanis gilli Bar-
nard. News. Lett. Limnol. Soc. South Afr. (21): 37-38.

Petrick, F. O. 1976. The anatomy of the freshwater fishes of the Transvaal VIII Gephyroglanis sclateri. Typescript report
to Provincial Fisheries Institute, Lydenburg. As revised (1976). 39 pp. 30 figs.

Roberts, T. C. 1975. Geographical distribution of African freshwater fishes. Zool. J. Linn. Soc. 57: 249-319.

Skelton, P. H. 1976. Preliminary observations on the relationships of Barbus species from Cape coastal rivers, South
Africa (Cypriniformes: Cyprinidae). Zool. Afr. 11 (2): 399-411.

Skelton, P. H. 1977. South African Red Data Book: Fishes. (South African National Scientific Programmes, Report
No. 14.) Pretoria. CSIR.

Skelton, P. H. 1980. Systematics and Biogeography of the redfin Barbus species (Pisces, Cyprinidae ) from southern
Africa. Unpubl. Ph. D. Thesis, Rhodes University, Grahamstown, South Africa.

Taylor, W. R. and Van Dyke, G. C. 1979. Revised procedures for staining and clearing small fishes and other verte-
brates for bone and cartilage study. Manuscript report, National Museum of Natural History, Washington, 21

pp.

Weitzman, S. H. 1962. The osteology of Brycon meeki a generalized characid fish with an osteological definition of the
family. Stanford ichthyol. Bull. 8 (1): 1-77.

Manuscript accepted for publication 2 March 1981.

249

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 13 PT 15, AUGUST 1981

APPENDIX 1

Material of Gephyroglanis gilli and G. sclateri referred to.

Gephyroglanis gilli — Olifants River System.

AM/P 893: K. van Rensburg, March 1965, Noordhoeks and Thee Rivers.

AM/P 1364: F. L. Farquharson, October 1964, Noordhoeks River.

AM/P 1369: F. L. Farquharson, April 1967, Noordhoeks River.

AM/P 1372: 3, F. L. Farquarson, April 1967, Matjies River.

AM/P 1851: 4, P. H. Skelton and D. Heard, January 1973, Jan Diessels River.

AM/P 1861: 6, P. H. Skelton, January 1973, Breekkrans River.

AM/P 1865: 6, P. H. Skelton, February 1973, Noordhoeks River.

AM/P 1879: 20, K. van Rensburg, March 1965, Noordhoeks River.

AM/P 2604: 2, P. H. Skelton and D. Heard, January 1973, Jan Diessels River.

AM/P 7646: 4, K. Hamman, S. Thorne, August 1979, Noordhoeks River.

AM/P 7673: 3, S. Thorne, G. Gabriel, March 1980, Noordhoeks River.

AM/P 7677: 2, S. Thorne, G. Gabriel, March 1980, Thee River.

AM/P 7688: 9, S. Thorne, G. Gabriel, March 1980, Rondegat River.

AM/P 7690: 4, S. Thorne, G. Gabriel, March 1980, Rondegat River.

AM/P 7710: 6, S. Thorne, G. Gabriel, March 1980, Noordhoeks River.

AM/P 8201: 4, P. H. Skelton, S. Thorne, January 1981, Noordhoeks River.

AM/P 8205: 2, P. H. Skelton, S. Thorne, January 1981, Thee River.

AM/P 8208: 7, P. H. Skelton, S. Thorne, January 1981, Oudste River.

AM/P 8211: 3, P. H. Skelton, S. Thorne, January 1981, Boskloof River (trib. Boontjies
River).

SAM 19359: 11, K. H. Barnard, C. W. Thorne, February 1939, Olifants River at Keerom &
Noordhoeks River (Paralectotypes).

SAM 22467: 12, K. H. Barnard, C. W. Thorne, A. C. Harrison, April 1937, Jan Diessels
River & Boontjies River (Paralectotypes).

SAM 22477: 6, K. H. Barnard, April 1949, Thee River.

SAM 29231: 1, Same as SAM 22467 (Lectotype).

Gephyroglanis sclateri — Orange River System

AM/P 1381: Cape Department of Nature & Environmental Conservation, September 1966,
Orange River at Venterstad, Cape Province.

AM/P 1447: Cape Department of Nature & Environmental Conservation, September 1970,
Telle River.

AM/P 3921: P. H. Skelton, I. G. Gaiger, J. Cambray, September 1976, Serfontein Bridge on
Orange River.

AM/P 7840: P. H. Skelton, J. Cambray, September 1980, Blouputs, lower Orange.

250

^ sJ 5 B 0 W 8 » W 1 ^

g ~ J»*’ * ■”" - vs**?fs5

IS^SB iavnan~LIBRARiES SMITHSONIAN “"iNSTITUTION^NOIlfUllSNI __ NVINOSHillNS S3 ! dvy a n

N INSTITUTION NO*iniiISNI__ NVINOSHIIWS S3IMVMfln__ LIBRARIES SMITHSONIAN JNSTITUTION
ot 2 \ *2 — - ■’

O O o “

is2s3iavHan_LIBRAR!ES Smithsonian institution NouniiiSNi nvinoshiisns saiavuan

z f“ Z ®“ 2 8” vV . 2

o m 2 vra 2 m 2

CO

S > m*

2 %

m ^ •W5' m m

^“INSTITUTION^ NOIiniilSNl“NVINOSHimS S3 I H VH 9 ll“u B R AR I ES SMITHSONIAN INSTITUTION

2 cn Z "y "*

< AN'.- 2 .< 2<T*S_PV£

fj i |

usw/S3 1 ava a li b r ar i es'smithsonian- msTiTUTiON^NoiiniiiSNi^NviNosHiiiNs^sa i a vaan

— tn r* tfi

Q O

\N"JINSTITUTI0N*"N0llfUllSNl“*NVIN0SHlllNS:2S3 I HVB 0 11 LI B RAR 1 ES SMITHSONIAN INSTITUTION
r- z _ f“ v 2 2 ^

frj x^Vosv^X ^ x>«'y m )^>v 2 XjVA5>v*y m

MS^Sl iaVHail”L!BRAR! ES SMITHSONIAN^INSTITUTlON^NOlinillSNI NVIN0SH1MS S3 I aVH9 11
in z \ 52 _ . z \ (£ ’’’

\N INSTITUTION NO!inmSNI_NVINOSHllWS . S3 I H VH 3 11_ll B RAR I ES SMITHSONIAN JNSTITUTION

O Z XSy o

^ 2 -1 Z _

ass saiavasn libraries Smithsonian institution NoiiniiiSNi^NViNOSHiiws^ss iavu as

t-

in

AN = INSTITUT!ON05 NOlinillSNl"” NVINOSHillNS S3 I HVH 8 n“L! B RAR I ES SMITHSONIAN “INSTITUTION

(n z C/> Z - U1

2 «r ^IsoSTx S <

_ m ^ r ' p, g m

R I ESWSMIIHSON!AN“lNST!TUIION^NOIiniSISNl”NVINOSHl!WS S3 I HVd a II J-* B RAR I ES SMITHSOf
to z \ to z <£ Z '

E /^SS^v <

-■ /^^Wk0-^ -p __ _ _

X
CO

O

> i *NJJT “•" ' ' I 1' * ' -

I1SNI NVINOSHlIINS^SS I dVdail^LIBRAR I ES^SMITHSQNIAN INSTITUTION NOIlfUllSNI_NVINOSH

5 \ to — w zz ___ CO ^ 2

2^ CO ^ “ xg|jS|x W UJ CO

6 Q “

Z -I 2 J 2

RIES SMITHSONIAN INSTITUTION NOIlfUliSNI NVINQSHlllNS S3 I d VH 0 11 LI B R AR I ES SMITHSO

z T ^ z r~ .. z . 5

L^X £ .• m xSx 9. x^E5"x m 4ft « 2 x^^STX ro

m g x^oTTRS^ m 'Sft. ^ xiom^x m

II1SN|—NVIN0SHIIINS S3 I d Vd 9 I "1 L S B R AR 1 ES^SMITHSONIAN “INSTITUTION NOIlfUliSNI NVINOSH

• ./I S .,,/ dki ! ,4SS^ I Jk. S -v aA. ^
W

> _ „ __ ^ ^

I ES^SMITHSONIAN INSTITUTION NOIlfUliSNI NVINOSHlIIAIS^Sa I dVd 9 11 LIBRARIES SMITHSO
m 5 in ” . to X „ ^ 01

i.i — ><vtso»r^ Lii •*- .<vs . llb n: x<v»so/v3s. yj

o D<-_>X _ VOJLLS^X O

ItlSNl^NVINOSHillAIS^SB I dVd 0 ll^LI B RAR I ES^SMITHSONIAN”lNSTITUT!ON-NOIiniliSN! NVINOSH
r- v z r~ 2 r* z r >

iR I ES SMITHSONIAN INSTITUTION NOIlfUliSNI NVIN0SH1IIAIS S3ldVd9ll LIBRAR I ES SMITHSC

CO _ 2 -- C/5 2 CO 2 N

§ i X * ^ S /«C2S\ 1 I X. , 1 /g®

> 5 . 5 _ _

LIlSNI NVINOSHlIINS^Sa I dVd a ll\l B RAR I ES^SMITHSONIAN INSTITUTION ' NOIlfUliSNI _NVINOSE
to . = CO “ to 2

§

O o z o

z J 2 _J 2

\R I ES SMITHSONIAN INSTITUTION ^NOIlfUliSNI NVIN0SH1IINS S3 I d VH 9 11^ LI B R AR I ES^ SMITHS(

% I ^ , >5 I /S^5s S .^5 % P , , /

03
P3
>

;□ V*T
- Vr

m «; ro 'sr, ^ x^uus^x m

„ w £ co X „ to

1I1SNI NVINOSHlIINS S3ldVdail LIBRARIES SMITHSONIAN INSTITUTION NOIlfUliSNI NVINOSf
2 CO z to 2 ,v.- to 2