_ — o

_J 2: _J 2 -I Z

sni nvinoshiiws saiavaan libraries Smithsonian institution NounuiSNi nvinoshiiws

Z r- z •“ „ ^ z r-

ro Ar&vbiPx it y y co

>

30

m ^ m m w ~

co X £ to — to — in -

ES SMITHSONIAN INSTITUTION NOIlfUliSNI NVINOSH1IWS S3 IdVaa 11 LIBRARIES SMITHSONIAN

Z ,v,.. to Z to Z to

SNI NVINOSH1IWS S3IHVaan LIBRARIES SMITHSONIAN INSTITUTION NOliDillSNI NVINOSHiiWS
— in ^ oo =r to —

o 'X _ x^uuv>^ u uy __ QJ X^li^/ Q

E S'2 SMITHSON I AN”* INSTITUTION NOIlfUliSNI “WlNOSHlIiAIS^SS laVdail^UBRARI ES^ SMITHSONIAN
Z 9 Z 9 » Z r- z

o ^ m xfoo>oX 9 X5?U<oX „ av o — XtovSs o

* —

CO

z

m '^vos^tx r; m - cw ™ m

to £ to £ co _

sni nvinoshiiws sa lava an libraries Smithsonian institution NoiiniusNi nvinoshiiiais

Z to Z „ CO z » to z

< xr55^3x ? ^ S <

g g fei 31 § xJSk: i ° (if

, , =3

m o ^ -

-Co to pj

X "T~ » sT>t

, s WW / -y

-v5(4i3$£y £/

_ ‘ >0 s^y > 2

CO Z to "W Z CO z to

ES SMITHSONIAN INSTITUTION NOIlfUliSNI NVIN0SH1IWS S3IUVH8I1 LIBRARIES SMITHSONIAN

z \ 9 „ — . =; co

SNI NVINOSHIIIAIS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION NOIlfUliSNI NVIN0SH1IWS
r~ w Z r- z r~ Z t-

aW ^ “ .XTCvstWTs* ^ — - A A

m ^ x^Uiisyx m :z m zz x^ocos^x m

to \ — to ± to ± 00

ES SMITHSONIAN INSTITUTION NOimillSNI NVIN0SH1IWS SaiHVaan LIBRARIES SMITHSONIAN

v> Z .v to Z to Z -v • <0

z
o

to

t " '%;x z t: vc7|t

> 2 >
z co *- z co z to *. z

SNI NVINOSHIIIAIS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION NOIlfUliSNI NVINOSHIIIAIS

5 to 5 to 5 to ~

fz .^SsX tu It, ui ySvtsoi^x tu z

to ' eft

O >\ ~ xtfUusg^ o o.t^ _ O p q

i E$2 SMITHSONIAN^ INSTITUTION 2 NOIlfUliSNI ^NVINOSHII INS ^S 3 I BVd 8 II^LI B R AR I ES^SMITHSONIAN

m xjvasvv^- rn ^ n^jvasj^ m

£ to £ to £ co _

SNI NVINOSHIIIAIS SSIHVHan LIBRARIES SMITHSONIAN INSTITUTION NOIlfUliSNI NVIN0SH11WS
z to . z , to z <0 z

m LIBRARIES SMITHSONIAN INSTITUTION NOlinillSN! NVINOSH1IWS S3iava8ll LIBRARIES

g v . vro g w Jjfe °

'A ^ /r>/ >Ar x t /s3ffW^\ a

l)

“ V to ^

<o — <o ~ v> K t: to

ION NOlinillSNI NVINOSH1IWS S3IUVti8l1 LIBRARIES SMITHSONIAN INSTITUTION NOIlfllllSNi
co 2 ^ ^ w 2 ,... jo 2 ,v. oo

2 ,< 5 . . •< 5 xjKSvSTx <

2 > *W"' 2 '\ > X^iius^X 2

2 CO 2 CO *2 w “ 2

3 1 3 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS S3IBVBan LIBRARIES

— (/) ~ co — co

SvJso^X td 00 **• m UJ /

I’^Sr = 1C If 1 f€ Jl i Ibk' i

“ O ® 0 n§s; m xv/)iaS^x 0

ION ^NOlinillSNI^NVINOSHlUNS^SB I U V U S IT ^ L I B R A R I ES^SMITHSONIAN^iNSTITUTION ^ NOlinillSNI

£ X^X 5 ' % g 2 ® 2

i- awaa - X 33 *- <zmzs*. ~

> ISL

33

m r; \*as£>x rn X^os^x r; m

CO — CO £ CO

m LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IINS S3IBVaail LIBRARIES

2 , <30 2 CO 2 CO 2

< ^ 2 xSSwcX < X 2 2 4/Me, . <

2

«*'- O

co

X

35P I I ^ s'"' A 5 xw- 5' • • 2

co 2 CO ■'* 2 CO 2 CO

ON NOlinillSNI NVIN0SH1IIAIS S3iaVaaiT LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI
CO 2 \ 00 5 to — CO

A — VX>-?m a: - /?/0 ‘‘SkA o: Jp; V - (St

< 2 1*1
QL Ay/ </ * C~

CQ *■

CO

3n LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IINS S3 I B VB 3 IT LIBRARIES
~ 2 r- ^ 2 1-2

m ^ m .. xouus^x rn ^ x'Oiii.sgx m

to x co £ — w \ 2 — w

ION NOlinillSNI NVIN0SH1IINS S3IBVaan LIBRARIES SMITHSONIAN INSTITUTION NOlinillSN

to 2 .... CO 2 v CO

5 vi

> s >

3n LIBRARIES^SMITHSONIAN INSTITUTION N0liniliSNrNVlN0SHlllNSW,S3iavaan^LIBRARlES

CO — » CO “ m

tn

in

o XOxpCX " 0 XgA Dtjx “ XO)jiiS^X Q ^ o

ION NOlinillSNI "XlVINOSHillNS SBiavaail^LIBRARI E S 2 SMITHSONIAN^ INSTITUTION 2 NOlinillSNI
~ 2 r* 2 r* 2

lA.\ 3 /iff* ' ■ *=avA » b (*B?"r*L'A ^ ”

m :‘ W ^ m XtxosXJ/ co " |-rj

co — co £ to

3 11 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IIAIS S3IBVBan LIBRARIES

to 2 <0.2,

co

Annals of the

Cape Provincial Museums

Natural History

Ann. Cape Prov. Mus. (nat. Hist.)
Volume 14 Parts 1-8 1981-83

Published jointly by the Cape Provincial Museums
at the Albany Museum, Grahamstown, South Africa

ANNALS OF THE CAPE PROVINCIAL MUSEUMS

These Annals are published in two series. Natural History and Human Sciences, the latter
series covering cultural history, ethnology, anthropology and archaeology. They are
issued in parts at irregular intervals as material becomes available.

The primary objective of these Annals is to disseminate the results of research work
undertaken by staff members of the Cape Provincial Museums. However, a manuscript by
any author is eligible for publication provided it is based wholly or partly on the collection/s
of one or more of these museums or if the material on which the work is based is wholly
or partly deposited in one of these museums.

Parts are obtainable from the Librarian of the Albany Museum. Correspondence in con-
nection with these Annals should be addressed to the Editor, Albany Museum, Grahams-
town 6140.

Editor

DrF.W.GESS: 1978-

LIST OF CONTENTS

1. Some aspects of an ethological study of the aculeate wasps and the bees of a

karroid area in the vicinity of Grahamstown, South Africa. F. W. Gess 1

2. A selected bibliography of literature on Odonata from Africa and adjacent

islands. Part 2. B. C. Wilmot and L. P. Wilmot 81

3. On the taxonomic status of Heleophryne regis Hewitt, 1909 (Anura: Lepto-

dactylidae). Richard Boycott 89

4. Three new species and a new genus of tripterygiid fishes (Blennioidei) from the

Indo- West Pacific Ocean. Wouter Holleman 109

5. Some aspects of the ethology of Chalybion (Hemichalybion) spinolae (Lepe-

letier) (Hymenoptera: Sphecidae: Sphecinae) in the Eastern Cape Province of
South Africa. F. W. Gess, S. K. Gess and A. J. S. Weaving 139

6. Ethological studies of Isodontia simoni (du Buysson), I. pelopoeiformis (Dahl-

bom) and 1. stanleyi (Kohl) (Hymenoptera: Sphecidae: Sphecinae) in the East-
ern Cape Province of South Africa. F. W. Gess and S. K. Gess 151

7. Historical atlas of the diurnal raptors of the Cape Province (Aves: Falconi-

formes). A. F. Boshoff, C. J. Vernon and R. K. Brooke 173

8. On the Hydropsychidae (Trichoptera) of Southern Africa with keys to African

genera of imagos, larvae and pupae and species lists. K. M. F. Scott . . 299

INDEX TO AUTHORS

BOSHOFF, A. F., C. J. VERNON and R. K. BROOKE.

Historical atlas of the diurnal raptors of the Cape Province. (Aves: Falconiformes) (7) Febr. 1983 . 173-297

BOYCOTT. RICHARD.

On the taxonomic status of Heleophryne regis Hewitt, 1909 (Anura: Leptodactylidae). (3) July 1982 . 89-108

BROOKE, R. K. see BOSHOFF, A. F., C. J. VERNON and R. K. BROOKE.

GESS.F. W.

Some aspects of an ethological study of the aculeate wasps and the bees of a karroid area in the
vicinity of Grahamstown, South Africa. (1) Oct. 1981 1-80

GESS, F. W. and S. K. GESS.

Ethological studies of Isodontia simoni (du Buysson), I. pelopoeiformis (Dahlbom) and I. stanleyi
(Kohl) (Hymenoptera: Sphecidae: Sphecinae) in the Eastern Cape Province of South Africa. (6)

Nov. 1982 151-171

GESS, F. W., S. K. GESS and A. J. S. WEAVING.

Some aspects of the ethology of Chalybion (Hemichalybion) spinolae (Lepeletier) (Hymenoptera:
Sphecidae: Sphecinae) in the Eastern Cape Province of South Africa. (5) Oct. 1982 139-149

GESS. S. K. see GESS, F. W. and S. K. GESS.

GESS, S. K. see GESS, F. W., S. K. GESS and A. J. S. WEAVING.

HOLLEMAN, WOUTER.

Three new species and a new genus of tripterygiid fishes (Blennioidei) from the Indo-West Pacific
Ocean. (4) Oct. 1982 109-137

SCOTT, K. M. F.

On the Hydropsychidae (Trichoptera) of Southern Africa with keys to African genera of imagos,
larvae and pupae and species lists. (8) June 1983 299-422

VERNON, C. J. see BOSHOFF, A. F., C. J. VERNON and R. K. BROOKE.

WEAVING, A. J. S. see GESS, F. W.,S. K. GESS and A. J. S. WEAVING.

WILMOT, B. C. and L. P. WILMOT.

A selected bibliography of literature on Odonata from Africa and adjacent islands. Part 2. (2) July

1982 81-87

WILMOT, L. P. see WILMOT, B. C. and L. P. WILMOT.

NEW NAMES PROPOSED IN THIS VOLUME

GENERA

Cremnochorites (Pisces, Tripterygiidae) Holleman, 1982 125

SPECIES

clarkae (Eimeapterygius) Holleman, 1982 121

fuscopinna (Helcogramma) Holleman, 1982 115

ventermaculus (Enneapterygius) Holleman, 1982 123

INSTRUCTIONS TO AUTHORS

LAYOUT of recent issues should be examined for details of acceptable style and format.

MANUSCRIPTS should be submitted in triplicate, typed on A4 size paper, double-
spaced, on one side of the paper only and with wide margins all round.

Only generic and specific names should be underlined to indicate italics. All other mark-
ing up should be left to the editor.

Footnotes should be avoided as far as possible.

All measurements should be metric.

Systematic papers should follow the rules of the International code of zoological nomen-
clature or the International code of botanical nomenclature , as applicable.

ILLUSTRATIONS should bear the author’s name, title of paper and number in soft
pencil on the reverse side. Diagrams, line drawings and photographs are all regarded as
figures, should be termed as such and should be numbered in a single series.

In addition to the original illustrations a photocopy of each should be submitted.

Diagrams and line drawings should be reducible to a size not exceeding 14 x 16 cm
(17,5 cm including caption). Photographs for half-tone blocks should be final size, of high
contrast and printed on glossy paper. Colour illustrations are only printed under special
circumstances and their financing must be arranged by the author.

A scale should appear with all illustrations or the magnification should be given in the
caption.

A list of captions should be typed on a separate sheet.

TABLES should be typed on separate sheets. Foldouts and tables printed sideways should
be avoided.

REFERENCES should be listed alphabetically at the end of the paper. Abbreviations of
periodical titles should conform with those in the World list of scientific periodicals
(4th ed.). Layout must be as follows:

Alexander, J. 1969. The indirect evidence for domestication. In: Ucko, P. J. and
G. W. Dimbleby eds , The domestication and exploitation of plants and animals.
London: Duckworth, pp. 123-129.

De Winter, B. 1969. Plant taxonomy today. S.Afr.J.Sci. 65 (8): 229-242.

Jubb, R. A. 1967. Freshwater fishes of southern Africa. Cape Town: Balkema.

GALLEY PROOFS are sent (when practicable) to authors for checking printer’s errors.
Adjustments of language, style and punctuation and any other amendments of the original
manuscript, if made at this stage, may be charged to authors.

AUTHOR’S COPIES (50) are issued free of charge. When more than one author is
involved the fifty free copies will be sent to the senior author. Additional copies ordered
before printing are supplied at a reduced cost.

BD5276 — CTP Book Printers, Cape

Am

•'»’ nti fa. "r

Cape Prov .

•.v«iroiwTvmvtrofm74Tm7S7m7n,ff47TOwiHT«m««n

y. *.* ».
yfvy;/"

KmOWiU-WI'SiWW ^r~-Y'S“
mtwmwm Jlmt inmi

1 *«* *** «H iftv

PUBLISHED jd|iMY BY THE

CAPE PROVINCIAL MUSEUMS AT' THE ALBANY MUSEUM, GRAHAMSTOWN

. SOUTH' AFRICA . ]MM I fillll I lii 1 1 Si 111

fj

ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

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

\R o 198

-ifiRARitS

VOLUME 14 • PART 1
20th OCTOBER 1981

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

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

BD9S36

Some aspects of an ethological study of the aculeate wasps and the bees of a
karroid area in the vicinity of Grahamstown, South Africa

by

F. W. GESS

(Albany Museum, Grahamstown)

CONTENTS

Page

Abstract 1

Introduction 2

Aspects of the ecology of the study area

Location and topography 3

Geology and soil types 5

Climate 7

Vegetation 8

Forage plants 9

Seasonal cycles of wasps and bees 9

A classification of the aculeate wasps and of the bees of the study area on the basis of

their ethology 9

Annotated list of the aculeate wasps and of the bees of the study area, arranged accord-
ing to the classification based on their ethology 16

Discussion 36

Acknowledgements 78

References 79

ABSTRACT

The present paper is based upon a survey of the aculeate wasps and the bees of a karroid
area in the Eastern Cape Province of South Africa. The location, topography, geology, climate
and vegetation of the study area are outlined. An annotated list of 241 species arranged on the
basis of their ethology is given.

There follows a discussion of the nesting behaviour of the community as a whole showing
how the diversity of the ethology of such a large number of sympatric species results in their
ecological displacement and giving some indication of the phylogenetic pattern leading to this
diversity.

From a dissertation approved for the degree of Doctor of Philosophy, Rhodes University.

1

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981
INTRODUCTION

During the eleven year period 1970-1980, the aculeate Hymenoptera occurring at Hilton,
a farm situated near Grahamstown, South Africa, were the subjects of extensive study by the
author and his wife and frequent co-author, S. K. Gess.

The choice of Hilton for the study followed the sampling of insect populations in areas of
various differing vegetation types in the vicinity of Grahamstown. From this survey it was
concluded that a karroid area showed the greatest promise with respect to potentially interest-
ing species.

Fieldwork at Hilton embraced the formation of extensive collections in the form of fre-
quent samples taken of the populations associated with various ecological situations. Methods
of sampling, besides the use of hand nets, included the use of Malaise traps and trap-nests.
Concomitant with the sampling of the populations, field studies of the nesting of many indi-
vidual species belonging to a range of families of both wasps and bees were pursued. An
outcome of this work was the publication between 1974 and 1980 of a number of papers deal-
ing with the ethology of certain species of solitary wasps (Gess, 1978, 1980a and 1980b; Gess
and Gess, 1974, 1975, 1976a, 1976b, 1980a, 1980b, 1980c and 1980d). Further publications of a
similar nature are in preparation.

Among the species already reported upon in the above listed publications are two species
of Dichragenia, D. pulchricoma (Arnold) and D. neavei (Kohl) which are of particular interest
as, in the Pompilidae, they bridge the gap between fossorial nesters and mud-using aerial nes-
ters. Similarly, in two species of Bembecinus, B. cinguliger (Smith) and B. oxydorcus (Hand-
lirsch), the use of water in the excavation of the nest and the construction of a mud entrance
turret appears to represent a nest type previously unknown in the Sphecidae. Parachilus insig-
nis (Saussure) is shown to sub-divide the cell into an egg compartment and a pantry compart-
ment in which way it differs in behaviour from other species of Eumenidae. The ethological
account of the three species of Ceramius (Masaridae) clarifies some uncertainties and serious
misconceptions, most notably showing that C. lichtensteinii (Klug) is a mass provisioner and
not as was previously believed a progressive provisioner. The accounts of the nesting of Para-
chilus insignis (Saussure) (Eumenidae), Jugurtia confusa Richards (Masaridae) and Kohliella
alaris Brauns (Sphecidae) are the first for the three genera to which these species belong. That
of Holotachysphex turned (Arnold) (Sphecidae) greatly augments what little was previously
known of the ethology of its genus. Of particular interest with respect to Dasyproctus wester -
manni (Dahlbom) (Sphecidae) is the orientation of the pupae which appears to be governed by
gravity rather than by the position of the nest entrance which is usually the rule. For the genus
Cerceris (Sphecidae) twelve new prey associations give weight to the author’s hypothesis that
Cerceris is an Old World genus, that its original prey was hymenopterous and that the taking
of Coleoptera is derived.

The present paper represents the hitherto unpublished portion of a dissertation submitted
to Rhodes University and approved for the degree of Doctor of Philosophy. Other than in the
omission of the above listed previously published papers as well as three further papers of a
supporting nature (all bound into the dissertation) the present paper deviates from the dis-
sertation only in the following respects: in the substitution of an abstract for the original sum-
mary; in the present expanded introduction; in the modified list of references (omitting refer-
ences listed in and pertaining to the published papers); in the omission of an appendix listing
plant species characterizing the vegetation communities of Hilton.

The study is the first attempt in southern Africa to consider an entire community of acule-
ate wasps and bees and their interactions with their environment as manifested by their etholo-
gy. As far as the author has been able to ascertain it is in fact the most comprehensive of its
kind to have been undertaken anywhere. The only comparable account, that of Evans (1970),

2

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

is more restricted in its scope as it is mainly concerned with fossorial species and their associ-
ates. It moreover deals with a fauna already fairly well known whereas the present study deals
with a fauna which is largely unstudied ethologically. By its nature, the present paper serves
also to place in correct context within the community of which they are members those species
concerning which ethological studies have already been published and provides a matrix within
which the subjects of further, yet to be published, ethological studies may be fitted.

All the material, including voucher specimens, amassed during the course of the study is
housed in the collection of the Albany Museum, Grahamstown, in which institution are depo-
sited also all relevant fieldnotes.

ASPECTS OF THE ECOLOGY OF THE STUDY AREA
Location and topography

The farm Hilton, the site of the present study, is situated 18 kilometres WNW of Grahams-
town (33° 19'S., 26° 32'E.) in the Albany Division of the Eastern Cape Province of South
Africa very close to the midpoint of a straight line connecting the village of Riebeek East and
Grahamstown (Figs 1 and 2).

Fig. 1. Map of southern Africa showing position of study area.

3

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

Fig. 2. Topographical map showing position of study area.

Situated on the course of the New Year’s River, Hilton at a minimum altitude of 1 550
-1 650 feet (472-503 m) above sea level occupies low-lying land more or less surrounded by
rising ground forming the lower slopes of a series of sub-parallel mountain ranges running
roughly ESE. Within a radius of 25 kilometres of Hilton these ranges consist of an unnamed
range passing through the trigonometrical point Grootfontein at 3 058 feet (932 m) just to the

4

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

north of Riebeek East; another range, the Swartwatersberg, just south of Riebeek East which
rises to a height of 3 010 feet (917 m) at the beacon Riebeek; a third which passes just south of
Grahamstown and which, near Highlands at a point roughly 13 kilometres SSW of Hilton,
attains 2 727 feet (831 m) and finally Botha’s Hill ridge to the north of Grahamstown which
rises to 2 780 feet (847 m) at Driver’s Hill. South of the range through Highlands the topogra-
phy consists of a series of dissected peneplains gradually dropping in elevation southwards
towards the coast, 58 kilometres due S of Hilton. As will be shown below, the climate of
Hilton and especially the rainfall is influenced to a marked extent by the surrounding topogra-
phy, the ranges immediately to the south being of particular importance.

At Hilton itself the altitude of the study area varies from a minimum of just less than
1 550 feet (472 m) in the river bed at the foot of the cliff to the west to a maximum of about
1 850 feet (564 m) on the slopes of the E-W ridge to the south. Most of the study area,
however, is situated at an altitude of between 1 550 and 1 650 feet (472 and 503 m) (Fig. 3.)

The New Year’s River which rises just north of Grahamstown enters the study area from
the north-east and after a markedly meandering course across the low-lying flat area of Hilton
leaves the study area in the south-west shortly before winding through precipitous gorges on its
way to Alicedale where it joins the Bushman’s River. Along its course through Hilton the river
is impounded at two points by stone weirs with the result that upstream from these weirs water
is present at all seasons in two reaches of considerable length (1 300-1 500 m). Below the
weirs the river breaks up into isolated pools during the dry seasons. Several small seasonal
tributaries in wide shallow valleys join the New Year’s River along its course through Hilton.
Of particular importance with respect to the present study is a tributary, marked as the Iron
Put River on the Geological Survey Cape Sheet No. 9, which enters the study area from the
south-east, flows more or less parallel to the rising ground in the south and joins the river about
450 m upstream of the lower weir. Arising in the mountains (the third range listed above) and
draining an area S to SE of Hilton in the vicinity of Atherstone, this tributory (the Iron Put
R.) falls in altitude by 700 feet (213 m) over the distance of about 8 kilometres from its sources
to its confluence with the New Year’s River at the point indicated above. After an initial rapid
rate of descent the rate decreases progressively and over the last 2Vz kilometres before the
confluence the fall is of the order of 90 feet (27 m). In the study area its course is over only
very slightly sloping ground and is in the form of shallowly incised meanders cut into its flood
plain. Frequently it is completely dry or at most holds a few pools but following heavy rain in
its catchment area comes down in flood, overflowing its banks and spreading over its flood-
plain. About 2 kilometres above its confluence with the New Year’s River a dam across the
course of this tributary provides water for a furrow running below the 1 600 foot contour
across the central part of the study area.

Also in this part of the study area is a shallow earth dam fed after rain by run-off from
slightly sloping ground to the east and partly by water overflowing from a break in the above
furrow, the lower course of which passes immediately to the north of the dam. In both the dam
and the furrow water is seasonal and temporary. Water generally flows in the furrow for some
time after good rains have fallen in the catchment area but as the dry season advances the
water shrinks to stagnant pools and then dries up completely.

Geology and soil types

Geologically Hilton is situated upon three conformable strata: from south to north these
are the Lower Witteberg Quartzites, the Upper Witteberg Shales (both belonging to the Cape
Supergroup) and Dwyka Tillite (the oldest stratum of the overlying Karroo Supergroup).

The contact between the strata of the Cape and the Karroo Supergroups is situated im-
mediately outside the study area to the north more or less in line with the road to Riebeek

5

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

6

Fig. 3. Map of study area.

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

East where this passes the turn-off to the Hilton homestead. Strata within the study area are
thus limited to the two Witteberg formations.

The pale grey Witteberg Quartzites form the ridge at the south of the study area and
indeed are the materials of the entire series of previously listed sub-parallel mountain ranges of
which the ridge is a part.

The rather dark coloured Witteberg Shales on the other hand form the lower-lying land
between the ridge in the south and the Dwyka Tillite in the north. The cliff at the west of the
study area is cut across this shale formation (Fig. 9) which is exposed also in the river bed
below the lower weir and at a point where the farm road crosses the 1 600 foot contour on the
rise north of the earth dam. On this rise are also found blocks of a dark sandstone derived
from a band of this material in the shale.

Upon weathering the Witteberg Quartzites give rise to a light grey sandy soil whereas the
Witteberg Shales (as also the Dwyka Tillite) give rise to brownish and reddish clayey soils.

As the entire study area, excepting the ridge in the south, is situated upon Witteberg
Shales it follows that the reddish clayey soil derived from this formation is basically the domi-
nant soil of the area. Whereas the clayey soil is somewhat thin in places on slopes such as those
of the rising ground north-east and east of the earth dam where the bedrock may be barely
covered, considerable thicknesses of this soil are present in lower-lying areas such as that
below the 1 600 foot contour east of the earth dam (Fig. 6) At this juncture it may be noted
that in describing the clayey soil of this particular locality within the study areas Gess and Gess
(1974: 192 and 1975: 24) stated that it was derived from the Dwyka Series. This identification
was consequent upon the use of the older geological classification which regarded the shales in
question as belonging to the Lower Dwyka rather than to the Upper Witteberg.

In the southern half of the study area light coloured sandy soil occurs overlying the clayey
soil and is in places intermixed with it. This sandy soil derived from the weathering of the
Witteberg Quartzite has two separate origins. On the higher ground in the south of the study
area the overlying sand is derived from the ridge and has been transported downhill by sheet
erosion. In thickness this sand tails off downslope (i.e. northwards). On the lower ground
below the 1 600 foot contour a wide band of similar sand but of alluvial origin has been depo-
sited upon its flood plain by the tributary of the New Year’s River previously referred to as the
Iron Put River. Near the middle of the flood plain near the present incised course of the
tributary the depth of alluvial sand is of the order of several feet, sufficiently deep for the
establishment of a pit (Fig. 5) for the removal of building sand and for the excavation by
aardvarks (Orycteropus afer) of their deep burrows.

Climate

The Albany Division, situated between the winter and summer rainfall regions, receives
rain in moderate amounts throughout the year, the wettest periods being spring and autumn,
the driest mid-winter.

The overall rainfall pattern for the region has been indicated by Dyer (1937: 33-34) who
showed the marked effect of topography upon precipitation. That author showed that rain
clouds from the coast on coming into contact with higher altitudes precipitate moisture either
in the form of rain or mist but that they pass uninterruptedly over lower valley levels which are
therefore drier. Specifically, regions of greater rainfall were shown to be the previously de-
scribed series of sub-parallel ESE running mountain ranges and further north the Fish River
Rand whereas the low, broad Fish River Valley, situated between these two belts of higher
ground, was shown to be a region of low rainfall.

The distribution of rainfall at Hilton and its more immediate surroundings fits well into
the overall pattern. This may be illustrated by means of average annual rainfall figures for
several localities in the area concerned. At Clifton at 1 276 feet (389 m) above sea level on the

7

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

peneplain near Seven Fountains, 22 km S of Hilton on the coastal side of the ESE ranges the
average annual rainfall is 575 mm (22,6") (Weather Bureau, 1965). Proceeding northwards the
rainfall figures rise with increasing altitude till the crest of the ESE range is reached where it is
assumed the rainfall is highest. Unfortunately records for the crest are mostly unobtainable
for, as Dyer (1937) has pointed out, homesteads, the site of most rainfall stations, are general-
ly placed in more protected positions. Thus of four recording stations on the relevant part of
the range only one. Faraway near Coldsprings, is on the crest whereas the other three are
below the crest on the inland (drier) side.

Proceeding along the range from west to east, Atherstone at 2 263 feet (690 m) has an
average annual rainfall of 573 mm (22,55") (Dyer, 1937); Slaaikraal at 1 949 feet (594 m) has
652 mm (25,7") (Weather Bureau, 1965): Faraway at slightly above 2 400 feet (732 m) has circa
864 mm (34") (pers. comm. C. F. Jacot Guillarmod, Oct., 1978), and Grahamstown at 1 768
feet (539 m) has 688 mm (27,1") (Weather Bureau, 1965).

The average annual rainfall figure for Hilton: circa 356 mm (14") (pers. comm. T. C.
White, Febr., 1973) is in marked contrast to the above rainfall figures for localities on the
ranges. The low rainfall at Hilton is clearly attributable to the nature of the topography for at
1 600 feet (488 m) the study area lies 700-1 100 feet (213-335 m) below the crests of the ESE
ranges only 7-12 kilometres distant to the SW, S and SE and is therefore situated in a rain
shadow area similar to that of the Fish River Valley to the north.

The low rainfall in the rain shadow at Hilton raises to considerable ecological importance
the fact that the catchment areas of the water courses passing through the study area are in
regions of greater rainfall on higher ground. The introduction of water from outside the rain
shadow area results in the presence of semi-permanent to permanent water (apart from in-
pounded water such as in the earth dam) even when the study area itself is very dry.

Vegetation

From a consideration of “Veld Types of South Africa” (Acocks, 1953) and the accom-
panying map the study area is seen to lie on the tongue of False Karroid Broken Veld which
extends eastwards towards Grahamstown, principally along the valley of the New Year’s
River. As the land rises to the north and south there is a transition to Valley Bushveld merging
into Eastern Province Thornveld in the south-east. False Macchia is present in the highest
areas.

The study area is characterized by dwarf karroo scrub with, where sufficient water and
bare earth are available, an addition of thorn scrub and along the river banks an admixture of
small trees. The sandy foot of the E-W ridge in the south is characterized by grassveld which
gives way to scattered succulents, notably Aloe ferox, and xerophytic bushes on the dry, rocky
north-facing slope (Fig. 13). To the west of the study area the valleys become narrower and
steeper and are characterized by Valley Bushveld which gives way as the land rises in the
south-west to False Macchia. Further to the south-east and rising from the low-lying study area
there is a transition to Eastern Province Thornveld.

The dwarf karroo scrub may be divided into three main communities. These are:

(1) Pentzia incana Scrub which characterizes clayey soil which has not been disturbed by
cultivation (Gess, 1980b: Fig. 1).

(2) Pentzia incana — Chrysocoma tenuifolia Scrub which characterizes previously cultivated
clayey soils in the area extending from above the furrow to the New Year’s River below
the earth dam (Fig. 6).

(3) Chrysocoma tenuifolia Scrub which characterizes the previously cultivated sand flats of the
flood plain of the Iron Put River and extends across to its confluence with the New Year’s
River (Fig. 5).

All three of these communities are subject to trampling and grazing by sheep and cattle.

8

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

The thorn scrub which is characterized by Acacia karroo occurs within the dwarf karroo
scrub area in places where the soil has been exposed and where in addition the available
amount of water in the soil is greater than elsewhere in the area (Fig. 12). Such sites are those
where there is seepage from the furrow and earth dam and where the water run-off over
eroded land causes the formation of temporary standing water in erosion gullies and low-lying
areas.

The rivers represent older and more permanent water erosion gullies and their banks
therefore support the same species as are present in the thorn scrub community except that the
conditions being more favourable the vegetation no longer falls into the classification of scrub
but rather of small trees (Fig. 8 and Gess and Gess, 1975: Plate 3).

Along the New Year’s River where permanent and semi-permanent water bodies occur in
natural pools and for considerable reaches above the two weirs the water is fringed by beds of
Phragmites australis , “Common Reed” (Fig. 14). These reeds are also present in the earth dam
immediately within the retaining wall (Fig. 15). Sedges which characterize marshy ground are
present in clumps along the river banks, along the course of the furrow and fringe and earth
dam (Gess and Gess, 1974: Plate 2).

Forage plants

A large number of flowering plants in the study area are visited by aculeate wasps and by
bees of both sexes for the purpose of obtaining their own nutriment. Little specificity is shown,
the requirement generally appearing to be the availability of nectar, short tubed or open flow-
ers being favoured. Nearness to nesting sites is also of importance, suitable flowering plants
growing at any considerable distance from nesting sites not being visited.

Particularly attractive to a large number of species are the flowers of Maytenus linearis
(Celastraceae) and flowers and sticky young growth of Acacia karroo (Leguminosae). How-
ever, many species visiting these plants will also forage on low-growing plants, for example
Selago corymbosa (Selaginaceae), Melolobium candicans (Leguminosae), Senecio spp. and
Lasiospermum bipinnatum (both Compositae). Composites are often favoured at the begin-
ning of the summer when Acacia karroo and Maytenus linearis are not yet in flower.

When plants are used as a source of provision for the young, or as a source of nesting
materials, or their tissue is used as a subtrate in which to excavate a nesting gallery more
specificity is shown. These associations are given in the annotated list.

Seasonal cycles of wasps and bees

The majority of the aculeate wasps and the bees of the study area show marked seasons of
activity and quiescence. In these species the period of quiescence extends from the end of the
nesting season in the summer until the following spring or summer when activity recommences
with the emergence of the adults from the nests in which they have remained in the pre-pupal
or pupal stage throughout the cold, usually dry winter months. In most of these species there
appears to be one generation per year but in some species such as Ampulex sp. near cyanura
there may be two generations — a non-diapausing first generation and a diapausing second gen-
eration.

The minority of species such as the carpenter bees and the social wasps (Vespidae) which
emerge as adults before the end of the summer overwinter as adults (both males and females
in the bees, fertilized females in the vespids) and may show a certain limited activity on warm
days during the winter months.

A CLASSIFICATION OF THE ACULEATE WASPS AND OF THE BEES OF THE
STUDY AREA ON THE BASIS OF THEIR ETHOLOGY

The aculeate wasps and the bees collected in the study area are grouped on the basis of
three ethological characters, situation of the nest (I, II, III, IV) subdivided into IN and ON

9

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

(A, B), nature of substrate (a, b, c) and degree of participation in construction of nest (i, ii, iii)
(Fig. 4). This grouping results in the formation of twenty-seven categories numbered 1, 2, 3
. . . 27.

I. NESTING IN THE GROUND (Figs 5 and 6)

(a) In non-friable soils. (Typically compacted clayey soils.)

(i) In nest constructed entirely by the nester. 1

(ii) In pre-existing cavity modified by the nester. 2

(iii) In pre-existing cavity not modified by the nester. 3

(b) In friable soils. (Typically sandy soils, exceptionally disturbed and therefore
non-compacted clayey soils.)

(i) In nest constructed entirely by the nester. 4

(ii) In pre-existing cavity modified by the nester. 5

(iii) In pre-existing cavity not modified by the nester. 6

II. NESTING IN OR ON VERTICAL BANKS (Figs 7, 8 and 9)

A. IN VERTICAL BANKS

(i) In nest constructed entirely by the nester. 7

(ii) In pre-existing cavity modified by the nester. 8

(iii) In pre-existing cavity not modified by the nester. 9

B. ON VERTICAL BANKS

(i) In nest constructed entirely by the nester. 10

(ii) In pre-existing cavity modified by the nester. 11

(iii) In pre-existing cavity not modified by the nester. 12

III. NESTING ON STONES (Figs 10 and 11)

(i) In nest constructed entirely by the nester. 13

(ii) In pre-existing cavity modified by the nester. 14

(iii) In pre-existing cavity not modified by the nester. 15

IV. NESTING IN OR ON PLANTS (Figs 12, 13, 14 and 15)

A. IN PLANTS

(a) Within woody stems.

(i) In nest constructed entirely by the nester. 16

(ii) In pre-existing cavity modified by the nester. 17

(iii) In pre-existing cavity not modified by the nester. 18

(b) Within pithy stems.

(i) In nest constructed entirely by the nester. 19

(ii) In pre-existing cavity modified by the nester. 20

(iii) In pre-existing cavity not modified by the nester. 21

(c) Within hollow stems.

(i) In nest constructed entirely by the nester. 22

(ii) In pre-existing cavity modified by the nester. 23

(iii) In pre-existing cavity not modified by the nester. 24

B. ON PLANTS (Irrespective of whether plants are woody, pithy or hollow-
stemmed.)

(i) In nest constructed entirely by the nester. 25

(ii) In pre-existing cavity modified by the nester. 26

(iii) In pre-existing cavity not modified by the nester. 27

10

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

In the annotated list which follows those species for which nesting was definately con-
firmed are listed with the symbol***.

Those species which were observed searching for nesting sites or transporting nesting
materials or nest provision but for which no nests were discovered are listed with the sym-
bol**.

The remaining species listed with the symbol* have been allocated on the basis of where
they were collected in conjunction with morphological features and known biology of related

Fig. 4. Diagrammatic representation of the classification of the aculeate wasps and the bees of the study area on the basis of
their ethology. The numbers (i, ii, iii) in the outermost ring correspond to the same numbers in the classification, (i) In nest
constructed entirely by the nester. (ii) In pre-existing cavity modified by the nester. (iii) In pre-existing cavity not modified
by the nester. The magnitude of the segments is in proportion to the number of species involved.

11

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

Fig. 5. Area of friable soil: the sandpit

Fig. 6. Area of non-friable soil: below furrow.

12

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Fig. 7

Fig. 8

Fig. 9

Figs 7, 8 and 9. Vertical banks along course of New Year’s River: sand (Fig. 7), “sandstone” (Fig. 8) and shale (Fig. 9).

13

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

Fig. 10. Stone lying loose on ground and bearing nest of Auplopus rossi.

Fig. 11. Stone embedded in ground and bearing nest of Hoplitis jansei.

14

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Fig. 12. Woody stemmed plant: Acacia karroo on bank of
watercourse.

Fig. 13. Pithy stemmed plant: Aloe ferox on north-facing
slope.

Fig. 14 Fig. 15

Figs 14 and 15. Hollow stemmed plant: Phragmites australis fringing New Year’s River (Fig. 14). in earth dam (Fig. 15).

15

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

ANNOTATED LIST OF THE ACULEATE WASPS AND OF THE BEES OF THE
STUDY AREA, ARRANGED ACCORDING TO THE CLASSIFICATION BASED ON

THEIR ETHOLOGY.

Category 1.

I. NESTING IN THE GROUND
(a) In non-friable soils.

(i) In nest constructed entirely by the nester.

POMPILIDAE

*** Dichragenia neavei (Kohl). Uncommon. Nov. -Apr. Constructs turreted multicellular nest; each cell
provisioned with a single lycosid or sparassid spider. Nesting described by Gess and Gess (1976b).

*** Dichragenia pulchricoma (Arnold). Common. Oct. -Apr. Constructs turreted multicellular nest; each
cell provisioned with a single lycosid, pisaurid, sparassid or salticid spider. Nesting described by
Gess and Gess (1974).

EUMENIDAE

*** Antepipona scutellaris G. Soika. Common. Oct. -March. Constructs turreted one- or two-celled nest;
each cell provisioned with about 12 small pyralidid caterpillars.

*** Parachilus insignis (Saussure). Common. Oct.-Febr. Constructs one- or two-celled nest; each cell
provisioned with 8-17 psychid caterpillars. Nesting described by Gess and Gess (1976a).

*** Parachilus capensis (Saussure). Rare. Oct. -Jan. Constructs turreted one- or two-celled nest; each
cell provisioned with about 11 psychid caterpillars.

*** Pseudepipona erythrospila (Cameron). Rare. Oct. -March. Constructs collared one-celled nest; cell
provisioned with pyralidid caterpillars.

*** Eumenid sp. G. Rare. Jan. Constructs turreted one-(? or two-) celled nest; cell provisioned with
chrysomelid larvae.

*** Eumenid sp. (minute, black). Rare.

MASARIDAE

*** Ceramius capicola Brauns. Common. Nov. -March. Constructs turreted multicellular nest; each cell
provisioned with a mixture of pollen and nectar derived from Mesembryanthemaceae.

*** Ceramius lichtensteinii (Klug). Common. Oct. -April. Constructs turreted multicellular nest; each cell
provisioned with a mixture of pollen and nectar derived from Mesembryanthemaceae. Nests
perennial.

*** Ceramius linearis (Klug). Rare. Oct. -Jan. Constructs turreted multicellular nest; each cell pro-
visioned with a mixture of pollen and nectar derived from Mesembryanthemaceae.

* Jugurtia braunsiella (von Schulthess). Rare. Sept. -Dec.

*** Jugurtia confusa (Richards). Common. Sept.-Febr. Constructs turreted multicellular nest; each cell
provisioned with a mixture of pollen and nectar derived from Mesembryanthemaceae.

Nesting of above Masaridae described by Gess & Gess (1980c).

SPHECIDAE

*** Bembecinus cinguliger (Smith). Very common. Nov. -March. Constructs turreted one- or two-celled
nest; each cell provisioned with numerous Cicadellidae (and also Fulgoroidea). Nesting described
by Gess & Gess (1975).

*** Bembecinus oxydorous (Handlirsch). Common. Jan. -Apr. Constructs turreted one- or two-celled
nest; each cell provisioned with numerous Cicadellidae. Nesting described by Gess & Gess (1975).

16

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Category 2.

I. NESTING IN THE GROUND
(a) In non-friable soils.

(ii) In pre-existing cavity modified by the nester.

EUMENIDAE

*** Eumenid sp. Very rare. March. One record. Curved, lacy turret constructed surmounting turret of
empty burrow of Bembecinus oxydorcus. (Gess & Gess, 1975: 40).

SPHECIDAE

* Ampulex mutilloides Kohl. Very rare. Jan. One record. Circumstantial evidence suggests that nesting

may be in ground burrow of cockroach, Pilema sp.

*** Prionyx kirbii (Van der Linden). Common. Nov. -May. Nesting in old or abandoned burrows of
Parachilus insignis (Fig. 24). Single cell provisioned with 1-2 grasshoppers (Acrididae: Acrotylus,
Aiolopus, Anaeolopus, Calliptamulus and other genera).

* Tachysphex modestus Arnold. Rare. Dec.

*** Tachysphex sp. near modestus Arnold. Rare. Dec. -Jan. Nesting in old or abandoned burrows of
Parachilus insignis (Fig. 25). Single cell provisioned with 2 grasshoppers (Lentulidae).

** Pison allonymum Schultz. Rare. Oct. -May. Strong circumstantial evidence suggesting that nesting is
in old or abandoned burrows of Bembecinus cinguliger and B. oxydorcus. Prey consists of small
spiders (e.g. Salticidae). In Gess & Gess (1975: 40) incorrectly referred to as Pison ?montanum
Cam.

ME GA CHI LI DA E

** Megachile (Eutricharaea) aliceae Cockerell. Rare. Oct. -Dec. Nesting in old burrows of Parachilus
insignis. Cells constructed of ?leaves. (Gess & Gess, 1976a: 98).

*** Megachile (Eutricharaea) meadewaldoi Brauns. Uncommon. Sept. -March. Nesting in old or aban-
doned burrows of Bembecinus oxydorcus, Parachilus insignis and Antepipona scutellaris. Cells con-
structed of petals from low-growing flowers (Oxalis sp., Wahlenbergia sp.). (Gess & Gess, 1975: 40
and 1976a: 98).

** Megachile ( Eutricharaea ) semiflava Cockerell. Common. Sept. -March. Nesting in old or abandoned
burrows of Parachilus insignis. Cells constructed of leaves. (Gess & Gess, 1976a: 98).

*** Megachile ( Eutricharaea ) stellarum Cockerell. Common. Sept. -March. Nesting in old or abandoned
burrows of Dichragenia pulchricoma, Parachilus insignis and Ceramius lichtensteinii. Cells con-
structed of leaves. (Gess & Gess, 1974: 204-206, Fig. 8; 1976a: 98; 1980c: 78).

*** Creightoniella dorsata (Smith). Common. Nov. -March. Nesting in old or abandoned burrows of
Parachilus insignis. Cells constructed of leaves. (Gess & Gess, 1976a: 98).

ANTHOPHORIDAE

* Tetralonia minuta Friese. Uncommon. Nov. -Dec. Nesting in old or abandoned burrows of Bembeci-

nus cinguliger and Parachilus insignis. (Gess & Gess, 1975: 40 and 1976a: 98). For account of
nesting see also Rozen (1969a).

17

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

Category 3.

I. NESTING IN THE GROUND
(a) In non-friable soils.

(iii) In pre-existing cavity not modified by the nester.

CHRYSIDIDAE

*** Allocoelia bidens Edney. Common. Nov. Parasitic in the nests of Jugurtia confusa. (Gess & Gess,
1980c: 76)

Allocoelia capensis Smith. Not yet recorded from Hilton but highly likely to occur there. Parasitic in
the nests of Ceramins lichtensteinii. (Gess & Gess, 1980c: 76).

Allocoelia latinota Edney. Not yet recorded from Hilton but highly likely to occur there. Parasitic in
the nests of Ceramius capicola and Ceramius linearis. (Gess & Gess, 1980c: 76).

*** Octochrysis vansoni (Brauns). Common. Nov. -Jan. Parasitic in the nests of Parachilus insignis.
(Gess & Gess, 1976a: 97).

*** Pseudospinolia arderis (Mocsary). Uncommon. Nov. -Dec. Parasitic in the nests of Parachilus insig-
nis. (Gess & Gess, 1976a: 97).

MUTILLIDAE

* Brachymutilla gynandromorpha Andre. Rare. Oct. Found in nesting area of Jugurtia confusa.

* Dasylabris stimulatrix (Smith). Rare. Nov.-Febr. Found in nesting areas of Jugurtia confusa and

Parachilus insignis. (Also found in association with friable soils.)

*** Dasylabroides caffra (Kohl). Common. Oct. -May. Found in nesting areas of Dichragenia pulchrico-
ma, Parachilus insignis, Bembecinus oxydorcus, Jugurtia confusa and Ceramius lichtensteinii.
Reared from the cells of C. lichtensteinii in which parasitic. (Gess & Gess, 1980c: 76).

(Rarely found in association with friable soils and vertical banks.)

* Glossotilla speculatrix (Smith). Rare. March. Found in nesting area of Bembecinus oxydorcus. (Very

commonly found in association with friable soils.)

* Mutilla scabrofoveolata Sichel & Radoszkowski. Rare. Nov. -April. Found in nesting areas of

Dichragenia pulchricoma and Parachilus insignis. (Also rarely found in association with vertical
banks.)

* Smicromyrme sp. Rare. Found in nesting area of Jugurtia confusa.

* Smicromyrme hecuba (Peringuey). Common. Dec. -April. Found in nesting areas of Parachilus insig-

nis and Bembecinus cinguliger and B. oxydorcus. Circumstantial evidence strongly indicates that
parasitic in the nests of Bembecinus spp. (Common also in association with friable soils.)

POMPILIDAE

*** Ceropales punctulatus Cameron. Rare. Nov. -March. Parasitic in the nests of Dichragenia pulchrico-
ma. (Gess & Gess, 1974: 202, 204.)

MEGA CHILIDAE

** Coelioxys (Liothyrapis) lativentroides Brauns. Rare. Dec. Circumstantial evidence very strongly indi-
cates that this species is parasitic in the nests of Megachilidae ( Megachile and Creightoniella spp.)
nesting in old or abandoned burrows of Parachilus insignis.

18

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Category 4.

I. NESTING IN THE GROUND
(b) In friable soils.

(i) In nest constructed entirely by the nester.

POMPILIDAE

** Batozonellus fuliginosus sepulchralis (Smith). Uncommon. Nov. -Apr. Observed searching for prey
on Acacia karroo and dragging prey, Caerostris sp., across sandpit. Known to dig shallow, sloping
one-celled nest and to provision cell with a single large argiopid spider. Araneus sp. or Caerostris
sp. (Gess & Gess, 1980b).

** Cyphononyx flavicornis antennatus Smith. Common. Nov. -Mar. Observed dragging prey, Palystes
sp. (Sparassidae) across ground.

* Tachypompilus ignitus (Smith). Uncommon. Nov.-Febr. Known to dig nest in dry sand in sheltered

situation and to provision cell with a single large sparassid spider, Palystes natalius Karsch (Gess &
Gess, 1980b).

SPHECIDAE

*** Prionyx kirbii (Van der Linden). Common. Nov. -May. In friable soils this species appears to dig its
own nest. (In non-friable soils nests in old or abandoned burrows of Parachilus insignis.)

*** Podalonia canescens (Dahlbom). Common. Aug. -May. Digs simple one-celled nest; cell provisioned
with a single hairless caterpillar of the cutworm type (Noctuidae).

*** Ammophila beniniensis (Palisot de Beauvois). Common. Oct. -Apr. Nesting in disturbed, loose and
crumbly clayey soil.

* Ammophila bonaespei Lepeletier. Rare. Dec., Mar. Known to provision with hairless caterpillars

(Geometridae).

* Ammophila conifera (Arnold). Uncommon. Nov. -Mar.

*** Ammophila ferrugineipes Lepeletier. Very common. Oct. -May. Nesting in fine but consolidated
sand. Digs a simple burrow terminating in single cell; cell provisioned with several hairless caterpil-
lars (Geometridae) carried to nest in flight. Parasitized by Stylopidae (Strepsiptera).

* Ammophila insignis litoralis (Arnold). Uncommon. Febr.-Apr.

* Ammophila vulcania du Buysson. Rare. Nov. and Dec.

* Diodontus sp. Rare. Oct. and Nov. Species of this genus are known to nest in soil and to provision

with aphids. (Bohart & Menke, 1976: 178).

*** Astata fnscistigma Cameron. Rare. Nov. -Apr. Nesting in compacted sand. Astata species are known
to provision with bugs, especially Pentatomidae (Bohart & Menke, 1976: 211-212).

* Dryudella fiavoundata (Arnold). Rare. Dec. Prey known to be a “small heteropteran” (Arnold,

1924: 38).

* Liris spp. (Some species may belong here, others may belong in category 2). Known to provision

with crickets (Gryllidae).

* Tachytella aureopilosa Brauns. Rare. Jan. Only species in genus; biology unknown. Presence in

female of foretarsal rake and of pygidial plate indicative of sand-nesting. Provisioning likely to be
with Orthoptera (sensu lato).

* Tachysphex aethiopicus Arnold. Rare. Jan.

*** Tachysphex albocinctus (Lucas). Common. Dec. -Mar. Digs simple one- or two-celled nest, each cell
provisioned with 1-3 mantids (Mantidae), usually nymphs.

*** Tachysphex fugax (Radoszkowski). Common. Sept. -May. Digs simple two-celled nest; each cell pro-
visioned with several grasshoppers (Lentulidae), adults and nymphs.

* Tachysphex karroensis Arnold. Uncommon. Dec.-Febr.

19

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

*** Tachysphex panzeri pentheri Cameron. (= caliban Arnold). Common. Oct. -Mar. Digs simple two-
celled nest; sealed cell examined contained two prey; grasshopper nymphs (Acrididae).

* Tachysphex schoenlandi Cameron. Common. Dec. -Mar. Known to provision with nymphal mantids

(Mantidae).

* Tachysphex sericeus (Smith). Uncommon. Nov. -May.

* Parapiagetia vernalis Brauns. Rare. Nov. Biology of genus largely unknown. Presence in female of

foretarsal rake and of pygidial plate indicative of sand-nesting. Immature Acrydiidae recorded as
prey of the Madagascan P. longicornis Arnold (Arnold, 1945: 94).

*** Kohliella alaris Brauns. Common. Dec. -Mar. Digs shallow nest with one to several cells; each cell
provisioned with several nymphs of the Tree Cricket, Oecanthus capensis Saussure (Gryllidae:
Oecanthinae). Nesting described by Gess & Gess (1980a).

*** Palarus latifrons Kohl. Common. Nov. -Mar. Unicellular nest dug in firm sand overlain by a layer of
loose sand. Provisions with honey bees. Apis mellifera L. (Apidae). Brauns (1911: 117) in addition
to honey bees recorded the following prey:

Mesa cf (cited as Elis) (Tiphiidae), Ceramius capicola Brauns (Masaridae) and small bees. In
South Africa known as the Banded Bee Pirate and is a nuisance to beekeepers (Mally, 1908:
206-213; Brain, 1929: 396; Taylor, 1939: 103; and Smit, 1964: 356).

* Palarus oneili Brauns. Rare. Jan. and Febr. Prey known to include Meria $ (cited as Myzine)

(Tiphiidae) and small bees (Brauns, 1911: 117).

* Oxybelus acutissimus propinquus Arnold. Rare. Nov. -Mar.

* Oxybelus aethiopicus Cameron. Rare. Jan.

* Oxybelus imperialis Gerstaecker. Rare. Nov.

*** Oxybelus lingula Gerstaecker. Common. Nov. -Mar. Nesting in loose sand. Like ail Oxybelus spp.
provisions with adult Diptera. Prey recorded at Hilton: Stomorhina lunata (F.) (Calliphoridae) and
Musca sp. (Muscidae). Other prey known for this species: Sarcophaga sp. (Calliphoridae).

* Oxybelus peringueyi Saussure. Rare. Oct., Nov. and Mar.

* Oxybelus rubrocaudatus Arnold. Uncommon. Oct. -Jan.

* Oxybelus ruficaudis Cameron. Uncommon. Oct., Nov. and Jan.

* Dienoplus vicarius karrooensis (Brauns). Uncommon. Nov. -Mar. Species of the genus are known to

dig simple nests in sandy soil. There may be from 2-15 cells per nest; cells are provisioned with
Cicadellidae and Cercopidae (Bohart & Menke, 1976: 495).

* Oryttus kraepelini (Brauns). Uncommon. Dec. and Mar. The well developed foretarsal rake in the

female is indicative of nesting in friable soil. Species of the genus are known to provision with
Cicadellidae and Fulgoridae (Bohart & Menke, 1976: 507).

** Hoplisoides aglaia (Handlirsch). Uncommon. Oct. -Mar. Associated with sandy soil. Provisions with
Membracidae.

* Hoplisoides thalia (Handlirsch). Uncommon. Dec., Jan. and Apr.

* Stizus imperialis Handlirsch. Rare. Dec., Jan. Stizus species are known to provision mostly with

grasshoppers.

*** Bembecinus braunsii (Handlirsch). Very common. Oct. -Apr. Nesting in loose dry fine sand. Digs
one- or two-celled nest; each cell provisioned with numerous Cicadellidae (incl. Macropsis octo-
punctatus China and Macropsis chinai Metcalf), exceptionally with one or two Fulgoroidea in addi-
tion.

* Bembecinus dentiventris (Handlirsch). Very rare. Nov., Dec.

*** Bembecinus haemorrhoidalis (Handlirsch). Very common. Oct. -Apr. Nesting in loose dry fine sand.
Digs one- or two-celled nest; each cell provisioned with numerous Cicadellidae (inch Coloborrhis
corticina Germar, Exitianus nanus (Distant), Macropsis octopunctata China, Macropsis chinai Met-
calf, Macropsis sp. nov., Idioscopus sp. and Batracomorphus subolivaceus (Stal)), exceptionally
with one or two Fulgoroidea or Membracidae in addition.

20

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

*** Bembix albofasciata Smith. Common. Nov.-Febr. Digs a one-celled nest in sandy soil; cell pro-
visioned progressively with adult Diptera belonging to the following families:

Stratiomyidae, Tabanidae ( Chrysops obliquefasciata Macq.), Bombyliidae (several spp., inch
Lomatia pictipennis Wied.), Asilidae (several spp. inch Stenopogon dilutus Walker), Conopidae
( Conops sp.), Muscidae (2 spp. inch Musca lusoria Wied.), Calliphoridae (inch Sarcophaga sp.
and Chrysomyia sp.) and Tachinidae.

* Bembix cameronis Handlirsch. Uncommon. Nov. -Jan.

* Bembix capensis Lepeletier. Uncommon. Dec., Jan. Single prey obtained was an adult fly (Tachin-

idae).

* Bembix fusciipennis Lepeletier. Rare. Mar.

*** Bembix melanopa Handlirsch. Uncommon. Dec.-Febr. Both nests located were in steeply sloping
firm and compacted sand. Known to provision with adult Diptera (incl. Muscidae).

* Bembix sibilans Handlirsch. Uncommon. Nov. -Mar.

* Philanthus loefflingi Dahlbom. Rare. Nov., Dec.

*** Philanthus triangulum Fabricius. Uncommon. Oct. -Jan. Nesting in clayey yet friable soil. Provision-
ing with honey bees. Apis mellifera L. (Apidae) captured at their foraging flowers.

*** Cerceris sp. A. Rare. Dec. Nest excavated in disturbed clayey soil. Shallow multicellular nest; each
cell provisioned with numerous Pteromalidae.

* Cerceris amakosa Brauns. Rare. Dec. -Mar. Associated with sandy soil.

* Cerceris armaticeps caffrariae Empey. Very rare. Febr.

* Cerceris diodonta diodonta Schletterer. Rare. Dec.

* Cerceris discrepans discrepans Brauns. Very rare. Nov.

* Cerceris dominicana Brauns. Very rare. Nov., Dec.

* Cerceris erythrosoma Schletterer. Very rare. Dec. Associated with sandy soil. Known to provision

with Curculionidae.

*** Cerceris holconota holconota Cameron. Common. Nov. -Apr. Nesting in sandy soil; provisioning
cells with Hymenoptera of various families (Braconidae, Bethylidae, Tiphiidae, Mutillidae, Formi-
cidae and Halictidae).

* Cerceris hypocritica Brauns. Rare. Dec.-Febr. Associated with sandy soil.

*** Cerceris languida languida Cameron. Very common. Oct. -Apr. Nesting in sandy soil; provisioning
cells with Phalacridae (Olibrus sp.).

*** Cerceris latifrons latifrons Bingham. Very common. Nov. -Mar. Nesting in sandy soil in very deep
nests; provisioning cells with Scarabaeidae (Melolonthinae: Hopliini).

*** Cerceris lunigera Dahlbom. Uncommon. Dec. -Mar. Nest excavated in disturbed clayey soil; pro-
visioning cells with Halictidae.

* Cerceris nasidens obscura Schletterer. Uncommon. Nov., Dec., Mar. and Apr. Associated with

sandy soil. Known to provision its cells with Scarabaeidae (Melolonthinae).

* Cerceris nigrifrons nigrifrons Smith. Uncommon. Nov.-Febr. Associated with sandy soil. Known to

provision its cells with Buprestidae (Brauns, 1926a: 320).

*** Cerceris oraniensis Brauns. Uncommon. Dec. -Mar. Nesting in sandy soil in deep nests; provisioning
with Curculionidae (Protostrophus sp.) (Fig. 17).

* Cerceris pearstonensis pearstonensis Cameron. Common. Dec. -Apr. Nest excavated in disturbed

clayey soil. Known to provision its cells with Curculionidae.

* Cerceris pictifacies Brauns. Rare. Dec.

*** Cerceris ruficauda ruficauda Cameron. Uncommon. Oct. -Apr. Nesting in disturbed clayey soil; pro-
visioning with Chrysomelidae.

*** Cerceris rufocincta polychroma Gribodo. Very common. Nov. -Apr. Nesting in sandy soil; provision-
ing with Tiphiidae and Mutillidae.

*** Cerceris spinicaudata spinicaudata Cameron. Common. Nov. -Mar. Nesting in sandy soil; provision-
ing with Halictidae.

21

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

Ethological notes concerning above Cerceris spp. given in Gess (1980a).

COLLETIDAE

* Colletes sp. Uncommon.

HALICTIDAE

* Halictus sp. A. Common. Oct. -Apr. Associated with sandy soils.

* Halictus sp. B. Common. Nov. -Mar. Associated with sandy soils.

* Lasioglossum spp. Uncommon. Associated with sandy soils.

*** Nomiodes ?halictoides Bliithgen. Common. Oct. -Mar. Nesting in sandy soil.

*** Nomia sp. Common. Nov.-Febr. Nesting in sandy soil.

Category 5.

I. NESTING IN THE GROUND
(b) In friable soils.

(ii) In pre-existing cavity modified by the nester.

SPHECIDAE

* Liris sp. Uncommon. May belong here or in category 4. Known to provision with crickets (Gryll-
idae).

*** Megachile (Eutricharaea) semiflava Cockerell. Common. Sept. -Mar. Frequently seen flying low over
the ground in the sandpit carrying green leaf-discs; on several occasions seen to be nesting in holes
on the floor of the sandpit. Identity of original excavator of holes not established.

Category 6.

I. NESTING IN THE GROUND
(b) In friable soils.

(iii) In pre-existing cavity not modified by nester.

CHRYSIDIDAE

A total of seven species was recorded in association with sandy soils — in the sandpit. Of these, six
species (Chrysidea africana Mocsary, Hedychrum coelestinum Spinola, H. gonomaculatum Edney,
Octochrysis laminata (Mocsary), O. mucronifera (Mocsary) and Pyriachrysis stilboides (Spinola))
were rare and were mostly more commonly found associated with vertical banks.

One species, however, was both restricted to and common in the sandpit, namely:

Hedychrum sp. (near comptum Edney). Nov. -Mar. Associated with the nests of Bembecinus braunsii
and B. haemorrhoidalis and believed to be parasitic in them.

TI PH II DAE

A total of seventeen species was recorded in areas of friable, mostly sandy soil in and around the
sandpit. Sixteen of these species (Anthobosca sp., Meria spp. (10), Mesa spp. (2) and Tiphia spp. (3))
are believed to be predaceous upon soil-inhabiting larvae of Scarabaeidae. The remaining species
was:

Methocha mosutoana Peringuey. Rare. Jan. Methocha spp. are known to be predaceous upon the
burrow-inhabiting larvae of tiger-beetles (Cicindelidae). The prey of the present species is certain
to be Cicindela brevicollis Wied., adults of which are present in very large numbers in the sandpit
throughout the summer months and well into the autumn or even early winter.

22

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

MUTILLIDAE

Fourteen species of Mutillidae were collected in association with the friable sandy soil of the sandpit.
The most common species by far was Glossotilla speculatrix (Smith) (46 females and 67 males col-
lected) followed by Smicromyrme hecuba (Peringuey) (29 females and 11 males collected), Mimeco-
mutilla renominanda Bischoff, Dasylabris sp. nr. danae (Peringuey) and Dasylabris mephitis (Smith).
Also recorded but uncommon were Antennotilla phoebe (Peringuey), Dasylabris sp. nr. bassutorum
(Andre), Dasylabris sp. nr. cryentocincta Andre, Dasylabris stimulatrix (Smith), Dasylabris thais
(Peringuey), Dasylabroides caffra (Kohl), Labidomulilla tauriceps (Kohl), Psammotherma flabellata
(Fabricius) and Viereckia sp.

With the exception of Dasylabroides caffra which is a species characteristic of clayey rather than
sandy areas and Smicromyrme hecuba which is common in both sandy and clayey areas, all the above
listed species, with the exception of Antennotilla phoebe, were more common in the sandy areas than
elsewhere and many appeared restricted to such areas.

Glossotilla speculatrix is highly characteristic of the sandy areas and in contrast to the larger
numbers caught in the sandpit only two specimens were collected in bare patches in clayey areas.
Though not reared from any nests circumstantial evidence indicates that the species is parasitic in the
nests of Bembecinus species though other species nesting in the same situations are doubtless also
parasitized. Its flight period is from December to April.

Smicromyrme hecuba appears like Glossotilla speculatrix to be associated with Bembecinus spe-
cies and the two species of mutillid are usually found together within the areas occupied by Bembeci-
nus pseudocolonies in the sandpit. Its flight period is similarly from December to April.
SCOLIIDAE

Eleven species of Scoliidae, believed to be associated with friable soils, were collected, many on
flowers. By far the most common species was Cathimeris capensis (Saussure) (28 males and 13
females collected) which was present in the sandpit from September to May. Less common or rare
were Campsomeriella sp., Micromeriella spp. (2), and Scolia spp. (7). All are believed to be pre-
daceous upon soil-inhabiting larvae of Scarabaeidae.

SPHECIDAE

* Nysson braunsi Handlirsch. Rare. Nov. -Jan. Species of Nysson are known to be cleptoparasitic in

the nests of various gorytin genera including Oryttus, Dienoplus and Hoplisoides (Bohart & Menke,
1976: 468). Representatives of these genera associated with friable soils at Hilton are O. kraepe-
lini , D. vicarius karrooensis, H. thalia and H. aglaia and it may be assumed that the present species
of Nysson is associated with one or more of these potential host species.

HALICTIDAE

*** Sphecodes sp. Common. Sept.-Febr. In the sandpit frequently observed inspecting holes in the
ground and entering them. Found associated with Nomia sp. which species is believed to have
been the excavator of the nests concerned and thus to be the host of this Sphecodes sp. The genus
Sphecodes is known to be cleptoparasitic in the nests of other Halictidae (Halictus and Lasioglos-
sum) (Michener, 1944: 250) and also certain Andrenidae and Colletidae (Krombein et. al. 1979:
1974).

MEGACHILIDAE

* Coelioxys ( Lyothyrapis ) bruneipes Pasteels. Very rare. Jan. Possibly parasitic in the nests of

Megachile (Eutricharaea) semiflava Cockerell.

ANTHOPHOR1DAE

* Epeolus amabilis Gerstaecker. Uncommon. Oct. -Apr. The species of Epeolus are known to be para-

sitic in the nests of Colletes (Michener, 1944: 279) and of Tetralonia and Colletes (Arnold, 1947:
218). Mostly collected in the sandpit where it probably parasitizes the Colletes sp.

* Thyreus sp. A. Uncommon in sandpit, more usually found associated with “sandstone” bank. In

sandpit female observed inspecting holes in the ground.

23

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

Category 7.

II. NESTING IN OR ON VERTICAL BANKS
A. IN VERTICAL BANKS

(i) In nest constructed entirely by the nester.

POMPILIDAE

** Dichragenia jacob (Arnold). Uncommon. Apr. -May. Nine females, three with prey caught, at the
“sandstone” bank with which this species, which was not found elsewhere, was definately associ-
ated. No nests located but females seemed to be coming and going to and from a small cave so
situated on the bank that it could not be reached by the author. Assumed to construct its nest itself
as do other Dichragenia species. Prey: lycosid spiders.

SPHECIDAE

*** Chalybion (Hemichalybion) spinolae (Lepeletier). Rare. Nov. -May. Nesting in the “sandstone”
bank, circumstantial evidence indicates that the wasp excavates cavities itself, rather than utilizing
pre-existing ones. Nesting cavity mud-lined. Prey: known to be small spiders.

MEGACHILIDAE

*** Hoplitis anthodemnion Michener. Common. Nov. -Feb. Nesting in the “sandstone” bank. Nest con-
sists of excavated inclined burrow terminated by a single excavated cell containing a cell con-
structed from discs cut from petals (including those of Barleria pungens). Nest closure, occupying
space between provisioned cell and nest entrance, consists of tightly packed sand.

Nesting described in detail by Michener (1968).

ANTHOPHORIDAE

*** Anthophora sp. Common. Oct., Nov. and Mar. Nesting in sand bank and “sandstone” bank. Prob-
ably the most common species at the “sandstone” bank, and its old and abandoned nests are
probably those used most often by species nesting in this situation in pre-existing cavities (that is
species of category 8).

Category 8.

II. NESTING IN OR ON VERTICAL BANKS
A. IN VERTICAL BANKS

(ii) In pre-existing cavity modified by the nester.

POMPILIDAE

* Auplopns carinigena Cameron. Rare. Apr. -May.

* Auplopus ferruginea (Magretti). Rare. Apr. -May.

** Auplopus personata ornaticollis (Cameron). Uncommon. Apr. Prey known to consist of errant
spiders of the families Clubionidae, Salticidae and Sparassidae.

All three species were collected on the face of the “sandstone” bank. A. personata was observed
hunting, boldly entering and searching crevices containing spider-spinnings, at times pushing its way
through veils of spinnings at the entrance of the spiders’ lairs; a prey-carrying female was furthermore
seen entering a pre-existing cavity and it is consequently believed that this species (and the other
two) may nest in such cavities. Support for this belief comes from Taylor (1968:72) who recorded
Auplopus mazoensis (Arnold) nesting in a trap-nest and using mud to form cell partitions.

24

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

EUMENIDAE

A number of species, collected in association with vertical banks, is included in the present section,
some having been observed nesting in pre-existing cavities, others being suspected of doing so. The
species include:

*** Eumenid A. Uncommon. Nov. Nesting in pre-existing cavities in the “sandstone” bank, in cracks
within and between the layers of the near-horizontally bedded shale of the shale bank, and in
trap-nests inserted into larger crevices in the latter. Nest entrance furnished with a lacy down-
turned mud turret, nest a linear series of cells separated by mud partitions, each cell provisioned
with numerous small bright green caterpillars (Pyralididae).

* Anterhynchium natalense (Saussure). Rare. At “sandstone" bank.

* Antodynerus radialis oogaster (Gribodo). Rare. Feb. -Apr. At “sandstone” bank.

* Euodynerus euryspilns (Cameron). Rare. Nov. -Feb. At “sandstone” bank. Recorded in a later

category (17) nesting in trap-nests tied to Acacia karroo.

*** Rynchium marginellum sabulosum (Saussure). Uncommon. Feb. Nesting in trap-nests inserted into
crevices in the “sandstone” bank and thus highly likely to be nesting in pre-existing cavities in the
bank itself. Nest a linear series of cells separated by mud partitions; each cell provisioned with
several (4-13) caterpillars (Pyralididae). Known in Grahamstown to nest in cut culms of Arundo
dortax used for vegetable frames.

*** Tricarinodynerus guerinii (Saussure). Uncommon. Jan. and Feb. Nesting in pre-existing cavities in
sandbank, “sandstone” bank and shale bank as also in a trap-nest inserted into a larger crevice in
the latter. Nest entrance furnished with a downturned mud turret (Fig. 28), nest a linear series of
cells separated by mud partitions; each cell provisioned with several caterpillars, probably Tortric-
idae. Anthrax ?tetraspilus Hesse (Bombyliidae) reared from cells.

SPHECIDAE

* Chalybion ( Chalybion ) tibiale (Fabricius). Rare. Nov. -Apr. Circumstantial evidence strong that this

species nests in pre-existing cavities in the “sandstone” bank. Nesting described in a later category
(17).

* Pison allonymum Schultz. Uncommon. Dec. -May. Nesting recorded and described in previous

category (2).

* Pison montanum Cameron. Rare. Oct. -Mar. Nesting recorded and described in a later category

(17).

* Trypoxylon sp. Rare. Oct. -Mar. Nesting recorded and described in later category (17).
MEGACHILIDAE

*** Anthidiellum (Pygnanthidiellum) kimberleyanum Friese. Uncommon. Oct. and Nov. Nesting in
“sandstone” bank.

*** Branthidium braunsii (Friese). Uncommon. Oct., Nov. and Apr. Nesting in “sandstone” bank.

*** Chalicodoma ( Pseudomegachile ) schulthessi (Friese). Uncommon. Sept, and Oct. Nesting in “sand-
stone” bank. Nest entrance furnished with a small mud collar, nest a linear series of cells separated
by mud partitions, final closure also of mud.

* Heriades ?freygessneri Schletterer. Uncommon. Nov., Dec., Apr. and May. Thought to nest in

“sandstone” bank. H. freygessneri is known to nest readily in trap-nests (Taylor, 1962b, 1965 and
1968).

*** Immanthidium junodi (Friese). Common. Nov. -Apr. Nesting in “sandstone” bank. A carder bee.

Nesting described by Michener (1968), Skaife (1950) and Taylor (1962a).

*** Megachile ( Eutricharaea ) gratiosa Gerstaecker. Uncommon. Oct. -Mar. Nesting in trap-nests inserted
into crevices in the shale bank and presumably also in cracks within and between layers of the
shale. Nest a linear series of cells constructed of pieces cut from green leaves. Parasitized by
Zonitoschema eborina (Fabr.) (Meloidae). Nesting described by Taylor (1963, 1965 and 1968).

25

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

*** Megachile (Paracella) spinarum Cockerell. Uncommon. Oct. -Apr. Nesting in “sandstone” bank and
also in trap-nest inserted into crevice in the above. Nest a linear series of cells constructed of
pieces cut from green leaves of Maytenus heterophylla.

Category 9.

II. NESTING IN OR ON VERTICAL BANKS
A. IN VERTICAL BANKS

(iii) In pre-existing cavity not modified by the nester.

CHRYSIDIDAE

Seven of the eight species of cuckoo wasps collected on the vertical banks are believed to be associ-
ated with wasps nesting in the banks. The seven species, none of which was common, are: Chrysidea
africana (Mocsary), Chrysidea ghiliani Gribodo, Chrysis antiqua Brauns, Chrysis capitalis Dahlbom,
Hedychrum coelestinum Spinola, Octochrysis laminata (Mocsary), and Octochrysis mucronifera
(Mocsary). Hosts are known for two species:

* Chrysidea africana Mocsary. Known to be parasitic in the nests of Trypoxylon (see category 18).

*** Octochrysis laminata (Mocsary). Reared from the nests of Tricarinodynerus guerinii (Saussure)

(2 specimens).

MUTILLIDAE

Six species of Mutillidae were collected on the vertical banks, namely Antennotilla phoebe (Pering-
uey), Dasylabroides caffra (Kohl), Mimecomutilla renominanda Bischoff, Mutilla scabrofoveolata
Sichel & Radoszkowski, Ronisia trispilota (Radoszkowski) and Stenomutilla sp. No species was com-
mon and no species can be said to be typical of this situation. One species only was definitely associ-
ated with a wasp nesting in the bank, namely

*** Stenomutilla sp. Reared from the nest of Eumenid A. constructed within open crack in shale bank. (1
specimen).

SAPYGIDAE

* Sapyga (Sapygina) simillima Arnold. Uncommon. Oct., Nov. and Apr. Known to be parasitic in the

nests of Heriades sp. (see category 18. Possibly to be associated with Heriades ?freygessneri
Schletterer nesting in “sandstone” bank.

* Sapyga (Sapygina) undulata Gerst. Rare. Apr.

MEGACHILIDAE

** Euaspis abdominalis (Fabricius). Very rare. Nov. Circumstantial evidence indicates that this parasitic
bee may be associated with the nests of Chalicodoma schulthessi (Friese) in the “sandstone” bank.

* Coelioxys (Coelioxys) lucidicauda Cockerell. Very rare. Febr.

* Coelioxys (Lyothyrapis) bruneipes Pasteels. Very rare. Nov. Possibly parasitic in the nests of

Megachile (Paracella) spinarum Cockerell.

** Coelioxys (Lyothyrapis) lativentris Friese. Rare. Oct., Nov. Recorded as associated with an
Anthophora sp. and as probably parasitic in its nest (Rozen, 1969b: 55). In the present community
thus probably also associated with Anthophora.

** Coelioxys ( Lyothyrapis ) lativentroides Brauns. Rare. Oct.-Dec. Seen entering and leaving nests of
Chalicodoma schulthessi (Friese) and therefore believed to be parasitic in the nests of this species.
Supporting evidence for this belief supplied by Brauns (1930) who recorded this species as a para-
site of Chalicodoma latitarsis Friese.

ANTHOPHORIDAE

* Thyreus sp. Common. Nov. -May. A species of Thyreus was recorded as parasitic in the nests of an

Anthophora species (Rozen, 1969b) and the same association may exist in the present community.

26

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Category 10.

II NESTING IN OR ON VERTICAL BANKS
B. ON VERTICAL BANKS

(i) In nest constructed entirely by the nester.

EUMENIDAE

* Synagris abyssinica Guerin. Rare. Feb. and Apr. Flying about in front of “sandstone” bank. Known

to make aerial nests of mud and to provision with caterpillars.

* Synagris analis H. de Saussure. Rare. Mar. and Apr. As above.

*** Eumenes lucasius Saussure. Rare. Dec. & Apr. Urn-shaped mud cells found in sheltered positions on
shale bank. Prey: caterpillars. Parasite: Osprynchotus violator (Thunberg) (Ichneumonidae).

VESPIDAE

*** Ropalidia sp. A. Rare. One nest with foundress female found (Jan.) in a fist-sized cavity in the sand
bank. Nest aerial, constructed of wood pulp (“wasp paper”) and in the form of a naked paper
comb suspended from the roof of the cavity by a short pedicel. Social species: larvae fed directly
on macerated insects and nectar.

SPHEC1DAE

*** Sceliphron spirifex (L.) One nest with female found (Apr.) in a fist-sized cavity in the “sandstone”
bank and several found on shale bank. Nest aerial, constructed of mud and consisting of several
cells arranged parallel to one another and covered with a common envelope of mud applied after
the last cell has been sealed. Each cell provisioned with numerous small spiders. Parasite: Ospryn-
chotus violator (Thunberg) (Ichneumonidae).

Category 11.

II. NESTING IN OR ON VERTICAL BANKS
B. ON VERTICAL BANKS

(ii) In pre-existing cavity modified by the nester.

EUMENIDAE

*** Eumenid A. Rare. Nov. Nesting in old urn-shaped mud cells of Eumenes lucasius Saussure found in
sheltered positions on the shale bank. Prey: Pyralididae (caterpillars).

Category 12.

II. NESTING IN OR ON VERTICAL BANKS
B. ON VERTICAL BANKS

(iii) In pre-existing cavity not modified by the nester.

CHRYSIDIDAE

*** Chrysis sp. One specimen reared from the urn-shaped mud cell of Eumenes lucasius Saussure found
in a sheltered position on the shale bank.

MUTILLIDAE

* Stenomutilla sp. Rare. Known to be parasitic in the nests of Eumenid A. (see category 9).

27

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

Category 13.

III. NESTING ON STONES

(i) In nest constructed entirely by the nester.

POMPILIDAE

*** Auplopus rossi nigricomis (Arnold). Rare. One nest was found, consisting of nine contiguous mud
cells attached to the underside of a large flat stone where there was a shallow space between it and
the ground on which it rested. One of the cells when opened was found to have been provisioned
with a single spider of the family Lycosidae. Two male wasps and a female emerged from the cells
during Sept. -Oct.

MEGA CHILIDAE

*** Hoplitis jansei (Brauns). Uncommon. Nests constructed upon the exposed portions of partially
buried quartzite stones and boulders. Nests consist of independent cells closely grouped together
and constructed of small, irregularly shaped pebbles of quartzite cemented in a matrix of resin.
Emergence of adult bees from Aug. -Nov. Brauns (1926b: 220) recorded the species foraging on
low composites at Willowmore.

Category 14.

III. NESTING ON STONES

(ii) In pre-existing cavity modified by the nester.

MEGACHILIDAE

*** Megachile (Eutricharaea) gratiosa Gerstaecker. On one occasion was found to have utilized old open
cells of Hoplitis jansei (Brauns) for purposes of nesting. Old Hoplitis cells each contained a single
M. gratiosa cell constructed of pieces cut from green leaves. A male reared from such a cell
emerged during Nov. M. gratiosa nests more commonly in pre-existing cavities associated with
vertical banks and with woody plants (categories 8 & 17).

Category 15.

III. NESTING ON STONES

(iii) In pre-existing cavity not modified by the nester.

MUTILLIDAE

*** Antennotilla phoebe (Peringuey). Rare. Parasitoids in the cells of Auplopus rossi nigricomis
(Arnold). Three males reared from these cells emerged during early November.

28

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Category 16.

IV. NESTING IN OR ON PLANTS
A. IN PLANTS

(a) Within woody stems.

(i) In nest constructed entirely by the nester.

ANTHOPHORIDAE

* Xylocopa caffra L. Rare and all sightings confined to the riverine bush. Old abandoned galleries

found in dead trunks and branches of Rhus and Pinus located in the riverine bush believed to have
been those of this species. Known to nest in dead branches and structural timbers of Pinus spp.

* Xylocopa divisa Klug. Rare and all sightings confined to riverine bush. Known to nest in dead bran-

ches.

Category 17.

IV. NESTING IN OR ON PLANTS
A. IN PLANTS

(a) Within woody stems.

(ii) In pre-existing cavity modified by the nester.

EUMENIDAE

*** Euodynerus euryspilus (Cameron). Rare. Found nesting (Nov., Febr.) in four trap-nests tied to Aca-
cia karroo. Nest a linear series of cells, cell partitions and closing plug composed of mud. Each cell
provisioned with 7-13 caterpillars of family Pyralididae, in two nests Loxostege frustalis Zeller, the
Karroo Caterpillar, a pest of Pentzia incana.

SPHECIDAE

*** Ampulex sp. (near cyanura Kohl) (Fig. 36). Uncommon. Found nesting (Nov. -March) in thirty trap-
nests tied to Acacia karroo. Examination of dead finger-thick branches of this shrub or small tree
revealed that the natural pre-existing cavities used by this wasp were abandoned larval galleries of
Ceroplesis hottentota (Fabricius) (Cerambycidae, Coleoptera) (Fig. 35). In both natural and arti-
ficial nesting cavities nest consists of a single cell plugged with loosely compacted detritus. Each
cell provisioned with a single large nymph or apterous female of Bantua dispar (Burmeister) (Blat-
tidae) (Fig. 36).

*** Ampulex denticollis Cameron. Rare. Found nesting (Febr.) in one trap-nest tied to Acacia karroo.
Nest single-celled, plugged with detritus and provisioned with a single blattid nymph.

*** Chalybion (Chalybion) tibiale (Fabricius). Rare. Found nesting (Jan. and Febr.) in three trap-nests
tied to dead tree-stump in riverine bush. Nest a linear series of cells; cell partitions and closing
plug composed of mud; outer surfaces of partitions and closing plug whitened with uric acid from
bird droppings. Each cell provisioned with 14-34 small spiders of the families Argiopidae
(Araneus, Argiope, Caerostris, Cyclosa, Isoxya and Nephila). Thiridiidae (Rhomphaea) and Zodari-
idae. Nesting described by Gess & Gess (1980d).

*** Isodontia pelopoeiformis (Dahlbom). Common. Found nesting (Jan. -Feb.) in over eighty trap-nests
tied to Acacia karroo, Maytenus linearis and Rhus lancea in thorn scrub and riverine bush. Nest a
linear series of cells; cell partitions constructed of plant “fluff” derived from fruiting inflorescences
of Lasiospermum bipinnatum and Senecio spp.; preliminary plugs and closing plugs of the above

29

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

materials as well as clods of earth and debris (Fig. 32). Each cell provisioned with several katy-
dids—'Tettigoniidae: Phaneropterinae (Terpnistria zebrata nymphs, Eurycorypha prasinata

nymphs, and Phaneroptera spp. — nymphs and adults). Nesting of this species in Rhodesia de-
scribed in Smithers (1958).

* Isodontia simoni (du Buysson). Rare. Nov. & Dec. Not found nesting.

*** Isodontia stanleyi (Kohl). Uncommon. Found nesting (Dec.-Febr.) in eighteen trap-nests similarly
situated to those of I. pelopoeiformis . Nest a linear series of cells; cell partitions constructed of
short lengths of grass leaf blades transversely arranged; preliminary plugs and closing plugs of grass
inflorescences, transversely coiled and longitudinally arranged respectively (Fig. 33). Grasses used;
Danthonia curva, Diplachne fusca, Eragrostis sp. (probably E. curvula ), Melica racemosa and
Sporobolas sp. (probably S. fimbriatus). Each cell provisioned with several katydids, apparently
the same species taken by I. pelopoeiformis but smaller (younger) individuals. One cell in addition
contained two tree crickets (Oecanthus capensis Saussure (Gryllidae: Oecanthinae)).

*** Holotachysphex turneri (Arnold). Rare. Found nesting (Dec.-Febr.) in five trap-nests tied to Acacia
karroo and Salix mucronata. Nest a linear series of cells; cell partitions constructed of coarse
detritus and clayey earth; closing plug of coarse detritus. Each cell provisioned with a variable
number (4-32) of ?Pyrgomorphella sp. nymphs (Pyrgomorphidae, Acridoidea). Nesting described
by Gess (1978).

*** Pison montanum Cameron. Rare. Found nesting (Dec.-Febr.) in eight trap-nests tied to Acacia
karroo and Maytenus linearis. Nest a linear series of cells; cell partitions and closing plug con-
structed of mud. Each cell provisioned with a variable number (7-47) of small spiders of the family
Argiopidae.

*** Trypoxylon sp. Common. Found nesting (Oct.-Febr.) in twenty-five trap-nests tied to Acacia karroo
and Rhus lancea. Nest a linear series of cells; cell partitions and closing plug constructed of mud.
Each cell provisioned with a variable number (5-11) of small spiders of the following families:
Lycosidae (incl. Hippasa sp.), Pisauridae ( Euphrosthenops sp.), and Argiopidae (inch Araneus
spp., Larinia sp.).

COLLETIDAE

*** Hylaeus braunsi (Alfken). Rare. Dec. Found nesting in two trap-nests tied to Acacia karroo. Nest a
linear series of cells; cell partitions and nest closure of very shiny and thin transparent “cel-
lophane"-like material.

* Hylaeus spp. (at least two others).

ME GA CHI LI DA E

*** Immanthidium junodi (Friese). Common. Nov. -May. Commonly found nesting in trap-nests tied to
Acacia karroo. A carder bee constructing cell partitions and closing plug of cottonwool-like mate-
rial. Nesting of this species described or commented upon by Skaife (1950), Taylor (1962a) and
Michener (1968).

*** Heriades sp. Rare. Oct. Found nesting in one trap-nest tied to Maytenus linearis.

*** Chalicodoma (Pseudomegachile) fulva (Smith). Rare. Found nesting (Dec.) in two trap-nests tied to
Acacia karroo. Nest a linear series of mud cells.

*** Chalicodoma (Pseudomegachile) sinuata (Friese). Rare. Found nesting (Jan., Febr.) in three trap-
nests tied to Acacia karroo. Nest a linear series of mud cells. Parasitized by Zonitoschema eborina
(Fabr.) (Meloidae).

*** Megachile (Eutricharaea) gratiosa Gerst. Common. Found nesting (Dec. -Feb.) in trap-nests tied to
Acacia karroo. Nest a linear series of cells constructed of pieces cut from green leaves. Parasitized
by Zonitoschema eborina (Fabr.) (Meloidae).

*** Megachile (Paracella) spinarum Cockerell. Common. Found nesting in many trap-nests tied to Acacia
karroo. Nest a linear series of cells constructed of pieces cut from green leaves of Maytenus hetero-
phylla.

30

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Category 18.

IV. NESTING IN OR ON PLANTS
A. IN PLANTS

(a) Within woody stems.

(iii) In pre-existing cavity not modified by the nester.

CHRYSIDIDAE

*** Chrysidea africana Mocsary. Uncommon. Parasitic in the nests of Trypoxylon sp. nesting in trap-
nests. (Seven females reared.)

*** Chrysis sp. (near purpuripyga Edney). Uncommon. Parasitic in the nests of Trypoxylon sp. nesting in
trap-nests. (Nine individuals reared.)

*** Chrysis inops Gribodo. Rare. Parasitic in the nests of Pison montanum Cameron nesting in trap-
nests. (Three males and a female reared.) Previously recorded as parasitic in the nests of Pison
transvaalensis Cameron in Natal (Taylor, 1968).

*** Octochrysis hoplites (Mocsary). Rare. Parasitic in the nests of Euodynerus euryspilus (Cameron)
nesting in trap-nests. (Six individuals reared.)

SAPYGIDAE

*** Sapyga (Sapygina) simillima Arnold. Rare. Parasitic in the nest of Heriades sp. nesting in a trap-nest.
(One individual reared.)

ME GA CHI LID A E

*** Coelioxys (Coelioxys) penetratrix Smith. Rare. Parasitic in the nests of Megachile (Paracella) spina-
rum Cockerell nesting in trap-nests. (2 females and one male reared.)

Category 19.

IV. NESTING IN OR ON PLANTS
A. IN PLANTS

(b) Within pithy stems.

(i) In nest constructed entirely by the nester.

EUMENIDAE

* Raphiglossa natalensis Smith. Very rare. March. Not found nesting. Recorded at Willowmore (as R.

flavo-ornata Cameron) as excavating its cells in dry pithy stems (chiefly those of Liliaceae) and
provisioning its cells with small caterpillars (Lepidoptera) (Meade-Waldo, 1913: 45 and Bequaert,
1918: 29-30).

SPHECIDAE

* Dasyproctus bipunctatus Lep. & Brulle (inch the colour forms bipunctatus Lep., lugubris (Arnold)

and simillimus (Smith)). Flight period: Oct. -March. Common. Known to nest in the inflorescence
stems of Amaryllidaceae, Iridaceae and Liliaceae (Bowden, 1964: 425-437) and to provision with
adult Diptera.

* Dasyproctus dubiosus (Arnold). Very rare. Dec., Jan.

* Dasyproctus immitis (Saussure). Rare. Oct. -March.

* Dasyproctus ruficaudis (Arnold). Rare. Dec.-Febr.

*** Dasyproctus westermanni (Dahlbom). Common. Oct. -March. Found nesting during Dec. and Jan. in
inflorescence stems of Urginea altissima (Liliaceae); nest a linear series of cells in an ascending and

31

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

a descending gallery; cells provisioned with numerous small adult Diptera (Simuliidae, Stra-
tiomyidae, Bombyliidae, Empididae, Syrphidae, Otitidae, Chamaemyiidae). Parasites: Perilampus
sp. (Perilampidae, Chalcidoidea, Hymenoptera) and Phoridae (Diptera). Evidence of nesting by
Dasyproctus spp. (unidentified) also in inflorescence stems of Gasteria spp. (Liliaceae) and in
stems of Berkheya decurrens Compositae).

Nesting of Dasyproctus spp. (especially D. westermanni) described by Gess (1980b).

ANTHOPHORIDAE

*** Xylocopa (Gnathoxylocopa) sicheli Vachal. Common. At Hilton found nesting exclusively in dry but
still attached inflorescence stems of Aloe ferox (Liliaceae) (Fig. 30). Nest consists of a descending
gallery and an ascending gallery, both unbranched and subdivided serially into a number of cells
each sealed with a pithy plug made of material rasped from the gallery walls. Bees are present
throughout the year; nesting takes place during the summer. Parasite: Coelopencyrtus sp. (Encyrt-
idae, Chalcidoidea). Cleptoparasites: Gasteruption robustum Kieffer (Gasteruptionidae) and
Synhoria hottentota Peringuey (Meloidae).

*** Ceratina sp. A. Common. Commonly found nesting in the thinner branches of dry but still attached
inflorescence stems of Aloe ferox (Liliaceae), less commonly nesting in the stems of Gasteria spp.
(Liliaceae) and infrequently in stems of Datura stramonium (Solanaceae). Nest plan like that of X.
sicheli but all dimensions much smaller. Bees are present throughout the year; nesting takes place
during the summer.

*** Ceratina sp. B. Eleven further species known to occur at Hilton. C. sp. B found nesting in inflores-
cence stem of Aloe ferox (1 nest); One male of C. sp. C found sheltering in bored dry stem of
Berkheya decurrens (Compositae); one female and two males of C. sp. D. found sheltering in
bored dry inflorescence stem of Apicra sp. (Liliaceae).

Category 20.

IV. NESTING IN OR ON PLANTS
A. IN PLANTS

(b) Within pithy stems.

(ii) In pre-existing cavity modified by the nester.

SPHECIDAE

*** Isodontia stanleyi (Kohl). Three nests found within old galleries of Xylocopa sicheli in inflorescence
stems of Aloe ferox (Fig. 38). (For details of nesting of this wasp see previous entry for this species
in category 17.)

*** Trypoxylon sp. Nests found within old galleries of Ceratina sp. A. in inflorescence stems of Aloe
ferox, within old galleries of Ceratina and Dasyproctus spp. in inflorescence stems of Gasteria sp.,
and within galleries of Dasyproctus westermanni in inflorescence stems of Urginea altissima (Gess,
1980b). (For details of nesting of Trypoxylon see previous entry for Trypoxylon sp. in category
17.)

ANTHOPHORIDAE

*** Allodape rufogastra Lep. & Serv. OR Allodape exoloma Strand. Nests found within old galleries of
Dasyproctus sp. in stems of Berkheya decurrens. Bee nest characterized by lack of cell partitions
and by the eggs being cemented to the gallery wall.

MEGACHI LI DAE

*** Capanthidium capicola (Friese). Two nests found within old galleries of Ceratina sp. A. in inflores-
cence stem of Aloe ferox. A carder bee constructing cell partitions of cotton-wool-like material.

32

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

*** Immanthidium junodi (Friese). Nests found within old galleries of Ceratina sp. A. in inflorescence
stems of Aloe ferox, within old galleries of Ceratina and Dasyproctus spp. in inflorescence stems of
Gasteria sp., and within old galleries of Ceratina sp. in stems of Datura sp. (For details of nesting
of this bee see previous entry for this species in category 17.)

*** Heriades spiniscutis (Cameron). Two nests found within old galleries of (?) Dasyproctus sp. in stems
of Berkheya decurrens and one nest found within old gallery of Dasyproctus sp. in inflorescence
stem of Gasteria sp. Nest in linear series of cells; cell partitions (when present) and closing plug of
dark resin-like substance. Cell provisioned with bright yellow, rather dry pollen. The nesting of H.
spiniscutis has been described in detail by Michener (1968).

*** Chalicodoma (Pseudomegachile) sinuata (Friese). Six nests found within old abandoned galleries of
Xylocopa sicheli in inflorescence stems of Aloe ferox (Fig. 40). Nest in linear series of mud cells.

*** Megachile (Paracella) spinarum Cockerell. Six nests found within old abandoned galleries of Xyloco-
pa sicheli in inflorescence stems of Aloe ferox (Fig. 39). Nest a linear series of cells constructed of
pieces cut from green leaves of Maytenus heterophylla. Parasites: Leucospis africana Cameron
(Leucospidae) (three females reared from above nests) and a species of Cleridae (ITrichodes au-
licus Klug) (one larva found in one of the above nests). Also Coelioxys (Coelioxys) penetratrix
Smith (see below).

Category 21.

IV. NESTING IN OR ON PLANTS
A. IN PLANTS

(b) Within pithy stems.

(iii) In pre-existing cavity not modified by the nester.

No species were recorded in this category from the above nests within pithy stems and branches of
non-woody plants. However, parasites recorded from the nests of species nesting in trap-nests may
be expected to occur also in the nests of the same host species nesting in pithy stems and branches.
The following species are thus listed on the strength of this assumption:

CHRYSIDIDAE

Chrysidea africana Mocsary. Parasitic in the nests of Trypoxylon sp.

Chrysis sp. (near purpuripyga Edney). As above.

ME GA CHI LI DA E

Coelioxys ( Coelioxys ) penetratrix Smith. Parasitic in the nests of Megachile (Paracella) spinarum
Cockerell.

Category 22.

IV. NESTING IN OR ON PLANTS
A. IN PLANTS

(c) Within hollow stems.

(i) In nest constructed entirely by the nester.

ANTHOPHORIDAE

*** Xylocopa caffrariae Enderlein. Common. Nesting exclusively in dry but still attached hollow culms of
Phragmites australis (Gramineae), “Common Reed”, growing along the course of the New Year’s
River and immediately within the retaining wall of the earthen dam. Nest consists of a descending

33

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

and an ascending gallery subdivided serially into a number of cells each sealed with a pithy plug
made of material rasped from the gallery walls. Access to the hollow internode is through an
entrance hole cut by the bee through the 1 mm thick side wall of the culm (Fig. 31). Bees are
present throughout the year; nesting takes place during the summer. Cleptoparasite; Gasteruption
robustum Kieffer (Gasteruptionidae).

Category 23.

IV. NESTING IN OR ON PLANTS
A. IN PLANTS

(c) Within hollow stems.

(ii) In pre-existing cavity modified by the nester.

SPHECIDAE

*** Isodontia stanleyi (Kohl.) Two nests found within old galleries of Xylocopa caffrariae in bored hollow
culms of OHHRAGMITES AUSTRALIS. (For details of nesting of this wasp see previous entry
for this species in category 17).

*** Holotachysphex turneri (Arnold). Two nests found within old galleries of Xylocopa caffrariae in
bored hollow culms oe Phragmites australis. (For details of nesting of this wasp see previous entry
for this species in category 17, Gess (1978) and Gess & Gess (1980a: 52)).

*** Trypoxylon sp. A few nests found within old galleries of Xylocopa caffrariae in bored hollow culms
of Phragmites australis, a few others found in hollow culms of the same plant bored by lepidopter-
ous larvae. (For details of nesting of Trypoxylon see previous entry for Trypoxylon sp. in category
17).

MEGACHILIDAE

*** Chalicodoma (Pseudomegachile) sinuata (Friese). One nest found within old gallery of Xylocopa
caffrariae in bored hollow culm of Phragmites australis. Nest a linear series of mud cells.

Category 24.

IV. NESTING IN OR ON PLANTS
A. IN PLANTS

(c) Within hollow stems.

(iii) In pre-existing cavity not modified by the nester.

No species were recorded in this category from the above nests within hollow stems of hollow-
stemmed plants. However, parasites recorded from the nests of species nesting in trap-nests may be
expected to occur also in the nests of the same host species nesting in hollow stems. The following
species are thus listed on the strength of this assumption:

CHR YS1DIDA E

Chrysidea africana Mocsary. Parasitic in the nests of Trypoxylon.

Chrysis sp. (near purpuripyga Edney). As above.

34

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Category 25.

IV. NESTING IN OR ON PLANTS

B. ON PLANTS (Irrespective of whether woody, pithy or hollow-stemmed),

(i) In nest constructed entirely by the nester.

EUMENIDAE

*** Eumenes lucasius Saussure. Rare. Urn-shaped aerial mud cell found under loose bark of fallen tree.
Provisioned with caterpillars. A few other aerial mud nests constructed by Eumenidae were found
but in all cases they were old, the wasps having already emerged, and identification of the species
involved was not possible.

MASARIDAE

* Celonites capensis Brauns. A rare species at Hilton; not found nesting. Known to construct small

aerial mud cells. Provisions with a mixture of pollen and nectar.

VESPIDAE

*** Polistes smithii Saussure. One nest (with foundress female) found (Nov.) attached to branches of
Lycium overhanging “sandstone” bank. Nest aerial, constructed of wood pulp (“wasp paper”) and
in the form of a naked paper comb suspended from the branch by a short pedicel. Social species;
larvae fed directly on macerated insects and nectar. Adult males occasionally stylopized.

*** Ropalidia sp. B. One nest (with foundress female) found (Nov.) attached to branches of Cadaba
aphylla overhanging “sandstone” bank. Form of nest similar to that of Polistes smithii. Adult
males likewise occasionally stylopized.

SPHECIDAE

* Sceliphron quartinae (Gribodo). Rare. Nov., Febr. Not found nesting at Hilton. Known to construct

its cells of cow dung or clayey mud and to attach them singly or in twos to grass culms or other
plant stems at some height above the ground. Provisioning is with small spiders. (Brauns, 1911;
119 and Jacot Guillarmod, pers.comm.) At Hilton found at puddles, presumably collecting mud
for nest-building purposes.

Category 26.

IV. NESTING IN OR ON PLANTS

B. ON PLANTS (Irrespective of whether woody, pithy or hollow-stemmed.)

(ii) In pre-existing cavity modified by the nester.

No representatives recorded for this category.

Category 27.

IV. NESTING IN OR ON PLANTS

B. ON PLANTS (Irrespective of whether woody, pithy or hollow-stemmed.)

(iii) In pre-existing cavity not modified by the nester.

No representatives recorded for this category.

35

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

DISCUSSION

Evaluation of the completeness and representativeness of the sample

The classification of the aculeate wasps and of the bees of the study area on the basis of
their ethology encompasses an annotated list of a total of 241 species. The strengths of the
specific representation of the families involved are shown in Table 1.

Table 1.

The strengths of the specific representation of the families included in the annotated list.

Family

No. of spp.

Family

No. of spp.

Chrysididae

19

Vespidae

3

Tiphiidae

17

Pompilidae

11

Mutillidae

20

Sphecidae

91

Scoliidae

11

Colletidae

3

Sapygidae

2

Halictidae

6

Masaridae

6

Megachilidae

25

Eumenidae

16

Anthophoridae

11

It is inevitable that in a survey of the present kind not all the species present in a given
area (in the present instance, the farm Hilton) will be collected and identified and that it will
not be possible to allocate to any particular nesting association those species for which etholo-
gical information is lacking.

It is axiomatic, however, that in any area, the most common species are the most likely to
be noticed and recorded — that is most likely to be collected and most likely to be found nest-
ing or at least intimately associated with a particular nesting association (e.g. a particular type
of soil) in which nesting may be assumed to take place. It may therefore be expected that the
present account of the ethology of the aculeate wasps and the bees of Hilton includes amongst
the 241 species listed most if not all of the common and therefore most important species
occurring there. Judged on the basis of the inclusion in the sample of these common and
important species the sample may therefore be considered as representative of the entire com-
munity of the above defined insects.

Although further study of the whole area of Hilton would undoubtedly swell the lists of
some if not all the various categories of the present classification with further species, it is
believed that this quantitative change would not be accompanied by a qualitative change of
any real importance.

For some families at least it is possible to obtain some indication of how representative the
number of species listed above is of the total number of species occurring at Hilton. For
example, with respect to the Sphecidae, the predominant family of the sample, the 91 species
listed are believed to represent 80-90% of the total number of species of the family present at
Hilton, and the five and six species listed respectively for the Vespidae and the Masaridae are
believed to represent 60% and 100% of their total numbers, however) the nineteen listed
species of Chrysididae are believed to represent only 45% of their total number.

An evaluation of the classification of the aculeate wasps and the bees of the study area on the
basis of their ethology.

In the classification of the aculeate wasps and the bees of the study area on the basis of
their ethology 27 categories are recognized. When the 241 species are allocated to these cate-

36

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

gories it is found that 209 species occur in only one category each and 32 in more than one
category each. If, however, those species which construct their nests entirely by themselves are
considered, it is found that of the 118 species in this category only one (0,85%) occurs in two
categories, both of which offer the species in question, Eumenes lucasius , an aerial nester, a
support for its nest.

The species which modify pre-existing cavities and therefore do not have to excavate their
cavities themselves show, as one would expect, less specificity. Of the 44 species involved 13
species (29,5%) occur in more than one category. Of these 13 species 10 nest only above the
ground either in vertical banks or in plants and of these three were restricted to plants. One
species. Megachile semiflava , was restricted to cavities in the ground regardless of the nature of
the soil.

Of the 79 species which nest in pre-existing cavities which they do not themselves modify
17 (21,5%) occur in more than one category, however, these species, belonging to the Chrysi-
didae, Mutillidae, Sapygidae and Megachilidae such as Coelioxys , are “parasitic” in one way
or another and the categories in which they occur are therefore determined by those of their
hosts. These species are therefore 100% restricted on the basis of host.

It is clear that each community determined by habitat is characterized by those species
which construct their nests entirely by themselves but that there is, not surprisingly, overlap in
species using pre-existing cavities and a resultant overlap in species which are hosted by the
latter.

Having established the validity of the ethological classification nesting in the four main
habitat categories; ground, vertical banks, stones and plants; will be discussed.

Discussion of nesting in the four main habitat categories.

The Ground

Soils may be divided into two different types — friable and non-friable. By definition, the
word “friable” means easily crumbled and a friable soil therefore is one in which it is easy to
dig, the individual particles being relatively loosely aggregated and not difficult to part from
one another. “Non-friable” means not easily crumbled and a non-friable soil is therefore one
in which it is difficult to dig, the individual soil particles being closely aggregated and difficult
to part from one another. The character of a non-friable soil may, however, vary greatly with
the amount of moisture which it contains, a wet non-friable soil being more easily worked than
a dry one.

At Hilton the friable soil is mostly sandy in nature and is derived from the weathering of
Witteberg Quartzite and the non-friable soil is clayey in nature and is derived from the weath-
ering of Witteberg Shale. In certain small areas recent disturbance during farming activities has
broken down the structure of the clayey soil and has temporarily rendered it partially friable.

At Hilton, the number of species recorded as nesting in the ground exceeds the number
recorded from the three other situations combined: 167 species as against 74. Of the 167 spe-
cies recorded from the ground, 126 (75,5%) were associated only with friable soil, 35 (21%)
were associated only with non-friable soil and 6 (3,5%) were associated with both friable and
non-friable soils.

On account of the species which were recorded in both soil types, the number of species/
substrate associations exceeds the actual number of species by 6 and therefore totals 173. An
analysis of this figure according to the degree of participation of species of wasps and of bees in
the construction of their nests is given in Table 2 and is shown graphically and in greater detail
in the Pie-diagram (Fig. 16).

37

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

Table 2.

Number of species of aculeate wasps and of bees nesting in the ground.

In nests con-
structed entirely

In modified

In unmodified

by the nesters

pre-existing

pre-existing

themselves

cavities

cavities

In non-friable soils

15

12

14

In friable soils

76

2

54

It may be seen immediately that with respect to those species which excavate their nests
themselves there are many more (76 as against 15) that do so in friable soils than in non-friable
soils. It should be noted that this marked difference in the numbers of species nesting in the
two soil types is not a reflection of the extent of the availability of these soils for nesting, the
density of nesting in areas of friable soil being far greater than in areas of non-friable soil with
readily available water sources. Furthermore this discrepancy is not a peculiarity of the Hilton
population but bears out the picture which emerges from a consideration of ground-nesting
aculeate wasps as a whole.

It can be clearly seen from the Pie-diagram (Fig. 16) that, at Hilton, at the family level the
composition of the communities of species which construct their nests themselves are strikingly
different in friable and non-friable soils. Whereas in friable soils the vast majority of species,
67 (88%), are members of the family Sphecidae in non-friable soils this family is only repre-
sented by two (13%) of the species. The majority of species, 11 (74%), in non-friable soils are
members of the Vespoidea, five species of Masaridae and six species of Eumenidae. There are
no representatives of the Vespoidea nesting in friable soils. Species of Pompilidae excavate
their nests in both soil types. Whereas the two species of Dichragenia appear to be the only
pompilids associated in this way with non-friable soils, the number of species, 3, recorded for
friable soils is in all probability too low. Bees represented, albeit poorly, amongst the nest
excavators in friable soil are absent amongst those in non-friable soil.

The construction of nests in the two very different soil types clearly demands different
excavation techniques and therefore differences in the structure and behaviour of the wasps
involved.

Characteristic of the species excavating nests in friable soil is that there is no modification
of the physical nature of the substrate by the addition of water. Excavation of the nests is by
digging, the organs involved being the mandibles, the fore-legs and in some species the pygi-
dium.

Initial loosening of the soil at the working face of the excavation is frequently done with
the mandibles. These may be used simply to bite away the soil but in some species nesting in
compacted soil they are aided in their work by the wasp’s manipulation of its flight mechanisms
which produces vibrations transmitted by the mandibles to the substrate. Evidence of this
method of loosening the soil was the very noticeable buzzing sound that could frequently be
heard being made by excavating females of the sphecids Podalonia canescens and Ammophila
ferrugineipes. Frequently this buzzing sound emanating from the ground was the first indica-
tion that nest excavation by one of these species was under way in the immediate vicinity.

Removal of the loosened soil may be effected in several ways, on the basis of which
Olberg (1959) divided the digger wasps into “rakers”, “pullers”, “carriers” and “pushers”. A

38

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Fig. 16. Diagram showing the structure of the communities of the aculeate wasps and the bees nesting in the ground in
the study area. Numbers accompanying the family names represent numbers of species. The numbers (i, ii, iii) in the
outermost ring correspond to the same numbers in the classification of the aculeate wasps and the bees of the study area
on the basis of their ethology, (i) In nest constructed entirely by the nester. (ii) In pre-existing cavity modified by the
nester. (iii) In pre-existing cavity not modified by the nester.

good account of the different modes of digging based upon the above division is given by
Evans and Eberhard (1970) and will only be briefly outlined here, with the addition of exam-
ples drawn from the present study.

39

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

The “rakers”, constituting the majority of the fossorial wasps, are characterized by having
a strongly developed series of stout spines on the front tarsus, forming the so-called “tarsal
comb” or “sandrake”. In digging, the fore tarsi which are bent towards the midline of the
body are moved repeatedly backwards and forward, each backstroke throwing back a load of
soil which passes beneath the upheld abdomen to behind the body. By walking backwards
whilst so raking, loose soil can rapidly be swept from the excavation. Many of the species
observed nesting in friable soil at Hilton are included in the “rakers”, for example Bato-
zonellus fuliginosus (Pompilidae) (Gess and Gess, 1980b), Tachyphex spp., Kohliella alaris
(Gess and Gess, 1980a), Palarus latifrons, Bembecinus braunsii, Bembecinus haemorrhoidalis
and Bembix albofaciata (Sphecidae). In some species the two front legs move alternately (e.g.
Batozonellus), in other species synchronously (e.g. Bembix). By virtue of the fact that it is not
possible to rake or sweep loose sand vertically upwards, all the burrows of “rakers” examined
at Hilton were found to be inclined, often at a fairly small angle with the horizontal (for
example see nest plans of Kohliella alaris — Gess and Gess, 1980a: Figs 4 and 5).

The “pullers” like the “rakers” use the fore legs to remove the loosened soil from the
excavation. However, instead of raking out the soil, the latter is gathered together to form a
load held between the underside of the head and prothorax and the basal parts of the fore legs
and, the wasp walking backwards, this soil is then pulled out of the excavation and deposited
at its entrance. At Hilton a common “puller” is Podalonia canescens. However, this wasp does
also act as a “raker” for the small heap of excavated soil pulled to the nest entrance is from
time to time dispersed by raking. Raking of soil is also practised in nest closure. Soil pulling
unlike soil raking does not necessarily require the burrow to be inclined and the burrow of
Podalonia canescens is frequently subvertical, at least initially.

The “carriers” are like the “pullers” in their mode of removing the loosened soil from the
excavation. However, unlike the “pullers” they do not deposit the soil at the nest entrance but
drop it at a distance from the nest entrance, having transported it there either on foot or in
flight. At Hilton the most common carrier nesting in friable soil is Ammophila ferrugineipes
which carries the soil to a distance of 1-1,5 m from the nest before dropping it. Generally each
load of soil is dropped in the same area, mostly into a low bush. Despite the fact that each load
is dropped in the same place, the discarded sand is not noticeable for, being dropped from a
height into a bush by a wasp which is moving in flight, it is well scattered and concealed. Like
the burrow of Podalonia canescens , the burrow of Ammophila ferrugineipes is initially subver-
tical. The “pushers” differ from the “rakers”, “pullers” and “carriers” in their use of the end
of the abdomen in clearing soil from their excavations. Soil loosened by the mandibles and
fore legs is moved back by the legs to behind the wasp which then backs up its burrow pushing
the soil before it, in an action comparable to that of a piston in its cylinder. The common
“pushers” in friable soil at Hilton are the many species of Cerceris which for the purpose of
pushing soil with the end of the abdomen have a well developed pygidial plate. Characteristic
of Cerceris burrows is that they are vertical or subvertical and that the excavated soil in the
form of “sand sausages” forms a conical heap surrounding and surmounting the nest entrance
(Fig. 17). The length of the “sand sausage” indicates that this method of nest excavation
allows a great amount of loosened material to be brought to the surface at any one time — more
than can be moved at any one time by a “puller”. It is therefore an efficient and time-saving
method of excavation and allows the construction of nests of great depth. At Hilton, Cerceris
latifrons, for example, constructs a burrow far deeper than that of any other species examined
there — at a depth of 600 mm, the reach of the excavator’s arm, the burrow still continues
downwards.

Characteristic of those species constructing original nests in non-friable soils is that ex-
cavation of the nest is affected by the use of water used to soften the soil thus rendering it
workable. The water, collected by the nesting female from a water source, usually a pool or

40

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Fig. 17. Nest entrance of Cerceris oraniensis showing excavated soil in the form of “sand sausages”, (x 1)

puddle in the vicinity of the chosen nesting site, is carried to the nesting site in the crop. The
mechanics of excavation consist of the regurgitation of a droplet of water from the crop onto
the working face of the excavation, the working of this water into the soil by the mandibles,
and the formation of the resultant mud into a pellet which is carried from the excavation by
means of the mouthparts. Several such pellets may be formed from a single crop-full of water:
when the supply is depleted the wasp returns to the water source for a further crop-full. Whilst
the above outlined method of softening the soil is unique to those wasps nesting in non-friable
soils, the use of the mandibles for carrying out the pellets of mud from the excavation involves
the same behaviour as shown by many nesters in friable soils which may use their mandibles
for carrying out pebbles and other objects from their excavations though their digging is other-
wise done by raking or pulling the soil. Another facet of behaviour shared by certain excava-
tors in both friable and non-friable soils is the use of vibrations generated by the manipulation
of the flight mechanisms and transmitted by the mandibles to the substrate for the purpose of
loosening the latter. Recorded above with respect to the sphecids Podalortia canescens and
Ammophila ferrugineipes (“pullers” and “carriers” respectively), it is a feature also of the
excavating behaviour of the sphecid Bembecinus cinguliger (and probably also Bembecinus
oxydorcus) .

The use of water for nest excavation has been described in detail for several species of
wasps nesting in clayey soils at Hilton: Dichragenia pulchricoma (Arnold) (Gess and Gess,
1974) and Dichragenia neavei (Kohl) (Gess and Gess, 1976b) (both Pompilidae); Parachilus
insignis (Saussure) (Gess and Gess, 1976a) (Eumenidae); Ceramius capicola Brauns, Cera-
mius lichtensteinii (Klug), Ceramius linearis Klug and Jurgurtia confusa Richards (Gess and
Gess, 1980c) (all Masaridae); Bembecinus cinguliger (Smith) and Bembecinus oxydorcus

41

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

(Handlirsch) (Gess and Gess, 1975) (both Sphecidae). In addition to these species a further five
species, all Eumenidae, have been recorded using this method of nest excavation: included are
Antepipona scutellaris Giordani Soika, Parachilus capensis (Saussure) and Pseudepipona
erythrospila (Cameron).

The above species, in marked contrast to those excavating their nests in friable soils, make
no use of the fore-legs in manipulating the nesting substrate and these limbs are consequently
devoid of any modifications such as “sandrakes” used for digging. This is particularly striking
with respect to the sphecid genus Bembecinus , represented at Hilton by four very common
species, B. braunsii and B. haemorrhoidalis nesting in friable soil (in the sandpit) and B.
cinguliger and B. oxydorcus nesting in non-friable, clayey soil. Typically the members of this
large cosmopolitan genus are sand-raking and have the fore-tarsi furnished with long spines for
this purpose as in B. braunsii and B. haemorrhoidalis. In B. cinguliger and B. oxydorcus , by
contrast, these spines forming the “sandrake” are totally absent, the fore-tarsi having instead a
dense row of short spines (Figs 18 and 19).

Fig. 19. Left fore-tarsus of Bembecinus
haemorrhoidalis showing long spines forming
“sandrake". (x 32)

Fig. 18. Left fore-tarsus of Bembecinus cinguli-
ger showing dense row of short spines, (x 32)

Within the Pompilidae, the unmodified fore-tarsi of the two species of Dichragenia nesting
in non-friable soil may similarly be contrasted with the modified fore-tarsi of Batozonellus
fuliginosus nesting in friable soil (Figs 20 & 21).

It might be expected with respect to the clay-nesters that as all work associated with the
manipulation of the substrate and of the excavated material (mud pellets) is performed by the
mouthparts, principally the mandibles, some modifications associated with these organs might
be present. This does indeed appear to be the case in the two Dichragenia species, females of
which possess a brush of long, stiff, forwardly directed bristles emitted from the base of the

42

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Fig. 20. Left fore-tarsus of Dichragenia pulchricoma Fig. 21. Left fore-tarsus (distal four joints only) of

showing very short spines not forming a “sand- Batozonellus fuliginosus showing long spines form-

rake”. (x 28) ing “sandrake”. (x 24)

mentum (Figs 22 & 23 and Arnold, 1934: Figs 1, la), it is believed that these bristles may by
supporting it from below aid the mandibles in holding and manipulating a mud pellet.
Support for this belief may be drawn from two auplopodine genera closely allied to
Dichragenia — Phanagenia and Auplopus which possess similar brushes of mental bristles and
which use mud pellets to construct aerial mud cells.

43

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

Fig. 22. Underside of head of Dichragenia pulchricoma
showing stiff, forwardly directed bristles emitted from the
base of the mentum. (x 56)

Fig. 23. Underside of head of Batozonellus fuliginosus
showing absence of forwardly directed bristles, (x 24)

In the brief outline given above of the excavating methods used by nesters in friable soils,
it was shown that whereas soil raking allowed the construction only of inclined burrows, soil
pulling and the development of soil pulling — soil carrying — allowed the construction of vertical
burrows. As the species excavating their nests in non-friable soils all remove the excavation
spoils in the form of mud pellets held by the mandiblss it follows that they may be considered
soil carriers and it is not surprising that the nests excavated are without exception vertical or
subvertical (Gess and Gess, 1974: Figs 2-5; 1975; Figs 10-17; 1976a: Figs 1-4; 1976b: Figs
1-2; 1980c: Figs 1-5). The genus Bembecinus may once again be used as an illustration, in this
instance to show the differences in the nest plan resulting from different excavation techniques
employed in the two substrates. Thus, whereas the nest of several sandraking species has been
described by Evans (1966: 137-138) as being without exception oblique, a description also
fitting the nests of B. braunsii and B. haemorrhoidalis , the nest of the mud pelletcarrying
B. cinguliger and B. oxydorcus is vertical or subvertical. The distinction can be clearly seen by
comparison between Evans (1966: Fig. 76) and Gess and Gess (1975: Figs 10-17). The same
picture emerges from a comparison of the nest plan of Batozonellus fuliginosus (Gess and
Gess, 1980b: 5) with those of the two Dichragenia species.

Characteristic of all the species recorded excavating their nests in non-friable clayey soil is
that at least some of the excavated material — in the form of mud pellets — is used to build an
aerial superstructure encircling and surmounting the nest entrance. Among the species in-
volved, the two Dichragenia species are unique in that all the pellets removed from their
burrows are added to the superstructures. All other species stop adding pellets to the super-
structure once this has attained a certain size, and further pellets are discarded at varying
distances from the nest, depending upon the species.

44

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Each species builds a superstructure of definite form so that the identity of the builder
may readily be established from the style of its architecture.

The largest and most striking superstructures are constructed by the two Pompilidae: that
of Dichragenia pulchricoma is retort-shaped (Gess and Gess, 1974: 195-196, Pis 5-8 and
1976b: Fig. 3) whereas that of D. neavei is funnel-shaped (Gess and Gess, 1976b: PI. 1 and
Figs 1 and 2).

The nest superstructures of the Eumenidae are all of a single basic type or developments
thereof: a low collar encircling the nest entrance in Parachiius insignis (Gess and Gess, 1976a:
88, Fig. 1 and Pis 3-8) and in Pseudepipona erythrospila; a very low turret with a downwardly
recurved lip in Eumenid sp. G; a short vertical flared tube in Antepipona scutellaris ; and a
vertical funnel-shaped turret in Parachiius capensis.

The four Masaridae studied all construct nest superstructures in the form of cylindrical
turrets. That of Jugurtia confusa is characteristically very short (Gess and Gess, 1980c: 66, Figs
1 and 7) whereas those of the three species of Ceramius are longer: those of C. linearis (Gess
and Gess, 1980c: 66, Fig. 3) and C. lichtensteinii (Gess, 1973: 117, PI. 2; and Gess and Gess,
1980c: 66, Figs 4, 5 and 9) are initially vertical or subvertical but may, if of any considerable
length, curve over and that of C. capicola (Gess, 1973: 117, PI. 1; and Gess and Gess, 1980c:
66, Figs 2 and 8) is always downcurved and may continue in a horizontal plane close to the
ground but always free from it.

The nest superstructures of the two Specidae though superficially very different are on
closer examination dearly of similar plan and may readily be homologized (Gess and Gess,
1975: 33): that of Bembecinus cinguliger is elongated and is applied to the surface of the
ground (Gess and Gess, 1975: Figs 6-9 and Pis 5 and 6) whereas that of B. oxydorcus is
shortened and raised above the ground (Gess and Gess, 1975: Figs 2-5 and Pis 7-13).

A feature common to the nest superstructures of the Pompilidae, Eumenidae and Masar-
idae enumerated above is that of radial symmetry, at least in the initial stages of their construc-
tion. Whereas radial symmetry is characteristic of the completed superstructures of all the
Eumenidae and of Jugurtia confusa as well as of large superstructures of Dichragenia neavei, it
is lost in those of the Ceramius species and in large superstructures of D. pulchricoma. The
nest superstructures of the two species of Bembecinus (Sphecidae), however, are never radially
symmetrical even initially.

In Ceramius lichtensteinii at least, it has been established (Gess, 1973: 117) that ex-
perimental removal of a completed nest turret is followed by its replacement by the nesting
w'asp with a new one of similar design to and size as the first. The response by D. pulchricoma
to a damaged nest superstructure is different and has been described in detail by Gess and
Gess (1974: 196).

That the possession of an aerial superstructure surmounting the nest entrance is of surviv-
al value to species excavating their nests in non-friable clayey soils is evident by the fact that all
the species listed as belonging to this category of ground nesters construct such superstruc-
tures. This is all the more striking when it is considered that the species involved belong to
four different families and that the technique of nesting in clay has evolved independently in
the Pompilidae, Eumenidae-Masaridae, and Sphecidae. Ways in which these superstructures
may be of value to the wasps constructing them will be discussed later.

All the species of wasps excavating their nests in the ground exhibit the same behavioural
elements. Differences in the nesting behaviour result from differences in the sequence in which
these behavioural elements are performed and from the repetition in some species of the per-
formance of one or more of the behavioural elements within the sequence.

The basic and most primitive ethologica! type found in nest-excavating wasps is character-
ized by the sequence:

Hunting — Paralysis of the prey (by stinging) — Transportation of the prey — Excavation of

45

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

the nest (consisting of an entrance burrow and a cell) — Oviposition — Closing of the cell

and of the nest.

This basic type is found amongst species excavating their nests in friable soils and is typical of
the vast majority of Pompilidae, examples at Hilton being Batozonellus fuliginosus and prob-
ably Tachypompilus ignitus (Gess and Gess, 1980b). A few Sphecidae such as the noctuid-
caterpillar-hunting Podalonia canescens also exhibit this primitive nesting type.

The wasp after transporting the prey into the vicinity of the chosen nesting site, often
close to the point of prey capture, is forced temporarily to deposit the former on the ground or
in some place of concealment — frequently under a leaf or in a small weed — while a shallow
burrow for its reception is rapidly excavated. Despite the fact that the wasp may frequently
interrupt her digging to check upon the prey, the latter is subject to misadventure and may
either be carried away by other insects such as ants or may be oviposited upon by cleptoparasit-
ic insects such as the pompilid Ceropales.

Excavation of the cell having been completed the wasp retrieves the prey, transports it
into the cell, oviposits upon it and then closes the cell by filling the burrow leading to it with
some of the previously excavated soil.

In the behaviour of the wasps belonging to this basic ethological type characterized by the
fact that hunting precedes nest excavation there is never any repetition of a behavioural ele-
ment within' a single sequence and it therefore follows that not only is a cell provisioned with
only a single prey but each nest consists of only a single cell. With the closure of the cell the
association with the nest is ended and capture of a further prey by the wasp is followed by the
excavation of a new nest.

More specialized ethological types found amongst species excavating their nests in the
ground are all characterized by the fact that hunting and nest excavation have been transposed
in the sequence in which the behavioural elements are performed — that is nest excavation now
precedes hunting. Of immediate benefit to all wasps having this more advanced sequence is
that the prey may be taken directly into the cell with greatly reduced risks of its being stolen or
furnished with an egg by a cleptoparasite.

Nest excavation preceding hunting is the sequence common to the majority of those wasps
recorded at Hilton as excavating their nests in friable soil and is characteristic of all those
species excavating their nests in the non-friable clayey soil.

As has been noted above, wasps which hunt before excavating nests construct these close
to the site of capture of the prey. With the adoption of the new sequence of nest excavation
before hunting, the nest site is not chosen as a matter of expediency but on the basis of its
overall suitability and as a consequence there is a tendency for successive nests excavated by a
wasp to be localized in one area which may moreover be removed from the hunting area.

The grouping by a wasp of successive nests in the same nesting area is clearly pre-adaptive
for a labour and time saving shortcut in which there is a substitution of a smaller number of
nests with two or more cells for a larger number of nests each with a single cell. This change
from single-celled to two-celled or multi-celled nests is brought about within the new sequence
of nest excavation before hunting by a delay in carrying out the final element of the se-
quence— that of nest closure (as opposed to cell closure) and by a repetition after cell closure
of all the preceding elements in the sequence other than the initial one — the excavation of the
entrance burrow.

The amended sequence may be expressed as:

Excavation of the nest entrance burrow — [Excavation of the cell — Hunting — Paralysis of the
prey — Transportation of the prey — Oviposition — Cell closure ]n — Nest closure.

(The portion of the sequence in square brackets is that subject to repetition “n” times.)

At Hilton, examples of wasps belonging to this ethological type are the two pompilids
excavating their nests in clayey soils — Dichragenia pulchricoma and D. neavei (Gess and Gess,

46

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

1974 and 1976b, respectively). In both these species the final nest closure has never been found
to occur and it therefore seems that it is not only delayed but is omitted altogether.

The sequence of nest excavation preceding hunting is pre-adaptive for the substitution in
the provisioning of a cell of several smaller prey for a single large one. This is clearly of
advantage with respect to prey transport, it being easier to transport several small prey, one at
a time, than a single large prey. The advantage can easily be seen with respect to wasps
provisioning their cells with spiders (Table 3). Like all Pompilidae, Batozonellus fuliginosus
and Tachypompilus igniius (Gess and Gess, 1980b) provision each cell with a single spider,
which on account of its weight relative to that of the wasp transporting it — up to around eight
times — cannot be transported other than by being dragged laboriously over the ground. Fre-
quently the wasp has difficulty in overcoming the obstacles on the way to the nest. By contrast,
in the Sphecidae Chalybion tibiale (Gess and Gess, 1980d) and Pison montanum provision
each cell with numerous small spiders, which on account of their light weight — for each indi-
vidual a mere fraction (about one-third) of that of the wasp — are able to be transported effort-
lessly, rapidly and directly to the nest by the wasp in flight.

Table 3

The relationship for some species of Pompilidae and Sphecidae between the weight of the spider
or spiders forming the provision of a cell and the weight of the female wasp provisioning the cell.

Wasp

No. of

No. of

Provision

weight

Provision

weight

Individual
prey wt.

family &
species

Size of
sample

prey/cell

(range)

prey/cell

(average)

wasp wt.
(range)

wasp wt.
(average)

wasp wt.
(average)

POMPILIDAE

Batozonellus fuliginosus

2

1

1

2,25-7,61

4,93

4,93

Tachypompilus ignitus

16

1

1

1,00-8,52

4,2

4,2

SPHECIDAE
Chalybion tibiale

3

15-21

17,4

4,65-6,52

5,87

0,37

Pison montanum

20

7-23

14,5

2, 5-7,0

4,9

0,33

The substitution of several smaller prey for a single large one, like the substitution of
two-celled or multi-celled nests for single-celled nests discussed above, is achieved by repeti-
tion of some of the elements of the behavioural sequence — in this case the elements concerned
with hunting, paralysis of the prey, and transport of the latter to the nest.

It is clear that though the number of prey introduced as provision into a single cell may be
increased from one to several or many, oviposition within the cell must be limited to a single
egg and cannot automatically follow the introduction of each prey into the cell. However,
whereas multiple oviposition is universally suppressed, there is variation in the point in time
during the provisioning of a cell when oviposition of the single egg occurs. Oviposition may be
fixed to be either on the first prey to be introduced into the cell or may be on the last (as in
Cerceris species) or it may not be determined by order of prey introduction at all but rather by
the size or some other particular attribute of one of the prey. This appears to be the case in the
sand-nesting sphecid Kohliella alaris for which size of prey appears to be a factor in the deter-
mination of the prey chosen for oviposition (Gess and Gess, 1980a: 52).

Oviposition upon the first prey to be introduced into a cell may be followed by two dis-
tinct forms of provisioning differing in the time span during which the subsequent provisioning
(that is introduction of additional prey) is performed. Thus, if rapidly performed, the cell may

47

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

be fully provisioned and sealed before the larva hatches from the egg (= mass provisioning).
However, if subsequent provisioning is delayed, the larva may hatch before provisioning is
completed. Whereas this may be the consequence of a shortage of prey or of weather un-
favourable for hunting, some wasps habitually practise delayed provisioning even under opti-
mal conditions. In this case, introduction of the second prey into the cell is delayed until the
young larva has at least partially consumed the first (on which the egg was laid) and subse-
quent prey are similarly introduced only when the larva is ready to consume them (= progres-
sive provisioning). Such is the case in the nesting of Bembix albofasciata , at Hilton excavating
its one-celled nests in sandy soil.

Whereas oviposition in all the ethological types hitherto discussed has been upon the
provision and has thus followed the introduction of one or more prey into the cell, some of the
more specialized wasps exhibit transposition of hunting and oviposition and oviposit into an
empty cell.

This behaviour is uncommon in the Sphecidae but is known in the genus Bembecinus ,
amongst others, and at Hilton has been described for B. cinguliger and B. oxydorcus (Gess
and Gess, 1975). Oviposition into an empty cell is, however, typical of all Vespoidea and
amongst the ground nesting wasps of Hilton has been described for the eumenid Parachilus
insignis (Gess and Gess, 1976a) and for masarids of the genera Jugurtia and Ceramius (Gess
and Gess, 1980c).

Oviposition into an empty cell, like oviposition onto the first prey to be introduced into a
cell, allows either mass or progressive provisioning. Both options are shown by the above
wasps. The two Bembecinus species practise progressive provisioning at least initially (Gess
and Gess, 1975: 38-39) whereas Parachilus insignis practises mass provisioning. Mass pro-
visioning is the method used by the Masaridae studied. The contrary belief held by some
authors in the past that one of these species ( Ceramius lichtensteinii ) practises progressive
provisioning has been exhaustively discussed and shown (Gess and Gess, 1980c: 80-81) to
have been based upon incorrect interpretation of delayed provisioning due to external causes.
In passing it may be mentioned that the Masaridae are unique amongst the wasps in that (with
the exception of one genus) all provision not with insect or spider prey but with a mixture of
pollen and nectar. In this facet of their behaviour the Masaridae therefore parallel the bees.

The nesting of the solitary bees recorded excavating their nests in the ground: Colletes
(Colletidae), Halictus, Lasioglossum, Nomioides and Nomia (Halictidae) is similar in many
respects to that of the more specialized wasps and follows a sequence of behaviour found also
amongst the latter. Nesting is initiated by the excavation of an entrance burrow and of a cell.
Provisioning of the cell (with a mixture of pollen and nectar) follows and is completed before
oviposition on or near the provision takes place; thereafter the cell is sealed and the next cell is
excavated. It will be noted that the relative order of provisioning and ovipositioning in these
bees and in the pollen and nectar provisioning wasps (Masaridae) is reversed.

When the nesting of the relevant species of wasps recorded at Hilton is considered with
respect to the foregoing account of the ethological types represented in the nesting of the wasps
and bees excavating original nests in the soil, an interesting distinction between those species
associated with friable and non-friable soils may be discerned. Whereas a wide range of etholo-
gical types from the most basic (hunting before nest excavation) to the most advanced is found
amongst those species associated with friable soils, only the more advanced types are found
amongst those species associated with non-friable soils. Illustrative of this is that in all the
species associated with non-friable soils hunting is preceded by nest excavation and that two-
celled or multi-celled nests are the rule (though Bembecinus and Parachilus may occasionally
fail to excavate the second cell). Furthermore that in all the species, with the exception of the
two Dichragenia species, provisioning of each cell is with numerous prey and that oviposition
in the empty cell is practised.

48

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

It must therefore be postulated that, amongst the wasps and bees excavating their nests in
the soil, the association with friable soil is the original one and is primitive and that the asso-
ciation with non-friable soil is secondary and is derived and advanced.

Certainly, the difficulties attendant upon the excavation of nests in non-friable soil pre-
cludes those species doing so from possessing the sequence of behavioural elements (hunting
before nest excavation) characteristic of the basic or primitive ethological nesting type. On the
other hand, the more advanced ethological nesting types seen amongst the species nesting in
friable soil are pre-adaptive in overcoming the difficulties imposed by nesting in non-friable
soil.

With respect to the relative sequence in which hunting and nest excavation are performed,
the primitive sequence (hunting preceding nest excavation) is possible only if the delay in the
introduction into the nest of the prey, already at hand and waiting, can be kept to a minimum.
This is possible if the nest is being excavated in friable soil in which digging is easy and there-
fore rapid. However, in non-friable soil where excavation is difficult and therefore slow and
furthermore involves the periodic absence of the wasp from the site when fetching water the
prey would be exposed for a longer period which would increase the possibility of desiccation,
theft or parasitism. Nest excavation preceding hunting is therefore a prerequisite for nesting in
non-friable soils.

The change from single-celled to two-celled or multi-celled nests sharing a single entrance
burrow common to both or all the cells, though in terms of saved energy and time advan-
tageous to nesters in friable soils, is of even greater advantage to nesters in the more difficult
non-friable soils, and must be considered pre-adaptive to nesting in the latter substrate.

Similarly, the substitution in the provisioning of each cell of several smaller prey for a
single large one, besides being of importance with respect to ease of transport, must be seen as
bringing about a saving in energy and time expended in nest excavation in that the bore of the
entrance burrow and the overall size of the cell can thereby be reduced. This again is of
greatest advantage to excavators in non-friable soil and can be seen as pre-adaptive to nesting
in this substrate.

A comparative reduction in the bore of the entrance burrow and in the size of the excav-
ated cell is achieved also by the two species of Dichragenia despite the fact that, like in all
Pompilidae, provisioning of each cell is with but a single large spider. In this case reduction of
the dimensions of the nest is made possible by the amputation of the legs of the prey. This
habit of amputation of the legs of the prey, characteristic of the pompilid tribe Auplopodini,
like the change seen in more highly evolved wasps from a single large prey to several smaller
ones per provisioned cell, probably serves the primary purpose of greater ease of prey trans-
port but is also pre-adaptive for nesting in non-friable soils.

A prime requisite for the nesting of all nest-excavating wasps, whether nesting in friable
or non-friable soils, is the presence of areas of soil partially or totally denuded of plant cover.
Thus, at Hilton the most favoured areas for ground-nesting are those which have been modi-
fied by man’s activities and take the form of paths, car-tracks, a sand-pit, and tracts once
denuded of their original plant cover (by ploughing and cultivation) and subsequently only
partially covered by mostly pioneer species.

Over the decade during which the nesting of wasps and bees has been studied at Hilton, it
has been noticeable that available nesting sites both in friable and in non-friable soils have
shrunk due to an increase of plant cover, a trend which, if it continues, will eventually lead to
a marked reduction in the size of the nesting communities.

It is obvious that the nest-excavating species associated with friable, sandy soil are able to
excavate their nests in this substrate where ever bare patches of it occur within a given area.
Species associated with non-friable, clayey soil, however, have no such freedom but are res-
tricted to bare patches of this substrate in those parts of a given area that lie within a certain

49

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

maximum distance of a source of water required for soil softening. This distance, which may
vary from one species to another, is determined by how far it is economical, in terms of
expenditure of energy and time, for a wasp to have to fly to collect this water. In effect, it has
been observed at Hilton to limit the nesting areas of the species concerned to tracts of clayey
soil in fairly close proximity to water sources (Gess and Gess, 1974: 192; 1975: 24; 1976a: 86;
1976b: 131; 1980c: 64). Other tracts of bare clayey soil, apparently identical in all respects to
the above but further removed from water sources are notable for the absence of any nesting
wasps.

The water sources most commonly used at Hilton by excavators in non-friable clayey soils
are temporary in nature and consist of pools formed after rain in small depressions, in erosion
gullies and in a man-made furrow flanking the chief nesting area (Gess and Gess, 1974: Pis
1-4). Infrequent use is made of small puddles left in the river bed but the larger and more
permanent water bodies, the dams, are unused. It follows therefore that nesting is dependent
upon the occurrence of rain and of run-off filled water holes and that it can continue only for
as long as the water supply lasts. In contrast therefore to those wasps excavating their nests in
friable sandy soil, which as a community may be found in any one year to nest continuously
over a long period, from early spring to late autumn, the wasps excavating their nests in
non-friable clayey soil and requiring water for nest excavation are as a community in any one
year frequently restricted to a very much shorter nesting period, or have an interrupted nesting
period, determined by the time and abundance of rainfall and by the availability of water in
temporary pools. Though these wasps individually and as a community have the potential for a
long nesting period, like that seen in the sand nesters, it is frequently not realized and in
several summer seasons it has been seen to have been arrested at or near its height by the
failure of the water supply. This is particularly noticeable with respect to Bembecinus cinguli-
ger.

In the light of the physical difficulties attendant upon the excavation of nests in non-friable
clayey soils and the spatial and temporal restrictions imposed upon nesting in such soils due to
the requirement that water be available to the wasps where and when nesting is undertaken, it
must be asked what led certain groups of wasps to leave a friable substrate in favour of a
non-friable one.

The answer must be postulated to lie on the one hand in a response to the pressure of
interspecific competition for available nesting sites and prey with respect to one substrate and
on the other hand in a response to the existence of another, unexploited substrate with its
associated community of unexploited potential prey species.

The ability of certain wasps to adapt to excavating nests in the previously unexploited
non-friable clayey soils allowed the spread of ground-nesting wasps into tracts of country which
had such soils and from which the latter had hitherto been debarred. Moreover, in situations
as at Hilton where both soil types occur in close proximity a given area including both soil
types can support a greater number of soil-excavating species than could the same area if only
one soil type were included.

The presence at Hilton of both soil types must be seen as the key to the co-existence there
of large nesting populations of four species of Bembecinus , two, B. cinguliger and B. oxydor-
cus , associated with non-friable soil and two, B. braunsii and B. haemorrhoidalis , associated
with friable soils. Whereas it is immediately apparent that there is no competition between the
two pairs of species for nesting sites, there is likewise no competition for prey. Examination of
the prey recovered from the cells of the four wasps shows that whereas provisioning by all the
species is with numerous species of (mostly) Cicadellidae and that both nymphs and adults,
males and females, large and small individuals are utilized, the specific composition of the prey
of the two pairs of Bembecinus species is completely different though within a pair of species it
is similar. This situation is a reflection of the observed fact that the two soil types support

50

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

different plant communities which in turn have feeding upon them different complexes of
Cicadellidae and furthermore that the Bembecinus species do not hunt at any great distance
from their nests and therefore exploit those prey found on vegetation growing on the soil type
to which they (the wasps) are themselves restricted.

Whereas the use for nest excavation of non-friable soils appears to have been adopted
only very rarely and sporadically by the Pompilidae and the Sphecidae, it is typical of almost
all the ground-nesting Eumenidae and of all the ground-nesting species of the allied Masar-
idae. This has in effect freed the Vespoidea from competing with the vast majority of the
Pompilidae and Sphecidae for nesting sites in the ground.

Of interest are those Eumenidae which, atypically, do excavate their nests in friable,
sandy soil — species of Pterocheilus , divided by Bohart (1940) into a number of subgenera. These
sand-nesting Eumenidae, unlike the sand-nesting Pompilidae and Specidae, do not possess
sandrakes on the fore-legs, which limbs do not differ in structure from those of the clay-nesting
and water-utilizing Parachilus insignis. Instead, for the removal from the burrow of the sandy
material loosened by the mandibles, the wasp possesses a “sand basket" formed by long hairs
fringing the outer edges of the mandibles and by others fringing the second and third segments
of the long pendulous labial palps (Evans and Eberhard, 1970: Fig. 55).

It must be postulated that the Vespoidea evolved from forms associated with friable soils,
that at an early stage in their evolution they forsook nesting in that soil type in favour of
nesting in non-friable soil and that forms such as Pterocheilus now found nesting in friable soil
secondarily returned to that substrate. Flowever, the Eumenidae having as a group earlier lost
any fore-tarsal digging organs, removal of excavation spoils by the sand-nesting species is by
means of the mouthparts as in the pellet-carrying, water-utilizing, clay-nesting species from
which the sand-nesters are probably derived.

It has been shown above that the substitution in the provisioning of each cell of several
smaller prey for a single large one was advantageous in terms of greater ease of transport of
the prey and greater economy of labour consequent upon the construction of nests of smaller
dimensions. However, a disadvantage introduced with the substitution is that cell closure has
to be delayed until after the last of several or many prey has been placed in the cell and that
the partially provisioned cell is left unattended by the wasp during the time the latter is away
hunting for additional prey. During this time, the open, partially provisioned cell is in danger
of attention by parasites of one sort or another unless some means are employed to exclude
them.

This exclusion of at least a certain proportion of parasites is effected by many of the
species excavating their nests within friable soil by the maintenance of a temporary closure at
the nest entrance. In the present study, this is described for the sphecid Kohliella alaris (Gess
and Gess, 1980a: 52) which rakes sand respectively into or out of the nest entrance each time it
leaves or returns to the nest.

So easy and quick a way of closing and opening a nest is clearly possible only in friable soil
and it must be asked how those wasps nesting in non-friable soil have solved the matter of the
exclusion of parasites. The answer may be sought in a comparative study of certain aspects of
the nesting of a genus including both sand- and clay-nesters — in the present instance the genus
Bembecinus.

All four species of Bembecinus studied at Hilton construct temporary nest closures but the
frequency with which these are fashioned differs between the sand-nesting and clay-nesting
species.

In the sand-nesting B. braunsii and B. haemorrhoidalis temporary nest closures are main-
tained throughout the wasps' working day whenever the wasps are not within their nests. In
the clay-nesting B. cinguliger and B. oxydorcus, however, temporary closures, in the form of
mud plugs sited in the entrance shafts (and in B. oxydorcus at the turret opening as well) (Gess

51

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

and Gess, 1975: Figs 11 and 16), are constructed only at the end of each working day and the
first action by the wasps on returning to the nesting sites at the beginning of the following
working day is to remove these mud plugs. Therefore, whereas the nests of B. braunsii and
B. haemorrhoidalis are open only for short periods during which the wasps are introducing
prey, those of B. cinguliger and B. oxydorcus are open for the full duration of the wasps’
working day, whether the wasps are present in the nests or not. As it is not an unreasonable
assumption that parasite pressure on Bembecinus should be similar in the two nesting sub-
strates— indeed at Hilton species of Mutillidae, notably Smicromyrme Hecuba, are commonly
found associated with all four species — the lack of a temporary closure during the working day
of B. cinguliger and B. oxydorcus must be compensated for by some other factor. In the
absence at the time of cell provisioning of any other behavioural difference between the two
pairs of species, the factor can only be an additional difference in the physical nature of the
nest, namely the presence in those species omitting the temporary closure of the mud super-
structures surmounting the nest entrances. The hypothesis is therefore put forward that in
Bembecinus at least, the mud superstructures serve as a protection against the entry of para-
sites into the nest.

Possibly a similar role may be assigned to the nest superstructures built by at least some of
the other species. Oldroyd (1964: 132) has pointed out that Bombyliidae seen hovering close to
the ground are not always looking for flowers but may be actively egg laying, the egg being
dropped near the burrow of a solitary bee or wasp. Furthermore, Painter (1932, as reported by
Clausen, 1940: 377) recorded experiments in inducing oviposition by a hombyliid. Villa sp.,
which develops in the cells of solitary bees. The eggs were reported to be “readily projected
into glass vials buried in the soil to simulate the nest openings of the host”. It would appear
that the stimulus to oviposit by the hovering fly is visual and consists of the sight of a small,
dark, round hole in the ground. It does not seem unreasonable therefore to suggest that the
removal of the stimulus to oviposit by the concealment of the dark hole in the ground would
be of survival value to those wasps which leave their nest entrances open — that is, those wasps
that nest in clayey soils. In many of these wasps concealment from above of the burrow en-
trances is effectively achieved by covering them with curved turrets and it is therefore post-
ulated that these superstructures are a defence against parasitization by Bombyliidae. It is
perhaps significant that the only bombyliid larva found during the course of the excavation of
very many nests of a variety of species nesting in clayey soils was in a cell of Parachilus insignis
(Gess and Gess, 1976a: 97), a species which surmounts its nest entrance not with a curved
turret but with a low collar which does not conceal the “dark, round hole”.

In addition to the probable value of at least some nest superstructures in excluding certain
categories of parasites, it is certain that all superstructures, even those consisting only of a low
collar surrounding the nest entrance serve to prevent loose, powdery clay soil and other fine
wind-blown debris from entering the nest shafts. This may well be of value to wasps which lack
both the physical equipment (tarsal sand rakes) and the behaviour requisite for the removal of
fine, loose material from their burrows.

Finally, the more extensive of the superstructures covering nest entrances serve to protect
the nests against flooding for upon being well wetted by rain or surface run-off the superstruc-
tures collapse upon the nest entrances, effectively blocking them. The nest collar of Parachilus
insignis is clearly too small to serve this function. However, as described in detail by Gess and
Gess (1976a: 99) the wasp nevertheless protects its nest from flooding by blocking the nest
shaft with its own body and by allowing a plug of water-borne mud to form above it.

Among the 167 species of ground-nesting wasps and bees recorded during the survey at
Hilton, thirteen species were found to nest not in burrows of their own excavation but in
pre-existing cavities which they, however, modify by the construction within them of a cell or
cells. The thirteen species concerned are listed in the annotated list of species under categories

52

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

2 and 5. Four families were represented: Eumenidae (1 species), Sphecidae (6 species),
Megachilidae (5 species) and Anthophoridae (1 species).

A necessary pre-condition for the nesting of these species is the presence in the ground of
pre-existing cavities of a bore suited to their individual requirements. The pre-existing cavities
most often used are burrows excavated for nesting purposes by other, nest excavating, acule-
ates. Usually these burrows are no longer being used by their excavators but are newly excav-
ated burrows that have been abandoned for some reason or another (such as the death of the
builder) or are old burrows, excavated in a previous nesting season, from which the wasps that
developed within them have emerged. There is very little evidence that competition for bur-
rows occurs between the original burrow excavators and the users of pre-existing cavities — that
is, burrows do not appear to be usurped if still in use.

It will be seen from Table 2 and from the Pie-diagram (Fig. 16) that, whereas in non-
friable soil the number of species nesting in modified pre-existing cavities approximates the
number of species excavating original nests, in friable soil the number nesting in modified
pre-existing cavities is very low indeed. This is all the more striking on account of the far
greater number of species excavating original nests in friable soil than in non-friable soil.

The reason for the difference which is clearly substrate-determined is probably due largely
to the fact that a burrow excavated in non-friable clayey soil is stable and long-lasting whereas
one excavated in friable sandy soil tends if it is not maintained to collapse or fill with loose
material after even a short time. This is of particular relevance to old burrows from which
wasps or bees have emerged. With respect to newly excavated but abandoned burrow's, it must
be remembered that wasps excavating in friable soil frequently maintain a temporary closure at
the nest entrance which would render such burrows inaccessible to seekers of pre-existing
cavities even when the original “owners” of such nests had suffered some misadventure during
hunting or foraging. The opposite w'ould pertain to burrows excavated in non-friable soil parti-
cularly where the clay next superstructures are incomplete or broken.

It appears that the thirteen species recorded as making use of pre-existing cavities (bur-
rows) in the ground may be divided into those for which such behaviour is obligatory and those
for which it is facultative.

Among the former are the five megachilid bees belonging to the genera Megashile and
Creightoniella. One of these bees. Megachile semiflava, has been recorded as making use of
pre-existing burrows in both the non-friable clayey soil and the friable sandy soil and there
seems no reason why the other species should not show a similar lack of restriction to a par-
ticular soil type. All five species construct their cells within the pre-existing cavities with
foreign materials — green leaves or in the case of M. meadewaldoi the petals of low-growing
flowers.

Other species for which nesting in pre-existing cavities is obligatory are the sphecid Pison
allonymum and an unidentified eumenid known only by its turret which it had constructed
surmounting that of Bembecinus oxydorcus, the burrow of which species it was using. Both
species are probably derived from forms which used water to excavate nests in non-friable
soils. Though the ability to excavate nests has been lost, both species remain associated with
clayey soils due to their use of mud in modifying the burrows in which they nest.

The nesting of all the above species involves the modification of the chosen pre-existing
cavities by the introduction into them of foreign materials for cell construction prior to the
gathering and introduction of provision.

Very primitive in comparison is the presumed nesting behaviour of Ampulex mutilloides
which though not actually found nesting is included in the present category on considerable
circumstantial evidence.

Ferruginous and metallic blue and thus atypical in colouration, this Ampulex looks very
mutillid-like (as attested by its specific epithet), a similarity that is enhanced by the fact that in

53

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

the experience of those who have seen the species in the field (the present author and the late
C. F. Jacot-Guillarmod — pers. com.) it is only met with on the ground. The fact that the
Hilton specimen, a freshly emerged female, had clayey soil adhering to her head strengthens
the view that this species nests in the ground. Probable prey is Pilema thoracica , a cockroach
excavating burrows in clayey soils in the area in which the wasp was found walking on the
ground.

In common with other species of Ampulex , nesting would be commenced with hunting
which would be followed by the introduction of the single prey into an unmodified pre-existing
cavity which would then be closed with detritus. If the above interpretation of the nesting of
A. mutilloides is correct, this wasp would be restricted to nesting in clayey soils as it is there
that its prey occurs. Furthermore, it is possible that the pre-existing cavities used by the wasp
are the burrows excavated in the soil by the cockroach.

Species for which nesting in pre-existing cavities in the ground appears to be facultative
and determined by the nature of the substrate are Prionyx kirbii, Tachysphex sp. near mo-
destus and possibly Tachysphex modestus itself (all Sphecidae) and Tetralonia minuta
(Anthophoridae).

Prionyx kirbii and Tachysphex sp. near modestus both have well developed tarsal sand
rakes and may therefore be expected to excavate original burrows in friable soil. This indeed is
the case with the common P. kirbii which has been observed at Hilton to excavate original
burrows in the sandpit. Tetralonia minuta also is probably capable of excavating original bur-
rows. Though this is not actually stated it may be inferred from the account of the nesting of
this bee given by Rozen (1969a).

What are these species doing then in pre-existing cavities? The only explanation is that
these species have some plasticity of behaviour and that, in non-friable soil, burrow excavation
is intitiated not at the surface of the soil but at the bottom of a pre-existing cavity: the old or
abandoned burrows of Parachilus insignis in the case of P. kirbii and T. sp. near modestus', the
old or abandoned burrows of Parachilus insignis or of Bembecinus cinguliger in the case of
T. minuta. In some instances at least (as with P. kirbii and T. sp. near modestus ) the pre-
existing cavity may already approximate in dimensions the burrow excavated in friable soil so
that little additional excavation need be done. With respect to T. minuta which appears to
make deeper and more complex nests than do the two sphecids it is probable that the initial
use of a pre-existing cavity enables the bee to start its own excavation at a depth at which the
soil may be moister and consequently softer and where it is therefore easier to excavate.

In all cases, the use of pre-existing cavities as a starting point for nest excavation allows
these species to extend their distribution into areas of non-friable soil where nesting in the
manner normal to them would be precluded or would at best be very difficult. That the use of
pre-existing cavities by these species is simply a short cut in nest excavation adopted in hard
clayey soil and does not represent any radical change of behaviour is shown by the fact that
both P. kirbii and T. sp. near modestus do remove considerable quantities of soil from the
bottom of the pre-existing cavity. P. kirbii arranges the material removed from the cavity in a
circle around the burrow opening (Fig. 24); T. sp. near modestus places it only to one side of
the opening (Fig. 25). Both species prepare the burrow before hunting and P. kirbii carefully
positions three or four small clods of earth as a temporary closure at the entrance of the
modified pre-existing cavity before she leaves to hunt.

Sixty-four of the 167 species of ground-nesting wasps and bees recorded during the survey
at Hilton are listed, under categories 3 and 6, as developing within pre-existing cavities which
they do not themselves modify. Ten species were recorded from non-friable soils, 50 species
were recorded from friable soils and four species, all Mutillidae, were recorded from both
friable and non-friable soils. It must be pointed out, however, that the above figures pertaining
to the two soil types are not directly comparable for, whereas all 14 species listed for non-

54

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

St »

Fig. 24

Fig. 25

Figs 24 and 25. Entrances of old burrows of Parachilus insignis used for nesting by Prionyx kirbii (Fig. 24) and by
Tachysphex sp. near modestus (Fig. 25), showing arrangement of material removed from the pre-existing cavities by

these wasps, (x 0,7)

55

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

friable soils are “parasitic” in one way or another in what are to them pre-existing cavities —
cells constructed and provisioned by other wasps and bees and occupied by the eggs or imma-
ture stages of these species, this is true of only 26 of the 54 species listed for friable soils. The
remaining 28 species listed for friable soils are associated with the larvae of Coleoptera, with
one exception (Methocha mosutoana) those of Scarabaeidae, which are sought for and attack-
ed in their burrows, often deep in the ground, where, after they have been oviposited upon,
they are usually left in situ.

Those species which are “parasitic” in the cells of other aculeate Hymenoptera and are
therefore dependent upon the occurrence of the latter for their own presence belong primarily
to the families Chrysididae (12 species) and Mutillidae (17 species). A few other families are
represented as well, namely Pompilidae (Ceropales punctulatus) , Sphecidae (Nysson braunsi),
Halictidae ( Sphedes sp.), Megachilidae (Coelioxys bruneipes and C. lativentroides) and
Anthophoridae ( Epeolus amabilis and Thy reus sp.).

Table 4.

Secondarily cleptoparasitic ground-nesting species and their established or presumed hosts. (For

authorities see annotated list.)

CLEPTOPARASITE

HOST

Genus & Species

Family

Genus & Species

Family

Ceropales punctulatus

Pompilidae

Dichragenia pulchri-
coma

Pompilidae

Nysson braunsi

Sphecidae

Nyssoninae

Oryttus kraepelini
OR

Dienoplus vicarius
karrooensis

OR

Hoplisoides aglaia
OR

Hoplisoides thalia

Sphecidae

Nyssoninae

Specidae

Nyssoninae

Sphecidae

Nyssoninae

Sphecidae

Nyssoninae

Sphecodes sp.

Halictidae

Nomia sp.

Halictidae

Coelioxys bruneipes

Megachilidae

Megachile semiflava

Megachilidae

Coelioxys lativentroides

Megachilidae

Megachile sp.

OR

Creightoniella sp.

Megachilidae

Megachilidae

Epeolus amabilis

Anthophoridae

Colletes sp.

Colletidae

Thy reus sp.

Anthophoridae

?

?

56

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Ethologically the species belonging to families other than Chrysididae and Mutillidae are
possibly the most interesting for all are derived from non-parasitic forms and the cleptoparasit-
ism now practised by them is therefore secondary or derived. As may be seen from Table 4
most of the species are related to their hosts and in each case host and cleptoparasite must
therefore have a common origin.

In all cases the cleptoparasitic way of nesting has been arrived at by a reduction in the
number of elements in the behavioural sequence performed by the parasite, the “missing”
elements being supplied by the host. For example, the pompiiid Ceropales punctulatus , re-
corded as cleptoparasitic in the nests of Dichragenia pulchricoma (Gess & Gess, 1974: 202,
204) and assumed to behave like other species of its genus, restricts its nesting activity to
hunting and oviposition. Hunting, however, is not for a free active spider but for one already
captured and stung by the host pompiiid and oviposition (inside the booklungs) takes place
before the host has installed the spider in its cell. There the Ceropales egg hatches before that
of its host and the young foreign larva after destroying the latter feeds upon the spider. In the
case of Ceropales most of the labour involved with nesting is therefore supplied by its host,
Dichragenia , namely construction of the cell, hunting of the spider prey, stinging of the prey,
transport of the prey to the cell, and closure of the cell.

The cleptoparasitic bees behave similarly in so far that all the labour involved in nest
construction and provisioning is supplied by their hosts. Concomitant with the adoption of a
cleptoparasitic way of life has been the secondary loss of the pollen-collecting apparatus and
therefore the ability to provision cells. The search for cells already provisioned with a mixture
of pollen and nectar by the host bees therefore replaces the search for flowers from which to
obtain these substances and flower-visiting by cleptoparasitic bees is therefore restricted to the
purpose of obtaining their own nutriment.

In contrast to the above discussed species, the species of Chrysididae and Mutillidae, all of
which are “parasitic”, are not related to their hosts.

The Chrysididae show similarities in their ethology to the cleptoparasitic Pompilidae,
Sphecidae and Apoidea in so far that the majority of species develop on the provision stored
by their hosts, the egg being introduced into the cell while this is being provisioned. Such
appears to be the case with respect to Octochrysis vansoni which was recorded (Gess & Gess,
1976a: 97) as attending the open nests of Parachilus insignis , inspecting these when they were
left unguarded and occasionally entering them, presumably in order to oviposit in the cells. In
behaviour O. vansoni is therefore clearly cleptoparasitic.

The Mutillidae on the other hand do not develop on the stored provision but limit their
attack to fully fed, diapausing larvae or pupae, the female mutillid breaking into a cell contain-
ing a cocoon into which she introduces her egg. Feeding upon the host within its cocoon is
followed by the mutillid larva spinning its own cocoon within that of its host, as recorded (Gess
& Gess, 1980c: 76) for Dasylabroides caffra in the cells of Ceramius lichtensteinii. In behaviour
D. caffra, like other Mutillidae, is therefore clearly a parasitoid. This term has been used
(Evans & Eberhard, 1970: 9) with respect to a species which cannot be considered a true
parasite (because it kills its host) nor a true predator (because it is confined to a single prey
individual).

The remaining 28 species listed as developing within pre-existing cavities which they do
not themselves modify are species of Tiphiidae and Scoliidae. As already stated all are associ-
ated with the larvae of Coleoptera, the majority with those of Scarabaeidae. All were found in
association with friable soil for it is only in such soil that the beetle larvae and the female
wasps hunting them are able to dig. In no case was the nesting investigated and only one
species was associated with its prey: Methocha mosutoana (Tiphiidae) with Cicindela brevicollis
(Cicindelidae).

57

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

Vertical banks

The vertical banks studied at Hilton are situated at the sandpit and in places along the
course of the New Year’s River. At the former locality the banks are man-made and constitute
the actual walls of the sandpit whereas in the other localities they have been cut by the river
and constitute the banks thereof. Mostly the latter are situated at a height above the river bed
reached by water only during times of flood.

The banks of the sandpit and a section of the river bank near the confluence of the Iron
Put River with the New Year’s River are composed of firmly compacted very fine sand. Going
downstream along the course of the latter river the physical nature of the banks changes,
increasing amounts of clay being mixed with the sand. In places the mixture of sand and clay
has apparently been affected by recent mineralization and has assumed the character of a weak
and crumbly sandstone-like material. Still further downstream the western bank of the river is
formed by the face of a low cliff cutting across roughly horizontally bedded shale.

Though the sand and shale banks clearly constitute very different substrates, those banks
composed of mixtures of sand and clay and including both mineralized and unmineralized
portions, cannot, on account of their variability, be placed in a single class. It seems preferable
therefore not to attempt to further classify the banks on the basis of the physical nature of the
materials of which they are composed but to deal with all banks together. However, where it is
relevant, attention will be drawn to substrate-determined differences in the nesting of the
wasps and bees concerned.

Wasps and bees which nest in association with vertical banks may be divided into those
which nest within the banks and those which nest on the banks and use the latter solely as a
raised support to which to attach their aerial nests.

At Hilton the number of species recorded as nesting in association with vertical banks
totals 51, made up of 43 species nesting within the banks, 6 nesting in aerial nests on the banks
and 2 nesting both in the banks and in aerial nests on the banks. On account of the latter
species the number of species/substrate associations exceeds the actual number of species by 2
and therefore stands at 53. An analysis of this figure according to the degree of participation of
species of wasps and of bees in the construction of their nests is given in Table 5 and is shown
graphically in the Pie-diagram (Fig. 26).

Table 5.

Number of species of aculeate wasps and of bees nesting in or on vertical banks.

In nests con-
structed entirely

In modified

In unmodified

by the nesters

pre-existing

pre-existing

themselves

cavities

cavities

In vertical banks

4

20

21

On vertical banks

5

1

2

With regard to the nesting of wasps and bees a vertical bank resembles level ground with
respect to the nature of the substrate which it offers but differs with respect to the angle at
which this substrate is presented.

58

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Fig. 26. Diagram showing the structure of the communities of the aculeate wasps and the bees nesting in or on vertical
banks in the study area. Numbers accompanying the family names represent numbers of species. The numbers (i, ii, iii)
in the outermost ring correspond to the same numbers in the classification of the aculeate wasps and the bees of the
study area on the basis of their ethology, (i) In nest constructed entirely by the nester. (ii) In pre-existing cavity modified
by the nester. (iii) In pre-existing cavity not modified by the nester.

The difference in the angle of presentation of the soil surface — vertical as opposed to
horizontal — has a very profound influence upon nesting. Illustrative of this is that, with the
possible exception of Pison allonymum, there is no sharing of species between level ground
and vertical banks. Indeed, both nesting situations are characterized by communities of species

59

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

unique to themselves. It is furthermore very noticeable that in the two nesting situations there
is a marked difference between the proportion of species excavating their nests and those
modifying pre-existing cavities. In level ground as has already been shown, nest excavators
greatly outnumber nest modifiers (91 as against 14) whereas in vertical banks the opposite is
true, nest excavators being outnumbered by nest modifiers (4 as against 20).

In view of the great dissimilarity between the communities nesting in vertical banks and in
level ground, it is of interest that some points of similarity exist between the former commun-
ity and that nesting within plant tissue, at least with respect to those species which do not
excavate or hollow out their nests themselves but modify pre-existing cavities. Thus at Hilton
at least seven species nesting in pre-existing cavities are shared between vertical banks and
plant tissue, namely Euodynerus euryspilus (Eumenidae), Chalybion (Chalybion) tibiale, Pison
montanum and Trypoxylon sp. (Sphecidae), and Immanthidium junodi, Megachile gratiosa and
Megachile spinarum (Megachilidae). Rynchium marginellum sabulosum (Eumenidae) at Hilton
found nesting in trap-nests inserted into crevices in the “sandstone” bank but known in Gra-
hamstown to nest in cut culms of the reed Arundo donax used for vegetable frames may
probably be added to the above listed species.

Furthermore, the preponderance of nest-modifiers over nest excavators characteristic of
the vertical banks is evident also with respect to plant tissue.

Characteristic of the community nesting within the vertical banks studied is the dominant
role played by bees, both in terms of number of species and number of individuals. Thus, of
the four species of wasps and bees listed as excavating their nests in the vertical banks, only
the two bees, Hoplitis anthodemnion (Megachilidae) and Anthophora sp. (Anthophoridae) are
of any real importance. Common to the degree of being by far the most noticeable species,
their presence is fundamental to the existance of a large part of the rest of the community as a
whole, for it is their abandoned galleries that constitute the pre-existing cavities which are
sought for and modified by a much larger number of diverse species of wasps and bees which
do not excavate their nests themselves (Fig. 27). Furthermore, among the latter cavity-

Fig. 27. Small portion of the surface of the “sandstone” bank showing the abundance of old nesting galleries available to

nesters in pre-existing cavities.

60

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

modifying species, twenty in number, the Megachilidae with seven species is the best repre-
sented family. Involved are Anthidiellum kimberleyanum, Branthidium braunsii, Chalicodoma
schulthessi, Heriades ?freygessneri, Immanthidium junodi, Megachile gratiosa and M. spinar-
um. Though in terms of numbers almost equally by the Eumenidae the bee family nevertheless
considerably surpasses the latter in number of individuals. Bees are well represented also
amongst those species which nest parasitically in the nests of other species, five Megachilidae,
namely Euaspis abdominalis, Coelioxys lucidicauda, C. bruneipes, C. lativentris and
C. lativentroides , and one species of Anthophoridae, Thyreus sp., being listed.

Taken altogether, bees represent 33% (15 out of 45) of the total number of species listed
as nesting within vertical banks. Comparative figures for nesters within level ground and with-
in plant tissue are 10% (17 out of 167) and 46% (18 out of 39) respectively. It may be seen
that, gauged with respect to the relative numerical representation of species of bees and wasps
in the different nesting substrates, vertical banks once again appear to have more in common
with plant tissue than with level ground.

With regard to the different substrates offered by vertical banks at Hilton, that consisting
of recently mineralized mixtures of sand and clay and referred to as “sandstone” supports the
largest community, both in terms of number of species and number of individuals. This is on
account of the fact that it is in this sandstone-like material that the principal nest excavators,
Anthophora sp. and Hoplitis anthodemnion, are most active; indeed H. anthodemnion appears
to be restricted to this substrate. Anthophora sp. on the other hand also nests in sand banks
but not as commonly. Clearly no nest-excavators are able to nest in the shale bank.

Whereas many of the wasps and bees nesting in modified pre-existing cavities are re-
stricted to the sand and in particular to the “sandstone” banks, probably on account of their
dependence upon the presence of old abandoned galleries of the above two nest-excavating
bees, a few species have in addition colonized the shale bank where they make use of pre-
existing cavities in the form of open cracks between and within individual layers of the shale.
Such is the case with respect to Eumenid A and Tricarinodynerus guerinii (both Eumenidae)
and probably also Megachile gratiosa. At Hilton T. guernii has been found nesting in vertical
banks presenting all three major substrates and is possibly the most characteristic species of
vertical banks (Fig. 28). Its nest which is furnished with a downturned mud turret is invariably
sited in a protected situation such as under an overhang where it is protected from the weather
and particularly from rain.

Protected situations are sought also by those species which nest not within banks but
construct aerial nests upon them, as was demonstrated by the location of recorded nests: a
naked comb made of “wasp paper” by Ropalidia sp. A. was sited in a fist-sized cavity in a sand
bank; a mud nest of Sceliphron spirifex was built in a similarly sized cavity in the “sandstone”
bank; and urn-shaped mud cells of Eumenes lucasius were found attached to the face of the
shale bank under overhangs.

On stones lying upon the ground

At Hilton, stones lying upon the ground were found to represent the least used of all four
nesting situations with only four of the 241 species included in the annotated list associated
with it. It is clear therefore that this nesting situation is of very little importance with respect to
the nesting of the community as a whole. However, it must be taken into consideration that all
nests associated with stones are aerial in nature and that a true assessment of the status of this
nesting situation can therefore only be arrived at by comparison with the status of aerial nest-
ing associated with other nesting situations. Thus it may be seen that the number of species (2)
constructing original nests on stones though smaller is nevertheless comparable to the numbers
of species constructing original nests on vertical banks (5) and on plants (5).

61

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

Fig. 28. Section through “sandstone" bank showing pre-existing cavity modified for nesting and furnished with a down-
turned mud turret by Tricarinodynerus guerinii. ( X 1,6)

The two species constructing original nests on stones differ in the placement of their cells.
Whereas Auplopus rossi (Pompilidae) builds its cells in a sheltered position on the underside
of large stones lying loose upon the ground (Fig. 10), Hoplitis jansei (Megachilidae) builds its
cells fully exposed on the upper surface of partially buried stones and boulders (Fig. 11). In the
case of the former species a sheltered position for the placement of the cells is essential as
these are made of dried mud, are thin-walled and fragile, and would consequently melt and
collapse if exposed to rain. The cells of Hoplitis , by contrast, are made of plant resin in which
are set numerous small pebbles and are consequently waterproof and of great mechanical
strength.

The cells of Hoplitis being fully exposed on the upper surface of stones and boulders, are
clearly potentially subject to extremes of temperature, both high and low, which might well be
detrimental to the bee young within them. There may therefore be a good reason with respect
to survival why the bee builds only on those stones which are partially buried in the ground for
these stones by virtue of the fact that heat can be exchanged between them and the ground
would not become as hot or as cold as similar stones lying loose upon the ground with as a
consequence a reduced capacity for heat exchange.

The nesting of Auplopus rossi in aerial mud cells which though separate are built in close
proximity to one another is of the most advanced and specialized ethological type exhibited by
the Pompilidae. It has been shown by Evans (1953: 161) that this ethological type may be
derived from that exhibited by those ground-nesting Pompilidae which prepare the cell before
hunting. This advanced sequence of behaviour has already been discussed with reference to
the two turret-constructing, clay-soil nesting Dichragenia species occurring at Hilton. The ma-
stery of the technique of using water to work clayey soils shown by species such as the latter
may be seen as presaging the eventual abandonment by some species of the ground as a nest-
ing substrate in favour of nesting aerially. There is thus a change from mud being merely a

62

GESS. ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

by-product of nest excavation to be used in turret construction to mud being purposively mixed
to provide the material used for cell construction.

Only one species was found to modify pre-existing cavities associated with stones on the
ground: Megachile gratiosa (Megachilidae) which constructed its own cells, made of pieces of
leaves, within old, abandoned and therefore open resin and pebble cells of Hoplitis jansei.

Antennotilla phoebe (Mutillidae), a parasite, was reared from the unmodified cells of its
host, Auplopus rossi.

Plants

Wasps and bees which nest in association with plants may be divided into those which nest
within plants, in galleries within plant tissue, and those which nest on plants and use the latter
solely as a raised support to which to attach their aerial nests.

With respect to those species which nest within plant tissue, it is necessary that the parts
of the plants concerned should be thick enough to contain the nesting galleries and that the
hollow parts should retain their physical structure for at least as long as it takes the species
nesting within them to complete their development from egg to adult and to emerge from the
nest. Parts of plants fulfilling these requirements are certain stems and, in the case of larger
shrubs and trees, branches.

Just as soils are divisible on the basis of whether they are sandy or clayey, so plant stems
may be divided on the basis of their internal cellular development and may be classified as
woody, pithy or hollow. The nature of each stem determines which species may be able to
prepare galleries within it and each of the three stem types therefore has associated with it
species peculiar to itself. The internal nature of a stem or branch, however, has no direct
bearing upon those wasps and bees using plants merely as sites to which to attach aerial nests.
Therefore, whereas the nesting of those species nesting within plant tissue is dealt with with
reference to the nature of the stems, the nesting of those species attaching aerial nests onto
plants is dealt with irrespective of the natures of the plants concerned.

At Hilton, the number of species recorded as nesting in or on plants totals 44, made up of
39 nesting within plant tissue and 5 nesting in aerial nests on plants. Of those nesting within
plant tissue, 17 (43,5%) were associated only with woody stems, 12 (31%) were associated
only with pithy stems, and 1 (2,5%) was associated only with hollow stems. Nine species
(23%) were associated with two or all three stem types.

On account of the species which were recorded in two or three stem types, the number of
species/substrate associations exceeds the actual number of species by 14. The number of spe-
cies/substrate associations with respect to those species nesting in plant tissue is therefore 53
and the total number of species/substrate associations for both those species nesting within and
on plants is 58. An analysis of this figure according to the degree of participation of species of
wasps and of bees in the construction of their nests is given in Table 6 and is shown graphically
in the Pie-diagram (Fig. 29).

With respect to those species which themselves prepare galleries in plant stems there is
complete specificity as regards stem type. Pithy stems are by far the most commonly utilized
and the number of species recorded (nine) could probably be more than doubled. Besides the
two species of Ceratina listed, approximately another ten species belonging to the genus are
known to occur at Hilton.

Species preparing their galleries themselves belong chiefly to the Anthophoridae ( Xyloco -
pa and Ceratina ; large and small carpenter bees, respectively) and the Sphecidae (Dasyproctus).
Raphiglossa natalensis of the Eumenidae, known to occur at Hilton but to date not found
nesting, is included among the gallery-excavators on the authority of Meade-Waldo (1913) and
Bequaert (1918).

63

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

Table 6.

Number of species of aculeate wasps and of bees nesting in or on plants.

In nests con-
structed entirely
by the nesters
themselves

In modified
pre-existing
cavities

In unmodified
pre-existing
cavities

In woody stems

2

18

6

In pithy stems

9

8

3

In hollow stems

1

4

2

On plants

5

—

—

In their choice of nesting substrates it appears that these bees and wasps exhibit marked
ecological displacement. In the present study this can best be illustrated with reference to the
nesting of the four species of large carpenter bees recorded at Hilton: Xylocopa caffra,
X. caffrariae, X. divisa and X. sicheli.

X. caffra and X. divisa have been observed to be confined to thick riverine bush where as
far as can be established they make their nests in dead, dry, woody branches of the trees
occurring in that situation. As the two species are very different in size it is likely that each is
confined for its nesting to branches of a certain thickness and interspecific competition between
the two for nesting sites is therefore unlikely despite the fact that they occur in the same
habitat.

X. caffrariae is also a riverine species but occurs not in thick bush but in open tracts along
the banks of permanent and semi-permanent water bodies where it nests exclusively in the dry
culms of Phragmites australis, a reed forming beds in such situation (Figs 14 & 15).

X. sicheli, by contrast is found only in open country and is associated with Aloe ferox
growing on higher ground, especially on the north facing slope of the E-W ridge to the south
of the study area. Nesting is restricted to the old, dry but still attached, inflorescence stems of
this plant (Figs 13 & 30).

The small carpenter bees, Ceratina species, appear also to show considerable ecological
displacement. One of these species, Ceratina sp. A. is most commonly found nesting with
X. sicheli in Aloe ferox inflorescence stems. Though galleries of both bees frequently occur in the
same inflorescence stems they never occupy the same parts: X. sicheli galleries are restricted to
the thicker basal parts above and below the point where the inflorescence branches whereas
Ceratina sp. A. galleries are restricted to the thinner terminal portions of the branches them-
selves. Maximum use is therefore made of the nesting substrate offered by the inflorescence
stem without any interspecific competition for gallery space arising.

With respect to species of Dasyproctus it may be seen from the review of their known
ethology (Gess, 1980b) that considerable specificity exists in their choice of pithy stems in
which to nest. At Hilton D. westermanni was found to nest exclusively in galleries hollowed
out within green inflorescence stems of Urginea altissima, which stems were, however, not
utilized by any of the other four Dasyproctus species. Evidence of the nesting of the latter was,
however, found in the form of old galleries in inflorescence stems of Gasteria species and in
stems of Berkheya decurrens (Gess, 1980b).

64

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Fig. 29. Diagram showing the structure of the communities of the aculeate wasps and the bees nesting in or on plants in
the study area. Numbers accompanying the family names represent numbers of species. The numbers (i, ii. iii) in the
outermost ring correspond to the same numbers in the classification of the aculeate wasps and the bees of the study area
on the basis of their ethology, (i) In nest constructed entirely by the nester. (ii) In pre-existing cavity modified by the
nester. (iii) In pre-existing cavity not modified by the nester.

Xylocopa, Ceratina and Dasyproctus share certain characteristics with respect to their
nesting: all initiate nesting with the preparation of galleries within plant tissue, all construct
multicellular nests the cells of which are serially arranged and mass provisoned, and all con-
struct the cell partitions of materials derived from within the plant stem — plant tissue rasped

65

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

from the walls of the gallery. In no case is foreign nesting material introduced into a gallery
from without.

Xylocopa caffrariae differs from the other species in that it does not itself hollow out a
nesting gallery but makes use of the naturally formed hollow internode of Phragmites . The bee
must nevertheless be classed with the nest constructors and not the tube renters in that it does
not seek out broken off reeds in order to enter the internodes at an open end but habitually
cuts an entrance hole through the hard wall of the culm and thus gains access to an intact
internode from the side (Fig. 31). Once through the culm wall, however, it is spared the
trouble of hollowing out a gallery as this already exists due to the natural breakdown and
disappearance of the pith concurrent with the maturing of the culm.

The nesting by the genera Xylocopa , Ceratina and Dasyproctus in galleries constructed by
themselves within plant tissue must be seen as advanced and derived from nesting in the
ground. Supporting this view is the fact that most nest-constructing (that is, non-parasitic)
Anthophoridae are ground-nesting (Krombein et alia , 1979: 2082) as are many of the more
primitive genera of the Crabronini (the sphecid tribe of which Dasyproctus is an advanced
genus) (Bohart and Menke, 1976: 371). These genera therefore substituted the excavation of
galleries in plant tissue for the more basic excavation of burrows in friable soil, thereby both
freeing themselves from competition for nesting sites in the soil and, more importantly, gaining
the ability to invade previously unutilized habitats.

Dasyproctus differs from the Anthophoridae in so far that it hollows out its galleries in
green plant tissue whereas this is not the case with respect to the two genera of bees. The
statement by Brauns (1913: 117-118) repeated by Hurd and Moure (1963: 15 and 185) and by
Hurd (1978: 1) to the effect that Xylocopa sicheli excavates its galleries in Aloe inflorescences
which are still green and juicy could not be substantiated by the field work of the present
author and must be recognized as being incorrect.

A danger inherent in nesting in green plant stems is that these stems with any nests con-
tained within them may be ingested by browsing herbivores. The ways in which the nests of
several Dasyproctus species, including that of D. westermanni in Urginea altissima, are pro-
tected by physiological or morphological attributes of the stems or by the protected habitats of
the latter is discussed by Gess (1980b: 105).

Among the 39 species recorded at Hilton as nesting within plant tissue, 21 were found to
nest not in galleries of their own excavation but in pre-existing cavities which they, however,
modified by the construction within them of a cell or cells. The species concerned are listed in
the annotated list of species under categories 17, 20 and 23. Four families were represented:
Eumenidae (1 species), Sphecidae (9 species), Colletidae (2 species) and Megachilidae (9 spe-
cies).

There seems little restriction by these insects to any one particular stem type, three of the
species (Holotachysphex turneri, Immanthidium junodi and Megachile spinarum) having been
recorded in two stem types each and another three species (Isodontia stanleyi, Trypoxylon sp.
and Chalicodoma sinuata) in all three stem types each. Further field work would doubtless
augment the number of species recorded in more than one stem type, all the more so as many
of these species are not even restricted to plant stems but are found nesting also in pre-existing
cavities in vertical banks (Euodynerus euryspilus , Chalybion tibiale, Pison montanum, Try-
poxylon sp., Immanthidium junodi, Megachile gratiosa and Megachile spinarum).

Due to the catholicity which these wasps and bees exhibit with respect to the nature of the
material in which the pre-existing cavities utilized by them occur, it is not surprising that they
should readily accept trap-nests of the types described by Krombein (1967 and 1970).

With the exception of one species (Isodontia simoni), all the wasps and bees listed in
category 17 as nesting in woody stems were in fact nesting in trap-nests tied to woody branches
of trees such as Acacia karroo, Maytenus linearis and Rhus lancea. The fact that these trap-

66

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Fig. 31. Dry culm of Phragmites australis showing nest en-
trance hole cut by Xylocopa caffrariae.

Fig. 30. Downcurved dry inflorescence stem of Aloe ferox
showing nest entrance hole cut by Xylocopa sicheli.

Fig. 32. Completed nests of Isodontia pelopoeiformis in trap-nests, showing prey, nesting materials and immature stages

of wasp, (x 0,65)

Fig. 33. Completed nest of Isodontia stanleyi in trap-nest, showing prey, nesting materials and immature stages of wasp.

(x 0,52)

67

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

nests were so readily and frequently made use of (for example, over a period of three summer
seasons 85 trap-nests were occupied by Isodontia pelopoeiformis alone) shows that the woody
branches to which they were attached were in fact being carefully searched for the presence of
pre-existing cavities by the wasps and bees concerned. The inclusion of the trap-nests as exten-
sions of the trees themselves is therefore validated. Further proof of the validity of the
approach is furnished by the nesting of Ampulex sp. near cyanura. It having been noticed that
the wasp nested only in those trap-nests tied to Acacia karroo , a careful physical investigation
was made of this tree species which culminated in the discovery within its branches of nests
constructed within naturally occurring pre-existing cavities.

All nests consisting of cells built within pre-existing cavities may immediately be recog-
nized as such and may readily be distinguished from nests consisting of cells built within a
gallery prepared by the nesting wasp or bee itself. This is possible due to the fact that all users
of pre-existing cavities in plant tissue introduce foreign materials into the former for the con-
struction of cell partitions and nest closures, whereas, as already noted above, species hollow-
ing out galleries themselves utilize materials rasped from the gallery walls.

Of all the species nesting in pre-existing cavities in plant tissue at Hilton, the two species
of Ampulex , A. sp. near cyanura and A. denticollis exhibit the most primitive behaviour. Nest-
ing in both species is commenced with hunting. After the introduction of the single large prey
(a cockroach) into the pre-existing cavity, oviposition onto the prey takes place after which the
cavity is sealed with detritus. It follows that neither Ampulex ever introduces foreign nesting
material into a cavity prior to the introduction of the prey. Preliminary plugs, as found at or
near the blind ends of cavities used for nesting by many of the behaviourally more advanced
wasps, are therefore unknown in the Ampulex nests. Furthermore, each nest consists of a
single cell and where two or more cells are found within a single cavity these represent as many
separate nests, not necessarily even made by the same female.

The great majority of species, however, are behaviourally advanced and commence nest-
ing with the search for a suitable pre-existing cavity. Thereafter some foreign nesting material
is frequently introduced in connection with cell construction. In the wasps, a preliminary plug
at or near the inner end of the cavity may or may not be constructed to form the inner end-
wall of the first cell as shown for Holotachysphex turneri (Gess, 1978: Fig. 1) and for Chalybion
tibiale (Gess and Gess, 1980d: Fig. 1) and additional material may be used to form a tempor-
ary cell closure as in species of Isodontia. In at least some of the bees the introduced foreign
materials may be used to fashion not merely an inner end-wall of a cell but a complete cylin-
drical cell within the walls of the cavity. Such is the case with the leaf-cutter bees. Megachile
gratiosa and M. spinarum , and the mason-bees, Chalicodoma fulva and C. sinuata. Only after
some form of cell preparation has been undertaken does provisioning commence.

If the pre-existing cavity is of a length that allows it, several serially arranged cells are
constructed. Successive provisioned cells are separated one from another by partitions of
foreign material, the outer end-wall (i.e. the cell closure) of one cell forming the inner end-
wall of that immediately succeeding it. Frequently the nest is completed by the construction
near the cavity opening of a closing plug as shown for Holotachyphex turneri (Gess, 1978:
Fig. 1).

Exceptionally no foreign nesting materials are introduced into the pre-existing cavity and
there is therefore a complete absence of any cell partitions. Such is the case in the nests of
Allodape sp. ( rufogastra or exoloma) found at Hilton within old Dasyproctus galleries in Berk-
heya decurrens stems. This bee is unusual also in that it practises progressive provisioning of
the larvae whereas all the other bees recorded at Hilton as nesting in modified pre-existing
cavities in plant tissue practise mass provisioning of the cells, an egg being laid on the provi-
sion of each cell prior to the closure of the latter. The biology of species of Allodape and of
related genera has been reported in detail by Michener (1971b).

68

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

The nesting of the species modifying pre-existing cavities in plant tissue, like that of the
species excavating original galleries in this substrate may be seen as being derived from nesting
in the ground. At Hilton this derivation is clearly reflected in those species for which the
foreign material introduced into the pre-existing cavities in plant tissue consists wholly or par-
tially of earth. The earth used is in all cases of a clayey nature and may be introduced into the
cavity either in the form of mud or in the form of dry clods. In the latter instance small pieces
of twig, bark and other detritus found lying upon the surface of the ground are utilized in
addition.

Such is the case with Holotachysphex turneri, found nesting both in trap-nests tied to
woody trees (Gess, 1978: 209) and in old abandoned galleries of Xylocopa caffrariae in the dry
culms of Phragmites (Gess and Gess, 1980a: 52), which, as has been discussed by Gess (1978:
212-214), may be considered to have developed from a Tachysphex- like form. The species
differs ethologically from typical Tachysphex species in that excavation of a nest in friable soil
has been abandoned in favour of the utilization for nesting of a pre-existing cavity in plant
tissue. Concomitant with this ethological change has been the secondary loss of both the fore-
tarsal rake and the pygidial plate.

The nesting of Tachysphex sp. near modestus in pre-existing cavities (abandoned wasp
burrows) in non-friable clayey soil, discovered since the publication of the account of the
nesting of Holotachysphex turned but described in some detail earlier in the present discussion,
may be seen to represent an intermediate but very important pre-adaptive step that will have
had to have been taken by the progenitor of H. turned. Clearly, once the basic change had
been made from the excavation of a nest in friable soil to the facultative use of a pre-existing
cavity in non-friable soil (as in Tachysphex sp. near modestus) the way was open to a change to
obligatory use of such a pre-existing cavity in the ground and for its eventual abandonment in
favour of a pre-existing cavity above the ground in plant tissue. The use by Holotachysphex
turneri of dry clods of clayey earth and pieces of detritus collected on the ground must there-
fore be interpreted as relict behaviour envolved in response to nesting in pre-existing cavities
in non-friable clayey soils.

Dry clods of clayey earth and pieces of detritus (including bits of stick, vetch burs, insect
remains and small mammal droppings) picked up off the ground are also the main foreign
nesting materials introduced into pre-existing cavities in plant tissue by Isodontia pelopoeifor-
mis , according to Bohart and Menke (1976: 121) one of the more structurally primitive species
of the genus Isodontia. The same interpretation as that given to the use of these nesting mate-
rials by Holotachysphex turneri may be applied to their use by Isodontia pelopoeiformis which
therefore is seen similarly as having evolved from a nester in pre-existing cavities in non-friable
soils. It is therefore of interest that amongst the nesting situations recorded for species of
Isodontia by Bohart and Menke (1976: 121) there should be that of abandoned bee burrows in
the ground.

Isodontia pelopoeiformis , however, in addition to the above foreign nesting materials uses
“fluffy” plant material, at Hilton derived from the fruiting inflorescences of the Composites
Lasiospermum bipinnatum and Senecio species (including S. leptophyllus ) (Fig. 32). Cell parti-
tions are almost exclusively constructed of this material whereas preliminary plugs and nest
closing plugs are constructed of earth, detritus and “fluff” in distinct layers. The “fluffy” plant
material which is collected directly from the plant inflorescences represents “new” nesting
material as opposed to the earth and the detritus which as has been shown to represent “old”
or “relict” materials. Presumeably pieces of this “fluff” were initially picked up off the ground
like any other detritus but as the use of this particular nesting material became more estab-
lished the wasp took to collecting it from its source.

Isodontia stanleyi, a structurally more advanced species, is also ethologically more ad-
vanced for it has abandoned the legacy of its past in the form of “old” or “relict” nesting

69

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

materials and uses only ‘'new” materials (Fig. 33). These consist of short lengths of grass leaf
blades and longer lengths cut from grass inflorescences including lengths of unbranched culm
and the branched portions including the flower heads. The cell partitions are constructed of
short lengths of grass leaf blades, transversely arranged; preliminary plugs and the inner parts
of the closing plugs are composed of the longer lengths of grass inflorescences, transversely
coiled; and the outer parts of the closing plugs are formed of the latter material arranged
longitudinally, the stems lying parallel and projecting from the nest entrance as a broom-like
tuft.

That the use of grass by I. stanleyi is homologous with the use of Composite “fluff” by
I. pelopoeiformis is evident from the very occasional and exceptional use of lengths of very soft
grass leaf blades by the latter species when its usual “new” nesting material is not available.

An interesting piece of behaviour practised by both I. pelopoeiformis and I. stanleyi is that
of “buzzing” in the nesting cavity in order to consolidate the introduced materials forming the
preliminary plug and the cell partitions. The “buzzing” is the same as that already commented
upon in connection with other Sphecidae, the ground-nesting Padalonia canescens and
Ammophila ferrugineipes which use it to loosen soil in the course of the excavation of their
burrows. This behaviour may be seen as additonal evidence for deriving the plant-nesting
Isodontia species from ground-nesting, possibly nest-excavating, forms.

Earth in the form of mud for the construction of cell partitions and nest closures is intro-
duced into pre-existing cavities in plant tissue by Euodynerus euryspilus (Eumenidae) and
Chalybion tibiale, Pison montanum and Trypoxylon sp. (all Sphecidae).

The use of mud as the nesting material suggests that these wasps are derived from species
which employed water to aid them in the excavation of original nests in non-friable clayey soils
in the manner already described for other species of Eumenidae ( Parachilus , Pseudepipona
and Antepipona) and Sphecidae (Bembecinus) . From the excavation of original nests in the
ground to the use of pre-existing cavities in plant tissue two changes of behaviour would have
been required — firstly the abandonment of the excavation of an original nest in the ground in
favour of the use of pre-existing cavity in the same substrate, secondly the abandonment of
nesting in this pre-existing cavity in the ground in favour of nesting in a similar pre-existing
cavity above ground in plant tissue (or for that matter in a vertical bank). The use of mud for
cell partitions and nest closures, initially a consequence of using water in nest excavation, is
retained by those species nesting in pre-existing cavities and provides a continuous thread
linking the three nesting types.

The three nesting types outlined above as necessary to the derivation may all be found to
occur within the genus Pison. Thus, excavation with the aid of regurgitated water of an origin-
al burrow in non-friable soil has been reported for the South American Pison chilense (Janvier,
1928, as reported by Bohart and Menke, 1976: 335). Use of pre-existing cavities (abandoned
burrows of bees and wasps) in clayey soils has been reported for the Micronesian Pison nigel-
lum (Krombein, 1950: 139) and is recorded in the present account for Pison allonymum at
Hilton. Use of pre-existing cavities in plant tissue has been reported for several species (see
Bohart and Menke, 1976: 335) and is the habit at Hilton of Pison montanum , one of the
species presently being discussed.

Great specificity is shown in the nature of the foreign materials introduced into pre-
existing cavities in plant tissue by various pollen-collecting (that is, non-parasitic) bees. In the
family Megachilidae, Megachile gratiosa and Megachile spinarum construct their cells of pieces
of green leaves cut to definite shapes and sizes and arranged overlappingly (Fig. 39). Leaves of
only one plant species are used by these two bees at Hilton, namely those of Maytenus hetero-
phylla (Celastraceae) which would appear therefore to be those best suited for the purpose at
that locality. Elsewhere M. gratiosa has been found to use not only green leaves but, less
commonly, flower petals also (Taylor, 1963, 1965 and 1968). Other bees of the same family,

70

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Capanthidium capicola and Immanthidium junodi , both so-called carder bees, use cotton wool-
like plant down, Heriades species use plant resin, and Chalicodoma fulva and Chalicodoma
sinuata use mud (Fig. 40).

Hylaeus species including Hylaeus braunsi (family Colletidae), by contrast, construct their
cells not of materials which they have collected but of a cellophane-like material secreted by
their salivary glands.

The construction by many of the bees of complete cells within the pre-existing nesting
galleries as opposed to the construction only of cell end walls (that is cell partitions) as by the
wasps is clearly linked to the nature of the provision. The damp nature of the pollen and
nectar mixture clearly requires that it be protected not only from desiccation through loss of
moisture to the nesting substrate but also that it be protected from contamination from the
nesting substrate. The cell walls be they of leaves, plant down or mud thus serve to isolate the
provision from the nesting substrate. This is of particular importance in the Hylaeus species the
nectar-pollen mixture of which is very liquid and can only be protected and contained by the
impermeability of the cellophane-like “bags” forming the cells. On the other hand the pollen
stored by Heriades species, including H. spiniscutis at Hilton nesting in old Dasyproctus galler-
ies in Berkheya and Gasteria stems, is very dry which may account for the fact that no cell
walls are constructed and that the introduced nesting material, resin, is used only for the
construction of cell partitions and nest closures.

It may be accepted that foreign materials for the construction of cells within pre-existing
burrows were used already by the ground-nesting antecedents of the bees now nesting above
ground and that the change in nesting situation was not accompanied by any change in the
nature of the nesting materials. Certainly, there is no difference in the nature of the leaf cells
constructed at Hilton by, on the one hand, the ground-nesting Megachile semiflava and
M. stellarum and, on the other hand, the above-ground-nesting M. gratiosa and M. spinarum.

In passing it may be remarked that the apparently successful co-existence at Hilton of six
species of Megachile must be due at least in part to the fact that the search for suitable pre-
existing cavities in which to nest is not restricted to one situation but is divided between the
ground (M. aliceae, M. meadewaldoi, M. semiflava and M. stellarum) and situations above the
ground (M. gratiosa and M. spinarum). Within each of the two major nesting situations com-
petition is probably further reduced by the fact that the species concerned are of different sizes
and therefore seek pre-existing cavities of commensurate bore. For example, with respect to
the two species nesting above the ground and thus found also in trap-nests, the 8 mm long
M. gratiosa was found to be restricted to small bore (6,4 mm) trap-nests whereas the 13 mm
long M. spinarum was found to use mostly larger bore (9,5 and 12,7 mm) trap-nests.

Six of the thirty-nine species recorded at Hilton as nesting within plant tissue were found
doing so in pre-existing cavities which they themselves had not in any way modified. AH were
reared from cells constructed by other aculeate Hymenoptera and all may be classed as clepto-
parasites as they feed upon the provision stored by their hosts.

Three families were involved, namely Chrysididae, Sapygidae and Megachilidae. The four
cuckoo wasps and their hosts were: Chrysidea africana and Chrysis sp. (near purpuripyga),
both in the cells of Trypoxylon; Chrysis inops in the cells if Pison montanum ; and Octochrysis
hoplites in the cells of Euodynerus euryspilus. Sapyga simillima (Sapygidae) and Coelioxys
penetratrix (Megachilidae) developed in the cells of Heriades sp. and Megachile spinarum re-
spectively.

All the above cleptoparasitic species ahear to be highly host specific with regard to the
nests in which they develop and their occurrence is therefore linked with that of their hosts.
Chrysidea africana , besides being reared from Trypoxylon cells in trap-nests tied to bushes and
trees, was found also in association with Trypoxylon nests in cavities in vertical banks. On the

71

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

face of the latter the cuckoo wasp was frequently seen walking about inspecting holes, then
flying up and settling elsewhere after which the inspection of holes was resumed.

With respect to the association of Coelioxys with Megachile it is of interest that the two
genera are closely related. It appears that both are probably derived from the same nest-
constructing and pollen-collecting forms and that the cleptoparasitic behaviour of Coelioxys is
secondary and specialized, being arrived at by the abandonment by this genus of nest construc-
tion and pollen collection in favour of the utilization for nesting of the nest constructed and
provisioned by Megachile.

Five species, all wasps, were listed at Hilton as constructing aerial nests on plants, namely
Eumenes lucasius (Eumenidae), Celonites capensis (Masaridae), Polistes smithii and Ropalidia
sp. B. (Vespidae), and Sceliphron quartinae (Sphecidae).

The nesting of all these species, like those species constructing nests above ground but
within plant tissue, may be seen as derived from nesting in the ground. In contrast to those
species which nest within plant tissue and which in their nesting therefore are subject to the
constraints imposed either by the availability of plant tissue suitable for gallery excavation or
by the availability of pre-existing cavities of suitable dimensions, the species nesting on plants
are free of such limitations and for their nesting require only that the plant should offer mecha-
nical support for the nest. Despite this freedom with regard to the surface to which the nest
may be attached, the species cannot nest anywhere for all are dependent upon the presence of
water for the construction of their cells.

With respect to the four Vespoidea, water, collected in the crop and regurgitated from
there, is required to transform dry clayey earth into mud and dry rasped-off wood fibres into
wood pulp, materials used for cell building by Eumenes and Celonites and by Polistes and
Ropalidia respectively. Sceliphron quartinae, in contrast to the above two water-collecting,
mud-mixing species, probably behaves like the mud-daubing Sceliphron spirifex and related
species which gather already existing mud from beside pools and puddles.

Nesting is therefore restricted to within a certain distance of water, and furthermore, in
the case of those species making mud cells, to areas of clayey soils. However, Sceliphron
quartinae, which has been reported by Brauns (1911: 119) to make its cells either of clayey
mud or cow dung, will by its use of the latter material (when fresh and wet) be able to nest in
areas where mud suitable for cell construction is unavailable, either because the soil is not of a
clayey nature or because pools and puddles are absent, or both.

With respect to wasps and bees nesting within plant tissue, two associations between some
species and certain plants are of particular interest. These associations will be dealt with in
further detail in order to illustrate the primary role played by the plant in question and to show
the chains of dependence existing amongst the insects constituting a community upon such a
plant. It is obvious that such chains of dependence within a community are not the monopoly
of those species nesting within plant tissue but are a feature of all communities whatever their
nesting substrates and nesting situations. The present two examples are chosen as they have
the added interest of having as the nesting substrate an additional living organism, the plant.

The first example pertains to a complex of three insect species associated with and re-
stricted to the shrub or small tree. Acacia karroo , and concerns the nesting of the sphecid,
Ampulex sp. near cyanura, listed in category 17 of the annotated list as being a modifier of
pre-existing cavities in woody stems. As will be seen in Fig. 34 in which the association under
discussion is diagrammatically portrayed the other two insects concerned are Ceroplesis hotten-
tota (Cerambycidae) and Bantua dispar (Blattidae) (Figs 35 & 36).

First in the three-linked chain is the beetle, C. hottentota , the larvae of which in the course
of their feeding bore galleries in finger-thick branches of A. karroo. The consequences of the
activity of these larvae are three-fold and make possible the existence of the other two species
in the chain. Firstly, damage to the living branches of the tree results in a copious amount of

72

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Fig. 34. Diagram showing the structure of a complex of three insect species, Ceroplesis hottentota (Cerambycidae),
Bantua dispar (Blattidae) and Ampulex sp. near cyanura (Sphecidae) associated with and restricted to the shrub or small

tree Acacia karroo (Leguminosae).

gum being exuded from the wounds. Secondly, and at a later stage, the bored branches die
with the result that their bark separates from the wood. Thirdly, after the beetles have com-
pleted their larval feeding, have pupated and have metamorphosed into adults, they break out
and abandon their galleries with the result that the latter become available for the use of other
insects which cannot themselves hollow out such cavities.

73

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

Fig. 35. Ceroplesis hottentota , the larvae of which bore galleries in finger-thick branches of Acacia karroo. (X 3)

Fig. 36. Bantua dispar and Ampulex sp. near cyanura. The wasp having subdued the cockroach has cut off the end of an
antenna (seen lying at bottom right-hand corner of photograph) and is malaxating prior to leading prey to nesting

gallery, (x 3)

74

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Second in the chain of species is the cockroach, B. dispar , which appears to be restricted
to those A. karroo shrubs and trees harbouring C. hottentota. A nocturnal species, it spends
the day in hiding in old abandoned beetle galleries and under the loosened bark of dead
branches and comes out at night to feed upon the gum exuding from living branches being
bored by the beetle.

Third and last in the chain is the wasp, Ampulex sp., which preys exclusively upon
B. dispar for which it may be seen hunting on foot on the branches of the tree. Having flushed
a B. dispar from its hiding place, the wasp stings it and then, holding the subdued cockroach
by one of its antennae, leads it to an old C. hottentota gallery (Fig. 36). Once arrived there the
wasp installs the cockroach within the cavity, oviposits on it and then seals the gallery with
detritus collected on or immediately under the tree.

The existence of Ampulex sp. near cyanura within any locality is therefore inextricably
linked with the presence of not only its prey but also with that of the beetle and ultimately with
that of Acacia karroo. As the wasp was never collected anywhere other than on A. karroo it
would appear that its entire life is spent on this plant.

The second example, presented diagrammatically in Fig. 37, pertains to a community of
species referable to categories 19, 20 and 21 of the annotated list and associated with the dry
inflorescence stems of Aloe ferox.

The community differs from the complex of three species concerned in the first example in
that the species at the core of the community are hymenopterous, not coleopterous. Fur-
thermore, species are included that exhibit all three degrees of participation in nest construc-
tion: those which construct their nests entirely by themselves, those which modify pre-existing
cavities (Figs 38, 39 & 40) and those which nest parasitically in pre-existing cavities which they
do not themselves modify.

Most strikingly different, however, is the far looser association with the plant, only the
two gallery excavators at the core of the community, Xylocopa sicheli and Ceratina sp. A.
being apparently restricted to the aloes. However, even these species are dependent upon
these plants only in so far that they offer a suitable pithy substrate in which nesting galleries
may be excavated and in areas where A. ferox does not occur, inflorescence stems of other
species of Aloe are known to be used. Foraging for pollen and nectar with which to provision
the cells within the galleries takes place on other plants for at the time of nesting the aloes are
not in flower.

At the level of those species which modify pre-existing cavities and are recorded as using
the abandoned borings of the two gallery excavating bees there is no dependence on the pre-
sence of aloes at all nor on the presence of the bees. As has been shown species like Isodontia
stanleyi, Trypoxylon sp., Megachile spinarum, Chalicodoma sinuata and Immanthidium junodi
will nest in a variety of pre-existing cavities provided these are situated up above the ground. It
is clear, however, that if these species are to nest in Aloe inflorescence stems then they can do
so only in the presence of old Xylocopa and Ceratina borings.

At the level of those species which do not modify pre-existing cavities but nest parasitical-
ly the association is clearly with their hosts only and the occurrence of these species is there-
fore determined by that of their hosts.

Communities of loosely associated species of wasps and bees, as exemplified by the com-
munity nesting in Aloe ferox inflorescences, may be seen to exist in all four nesting situations —
the ground, vertical banks, stones and plants and in the subdivisions of these primary nesting
situations. It may therefore be concluded that the picture that emerges of the interaction of
species of wasps and bees in a community such as that described above is typical of the entire
population of these insects at Hilton.

Within any community individual species, as exemplified by Ampulex sp. near cyanura of
the community nesting in Acacia karroo stems, may be associated not with other species of

75

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

Fig. 37. Diagram showing the structure of a community of insects associated with the dry inflorescence stems of Aloe
ferox (Liliaceae). With respect to the aculeate wasps and the bees the degree of participation in the construction of their
nests is indicated by the numbers (i, ii, iii) as in the classification of these insects on the basis of their ethology, (i) In
nest constructed entirely by the nester. (ii) In pre-existing cavity modified by the nester. (iii) In pre-existing cavity not

modified by the nester.

76

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

Fig. 39

Fig. 38

Figs 38 and 39. Dry inflorescence stems of Aloe ferox cut longitudinally to show old galleries bored by Xylocopa sicheli
and containing nests of Isodontia stanleyi (Fig. 38) and Megachile spinarum (Fig. 39).

wasps or with bees, except perhaps peripherally (e.g. in competition for pre-existing cavities),
but may rather be part of the community by virtue of their association with non-
hymenopterous species.

The circumscribed association of Ampulex sp. near cyanura with the beetle and the cock-
roach making up the three-species complex on Acacia karroo appears to be exceptional and no
similar example was found during the course of the present study at Hilton.

77

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1, OCTOBER 1981

Fig. 40. Dry inflorescence stem of Aloe ferox cut longitudinally to show old gallery bored by Xylocopa sicheli and

containing nest of Chalicodoma sinuata.

The study of nesting ethology of the aculeate wasps and the bees of Hilton has apart from
adding considerably to the knowledge of these groups in Africa provided a clear illustration of
their remarkable propensity for behavioural adaptability which has led to the formation of
communities of large numbers of sympatric species showing marked ecological displacement.

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 for identification of material are due to the following: Mrs E. Brink and Miss G.

78

GESS: ASPECTS OF STUDY OF ACULEATE WASPS AND THE BEES

V. Britten of the Albany Museum Herbarium — Plants; Dr D. J. Brothers of Natal University,
Pietermaritzburg — Mutiilidae; Mr C. F. Jacot-Guillarmod, lately of the Albany Museum, Gra-
hamstown — Scoliidae and Tiphiidae; Dr J. G. Theron of the University of Stellenbosch —
Cicadellidae. Thanks are due also to Mr R. H. M. Cross and Mr A. G. Bruton of the Electron
Microscopy Unit, Rhodes University, for producing the scanning electron micrographs repro-
duced in Figs 18-23.

The author is grateful to Dr G. B. Whitehead of Rhodes University for his encourage-
ment. Sarah Gess who has acted as the author’s research assistant during the past eleven years
is thanked for her invaluable support and encouragement.

Gratitude is expressed to the C.S.I.R. for running expenses grants for field work during
the course of which the present study was made.

REFERENCES

Acocks, J. P. H. 1953. Veld types of South Africa. Mem. bot. Surv. S. Afr. 29: i-iv, 1-192.

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

Arnold. G. 1934. The Psammocharidae of the Ethiopian Region. Part 3. Ann. Transv. Mus. 15: 283-400.

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

Arnold, G. 1947. A key to the African genera of the Apidae. J. ent. Soc. sth. Afr . 9: 193-218.

Bequaert, J. 1918. A revision of the Vespidae of the Belgian Congo based on the collection of the American Museum
Congo Expedition, with a list of Ethiopian Diplopterous wasps. Bull. Amer. Mus. nat. Hist. 39: 1-384.
Bohart, R. M. 1940. A revision of the North American species of Pterocheilus and notes on related genera (Hymenop-
tera, Vespidae). Ann. ent. Soc. Am. 33 (1): 162-208.

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

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

Brain, C. K. 1929. Insect pests and their control in South Africa. Cape Town. Nasionale Pers.

Brauns, H. 1911. Biologisches liber sudafrikanische Hymenopteren. Z. wiss. InsektBiol. 7: 16-19, 90-92, 117-120,
238-240.

Brauns, H. 1913. Biologie siidafrikanischer Apiden. Z. wiss. InsektBiol. 9: 116-120, 190-193.

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

Brauns, H. 1926b. V. Nachtrag zu Friese, Bienen Afrikas. Zool. Jb., Syst. 52: 187-230.

Brauns, H. 1930. Neue Bienen aus Siid-Afrika. Konowia 9: 43-54.

Clausen, C. P. 1940. Entomophagous Insects. New York: McGraw-Hill.

Dyer, R. A. 1937. The vegetation of the divisions of Albany and Bathurst. Mem. bot. Surv. S. Afr. 17: 1-138.

Evans, H. E. 1953. Comparative ethology and the systematics of spider wasps. Syst. Zool. 2 (4): 155-172.

Evans, H. E. 1966. The Comparative Ethology and Evolution of the Sand Wasps. Cambridge, Mass. Harvard University
Press.

Evans. H. E. 1970. Ecological-behavioral studies of the wasps of Jackson Hole. Wyoming. Bull. Mus. Comp. Zool.
Harvard 140: 451-511.

Evans, H. E. and Eberhard, M. J. W. 1970. The Wasps. Ann Arbor. University of Michigan Press.

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. 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. 1980a. Prey and nesting sites of some sympatric species of Cerceris (Hymenoptera: Sphecidae) with a
review and discussion of the prey diversity of the genus. Ann. Cape Prov. Mus. (nat. Hist.) 13 (7): 85-93.
Gess, F. W. 1980b. Some aspects of the ethology of Dasyproctus westermanni (Dahlbom) (Hymenoptera: Sphecidae:
Crabroninae) in the Eastern Cape Province of South Africa. Ann. Cape Prov. Mus. (nat. Hist.) 13 (8): 95
-106.

Gess. F. W. and Gess, S. K. 1974. An ethological study of Dichragenia pulchricoma (Arnold) (Hymenoptera: Pompil-
idae), 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. 1976a. 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.

79

ANN. GAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 1. OCTOBER 1981

Gess, F. W. and Gess, S. K. 1976b. Ethological notes on Dichragenia neavei (Kohl) (Hymenoptera: Pompilidae), an
African spider-hunting wasp building a turreted, subterranean nest. Ann. Cape Prov. Mus. (nat. Hist.) 11 (8):
129-134.

Gess, F. W. and Gess, S. K. 1980a. 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.

Gess, F. W. and Gess, S. K 1980b. Spider vanquishers: the nesting of Tachypompilus ignitus (Smith) and Batozonellus
fuliginosus (Klug). The Eastern Cape Naturalist 69: 4-7.

Gess, F. W. and Gess, S. K. 1980c. 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. Ann. Cape Prov. Mus. (nat. Hist.) 13 ( 6): 63-83.

Gess, F. W. and Gess, S. K. 1980d. The whited sepulchre — the nesting of Chalybion tibiale (Fabr.). The Eastern Cape
Naturalist 70: 11-14.

Hurd, P. D. 1978. An annotated catalog of the carpenter bees (Genus Xylocopa Latreille) of the Western Hemisphere
(Hymenoptera: Anthophoridae). Washington, D. C. Smithsonian Institution Press.

Hurd, P. D. and Moure, J. S. 1963. A classification of the large carpenter bees (Xylocopini) (Hymenoptera: Apoidea).
Univ. California Pub. Entomol. 29: 1-365.

Janvier, H. 1928. Recherches biologiques sur les predateurs du Chili. Ann. Sci. Nat. Zool. (10) 11: 67-207.

Krombein. K. V. 1950. The aculeate Hymenoptera of Micronesia. II. Colletidae, Halictidae, Megachilidae, and Apidae.
Proc. Hawaiian Ent. Soc. 14: 101-142.

Krombein, K. V. 1967. Trap-nesting Wasps and Bees: Life Histories, Nests, and Associates. Washington, D.C. Smithson-
ian Press.

Krombein, K. V. 1970. Behavioral and life-history notes on three Floridan solitary wasps (Hymenoptera: Sphecidae).
Smithson. Contr. Zool. 46: 1-26.

Krombein, K. V. et al. 1979. Catalog of Hymenoptera in America North of Mexico. Vol. 2 Apocrita (Aculeata).
Washington, D.C. Smithsonian Press.

Mally, C. W. 1908. Bee Pirates. Agric. J. Cape G. H. 33 (2): 206-213.

Meade-Waldo, G. 1913. New species of Diploptera in the collection of the British Museum. Ann. Mag. nat. Hist. (8)
11: 44-54.

Michener, C. D. 1944. Comparative external morphology, phylogeny and a classification of the bees (Hymenoptera).
Bull. Amer. Mus. nat. Hist. 82: 151-326.

Michener, C. D. 1968. Nests of some African Megachilid bees, with description of a new Hoplitis (Hymenoptera,
Apoidea). J. ent. Soc. sth. Afr. 31 (2): 337-359.

Michener, C. D. 1971b. Biologies of African allodapine bees (Hymenoptera, Xylocopinae). Bull. Amer. Mus. nat. Hist.
145 (3): 219-301.

Olberg, G. 1959. Das Verhalten der solitdren Wespen Mitteleuropas (Vespidae, Pompilidae, Sphecidae). Berlin. Deut-
scher Verlag der Wissenschaften.

Oldroyd, H. 1964. The Natural History of Flies. London. Weidenfeld and Nicolson.

Painter, R. H. 1932. The Bombyliidae of China and nearby regions. Lingnan Sci. Jour. 11: 341-374.

Rozen, J. G. 1969a. Biological notes on the bee Tetralonia minuta and its cleptoparasite, Morgania histrio transvaalensis
(Hymenoptera: Anthophoridae). Proc. ent. Soc. Wash. 71 (1): 102-107.

Rozen, J. G. 1969b. The biology and description of a new species of African Thyreus, with life history notes on two
species of Anthophora (Hymenoptera: Anthophoridae). J. New York ent. Soc. 77: 51-60.

Skaife, S. H. 1950. Some South African solitary bees. S. Afr. J. Sci. 46: 183-187.

Smit, B. 1964. Insects in southern Africa: how to control them. Cape Town. Oxford University Press.

Smithers, C. N. 1958. Notes on the life history of Sphex ( Isodontia ) pelopoeiformis Dhlb. in Southern Rhodesia.
Entomologist’s mon. Mag. 94: 193-197.

South Africa (Republic) Dept, of Transport. Weather Bureau. Climatology Branch. 1965. Climate of South Africa. Part
9. Average monthly and annual rainfall and number of rain-days up to the end of 1960. Pretoria. Government
Printer.

Taylor, F. 1939. Beekeeping for the beginner. Union of South Africa Dept. Agric. and Forestry Bulletin 199.

Taylor, J. S. 1962a. A note on the carder bee Anthidium junodi melanosomum Cameron. Pan-Pacif. Ent. 38: 241-243.

Taylor, J. S. 1962b. Notes on Heriades freygessneri Schletterer (Hymenoptera, Megachilidae). J. ent. Soc. sth. Afr. 25:
133-139.

Taylor, J. S. 1963. Notes on the Leaf-cutter Bee Megachile (Eutricharaea) gratiosa Gerstaecker (Hymenoptera:
Megachilidae). Pan-Pacif. Ent. 39 (2): 129-134.

Taylor, J. S. 1965. Notes on some South African bees (Hymenoptera: Colletidae, Megachilidae). Pan-Pacif. Ent. 41
(3): 173-179.

Taylor, J. S. 1968. Notes on some South African bees and wasps. Entomologist's Record 80: 69-72.

80

art

i

Ml3

MH

ANNALS,

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

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

26th JULY 1982

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

A selected bibliography of literature on Odomata from Africa and adjacent

islands Part 2

by

B. C. WILMOT and L. F. WILMOT
(Albany Museum, Grahamstown)

This bibliography, like the first compiled by Wilmot & Wilmot (1978), is intended to
supplement that of Pinhey (1962) which included mostly systematic references to African Odo-
nata up to December 1959. Though the vast majority of articles included in the present bib-
liography were published subsequent to 1959 a few earlier non-systematic references omitted
by Pinhey and previously overlooked by the compilers have been included. All articles in
which reference is made to Odonata have been included, whether of a systematic, ecological,
zoogeographical, biological or bibliographical nature.

The geographical coverage of articles is indicated by a code letter or letters placed after
each reference. The code letters and their meanings, after Scott & Scott (1969), are as follows:

N. North Africa, i.e. Africa north of the Sahara, including oases.

C. Africa south of the Sahara but north of the Zambesi. This is the largest area, includ-
ing all the central African countries from the Sudan to Zambia and Senegal to Soma-
lia.

S. Southern Africa, i.e. Africa south of the Zambesi. This includes South Africa, Nami-
bia, Zimbabwe, Lesotho, Swaziland, Botswana and the southern part of Mozambi-
que.

I. Islands off the coast of Africa with the exception of those in the Mediterranean Sea
which are excluded.

A. This indicates Africa as a whole and means that the article concerned covers the
continent in general.

The relative importance of an article is indicated by the number of asterisks placed after
the geographical code of that reference. Following Davies & Davies (1976) the ratings are as
follows:

* only passing reference made to Odonata;

** a part of the article deals with Odonata;

*** the article deals exclusively with Odonata.

An article dealing exclusively with Odonata but with only passing reference to the African
fauna is rated as single asterisk.

The abbreviations of publication titles cited follow the World List of Scientific Periodicals
(1963-1965). Titles of publications for which there are no acceptable abbreviations are written
out in full.

81

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. HPT 2, JULY 1982
BIBLIOGRAPHY

Aguesse, P. 1968. Faune de I'Europe et du Bassin Mediterraneen. 4. Les odonates de I'Europe occidental, du Nord de
I'Afrique et des ties atlantiques. Centre National de la Recherche Scientifique, Masson et Cie, Paris, pp.
1-255, 106 figs, 7 plates, 6 tables, 10 maps. N J. **

Allanson, B. R. et al. 1966. An estuarine fauna in a freshwater lake in South Africa. Nature 209(5022):532-533. S. *
Bailey, R. G., S. Churchfield and R. Pimm 1978. Observation on the zooplankton and littoral macroinvertebrates of
Nyumba ya Mungu reservoir, Tanzania. Biol. J. Linn. Soc. 10:93-107. C. *

Balirwa, J. S. 1979. A contribution to the study of the food of six cyprinid fishes in three areas of the Lake Victoria
Basin, East Africa. Hydrobiologia 66(l):65-72. C. *

Bell-Cross, G. 1973. The fish fauna of the Buzi River System in Rhodesia and Mozambique. Arnoldia Rhod. 6(5):
1-14. S. *

Bell-Crqss, G. 1974. A fisheries survey of the Upper Zambezi River System. Occas. pap. Natl. Mas. Rhod., B., Nat.
sci. 5(5):279— 338. S,C. *

Belyshev, B. F. and A. Yu. Haritonov 1978. The distribution of the cosmopolitan genus Anax Leach and its probable
origin (Anisoptera: Aeshnidae). Odonatologica 7(2) : 1 15— 121 . A. **

Belyshev, B. F. and A. Yu. Haritonov 1980. [On the latest faunistic contacts of Ethiopian and Australian regions in
Odonata.] [Russian with English summary] Ent. Obozr. S9(l):89— 91. S. **

Belyshev, B. F. and A. Yu. Kharitonov 1979. On the Atlantic Gap in the ranges of dragonflies (Odonata) in the south
parts of Africa and America. Zool. Zh. 58(4):610-611. S. **

Blake, B. F. 1977. Food and feeding of the mormyrid fishes of Lake Kainji, Nigeria, with special reference to seasonal
variation and interspecific differences. J. Fish. Biol. 11:315-328. C. *

Boon von OchssSe, G. A. 1979. Dragonflies in the diet of the teleostean fish in the Comoe River, Upper Volta, West
Africa. Notul. odonatol. 1(3) :46— 47 . C. ***

Boyes, J. W., J. M. van Brink and B. Kiauta 1980. Sixteen dragonfly karyotypes from the Republic of South Africa
and Swaziland, with evidence on the possible hybrid nature of Orthetrum julia falsum Longfield (Anisoptera:
Libellulidae). Odonatologica 9(2):131-153. S. ***

Bruton, M. N. 1979a. The breeding biology and early development of Clarias gariepinus (Pisces: Clariidae) in Lake
Sibaya, South Africa, with a review of breeding in species of the subgenus Clarias (Clarias). Trans zool. Soc.
Lond. 35:1-45. S. *

Bruton, M. N. 1979b. The food and feeding behaviour of Clarias gariepinus (Pisces: Clariidae) in Lake Sibaya, South
Africa, with emphasis on its role as a predator of cichlids. Trans zool. Soc. Lond. 35:47-114. S. *

Bruton, M. N. 1979c. The fishes of Lake Sibaya. In: Allanson, B. R. ed.. Lake Sibaya. Monographiae biol. 36:162-
245. S. *

Bruton, M. N. and H. M. Kok 1980. The freshwater fishes of Maputoland. In: Bruton, M. N. and K. H. Cooper eds,
Studies on the Ecology of Maputoland. Rhodes University, Grahamstown and Natal Branch of the Wildlife
Society of Southern Africa, pp. 210-244. S. *

Burton, G. J. and T. M. McRae 1972. Phoretic attachment of Simulium larvae and pupae to mayfly and dragonfly
nymphs. Mosquito News 32(3):436-443. C. **

Buxton, D. R. 1937. Insects of the Lake Rudolf Rift Valley Expedition, 1934. - 1. Orders other than Coleoptera. Ann.
Mag. not. Hist., Ser. 10, 20:579-588. C. *

Cammaerts, R. 1978. Mission entomologique de Musee Royale de I'Afrique Centrale aux Monts Uluguru, Tanzanie
(L. Berger, L. Leleup et J. Debecker, V-VIII. 1971). 24. Odonata. Rev. Zool. Afr. 92( 1 ) : 1 1—36. C. ***
Carlsson, G. 1968. Benthonic fauna in african watercourses, with special reference to black-fly populations. Scandina-
vian Institute of African Studies, Uppsala, 13 pp. [only separatum seen] C. *

Carmouze, J. -P. et al. 1972. Grandes zones ecologiques du lac Tchad. Cali. O.R.S.T.O.M., ser. Hydrobiol. 6(2):103-
169. C. *

Coetzee, D. J. 1980. Zooplankton and environmental conditions in Groenvlei, Southern Cape, during 1976. J . Limnol.
Soc. sth. Afr. 6(1) : 5— 1 1 . S. *

Coetzer, A. 1978. The Invertebrate Fauna and Biotic Index Value of Water Quality of the Great Berg River, Western
Cape. J. Limnol. Soc. sth. Afr. 4(1): 1-7. S. *

Consiglio, C. 1978. Odonata collected in Ethiopia by the expeditions of the Accademia Nazionale dei Lincei. 11. New
and rare species of Notogomphus from Ethiopia (Odonata: Gomphidae). Problemi att. Sci. Cult. (Ill)
243:53-58. C. ***

Corbet, P. S. 1974a. Entomological reflections. Bull. ent. Soc. Can. 6(3):70-75. C. *

C [orbet], P. S. 1974b. Odonata. In: Encyclopedia Britannica (15th ed.). Hemingway Beuton, pp. 507-509. A. **
Corbet, P. S. 1976. A. E. Gardner, [obituary, includes bibliography] Odonatologica 5(4):387-390. I. *

Corbet, P S. 1980. Biology of Odonata. Ann. Rev. Ent. 25:189-217. A. **

Corbet, Sarah A. 1977. Gomphids from Cameroon, West Africa (Anisoptera: Gomphidae). Odonatologica 6(2):55-68.

Q * * *

Dejoux, C. and J. M. Elouard 1977. Action de l’Abate sur les invertebres aquatiques. Cinetiquc de decrochement a
court et moyen terme. Cab. O.R.S.T.O.M., ser. Hydrobiol. 1 1(3 ) :217— 230. C. *

82

WILMOT & WILMOT: BIBLIOGRAPHY ON ODONATA OF AFRICA: PART 2

Dejoux, C., L. Lauzanne and C. L6v£que 1971. Prospcction hydrobiologique du lac de Lere (Tchad) et des mares
avoisinantes IV. - Faune benthique. Cah. O.R.S.T.O.M., ser. Hydrobiol. 5(2): 179—188. C. *

Dejoux, C. and L. Saint-Jean 1972. Etudes des communantes d’invertebres d’herbiers du lac Tchad: recherches preli-
minaires. Cah. O.R.S.T.O.M., ser. Hydrobiol. 6( 1):67— 83. C. *

Dekeyser, P. L. and A. Villiers 1956. Contribution a l’etude du peuplement de la Mauritanie. Notations ecologiques
et biogeographiqaes sur la faune de l’Adrar. Mem. Inst, franc. Afr. noire 44, 222 pp. N,C. *

Dumont, H. J. 1976. Odonata from South Morocco, Rio de Oro and Mauretania, with biogeographical notes. Odonato-
logica 5(2) : 107— 117. N. ***

Dumont, H. J. 1977a. On migrations of Hemianax ephippiger (Burmeister) and Tramea basilaris (P. de Beauvois) in
West and North-West Africa in the winter of 1975/1976 (Anisoptera: Aeshnidae, Libellulidae). Odonatologica
6(1):13— 17. C. ***

Dumont, H. J. 1977b. Orthetrum abbotti Calvert, 1892, a new Ethiopian representative in the palaearctic fauna (Ani-
soptera: Libellulidae). Odonatologica 6(3):199-203. C,S. **

Dumont, H. J. 1977c. Odonata from Mali, West Africa (Insecta). Rev. Zool. afr. 91 (3) :573— 586. C. ***

Dumont, H. J. 1977d. An analysis of the Odonata of Tunisia. Bull. Ann. Soc. r. beige Ent. H3(l-3):63-94. N. ***
Dumont, H. J. 1978a. On confusion about the identity of Pseudagrion acaciae Foerster, 1906, with the description of
P. niloticum spec, nov., and on the identity of P.hamoni Fraser, 1955 (Zygoptera: Coenagrionidae). Odonato-
logica 7(2): 123-133. A. ***

Dumont, H. J. 1978b. Odonata from Nigeria with special reference to the Air Mountains. Rev. Zool. afr. 92(2):
303-316. C. ***

Dumont, H. J. 1978c. Additions a la faune des Odonates de Mauritanie. Bull. Ann. Soc. r. beige Ent. 1 14( 1— 3):29— 34.

JiyJ £ * * *

Dumont, H. J. 1978d. Odonates d’Algerie, principalement du Hoggar et d’ oasis du Sud. Bull. Ann. Soc. r. beige Ent.
114(4-6): 99- 106. N. ***

Dumont, H. J. 1978e. Les odonates du Parc national du Niokolo-Koba et du lac de Guiers (Senegal) pendant la saison
seche. Bull. Inst. fr. Afr. noire (A) 40(4):847-851. C. ***

Dumont, H. J. 1978f. Neolithic hyperarid period preceded by the present climate of the Central Sahel. Nature
274(5669):356-358. N,C. **

Dumont, H. J. 1979. Limnologie van Sahara en Sahel: bijdrage tot een beter begrip van de klimaats-veranderingen van
het laat-Pleistoceen en Holoceen. Fac. Sci., Univ. Ghent, Gent, 557 pp. N. **

Dumont, H. J. 1980. The dragonfly fauna of Egypt and the role of the Nile in its origin and composition. Water Supply
& Manage. 4(l/2):29-34. N,C. ***

El Rayah, M. 1975. Dragonfly nymphs as active predators of mosquito larvae. Mosquito News 35(2):229— 230. N. ***
El Rayah, M. and F. T. Abu Shama 1977. The predatory capacity of nymphs of the dragonflies: Trithemis annulata
scortecii Nielsen (Odonata: Anisoptera) and Ischnura senegalensis Rambur (Odonata: Zygoptera). Rev. Zool.
afr. 91(2):381— 384. N. ***

El Rayah, M. and F, T. Abu Shama 1978. Notes on morphology and bionomy of the dragonfly Trithemis annulata
scortecii (Odonata, Anisoptera) as a predator on mosquito larvae. Z. angew. Ent. 85( 1 ):81— 86. N. ***
Emsley, M. 1980. Insect Magic. Reader’s Digest (South African edition) 116 (February): 140-147. C. *

Frith, D. W. 1979. A twelve month study of insect abundance and composition at various localities on Aldabra Atoll.
Phil. Trans. R. Soc. Lond. (B) 286(1011): 119-126. I. *

Flint, O. S., Jr., A. S. Menke, F. C. Thompson and R. A. Ward [Eds.] 1976. The United States national entomological
collections. Smithsonian Inst. Press, Washington, D.C., 48 pp. A. *

Forbes, A. T. 1972. Biological survey of the Sundays River with reference to the macro-invertebrate fauna. In: Scott
K. M. F., B R. Allanson and F. M. Chutter, Orange River Project Working Group for ORP hydrobiology of
the Fish and Sundays Rivers. CSIR Res. Rep. 306:29-34. S. *

Forge, P., C. Dejoux and J. M. Elouard 1977. Microdistribution des Invertebres d’un bief. I. le N’zi. Rapp.

O.R.S.T.O.M., Bouake , no. 11, 60 pp. [mimeographed], C. *

Fraser, F. C. 1962. Report on the Austrian Expedition to Madagascar, 1958. Order Odonata. Mem. Inst. sci. Madag.,
Tananarive 13:147-159, 10 figs. I. ***

Fry, C. H. 1972. The biology of African bee-eaters. Living Bird 11:75-112. A. *

Gambles, R. M. 1979a. Paulianagrion platystictoides Fraser and Pseudagrion risi Schmidt (Zygoptera: Coenagrionidae).
Odonatologica 8(3):215— 219. C. ***

Gambles, R. M. 1979b. West African species of Macromia (Odonata: Corduliidae) belonging to the picta and sophia
groups. Syst. Entomol. 4(4):389— 407. C. ***

[Gambles, R. M.] 1980. Odonata. In: Medlar, J. T. ed.. Insects of Nigeria - checklist and bibliography. Mem. Am. ent.
Inst. 30:24-28 excluding bibliography. C. ***

Gardner, A. E. 1968. Notes on insects collected in Madeira in 1966. Bolm Mus. munic. Funchal 22(102):35-37. I. ***
Gee, J. M. 1969. A comparison of certain aspects of the biology of Lates niloticus (Linne) in endemic and introduced
environments in East Africa. In: Obeng, L. E. ed.. Man-made Lakes: The Accra Symosium. Ghana Universi-
ties Press, Accra, pp. 251-260. C. *

Gee, J. M. 1975. The genus Xenoclarias (Pisces, Siluriformes) in Lake Victoria, with a redescription of Xenoclarias
eupogon and notes on the biology of the species. J. Zool., Lond. 175:201-217. C. *

83

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT2, JULY 1982

Green, J. 1979. The fauna of Lake Sonfon, Sierre Leone. J. Zool., Lond. 187:113-133. C. *

Green, J., A. I. el Moghraby and O. M. M. Ali 1979. Biological observations on the crater lakes of Jebel Marra,
Sudan. ./. Zool., Lond. 189(4):493-502. C. *

Greenwood, P. H. 1980. A new species of cichlid fish from the Malagarasi swamps and river (Tanzania, East Africa).

Bull. Br. Mus. nat. Hist. (Zool.) 38(3): 159- 163. C. *

Hammad, S. M. 1974. Additions to the insect fauna of Libya. Bull. Soc. ent. Egypte 58:207-211. N. *

Haritonov, A. Yu. and B. F. Belyshev 1980. [The peculiarities in the geographic distribution of the genus Enallagma
Charp., 1840 (Insecta, Odonata).] [Russian] Trudy biol. Inst. sib. Otdel. Akad. Sci. SSSR, Novosibirsk 40:80
-84. A. **

Harrison, A. D. 1978. Freshwater Invertebrates (except molluscs). In: Werger, M. J. A. ed., Biogeography and Ecol-
ogy of Southern Africa Part 2. Junk, The Hague, pp. 1139-1152. S. **

Hassan, A. T. 1975. Studies on the larval development of Palpopleura lucia lucia, Acisoma panorpoides inflatum and
Urothemis assignata (Anisoptera: Libellulidae) in a semi-natural environment. Nigerian J. Ent. 1(2):143-146.

£ ***

Hassan, A. T. 1976. Studies on the roosting behaviour of Palpopleura lucia lucia (Drury) and Acisoma panorpoides
inflatum Selys (Anisoptera: Libellulidae). Odonatologica 5(4):323— 329. C. ***

Hassan, A. T. 1977a. Longevity of three libellulid dragonflies under semi-natural conditions (Anisoptera: Libellulidae).
Odonatologica 6(1): 1—5. C. ***

Hassan, A. T. 1977b. The larval stages of Urothemis assignata (Selys) (Anisoptera: Libellulidae). Odonatologica
6(3): 151—161 . C. ***

Hassan, A. T. 1978. Reproductive behaviour of Acisoma panorpoides inflatum Selys (Anisoptera: Libellulidae). Odo-
natologica 7(3):237-245. C. ***

Hey, D. 1945. The Establishment and Maintenance of Freshwater Fishes in South Africa. S.A. Forestry Association,
Stellenbosch, 115 pp. S. *

Holden, M. J. 1970. The feeding habits of Alestes baremose and Hydrocynus forskali (Pisces) in Lake Albert, East
Africa. J. Zool., Lond. 161:137-144. C. *

Howard-Williams, C. etal. 1972. Lake Chilwa, Malawi. Studies on a tropical freshwater ecosystem. 40pp. Report to the
IBP/UNESCO Symposium on Synthesis of PF Results. 12-18 Sept. 1972 at Reading, England,
[mimeographed] C. *

Jacquemart, S. 1974. Resultats de la Mission anthropologique Beige au Niger. Collemboles nouveaux du Sahara. Bull.
Inst. r. Sci. nat. Belg. (Ent.) 50(6):l-46, 9 plates. C. *

Jocque, R. 1977. Une etude sur Clarias senegalensis Val. dans la region du Lac Kossou (Cote d’Ivoire). Hydrobiologia
54( 1):49— 65. C. *

Kiauta, B. and G. A. Boon von Ochssee 1979. Some dragonfly karyotypes from the Voltaic Republic (Haute Volta),
West Africa. Odonatologica 8( 1) :47— 54. C. ***

Koch, S. 1979. Libellenfunde in Algerien and Tunisien (Odonata). Entomol. Z. 89(7):77-80. N. ***

Lamoot, E. H. 1977. The food of the damselfly larvae of a temporary tropical pond (Zygoptera). Odonatologica
6( 1 ) :21— 26. C. ***

Larochelle, A. 1977. [The life and work of Philip S. Corbet Odonatologist] [French] Cordulia 3(4):125-130. A. **
Lauzanne, L. 1973. Etude qualitative de la nutrition des Alestes baremoze (Pisces, Characidae). Cah. O.R.S.T.O.M.,
ser. Hydrobiol. 7(1):3-15. C. *

Lavoie, J., M. A. Ali and J.-S. Pilon 1981. Bibliographic analytique de la vision chez les odonates. Odonatologica
10(1) : 5— 28. A. *

Legrand, J. 1975. Contribution a la Faune du Gabon. Odonates: Ire note. Ann. Soc. ent. Fr. (N.S.) ll(4):679-696.

C. ***

Legrand, J. 1976. A propos de Lestes simulans Martin, 1910: larve et imago (Zygoptera: Lestidae). Odonatologica
5(4):375— 381 . C. ***

Legrand, J. 1977a. Deux Tetrathemis Brauer nouveaux du Gabon et la larve de l’un d'eux (Anisoptera: Libellulidae).
Odonatologica 6(4):245-251. C. ***

Legrand, J. 1977b. Description des larves de quatre especes de Calopterygidae du Gabon (Odonata). Ann. Soc. ent.
Fr. (N.S.) 13(3):453— 467. C. ***

Legrand, J. 1979a. Morphologie, biologie et ecologie de Malgassophlebia aequatoris n.sp., nouveau Tetratheminae du
Gabon (Odonata: Libellulidae). Rev. fr. Ent. (N.S.) 1(1 ):3— 12. C. ***

Legrand, J. 1979b. Palpopleura albifrons n.sp., nouveau Diastatopidinae de la foret gabonaise (Odonata Libellulidae).
Rev.fr. Ent. (N.S.) 1 (4 ) : 1 79- 18 1 . C. ***

Legrand, J. 1980. Macromia hervei n.sp. precedemment confondue avec M. aequatorialis (Martin, 1906) (Odonata:
Libellulidae). Rev. fr. Ent. (N.S.) 2(1):10-14. C. ***

LUveque, C. et al. 1972. Recherches sur les biomasses et la productivite du Lac Tchad. In: Kajak, Z. and A. Hillbricht-
Ilkowska eds, Productivity problems of freshwaters. Polish Sci. Publications, Warszawa - Krakow, pp.
165-181. C. *

[Lieftinck, M. A.] 1979. Libellen (Odonata). In: Guldemond, J. A., R. Leys, J G. M. Notenboom and A. W. Wesselo
eds, Biospeleologische Expeditie Algerije 1978. Biospeleologische Werkgroep, Wageningen. p. 108. N. **

84

WILMOT & WILMOT: BIBLIOGRAPHY ON ODONATA OF AFRICA: PART 2

Lindley, R. P. 1976. A new species of Ethiopian Elailoneura Cowley, 1935 from the Central African Republic, and
some notes on the genus (Zygoptera: Protoneuridae). Odonatologica 5(2) : 153— 164. C. ***

Lindley, R. P. 1980. Macromia gamblesi spec. nov. from the Central African Republic (Anisoptera: Corduliidae).
Odonatologica 9(2) : 189—195 . C. ***

Lohmann, H. 1981. Zur Taxonomie einiger Crocothemis - Arten, nebst Beschreibung einer neuen Art von Madagaskar
(Anisoptera: Libellulidae). Odonatologica 10(2): 109— 1 16. I. **

Lounibos, L. P. 1980. Larval odonata in water-containing treeholes at the Kenya coast. Notul. odonatol. 1 (6) :99— 1 00.

C. ***

Machado, A. 1976. Catalogo preliminare de la bibliografia entomologica Canaria. Dept. Zool. Cienc. marinas, Fac.
Ciencias, Univ. de la Laguna, Tenerife, X + 47 pp. I. *

Magadza, C. H. D. 1978. Field observations on the environmental effect of large-scale aerial applications of endosulfan
in the eradication of Glossina morsitans centralis Westw. in the Western Province of Zambia in 1968. Rhod.
J. agric. Res. 16(2):2U-220. C. *

Mamet, J. R. 1978. Contribution a la connaissance de la faune entomologique d' Agalega (Ocean Indien). Bull. Soc.
ent. Fr. 83(5-6) :97-107. I. *

Marlier, G. 1979. Une mission hydrobiologique aux Seychelles. Naturalistes belg. 60(l):44-58. I. *

Marshall, A. G. and R. M. Gambles 1977. Odonata from the Guinea Savanna Zone in Ghana. J. Zool., Lond.
183(2): 177-187. C. ***

Marshall, B. E. 1978. Aspects of the ecology of benthic fauna in Lake Mcllwaine, Rhodesia. Freshwater Biol. 8:241-
249. S. *

Matthes, H. 1964. Les poissons du lac Tumba et de la region d’lkela. Etude systematique et ecologique. Annls Mus. r.

Afr. cent. Ser. 8vo. (Sciences zoologiques) 126:204 pp., 6 plates, 1 map. C. *

McLachlan, A. J. 1965. The ecology of the bottom fauna of Lake Kariba. In: Record of a Kariba Research Symposium
held at the Fisheries Research Institute on 8th-9th June 1965. Lake Kariba Fisheries Research Institute, Kariba,
pp. 58-59. C,S. *

McLachlan, A. J. 1975. The Role of Aquatic Macrophytes in the Recovery of the Benthic Fauna of a Tropical Lake
after a Dry Phase. Limnol. Oceanogr. 20( 1 ):54— 63. C. *

Mead, A. P. 1978. A rhabdocoele turbellarian predator on the aquatic stages of mosquitoes. Ann.trop . Med. Parasitol.
72(6):591— 594. C *

Mok, M. 1975. Biometrie et biologie des Schilbe (Pisces, Siluriformes) du bassin Tchadien. He partie. Biologie com-
paree des deux especes. Cah. O.R.S.T.O.M., ser. Hydrobiol. 9( 1 ) :33— 60. C. *

Noble, R. G. and J. Hemens 1978. Inland water ecosystems in South Africa - a review of research needs. South African
National Scientific Programmes Report No. 34, X + 50 pp.. S. *

Obeng, L. E. 1965. Preliminary observations on the aquatic fauna of the Volta Basin. Ghana J.Sci. 5(3 & 4):273— 274.
C. *

Obeng, L. E. 1969. The invertebrate fauna of aquatic plants of the Volta Lake in relation to the spread of helminth
parasites. In: Obeng, L.E. ed.. Man-made Lakes: The Accra Symposium. Ghana Universities Press, Accra,
pp. 320-325. C. *

Olatunde, A. A. 1978. The food and feeding habits of Eutropius niloticus (Ruppell), family Schilbeidae (Osteichthyes:
Siluriformes) in Lake Kainji, Nigeria. Hydrobiologia 57(3 ) : 197—207 . C. *

Parr, M. J. 1977a. A second species of Pentaphlebia Foerster (Zygoptera: Amphipterygidae), from the Nigerian-
Cameroun border. Odonatologica 6(2):77-82. C. ***

Parr, M. J. 1977b. The present status of Nesciothemis nigeriensis Gambles, 1966 (Anisoptera: Libellulidae) in Northern
Nigeria. Odonatologica 6(4):271-276. C. ***

Parr, M. J. 1980. Territorial behaviour of the African libellulid Orthetrum julia Kirby (Anisoptera). Odonatologica
9( 1 ) :75— 99. C. ***

Parr, M. J. and R. P. Lindley 1980. To Mr Robert Gambles on his 70th Birthday. Odonatologica 9(4):279— 283. C. **
Pinhey, E. C. G. 1966. Odonata. Explor. Parc. natn. Garamba Miss. H de Saeger 45:1-114. C. ***

Pinhey, E. C. G. 1973. Department of Entomology. Rhodesia Sci. News 7(10): 293 . S. **

Pinhey, E. C. G. 1974. Dragonfly collecting on Falcon College expeditions in the Okavango. Falcon, Rhod. 5(3):69— 70.

S. ***

Pinhey, E. C. G. 1975a. Dragonflies: Falcon College Expedition. Falcon, Rhod. 5(4 ) : 63 . S. ***

Pinhey, E. C. G. 1975b. Insects at the water’s edge. Wild Rhodesia 8:21-23. S. *

Pinhey, E. C. G. 1975c. The Insects. In: Phillipson, D. W. ed, Mosi-oa-Tunya. A handbook to the Victoria Falls region.
Longman, Salisbury, pp. 201-218, 14 plates. C,S. *

Pinhey, E. C. G. 1976a. Dragonfly (Odonata) of Botswana, with ecological notes. Occas.pap. Natl. Mus. Rhod., B ,
Nat. sci. 5( 10) :524— 601 . S. ***

Pinhey, E. C. G. 1976b. Further notes on the Odonata of Mauritius. Arnoldia Rhod 8(9): 1—6. I. ***

Pinhey, E. C. G. 1976c. The completion of a research project. Falcon, Rhod. 6( 1 ):71— 72. S. ***

Pinhey, E. C. G. 1976d. Entomofauna from Cabora Bassa. Results of the Entomological Brigade of the IICM. II.

Odonata. Garcia de Orta, Ser. Zool. Lisboa 5(2): 15—24. C, S. ***

Pinhey, E. C. G. 1978a. A new species of Pseudagrion Selys, its separation and comparisons (Odonata: Coenagrioni-
dae). Arnoldia Rhod. 8(22): 1 — 10. C,S. ***

85

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 2, JULY 1982

Pinhey, E. C. G. 1978b. Comparative notes on African species of Trithemis Brauer (Odonata: Libellulidae) and its
cogeners. Arnoldia Rhod. 8(26): 1—8. C,S. ***

Pinhey, E. C. G. 1978c. Odonata. In: Werger, M. J. A. and A. V. van Bruggen eds, Biogeography and Ecology of
Southern Africa. Junk, The Hague, pp. 723-731. S. ***

Pinhey, E. C. G. 1979a. Rediscovery of an elusive S.W. Cape dragonfly (Odonata). Arnoldia Rhod. 8(32):1— 3.

g ** *

Pinhey, E. C. G. 1979b. The status of a few well-known African anisopterous dragonflies (Odonata). Arnoldia Rhod.
8(36): 1-7. A, I. ***

Pinhey, E. C. G. 1979c. Examples of Anisopteran swarms (Odonata). Arnoldia Rhod. 8(37):1— 2. S. ***

Pinhey, E. C. G. 1979d. Additions and corrections to the 1966 checklist to dragonflies (Odonata) from Malawi. Arnol-
dia Rhod. 8(38) : 1 — 14. C. ***

Pinhey, E. C. G. 1979e. Falcon College Expedition to Soutpansberg, December 1978: Entomology: Odonata and Lepi-
doptera. Falcon. Rhod. 6(4): 119—123. S. **

Pinhey, E. C. G. 1979f. Odonata. In: Skaife, S. H., African insect life (2nd edition revised by J. Ledger and
A. Bannister). Struik, Cape Town, pp. 35-41. S. ***

Pinhey, E. C. G. 1980a. A revision of African Lestidae (Odonata). Occas. pap. Natl. Mus. Rhod., B., Nat. sci. 6(6):
327-479. A, I. ***

Pinhey, E. C. G. 1980b. A melanic morph amongst Transvaal Chlorolestidae (Odonata). Arnoldia Rhod. 8(39):1— 4.

^ ***

Pinhey, E. C. G. 1980c. A review of the Metacnemis group (Odonata: Platycnemididae). Arnoldia Zimbabwe 9(2) : 1—13.
A, I. ***

Pinhey, E. C. G. 1980d. Entomology, Odonata and Lepidoptera. Falcon, Zimbabwe 7(1):76— 78. S. **

Pinhey, E. C. G. and F. C. De Moor. 1973. Entomology Reports. Bukwa R. S. E. S., pp. 32-39. S. **

Reynolds, D. R. and J. R. Riley 1979. Radar observations of concentrations of insects above a river in Mali, West
Africa. Ecol.Entomol. 4(2): 161-174. C. *

Riek, E. F. 1976. A new collection of insects from the Upper Triassac of South Africa. Ann. Natal Mus. 22(3):791-820.

S. ***

Robertson, H. M. 1979. Courtship in the Chlorocyphidae (Odonata). Abstr. Pap. Symp. Anim. Communication, Zool.
Soc. Southern Afr., Cape Town. pp. 32-33. S. ***

Russell-Smith, A, and E. Ruckert 1981. The effects of aerial spraying of endosulfan for tsetse fly control on aquatic
invertebrates in the Okavango Swamps, Botswana. Environ. Pollut. Ser. A. 24:57-73. S. *

Rz6ska, J. 1976. Notes on the benthos of the Nile system. In: Rzoska, J. ed., The Nile, biology of an ancient river.
Junk, the Hague, pp. 345-351. N, C. *

Schmidt, E. 1959. Versuch einer Analyse der Libelle Agrion caerulescens Fonsc. 1838 Bull. Inst. Sci. nat. Belg. 35(29):
1-20, 5 figs. N. **

Scott, K. M. F., A. D. Harrison and W. Macnae 1952. The ecology of South African estuaries. Part 2. The Klein
River estuary, Hermanus, Cape. Trans, roy. Soc. S. Afr. 33(3):283-331, 2 plates. S. *

Seaman, M. T. et al. 1978. A limnological investigation of Lake Liambezi, Caprivi. J. Limnol. Soc. sth. Afr.
4(2): 129— 144. S. *

Sevastopulo, D. G. 1980. Possible migration of Deilephila nerii L. (Lep., Sphingidae). Ent. mon. Mag. 115(1376/
1379):76. C. *

Sheppe, W. 1975. Observations on the animal life of some Zambian hot springs. Ohio J. Sci. 75(l):26-29. C. *

Smith, V. W. 1972. Trans-Sahara 1966. Nigerian Field 37( 1):5— 20. N. *

Starmuhlner, F. 1962. Voyage d’ etudes hydrobiologiques a Madagascar, 1958. NaturaUste malgache 13( 1):53— 83. I. *
Starmuhlner, F. 1976. Contribution to the knowledge of the fauna of running waters of Mauritius. Bull. Mauritius Inst.
8(2)405-128, 19 plates. 1. *

Starmuhlner, F. 1977. Contribution to the knowledge of the freshwater fauna of La Reunion (Mascarene). Cah.
O.R.S.T.O.M., ser. Hydrobiol. ll(3):239-250. I. *

Statzner, B. 1975. Zur Longitudinalzonierung eines zentralafrikanischen Fliessgewassersystems unter besonderer Be-
rticksichtigung der Kocherfliegen (Trichoptera, Insecta). Arch. Hydrobiol. 76(2)453-180. C. *

Thiollay, J.-M. 1973. Place des oiseaux dans les chaines trophiques d'une zone preforestiere en Cote-d’Ivoire. Alauda
41(3):273-300. C. *

Tawfik, M. F. S., M. T. Kira and S. M. I. Metwally 1974. A survey of the insect fauna of corn fields in Egypt. Bull.
Soc. ent. Egypte 58:145-152. N. *

Tsacas, L. and J. Legrand 1979. Les pontes d’odonates, gite larvaire nouveau pour une Drosophile africaine inedite:
Drosophila libellulosa n.sp. (Odonata: Libellulidae; Diptera: Drosophilidae). Rev. fr. Ent. 1(1)43-22. C. **
Vidy, G. 1976. Etude du regime alimentaire de quelques poissons insectivores dans les rivieres de Cote d’ Ivoire.

Recherche de l’influence des traitements insecticides effectues dans le cadre de la lutte contre l’Onchocercose.
Rapp. O.R.S.T.O.M., Bouake, no. 2, 36 pp. [mimeographed], C. *

Waterston, A. R. 1976. On the genus Cordulegaster Leach, 1815 (Odonata) with special reference to the Sicilian
species. Trans. R. Soc. Edinb. 69( 19):457-466. N. *

Weaving, A. 1977. Insects, a review of insect life in Rhodesia. Regal, Salisbury, 179 pp. S. **

86

WILMOT & WILMOT: BIBLIOGRAPHY ON ODONATA OF AFRICA: PART 2

Weir, J. S. 1972. Diversity and abundance of aquatic insects reduced by introduction of the fish Clarias gariepinus to
pools in central Africa. Biol. Conserv. 4(3): 169—175. S. *

Youdeowei, A. 1977. A laboratory manual of entomology . Oxford Univ. Press Nigeria, Ibadan, XVII + 208 pp. C. *

ACKNOWLEDGEMENTS

Thanks are due to Mr P. B. N. Jackson of the J. L. B. Smith Institute of Ichthyology,
Grahamstown, for allowing the authors to consult his reprint collection and to Dr F. W. Gess
of the Albany Museum for assistance in resolving bibliographical problems.

REFERENCES

Davies, T. and B. R. Davies 1976. A bibliography of the Chironomidae (Diptera) of Africa and adjacent islands: Part I.

Institute for Freshwater Studies, Rhodes University, Grahamstown, RSA, 15 pp.

Pinhey, E. C. G. 1962. A descriptive catalogue of the Odonata of the African Continent (up to December 1959). Part I.
Publcoes cult. Co. Diam. Angola 59:1-153, 5 plates.

Scott, K. M. F. and Patricia E. Scott 1969. A bibliography of literature on Trichoptera from Africa and adjacent
islands. J. ent. Soc. sth. Afr. 32(3) :399— 41 1 .

Wilmot, B. C. and Leslie P. Wilmot 1978. A selected bibliography of literature on Odonata from Africa and adjacent
islands. Ann. Cape Prov. Mus. (nat. Hist.) 1 1( 10) : 195—208.

World List of Scientific Periodicals published in the years 1900-1960. 4th ed. , Butterworth, London, 1963-1965, 1824 pp.

87

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

BD0410

ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

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

VOLUME 14 • PART 3
26th JULY 1982

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

I !

CO

On the taxonomic status of Heleophryne regis Hewitt, 1909 (Anura:

Leptodactylidae)

RICHARD BOYCOTT

Cape Department of Nature and Environmental Conservation, Jonkershoek Nature Conserva-
tion Station, Stellenbosch, Cape Province, South Africa

CONTENTS

Page

Abstract 89

Introduction 89

Vocalisation 90

Past taxonomic history of Heleophryne regis 93

Review of taxonomic characters 93

Webbing 94

Inner metatarsal tubercle 95

Banding on the hindlimbs 96

Relative body length 96

(a) Tibio-tarsal articulation 97

(b) Body length/hindlimb length ratio 98

Interorbital space 98

Size and sexual dimorphism 99

Habitat and distribution 101

Conclusion 105

Museum abbreviations 107

Acknowledgements 107

References 107

ABSTRACT

Heleophryne regis, which has for some time been regarded as the eatern subspecies of the
H. purcelli complex in the Cape Folded Mountain Belt, has a mating call distinct from typical
H. purcelli and is therefore reinstated to specific rank. It is geographically isolated from other
species in the southern Cape Province. An account of the species’ past taxonomic history is
presented. Certain morphological characters hitherto employed to distinguish H. regis from its
congeners do not withstand critical scrutiny.

INTRODUCTION

The genus Heleophryne, endemic to southern Africa, was described by Sclater in 1899.
The description of the type-species, H. purcelli, was based on a single adult specimen. H. regis

89

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 3, JULY 1982

was also described from one adult specimen by Hewitt in 1909. During the next seventeen
years three further species were described by the same author, viz. H. natalensis, H. rosei and
H. sylvestris (Hewitt, 1913, 1925, 1926). Du Toit (1934) in his revision of the genus recognized
four species H. purcelli, H. rosei, H. regis and H. natalensis ; the last species incorporated
Hewitt's H. sylvestris. Poynton (1964) suggested that H. regis could be merely the eastern sub-
species of a purcelli gradient extending from the south-western Cape eastwards along the Cape
Folded Mountain Belt to the vicinity of Humansdorp. Van Dijk (1966) pointed out that a
number of H. p. regis tadpoles in the Transvaal Museum collection differed from more typical
purcelli from the species’ western range and suggested that the relationships of purcelli and
regis warranted re-examination.

In 1972 the Cape Department of Nature and Environmental Conservation initiated a pro-
ject to establish the conservation status of Cape Province anurans. One member of the genus
Heleophryne, H. rosei , was considered as a candidate for critical examination but its rela-
tionship to other members of the genus had first to be determined. This investigation led to the
findings reported in the present publication.

VOCALISATION

Unequivocal species of frogs are generally recognized by distinctive adult external mor-
phology and unique male mating call, however, morphological distinctness is not an essential
requisite and many anuran species are so similar in appearance that initial recognition was
based solely on the mating call (Littlejohn, 1971). It is now accepted that mating calls in
amphibians are species-specific and that the voice is the most diagnostic feature of any frog
species (Blair, 1958; Pengilley, 1971; Passmore and Carruthers, 1975; Passmore and Car-
ruthers, 1979). Often it is the only character that can be safely used to distinguish between
cryptic species.

Rose (1926) stated for H. rosei that “ . . .on one occasion a H. rosei was heard to give a
chirp, but no other vocal sound has been heard . . .”. FitzSimons (1946) described the call of
H. p. orientalis as “ . . .a somewhat high-pitched, short, clinking whistle (not unlike that of
an Anvil Bat when heard at a distance) . . The call of H. natalensis has been described as
“ . . . a soft melodious, high-pitched, bell-like ’ting’, repeated slowly about ten times, and
sometimes prefaced by a ‘currick’ . . (Wager, 1965). That of H. p. purcelli has been de-
scribed as “. . .a clear, high-pitched ringing note, produced at regular intervals of one per
second . . .” (Passmore and Carruthers, 1979). It is this last call with which the present author
is most familiar. It is heard from September to January in typical habitat of H. p. purcelli.
During 1978 very different “creaking calls” were heard in typical habitat of H. p. regis in the
southern Cape Province. In December 1979 the same calls were heard in the Kareedouw
Mountains where they were emitted by breeding males. Typical calls of H. p. purcelli were not
heard on these occasions. Because of the striking call difference between H. p. regis and H. p.
purcelli the calls were, recorded for analysis and a series of specimens of both “subspecies”
were collected to investigate possible morphological differences between them.

The recordings were made in the field on a Philips Automatic N2203 portable cassette
recorder. The recorded calls were analysed on a sound spectograph (KAY 7029A spectrum
analyser) within the frequency range 80-8000 Hz using a wide-band (300 Hz) filter.

Heleophryne purcelli regis (sensu Poynton) produces a harsh, low-pitched, creaking
(almost rasping) call that differs markedly from the distinctive high-pitched ringing note of
Heleophryne purcelli purcelli of the south-western Cape.

The sonograms of the calls of H. p. purcelli and H. p. regis (Figs 1 to 4) show clearly that
the call of the former species is distributed over a relatively narrow range of frequencies (2-2,5
kHz) whereas that of the latter species is distributed over a wider range (1-2,5 kHz). The

90

FREQUENCY (kHz) 3 FREQUENCY (kHz)

BOYCOTT: TAXONOMIC STATUS OF HELEOPHRYNE REGIS HEWITT, 1909

8-

6-

4-

I I M E (seconds)

1. Heleophryne purcelli — sonogram of 3 successive calls using wideband (300 Hz) filter. (Locality: Donkerkloof,

Klein Drakenstein Mountains).

8-

6-

4-

2 -

*

*

Fig. 2. Heleophryne purcelli — sonogram of 2 successive calls using wideband (300 Hz) filter. (Locality: Uitkyk River,

Cedarberg Mountains).

91

FREQUENCY (kHz) 3 FREQUENCY (kHz)

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 3, JULY 1982

;. 3. Heleophryne regis — sonogram of 2 successive calls using wideband (300 Hz) filter. (Locality: Near Wolfkop,

Kareedouw Mountains).

92

BOYCOTT: TAXONOMIC STATUS OF HELEOPHRYNE REGIS HEWITT, 1909

background noise of rushing water has been left intact in Figs 1 and 2. There is no extensive
background noise in the H. p. regis recordings presented here (Figs 3 and 4).

The call of H. p. purcelli is of short duration. The resulting high-pitched bell-like call
stands out clearly against the background noise of rushing water. In H. p. regis the call is of
longer duration and has a wider frequency range. The call duration varies from 0,03 to 0,05
seconds in H. p. purcelli and from 0,08 to 0,11 seconds in H. p. regis. The interval between
successive calls varies from 0,61 to 0,88 seconds in the former and from 0,93 to 1,10 seconds in
the latter. H. p. purcelli calls are non-pulsatile whereas the calls of H. p. regis are divided into
a series of pulses. In summary, the call of H. p. regis differs substantially from that of H. p.
purcelli in duration, frequency structure and in being pulsatile. It appears therefore that H. p.
regis must be reinstated to specific rank as Heleophryne regis Hewitt, 1909.

The Heleophryne purcelli calls were recorded at Donkerkloof in the Klein Drakenstein
Mountains (3319CC Franschhoek) on 21 October 1977 and at the Uitkyk River in the Cedar-
berg Mountains (3219AC Wuppertal) on 2 November 1978 (Figs 1 and 2). The Heleophryne
regis calls were recorded near Wolfkop in the Kareedouw Mountains (3424AB Clarkson) on 1
December 1979 and at Noghorakloof in the Tsitsikamma Mountains (3324CC Witelsbos) on 14
December 1979 (Figs 3 and 4). On 15 January 1982 a typical call of H. regis was recorded at
the Perdeberg River in the Outeniqua Mountains (3322CC Jonkersberg) and was found to be
no different from the calls recorded above. During an earlier fieldtrip on 30 October 1981 H.
purcelli-\ike calls were heard (but not recorded) at Garcia Pass in the Langeberg Mountains
(3321CC Muiskraal). Although slight differences between the calls of H.p. purcelli and H.p.
orientalis may be detected by detailed analysis, differences between the calls of H.p. orientalis
and H. regis are readily detected as they are vast.

PAST TAXONOMIC HISTORY OF HELEOPHRYNE REGIS

The original diagnoses of H. purcelli (Sclater, 1899) and H. regis (Hewitt, 1909) have been
condensed and presented in Table 1. In many of the characters used the two species are similar
and it is therefore difficult to distinguish them on morphological grounds. Only three of the
characters appear to distinguish the species from one another. These are italicized in Table 1.
Of these, the webbing has received the most attention and was Hewitt’s chief diagnostic char-
acter separating the two species (Hewitt, 1909, 1911, 1913, 1926).

Du Toit (1934) who had more Heleophryne material at his disposal than did previous
workers carried out the first revision of the genus. He clarified certain aspects but at the same
time confused others. In his key to the genus, H. purcelli and H. regis show few differentiating
characters. This is perhaps not surprising as more than half of du Toit’s regis material (exclud-
ing tadpoles) is referable, in the present author’s opinion, to H. purcelli. Only one of du Toit’s
H. regis localities, “Knysna”, falls into the range of H. regis as proposed in the present paper
(15 of his 34 specimens came from this locality). No clear-cut morphological distinction be-
tween the two forms is therefore likely to result from using du Toit’s key.

Poynton (1964), who carried out the second revision of the genus, suggested that H. regis
be considered the eastern subspecies of a H. purcelli dine consisting of, from west to east, H.p.
purcelli, H.p. depressa FitzSimons, H.p. orientalis FitzSimons and H.p. regis. He concluded
that “. . . The forms of Heleophryne occurring in the Cape Folded Range are very similar to
one another, and further collecting in intermediate areas may well make it impossible to dis-
tinguish between these forms . . .”. Poynton’s diagnoses of H.p. purcelli and H.p. regis are
presented in Table 2.

REVIEW OF TAXONOMIC CHARACTERS

A series (n = 25) of adult specimens of H.p. purcelli was compared with a series (n = 25)
of adult specimens of H.p. regis (sensu Poynton) in terms of characters held by Hewitt (1909),

93

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 3, JULY 1982
Table 1.

A comparison of diagnoses for Heleophryne purcelli Sclater and Heleophryne regis Hewitt from
the original descriptions. (The italicized characters are those which appear to hold most prom-
ise for distinguishing the one form from the other).

HELEOPHRYNE PURCELLI

HELEOPHRYNE REGIS

Based on SAM 1313 from Jonkershoek
Snout obtuse

Vomerine teeth in two linear groups
Interorbital space a little broader than upper
eyelid

Tympanum not visible

Fingers free - ending in broad, flat, somewhat
truncate triangular expansions or discs
Toes completely webbed
Terminal discs on toes well developed but not
as large or triangular as those on fingers
Subarticular tubercles oval and fairly large

Inner metatarsal tubercle oval and well de-
veloped

Outer metatarsal tubercle absent
Tibio-tarsal joint reaches the end of snout
Skin smooth above and granular below
Colour purplish with darker round, faintly
white edged spots of the same colour
Hindlimbs cross-barred with darker purple
bands

Based on TM 10093 from Knysna

Snout rounded

Vomerine teeth in two transverse groups

Interorbital space a trifle narrower than upper
eyelid

Tympanum indistinct

Fingers free - ending in triangular discs

Toes half-webbed

Terminal discs not as large as those on fingers

Subarticular tubercles oval and well-
developed

Inner metatarsal tubercle elliptic and of moder-
ate size

Outer metatarsal tubercle absent

Tibio-tarsal joint reaches the end of snout

Skin smooth above and granular below

Colour purplish above with irregular dark
spots

Limbs with indistinct dark cross-bands

Yellowish white below

Yellowish white below

du Toit (1934) and Poynton (1964) to be diagnostic of the two forms. The purcelli series
originated from: Bain’s Kloof (Limietberge); Donkerkloof (Klein Drakenstein Mountains);
Meerluskloof (Riviersonderend Mountains).

Mus. cat. nos.: SAM 1313 (Holotype); AM A Nos. 469, 470, 480-483, 502, 505-509, 511,
513, 569-573, 584-588, 664.

The regis series originated from: Perdeberg River, Deepwalls (Outeniqua Mountains);
Lottering River, Heuningkloof, Sanddrif River, Witelsbos (Tsitsikamma Mountains); near
Wolfkop (Kareedouw Mountains).

Mus. cat. nos.: TM 10093 (Holotype); PEM A Nos. 455, 456; AM A Nos. 516, 517,
521-523, 602, 604-606, 608, 612, 616, 618, 619, 623, 627, 632-635, 652, 653, 700.

Webbing

In the purcelli series males (n = 15) had 2 to V-h (3 in subadults) phalanges of the fourth
toe free of webbing and all females (n = 10) had 2 to 3 phalanges of the fourth toe free of

94

BOYCOTT: TAXONOMIC STATUS OF HELEOPHRYNE REGIS HEWITT, 1909

webbing. In adult males the webbing extends to the tip of the fifth toe or to a point within 1
phalanx of the tip (1 to IV2 in subadults). The females usually have 1 to 2 phalanges of the fifth
toe free of webbing although in exceptional cases webbing may extend to within V2 a phalanx
of the tip. It appears as if juvenile and subadult males have less webbing than adults.

In the regis series all males (n = 15) had 2 to 3 phalanges (exceptionally 3lA) of the fourth
toe free of webbing and females (n = 10) had 3 to 3V2 phalanges of the fourth toe free of
webbing. In males the webbing extends to the tip of the fifth toe or to a point within V2, 1 or
IV2 phalanges of the tip (a variable character in this species). The females had 1 to 2 phalanges
of the fifth toe free of webbing.

In both species the extent of webbing is greater in males than in females (although there is
some overlap between the sexes). Although it is also true that the extent of webbing in H.
purcelli tends to be greater than that in H. regis, it is necessary to point out that the holotypes
of H. purcelli and H. regis are respectively an adult male and an adult female. This could
explain why the extent of the webbing was Hewitt’s chief diagnostic character separating the
two species. However, the overlap between the sexes and between the species renders the
extent of webbing unreliable as a diagnostic character.

Inner metatarsal tubercle

Hewitt (1909) stated that the inner metatarsal tubercle was “oval and well-developed” in
H. purcelli but “elliptical and of moderate size” in H. regis. Poynton (1964) considered the
tubercle to be narrowly elliptical in both forms (Table 2) and made no mention of relative size.
In the fifty specimens examined in the present study no difference in shape was found. The
tubercle of H. regis is slightly smaller than that of H. purcelli, however, it is only a proportion-
ate size difference as H. regis is smaller than H. purcelli. A subadult H. purcelli of the same
size as an adult H. regis does not have a significantly larger inner metatarsal tubercle. The
shape and size of this tubercle is not a clear-cut distinguishing character.

Table 2.

A comparison of Poynton’s (1964) diagnoses for Heleophryne purcelli purcelli Sclater and
Heleophryne purcelli regis Hewitt. (Apparently significant distinguishing characters are empha-
sized by italics).

HELEOPHRYNE PURCELLI PURCELLI
Based on specimens from: Stellenbosh, Frans-
chhoek Pass, Bain’s Kloof, Rivierson-
derend.

HELEOPHRYNE PURCELLI REGIS
Based on specimens from: Knysna, Deep-
walls, Jonkersberg, Humansdorp district.

Tibio-tarsal articulation not or only slightly
extending beyond eye

Outer metacarpal tubercle elliptical, long axis
approximately continuous with axis of first
finger

Inner metatarsal tubercle narrowly elliptical
2 phalanges of 4th toe free of web
Web reaching almost to tip of 5th toe

Tibio-tarsal articulation reaching from eye to
beyond nostril

Outer and inner metacarpal tubercle elliptic-
al, long axis approximately continuous with
axis of first finger

Inner metatarsal tubercle narrowly elliptical
2V. 2 to 3 phalanges of fourth toe free of web
Last phalanx of 5th toe free to nearly free of
web

95

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 3, JULY 1982
Banding on the hindlimbs

Sclater (1899) describes the hindlimbs of H. purcelli as “cross-barred with darker purple
bands”. In all but one of the H. purcelli specimens examined in the present study the cross-
banding of the hindlimbs was distinct (Fig. 5). Hewitt (1909) states that the bands on the
hindlimbs of H. regis are “indistinct” and FitzSimons (1946) describes the hindlimbs as “more

Fig. 5. Heleophryne purcelli Sclater — adult male specimen (AM A 513) from Donkerkloof, Klein Drakenstein Moun-
tains (3319CC Franschhoek).

or less banded”. Poynton’s (1964) photograph of H.p. regis shows clearly defined transverse
bands, contrasting sharply with Hewitt’s observation. In two of the H. regis specimens from the
series of 25 banding was absent, in 10 it was reduced to blotches (Figs 6 and 8) and in the
remaining 13 it was present (being noticeably more distinct in some than in others). Fig. 7
shows the rare, scarcely-banded pattern variant of H. purcelli and Fig. 8 shows a similarly
unbanded H. regis. It is clear that the banding of the hindlimbs cannot be considered to be a
reliable distinguishing character.

Relative body length

Two methods have been used to determine the ratio of body length to hindlimb length in
Heleophryne . The first and more satisfactory method is to measure body length (snout/vent)
and hindlimb length (vent/tibio-tarsal joint) independently. The second is to bend the hindlimb

96

BOYCOTT: TAXONOMIC STATUS OF HELEOPHRYNE REGIS HEWITT, 1909

Fig. 6. Heleophryne reeis Hewitt — adult male specimen (AM A 632) from near Wolfkop, Kareedouw Mountains

(3424AB Clarkson).

round until the tibio-tarsal joint is adpressed to the head region and to relate its position to the
tympanum, eye, nostril or snout. Adpression of the hindlimb (“tibio-tarsal articulation”
method) is the less accurate method since dislocation or breakage of the femur or tibia is not
infrequent and different workers apply differing degrees of pressure. Furthermore, the dis-
tances between tympanum, eye, nostril and snout are often very small.

Sclater (1899), Hewitt (1909) and Poynton (1964) used the tibio-tarsal articulation
method. Their data are therefore not directly comparable with those of du Toit (1934) and
FitzSimons (1946) who used independent measurements of body and hindlimb length.

(a) Tibio-tarsal articulation

Both Sclater (1899) for H. purcelli and Hewitt (1909) for H. regis state that the
tibio-tarsal joint of the adpressed hindlimb “reaches the end of the snout” (Table 1).
For the 15 H. p. purcelli specimens examined by Poynton (1964) the tibio-tarsal joint
is given as “not or only slightly extending beyond the eye” whereas for his 78 H. p.
regis specimens it “reaches from the eye to beyond the nostril” (Table 2).

The tibio-tarsal articulation of the adpressed hindlimb does not extend beyond
the tip of the snout in any of the 25 H. purcelli and 25 H. regis specimens examined
for this paper. In both species, irrespective of sex, the tibio-tarsal articulation may
extend to the eye, to a point between the eye and the nostril, or to a point between
the nostril and the tip of the snout.

97

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 3, JULY 1982

Fig. 7. Heleophryne purcelli Sclater — adult female specimen (AM A 546) from Donkerkloof, Klein Drakenstein Moun-
tains. Note the absence of dorsal patches and limb-banding; this pattern morph is of rare occurrence in this species.

(b) Body length / hindlimb length ratio

Du Toit (1934) stated that H. purcelli has a proportionately greater snout/vent
length than has H. regis. The present study contradicts his findings, however, it must
be remembered that more than 50% of his “regis” material is in fact referable to H.
purcelli. In the females of both species the body length was greater than, or equal to,
the hindlimb length (SV/HL ratio 1,01-1,11 in purcelli; 1,00-1,16 in regis). In the
males of both species, the body length was usually less than, or equal to, the hind-
limb length (SV/HL ratio 0,93-1,05 in purcelli ; 0,95-1,03 in regis). There was no
difference between the two species. However, there is obviously considerable varia-
tion in this character particularly in immature specimens.

The possibility that there is a clear-cut sexually dimorphic distinction is discussed
in the section on size and sexual dimorphism below.

Interorbital space

Sclater (1899) states for H. purcelli that the interorbital space is “a little broader than”
(presumably the breadth of) the upper eyelid. Hewitt (1909), in his description of H. regis,
considers it to be “a trifle narrower than” the upper eyelid, whereas FitzSimons (1946) states
that in H. regis the interorbital space is “broad, equal to or greater than breadth of upper

98

BOYCOTT: TAXONOMIC STATUS OF HELEOPHRYNE REGIS HEWITT, 1909

Fig. 8. Heleophryne regis Hewitt — adult female specimen (AM A 700) from Noghorakloof, Tsitsikamma Mountains.
Note the absence of dorsal patches and limb-banding; this pattern morph appears to be more common in H regis than

in H. purcelli.

eyelid”. For H. p. depressa FitzSimons gives the interorbital space as “narrow, less than
breadth of upper eyelid” and for H. p. orientalis “a little narrower than upper eyelid”.

Sclater, Hewitt and FitzSimons did not describe the method employed to obtain these
measurements. The interorbital distance (= space) could have been measured at the narrowest
point between the orbits or alternatively at the anterior or posterior corners. The breadth of the
upper eyelid has not been accurately defined. It is often difficult to determine exactly where the
orbit merges into the dorsal surface of the head, especially in poorly preserved specimens such
as the holotypes of H. purcelli and H. regis.

In view of these problems the ratio of interorbital distance to eyelid breadth is of doubtful
value as a diagnostic character.

SIZE AND SEXUAL DIMORPHISM

Poynton (1964) gives the maximum snout/vent length for H. p. purcelli as 53 mm and for
H. p. regis as 47,5 mm. The largest specimens (both females) of H. purcelli and H. regis
examined during the present study were 56,3 mm and 49,1 mm respectively. The largest male
purcelli measured 47,3 mm and the largest male regis 42,6 mm. Females of both species attain
a larger size than males (Table 3).

99

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 3, JULY 1982
Table 3.

Comparison of the total length (snout to vent) and hindlimb length (vent to tibio-tarsal joint)
measurements (mm) in Heleophryne purcelli Sclater and Heleophryne regis Hewitt.

Species

Sex

N

Total length

Hindlimb length

Range

Mean

S.D.

Range

Mean

S.D.

H. purcelli

cfcf

15

36,6-47,3

42,60

3,72

35,0-46,4

42,27

3,82

9?

10

40,6-56,3

47,50

5,33

39,0-53,0

45,26

4,50

H. regis

cfcf

15

36,2-42,6

38,40

1,93

36,2-43,0

39,22

1,95

99

10

43,2-49,1

46,72

1,83

40,2-49,2

44,82

2,95

The hindlimb length was expressed as a percentage of the snout/vent length and subjected
to the t-test procedure for a comparison of two sample means (Steel and Torrie 1960). In both
H. purcelli (P < 0,01) and H. regis (P < 0,001) females had proportionately shorter hindlimbs
than males (respective values for both species: t = 3,03 and 4,62; df = 23 and 23).

Although there is some overlap between the sexes it appears that the webbing between the
toes of the hindfeet is more extensive in males than in females (see review of taxonomic
characters). The males of H. purcelli and H. regis can be distinguished from the females by the
presence of a fleshy “anal flap”. The cloacal orifice is ventrally directed in males and posterior-
ly directed in females.

During the breeding season both sexes develop secondary sexual characters which are
more pronounced in males than in females. In males the forearms become thicker and swollen
and bear elongated nuptial pads, dorsal skin folds develop and numerous asperities or spines
appear on the surface of the skin. In some specimens asperities cover the entire dorsal surface.
Asperities may develop on the skin in females but never to the same extent as in males. The
number and stage of development of asperities varies considerably amongst individuals of the
same population and is probably determined by the sexual maturity of each individual. At the
termination of the breeding season a recession of the secondary sexual characters occurs.

The distribution and concentration of asperities in breeding males may be of some use as a
taxonomic character in distinguishing those species which differ markedly from one another
(viz. H. rosei; H. natalensis; H. purcelli/ H. regis). In H. natalensis for example the spines are
concentrated on the fingers and in the axilla (Passmore and Carruthers, 1979) whereas in H.
purcelli and H. regis the asperities are more widely distributed and are concentrated along the
margin of the lower jaw, on the inside surface of the forearms and on the central portion of the
chest. H. purcelli and H. regis individuals at an advanced stage of development of secondary
sexual characters possess concentrations of asperities around the vent and along the outer
edges of the ventral surface between the fore- and hindlimbs. There is no difference in the
distribution and concentration of asperities between H. purcelli and H. regis.

In H. rosei the asperities are concentrated on the outside surfaces of the forearms, on the
back and sides and on the dorsal surfaces of the hindlimbs. A few scattered asperities are
situated along the supratympanic fold. A clearly defined dorsal patch, in which asperities are
absent, extends posteriorly from the nasal region to the sacral region. In H. rosei the asperities
on the outer edges of the upper and lower jaws are, when compared with those of H. purcelli,
poorly developed and are most distinctive in the region between the eye, the angle of the jaw

100

BOYCOTT: TAXONOMIC STATUS OF HELEOPHRYNE REGIS HEWITT, 1909

and the edge of the upper jaw. These observations have been made on a limited number of
individuals and further collections of adults in breeding condition will have to be made for
confirmation.

HABITAT AND DISTRIBUTION

The Cape species of Heleophryne are restricted to mountainous terrain with high rainfall
(600 to 3 000 mm p.a.) and clear, swift-flowing, perennial mountain streams. The length of the
larval phase has never been determined for any species of Heleophryne. However, it has been
suggested that it lasts up to two years (Wager, 1965) and from personal observations is certain-
ly in excess of 12 months. The genus is therefore prevented from colonizing non-perennial
streams.

The rainfall patterns in the south-western and southern parts of the Cape Province differ
considerably from west to east. The western species, H. purcelli, occurs in a predominantly
winter rainfall region. In the southern Cape, H. regis habitats receive rain throughout the year.
The range of H. purcelli (west of the 20° E line of longitude) has a rainfall peak in winter
(June, July and August). The summers in this region are markedly dry. However, from Swel-
lendam eastwards to the eastern limit of the Langeberg range, which corresponds approximate-
ly with the range of H. p. orientals, rain falls throughout the year with peaks in spring and
autumn. The range of H. regis is likewise subjected to year round rainfall with peaks in March/
April/May and August/September/October.

The breeding seasons of H. purcelli and H. regis appear to be influenced by these rainfall
peaks. Although H. purcelli normally breeds from October to January, a few individuals were
heard calling at Dassieshoek (3319DB Koo) on 19 September 1980, which suggests that breed-
ing may commence earlier. The southern Cape species, H. regis , usually breeds from Decem-
ber to February. The breeding season of this species may also start earlier as a single individual
was heard calling in Noghorakloof (3324CC Witelsbos) on 3 November 1980. H. regis breeds
when southern Cape streams are almost at their lowest level. This ensures survival of larvae as
any stream that has water at this time will certainly be perennial. Furthermore the survival of
eggs and young tadpoles would be greater when river flow is reduced.

In the south-western and southern Cape Province the genus Heleophryne is confined to the
Macchia, False Macchia and Knysna Forest veld types (Acocks, 1975). The distribution range
of H. purcelli is included in the Macchia and False Macchia veld types. Although H. regis may
occur marginally in False Macchia the river courses throughout its range are essentially dense-
ly-forested. There may be a greater tendency towards forest-dwelling in H. regis than in H.
purcelli , however, H. p. orientalis does occur in heavily forested gorges and ravines on the
southern slopes of the Langeberg range (e.g. at Swellendam and Grootvadersbosch). All H.
regis material collected to date (from 21 localities) has originated from forested areas. Typical
habitat of H. regis is shown in Fig. 9.

Poynton (1964) lists five localities for H. p. regis. One of these is based on a C.N.D. (Cape
Nature Conservation Department) specimen from “Humansdorp district” and Poynton places
this in the locus 3424BB Humansdorp. The only suitable habitat in the Humansdorp district for
Heleophryne is in the Kromrivier Forest Reserve in the Kareedouw Mountains (3424AB Clark-
son) approximately 45 km west of the town. It can therefore be assumed that the C.N.D.
specimen, in fact, originated from the Kromrivier Forest Reserve. This specimen was never
lodged in a museum and has since been lost (Poynton in litt.). Nevertheless, the presence of
Heleophryne in the Kromrivier Forest Reserve has been confirmed by the recent collection of
material from three localities in the Kareedouw Mountains (Table 4). The Kareedouw Moun-
tains appear to be the eastern limit of H. regis (Fig. 10). The other four localities recorded by
Poynton (Knysna, Deepwalls, Jonkersberg, Montagu Pass) are all acceptable for Heleophryne
regis as understood in the present paper.

101

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 3, JULY 1982

Fig. 9. Habitat of Heleophryne regis — Noghorakloof, Tsitsikamma Mountains (3325CC Witelsbos).

The western limit of H. p. regis is Jonkersberg (Poynton, 1964). This material is clearly
assignable to H. regis as is material collected during the present study from the Perdeberg
River (3322CC Jonkersberg) near the top of the Robinson Pass, approximately 20 km west of
Jonkersberg (Table 4).

The area between Robinson Pass and the Gourits River was surveyed during March 1980
and November and December 1981. Many of the localities visited appeared unsuitable for
Heleophryne with the exception of the Huis River (3322CC Jonkersberg), Attakwaskloof Riv-
er, the upper reaches of the Meul River, the Goliats River (3321DD Attakwaskloof), the
Jakkals River and the Dwars River (3321DC Langeberg). Of these six localities all but the first
provide suitable habitat along only certain stretches. The Huis River is situated approximately
2,5 km south-west of Ruitersbos Forest Station, the Attakwaskloof River on the northern side
of the Attaqua Mountains, the Meul and Goliats Rivers respectively south and south-west of
the Attaqua Mountains and the Jakkals and Dwars Rivers at the eastern extremity of the
Langeberg Mountains. There is in this region a marked difference between the types of habitat
available for Heleophryne. The streams in this region, excluding the Huis River, are not for-
ested and appear more suitable for H. purcelli than for H. regis, if it is assumed that the latter
is more of a forest-dwelling species than H. purcelli. The area between the Gourits River and
the Attaqua Mountains has lower rainfall (500-700 mm p.a.) than the areas to the east and
west which have rainfall figures in excess of 700 mm (800-1 000 mm p.a.). This could account

102

BOYCOTT: TAXONOMIC STATUS OF HELEOPHRYNE REGIS HEWITT, 1909

Table 4.

Distribution records for Heleophryne regis Hewitt

Locality

Locus

Mus. cat. No.

Notes

Knysna, Outeniqua Moun-
tains.

3423 AA KNYSNA

TM 10093

Type-locality

Recorded by Poynton (1964)

Forested stream north of
George, Outeniqua Moun-
tains.

3322CD GEORGE

AM A 720

Collected 30/vii/1972
Poynton (1964) records speci-
mens from this locus.

Forested stream near Jonkers-
berg Forest Station, Out-
eniqua Mountains.

3322CD GEORGE

AM A 436

Collected J. C. Greig
31/xii/1973

Poynton (1964) records speci-
mens from this locus.

Perdeberg River near Robin-
son Pass, Outeniqua Moun-
tains.

3322CC JONKERS-
BERG

AM A 516-525
AM A 703-709
AM A 727-730

Collected 3/i/1977
Collected 14, 15/xii/1981
Collected 15/i/1982
Poynton (1964) records speci-
mens from this locus.

Outeniqua Pass, Outeniqua
Mountains.

3322CD GEORGE

AM A 595

D.O.R. collected 7/viii/1979
Poynton (1964) records speci-
mens from this locus.

Montagu Pass, Outeniqua
Mountains.

3322CD GEORGE

AM A 596
AM A 712-715

Collected 7/viii/1979
Collected 18, 19/xii/1981
Poynton (1964) records speci-
mens from this locus.

Near Wolfkop, Kareedouw
Mountains.

3424AB CLARK-
SON

AM A 597
AM A 632-635

Collected 28/viii/1979
Collected l/xii/1979
New distribution record
The recording illustrated in
Fig. 3 was obtained here.

Tributaries of the Groot Riv-
er, Tsitsikamma Mountains.

3324CC WITELS-
BOS

AM A 601-607

Collected 5, 6/x/1979
New distribution record

Noghorakloof west of Witels-
kop, Tsitsikamma Moun-
tains.

3324CC WITELS-
BOS

PEM A 455, 456
AM A 608-613
AM A 647-650
AM A 700, 701

Collected 6/x/1979
Collected 6/x/1979
Collected 14/xii/1979
Collected 3/xi/1980
The recording illustrated in
Fig. 4 was obtained here.

Sanddrif River, Tsitsikamma
Mountains.

3323DD JOUBER-
TINA

AM A 615-620

Collected 13/x/1979
New distribution record

Lottering River, Tsitsikamma
Mountains.

3323DC NA-

TURE'S VAL-
LEY

AM A 623

Collected 26/x/1979
New distribution record

103

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 3, JULY 1982

Locality

Locus

Mus. cat. No.

Notes

Heuningkloof, Tsitsikamma
Mountains.

3323DD JOUBER-
TINA

AM A 624-627

Collected 27/x/1979

Jack-se-bos, Kareedouw

Mountains.

3424AB CLARK-
SON

AM A 631

Collected 30/xi/1979

Forested stream north of
Clarkson, Kareedouw

Mountains.

3324CD KAREE-
DOUW

AM A 636-646

Collected 2/xii/1979
New distribution record

Tierkloof, Outeniqua Moun-
tains.

3322CD GEORGE

SZC 311

Collected P. W. de Kock
27/xi/1979

Poynton (1964) records speci-
mens from this locus.

Plaat River near Karatara,
Outeniqua Mountains.

3322DD KARA-
TARA

SZC 312, 313

Collected P. W. de Kock
5/xii/1979

New distribution record

Groenkop near Saasveld, Out-
eniqua Mountains.

3322DC WILDER-
NESS

SZC 36

Collected J. C. van Dalen
26/viii/1977

New distribution record

Diepwal-se-Loop near Deep-
walls Forest Station, Out-
eniqua Mountains.

3323CC KRUIS-

VALLEI

AM A 652, 653

AM A 716-718, 721
-724

SAM 45192
SAM 45193

Collected N. I. Passmore
4/M980

Collected 19/xii/1981

Collected 19/xii/1981
Collected 12/i/1982
Poynton (1964) records speci-
mens from this locus.

Swart River near Saasveld,
Outeniqua Mountains.

3322DC WILDER-
NESS

AM A 702

Collected P. J. Louwrens
24/viii/1981

Poynton (1964) records speci-
mens from this locus.

Huis River near Robinson
Pass, Outeniqua Moun-
tains.

3322CC JONKERS-
BERG

AM A 710, 711

Collected 16/xii/1981
Poynton (1964) records speci-
mens from this locus.

Klein-Witelskloof near Keur-
booms River Forest Sta-
tion, Outeniqua Mountains.

3323CD THE
CRAGS

AM A 725

Collected ll/i/1982
New distribution record

Watervalkloof near Jonkers-
berg Forest Station, Out-
eniqua Mountains.

3322CC JONKERS-
BERG

AM A 726

Collected 13/i/1982
Poynton (1964) records speci-
mens from this locus.

104

BOYCOTT: TAXONOMIC STATUS OF HELEOPHRYNE REGIS HEWITT, 1909

for the absence of forested habitat. H. regis has only been found in forested areas that receive
900-1 200 mm of rain per annum.

Tadpoles and newly metamorphosed froglets from the Goliats and Jakkals Rivers and
three adults from the Dwars River were collected in March 1980. In November 1981 three
more adult females were collected from the Dwars River and one subadult from the Jakkals
River. During December 1981 an adult female was collected from the Attakwaskloof River
and a series of 10 tadpoles was collected from the Meul River. During the same fieldtrip
tadpoles and newly metamorphosed froglets and a juvenile frog were collected from the Huis
River. The eight adult specimens from the Dwars, Jakkals and Attakwaskloof Rivers have
been tentatively identified as H. p. orientalis and the juvenile from the Huis River as H. regis.
No adults have been collected from the Goliats River but is is likely that once they have been
discovered they too will prove to be H. p. orientalis. Newly metamorphosed froglets of
H. purcelli are usually found moving out of rivers and streams during March and April. H. regis
froglets, on the other hand, have only been found leaving the river in November, December
and January. If this behaviour pattern is characteristic of the two species, the finding of frog-
lets in the Goliats and Jakkals Rivers would substantiate the assumption that these localities
support H. p. orientalis populations. However, the final proof rests with the acquisition of
tape-recordings of calls from these localities. Until then, the eastern limit of H. p. orientalis
and the western limit of H. regis will remain speculative. At present H. regis calls have been
acquired as far west as the Perdeberg River.

H. regis is endemic to the southern Cape Province. Its known distribution range extends
from the Huis and Perdeberg Rivers near Robinson Pass in west, eastwards along the Outeni-
qua and Tsitsikamma mountain ranges to the Kareedouw Mountains in the east (Fig. 10). A
list of collecting localities for H. regis is provided in Table 4.

CONCLUSION

The genus Heleophryne Sclater is confined to the mountain ranges of South Africa from
the Cape Peninsula to the eastern Transvaal. One species, H. natalensis, occurs in the moun-
tain streams of the Drakensberg and the plateau slopes in Natal and the Transvaal. In the Cape
Province the isolated nature of the folded mountain belt on with the genus occurs has given
rise to considerable taxonomic complexity (Poynton and Broadley, 1978). Considering
topography, rainfall and other environmental factors it becomes apparent that ghost frogs are
restricted to a specific type of habitat. As these mountains are not continuous, speciation in
allopatry has occurred, giving rise to the four species recognized in the present paper. A new
species is currently being described form the isolated Elandsberg range north-west of Port
Elizabeth (Boycott, in prep.) and an isolated and possibly distinct form has recently been
discovered in the Kammanassie Mountains. The morphological differences between the species
are not conspicuous and the composition of the Cape species can be regarded as a complex of
sibling species.

On the various forms occurring in South Africa, H. rosei and H. natalensis are the most
distinctive and are well separated geographically and morphologically from the H. purcelli /
H. regis complex. H. rosei is separated from H. purcelli by the Cape Rats and H. natalensis is
separated from the species in the southern Cape mountains by a swathe of unsuitable habitat in
the eastern Cape Province. Although differences between the tadpoles of H. purcelli and
H. regis have been suspected (van Dijk, 1966), definitive morphological and behavioural differ-
ences between the two forms have yet to be detected.

The disjunct distribution of ghost frogs in South Africa is probably a reflection of their
specialised habitat requirements. The question posed by Watson, Loftus-Hills and Littlejohn
(1971) as to whether disjunct allopatric forms would remain distinct, if their ranges came into
contact, is also appropriate in this case. Telford and Passmore (1981) have demonstrated the

105

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 3, JULY 1982

Fig. 10. The distribution of Heleophryne Sclater in South Africa.

# Heleophryne rosei Hewitt, □ Heleophryne purcelli Sclater,

■ Heleophryne regis Hewitt, O Heleophryne natalensis Hewitt.

The distribution records for H. natalensis have been taken from Poynton (1964).

importance of the male mating call as the basis for conspecific recognition. Because of their
role in maintaining species distinctness, pre-mating isolating mechanisms are the most impor-
tant taxonomic characters that a species possesses (Littlejohn, 1968). In H. purcelli and
H. regis no substantial morphological differences are apparent. Littlejohn (1968) goes on to say,
however, that this does not necessarily mean that they are not both good biological species
since the distinction depends on their basic genetic differences and their failure to interbreed in
nature. Morphological differences, unless associated with ethological isolation, are of secon-
dary taxonomic importance at the species level (Littlejohn, 1968). The fact that H. purcelli
and H. regis have little or no apparent morphological differences suggests that they be
classified as sibling species. The call difference described in the present paper is at this stage
the only reliable character separating them unequivocally from one another.

106

BOYCOTT: TAXONOMIC STATUS OF HELEOPHRYNE REGIS HEWITT, 1909

MUSEUM ABBREVIATIONS

AM Albany Museum, Grahamstown, South Africa.

PEM Port Elizabeth Museum, Port Elizabeth, South Africa.

SAM South African Museum, Cape Town, South Africa.

SZC Saasveld Zoological Collection, Saasveld Forest Research Station, George, South Afri-
ca.

TM Transvaal Museum, Pretoria, South Africa.

ACKNOWLEDGEMENTS

Thanks are due to the Department of Nature and Environmental Conservation of the
Cape Province for making this study possible. The assistance and co-operation of the Depart-
ment of Forestry, especially the Regional Directors of the Tsitsikamma, Southern Cape and
Western Cape regions, are also gratefully acknowledged. The Cape Department of Nature and
Environmental Conservation and the Board of Trustees of the Port Elizabeth Museum are
thanked for financial support. The author is particularly indebted to John Greig (formerly
Herpetologist, Cape Department of Nature Conservation). His frequent discussions were al-
ways stimulating and his advice and guidance were of tremendous assistance in the writing of
this paper. The author benefitted from discussions with W. R. Branch, C. J. Burgers, H. G.
Cogger, A. L. de Villiers, R. A. Haynes, F. J. Kruger, M. J. Littlejohn, P. H. Lloyd, A. A.
Martin, and M. J. Tyler. Messrs. W. Bond, C. Burgers, A. de Villiers, W. J. Frey, E. Groom-
er, R. Haynes, D. Heard, A. Jones, C. McDowell, H. Swanevelder and J. van der Velden are
thanked for assistance in the field. N. I. Passmore made the sonograms and was generous with
advice. J. C. Greig and N. I. Passmore read earlier drafts of the manuscript. Finally the author
would like to thank the Director of Nature and Environmental Conservation of the Cape
Province for permission to publish this paper.

REFERENCES

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

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

Boycott, R. C. (in prep.). Description of a new species of Heleophryne Sclater, 1899 from the Cape Province, South
Africa (Anura: Leptodactylidae).

Du Toit, C. A. 1934. A revision of the genus Heleophryne. Ann. Univ. Stellenbosch 12: 1-26.

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. Transv. Mus. 20 (4): 351-377.

Hewitt, J. 1909. Description of a new frog belonging to the genus Heleophryne and a note on the systematic position of
the genus. Ann. Transv. Mus. 2: 45-46.

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 Mas. 2: 189-228.

Hewitt, J. 1913. Description of Heleophryne natalensis , a new batrachian from Natal; and notes on several South
African batrachians and reptiles. Ann. Natal Mus. 2: 475-484.

Hewitt, J. 1925. On some new species of reptiles and amphibians from South Africa. Rec. Albany Mus. 3: 343-368.
Hewitt, J. 1926. Descriptions of some new species of batrachians and lizards from S. Africa. Ann. Natal Mus. 5 (3):
435-448.

Littlejohn, M. J. 1968. Frog calls and the species problem. Aust. Zoo!. 14 (3): 259-264.

Littlejohn, M. J. 1971. Amphibians. In Amphibians and Reptiles of Victoria, pp. 1-11. Reprinted from the Victorian
Year Book No. 85.

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

Passmore, N. 1. and Carruthers, V. C. 1979. South African Frogs. Johannesburg: Witwatersrand University Press.
Pengilley, R. K. 1971. Calling and associated behaviour of some species of Pseudophryne (Anura: Leptodactylidae).
J. Zool. Lond. 163: 73-92.

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

Poynton, J. C. and Broadley, D. G. 1978. In Herpetofauna. In Biogeography and Ecology of southern Africa,
M. J. A. Werger, ed., pp 925-948. Dr W. Junk The Hague.

107

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 3, JULY 1982

Rose, W. 1926. Some field notes on the Batrachia of the Cape Peninsula. Ann. S. Afr. Mus. 20: 433-450.

Sclater, W. L. 1899. List of the reptiles and batrachians of South Africa, with descriptions of new species. Ann. S. Afr.
Mus. 1: 95-111.

Steel, R. G. D. and Torrie, J. H. 1960. Principles and procedures of statistics. New York: McGraw-Hill Book Co., Inc.
Telford, S. R. and Passmore, N. I. 1981. Selective phonotaxis of four sympatric species of African reed frogs (Genus
Hyperolius) . Herpetologica 37 (1): 29-32.

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.

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

Watson, G. F., Loftus-Hills, J. J. and Littlejohn, M. J. 1971. The Litoria ewingi complex (Anura: Hylidae) in
south-eastern Australia. 1. A new species from Victoria. Aust. J. Zool. 19: 401-416.

108

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

BD0429

QH

I ^

m

ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

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

29th OCTOBER 1982

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

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

BD0759

Three new species and a new genus of tripterygild fishes (Blennioidei) from

the Indo-West Pacific Ocean

by

WOUTER HOLLEMAN

Albany Museum, Grahamstown, South Africa

CONTENTS

Page

Abstract 109

Introduction 109

Methods 110

Description of species 115

Helcogramma McCulloch & Waite 115

Helcogramma fuscopinna sp. nov 115

Enneapterygius Riippell 120

Enneapterygius clarkae sp. nov 121

Enneapterygius ventermaculus sp. nov 123

Cremnochorites gen. nov 125

Cremnochorites capense (Gilchrist & Thompson) comb, nov 131

General Discussion 132

Acknowledgements 136

Literature Cited 137

ABSTRACT

A new species of Helcogramma, H. fuscopinna, which ranges across the Indo-West Pacific
(excluding the Red Sea) from as far south as Durban to southern Japan, is described. Two new
species of Enneapterygius, E. clarkae from the western Indian Ocean including the Red Sea,
and E. ventermaculus from Zululand, Aden and Pakistan, are described. The possibility that
Enneapterygius may not be monophyletic is discussed. Gillias capensis (Gilchrist & Thompson)
is assigned to a new genus, Cremnochorites , and the species is redescribed. Reference is made
to certain osteological features of the family, particularly the presence of a free pterygiophore
between the second and third dorsal fins and the occurrence of a septal bone in certain tripte-
rygiid genera.

INTRODUCTION

Recent collecting of inshore fishes from the Zululand coast revealed the presence of sev-
eral undescribed species in the family Tripterydiidae. These collections formed the basis for a
revision of the South African species by Holleman (1978).

With the exception of Clark’s (1979) revision of the Red Sea species and Holleman’s
(1978) unpublished dissertation of the South African taxa, little is known of the taxonomy of
the family. Confusion abounds, particularly in the assignation of nominal Indo-Pacific species
and genera.

Pending revisions of the two largest genera, Helcogramma and Enneapterygius which are
currently in progress (by Hanson and Holleman respectively), the new species are here de-

109

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 4, OCTOBER 1982

scribed without providing keys to the species of each genus. Diagnoses are provided for the
genera, and the species are distinguished from apparently closely related species. Comments
are also made on the status of Enneapterygius since an investigation of the osteology of this
genus (Holleman, 1978) casts some doubt on its monophyly.

Gillias capensis (Gilchrist & Thompson) was placed in Gillias Everman & Marsh, 1899 by
Barnard (1927). Rosenblatt (1960) synonynrised Gillias with Enneanectes but did not refer any
of the South African or Australian species ascribed to Gillias to other genera. Gillias capensis
is not referable to Enneanectes. The latter genus is reputedly endemic to the eastern Pacific
Ocean and Caribbean (Rosenblatt, 1960). Enneanectes lacks the characteristic scalation of G.
capensis. Clark (1979) suggests that G. capensis may be referable to Norfolkia Fowler. This is
not the case, however, as the two genera differ considerably in osteology and scalation. A new
genus, Cremnochorites , is thus proposed for G. capensis. Cremnochorites appears to be mono-
typic and endemic to the southern coast of South Africa.

METHODS

The methods of taking measurements follow those outlined by Hubbs & Lagler (1958).
All fin elements were counted following Rosenblatt (1960) and Springer (1968). The last dorsal
and anal fin rays are almost without exception divided to the base and are counted as a single
element. The caudal fin of tripterygiids, with the exception of the highly specialized Notoclinus
which has ten principal rays, invariably has seven dorsal and six ventral segmented principal
rays, with the upper- and lowermost rays unbranched and the remainder bifurcate. There is
also a variable number of dorsal and ventral unsegmented procurrent rays.

All species here considered have two undivided segmented rays and one short hidden
spine in each pelvic fin. The rays are united by a membrane for part of their length. The inner
ray is always the longest.

The following measurements and counts are given in Tables 1-3: standard length (SL);
head length; horizontal eye diameter; snout length; upper jaw length; snout angle (measured
as shown in Fig. 1 — the “angle of head profile” of Zander & Heymer (1979)); counts for all
fins except pelvic fins; number of caudal and precaudal vertebrae; lateral line counts; trans-
verse scales, and total lateral scales.

110

HOLLEMAN: THREE NEW SPECIES AND NEW GENUS OF TRIPTERYGIID FISHES (BLENNIOIDEI)

The numbers of caudal and precaudal vertebrae were determined from cleared and
stained specimens and from radiographs. The choice as to which vertebra to count as the first
caudal vertebra poses a problem. Springer (1968) considered the first vertebra bearing a well-
developed haemal spine as the first caudal vertebra for Blenniidae, where the first proximal
anal pterygiophore is associated with the first haemal spine. Certain authors consider the first
vertebra with closed haemal arch as the first caudal vertebra. The first “well-developed” hae-
mal spine of most tripterygiids is distally forked (Fig. 2) and the subsequent centrum has a con-
ventional haemal spine (Fig. 2). The first two proximal anal pterygiophores fall between the
forked haemal spine and the first proper haemal spine and the third pterygiophore is associated
with the latter. The first closed haemal arch is usually coincident with the forked haemal spine
but closure may occur in one or two preceding vertebrae. As closure of the haemal arch is ex-
tremely difficult to determine from radiographs, it is expedient to consider the centrum with
the forked haemal spine as the first caudal vertebra. This is most easily determined as the ver-
tebra with a haemal spine immediately anterior to or in contact with the first anal pterygio-
phore. This criterion for the first caudal vertebra is thus adopted here.

Clark (1979) does not state which vertebra she considers the first caudal vertebra. Ver-
tebral counts given by Clark and herein are therefore not necessarily comparable.

Where the lateral line is divided into two portions, one of pored and one of notched
scales, these are referred to as “anterior” and “posterior series”. Counts of total lateral scales
were made from the first scale in the pectoral fin axil to the last scale on the caudal peduncle,
not counting scales on the caudal fin. Transverse scale counts were taken along a diagonal from
the first spine of the second dorsal fin to the base of the anal fin and the count is given as xly
where x is the number of scales above the lateral line and y the number of scales below the
lateral line.

Osteological observations were made on specimens cleared and stained by the trypsin-ali-
zarin technique of Taylor (1967).

Abbreviations — BMNH = British Museum of Natural History; BPBM = Bernice P.
Bishop Museum; LACM = Los Angeles County Museum of Natural History; RlJSf = J. L. B.
Smith Institute of Ichthyology; SAM = South African Museum; USNM = United States Nat-
ional Museum of Natural History.

Table 1

Frequency distribution of number of vertebrae.

Precaudal

Caudal vertebrae

Species

9

10

11

22

23

24

25

26

27

28

Helcogramma

12

3

8

1

fuscopinna

Enneapterygins

10

2

8

clarkae

E. ventemaculus

12

1

3

8

Cremnochorites

23

13

10

capense

Ill

Frequency distribution of meristic data.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 4, OCTOBER 1982

Anal

22

<N

SO

m (N

20

NO (N

ON

2

42

00

i n ,

r-

m i

NO

ON

<n

a

25

74

22

60

Third Dorsal

<— < (S rj-

NO <N

O

t— < rt (N

1

ON

oo m

00

CO H

Second Dorsal

XV

XIV

68

23

XIII

rf 00

XII

OO

r-H rt

><

T—t Tf

First Dorsal

>

26

III

75

22

60

Helcogramma

fuscopinna

Enneapterygius

clarkae

E. ventemaculus
Cremnochorites
capense

0

24

—

O ^

23

<N 2

22

2 O <N

<N

m (N t-h

20

r-H (N

ON

CN

OO

(N

r-

m

NO

-

~28]

-

0

(N

~26j

r-

25

r-

24

17

3

23

r- On

22

Tt IT)

r^i

m

o

ON

00

r-

i

NO

00

i/~>

27

m

t-h m

<N

(N

NO

Helcogramma

fuscopinna

Enneapterygius

clarkae

E. ventemaculus
Cremnochorites
capense

w

cq

<

H

112

HOLLEMAN: THREE NEW SPECIES AND NEW GENUS OF TRIPTERYGIID FISHES (BLENNIOIDEI)

Fig. 2. Last precaudal and first two caudal vertebrae of Enneapterygius clarkae with associated anal pterygiophores.

113

io<y i to* i4Q°

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 4, OCTOBER 1982

3

S

T3

a>

c

<u

-c

C/D

T3

'5b

Er

<U

o.

•c

CO

6

£

114

ventermaculus ; *-E. clarkae; QCremnochomes capensis

HOLLEMAN: THREE NEW SPECIES AND NEW GENUS OF TRIPTERYGIID FISHES (BLENNIOIDEI)

DESCRIPTION OF SPECIES
Genus Helcogramma McCulloch & Waite

Helcogramma McCulloch & Waite 1918: 51 (Type-species, Helcogramma decurrens Mc-
Culloch & Waite, 1918, by original designation).

Diagnosis

First dorsal fin with three spines; anal fin with one spine. Lateral-line a continuous series
of pored scales curving down behind pectoral fin base and continuing midlaterally to below sec-
ond or third dorsal fin or onto caudal peduncle. Head naked, nape of some species scaled;
body, except abdomen and pectoral fin bases, with ctenoid scales. Scalation at base of first and
second dorsal fins reduced in some species. Two or three rows of conical teeth on vomer and
anterior ends of palatines.

Description (Characters in diagnosis not repeated)

Dorsal fins III+X-XVI + 6-12. Anal fin 1+14-21. Pectoral fin rays 15-17, lowermost 6-7
simple and thickened, uppermost 1-5 simple, remainder bifurcate. Caudal fin 8-10 dorsal,
7-10 ventral procurrent rays. Small, simple orbital and anterior nasal tentacles usually present.
Mandibular sensory canals confluent, opening as single or double pore just posterior to lower
jaw symphysis. Lateral-line a continuous series of 17-37 pored scales. Jaws with slightly re-
curved, fixed teeth in bands in front, decreasing posteriorly to a single row; teeth in outer rows
enlarged; 2-3 rows of conical teeth on vomer and anterior ends of palatines. Septal bone (see
under General Discussion) present; cephalic lateralis canals not covered by bone; hypural 5
small and unossified, one or two epurals.

Discussion

Helcogramma is confined to the Indo-Pacific Oceans as far east as Hawaii. There appear
to be in excess of 17 species which can be ascribed to the genus. Clark (1979) states that about
10 species are known from the Indo-Pacific and Red Sea. Most of Fowler’s (1964; 1958)
species, which he placed first in Enneapterygius and then Tripterygion , appear to belong in
Helcogramma.

Helcogramma fuscopinna. sp. nov. Fig. 4.

Diagnosis

Second dorsal fin usually with 14 rays; anal fin usually with 21 rays. First two dorsal fins
and anal fin conspicuously dark to black; a distinct blue-white line finely stippled with melano-
phores extends from upper lip, below eye, to posterior margin of preopercle.

Description (characters for holotype in parenthesis).

Dorsal fins 1II+X11I-XV+10-11 (III+XIV+11), usually III+XIV+11. First dorsal fin
slightly lower than, or equal to, second dorsal, slightly higher in males than in females; anal fin
1+19-21(1+21), usually 21; pectoral fin 16-17 upper 3-4 simple, lowermost 7 simple and thick-
ened, remainder bifurcate; caudal fin 9-10,7+6,9-10; pelvic fin rays united by membrane for
half to third of their length. Lateral-line 22-33 (25), usually 24-25, pored scales ending below
front of third dorsal fin; transverse scales 6/8; lateral scales 38-41 (40), usually 39. Precaudal
vertebrae 10; caudal 26-28, usually 27. Head 3,2-3,7 in SL; eye 2, 6-3, 4, upper jaw 2, 0-2, 4,
snout 2, 9-4,2 in head. Snout angle 68°-73°. Body scales do not extend to bases of first and
second dorsal fins.

Colour

Freshly caught adult males with orange-pink body and conspicuously dusky anal and first
two dorsal fins. Scales generally with row of small melanophores along posterior margin. Small

115

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 4, OCTOBER 1982

dusky rosettes scattered over body, generally more densely below midline. Darkly pigmented
specimens have a row of five or six grey-white blotches stippled with small melanophores
above and below midline. These may produce faint vertical banding. There may also be two
pairs of white blotches on either side of dorsum at posterior ends of second and third dorsal
fins. Distinct blue-white line stippled with very fine melanophores extends from upper lip, be-
low eye to posterior edge of preopercle. This line may be continued as two or three spots on
opercle and upper pectoral fin base. Head below blue-white line, throat and chest heavily
stippled with dusky to black spots and rosettes; head above line pale. Orbital tentacle black;
nasal tentacle unpigmented. Nape and interorbital area pinkish. First two dorsal fins heavily
stippled; stippling on third dorsal lighter, occurring mainly on the fin margin. Anal fin darkly
pigmented; caudal fin slightly darker than third dorsal but not as dark as second. Melano-
phores decrease in size from fin base to margin and grade from dark brown at base to black at
margin. Lower half of pectoral fyis darker than upper, with oblique white blotch on middle of
base; melanophores on lower part of base tend to form a large dark blotch. Pelvic fins finely
stippled, darker basally than distally.

Juveniles and small females are virtually immaculate; larger females have lower half of
body lightly dusted with small melanophores, dusky margins to anal and first two dorsal fins
and stippling on lower part of pectoral fin base, mid-opercle, cheek, snout and nape. Orbital
tentacle dark. Sometimes white blotches at posterior ends of second and third dorsal fins.

10 mm

Fig. 4. Helco gramma fuscopinna sp.n. RUSI 954. Holotype, male 37,6 mm SL.

Material examined

Holotype — RUSI 954 (out of former RUSI 77-18) male, 37,6 mm SL; reef offshore
Sodwana Bay (27° 31'S, 32° 41'E), Zululand, South Africa: depth 10 m; June, 1977;
M. S. Christensen, W. Holleman, W. Devos; field number MSC 77-18.

Paratypes — RUSI 955 (2, 36,5 & 40,0 mm SL), taken with holotype. RUSI 956 (6,
32,0-41,7 mm SL), Bazaruto Isl., Mozambique; Sept., 1954; J. L. B. & M. M. Smith.

RUSI 489 (15, 26,0-36,9 mm SL), gully with vertical sides offshore Sodwana Bay; depth
10 m; 28 July, 1976; R. Winterbottom, M. S. Christensen, A. E. Louw (field number
RW 76-21).

RUSI 190 (28, 22,0-43,0 mm SL), as preceeding — remainder of former RUSI 76-8.

BPBM 21164 (5, 31,5-41,3 mm SL), reef 1 km off north of Sodwana Bay; depth 10-12 ip;
21 June, 1977; J. E. Randall, M. S. Christensen.

116

HOLLEMAN: THREE NEW SPECIES AND NEW GENUS OF TRIPTERYGIID FISHES (BLENNIOIDEI)

BMNH 1978.5.30. 1-2 (2, 31,4 & 40,5 mm SL) patch reef, 1 km offshore Sodwana Bay;
depth 10 m; July, 1976; R. Winterbottom, M. S. Christensen; A. E. Louw (both out of former
RUS1 76-8).

USNM 227738 (4, 28,5-39,9 mm SL); southwest shore just off Ch’uan-fan-shih, Taiwan
(21° 55' 48"N, 120° 48' 48"E); depth 8-9 m; 3 May, 1968; V. G. Springer et al.; field number
VGS 68-14.

USNM 227740 (33, 28,9-39,1 mm SL); rocky reef with some live coral, course sand bot-
tom off northern tip of St Brandon’s Shoals (16° 25'S, 59° 36'E); depth 6-11 m; 6 April, 1976;
V. G. Springer et al.; field number VGS 76-10.

USNM 227741 (33, 13,7-36,3 mm SL); dead coral channels, about half mile SW of tip of
North Island, St Brandon’s Shoals (10° 9'S, 56° 35'E); depth 0-8 m; 17 April, 1976;
V. G. Springer et al.; field number VGS 76-24.

USNM 227742 (3, 35,0-36,4 mm SL); west side Apo Island about | km north of South
end, Philippine Islands (09° 04' 25"N, 123° 16' 05"E); depth 0-6 m; 6 June, 1978; V. G. Spring-
er et al.; field number SP-78-34.

USNM 227743 (3, 33,2-40,3 mm SL); blind surge channel with many small caves, 100
yards off west side of Raphael, St Brandon’s Shoals (16° 26'S, 59° 36'E); depth 0-8 m, 2 April,
1976; V. G. Springer et al.; field number VGS 76-6.

USNM 227744 (2, 31,7 & 39,3 mm SL); northwest side (Cuyo Island) Putic Island, Pala-
wan, Philippine Islands (10° 55' 05"N, 121° 02' 03"E); depth 0-4,6 m; 22 May, 1978; V. G.
Springer et al.; field number SP-78-18.

USNM 227746 (19, 14,4-30,7 mm SL); about 1 km south of west of north end of North Is-
land, Agalega Island, St Brandon’s Shoals; depth 6-8 m; 19 April, 1976; V. G. Springer et al.;
field number VGS 76-29.

ROM 38782 (4, 27,3-32,8 mm SL), patch reef, inshore off Isle Fouguet, Peros Banhos,
Chagos Archipelago (05° 26' 40''S, 71° 41' 02"E); depth 0,5-4 m; 21 February, 1979; R. Win-
terbottom et al.; field number WE 79-33.

ROM 38783 (1, 21,2 mm SL), steep drop off with mixed coral and sand, on lagoon side of
Isle Mapua, Peros Banhos, Chagos Archipelago (05° 26' 44"S, 71° 47' 42"E); depth 3-7 m; 6
March, 1979; R. Winterbottom et al.; field number WE 79-55.

ROM 38784 (2, 30,9 & 23,6 mm SL), reef off Isle Anglaise, Peros Banhos, Chagos Archi-
pelago (05° 24' 40"S, 71° 46' 12"E); depth 5-7 m; 8 February, 1979; R. Winterbottom &
A. Emery; field number WE 79-10.

ROM 38785 (1, 26,8 mm SL), spur and groove formation, reef top on ocean side of SW
tip of Isle Boddam, Salomons, Chagos Archipelago (05° 21' 05"S, 72° 12' 12"E); depth 0-3 m;
18 March, 1979; A. Emery et al. ; field number WE 79-78.

WAM P26507-016 (2, 32 & 35 mm SL), Ko Similan, Similan Islands (8° 40'N, 97° 38'E);
depth 3 m; 13 February, 1979; G. Allen and R. C. Steene.

USNM 227747 (91, 24, 4-38, 4 mm SL), face of channels of reef along SE side of Grande
Passe, St Brandon’s Shoals (16° 28' S, 59° 40'E); depth 0-3 m; 5 April, 1976; V. G. Springer et
al.; field number VGS 76-9.

USNM 227739 (1, 43,0 mm SL), bay with rock and coral, SE of K’enting, SE Taiwan;
depth 0-3 m; 22 April, 1968; V. G. Springer et al.; field number VGS 68-1.

USNM 227745 (9, 29,8-40,8 mm SL), SW shore just off Ch’uan-fan-shih, Taiwan; depth
5-7 m; 28 April, 1968; V. G. Springer et al. ; field number VGS 68-9.

Comparisons

There are six other species of Helcogramma described from the Red Sea, Indonesia and
the Indian Ocean: H. ellioti (Herre, 1944), H. indicus Talwar & Sen (1971), H. obtusirostre
(Klunzinger, 1871), H. shinglensis Lai Mohan (1971), H. steinitzi Clark (1979), H. trigloides

117

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 4, OCTOBER 1982

(Bleeker, 1858). Salient comparative features of these six species and H. fuscopinna are given
in Table 4.

Herre (1944) provided a detailed description of the colour of H. ellioti. In males the head
and trunk are brilliant blue ventrally (other species except H. trigloides, are dark brown to
black); anal fin brilliant blue, caudal and dorsal fins lighter blue, and pectoral fin base with a
bright blue ocellus outlined in golden red. This essentially agrees with the colour pattern of
H. trigloides as described by Day (1876) from illustrations by Elliot. Bleeker (1858) gives an
anal fin count of 10 (rays?) for trigloides whereas Day recorded 18-20 and de Beaufort &
Chapman (1951) 19 rays. Talwar & Sen (1971) intimate that the species described by Day as
trigloides was indicus. These authors do not give a source for their data for trigloides. The dis-
tinctive colour pattern of trigloides and ellioti does, however, distinguish them from other
species.

H. fuscopinna is the only species with a norm of 14 spines in the second dorsal fin. It ap-
pears to be similar in colour pattern to H. steinitzi but lacks the scaled nape. It can also be
clearly distinguished from H. obtusirostre by colouration (particularly the distinctive stipple
line below the eye), relative eye size (the eye of H. fuscopinna is relatively larger than that of
H. obtusirostre — Fig. 5) and snout angle (Table 1). There appears to be little to distinguish
H. shinglensis from H. obtusirostre. Lai Mohan’s (1971) description of H. shinglensis is based
on only three specimens. This species is very similar in meristic data and colour pattern to
South African H. obtusirostre and it is possible that H. shinglensis is referable to H. obtusi-
rostre. Pending revision of the genus no further comments can be made on the status of the
other species.

118

Comparative features of seven species of Helcogramma from the western and central Indian Ocean.

HOLLEMAN: THREE NEW SPECIES AND NEW GENUS OF TRIPTERYGIID FISHES (BLENNIOIDEI)

(ucrndBiQ ?y
jjojnEag ap uiojj)
(J33)33ig uiojj)
sapio/Sw H

13 13

10 10

10 19

50? 50

4

2,5 2,8

naked

suauipads p

SlSU3]§UlljS /-J

12,13

9

20

20-22
3,0-3, 8
2, 7-3,0

naked

naked?

suauipads 88

1ZJ1U19JS //

12-14

(13)

10-12

(ID

29-21

(20)

21-27

(25)

3, 1-3,5

scaled

scaled

suauipads 901
aujsoujsmqo //

12-14
(13)
9-11
(10)
17-20
(18-19
19-29
(22-24)
3, 2-4,1
(3,6)
2, 8-3, 6
(3,2)
70-80°
(77°)
naked
naked

SU3Uip3ds f
snoipuj 7/

13

9-10

17

33-34

(34)

3, 0-3, 3
(3,3)
3, 2-3 ,4
(3,3)

naked

suauipads g/_
muudojsnj -jj

13-15

(14)

10-11

(11)

19-21

(20-21)

22-23
3, 2-3, 7
(3,4)
2, 6-3,1
(2,9)
68-73°
(70°)
naked
naked

mn3 h

13

8-10*

16-19*

(18)

34-35

2, 9-3,2

3, 3-3,4

“nearly*

vertical”

naked

C/D

4

, — . 03 , — .

v-. *-» »— i
1> — CD
xi xi

ef e

3 a 3

c £ c

o

!- "J U->

Q

o3 -

CQ

X> 3 £>

& E “ E

^3 U 3 J X3

2c|c5)^S^^
S’oSSSSS’oS-Saeu

3.S3M3«3mEuE2E
' ”” -C s'"^ C ^ ^ ' >v s ” c ^

< J PH W C/5 C/5

.’ob^

a c c

c

03

x:

o

s

s

o

5

3

03

<U

CQ

T3

TD

C

03

00

LTl

00

<u

PH

<u

-D

* S *

03 <D

g M

a a

«r «

~ e

c o
.2 *

E s-2

O E 33

* M 5

£

-± ^

2.5 aj

</> c b
'C o<2
a v> -a

c £ .
a3 .33 ^

J3 5 F?

;£ <*> os

> <U

> <U w

Sb c

2: 03 <l>
a .52 c/5

OXI-TU

* r; =y
JCE S3

bf £

w ^ 03

E

Os 52 O
its

c.S§

“o|

oci-3

^ c

a:

J 2 i-

D JO
03 _b

Os

-1- + rx

>__( Os

o^hh

C><34

a + S3

5aG

x S E
E« o

° -= £
i— o3

£ N ^

■C O 1— I

O T3 > IT)

> r/3 S ^

'sbaj g

"S o3

5 03X

Ssu

Q

119

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 4, OCTOBER 1982

Etymology

The name is derived from the combination of the Latin fuscus, dark coloured, and pinna ,
fin, referring to the dark dorsal and anal fins. It is to be treated as a noun in apposition.

Distribution and variation

Helcogramma fuscopinna appears to be a remarkably widely distributed species (Fig. 3).
The only other tripterygiid which is apparently as nearly widely distributed is H. obtusirostre
which has been recorded along the east coast of Africa from the Red Sea to the Transkei
(Klunzinger, 1871; Clark et al., 1968; collected by J. L. B. & M. M. Smith in East Africa dur-
ing the 1950’s; collected by the author in Zululand and on Natal Coast in 1976, 1977; collected
by R. Winterbottom in Transkei in 1975). H. fuscopinna is not represented in recent collec-
tions from Christmas Island (collection of G. R. Allen and R. G. Steene, 1978) and the Cocos-
Keeling Islands (collection of W. Smith-Vaniz et al., 1974). It is also not one of the many trip-
terygiids described by Fowler (1946) from the Ryu Kyu Islands. A few large specimens were,
however, collected by Hanson in southern Japan in 1978.

The specimens from Taiwan and Japan are considerably more heavily pigmented than
those from the Indian Ocean. Five male specimens collected by Springer in the Philippine Is-
lands also differ from the Indian Ocean specimens. Three of these (from Apo Islands) are very
darkly pigmented, like those from Taiwan and Japan, and two (from Putic Island, Palawan
Province) are considerably lighter. The three from Apo Island also have an unusually long first
dorsal fin spine— in two specimens it is twice the length of the first spine of the second dorsal
fin. Usually the difference in length is only about one third. The same three specimens also
have a sharper snout (58°-64° versus 68°-73°). All five specimens have lower lateral-line
counts, namely 20-23 (usually 24-25 in H. fuscopinna). In contrast the specimens from Taiwan
have higher lateral-line counts, namely 26-33, except one with a count of 23. Comparable data
are not available for the specimens from Japan. In all other respects the specimens from the
Philippines conform to the description of the species and, until further material is available,
are referred to H. fuscopinna.

Genus Enneapterygius Riippell

Enneapterygius Riippell 1835: 2 (Type species Enneapterygius pusillus Riippell, 1835, by
original designation).

Diagnosis

First dorsal fin with three spines; anal fin with one spine. Lateral-line divided into an ante-
rior series of pored scales which ends below the second dorsal fin and a posterior series of
notched scales from \ to 2 scales below end of anterior series to base of caudal fin. Small, sim-
ple orbital and anterior nasal tentacles present. Head and nape naked; body with ctenoid
scales, except abdomen and pectoral fin bases which are usually naked but with cycloid scales
in some species. Vomer with 1-3 rows of conical teeth, palatines edentate.

Description (Characters in diagnosis not repeated).

Dorsal fins III+XI— XIV+8— 11. Anal fin 1+17-22. Pectoral fin 13-16, lowermost 6-7
simple and thickened, uppermost 1-4 undivided, remainder bifurcate. Caudal fin with 7-8 dor-
sal, 6-7 ventral procurrent rays. Lateral-line anterior series of 9-15 pored scales ending below
second dorsal fin, posterior series of 21-28 notched scales from \ to 2 scale rows below end of
anterior series to base of caudal fin. Mandibular sensory canals confluent, opening as single
pore just posterior to lower jaw symphysis. Supratemporal sensory canal “U”-shaped, curving
around base of dorsal fin, or crescentic (Fig. 13). Jaws with slightly recurved conical teeth in

120

HOLLEMAN: THREE NEW SPECIES AND NEW GENUS OF TRIPTERYGIID FISHES (BLENNIOIDEI)

bands in front, decreasing to single rows at sides of jaws; teeth in outer rows enlarged; septal
bone present; cephalic lateralis canals not covered by bone; ascending and articular processes
of premaxilla partly or completely fused; hypural 5 present or absent; single epural; free ptery-
giophore between second and third dorsal fins ossified or unossified (see General Discussion).

Remarks

The monophyly of the genus as described above is in doubt. The disparity in the form of
the supratemporal sensory canal and associated structures, the ossification or non-ossification
of a free pterygiophore between the second and third dorsal fins, the degree of fusion between
the articular and ascending processes of the premaxilla and the presence or absence of hypural
5 suggests that the genus might potentially be divided into two. Pending a revision of the genus
currently in progress, all the species below are referred to Enneapterygius .

Enneapterygius clarkae sp. nov. Fig. 6

Enneapterygius n. sp. 2 Clark, 1979: 104.

Diagnosis

Abdomen scaled; single row of thin cycloid scales on pectoral fin base parallel to margin
of branchiostegal membrane. Body with four conspicuous vertical dark bars which divide ven-
trally and continue onto anal fin. Two less conspicuous bars under pectoral fin.

Description

(Except for colour pattern, characters in diagnosis not repeated).

Dorsal fins III+XI-XII + 8-10, usually III + XII +9; anal fin 1+16-17, usually 16; pectoral
fin 14-15, usually 15, with uppermost 1-3 undivided, lowermost 7 undivided and thickened, re-
mainder bifurcate; caudal fin 6-8, 7+6, 5-6. Pelvic fin rays united by membrane for less than
quarter of their length. Lateral-line anterior series 1 1—12, usually 12, pored scales, posterior
series 20-22, usually 22, notched scales from one scale below end of anterior series, overlap-
ping by 2-3 scales, to base of caudal fin; transverse scales 3/6; longitudinal scales 29-30. Verte-
brae, 10 precaudal and 22-23 caudal. Head 3, 3-3, 6 in SL; eye 3, 0-3, 3, upper jaw 2, 5-4,0,
snout 3, 8-4, 8 in head; snout angle 71°-73°. Supratemporal canal crescent shaped, free ptery-
giophore between second and third dorsal fins present. Orbital tentacle of same length as nasal
tentacle, about three times as wide, with serrated margin. Longest pectoral fin ray reaches first
ray of third dorsal fin.

Colour

No live or freshly dead specimens have been seen. Side of body marked with four con-
spicuous vertical dark bands, usually divided ventrally and continuing on to anal fin as 5-7
oblique bars. First band on body from middle of second dorsal fin, second from junction of
second and third dorsal fins, third from posterior half of third dorsal fin and fourth on caudal
peduncle. Last bar may be considerably darker than the others, particularly ventrally. Also
one or two less distinct dusky vertical bands on body under pectoral fin. Black pre-anal mark
present. Anterior half to two thirds of body and head dusted with melanophores. Abdomen
unpigmented in females. Lower portion of the head and base of the pectoral fin with irregular
bars. Pelvic fins unpigmented. First dorsal fin dusky, darker in males than females, with partial
black margin in males. Faint, irregular dusky bars on third dorsal fin, lower half of pectoral fins
and base of caudal fin.

121

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 4, OCTOBER 1982

i 10mm j

Fig. 6. Enneapterygius clarkae sp.n. RUSI 14175. Holotype, male, 23,5 mm SL.

Material examined

Holotype — RUSI 14175, male, 23,6 mm SL; coral reef off Barreira Vermelha, Inhaca Is-
land, Mozambique; 5 December, 1970; T. H. Fraser; field number THF-SA-14.

Paratypes — RUSI 14174 (4, 18,9-24,0 mm SL); taken with holotype.

RUSI 14176 (1, 21,9 mm SL); reef with coral, Sodwana Bay, Zululand, South Africa;
depth 15 m; 25 July, 1976; R. Winterbottom et al.; field number RW 76-15.

RUSI 7939 (2, 18,5 & 21,6 mm SL); Shimoni, Kenya; November, 1952; J. L. B. & M. M.
Smith.

RUSI 14177 (3, 11,3-24,9 mm SL); rock arch with coral and sponges over sand, Sodwana
Bay, Zululand, South Africa; depth 13 m; 24 July, 1976; R. Winterbottom et al., field number
RW 76-14.

RUSI 7938 (1, 20,5 mm SL); reef with coral, sponges, Sodwana Bay, Zululand; depth
14-17 m; 18 June, 1977; M. S. Christensen et al., field number MSC 77-20.

RUSI 9842 (2, 17,9 & 23,4 mm SL); reef with coral and sponges, Sodwana Bay, Zululand;
depth 8-10 m; 19 May, 1979; P. C. Heemstra et al., field number PCH 79-23.

RUSI 14629 (4, 17,4-25,2 mm SL); Malindi Marine Reserve, Malindi, Kenya; April,
1978; P. Wirtz.

LACM 31617-27 (1, 24,5 mm SL); Manda Island, Kenya; 27 November, 1970; P. Saw.

USNM 231380 (1, 20,4 mm SL); reef station. Chesterfield Island (16° 21'S, 43° 59'E);
depth 5 m; 16 October, 1964; Anton Bruun Cruise No. 8, Station 408F, International Indian
Ocean Expedition; field number LK 64-66.

USNM 231378 (1, 23,1 mm SL); fossil coral rock patch with young corals, blind surge
channel about 100 yards off Raphael on West side, St Brandon’s Shoals (16° 26'S, 59° 36'E);
depth 0-8 m; 2 April, 1976; V. G. Springer et al.; field number VGS 76-6.

USNM 231381 (1, 17,0 mm SL); coral reef off northwest shore, Albatross Island, St Bran-
don’s Shoals (ca 16° 15'S, 59° 35'E); depth 0-18 m; 14 April 1976; V. G. Springer et al.; field
number VGS 76-22.

USNM 231379 (1, 23,4 mm SL); Red Sea (27° 16' 46"N, 33° 46' 25"E); depth 0-3 m; Inter-
national Indian Ocean Expedition; 1 January 1965; L. Kornicker & H. A. Feldmann; field
number HA 29.

USNM 231382 (3, 13,5-21,7 mm SL); rock reef with live coral, channels, course sand bot-

122

HOLLEMAN: THREE NEW SPECIES AND NEW GENUS OF TRIPTERYGIID FISHES (BLENNIOIDEI)

tom, off northern tip of St Brandon’s Shoals (16° 25'S, 59° 36'E); depth 6-10 m; 6 April 1976;
V. G. Springer et al. ; field number VGS 76-10.

The larger two specimens of RUSI 14177 were cleared and stained.

Discussion and comparisons

Clark (1979) described five specimens of a species from the Red Sea which most probably
is E. clarkae. The specimens were lost prior to the publication of the description and the
species was consequently not named.

Clark was unsure whether these five specimens, described by her as Enneapterygius
n.sp.2, represented a new species or were merely large adults of E. destai Clark, 1979. The
South African specimens essentially agree with the description of Clark’s n.sp.2 and not with
that of destai. The only meristic difference between the two species is the number of pored
scales in the anterior lateral-line series (Table 5). However, Clark does not describe the scala-
tion of either her n.sp.2 or destai. The abdomen of clarkae is entirely covered with ctenoid
scales and there is a single row of cycloid scales on the pectoral fin base. Both these features
are absent in destai (Springer, pers. comm.). The two species also differ in colour pattern. The
body bars of clarkae and Clark’s n.sp.2 are vertical and very distinct, whereas those of destai ,
except the bar on the caudal peduncle, are faint and oblique. The peduncular bar of destai is
dark and constricted in the centre giving the impression of an hour-glass. Another species with
a very distinctive hour-glass-like peduncular is E. elegans (Peters, 1876). However, elegans has
a scaled abdomen, 17 anal fin rays and 17+17 lateral line scales. Finally, this series of clarkae
includes specimens smaller than the largest specimens of destai. The two species are thus clear-
ly distinct.

Etymology

Eugenie Clark was aware of the South African specimens of her Enneapterygius n.sp.2
prior to publication of the Red Sea revision. She, however, very kindly consented that 1 name
and describe the species. Eugenie Clark also provided me with a draft manuscript of her Red
Sea revision in 1976. This was of invaluable assistance in a revision of the South African trip-
terygiids which formed the basis of my Masters dissertation. It is thus fitting — and it gives me
great pleasure — to name the species for her.

Distribution

The distribution of E. clarkae is shown in Fig. 3.

Enneapterygius ventermaculus sp.nov. Fig. 7

Diagnosis

Third dorsal fin usually with 10 rays; anal fin usually with 19 rays. Supratemporal sensory
canal “U”-shaped. Row of 5-6 conspicuous black spots at base of anal fin and one anterior to
vent.

Description

(Characters in diagnosis not repeated; characters for holotype in parenthesis).

Dorsal fins III + XI-XIII+9 -10 (III+XII+9), usually III + X1I + 10; first dorsal fin height
equal to second dorsal fin in females, slightly higher than second in males; anal fin 1 + 17-20
(1+18), usually 19 rays; pectoral fin rays 14, upper 1-3 undivided, lower 7 thickened and undi-
vided, remainder bifurcate; caudal fin 7-8 7+6, 6—7 pelvic fin rays united by membrane for
about half their length. Lateral-line anterior series 13-16 (13), usually 15 pored scales ending
under last third of second dorsal fin; posterior series 21-25 (22), usually 23 notched scales from

123

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 4, OCTOBER 1982

two scale rows below end of anterior series, overlapping by 1-3 scales and continuing to base
of caudal fin; transverse scales 2/6; lateral scales 32-34, usually 33. Precaudal vertebrae 10,
caudal 24-26. Head 3, 4-4,0 in SL; eye 3,0-3, 7, upper jaw 2, 8-3, 3, snout 3, 0-3,4 in head;
snout angle 76°-78°. Head and pectoral fin bases naked; body except abdomen from line be-
tween upper angle of pectoral fin base to origin of anal fin with ctenoid scales. Free pterygio-
phore between second and third dorsal fin cartilagenous; hypural 5 absent.

Colour (in preservative)

No live or freshly dead specimens have been seen. The body is irregularly pigmented with
dark brown to black melanophores, the pigment normally occurring on the posterior margins
of the scales. There is a row of irregular blotches along the lateral midline, the darkest forming
a bar at the base of the caudal fin. This bar may be divided ventrally to form an inverted “Y”.
The head is lightly spotted with small clusters of melanophores on the cheeks. The lower half
of the pectoral fin bases have clusters of melanophores forming narrow bars. There is a dark
blotch on either side of the midline on the throat and near the base of the brachiostegals. The
abdomen is unpigmented except for a conspicuous black mark, frequently triangular in shape
with apex anterior, which lies just anterior to the anus (Fig. 7b). This marie may also be round-
ish or crescentic. The caudal and pectoral fins each have 4-5 irregular, faint dusky bars, with
the pigmentation on the rays only, giving the fins a spotted appearance. There are 5-6 irregu-
larly spaced dark spots at the anal fin base. These are continued as “bars” across the fin. Pig-
mentation occurs only on the rays so that when the fin is extended these spots form a dotted
line running obliquely forward from the basal spot. The first dorsal fin is irregularly dusky
whereas the second may have four broad, irregular bars, and the third three broad irregular
bars. There is considerable variation in the intensity of pigmentation in different individuals.
However, all specimens have a spotted appearance with the distinct preanal mark. No sexual
dichromatism is evident.

Fig. 7. Enneapterygius ventermaculus sp.n., RUSI 7943, Holotype, female, 24,8 mm SL.

Material examined

Holotype — RUSI 7943, female (24,8 mm SL); rock pool about 10 km south of Sodwana
Bay (27° 37' 30", 33° 40' 50"E), Zululand, South Africa; depth 0-3 m; July, 1976; R. Winter-
bottom et al. ; field number RW-76-8.

Paratypes— RUSI 7947 (1, 24,7 mm SL); from same collection as holotype.

RUSI 7944 (4, 17,8 & 25,7 mm SL); rock pool at six mile reef, Zululand; depth 0-2 m;
July 1976; R. Winterbottom et al.', field number RW-76-9.

124

HOLLEMAN: THREE NEW SPECIES AND NEW GENUS OF TRIPTERYGTID FISHES (BLENNIOIDEI)

RUSI 7945 (1, 23,5 mm SL); reef with corals about 2 km offshore Sodwana Bay, Zulu-
land; depth 8 m; June, 1877; M. S. Christensen et al. ; field number MSC-77-18.

RUSI 7946 (1, 27,1 mm SL); reef with corals and sponges about f km offshore Sodwana
Bay, Zululand; depth 11-13 m; June, 1977; M. S. Christensen et al.', field number
MSC-77-17.

LACM 38309-5 (6, 25,6-27,7 mm SL); tide pool with rock and algae, Beluji Point, Sind,
Pakistan; depth 20-60 cm; 27 January, 1979; C. C. Swift et al.; field number CCS 79-20.

LACM 38310-10 (43, 20,2-31,8 mm SL); tide pool with sand, rocky rubble and algae. Be-
luji Point, Sind, Pakistan; depth 20-60 cm; 27 January, 1979; C. C. Swift et al.; field number
CCS 79-21.

LACM 38320-9 (2, 27,2 & 31,8 mm SL); small cove with boulders and algae, 4,8 km west
of nuclear power plant, Karachi, Pakistan; depth 0-5 m; 13 February, 1979; C. C. Swift et al.;
field number CCS 79-34.

BMNH 1954.4.26.209-214 (6, 19,5-21,9 mm SL); rock pools, Aden; 1954; A. Fraser-
Brunner.

BMNH 1954.4.26.191-196 (6, 17,2-23,6 mm SL); rock pools, Mukalla, Aden; 1954;
A. Fraser-Brunner.

BMNH 1954.4.26.197-208 (12, 17,0-24,0 mm SL); tide pools, Alayu, Aden; 1954;
A. Fraser-Brunner.

Etymology

The name is a combination of the Latin venter, meaning “belly” and maculus, a spot or a
mark, and is given for the distinct black preanal mark. It is to be treated as a noun in apposi-
tion.

Distribution

On the east coast of Africa this species has only been taken in Zululand but may also oc-
cur further north. The species appears to be fairly common in Pakistan and there are a few
specimens known from Aden.

Comparisons

The only species which approaches E. ventermaculus in overall appearance is E. clarkae.
In the latter species the bars on the body are much more distinct; it has fewer scales in the an-
terior lateral line series (11-12 vs. 13—16), fewer anal fin rays (16-17 vs. 18-19), a crescent-
shaped supraoccipital sensory canal, and a scaled abdomen and pectoral fin base. The salient
features of six Indian Ocean species of Enneapterygius are compared in Table 5.

Genus Cremnochorites gen. nov.

Type-species Tripterygium capense Gilchrist & Thompson, 1908
Diagnosis

First dorsal fin with four spines; anal fin with two spines. Lateral line divided. Body heavi-
ly scaled with ctenoid scales; head with denticle like scales; row of “ctenii” around perimeter
of eye. Orbital tentacle large and multifid, nasal tentacle similar but smaller. Single row of con-
ical teeth on vomer and palatines.

Description

Dorsal fins IV+XIV— XV+10-11; some rays may be bifurcate and the last is usually divi-
ded to its base; first dorsal fin lower than second. Anal fin 11+21—22, the last ray usually divi-
ded to its base. Pectoral fins with 16 rays, upper 8 bifurcate, lower 8 simple and thickened.

125

Comparison of selected characters of five species of Enneapterygius

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 4, OCTOBER 1982

a.

£

o

GO

TD

<D

a

U

B

o

e

<u

M

^3

<L>

C/3

&

jC

H

126

South African specimens. (See methods).

HOLLEMAN: THREE NEW SPECIES AND NEW GENUS OF TRIPTERYGIID FISHES (BLENNIOIDEI)

Caudal fin 6 dorsal, 5 ventral procurrent rays. Lateral-line anterior series of 21-24 pored scales
running to below anterior of third dorsal fin; posterior series of 17-20 notched scales extending
from below end of anterior series onto caudal peduncle. Body heavily scaled, all scales with
relatively large, irregular cteni (Fig. 8 A-D); scales on abdomen with few cteni. A few cycloid
scales occur around vent and at base of pelvic fins. Scale rows somewhat irregular. Head and
pectoral fin bases covered with ctenoid scales; those on posterior edge of opercle about half
size of body scales, decreasing in size anteriorly to small denticle-like scales with few, large cte-
ni below eye and on cheeks (Fig. 9A). Scales on head and particularly on cheeks appear to be
situated on small pedestals, creating appearance of shark denticles. Small denticle-like spines
on nape and interorbital area, apparently ankylosed to cranial bones. Perimeter of eye with
ring of “cteni” (Fig. 9B); similar “cteni” on posterior end of maxilla. Skin of isthmus papillose
with single “cteni” or spines embedded in papillae (Fig. 9C).

Posterodorsal margin of post-temporal serrated; interorbital concave, with ridge over each
orbit; transverse depression behind orbits. Head broad with rounded profile. Large multifid or-
bital tentacle present; similar, small tentacles on posterior margin of anterior nostrils. Man-
dibular sensory canals confluent, opening as single pore posterior to lower jaw symphysis. Both
jaws with slightly recurved conical teeth, a patch in front grading to a single row at back of
jaw. Teeth unequal in size on lower jaw; upper jaw with outer row of large teeth and inner
band of small teeth. Vomer with single row of slightly recurved conical teeth which continues
into palatines. Septal bone present; cephalic lateralis canals covered by bone. Caudal skeleton
with large hypural 5 and two free epurals.

DISCUSSION

Cremnochorites capense was originally placed in Tripterygium (=Tripterygion) by Gilchrist
& Thompson (1908). It was later transferred to Gillias Evermann & Marsh, 1899 by Barnard
(1927) and retained there by Smith (1949). Rosenblatt (1960) placed Gillias in synonymy with
Enneanectes Jordan & Evermann, 1895 (a genus restricted to the eastern tropical Pacific and
the western tropical Atlantic). Rosenblatt (1960: 3) did not refer Indo-Pacific species in Gillias
to any other genus but merely stated that “none of the Australian or South African species re-
ferred to Gillias has anything to do with that genus”. Clark (1979) suggested that G. capense
be referred to Norfolkia Fowler, 1953 for these taxa share four first dorsal fin spines, two anal
fin spines and a scaled head. However, Cremnochorites differs from Norfolkia in a number of
characters: palatine teeth are present in Cremnochorites and absent in Norfolkia\ lateral-line
counts are reversed, 21-22 pored, 15-16 notched scales for the former compared with 13-17
pored, 21-23 notched scales for Norfolkia. The head scales of Cremnochorites are quite unlike
those of any other tripterygiid examined (see Description).

Body scales of the species in the two genera are also very different. Those of Cremnocho-
rites (Fig. 8) are heavily ctenoid and the cteni are generally unequal in size. Pored lateral-line
scales frequently have 2-3 rows of cteni in the centre of the row. Scale foci are close to the
posterior edge of the scale and the radii are thus comparatively long. (Radii occur only in the
anterior field). In Norfolkia (Fig. 10) cteni are smaller, more or less equal in size and always in
a single row. Scale foci are further away from the posterior margin of the scale and the radii
are consequently relatively shorter.

Cremnochorites also possesses a septal bone ( sensu Springer & Freihofer, 1976; Fig. 16)
which is absent in Norfolkia. Furthermore, the sensory canals in the infraorbitals and nasals of
Cremnochorites are complete, as are those portions of the preopercle, posttemporal and pte-
rotic which carry sensory canals. In Norfolkia these canals are open laterally, except the nasals
which may be narrowly bridged.

127

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 4, OCTOBER 1982

Fig. 8. Examples of body scales (A, B) and pored lateral-line scales (C, D) of Cremnochorites capensis.

128

HOLLEMAN: THREE NEW SPECIES AND NEW GENUS OF TRIPTERYGIID FISHES (BLENNIOIDEI)

S 2

o c
r- c

03 O

JD O

Oh C

S.O

o3 *->
O

— 0-H

<u

2 o

O 2-

<D r;

55 U

^ <
o w
£

129

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 4, OCTOBER 1982

Fig. 10. Examples of body scales (A, B) and pored lateral line scales (C, D) of Norfolkia springeri.

130

HOLLEMAN: THREE NEW SPECIES AND NEW GENUS OF TRIPTERYGIID FISHES (BLENNIOIDEI)

Specimens of all nominal species of Norfolkia have been seen. The species appear to form
a coherent group with little interspecific variation. There are no species which share the cu-
rious head scalation with Cremnochorites or have palatine teeth. The loss of palatine teeth and
the possible secondary loss of the septal bone in Norfolkia may represent synapomorphies for
the genus. However, pending revision of Norfolkia and an assessment of the septal in Triptery-
giidae, no further comments can be made about Norfolkia.

The species capensis can patently not be referred to Norfolkia. The septal and highly de-
rived head scalation is considered sufficient to warrant the placing of this species in a new gen-
us.

Cremnochorites capense is the only species thus far examined which can be ascribed to this
genus. It has been taken in False Bay, western Cape, at Skoenmakerskop (near Port Eliza-
beth) and at Port Alfred (gully at 15 m), eastern Cape, South Africa. Specimens have also
been taken at the Storms River mouth and in the mouth of the Knysna estuary. The four type
specimens were taken in shrimp trawls, two at 5 fms (10 m) and two at 14 fms (28 m). The
other specimens were taken at 1-20 m depth. With the exception of the types, for which there
are no other collection data, all other specimens were taken from relatively sheltered, vertical,
seaweed and Pyura covered rock faces.

Etymology

Cremnochorites is derived from the Greek kremnos, a cliff, and chorites , native or country
man. It is thus named because it has only been found associated with vertical rock faces. The
gender is masculine.

Cremnochorites capense (Gilchrist & Thompson) comb. nov. Fig. 11

Tripterygium capense Gilchrist & Thompson, 1908: 140 (Type locality False Bay, Cape,
South Africa); Thompson, 1918: 151; Gillias capensis Barnard, 1927: 827; Smith, 1949: 359.

Diagnosis

As for the genus.

Description

Dorsal fins IV+XIV-XV+ 10-11, usually IV+X1V+11, first dorsal fin lower than second,
rays except first branched once; anal fin 11+21-22, usually 21 rays; pectoral fin rays 16, lower 8
thickened and undivided, uppermost sometimes undivided, remainder divided once; caudal fin
6,6+7, 5. Lateral-line anterior series 21-24 pored scales to end of second or beginning at third
dorsal fin, posterior series 17-20, usually 19, notched scales from two scales below anterior se-
ries and frequently overlapping by 4 or more scales, to base of caudal fin; transverse scales
5/11; longitudinal scales 36-37. Precaudal vertebrae 10, caudal 27-28. Head, 3, 3-3, 7 in SL;
eye 3,0-3, 6, upper jaw 2, 0-2, 4, snout 3, 6-4,4 in head. Snout angle 71°-76°. Balance of de-
scription is as given for the genus. The only sexual dimorphism noted consists of a single, coni-
cal papilla at the posterior of the vent in males and a “rosette” around the vent of females.
This “rosette” may or may not protrude a short distance. Although these features appear to be
common to many tripterygiids, they cannot always be distinguished.

Colour

In preservative, these fishes are generally grey with a pale belly. There are six, irregular,
vertical dark bars on the body above the lateral midline. These may divide into eight or more
bars below the midline. The lower body bars may alternate with the upper bars. In juveniles
the bars are continuous across the body and are more distinct. The penultimate bar is across
the peduncle and the last bar lies at the base of the caudal fin. A dark bar extends across the

131

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 4, OCTOBER 1982

nape and continues on to the opercles. Another bar runs from below the midline of the eye to
the corner of the mouth. There is an irregular dark blotch on the lower portion of the opercle.
The orbital tentacles are dusky. The anal and caudal fins have irregular small dark blotches.
The first dorsal fin is dark whereas the second and third have dusky margins. Neither the pelvic
nor pectoral fins are pigmented although there are two short dark bars on each pectoral fin
base. Males have dusky brachiostegal membranes and a more darkly pigmented head than fe-
males. Freshly caught fishes are rust-coloured and the irregular blotching renders them cryptic
in brown and purple algae. One live specimen was seen in the mouth of the Knysna estuary. It
may have been a male in breeding dress as it had a bright yellow head and first dorsal fin.

Fig. 11. Cremnochorites capense. RUSI 75-21. male, 80 mm SL.

material examined

3 Syntypes, SAM 9900, females (46-53 mm SL), 10-28 m depth. False Bay, Cape, South
Africa, collected in a shrimp trawl; 1 Syntype, SAM 9901, male (54,7 mm SL), collected with
SAM 9900.

RUSI 7384 (49,3 mm SL), False Bay, Cape, South Africa, no date.

RUSI 75-22 (8, 46,9-61 mm SL), cliff with Pyura sp., False Bay, Cape, South Africa;
depth 6-8 m; 1 November, 1975; R. Winterbottom et al.; field number RW-75-22.

RUSI 75-21 (1, 80 mm SL), sheltered bay with vertical rock walls; Platboom, Cape,
South Africa; depth 2-3 m; 1 November, 1975; R. Winterbottom et al.; field number
RW-75-20.

RUSI 76-7 (12, 26,9-55,6 mm SL) from vertical rock wall with algae; Skoenmakerskop,
Eastern Cape, South Africa; depth 4 m; 19 January, 1976; R. Winterbottom et al.; field num-
ber RW 76-1.

RUSI 77-7 (2, 18,5 & 41,7 mm SL) same site as above; depth 5 m; 1 February, 1977; M.
Christensen et al.; field number MSC 77-4.

GENERAL DISCUSSION

Only one major revision of the family Tripterygiidae has ever been undertaken; that by
Rosenblatt (1959) as a Ph.D. dissertation. To date the major portion of this revision is unpub-
lished.

132

HOLLEMAN: THREE NEW SPECIES AND NEW GENUS OF TRIPTERYGIID FISHES (BLENNIOIDEI)

The concepts of systematics have undergone radical changes in the past 20 years. With the
introduction of the Hennigian approach to systematics it has become accepted to consider as
valid only those taxa that comprise entities that share one or more derived characters — synapo-
morphies. Only by identifying synapomorphic characters can taxa be properly defined and the
interrelationships of taxa assessed. No such study has been undertaken for the Tripterygiidae
and as such the family, as well as the genera within it, remain undefined.

Springer (pers. comm.) informs me that the one osteological character that serves to
define the family and delimit it from all other blennioid fishes is the derived loss of at least the
posteriormost spine of the second dorsal fin (Fig. 12C). All other blennioids have a spine im-
mediately anterior to the first ray (Fig. 12A) although it may be greatly reduced, as in Entoma-
crodus nigricans Gill (Fig. 12B). In many tripterygiid genera the first ray of the third dorsal fin
is also lost, leaving a free pterygiophore which supports nothing in the “gap” between the sec-
ond and third dorsal fins (Fig. 12D). This pterygiophore frequently becomes considerably re-
duced (Fig. 12E) and in some specimens of E. pusillus and E. ventermaculus a spine and two
rays have been lost, and the free pterygiophores are unossified (Fig. 12F). It is as yet not
known how consistent the degree of loss is within a genus. The minimal loss of the posterior-
most spine is found in representatives of 12 genera inspected and is synapomorphous for the
family.

The species E. pusillus and E. ventermaculus which have at least one unossified pterygio-
phore also have a “U”-shaped supratemporal sensory canal (Fig. 13A) whereas the species of
Enneapterygius which have an ossified free pterygiophore (E. abeli, E. clarkae , E. elegans)
have a crescentic supratemporal sensory canal (Fig. 13B). Furthermore, in those species which
possess a “U”-shaped supratemporal sensory canal, it would appear that the extreme curvature
of the canal around the first dorsal fin has been brought about by the forward movement of the
dorsal fins. This has resulted in a compression of the pterygiophores of the first dorsal fin as
they have moved forward over the back of the neurocranium (Fig. 14) and a concomittant de-
pression of the supraoccipital giving rise to a concave supraoccipital bone.

If these two states (the non-ossification of the free pterygiophore plus “U”-shaped
supraoccipital sensory canal and associated features as opposed to an ossified free pterygiophore
plus crescentic supratemporal sensory canal) are found to be consistent for all species ascribed
to Enneapterygius there may be sufficient reason to divide the genus into two separate taxa.

Springer & Freihofer (1976) described a de novo autogenous ossification in the interorbital
septum of Pholidichthys leucotaenia and named the bone the septal. Ruck (1977) found the
same bone in the tripterygiid Forsterygion varium; Springer (pers. comm.) found it present in
Gilloblennius, another tripterygiid, and Holleman (1978) found a septal in species of a number
of other tripterygiid genera viz. Cremnochorites, Enneapterygius , (Fig. 15) Helcogramma, No-
toclinops , Tripterygion and Vauclusella. It is absent in Brachynectes, Lepidoblennius, Norfolkia
and Notoclinus. Whether the septal was originally developed in those last four genera and sec-
ondarily lost, or not developed at all has still to be determined.

The presence of a septal in certain tripterygiid genera suggests some relationship between
the Tripterygiidae and Pholidichthyidae. Freihofer (pers. comm.) informs me that there are
considerable similarities between the trunk lateral line nerve patterns of Pholidichthys and
Forsterygion which is again suggestive of relationship between the two families. Springer &
Freihofer (1976) only considered Pholidichthys to show possible relationship with the tropical
blennioid fishes (superfamily Blennioidea) considered by George & Springer (1980) to include
the Clinidae, Blenniidae, Dactyloscopidae, Tripterygiidae, Labrisomidae and Chaenopsidae.

133

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 4, OCTOBER 1982

Fig. 12. Junction of spinous and rayed dorsal fins of: A — Pavoclinus laurentii (Clinidae); B — Entomacrodus nigricans
(Blenniidae); C — Norfolkia squamiceps (Tripterygiidae), D — Crenmochorites capensis (Tripterygiidae); E — Enneaptery-
gius elegans (Tripterygiidae), and F — Enneapterygius pusillus. S — last spine, R — first ray, P — free pterygiophore, U —
unossified free pterygiophore. (B — portion of Figure 10, Springer 1968).

134

HOLLEMAN: THREE NEW SPECIES AND NEW GENUS OF TRIPTERYGTID FISHES (BLENNIOIDEI)

supratemporal sensory canals in Enneapterygius spp.

A B

Fig. 14. First dorsal fin spines and pterygiophores of Enneapterygius spp. with (A) crescentic and (B) “U”-shaped supra-
temporal sensory canals.

135

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 4, OCTOBER 1982

B

Vomer . ^ . ,

Anterior myodome

Fig. 15. Septal of (A) Cremnochorites capensis and (B) Enneapterygius clarkae.

ACKNOWLEDGEMENTS

The author is indebted to a number of people and institutions which provided specimens
on loan or as gifts: G. R. Allen, Western Australian Museum; D. F. Hoese, Australian Nat-
ional Museum; W. Klausewitz, Senckenberg Museum; V. G. Springer, National Museum of
Natural History; C. C. Swift, Los Angeles County Museum of Natural History; R. Winterbot-
tom, Royal Ontario Museum; the British Museum (Natural History); the Hebrew University
of Jerusalem, and the South African Museum.

The illustrations of the species were executed by Mrs E. M. de Villiers (E. M. Tarr) and
R. E. Stobbs made the radiographs used in this study. The electron micrographs were produc-
ed by the Electron Microscope Unit of Rhodes University.

Portion of this study received financial support from Prof. M. M. Smith, the Research
Committee of Rhodes University and the University Research Division of the Council for
Scientific and Industrial Research.

Drafts of this manuscript were reviewed by Dr P. C. Heemstra, Dr V. G. Springer and Dr
R. Winterbottom.

136

HOLLEMAN: THREE NEW SPECIES AND NEW GENUS OF TRIPTERYGIID FISHES (BLENNIOIDEI)

LITERATURE CITED

Barnard, K. H. 1927. A monograph of the Marine Fishes of South Africa. Ann. S. Afr. Mus. 21 (2): 419-1065.

Bleeker, P. 1858. Vierde bijdrage tot de kennis der vischfauna van Biliton . Natuurk. Tijdschr. Ned. -Indie. 15: 219-240.

Clark, E. 1979. Red Sea Fishes of the Family Tripterygiidac with Descriptions of Eight New Species. Israel J. Zool. 28:
65-113.

Clark, E., A. Ben-Tuvia and H Steinitz 1968. Observations on a Coastal Fish Community, Dahlak Archipelago, Red
Sea. Bull. Sea Fish. Res. Stn Israel 49: 15-31.

Evermann, B. W. and M. C. Marsh 1899. Investigations of the Aquatic Resources and Fisheries of Porto Rico. Bull.
U.S. Fish Comm: 1-330.

Day, F. 1876. The Fishes of India. London.

de Beaufort, L. F. and W. M. Chapman 1951. The Fishes of the Indo-Australian Archepelago. IX. Percomorphi (con-
cluded), Blennioidea. Leyden: E. J. Brill.

Fowler, H. W. 1946. Collection of Fishes Obtained in Riu Kiu Islands by Captain Ernest R. Tinkham A. U.S. Proc.
Acad. nat. Sci. Philad. 98: 184-193.

Gilchrist, J. D. F. and W. W. Thompson 1908. The Blenniidae of South Africa. Ann. S. Afr. Mus. 6 (2): 1-212.

Gosline, W. A. 1968. The Suborders of the Perciform Fishes. Proc. U.S. natn. Mus. 124 (3647): 1-78.

Herre, A. C. W. T. 1944. Notes on Fishes in the Zoological Museum of Stanford University. XVII New Fishes from Jo-
hore and India. Proc. biol. Soc. Wash. 57: 45-52.

Holleman, W. 1978. The Taxonomy and Osteology of Fishes of the Family Tripterygiidae (Perciformes: Blennioidei) of
South Africa. Unpublished M.Sc. Dissertation, Rhodes University, Grahamstown.

Hubbs, C. L. and K. F. Lagler 1958. Fishes of the Great Lakes Region. Cranford Inst. Sci. Bull. 26: 1-213.

Klunzinger, C. B. 1871. Synopsis der Fische des Rothen Meeres. Verh. zoolbot. Ges. Wien 21: 441-688.

Lal Mohan, R. S. 1971. Helcogramma shinglensis, a new species of tripterygiid from the Gulf of Mannar, with a key to
the fishes of the family Tripterygiidae of the Eastern and Central Indian Ocean. Senckenberg. Biol. 52 (3/5):
219-223.

McCulloch, A. R. and E. R. Waite 1918. Some New and Little-known Fishes from South Australia. Rec. S. Aust.
Mus. 1 : 51-53.

Rosenblatt, R. H. 1959. A Revisionary Study of the Blennioid Fish Family Tripterygiidae. Unpublished Doctoral Dis-
sertation, University of California, Los Angeles.

Rosenblatt, R. H. 1960. The Atlantic Species of the Blennioid Fish Genus Enneanectes. Proc. Acad. nat. Sci. Philad.
112: 1-23.

Ruck, J. 1976. Studies on the development and osteology of some New Zealand inshore fishes. Unpublished Doctoral
Dissertation, Victoria University of Wellington, N.Z.

Ruppell, E. 1835. Neu Wirbelthiere zu der Fauna von Abyssinien gehorig, Fische des rothen Meeres. Frankfurt: Bron-
ner.

Smith, J. L. B. 1949. The Sea Fishes of Southern Africa. Fifth Edition 1965, Johannesburg: Central News Agency.

Springer, V. G. 1968. Osteology and Classification of the Fishes of the Family Blennndae. Bull. U.S. natn. Mus. 284:
1-80.

Springer, V. G. and W. C. Freihofer 1976. Study of the Monotypic Fish Family Pholidichthyidae (Perciformes). Smith-
son. Contr. Zool. 216: 1-43.

Talwar, P. K. and T. K. Sen 1971. On some fishes from the Madras Coast with a description of a new species of the
family Clinidae. Rec. Zool. Surv. India 65 (1-4): 243-251.

Taylor, R. W. 1967. Outline of a Method of Clearing Tissues with Pancreatic Enzymes and Staining Bones of Small
Vertebrates. Turtox News 45 (12): 308-309.

Zander, C. D. and A. Heymer 1970 Tripterygion tripteronotus (Risso, 1810) und Tripterygion xanthosoma n.sp: eine
okologische speziation. (Pisces, Teleostei). Vie Milieu 21 (24): 364-395.

137

8 «

k&

m

ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

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

29th OCTOBER 1982

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

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

BD0760

Some aspects of the ethology of Chalybion ( Hemichalybion ) spinolae
(Lepeletier) (Hymenoptera: Sphecidae: Sphecinae) in the Eastern Cape

Province of Sooth Africa

by

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

ABSTRACT

Some aspects of the ethology of Chalybion (Hemichalybion) spinolae (Lepeletier) are de-
scribed, particular attention being paid to the situation and nature of the nest, the mode of its
construction, provisioning and sealing. Published accounts concerning the nesting of species of
Hemichalybion are briefly reviewed. Data attributed to certain species, including H. spinolae ,
are shown to be misleading due to the confusion by some authors of species of Hemichalybion
with those of Sceliphron. The nesting of Hemichalybion (as exemplified by that of H. spinolae
here described) is discussed with reference to that of the Sceliphronini as a whole. Possible
specific interactions between three spider-hunting Sceliphronini nesting at Hilton are touched
upon as is resource partitioning with respect to their prey.

INTRODUCTION

That some confusion exists in the literature concerning the nesting of species of the sub-
genus Hemichalybion Kohl, 1918, has been clear to two of the authors (F. W. G. and S. K. G.)
for some years. Consequently it was with considerable interest that during the autumn of 1978
they twice witnessed part of the process of nest construction by C. (H.) spinolae (Lepeletier).
Unfortunately, both nests were abandoned by their builders prior to their being provisioned
and, though the site was visited frequently during the succeeding months, no further nesting
activity was observed. The consequent nesting data obtained were published in outline (Gess,
1981: 24).

Late in the summer of 1982 two females provisioning nests at a site a little removed from
that of the previous observations were discovered by the third author (A. J. S. W.) who exca-
vated the nests after their completion and thereby established the identity of the prey. Having
been alerted to the nesting of C. (H.) spinolae at the new site, F. W. G. and S. K. G. during
the week following closely observed a female constructing, provisioning and sealing a further
three nests. In their turn these were excavated and examined. The present account is therefore
based on a total of seven nests, five of which were completed.

The present paper is the twelfth of a series of publications dealing with the ethology of
certain solitary wasps occurring at Hilton, a farm situated 18 kilometres WNW of Grahams-
town (33° 19'S, 26° 32'E) in the Albany Division of the Eastern Cape Province of South Afri-
ca. A detailed account of various aspects of the ecology of Hilton has previously been given
(Gess, 1981: 3-9).

139

ANN. CAPE. PROV. MUS. (NAT. HIST.) VOL. 14, PT. 5, OCTOBER 1982

THE NESTING OF CHALYBION (HEMICHALYBION) SPINOLAE (LEPELETIER)
Description of the nesting sites

Two nesting sites of C. (H.) spinolae were located at Hilton. Both are north-facing vertical
banks. One is naturally occurring and is the bank of the New Year’s River. It is composed of a
mixture of sand and clay which, as a result of recent mineralization, has the character of a
weak and crumbly sandstone-like material and is referred to as the “sandstone” bank (Gess,
1981: 58 and Fig. 8). The other is man-made and is a wall of a sandpit dug in the flood plain of
the Iron Put River, a tributary of the New Year’s River (Gess, 1981: 58 and Fig. 5). The bank
(Fig. 1), created in 1981 by an enlargement of the sandpit in a direction away from the river,
differs from the other banks of the sandpit in that it is cut in non-friable soil. Although it is
composed largely of firmly compacted fine sand, it has a considerable content of colloidal ma-
terial as was demonstrated at the time of study by the presence of the nest turrets of eumenids
and, on small ledges, those of the pompilid Dichragenia pulchricoma (Arnold).

At both nesting sites C. (H. ) spinolae favoured slightly protected situations such as be-
neath overhangs caused by uneven weathering of the banks or by man’s undercutting of the
same (Figs 2 and 3). All nest entrances were located above the base of the banks, at heights
ranging from 16,5-70 cm.

Water, required by the wasp for nest construction, was available at the two sites as pools
in the river beds. At the time of nesting there was in addition a temporary rain water pool on
the floor of the sandpit.

Fig. 1. Hilton, vii. 1982. North-facing bank of sandpit, the location of Nests 3-7.

140

GESS ET AL.: ETHOLOGY OF CHALYBION (HEMICHALYBION) SP1NOLAE (LEPELETIER)

Fig. 3. Hilton, 6.iv.l978. Entrance of Nest 1 (ar-
rowed) with flanking it quarry from which cell-lining
material was obtained. Turret on lower left that of an
unidentified eumenid. (x c.1,0)

Fig. 3 inverted

Fig. 2. Hilton, 6.iv.l978. Situation of Nest 1 in shel-
tered position on north-facing “sandstone” bank.

Flight season and daily flight period

The flight season of C. (H.) spinolae at Hilton starts in early November. Males have been
collected through to early March and females to mid May. Nesting was observed from mid-
March through to the first week of June.

Both males and females were observed and collected near the water puddles, while forag-
ing and while flying about in front of the banks. Though suitable opportunities for mating oc-
curred it was not, however, observed.

The day’s nesting activities started at about 10 a.m. though, shortly before this, females
were observed sunning themselves near nests started the previous day. If a nest is completed in
the early afternoon a new nest may be initiated on the same day. Provisioning and sealing of
nests in all cases took place during late morning and early to mid afternoon.

Female C. (H.) spinolae do not shelter or sleep in or near their nests and their sheltering
places were not discovered.

141

ANN. CAPE. PROV. MUS. (NAT. HIST.) VOL. 14, PT. 5, OCTOBER 1982
Plants visited by adult wasps

Adult C. (H.) spinolae of both sexes have been observed visiting flowers for the purpose
of imbibing nectar. The following associations between this wasp and forage flowers in the
eastern Cape have been recorded: on Maytenus linearis (L.f.) Marais (Celastraceae) at Hilton,
Grahamstown, 9-11. xii. 1969, 82 ?, A6 6 (F. W. Gess) and 6.xii . 1972, 6 (F. W. Gess); on
Foeniculum vulgare Mill. (Umbelliferae) in Belmont Valley, Grahamstown, 20. i. 1970, 62 2,
26 6, 22. i. 1970, 2, 23. i. 1970, 2, 24T1970, 6, 5.ii.l970, 2? 2 , 28.iv. 1970, 6 (all F. W. Gess),
17—25 . i . 1970, 22 2 (J. G. H. Londt) and 244.1970, 6 ( C . F. Jacot-Guillarmod); on Zizyphus
mucronata Willd. (Rhamnaceae) at the Koonap River, 17 miles from Adelaide on Grahams-
town road, 20-22. xii. 1972, 2, c? (C. F. Jacot-Guillarmod); on Acacia karroo Hayne (Legum-
inosae) at Hilton, Grahamstown, 34.1977, 6 (D. W. Gess); and on Hedera helix L. (Arali-
aceae) in a garden, Grahamstown, 27. xii. 1981-44. 1982, 9,26 6 (R. W. Gess).

In addition one male was collected on young foliage of Acacia karroo at Hilton on
6. xii. 1977 (H. W. Gess).

Identification of the prey

Twenty-eight prey individuals were recovered from the five provisioned nests examined
(Table 1). All were spiders of the genus Latrodectus (Theridiidae), 26 being L. geometricus
Koch and the remaining two L. mactans (Fabr.). Four were adult females, two were subadult
females and the remaining 22 were immatures, apparently all females.

Description of the nest

The nest of C. (H.) spinolae consists of a short, slightly upwardly inclined burrow termin-
ating in a single ovate mud-lined cell and sealed at the entrance with a mud plug (Fig. 4). The
average total length of the burrow is 36 mm (sample of 7), that of the entrance passage 15 mm
(sample of 4) and that of the cell 23 mm (sample of 4). The average diameter of the entrance
passage is 8 mm (sample of 7) and that of the cell at its widest point is 12 mm (sample of 5).
The nest closure is concave towards the outside and is 2-3 mm thick at the centre and some-
what thicker around its perimeter. A sealed nest is characterized by the presence of two quar-
ries of similar diameter to the nest closure and situated with their edges 4-20 mm from it.

Method of nest construction, provisioning and oviposition

Nest construction follows three distinct phases: excavation of the burrow, lining of the
cell, nest closure. Water is required for working the substrate during all three phases. This is
collected from a nearby pool, the wasp, whilst filling her crop, standing on the earth at the
water’s edge.

142

GESS ET AL.: ETHOLOGY OF CHALYBION (HEMICHALYBION) SPINOLAE (LEPELETIER)

Table 1.

Details of the prey recovered from five provisioned nests of C. (H.) spinolae.

Date

Nest

no.

Prey (listed in decreasing order of weight)

No.

Identity: sp. of
Latrodectus

Developmental stage, sex,
individual bearing wasp’s

egg-

Individual
wt. (mg)

Total
wt. (mg)

18.3.82

3

4

geometricus

ad. 9 (one of these)

268

586

geometricus

subad. 9 (two with egg)

208

geometricus

imm. $

81

geometricus

imm. 9

29

19.3.82

4

3

mactans

ad. $

318

858

geometricus

ad. 9

300

geometricus

ad. 9 (with egg)

240

20.3.82

5

7

geometricus

imm. 9 (with egg)

194

723

geometricus

imm. 9

168

geometricus

imm. 9

136

geometricus

imm. 9

91

mactans

imm. 9

54

geometricus

imm. 9

41

geometricus

imm. 9

39

23.3.82

6

3

geometricus

subad. 9 (with egg)

207

391

geometricus

imm. 9

139

geometricus

imm. 9

45

25.3.82

7

11

geometricus

imm. 9

143

714

geometricus

imm. 9

98

geometricus

imm. 9 (with egg)

81

geometricus

imm. 9

81

geometricus

imm. 9

70

geometricus

imm. 9

55

geometricus

imm. 9

53

geometricus

imm. 9

43

geometricus

imm. 9

43

geometricus

imm. 9

32

geometricus

imm. 9

15

Excavation of the burrow is by wetting the chosen nest site or the working face of the de-
veloping cavity with regurgitated water, by working this moisture into the substrate with the
aid of the mandibles, and by forming the resultant mud into pellets. These are not carried
away to be disposed of at a distance but are simply dropped at the mouth of the burrow and
consequently collect at the foot of the bank immediately below the nest.

Burrow excavation having been completed, lining of the cell is commenced. For this pur-
pose mud, mixed in a quarry close to the burrow entrance (Fig. 3) is introduced into the nest.
At the quarry only a small quantity of water is regurgitated from the crop at any one time, just
enough for the formation of a single pellet. Holding this in her mandibles, the wasp walks to
the nest which she enters head first. Application of the introduced mud to the surface of the
excavated cell is aided by vibrations produced by manipulation of the flight mechanisms and

143

ANN. CAPE. PROV. MUS. (NAT. HIST.) VOL. 14, PT. 5, OCTOBER 1982

transmitted through the mandibles. These vibrations are accompanied by a buzzing sound
clearly audible from without the nest. The application of that pellet of mud having been com-
pleted the wasp backs out of the nest and on foot returns to the quarry to form another pellet.
The sequence is repeated until all the water in the crop is used up. The number of pellets pro-
duced per water load will vary with the nature and moisture content of the substrate. How-
ever, at Hilton in the sand pit 5-8 pellets were formed per water load. It is possible that, due
to the porous nature of the substrate, additional water may be regurgitated from the crop to
aid in the application and smoothing of the cell lining. Indicative of this is that a greater num-
ber of pellets (14) was formed per water load during the construction of the nest closure which
is in contact with the dry substrate only around its perifery. The water in the crop having been
exhausted, the wasp walks along the bank for a distance of 25-50 cm before taking off and
flying to the water source from which she returns rapidly, flying directly to the nest and alight-
ing at the entrance. Several such cycles of water collection, quarrying and mud ap-
plication/smoothing are required for the completion of the cell lining. Under the conditions at
the Hilton sandpit during the period of observation water collection took two minutes and
quarrying and cell lining five minutes in each cycle.

Cell filling having been completed the wasp flies off in search of prey. The length of time
which elapses before the wasp returns with the first prey is presumably dependent upon the
success of the hunt. During observations at Hilton it varied from 30 minutes to several hours.
However, a source of prey spiders having been discovered further prey are brought in in quick
succession. In one instance (Nest 7) eleven spiders were brought in within an hour, the first six
in the amazingly short time of 14 minutes which would indicate that they must have been found
all together in one place. The return with the seventh and eighth spiders took 13 and 16 min-
utes respectively and the remaining four 3-8 minutes each.

If the hunt is of considerable duration the wasp will return to inspect her nest at intervals
of about 15 minutes. During these inspections she touches the bank around the nest entrance
with her antennae. Furthermore, she appears to mark the nest for, before flying off again, she
turns round and rubs the tip of her abdomen around the rim of the nest entrance.

When flying in with prey the wasp alights on the bank immediately below and facing the
nest entrance and, walking, enters the nest rapidly. The prey spider is held beneath the wasp,
orientated dorsum-up and facing the direction of travel. The mandibles grip the paralysed prey
at its anterior end and the first pair of legs support it from below. After a very short time the
wasp emerges rapidly from the nest, pauses to groom her face and antennae, walks along the
bank as before and then flies off. Whereas during the construction of the nest the wasp takes
little notice of the presence of other insects or of a human observer, during provisioning they
were “buzzed” on her departure from the nest even if this meant changing her course of flight.

Within the nest the prey are positioned on their backs or sides, facing the inner end of the
cell. Oviposition takes place on one of the spiders during its positioning within the cell for dur-
ing provisioning the wasp enters the nest only when carrying prey. In the five provisioned cells
examined (Table 1) the spider bearing the wasp’s egg was mostly of large size though not
necessarily the largest. Adult, subadult or very large immature spiders were therefore chosen
though in Nest 7 oviposition was on a medium-sized immature. The order of prey introduction,
as deduced from the position of the individual spiders within the cell, was recorded in full de-
tail only for Nest 6. In that nest the first prey was the egg-bearing subadult 9 of 207 gms, fol-
lowed by the immatures of 45 gms and 139 gms. In Nests 3, 4 and 7 only the position of the
egg-bearing spider was noted. This in each case was at the innermost end of the cell. Lack of
such data for Nest 5 was due to the accidental disarrangement of the prey during cell opening.
From the evidence it appears that in at least four of the five nests examined oviposition was on
the first spider to be introduced into the cell. However, as it is likely that a spider has to be of
a certain minimum size if it is to be suitable as the initial food source of the newly-hatched

144

GESS ET AL.: ETHOLOGY OF CHALYBION (HEMICHALYBION) SPINOLAE (LEPELETIER)

wasp larva, it is probable that oviposition upon the first-introduced prey takes place only if it
attains or exceeds this size. If smaller, it is likely that oviposition is delayed until a large prey
individual is subsequently introduced.

The egg of C. (H.) spinolae is white, curved and measures 3,68 x 1,01 mm (average of 5).
It is attached medially on one side or other of the anterior part of the spider’s abdomen (Fig.

Fig. 5. Diagrammatic representation of immature 9 Latrodectus geometricus showing position of attachment of egg of
C. (H.) spinolae.

Provisioning of the cell having been completed, the wasp immediately flies off to collect
water for the sealing of the nest. Mud for this purpose is mixed in a new quarry which like the
first (formed by extraction of material for cell lining) is close to the nest entrance. Pellets for
sealing the nest are laid down spirally. The nest opening having been closed, more pellets are
added to thicken the plug which when complete is not flush with the surface of the bank but is
somewhat concave. In an observed instance of nest sealing which took 11 minutes, the plug
was formed of 28 mud pellets, 14 being produced and positioned for each of two water loads.
Whereas the nest as a whole is sealed by the construction of the above-described mud plug, the
cell itself is not sealed nor is the entrance passage filled. No temporary closures of the nest are
made at any time during the provisioning of the cell.

145

ANN. CAPE. PROV. MUS. (NAT. HIST.) VOL. 14, PT. 5, OCTOBER 1982

A nest having been completed, the wasp begins another that may be closely associated
with the first. Thus excavation of Nest 6 was initiated within the second quarry associated with
the sealed Nest 5, the distance between the two nests being therefore less than 1 cm. The next
nest excavated by the female, Nest 7, was in its turn initiated within the concavity represented
by the remains of the exposed and longitudinally cut away cell of Nest 6 after its excavation by
the authors. It is probable that under undisturbed conditions a female may construct a series of
closely associated nests, each excavated within a quarry of the nest immediately preceding it.

Life history

After removal from the cells and subsequent examination the contents of Nests 5, 6 and 7
were placed in cotton wool stoppered glass vials of similar dimensions to the wasp cells and
kept indoors at room temperature. Two eggs, those from Nests 5 and 6, were viable. Larval
hatching occurred 5 and 7 days after oviposition and feeding took place on the first prey for a
period of 8 days. A further 4 days were required to complete feeding and to reach larval matu-
rity. In feeding the larvae first consumed the abdomens of all the spiders making up the provis-
ion, then turned to the cephalothoraxes and legs previously left intact. Eventually only a few
leg fragments remained uneaten. Cocoon spinning commenced two or three days after the
cessation of feeding. The winter months were spent in a state of pre-pupal diapause.

The cocoon of C. (H.) spinolae is pale brown and translucent, the insect within being
clearly visible, and is attached to the walls surrounding it by sparse silken threads. It varies in
length from 23-25 mm and in breadth at its midlength from 7-8 mm, widens slightly towards
the rounded anterior end and narrows to 5, 5-6, 5 mm at the truncate posterior end at which
the meconium is deposited.

An attempt was made to establish the number of generations which occur during a season
by estimating the age of female specimens. As the number of specimens from Hilton was very
small, material from other localities in the eastern Cape was included in the analysis. The man-
dibles of females vary considerably in length due to wear resulting from nest excavation. In or-
der to eliminate errors due to variation in wasp size, the ratio of mandible length to face width
was used as a measure of age. These parameters showed a linear relationship on unworn speci-
mens (r = 0,91; n = 22). The resulting age distributions (Fig. 6) show that at the beginning of
the season only young wasps are present. Adults do not therefore survive the winter. By Janu-
ary, the proportion of young wasps is much lower and in February very few freshly-emerged

3 9 17 7 3 2

N D J F M A

Fig. 6. Percent age distribution from November to April of female C. (H.) spinolae from the eastern Cape. The width of
the base line of each histogram represents 100%. Sample size appears above each month. Classes 1 to 6 on the vertical
scale represent increasing age, as indicated by mandibular wear (face width/mandibular length), and have the following
intervals: 1(0,951-1,020), 2(1,021-1,090), 3(1,091-1,160), 4(1,161-1,230), 5(1,231-1,300), 6(1,301 + ).

146

GESS ET AL.: ETHOLOGY OF CHALYBION (HEMICHALYBION) SPINOLAE (LEPELETIER)

wasps are found. There is an indication that a second generation of wasps may emerge during
March. However, since the sample size is so small, much more data are required before this
conclusion can be substantiated. In addition, ages of both male and female specimens were es-
timated according to four categories of wing tatter. Although sample size was still small (?n =
41; dn = 26), both sexes showed similar age distributions which were moreover in agreement
with those obtained using mandibular wear in females.

DISCUSSION

The subgenus Hemichalybion Kohl, as presented by Bohart and Menke (1976: 103) is con-
stituted of five Old World species: the African clypeatum (Fairmaire) and spinolae (Lepeletier)
[= eckloni (Dahlbom)], the southern European and western Asian femoratum (Fabricius), and
the Asian fuscum (Lepeletier) and sumatranum (Kohl).

Biological data attributed to three species of Hemichalybion, clypeatum, femoratum and
spinolae, have been published by various authors. However, the similarity in general appear-
ance and in colour pattern between Hemichalybion clypeatum and H. spinolae, and the wide-
spread and common Sceliphron spirifex (L.) occurring sympatrically with them has, it would
appear, been the cause of at least two authors being confused about the identity of the species
upon which they were reporting.

According to van der Vecht (see Bohart and Menke, 1976: 102) the extensive papers by
Verlaine (1924, 1926) on a species called clypeatum actually pertain to S. spirifex. In the light
of the present account of the nesting of H. spinolae it is clear that a similar confusion of identi-
ty exists in the literature between this species and 5. spirifex. Arnold (1928: 243) following his
systematic account of H. spinolae stated that it is “a common insect, frequently seen in houses,
with an inconvenient partiality for building its mud nests at the angles of walls, or on furniture,
books and boxes”. The above quoted statement characterizes the nesting behaviour of S. spiri-
fex under synanthropic conditions and clearly does not pertain to H. spinolae. It is of interest
that in the same paper Arnold gave no biological notes following his account of S. spirifex. Ar-
nold’s continued misconception concerning the nesting of H. spinolae is evident from a state-
ment made by him in correspondence with van der Vecht (see van der Vecht, 1961: 256).

The nesting of H. femoratum has been the subject of two reports: Rudow (1912) recorded
it using a Eumenes mud nest, and Bonelli (1969) recorded it from an old mud nest which Bo-
hart and Menke (1976: 102) presumed was abandoned by a species of Sceliphron. A similar as-
sociation with aerial mud nests is recorded by van der Vecht (1961: 256) with respect to H. cly-
peatum for, according to a note in the Paris Museum, this species is a “commensal des nids de
la moyenne Synagris". The validity of these associations cannot either be proved or disproved
at this stage. However, in view of the confusion of specific identity attending other accounts,
there is a need for confirmation of the reported association of Hemichalybion species with the
aerial mud nests of other aculeate wasps.

No doubt influenced by the reports of Rudow, Bonelli and van der Vecht, Bohart and
Menke (1976: 102) expressed the view that “quite likely Arnold made the same mistake with
spinolae that the Peckhams and Rau (early papers) made in their observations of C. californi-
cum in not realizing that the wasp does not build its own mud nest, but takes empty Sceliphron
nests for its own”. This has now been shown to be incorrect.

Brauns (1911: 118) was nearer the truth when he wrote concerning H. spinolae : “In der
Nistweise nahert sie sich den Chalybion- Antn, indem sie ihre Zellen ebenfalls in Hohlungen,
innerhalb von Mauern, Erdwanden etc. verbirgt, welche dunkel sind and nur eine kleine Zu-
gangsoffnung aufweisen”. (In its manner of nesting it approaches the Chalybion species, in
that it likewise conceals its cells in hollows within walls, earth banks etc., which (hollows) are
dark and only exhibit a small entrance hole.) Though correct in saying that H. spinolae nests in
cavities in earth banks Brauns was mistaken in believing that these cavities were pre-existing

147

ANN. CAPE. PROV. MUS. (NAT. HIST.) VOL. 14, PT. 5, OCTOBER 1982

ones and in classing the species with species of Chalybion (sensu stricto) which are gallery rent-
ers, not gallery excavators.

That H. spinolae nests in vertical banks where it excavates cavities itself, rather than utiliz-
ing pre-existing cavities, was indicated by Gess (1981: 24) on the basis of information then
available and is confirmed in the present paper.

Both van der Vecht (1961) and Bohart and Menke (1963 and 1976) have attempted to elu-
cidate the relationship between Chalybion (sensu stricto), Hemichalybion and Sceliphron.
These attemps have, however, been largely thwarted by the overall dearth of information con-
cerning the biology of Hemichalybion and by the misleading nature of at least some of the lim-
ited data attributed to individual species.

The present study has shown that Hemichalybion (as represented by H. spinolae) shares
with Chalybion (sensu stricto) and Sceliphron the habit of constructing its nests in situations
raised up above the ground. It has, however, shown also that Hemichalybion differs from both
the other taxa in its behaviour with respect to nest construction. Thus whereas Hemichalybion
excavates cells within vertically-presented earthen banks, Chalybion seeks and modifies pre-
existing cavities and Sceliphron builds free aerial cells.

With respect to its nest building behaviour Hemichalybion may therefore be seen as the
least specialized of the three taxa. This view accords well with the statement by van der Vecht
(1961: 256) that morphologically Hemichalybion species are clearly very close to the hypo-
thetical ancestor of the Chalybion-Hemichalybion-Sceliphron complex. There is moreover no
difficulty in accommodating the suggested behavioural and morphological status of Hemichaly-
bion within the dendrogram given by Bohart and Menke (1976: Fig. 83) to express their
thoughts on the relationships of the genera in the tribe Sceliphronini.

Within the .Sceliphronini there appear to have been strong pressures to abandon nesting in
friable soil for it is only in the two most primitive genera, the mantid-hunting Stangeella and
the cricket-hunting Chlorion , that the females possess foretarsal rakes and excavate their nests
by digging. In all the other genera, divisible into two distinct lines — the predominantly cock-
roach-hunting Penepodium, Dynatus, Podium and Trigonopsis and the exclusively spider-hunt-
ing Chalybion (including Hemichalybion ) and Sceliphron , the females lack foretarsal rakes but
make use of mud in nest construction.

Behaviourally most primitive of these rake-less forms are probably those species of Pene-
podium, P. luteipenne (Fabricius) and P. haematogastrum (Spinola), the nesting of which was
studied by Williams (1928: 118-126 and Figs 179, 180, 183, 185-189, 192 and 193) in British
Guiana and in Brazil respectively. P. luteipenne , of which Williams gave the fuller account (as
Podium flavipenne Latreille), was found to nest in bare, hard, solid ground, either level or
sloping, in close proximity to water. Excavation of the simple, shallow, single-celled burrow
was effected by the jaws, regurgitated water being used to soften the soil. The excavated ma-
terial was deposited about the hole in the form of mud pellets. The nest was left open until
fully provisioned when it was sealed with a plug of mud mixed with the aid of regurgitated
water at a quarry site a few inches removed from the nest.

The nesting behaviour of Hemichalybion spinolae is in many respects very similar to that
of Penepodium luteipenne but does represent an advance for instead of nesting in the ground,
the surface presentation of which is horizontal or sloping, this species has left the ground and
nests in earth banks, the surface presentation of which is vertical. Furthermore H. spinolae
lines its burrow with a layer of mud, a habit not recorded for P. luteipenne.

Behaviourally more advanced are the species of Chalybion (sensu stricto) and their nest-
form homologues of the cockroach-hunting line — Penepodium goryanum (Lepeletier) and
species of the genus Podium and possibly Dynatus , the biologies of which have been reviewed
by Bohart and Menke (1976). These forms do not themselves hollow out nesting cavities but
seek and modify pre-existing ones in situations above the ground. Mud is used for cell parti-

148

GESS ET AL.: ETHOLOGY OF CLIALYBION (HEMICHALYBION) SPINOLAE (LEPELETIER)

tions and for nest closures. More advanced still are the species of Sceliphron and their nest-
form homologues, species of Trigonopsis , all of which construct free aerial nests made of mud.

Three species of spider-hunting Sceliphronini have been found nesting at Hilton: Chaly-
bion (Hemichalybion) spinolae (Lepeletier), Chalybion (Chalybion) tibiale (Fabr.) and Sceli-
phron spirifex (L.). Whereas C. (H.) spinolae is on account of the nature of its nest restricted
to vertical earthen banks the other two species are not so limited but may nevertheless nest
there also. Thus during the summer of 1982 S. spirifex constructed its aerial nests on the sand-
pit bank not far from the excavated nests of C. (H.) spinolae and all three species were pre-
viously recorded (Gess, 1981: 24, 25 and 27) in association with the “sandstone” bank. Under
such circumstances a certain degree of specific interaction may well occur. Though C. (H.)
spinolae and S. spirifex construct very different nests, interspecific competition for nesting sites
may arise when there is a scarcity of the overhangs and similar sheltered situations favoured by
both. C. (C.) tibiale clearly would not enter into such competition for nesting sites but may yet
be associated with them by virtue of the probable adoption and modification for its own use of
the old abandoned nests of the other two species.

With respect to the prey utilized by the three species for cell provisioning a high degree of
resource partitioning appears to pertain. As far as can be judged from the present paper C.
(H.) spinolae appears to be restricted to species of Latrodectas (Theridiidae). C. (C.) tibiale
nesting in trap-nests sited in riverine bush alongside the sandpit has previously been recorded
(Gess and Gess, 1980: 13) as provisioning principally with Argiopidae and in addition with a
small number of Theridiidae (Rhomphaea sp.) and Zodariidae. Most commonly represented
among the Argiopidae were species of Isoxya, Nephila and Cyclosa whereas present only in
small numbers were species of Araneus, Argiope and Caerostris. In contrast, Sceliphron spiri-
fex at Hilton and elsewhere in the Grahamstown district has been found to provision exclusive-
ly with Argiopidae of the genus Araneus.

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. Gratitude is expressed to the
C.S.I.R. by F. W. G. and A. J. S. W. for running expenses grants for field work during the
course of which the present observations were made.

REFERENCES

Arnold, G. 1928. The Sphegidae of South Africa. Part 10. Ann. Transv. Mns. 12 (3): 233-279.

Bohart, R. M. and Menke, A. S. 1963. A reclassification of the Sphecinae with a revision of the nearctic species of the
tribes Sceliphronini and Sphecini. Univ. Calif. Pubis Ent 30: 91-182.

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

Bonelli, P. 1969. Osservazioni biologiche sugli Imennotteri melliferi e predatori della Val di Fiemme. Boll. 1st. Ent.
Univ. Bologna 29: 149-154, 165-172.

Brauns, H. 1911. Biologisches fiber siidafrikanische Hymenopteren. Z. wiss. Insekt. Biol. 7 (4): 117-120.

Gess, F. W. 1981. Some aspects of an ethological study of the aculeate wasps and the bees of a karroid area in the vicin-
ity of Grahamstown, South Africa. Ann. Cape Prov. Mas. ( nat . Hist.) 14 (1): 1-80.

Gess, F. W. and Gess, S. K. 1980. The whited sepulchre— the nesting of Chalybion tibiale (Fabr ). The Eastern Cape
Naturalist 70: 11-14.

Rudow, F. 1912. Lebensweise und Nestbau der Raub-, Mord-, und Grabwespen, Sphegidae und Crabronidae. Ent. Z.

Frankf. a. M. 26:. 30-32, 35-36, 39-40, 42-44, 46, 54-55, 59-60, 64, 66-67, 70-72, 75-76.

Van der Vecht, J. 1961. fiber Taxonomie und Evolution der Grabwespengattung Sceliphron Klug. Verhandl. XI. Inter-
natl. Kong. Ent. Wien 1: 251-256.

Verlaine, L. 1924. L'instinct et l’intelligence chez les Hymenopteres 11. L'instinct de nidification chez le Pelopaeus cly-
peatus Kohl du Congo beige. Annls Soc. ent. Belg. 64: 197-237.

Verlaine, L. 1926. Les moeurs de Pelopaeus clypeatus , guepe magonne du Congo. Revue zool. afr. 13: 174-180.
Williams, F. X. 1928. Studies in tropical wasps— their hosts and associates (with descriptions of new species). Bull.
Hawaiian Sug. Plrs' Ass. Exp. Stn (Ent.) 19: 1-179.

149

ANNALS

OF THE APR2 8 1983

CAPE PROVINCIAL -A®? .arils,,..

MUSEUMS

NATURAL HISTORY

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

VOLUME 14 • PART 6
15th NOVEMBER 1982

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

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

BD0847

Ethological studies of Isodontia simoni (du Buysson), /. pelopoeiformis
(Dahlbom) and I. Stanley i (Kohl) (Hymenoptera: Spheeidae: Spheeinae) in
the Eastern Cape Province of South Africa

by

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

CONTENTS

Page

Abstract 151

Introduction 152

Description of nesting sites 152

Flight period 153

Young growth, flowers and dry inflorescences visited 153

Identification of the prey 155

Description of the nest 156

Method of construction of the nest, provisioning and oviposition 156

Life history 162

Sequence of sexes in nests 165

Cannibalism 166

Parasites and other associated insects 167

Sheltering and sleeping habits of adult wasps 167

Discussion 167

Acknowledgements 170

References 170

ABSTRACT

Some aspects of the ethology of three southern African species of Isodontia , I. simoni (du
Buysson), I. pelopoeiformis (Dahlbom) and I. Stanley i (Kohl) are described. Particular atten-
tion is paid to: the situation and nature of the nest; the nature of the materials introduced into
it and the manner of their arrangements; provisioning; oviposition. With respect to I. pelopoei-
formis, the most common species, accounts are given also of the life history, sequence of sexes
in the nests, cannibalism and parasites. Certain aspects of the nesting of a total of 14 Isodontia
species are briefly reviewed. I. simoni is shown to differ from the other species in that the pre-
existing cavity chosen for nesting is situated in horizontal ground and not in vertical banks, in
plants, or in other situations above ground level. On the basis of the situation of its nest and
the nature of the materials introduced into it, I. simoni is considered to be primitive, a conclu-
sion which agrees with that reached by other authors from a consideration of structural charac-
ters.

151

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 6, NOVEMBER 1982
INTRODUCTION

Fifty-four species of the genus Isodontia Patton (Sphecidae: Sphecinae) are listed by Bo-
hart and Menke (1976). Of this total, five species occur in and are restricted to the Afro-tropi-
cal Region. I. longiventris (Saussure), I. pelopoeiformis (Dahlbom), I. simoni (du Buysson)
and I. stanleyi (Kohl) are found on the African mainland and I. leonina (Saussure) on the is-
land of Malagasy.

Biological data pertaining to fourteen species, including two from Africa, have been pub-
lished. Smithers (1958) gave a detailed account of the nesting of I. pelopoeiformis in trap-nests
in Zimbabwe (formerly Southern Rhodesia) and the same species in Angola was the subject of
a note by Heinrich (1969). Gess (1981) gave brief accounts of the nesting in the Eastern Cape
Province of South Africa of both I. pelopoeiformis and I. stanleyi and suggested that a third
species, I. simoni, occurring sympatrically with the first two, nested in similar situations. The
subsequent discovery and study of nests of I. simoni has shown the suggestion to have been in-
correct and that this species nests in a situation fundamentally different from those recorded
for all other Isodontia species.

The present paper, a comparative account of the nesting of the three species of Isodontia
known to ocgur in Africa south of the Limpopo, is the thirteenth in a series of publications
dealing with the ethology of certain solitary wasps occurring at Hilton, a farm situated 18 kilo-
metres WNW of Grahamstown (33° 19'S, 26° 32'E) in the Albany Division of the Eastern
Cape Province of South Africa. Nesting of I. pelopoeiformis and I. stanleyi is described from
Hilton but that of I. simoni is described from a neighbouring farm Thursford as, although this
species occurs at Hilton, its nesting site on this farm was not located.

A detailed account of various aspects of the ecology of Hilton has previously been given
(Gess, 1981: 3-9).

DESCRIPTION OF THE NESTING SITES

Four nests of I. simoni , 85 nests of I. pelopoeiformis and 18 nests of I. stanleyi were stu-
died.

The nests of I. simoni are constructed in pre-existing cavities in the ground. The nesting
cavities investigated were subvertical burrows, 64-95 mm long and of average bore 12 mm.
They did not appear to be the burrows of a hymenopteran as they showed no evidence of any
previous cells. However, they agreed in appearance and dimensions with the burrows of the
cockroach, Pilema thoracica (Walker), which are common in the nesting area of I. simoni (Fig.
1). The investigated nests were in an eroded area of clayey non-friable soil sparsely covered by
dwarf karroid scrub (Fig. 2). The karroid scrub in the areas studied on both Thursford and Hil-
ton is characterized by Pentzia incana (Compositae), however, that on Thursford includes
many more succulent species than that on Hilton.

The nests of I. pelopoeiformis and of I. stanleyi are constructed in pre-existing cavities
above the ground. All the nests of I. pelopoeiformis were in trap-nests having cavities of length
155 mm and bore either 9,5 mm or 12,7 mm. These trap-nests were suspended horizontally at
heights of from 10-210 cm above the ground in shrubs and small trees of Acacia karroo (Legum-
inosae), Maytenus linearis (Celastraceae) and Rhus lancea (Anacardiaceae) in thorn scrub
and riverine bush (Fig. 3). Eight nests of I. stanleyi were in trap-nests of the same dimensions
as and similarly situated to those occupied by I. pelopoeiformis. The other ten nests of this
species were in naturally occurring pre-existing cavities which were abandoned nesting galleries
of carpenter bees (Anthophoridae: Xylocopinae), three of Xylocopa sicheli Vachal in dry inflor-
escences of Aloe ferox (Liliaceae) (Gess 1981: Fig. 38) and seven of Xylocopa scioensis Gribo-
do (as X. caffrariae Enderlein, a junior synonym of X. scioensis, in Gess, 1981) in dry culms of
Phragmites australis (Gramineae).

152

GESS & GESS: STUDIES OF ISODONTIA (HYMENOPTERA: SPHECIDAE: SPHECINAE)

Fig. 1. Burrow entrance of Pilema thoracica in I. simoni nesting site on Thursford.

The favoured nesting sites of I. pelopoeiformis and I. stanleyi would appear from trap-
nesting data to be within the riverine bush and thorn scrub. The nests of /. stanleyi in Aloe
ferox and Phragmites australis are probably on the fringes of the mam nesting areas as these
plants occur respectively on the slopes above the riverine bush in proximity to thorn scrub and
fringing a river bed and the associated riverine bush (Gess 1981: Figs 13-15). A search was
made for nests in woody stems with little success. However, branches of Acacia karroo bored
by the larvae of Ceroplesls hottentota (Fabricius) (Coleoptera: Cerambycidae) were found to
contain nesting materials of the type used by I. stanleyi.

FLIGHT PERIOD

The three species of Isodontia are proterandrous. /. simoni has been collected at Hilton
from the beginning of November to early December and at Thursford was nesting in January.
I. pelopoeiformis and /. stanleyi at Hilton fly from December to late February and nest from
late December and from mid December, respectively, to late February.

YOUNG GROWTH, FLOWERS AND DRY INFLORESCENCES VISITED

At Hilton a male of I. simoni was collected on flowers of Lasiospermum bipinnatum
(Compositae) (3.xi.l977) and three females were recorded on the flowers of Senecio pterophor-
us (Compositae) (29. xi. 1979, l.xii.1979 and 5.xii. 1979). Three males of I. pelopoeiformis were
collected on Acacia karroo , two on young growth (6.xii. 1972 and 5.xii. 1973) and one on flow-
ers (4. i. 1978). Numerous females were observed visiting flowers of Acacia karroo. However, as
there were trap-nests in the vicinity only one specimen was collected (3. i. 1977) to confirm the
identity. There are no records of 1. stanleyi visiting flowers.

153

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 6, NOVEMBER 1982

Fig. 2. A nesting site of I. simoni, an eroded area on Thursford.

Fig. 3. A nesting site of I. pelopoeiformis and I. stanleyi , riverine bush on Hilton. The figure indicates a bundle of trap-

nests suspended in Maytenus linearis.

154

GESS & GESS: STUDIES OF ISODONTIA (HYMENOPTERA: SPHECIDAE: SPHECINAE)

I. simoni and /. pelopoeiformis visit seeding plants to collect fluff for the construction of
nest plugs. It is likely that any suitable fluff of this type will be taken. At Thursford I. simoni
collected fluff derived from the fruit of plants of the family Asclepiadaceae which are common
in its nesting area and to a lesser degree made use of fluff from flowers of Compositae. I. pelo-
poeiformis at Hilton used only fluff from flowers of Compositae and was observed collecting
this from Lasiospermum bipinnatum and Senecio species. No fruiting plants of Asclepiadaceae
were present in the vicinity of its nests.

I. stanleyi visits flowering grasses (Gramineae) including Danthonia curva , D iplachne fus-
ed, Eragrostis sp. (probably E. curvuld), Melica racemosa and Sporobolus sp. (probably S.
fimbriatus) to collect materials, short lengths of leaf blades and longer lengths from inflores-
cences including stems and flowers, for the construction of nest plugs.

IDENTIFICATION OF THE PREY

The prey of all three species of Isodontia consists of various species of Tettigoniidae: Phan-
eropterinae (katydids). Nymphs of both sexes are most commonly taken but sometimes in the
case of the smaller prey species also adults of both sexes.

Species represented amongst the prey examined included the following: Terpnistria zebra-
ta (Serville) which is easily recognized by the highly characteristic saddle-shaped pronotum
(Fig. 4), Eurycorypha prasinata Stal and two species of Phaneroptera. One cell of /. stanleyi
was found to contain in addition to several Phaneropterinae two specimens (an adult male and
a nymphal female) of Oecanthus capensis Saussure (Gryllidae: Oecanthinae), a tree cricket.

Provisioning by /. pelopoeiformis appears to be effected by the wasp’s hunting whatever
species are common in the immediate vicinity of the nest. Marked differences are consequently
found in the composition of the prey used by this species in provisioning nests located in differ-
ent plant communities. Thus 372 of a total of 377 prey katydids in 83 cells constructed in trap-
nests suspended from Acacia karroo in thorn scrub were nymphs of Terpnistria zebrata where-
as only two of a total of 179 prey katydids in 37 cells constructed in trap-nests suspended from
Maytenus linearis and Rhus lancea in the riverine bush were of this species. In the riverine bush

Fig. 4. Provision of a single cell of I. pelopoeiformis , 5 individuals of Terpnistria zebrata all with antennae cut short.
Middle prey of upper row bears the wasp’s egg. (x 1,18).

155

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 6, NOVEMBER 1982

Eurycorypha prasinata (nymphs with one exception) and Phaneroptera species (nymphs and
adults) constitued the prey commonly used for provisioning.

Provisioning by I. Stanley i in the riverine bush, where most of its nests were located, ap-
peared to be with the same species as or similar species to those taken there by /. pelopoeiform-
is. However, I. stanleyi generally took smaller prey individuals than did I. pelopoeiformis .

Provisioning by I. simoni at Thursford was with small nymphs of Phaneroptera species.

DESCRIPTION OF THE NEST

The nests of /. simoni, I. pelopoeiformis and I. stanleyi consist of a number of serially ar-
ranged cells each sealed with a plug, the series frequently preceded by a preliminary plug and
always succeeded by a closing plug which fills the space between the last cell and the nest en-
trance (Figs 5-8).

The materials from which the plugs are constructed are in each species of two types. In
nests of I. simoni and I. pelopoeiformis the materials are “a”, clods of earth with, in the case
of I. pelopoeiformis , the addition of debris collected off the ground including pieces of stick,
bark, vetch burs, insect remains and small mammal droppings, and “b”, plant fluff. The pre-
liminary plug in the nests of I. simoni is a thin layer of material “b” and where present in nests
of I. pelopoeiformis is of either or both materials in distinct layers. The cell partitions in nests
of I. simoni are constructed of materials “a” with an admixture of a little of material “b”.
However, those of I. pelopoeiformis are constructed almost exclusively from material “b”. The
closing plug in the single completed nest of I. simoni consisted of material “b” sealed at the
ground surface with a thin layer of wet dung. The closing plugs of nests of I. pelopoeiformis are
constructed in two distinct layers, the first of “a” and the second of “b”. Exceptionally, when
material “b” is in short supply lengths of very soft grass leaf blades are used as a substitute by
I. pelopoeiformis in the construction of cell partitions. It may, however, be omitted from the
nest closure.

The materials from which the plugs in nests of I. stanleyi are constructed are “c”, short
lengths of grass blades, and “d”, longer lengths cut from grass inflorescences including lengths
of unbranched culm and the branched portions including the flower heads. The cell partitions
are constructed from material “c” transversely arranged, the preliminary plug and the inner
part of the closing plug from material “d” transversely coiled and the outer part of the closing
plug from material “d” arranged longitudinally, the stems lying parallel, and projecting from
the nest entrance.

In order that the cells should be of adequate volume their lengths vary according to the dia-
meter of the cavity used. The nests of I. pelopoeiformis and I. stanleyi in trap-nests were used
for analysis. Due to there being a considerable variation in the lengths of the cell closures the
lengths of the storage space of each cell was measured for comparison, not the length of the
entire cell (Table 1). Cell length is dependent also upon the sex of the wasp that will develop
within it. From Table 2 it can be seen that, in I. pelopoeiformis at least, cells which will cradle
female wasps are larger than those which will cradle male wasps and that the diameter of the
boring in the lower ranges becomes a limiting factor as only males are produced in 9,5 mm bor-
ings.

METHOD OF CONSTRUCTION OF THE NEST, PROVISIONING AND OVIPOSITION

Nesting activity is at its height at midday. Nest construction is initiated by the wasp’s sel-
ecting a suitable pre-existing cavity and cleaning out any small pieces of debris. A preliminary
plug may then be constructed at the inner end of the cavity, or in the case of the two trap-nest-
ing species, some distance from it if sonfe foreign object such as extensive spider-spinnings ob-
structs the end. In a sample of 81 I. pelopoeiformis nests in trap-nests preliminary plugs were
present in 60 and ranged in thickness from 1-36 mm and in a sample of eight I. stanleyi nests in

156

GESS & GESS: STUDIES OF ISODONTIA (HYMENOPTERA: SPHECIDAE: SPHECINAE)

PUNT FLUFF

CLODS OF EARTH

Fig. 5. Plans of four nests of I. simoni , a, b and d incomplete, c with final closure of dung.

trap-nests preliminary plugs were present in six and ranged in thickness from 1-32 mm. Mat-
erial, be it fluff, debris or grass, for the construction of this and succeeding plugs is collected
by the female with her mandibles and held by them is carried in flight by her to the nest. In all
three species, whilst the nesting female is compacting the materials, buzzing sounds can be
heard within the nest. Collecting of nesting materials from source by I. pelopoeiformis females
was observed. They were gathering debris from the ground in close proximity to their nests
and collecting fluff from inflorescence heads on which they alighted. Collection and positioning
of material by this species is very rapid, each cycle of departure from the nest, collection of
material, return to the nest and compaction of the material taking an average 1,3 minutes. The
wasp usually pauses on the threshold of the nest entrance to groom her antennae before each
excursion.

After the cavity has been prepared for the reception of the first prey, loosely packed mat-
erial, fluff in the case of /. simoni and /. pelopoeiformis and grass in the case of /. stanleyi , is
assembled at a distance between the nest entrance and the closed end of the cavity to form a
temporary closure through which the wasp will pass on entering and leaving the cell during
provisioning. Hunting is then commenced. A prey katydid is captured, subdued by stinging
and its antennae are pruned. The wasp then holds it with her mandibles and legs and flies with
it held beneath her, head forwards and dorsum uppermost. In the trap-nesting species it was
observed that having arrived at the nest the wasp puts down the prey within the entrance, en-

157

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 6, NOVEMBER 1982

PLANT FLUFF

Vv"

7;y,

LENGTHS OF FINE GRASS
LEAF BLADES

EARTH AND DEBRIS

CLODS

AND DISTINCT PIECES OF DEBRIS

. 100 mm

A

Figs 6 and 7. Plans and photographs of eight completed nests of I. pelopoeiformis in trap-nests.

158

GESS & GESS: STUDIES OF ISODONTIA (HYMENOPTERA: SPHECIDAE: SPHECINAE)

159

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 6, NOVEMBER 1982

1QQ mm

LENGTHS OF GRASS LEAF BLADE ARRANGED

TRANSVERSELY

LENGTHS OF GRASS
INFLORESCENCE ARRANGED

V

TRANSVERSELY

LONGITUDINALLY

Fig. 8. Plan and photograph of one completed nest of I. stanleyi in a trap-nest and plan of a second.

ters, turns around and pulls the prey in towards her. She positions it in the cell so that it faces
the inner end and lies either on its back or on one of its sides. She usually oviposits on this
prey before leaving the nest to hunt for a second prey. If an egg is not laid on the first prey to
be introduced into the cell, it will be laid on the second. In the four nests of /. simoni exam-
ined the prey were found to lie horizontally across the cells. Due to the smallness of the prey in
relation to the diameter of the cell this was probably inevitable as it would only be possible to
position larger prey vertically. It appeared that oviposition had been on one of the first prey.

The egg is yellowish and slightly curved in all three species. Dimensions are given in Table
3. In all three species the egg is cemented to the prey by its anterior end. It is positioned in
front of the right or left mesothoracic coxa and lies transversely across the venter extending be-
yond it from the side away from the point of attachment.

The number of prey per fully provisioned cell is governed by the size of the prey and the
total weight of provision required. Thus, if the prey are large a smaller number is required
than if the prey are small. In Fig. 9 the numbers of prey per completed cell are plotted against
frequency for 111 cells of I. pelopoeiformis in trap-nests. As the range in number of prey is
from 1-10 and a peak is shown at 5 it is clear that an average sized prey is preferred but that
when not available a single large prey may be taken or conversely at the other extreme ten
very small prey. The prey taken by I. simoni and /. stanleyi were consistently smaller than

160

GESS & GESS: STUDIES OF ISODONTIA (HYMENOPTERA: SPHECIDAE: SPHECINAE)

those taken by I. pelopoeiformis and therefore, although these two wasps are smaller and con-
sequently have a smaller prey requirement measured by weight the actual number of prey per
cell tended to be larger than in cells of I. pelopoeiformis (Table 4).

The full provision for a cell having been supplied the wasp closes the cell. I. simoni seals
the cell with compacted clods of earth mixed with a little fluff leaving the temporary closure in-
tact. However, I. pelopoeiformis and I. stanleyi close their cells by compacting the material
which formed the temporary closure and adding to this further material of the same nature. In
these two species if a further cell is to be constructed, a new temporary closure is formed. The
thickness of the cell closure varies. The range was 7-40 mm in a sample of 132 I. pelopoeiform-
is cells (63 per cent being between 10 and 20 mm), 1-32 mm in a sample of 7 I. stanleyi cells
and 3-11 mm in a sample of 4 I. simoni cells. After the closure of the last cell has been com-
pleted the remaining space between it and the cavity entrance is filled by the construction of
the closing plug. The length of the closing plug varies. The range was 26-100 mm in a sample
of 77 nests of I. pelopoeiformis, 10-96 mm in a sample of 4 nests of I. stanleyi and 48 mm in a
single sealed nest of I. simoni.

Table 1.

Measurements of all sealed cells of Isodontia pelopoeiformis and of Isodontia stanleyi con-
structed within trap-nests, showing the relationship between length of cell lumen and boring
diameter and how it differs between the two species.

Cells constructed by
Isodontia pelopoeiformis

Cells constructed by
Isodontia stanleyi

Boring

No. of

Range in

Average

No. of

Range in

Average

diameter

cells

length of

length of

cells

length of

length of

cell

cell

cell

cell

lumen

lumen

lumen

lumen

(mm)

(mm)

(mm)

(mm)

(mm)

9,5

47

28-67

40

3

18-26

21

12,7

89

10-62

31

13

8-30

18

161

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 6. NOVEMBER 1982

Table 2.

Measurements of sealed cells of Isodontia pelopoeiformis constructed within trap-nests and
from which wasps were reared, showing the relationship between length of cell lumen, boring
diameter and the sex of the wasp for which the cell was constructed.

Cells from which male
wasps were reared

Cells from which female
wasps were reared

Boring

No. of

Range in

Average

No. of

Range in

Average

diameter

cells

length of

length of

cells

length of

length of

cell

cell

cell

cell

lumen

lumen

lumen

lumen

(mm)

(mm)

(mm)

(mm)

(mm)

9,5

38

28-60

40

0

—

—

12,7

30

10-58

29

29

20-55

36

Table 3.

Dimensions of the eggs of the three Isodontia species.

Isodontia sp.

Dimensions of eggs

Average length
(mm)

Average diameter
(mm)

Size of sample

simoni

3,61

0,88

3

pelopoeiformis

4,01

0,94

12

stanleyi

4,0

0,87

1

LIFE HISTORY

The life history of I. pelopoeiformis has been described in detail by Smithers (1955).
Stages in the development from egg to fully developed larva are shown in Fig. 10. In the pres-
ent study hatching of the larva of this species was found to take place from two to four days
after oviposition with most larvae hatching after three days. From four to eight days were
taken by the larva to consume the cell’s provision and to reach the full-grown state. The outer
loosely woven cocoon of yellow silk was completed within a day after the cessation of feeding
and the inner cocoon of mahogany to black parchment-like material was completed within a
further one or two days. The total length of time from oviposition to the construction of the in-
ner cocoon by the full grown larva therefore ranged from eight to fifteen days. The scant data
for I. stanleyi indicated similar times for this species. No data are available for I. simoni.

162

GESS & GESS: STUDIES OF ISODONTIA (HYMENOPTERA: SPHECIDAE: SPHECINAE)

•^t

U4

►J

«

<

H

3

X)

CO

Ml

U

£

T3

C

03

Ci

"C

C

OJ

CX

B

o

0

<u

■C

Cu

3

00

c

x2

CO

43

.2P

’S

£

T3

G

03

}-t

<D

X)

G

z

163

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 6, NOVEMBER 1982

Fig. 10. Stages in development of I. pelopoeiformis from egg to fully developed larva ( x 2,2).

164

GESS & GESS: STUDIES OF ISODONTIA (HYMENOPTERA: SPHECIDAE: SPHECINAE)

Development periods, from oviposition to emergence of the adult, were calculated for a
total of 45 males and 12 females of I. pelopoeiformis reared from trap-nests taken from the
field when completed and thereafter kept indoors at room temperature. Wasps developing
from eggs laid during the period from late December through to February of one summer
season emerged as adults during the period from January to March of the following summer
season. Developmental periods of 321-407 days (average 365 days) were recorded for males
and 351-431 days (average 401 days) were recorded for females. From these figures, unsatis-
factory as they are (see below), it is apparent that I. pelopoeiformis is univoltine under the cli-
matic conditions prevailing in the Eastern Cape Province. This conclusion is supported by the
relatively short flight season observed for this species in the field. The greater part of the year,
including the winter months, is therefore spent in a state of prepupal diapause. This situation is
different from that reported by Smithers (1958) for I. pelopoeiformis in Southern Rhodesia
(now Zimbabwe) where the species was found to be bivoltine at least.

As I. pelopoeiformis in the field flies and nests during the same weeks each summer, it is
clear that under natural conditions the development period from oviposition to adult emer-
gence must be considerably shorter than 365 days to allow for the length of the pre-oviposition
and pre-nesting period and for at least part of the length of the nesting period itself. The devel-
opment periods obtained during the present study are therefore unnaturally long and must be
ascribed to the fact that the trap-nests containing the developing wasps were kept indoors and
were therefore not subjected to the temperature and humidity conditions prevailing in the
field. The figures nevertheless show that the developmental period for males is shorter than
that for females. This is confirmed by the proterandry observed in the field.

When handled, newly emerged specimens of I, pelopoeiformis , both males and females,
have a very strong odour, somewhat akin to that of guava fruits or to that of the leaves of some
Eucalyptus species.

SEQUENCE OF SEXES IN NESTS

Certain species of aculeates which construct serially arranged cells within pre-existing cav-
ities commonly show a fixed sequence of sexes in nests in which both males and females devel-
op. This has been shown by Krombein (1967: 29) for species of Vespidae and some Megachil-
idae in which females develop in cells at the inner or blind end of the cavity and males in cells
at the outer or open end. To establish whether this is true also of I. pelopoeiformis , an analysis
of completed nests constructed within 12,7 mm trap-nests and from which at least some wasps
were reared was undertaken (Table 5).

Of the 33 nests analysed, seven were one-celled and therefore yielded no information con-
cerning the correlation between the position of a cell within the nest and the sex of the wasp
produced within that cell. Of the 26 two- and three-celled nests half included cells in which the
wasp egg had not hatched or in which the larva died before spinning its cocoon. In the remain-
ing 13 nests, all larvae had developed at least to the cocoon-spinning stage and 21 of the 31
cells had yielded adult wasps which were available for sexing. Notwithstanding the failure of
the 10 other wasps to develop beyond the cocoon-spinning stage, these individuals were also
available for sexing by virtue of the difference in length between the inner cocoons of males
(21-24 mm) and of females (27-29 mm).

Six nests each produced wasps of one sex only, 3 two-celled nests producing only females
and 1 two-celled nest and 2 three-celled nests producing only males. Seven nests, 4 two-celled
and 3 three-celled produced wasps of both sexes. In these, the females had without exception
developed within cells constructed and provisioned before those from which males developed.
Female-producing cells were therefore sited towards the inner or blind end of the trap-nest and
male-producing cells towards the outer or open end of the trap-nest. Though incomplete, the
data from the 13 nests in which only some wasps had reached the cocoon-spinning stage shows

165

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 6, NOVEMBER 1982

Table 5.

Analysis of completed I. pelopoeiformis nests constructed within trap-nests having a boring
diameter of 12,7 mm and from which at least some wasps were reared. The correlation be-
tween the position of a cell within the nest and the sex of the wasp produced within that cell is
shown for two- and three-celled nests. Number of nests = 33; number of cells = 68; number of

sexed wasps = 53.

Nature of nest

Cells in positional order and sex
of wasp produced within each

Frequency
within sample

Cell 1

Cell 2

Cell 3

One-celled

9

—

—

2

3

—

—

5

Two-celled

9

9

3

9

3

—

4

9

?

—

6

?

6

—

3

3

3

—

1

Three-celled

9

9

9

0

9

9

6

1

?

9

3

1

9

?

3

1

9

6

3

2

9

?

?

1

?

6

?

1

3

3

3

2

the same trends and are in no instances contradictory. It follows from the above that, if the
first cell in a nest is male-producing, all subsequent cells constructed within that nest by the fe-
male will also be male-producing.

CANNIBALISM

Four of the 132 cells of I. pelopoeiformis examined were found to contain not one egg-
bearing katydid but two. In their dimensions and provisioning these cells were not unusual: all
were in 12,7 mm trap-nests, cell lumens ranged in length from 32-59 mm, and the number of
prey ranged from 4-8. In the 59 mm long cell, however, the arrangement of the prey was atyp-
ical in that the katydids were divided into two distinct groups of four each as if belonging to
separate cells from which the dividing partition had been omitted. One katydid of both groups
bore an egg.

In one of the four cells the first-laid egg was not viable. In the remaining three cells (in-
cluding the cell with two groups of prey) the first-laid egg, identified as such by the position
within the cell of the katydid to which it was cemented, hatched before the second-laid egg. In
two instances hatchings were separated by one day and in the third instance by three days. The
two larvae co-existed in the cells for as long as the older larva was feeding on its first katydid- —

166

GESS & GESS: STUDIES OF ISODONTIA (HYMENOPTERA: SPHECIDAE: SPHECINAE)

that is, the katydid to which the egg had been cemented. However, after leaving this prey and
before starting to feed on a second katydid the older larva in all three instances located the
younger larva and ate it. Thereafter the older larva proceeded to feed on the katydid with
which the younger larva had been associated. It would appear that a larva having devoured its
first katydid actively seeks out any conspecific larva in the cell for in all three instances, and in
the cell with the divided provision in particular, the larvae were well separated spatially.

PARASITES AND OTHER ASSOCIATED INSECTS

Parasitic and other associated insects were recorded from the nests of I. pelopoeiformis
only. Undoubtedly this was due to the fact that the number of nests of this species that was
available for examination was far greater than that of the other two species. Nine of the 85
nests (10,6 per cent) were affected.

Most common and affecting six completed and sealed nests was an unidentified species of
scavenger fly (Sarcophagidae: Miltogramminae). The cleptoparasitic maggots, up to 17 per
nest, developed on the provision stored in the cells and in multicellular nests generally moved
from one cell to the next thus destroying the contents of all. In one three-celled nest, however,
only the second and third cells were affected, the first in time yielding a wasp. Developmental
periods were short, the larvae feeding for about a week after which they pupated in or near the
nest closure. The adult flies emerged from the puparia after about a further fifteen days.

Also cleptoparasitic was a species of Chrysididae affecting one nest. Four adults were
reared, the development time, from oviposition to emergence of the adult, being about 325
days.

Two nests were invaded and robbed by small black ants of the genus Crematogaster.

SHELTERING AND SLEEPING HABITS OF ADULT WASPS

During the period of nest construction, when nesting activities had ceased for the day and
during unfavourable weather, females of I. stanleyi utilizing trap-nests were frequently found
sheltering in a temporary closure of an incomplete nest or between such a closure and the trap-
nest entrance. However, despite the far greater number of nests of I. pelopoeiformis obtained
in trap-nests, females of this species were never found sheltering in nests and would therefore
appear to shelter elsewhere. No information on this facet of behaviour is available for I. simoni
females nor for males of any of the three species.

DISCUSSION

Some aspects of the nesting of 14 of the 54 species of Isodontia recognized by Bohart and
Menke (1976) have been published to date. Grouped by zoogeographical region, the 14 species
and the publications relevant to each are:

Afrotropical Region — /. pelopoeiformis (Dahlbom) (Smithers, 1958; Heinrich, 1969; Gess,
1981) and I. stanleyi (Kohl) (Gess, 1981);

Nearctic Region — I. auripes (Fernald) (Packard, 1869; Rau and Rau, 1918; Rau, 1926 and
1928; Krombein, 1967 and 1970), I. elegans (Smith) (Ashmead, 1894; Davidson, 1899; Fer-
nald, 1906; Ainslie, 1924; Bohart and Menke, 1963; Parker and Bohart, 1966; Krombein,
1967), I. exornata Fernald (Bequaert, 1930), I. mexicana (Saussure) (often under the name
harrisi Fernald) (Ashmead, 1895; Hubbard, 1896; Jones, 1904; Engelhardt, 1929; Rau, 1935
and 1943; Suehiro, 1937; Swezey, 1947; Lin, 1962 and 1966; Medler, 1965; Krombein, 1967)
and I. philadelphica (Lepeletier) (Bohart and Menke, 1963; Krombein, 1967);

Neotropical Region — I. costipennis (Spinola) (Mayer and Schulthess, 1923; Bristowe, 1925;
Richards, 1937) and /. paranensis (Berland) (Berland, 1929);

Palaearctic Region — I. harmandi (Perez) (Tsuneki, 1963 and 1964), I. maidli (Yasumatsu)
(Tsuneki, 1957, 1963 and 1964), I. nigella (Smith) (Piel, 1933; Tsuneki, 1963 and 1964), I.

167

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14. PT. 6, NOVEMBER 1982

paludosa (Rossi) (De Stefani, 1896 and 1901; Rudow, 1912; Berland, 1929 and 1959) and I.
splendidula (A. Costa) (Nicolas, 1894; Berland, 1929).

In addition to the zoogeographic distributions given above, I. nigella occurs in the Oriental and
Australian Regions and I. mexicana in Central America. The latter has also become estab-
lished in Hawaii and France.

Characteristic of all the above species is that nesting takes place in pre-existing cavities
situated above ground level. The actual height above the ground appears to be immaterial as is
evidenced by the prositional range given above for trap-nests utilized by two species at Hilton.
Most frequently the records concern the use of cavities associated with plants, less frequently
with vertical banks and their man-made counterpart, walls.

Records pertaining to nesting in association with plants may be divided into three broad
categories. The first category concerns nests made within tubes formed by leaves — I. costipen-
nis in a curled-up leaf, I. mexicana in folded dead leaves of yucca, in the open-ended tube-like
leaves of pitcher plants (Sarracenia spp.) and in the exposed hollow mid-ribs of dead fronds of
Pandanus palms, and I. nigella in a holed leaf of Allium.

The second category concerns nests made in hollow stems the ends of which have been
broken off or in which holes have been made by other animals. Records include the use of
stems of dicotyledonous plants such as white sage by I. elegans, thistles by I. paranensis and
probably an umbellifer by I. exornata. However, of greater frequency are records concerning
the use of culms of robust perennial Gramineae such as Arundo donax by I. splendidula ,
Phragmites australis by I. Stanley i and in particular bamboo by I. costipennis, I. elegans , I. har-
mandi , I. mexicana and I. nigella. Most records for bamboo pertain to culms which have been
cut and incorporated into man-made structures.

The third category concerns nests constructed in pithy of woody stems (including timber)
within abandoned nesting galleries of Xylocopa species by I. auripes, I. mexicana and I. stan-
leyi , and in old borings of cerambycid larvae by I. auripes , I. harmandi , I. maidli, l. nigella and
I. stanleyi. /. philadelphica is recorded from borings of unspecified origin.

Records pertaining to nesting within vertical clayey banks concern the use of the aban-
doned burrows of mining bees, those of Anthophora and Melitoma species by I. auripes and
those of the former only by I. elegans. I. paludosa is recorded nesting in a cavity in a stone
wall.

In view of the variety of naturally occurring pre-existing cavities utilized by the wasps it is
not surprising that no fewer than eight species have been recorded as accepting trap-nests of
various designs; similarly that records exist of bizarre nesting situations such as the metal tubes
of laboratory gas burners chosen by I. nigella (Iwata, 1976: 206) and the shot-gun barrel and
bicycle handle bars used by 1. pelopoeiformis (Heinrich, 1969: 113 and present authors, re-
spectively).

With regard to wasps and bees which do not excavate or hollow out their nests themselves
but modify pre-existing cavities, a recent community study at Hilton (Gess, 1981) showed that
at least some species are shared between vertical banks and plant tissue. This finding is be-
lieved to be of general application and appears relevant with respect to the above reviewed
Isodontia species. These appear to exhibit little specificity with regard to the nature of the pre-
existing cavities chosen for nesting provided these cavities are situated above the ground level.
For example, individuals of a single population of I. auripes in Missouri were recorded by Rau
(1926: 200) as utilizing both the abandoned galleries of carpenter bees in the wooden rafters of
a porch above a clay bank and the old burrows of mining bees in the clay bank itself. All the
nesting situations listed in the above review may therefore be considered to be homologous as
to type and it may be expected that any one species would, given the opportunity, nest in the
situations recorded for any of the others.

The nesting of I. simoni as described for the first time in the present publication is of con-

168

GESS & GESS: STUDIES OF ISODONTIA (HYMENOPTERA: SPHECIDAE: SPHECINAE)

siderable interest in that the pre-existing cavities used by this species are situated in level (that
is, horizontal) ground, a nesting situation hitherto unknown for any Isodontia species. Level
ground as a nesting situation does not represent merely yet another site to be added to those
listed above but rather represents one that is fundamentally different. This is indicated by the
finding in the already mentioned community study that the difference in the angle of present-
ation of a soil surface, vertical as opposed to horizontal, has a very profound influence on the
nesting of wasps and bees. In the community studied it was found that, with the possible ex-
ception of one species ( Pison allonymum), there was no sharing of species between level
ground and vertical banks and that indeed both nesting situations are characterized by com-
munities of species unique to themselves. Nesting in pre-existing cavities in level ground must
therefore be considered to be characteristic for I. simoni and it is highly unlikely that the
species will ever be found nesting elsewhere. Certainly at Hilton it was very striking and, be-
fore the nesting site was established, indeed puzzling that, in contrast to I. pelopoeiformis and
I. Stanley i, /. simoni was never found to utilize trap-nests suspended in bushes and trees.

In their discussion of the tribe Sphecini Bohart and Menke (1976: 106-108) pointed out
that most wasps in this tribe are fossorial and that the genus Isodontia (as also Chilosphex)
which nests in pre-existing cavities is presumed to have evolved from fossorial forms. Similarly,
Tsuneki (1963: 71) is of the view that burrow- or tube-renters like Isodontia are more advanced
than burrowing species like Sphex. In order to derive the nesting typical of the majority of Iso-
dontia species from that of the fossorial type as exhibited by species of Sphex the following
evolutionary stages in nesting behaviour may be postulated:

Digging of burrow Use of pre-existing Use of pre-existing

in friable soil of — » burrow (or other — > burrow (or other

horizontal aspect. cavity) in preferably cavity) in a vertical

non-friable soil of bank,

horizontal aspect.

Within this evolutionary sequence the nesting of I. simoni clearly belongs to the second step
whereas that of all the other species here reviewed belongs to the third and fourth steps. On
this basis the nesting of I. simoni must be seen as being more primitive than that of the other
species.

In view of the suggested primitiveness of the nesting of I. simoni it is of interest that
Bohart and Menke (1976: 120 and 121) list this species, together with I. pelopoeiformis and
two other species for which details of nesting are lacking, as among the least specialized, struc-
turally primitive species.

A structural character used by the above authors in assessing the degree of specialization
of various species is the form of the female mandibular apex. They point out that there is a
progression from the bidentate (primitive) to the tridentate (specialized) condition.

As the form of the mandibular apex is determined by function it is relevant to examine the
identity of the nesting materials that are manipulated by the mandibles during their collection,
transport and incorporation into the structure of the nest. From the relatively scant data avail-
able it would appear that the use of grass leaf blades and grass inflorescences is typical of those
species with trifid mandibles, such as I. auripes , I. elegans , I. mexicana, I. nigella , I. splendidu-
la and I. stanleyi , though the first three of these species as also I. hermandi are known to use
also bark fibres. Moss is in addition used by I. hermandi and I. maidli.

Plant fluff from the fruits of Compositae and Asclepiadaceae on the other hand is col-
lected by those species with bifid mandibles such as I. simoni and I. pelopoeiformis though the
latter is known very occasionally and exceptionally to use also very soft grass leaf blades, prob-
ably when pappus is not available. With respect to I. paludosa which is recorded as using both

Use of a pre-exist-
ing burrow (or
other cavity) asso-
ciated with plants.

169

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14, PT. 6, NOVEMBER 1982

thistle pappus and grass inflorescences it is perhaps significant that the form of the mandibular
apex is described by Bohart and Menke (1976: 120) as being more nearly bidentate that triden-
tate and as illustrating the intermediate state. Recorded nesting materials of 1. costipennis are
similarly asclepiadaceous or apocynaceous seed hairs: it would seem likely that this species has
a bifid mandible but this could not be confirmed.

On the basis of the correlation between the form of the mandibular apex and the nature of
the plant-derived nesting materials it is apparent that the use of plant fluff is primitive and the
use of grass is advanced. The primitiveness of the nesting of I. simoni suggested by the situa-
tion of its nest is therefore confirmed by the nature of the plant material used for nest construc-
tion.

If, as is suggested, the nesting of Isodontia is to be derived from the fossorial type as ex-
hibited by Sphex , then the use of clods of earth by the ground-nesting 1. simoni is hardly sur-
prising and is clearly more primitive than its use of plant fluff. Similarly, the retention as nest-
ing materials by I. pelopoeiformis of clods of earth and the use of other objects picked up off
the ground can be seen as relict behaviour carried over from the original nesting site, a pre-
existing cavity in the ground, to the derived nesting site, a pre-existing cavity raised above the
ground.

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, they wish to thank
Mr G. J. H. C. Willetts of the farm Thursford for permission to investigate insects there.
Thanks are also due to Mr C. D. Eardley of the Plant Protection Research Institute, Pretoria
for furnishing the valid name for the carpenter bee nesting in Phragmites australis. Particular
thanks and recognition are accorded to the authors' youngest son, Robert, for discovering and
drawing attention to the nests of I. simoni and thereby ending what had been a fruitless search
of several years duration. 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

Ainslie, C. N. 1924. Note on the nesting habits of Chlorion elegans. Can. Ent. 56: 269-270.

Ashmead, W. H. 1894. The habits of the aculeate Hymenoptera. Parts 1-4. Psyche , Camb. 7: 19-26, 39-46, 59-66 and
75-79.

Bequaert, J. 1930. Nesting habits of Isodontia , a subgenus of Chlorion (Hymenoptera). Bull. Brooklyn ent. Soc. 25 (2):
122-123.

Berland, L. 1929. Notes sur les Hymenopteres de France. XIII. La nidification du Sphex paludosus Rossi et des Iso-
dontia en general. Bull. Soc. ent. Fr. 1929: 63-66.

Berland, L. 1959. La nidification du Sphex paludosus. Bull. Soc. ent. Fr. 64 (9/10): 195-197.

Bohart, R. M. and Menke, A. S. 1963. A reclassification of the Sphecinae with a revision of the Nearctic species of the
tribes Sceliphronini and Sphecini (Hymenoptera, Sphecidae). Univ. Calif. Pubis Ent. 30 (2): 91-182.

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

Bristowe, W. S. 1925. Notes on the habits of insects and spiders in Brazil. Trans. R. ent. Soc. Lond. 1925: 475-504.
Davidson, A. 1899. Sphex elegans. Ent. News 10: 179-180.

De Stefani, T. 1896. Sulla nidificazione e biologia dello Sphex paludosus Rossi. Naturalista sicil. (n.s.) 1: 131-136.

De Stefani, T. 1901. Ulteriori osservazioni sulla nidificazione delle Sphex paludosus. Monitore zool. ital. 12: 222-223.
Engelhardt, G. P. 1929. An observation on the breeding habits of Chlorion harrisi in Texas. Bull. Brooklyn ent. Soc.
23 (5): 269-271.

Fernald, H. T. 1906. The digger wasps of North America and the West Indies belonging to the subfamily Chlorioninae.
Proc. U.S. natn. Mus. 31: 291-423.

Gess, F. W. 1981. Some aspects of an ethological study of the aculeate wasps and the bees of a karroid area in the vicin-
ity of Grahamstown, South Africa. Ann. Cape Prov. Mus. (nat. Hist.) 14 (1): 1-80.

Heinrich, G. H. 1969. The charcoal digger wasp, Isodontia pelopoeiformis. Fla Ent. 52: 113.

Hubbard, H. G. 1896. Some insects which brave the dangers of the pitcher plant. Proc. ent. Soc. Wash. 3: 314-316.

170

GESS & GESS: STUDIES OF ISODONTIA (HYMENOPTERA: SPHECIDAE: SPHECINAE)

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.

Jones, F.M. 1904. Pitcher plant insects. Em. News 15: 14-17.

Krombein, K. V. 1967. Trap-nesting wasps and bees: life histories, nests, and associates. Washington, D.C. Smithsonian
Press.

Krombein, K. V. 1970. Behavioral and life-history notes on three Floridian solitary wasps (Hymenoptera: Sphecidae).
Smithson. Contr. Zool. 46: 1-26.

Krombein, K. V. et al. 1979. Catalog of Hymenoptera in America North of Mexico. Vol. 2 Apocrita (Aculeata). Wash-
ington, D.C. Smithsonian Press.

Lin, C. S. 1962. Biology and nesting habits of hunting wasp Isodontia harrisi Fernald. Tex. J. Sci. 14: 429-430.

Lin, C. S. 1966. Bionomics of Isodontia mexicana with a review of generic ethology. Wasmann J. Biol. 24: 239-247.
Mayes, J. and Schulthess, A. von 1923. Beobachtungen an Nestern von geselligen und solitaren Wespen. Mitt. Ent.
Zurich 6: 357-367.

Medler, I. T. 1965. Biology of Isodontia (Murrayella) mexicana in trap nests in Wisconsin. Ann. ent. Soc. Am. 58:
137-142.

Nicolas, M. 1894. Le Sphex splendidulus da Costa. Assoc. Francaise Avancement Sci., Compte Rendu 22 Sess. . Bes an-
con 1893 (2): 636-647.

Packard, A. S. 1869. Guide to study of insects. Salem.

Parker, F. D. and Bohart, R. M. 1966. Host-parasite associations in some tv/ig-nesting Hymenoptera from western
North America. Pan-Pacif. Ent. 42: 91-98.

Piel, O. 1933. Recherches biologiques sur les Hymenopteres du Bas Yang-tse (Chine). Sphex ( Isodontia ) nigellus F.
Smith. Annls Soc. ent. Fr. 102: 109-154.

Rau, P. 1926. The ecology of a sheltered day bank; a study in insect sociology. Trans. Acad. Sci. St Louis 25: 157-277.
Rau, P. 1928. Field studies in the behavior of the non-social wasps. Trans. Acad. Sci. Si Louis 25: 319-489.

Rau, P. 1935. The grass-carrying wasp, Chlorion (Isodontia) harrisi Fernald. Bull. Brooklyn ent. Soc. 30: 65-68.

Rau, P. 1943. The nesting habits of certain sphecid wasps of Mexico, with notes on their parasites. Ann. ent. Soc. Am.
36: 647-653.

Rau, P. and Rau, N. 1918. Wasp studies afield. Princeton, N. J. Princeton University Press.

Richards, O. W. 1937. Results of the Oxford University Expedition to British Guiana, 1929. Hymenoptera, Sphecidae
and Bembecidae. Trans. R. ent. Soc. Lond. 86 (7): 101-118.

Rudow, F. 1912. Lebensweise und Nestbau der Raub-, Mord-, und Grabwespen, Sphegidae und Crabronidae. Ent. Z.

Frankf. a. M. 26: 30-32, 35-36, 39-40, 42-44, 46, 54-55, 59-60, 64, 66-67, 70-72, 75-76.

Smithers, C. N. 1958. Notes on the life history of Sphex ( Isodontia ) pelopoeiformis Dhlb. in Southern Rhodesia. Ento-
mologist’s mon. Mag. 94: 193-197.

Suehiro, A. 1937. No title. (Note on Isodontia harrisi Fernald.) Proc. Hawaii, ent. Soc. 9 (3): 358.

Swezey, [O. H.j 1947. No title. (Note on Isodontia harrisi Fernald.) Proc. Hawaii, ent. Soc. 13 (1): 8.

Tsuneki, K. 1957. Biology of Sphex ( Isodontia ) maidli Yasumatsu (Hymenoptera, Sphecidae). Akitu 6: 25-26.

Tsuneki, K. 1963. Comparative studies on the nesting biology of the genus Sphex (s.l.) in East Asia (Hymenoptera,
Sphecidae). Mem. Fac. lib. Arts Fukui Univ., Ser. II, Natural Science, No. 13 (2): 13-78.

Tsuneki, K. 1964. Supplementary notes on the nesting biology of three species of Sphex (Isodontia) occurring in Japan
(Hymenoptera, Sphecidae). Etizenial: 1-14.

171

A3

i

ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

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

VOLUME 14 • PART 7
28th FEBRUARY 1983

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

, -jSs3KK*i

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

BD0968

Historical atlas of the diurnal raptors of the Cape Province (Aves:

Falconiformes)

by

A. F. BOSHOFF

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

C. J. VERNON

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

and

R. K. BROOKE

(Percy FitzPatrick Institute of African Ornithology, University of Cape Town, Rondebosch,

7700)

CONTENTS

Page

Abstract 174

Introduction 175

Survey methods 176

Distribution, habitat and status of individual taxa 180

Sagittarius serpentarius Secretary Bird 184

Gypohierax angolensis Palmnut Vulture 186

Gypaetus barbatus Lammergeyer 188

Neophron percnopterus Egyptian Vulture 190

Necrosyrtes monachus Hooded Vulture 192

Gyps coprotheres Cape Vulture 194

Gyps africanus Whitebacked Vulture 196

Torgos tracheliotus Lappetfaced Vulture 198

Trigonoceps occipitalis Whiteheaded Vulture 200

Milvus migrans migrans Black Kite 202

Milvus migrans parasitus Yellowbilled Kite 204

Elanus caeruleus Blackshouldered Kite 206

Aviceda cuculoides Cuckoo Hawk 208

Pernis apivorus Honey Buzzard 210

Aquila verreauxii Black Eagle 212

173

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983

Page

Aquila rapax Tawny Eagle 214

Aquila nipalensis Steppe Eagle 216

Aquila wahlbergi Wahlberg’s Eagle 218

Hieraaetus pennatus Booted Eagle 220

Hieraaetus fasciatus Hawk Eagle 222

Hieraaetus ayresii Ayres’ Eagle 224

Lophaetus occipitalis Longcrested Eagle 226

Polemaetus bellicosus Martial Eagle 228

Stephanoaetus coronatus Crowned Eagle 230

Circaetus cinereus Brown Snake Eagle 232

Circaetus gallicus Blackbreasted Snake Eagle 234

Terathopius ecaudatus Bateleur 236

Haliaeetus vocifer Fish Eagle 238

Buteo buteo Steppe Buzzard 240

Buteo tachardus Mountain Buzzard 242

Buteo rufofuscus Jackal Buzzard 244

Accipiter rufiventris Redbreasted Sparrowhawk 246

Accipiter ovampensis Ovambo Sparrowhawk 248

Accipiter minullus Little Sparrowhawk 250

Accipiter melanoleucus Black Sparrowhawk 252

Accipiter badius Little Banded Goshawk 254

Accipiter tachiro African Goshawk 256

Micronisus gabar Gabar Goshawk 258

Melierax canorus Pale Chanting Goshawk 260

Circus ranivorus African Marsh Harrier 262

Circus pygargus Montagu’s Harrier 264

Circus macrourus Pallid Harrier 266

Circus maurus Black Harrier 268

Polyboroides typus Gymogene 270

Pandion haliaetus Osprey 272

Falco peregrinus Peregrine 274

Falco biarmicus Lanner 276

Falco subbuteo Hobby 278

Falco cuvierii African Hobby 280

Falco chicquera Rednecked Falcon 282

Falco amurensis Eastern Redfooted Kestrel 284

Falco tinnunculus Common Kestrel 286

Falco rupicoloides Greater Kestrel 288

Falco naumanni Lesser Kestrel 290

Polihierax semitorquatus Pygmy Falcon 292

Discussion 294

Acknowledgements 296

References 296

ABSTRACT

Maps showing the past and present (1970s) distribution of 55 taxa of breeding and non-
breeding diurnal raptor occurring in the Cape Province, South Africa, are presented. The corn-

174

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)

pilation of the maps is such that any major range fluctuations which have taken place on a tem-
poral basis are indicated. Almost a quarter of the taxa show a decrease, most show no change
and only two indicate an increase. The value of this information for long-term monitoring stu-
dies and for conservation purposes is mentioned. The limitations of the survey are discussed.

UITTREKSEL

Kaarte word vertoon wat die vorige en huidige (1970s) verspreiding van 55 soorte telende
en nie-telende dags-aktiewe (diurnal) roofvoels wat in die Kaapprovinsie, Suid-Afrika, voor-
kom, aandui. Die samestelling van die kaarte is sulks dat enige grootskaalse gebiedsverande-
rings wat oor ’n tydperk plaasgevind het, aangedui word. Ongeveer ’n kwart van die soorte
toon ’n afname, die meeste wys geen verandering en slegs twee het toegeneem. Die waarde
van hierdie inligting vir lang-termyn monitorstudies en vir bewaringsdoeleindes word genoem.
Die beperkings van die opname word bespreek.

INTRODUCTION

This paper documents the past and present distribution and status of the breeding and
non-breeding diurnal raptors, all members of the Falconiformes, of the Cape Province. In
terms of the conservation requirements of the breeding birds of any area, it is necessary to
define and monitor their status. As a group raptors require detailed study as they are located at
the ends of various food chains and reflect major fluctuations taking place along such chains.
Thus they are considered to be reliable indicators of the state of the environment. There is
little doubt that the various agricultural practices employed throughout the Cape Province (and
elsewhere in southern Africa) have greatly altered the habitat and that this must have a mark-
ed effect on raptor populations. Due to their predatory and, in the case of vultures, carrion
eating habits, raptors suffer as a result of various forms of indirect persecution, e.g. poisoning
and trapping for jackals, Canis mesomelas , and other mammalian “vermin”. Whereas pesticide
levels appear to be low in southern African raptor populations, mortality caused by electrocu-
tion is a recent and increasing factor. In addition many raptors are accused of causing death
and damage to domestic livestock and poultry and as a result are directly persecuted by farm-
ers.

Whereas certain species may be able to adapt to these changes, viz. the “generalists”,
others have become extinct or face extinction, viz. the “specialists”. Some of the changes that
have already taken place in raptor populations in the eastern Cape Province have been de-
scribed by Skead (1967) and Vernon (1978). A further indication of their parlous conservation
state is the relatively large number of raptors included in the South African Red Data Book:
Aves (Siegfried et al. 1976).

It is hoped that an overall broad-based survey will allow for the ranking of the raptors of
the Cape Province in terms of priority for detailed research and conservation action. In par-
ticular the survey was seen as an attempt to update the two most frequently used works which
present distribution maps for the diurnal raptors of the Cape Province, viz. McLachlan &
Liversidge (1978) and Snow (1978).

One way of determining the dynamic status of a taxon is to monitor its distribution ( =
range). Here two categories could be applied. Firstly those with a drastically reduced range
and therefore also a reduced total population. This situation can be caused by both direct and
indirect persecution and in terms of conservation species in this category provide the most
cause for concern. Secondly there are those taxa with an unchanged or slightly reduced range
but with a reduced total population. This situation can be caused mainly by direct persecution
although indirect persecution remains an important factor. In the cases where there is no re-
duction in range and where population levels have remained stable there is little cause for con-
cern at present. In other words, it is the authors’ contention that an unaltered range may rep-

175

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983

resent a relatively stable population whereas a marked decrease in range may represent a real
decrease in numbers. Whereas it may be possible to obtain some idea of distributional changes
for certain taxa, it is virtually impossible to make any detailed statement on changes in popu-
lation levels because of the lack of data.

SURVEY METHODS

All available published and unpublished sources containing references to raptors in the
Cape Province were consulted, records were extracted and entered into the computer data-
bank scheme outlined by Boshoff et al. (1978). Where possible the following information was
coded: taxon, locality, date, relative and actual abundance, sex, age and breeding data. At-
tempts were made to ensure the accuracy of the contents of the databank, but it became clear
that many records, from both published and unpublished sources, were suspect. Where poss-
ible records were awarded a “reliability” category (cf. Boshoff et al. op cit.) but in many cases
this was not possible. In the event of an obvious error the record concerned was rejected; often
this was not easy and some subjective decisions were made. Taxa which had 50 or fewer records
for the entire 1700 - 1979 period were particularly carefully scrutinised. Maps were prepared
with symbols indicating presence per locus (= one “quarter-degree square”).

On a map one symbol per locus may represent one or more records for the particular lo-
cus. Where several records were made for one species in one locus for one period this has been

Fig. 1 The Cape Province of South Africa, showing the various regions referred to in the text, and the Transkei.

176

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AYES: FALCONIFORMES)

taken as a single observation. The term “resident”, as used in this paper, refers to a taxon liv-
ing and breeding in the Province, however, confusion is possible where a taxon is not seden-
tary but undergoes local migration or is nomadic. Degree of abundance has been limited to
three subjective categories, viz. “common”, “scarce” and “rare”.

The general regions of the Cape Province referred to in the text, viz. E. Cape, N. Cape,
N.W. Cape, S. Cape, S.W. Cape and Karoo, are outlined in Fig. 1. These regions were named
for convenience and coincide approximately with the major vegetation types of the Province
(Fig. 2). In Fig. 2 the broad veld types shown incorporate the following veld types, as de-
scribed by Acocks (1975):

Forest

Thornveld

Bushveld

Macchia

Karoo

Grassland

Nos 1,2, 4

Nos 7, 21, 22, 23

Nos 16, 17

Nos 47, 69, 70

Nos 24 to 43

Nos 44,48,50,59,60

Although no fieldwork has been carried out by the Cape Department of Nature and En-
vironmental Conservation in the Transkei since independence was granted to that territory in

177

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983

1976, all available data from that area, as well as from the Ciskei and Bophuthatswana, have
been included in the databank. This was done because much of the early ornithological work
was carried out in the Transkei, then part of the Cape Province.

Fig. 3. The number of published and unpublished raptor records (to nearest 100) for the Cape Province and Transkei,
per decade, up to 1979. 1700 - 1899 taken as one decade.

178

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AYES: FALCONIFORMES)

Table 1

The number of published and unpublished records of species of diurnal raptor per decade in the Cape Province and Transkei up to the 1970s.

1700 - 1899 taken as one decade.

1700 -
1899

1900 -
1909

1910 -
1919

1920 -
1929

1930 -
1939

1940 -
1949

1950 -
1959

1960 -
1969

1970 -
1979

Total

Secretary Bird

36

16

33

18

25

19

71

60

351

629

Palmnut Vulture

—

—

—

—

2

1

i

3

2

9

Lammergeyer

22

7

4

6

3

4

15

64

53

178

Egyptian Vulture

29

2

2

2

1

1

3

2

4

46

Hooded Vulture

3

1

—

1

—

—

3

—

1

9

Cape Vulture

41

15

13

21

51

25

91

207

483

947

Whitebacked Vulture

1

—

—

—

—

1

5

16

50

73

Lappetfaced Vulture

13

2

1

1

2

—

13

9

25

66

Whiteheaded Vulture

4

—

—

1

—

—

—

3

4

12

Black Kite

1

—

—

—

1

—

5

2

31

40

Yellowbilled Kite

13

8

45

39

73

25

88

43

500

834

Blackshouldered Kite

65

46

105

59

87

67

96

132

1 799

2 456

Cuckoo Hawk

3

6

14

5

2

4

3

10

20

67

Honey Buzzard

—

—

2

1

1

1

—

3

—

8

Black Eagle

41

12

10

10

8

13

82

108

666

950

Tawny Eagle

18

5

5

1

3

2

11

14

65

124

Steppe Eagle

—

—

—

—

—

1

—

—

4

5

Wahlberg's Eagle

1

—

—

—

1

—

2

3

5

12

Booted Eagle

13

5

5

3

5

—

4

38

470

543

Hawk Eagle

1

—

—

—

—

—

1

—

1

3

Ayres’ Eagle

2

2

1

—

—

—

—

1

—

6

Longcrested Eagle

21

8

33

2

2

1

6

3

16

92

Martial Eagle

30

15

18

20

22

18

38

59

303

523

Crowned Eagle

36

15

13

13

16

14

35

57

251

450

Brown Snake Eagle

3

—

—

1

—

—

3

1

8

16

Blackbreasted Snake Eagle

17

3

2

1

1

1

21

6

39

91

Bateleur

12

5

23

6

2

3

8

10

41

110

Fish Eagle

30

9

6

4

8

26

56

88

328

555

Steppe Buzzard

28

19

45

31

30

22

62

57

786

1 080

Mountain Buzzard

1

7

1

1

1

2

4

14

74

105

Jackal Buzzard

54

37

56

20

34

35

74

74

1 169

1 553

Redbreasted Sparrowhawk

38

15

21

2

4

10

23

13

82

208

Ovambo Sparrowhawk

—

—

—

—

—

—

1

—

1

2

Little Sparrowhawk

22

15

19

8

8

9

18

13

31

143

Black Sparrowhawk

15

4

4

1

—

6

4

26

27

87

Little Banded Goshawk

2

—

1

—

1

—

1

—

4

9

African Goshawk

47

34

20

24

10

21

43

48

155

402

Gabar Goshawk

16

3

2

6

2

4

14

12

32

91

Pale Chanting Goshawk

26

17

14

12

5

3

55

38

850

1020

African Marsh Harrier

25

7

10

5

9

16

88

60

297

517

Montagu’s Harrier

8

1

4

1

3

—

7

2

2

28

Pallid Harrier

12

3

14

18

27

7

5

6

6

98

Black Harrier

31

6

3

2

3

3

26

47

470

591

Gymnogene

18

11

13

5

6

4

28

33

151

269

Osprey

6

1

5

7

7

17

27

19

116

205

Peregrine

23

6

2

4

3

2

19

15

86

160

Lanner

27

23

18

10

9

17

52

48

318

522

Hobby

9

—

—

2

—

1

4

8

9

33

African Hobby

1

—

—

1

—

—

— t

1

2

5

Rednecked Falcon

2

—

—

—

—

—

5

5

4

16

Eastern Redfooted Kestrel

4

—

20

1

3

—

1

6

7

42

Common Kestrel

52

60

51

43

77

63

80

61

1 481

1 968

Greater Kestrel

10

12

4

16

2

2

35

24

169

274

Lesser Kestrel

2

3

64

29

25

28

52

24

221

448

Pygmy Falcon

10

—

-

—

1

—

7

35

41

94

Total

945

466

726

464

586

499

1 396

1631

12 111

18 824

179

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983

It should be emphasised that the concept of the present atlas was actively promoted only
from early 1978 and therefore field surveys connected with other projects prior to this date did
not involve the recording of distributional data for all raptor taxa. Thus on certain field surveys
prior to 1978 the very common taxa were not recorded in the field. Attention is drawn to this
in the relevant taxon accounts.

Nomenclature and order of taxa follow Clancey (1980). As much of the data interpreta-
tion methodology was decided on a post hoc basis this aspect is discussed below.

DISTRIBUTION, HABITAT AND STATUS OF INDIVIDUAL TAXA

In total some 18 824 records of 55 taxa were obtained from the various sources with the
1970s having 1,8 times the number of records as for the entire pre-1970s period (Table 1 and
Fig. 3). Of the 55 taxa surveyed, some 34 are considered to breed in the Province. On a post
hoc basis it was decided to present two maps for each taxon. These maps show, respectively,
the pre- 1970s and 1970s distribution of each taxon in the Cape Province and part of the Trans-
kei; Transkei records east of 29° E are not mapped here. On the pre-1970s maps three periods

Fig. 4. Loci in the Cape Province and part of the Transkei where raptors were recorded between 1700 and 1969.

180

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AYES: FALCONIFORMES)

Fig. 5. Loci in the Cape Province and part of the Transkei where raptors were recorded between 1970 and 1979. “X"
denotes loci visited (once) but where no raptors were recorded.

fire differentiated, viz. 1700 - 1899 (open triangle), 1900 - 1949 (open circle), 1950 - 1969
(open square). On the 1970s maps a solid square symbol is used to denote presence of a taxon.
Here 1949/1950 was taken as a cut-off point due to the general increase in the number of re-
cords in the 1950s and which carried through to the 1970s (Fig. 3).

The coverage obtained by the survey in terms of loci from which raptors were recorded for
the pre-1970s and 1970s is shown in Figs 4 and 5 respectively and in Table 2. A comparison of
the coverage between the two periods is also made (Table 2). For both periods the percentage
coverage for the S.W. Cape, S. Cape and E. Cape is markedly greater than that for the N.W.
Cape, N. Cape and Karoo. However, Table 2 and Figs 4 and 5 should only be referred to in
terms of the constraints mentioned in the discussion.

The general range trend of each of the 55 taxa between the two periods is listed in Table
3. Individual taxon accounts follow.

181

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983

Table 2

Number of loci in the Cape Province in which raptors were recorded in the 1700 - 1969 and 1970 - 1979 periods

Region

Total

N.

Cape

N.W.

Cape

Karoo

S.W.

Cape

S.

Cape

E.

Cape

Total number of loci

285

101

379

92

82

163

1 102

1700 - 1969

Loci where raptors recorded

52

43

109

70

54

141

469

% coverage

18

43

29

76

67

87

43

1970 - 1979

Loci where raptors recorded

159

46

252

87

79

154

775

% coverage

56

46

66

95

96

94

70

Trend or Change

♦ * *

♦ * *

* *

* *

*

*** substantial increase
** marked increase
* small increase

Table 3

General range trend of the diurnal raptors of the Cape
Province between the periods 1700 - 1969 and
1970 - 1979.

Species

No

change

Some

increase

Some

decrease

Marked

decrease

Secretary Bird

*

Palmnut Vulture

Lammergeyer

*

Egyptian Vulture

*

Hooded Vulture

*

Cape Vulture

*

Whitebacked Vul-

ture

*

Lappetfaced Vul-

ture

Whiteheaded Vul-

ture

*

Black Kite

*

Yellowbilled Kite

*

Blackshouldered

Kite

*

Cuckoo Hawk

*

Honey Buzzard

*

Black Eagle

*

Tawny Eagle

Steppe Eagle

Wahlberg’s Eagle

*

Booted Eagle

?

Hawk Eagle

*

182

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AYES: FALCONIFORMES)

Species

No

change

Some

increase

Some

decrease

Marked

decrease

Ayres’ Eagle

*

Longcrested Eagle

*

Martial Eagle

*

Crowned Eagle

*

Brown Snake Eagle

*

Blackbreasted

Snake Eagle

*

Bateleur

*

Fish Eagle

*

Steppe Buzzard

*

Mountain Buzzard

*

Jackal Buzzard

*

Redbreasted Spar-

rowhawk

*

Ovambo Sparrow-

hawk

*

Little Sparrowhawk

*

Black Sparrowhawk

*

Little Banded Gos-

hawk

*

African Goshawk

*

Gabar Goshawk

*

Pale Chanting Gos-

hawk

*

African Marsh Har-

rier

*

Montagu’s Harrier

*

Pallid Harrier

*

Black Harrier

*

Gymnogene

*

Osprey

*

Peregrine

*

Lanner

Hobby

*

African Hobby

*

Rednecked Falcon

*

Eastern Redfooted

Kestrel

*

Common Kestrel

*

Greater Kestrel

*

Lesser Kestrel

*?

Pygmy Falcon

*

Total (N = 55)

40

2

6

7

(72,7%) (3,6%) (10,9%) (12,7%)

183

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Secretary Bird Sagittarius serpentarius

Distribution (Figs 6 and 7)

The distribution for the pre-1970s and 1970s accords with that given by McLachlan &
Liversidge (1978) and Snow (1978).

Habitat

All veld types, except forest.

Fig. 6. Distribution of the Secretary Bird, 1700 - 1969.

184

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A scarce resident species which shows no change in range in the Province between the two
periods. The large number of records from the south-eastern part of the N. Cape in the 1970s
is due to increased sampling. Some decrease in numbers may have taken place in the N.W,
Cape and Karoo. S. serpentarius is sensitive to habitat degradation due to overgrazing.

Fig. 7. Distribution of the Secretary Bird, 1970 - 1979.

185

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Palmnut Vulture Gypohierax angolensis

Distribution (Figs 8 and 9)

1700-1969: This distribution accords with that given by Snow (1978).

1970-1979: This distribution accords with that given by McLachlan & Liversidge (1978).

Habitat

There is no suitable habitat for this species in the Cape Province. Elsewhere in Africa the
species is associated with palm trees, especially oil palms, Elaeis.

Fig. 8. Distribution of the Palmnut Vulture, 1700 - 1969.

186

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A rare vagrant for which there are only nine records over a 50 year period, from 1931 on-
wards. The birds recorded in the Cape Province were most often immatures and were presum-
ably dispersing birds from the breeding populations in Zululand (Natal) and Mozambique.

G. angolensis is peripheral in the Cape Province and Transkei. This species is listed as
“vulnerable” in South Africa by Siegfried et al. (1976). There are few records from the Trans-
kei and E. Cape from the 1970s but it is not known to what extent this reflects the status of the
southern African breeding populations. G. angolensis is not considered to have undergone a
change in status in the Province between the two periods.

Fig. 9. Distribution of the Palmnut Vulture. 1970 - 1979.

187

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Lammergeyer Gypaetus barbatus

Distribution (Figs 10 and 11)

1700-1969: This distribution accords with that given by Snow (1978).

1970-1979: This distribution accords with that given by McLachlan & Liversidge (1978).

Habitat

Mountainous grassland areas. Previously extending down to the coast but now restricted
to above 1 700 m above sea level.

Fig. 10. Distribution of the Lammergeyer, 1700 - 1969.

188

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS ( AVES: FALCONIFORMES)
Status

G. barbatus has undergone a drastic decrease in range in the Province between the two
periods. Little is known of the biology of this species in South Africa and the reasons for this
decrease are therefore not understood. There is a small resident population of about 120 pairs
in the Drakensberg in Lesotho and outliers in Natal, Cape Province and Transkei (Brown
1977). G. barbatus is classified as “rare and threatened” in South Africa by Siegfried et al.
(1976).

Fig. 11. Distribution of the Lammergeyer, 1970 - 1979.

189

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Egyptian Vulture Neophron percnopterus

Distribution (Figs 12 and 13)

1700-1969: This distribution accords with that given by Snow (1978). It is noteworthy that 29
of the 42 records for this period are between 1700 and 1900 (cf. Table 1).
1970-1979: McLachlan & Liversidge (1978) do not present a distribution map but list the
species as “now very rare in our area (South Africa)”.

Habitat

Highveld open grassland and mountainous areas; also arid and semi-arid areas.

Fig. 12. Distribution of the Egyptian Vulture. 1700 - 1969.

190

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

Reported to have bred at several localities in the Cape Province and Transkei in the past;
last breeding record 1923, in the Transkei. N. percnopterus occurred throughout the Cape
Province before 1900 but since then the number of records has decreased considerably with
only four observations from the 1970s (Table 1). Snow (1978) states that “there is some evi-
dence that the species spread south, colonizing South Africa between 1810 and the 1860s, and
then after maintaining itself began to disappear after the turn of the century”. The status of N.
percnopterus in southern Africa has been documented in detail by Mundy (1978), who consid-
ers that there may be a viable relict breeding population in the Transkei. The species is extinct
in the Cape Province or occurs only as a rare vagrant.

Brooke (1979) and Mundy (op cit.) have discussed the reasons for the decrease in range
and numbers of this vulture. Siegfried et al. (1976) list N. percnopterus as “endangered and
possibly no longer a part of the South African (avi)fauna”.

Fig. 13. Distribution of the Egyptian Vulture, 1970 - 1979.

191

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Hooded Vulture Necrosyrtes monachus

Distribution (Figs 14 and 15)

1700-1969: This distribution accords with that given by McLachlan & Liversidge (1978) and
Snow (1978).

1970-1979: Mills (1976) records this species as a vagrant in the Kalahari Gemsbok National
Park during the early 1970s. The 1970s distribution is more restricted than that
given by McLachlan & Liversidge (op cit.)

Habitat

Bushveld in the Cape Province. Elsewhere in Africa in savanna and both open and dense
woodland; well wooded moist lowveld areas. Commensal with man north of the equator.

Fig. 14. Distribution of the Hooded Vulture, 1700 - 1969.

192

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

Overall there is a decrease in the frequency with which the narrow-billed N. monachus
was recorded in the 1970s as compared to the pre-1970 period. The lack of suitable habitat in
the Province has prevented N. monachus from becoming a breeding resident. The species was,
and still is, peripherally represented in the Cape Province and its range is allopatric with the
other thin-billed vulture Neophron percnopterus , which is an ecological equivalent in so far as
feeding is concerned. The paucity of records for Necrosyrtes monachus from the Province in
the 1970s may indicate a decrease in range; the species is considered rare in the Cape Province.

Fig. 15. Distribution of the Hooded Vulture, 1970 - 1979.

193

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Cape Vulture Gyps coprotheres

Distribution (Figs 16 and 17)

1700-1969: This distribution accords with that given by Snow (1978).

1970-1979: Similar to the 1700 - 1969 period but more restricted in the Karoo. This range is
more limited than that given by McLachlan & Liversidge (1978).

Habitat

Breeds and roosts on cliffs. Forages over country with few trees; mainly grassland, thorn-
veld, macchia and also the karoo at the ecotone with grassland.

Fig. 16. Distribution of the Cape Vulture, 1700 - 1969.

194

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS ( AVES: FALCONIFORMES)
Status

The past and present distribution and status of G. coprotheres in the Cape Province is
dealt with in detail by Boshoff & Vernon (1980a). This resident species has undergone several
fluctuations in range and numbers during the past 150 years and at the present time the popu-
lation has stabilised at a low level which reflects the existing food supply. Additional mortality
factors posing a threat to the population are drowning, electrocution and indirect poisoning. In
1979 there were only 65 breeding pairs in the Province and G. coprotheres is considered a
scarce species in the Province. G. coprotheres is rightly listed as “threatened and vulnerable”
in the South African Red Data Book: Aves (Siegfried et al. 1976).

Fig. 17. Distribution of the Cape Vulture, 1970 - 1979.

195

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Whitebacked Vulture Gyps africanus

Distribution (Figs 18 and 19)

In view of the number of records from the 1970s, the paucity of records from the pre-
1970s is noteworthy. This is due to the area being opened to farmers and travellers as late as
the 1930s and 1940s, and also as a result of incorrect nomenclature (as G. rueppelli). The dis-
tribution of G. africanus for the pre-1970 and the 1970 to 1979 periods accords with that given
by McLachlan & Liversidge (1978) and Snow (1978).

Habitat

Bush veld in the Cape Province. Elsewhere in Africa in tropical and sub-tropical savanna
and woodland. Breeds and roosts in trees.

Fig. 18. Distribution of the Whitebacked Vulture, 1700 - 1969.

196

BOSHOFF ET AL. : HISTORICAL ATLAS OF DIURNAL RAPTORS ( AVES: FALCONIFORMES)
Status

G. Africanus was undoubtedly more widespread in the N. Cape than the map for the
1700 - 1969 period suggests. At present it is the commonest vulture species in that area but it
is not known whether a decrease in numbers has taken place; if so this would be due mainly to
reduction in available carrion. A breeding population remains in the Kalahari Gemsbok Nat-
ional Park, where carrion is still freely available. The species also breeds near Kimberley (at
least 75 pairs in 1979) and there is a group of unknown size which nests N.W. of Vryburg (P. J.
Mundy in lift.). Within the Province the distribution of G. africanus is closely linked to that of
the camelthorn Acacia erioloba which it finds very suitable for nesting in. There has been no
change in range in the Province between the two periods.

Fig. 19. Distribution of the Whitebacked Vulture, 1970 - 1979.

197

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Lappetfaced Vulture Torgos tracheliotus

Distribution (Figs 20 and 21)

1700-1969: Recorded as a breeding species from the S.W. Cape, S. Cape, Karoo, E. Cape
and N. Cape during the 19th century. This distribution accords with that given by
Snow (1978).

1970-1979: This distribution accords with that given by McLachlan & Liversidge (1978).
Habitat

Bushveld in the Cape Province. Elsewhere in Africa in tropical and sub-tropical savanna
and open woodland. Particularly adapted to arid and semi-arid areas. Breeds and roosts in
trees.

Fig. 20. Distribution of the Lappetfaced Vulture, 1700 - 1969.

198

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

T. tracheliotus , a former breeding species over most of the Cape Province, has exhibited a
major decrease in range and is now absent from the Province except for the N. Cape where it
is a scarce but still resident species. The reasons for this decrease are not understood but may
be linked to the extirpation of the indigenous game herds. Breeding is recorded from the 1970s
in the Kalahari Gemsbok National Park and near Schmidtsdrif (P. J. Mundy in litt.) and also
near Kimberley (AFB pers. rec.; R. Liversidge pers. comm.). T. tracheliotus should be includ-
ed in a revised South African Red Data Book: Aves.

Fig. 21. Distribution of the Lappetfaced Vuiture, 1970 - 1979.

199

ANN. CAPE PROV. MUS. (NAT. HIST. ) VOL. 14 PT 7, FEBRUARY 1983

Whiteheaded Vulture Trigonoceps occipitalis

Distribution (Figs 22 and 23)

McLachlan & Liversidge (1978) and Snow (1978) exclude T. occipitalis from the Cape
Province.

Habitat

Bushveld in the Cape Province. Elsewhere in Africa in tropical savanna and woodland.

Fig. 22. Distribution of the Whiteheaded Vulture, 1700 - 1969.

200

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

T. occipitalis is, and probably always has been, rare in the Cape Province, which is on the
periphery of its range. The most southerly breeding records for South Africa are from Zulu-
land, Natal. Little is known of the movements of the species in the southern parts of its range
but it is considered a rare vagrant in the N. Cape, outside of the Kalahari Gemsbok National
Park where adults do occur. The habitat in the Kalahari Gemsbok National Park does not suit
breeding. There is, however, a record of an adult on a nest in July 1964 (Broekhuysen et al .,
1968). There are only 12 records from 1700 to 1969 and only four from the 1970s (Table 1).
There has been no apparent change in status in the Province between the two periods.

Fig. 23. Distribution of the Whiteheaded Vulture, 1970 - 1979.

201

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Black Kite Milvus migrans tnigrans

Distribution (Figs 24 and 25)

The present range (1970s) is more restricted than that given by McLachlan & Liversidge
(1978). Snow (1978) does not separate M. migrans and M. migrans parasitus.

Habitat

Wide variety of habitat, with preference for open wooded country in vicinity of water.

Fig. 24. Distribution of the Black Kite, 1700 - 1969.

202

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS ( AVES: FALCONIFORMES)
Status

A scarce non-breeding Palaearctic migrant which is present during the austral summer.
There has been no change in status in the Province between the two periods. The E. Cape and
Transkei records are mapped here, however, there is a strong possibility that in these areas
there have been some misidentifications as juvenile M. m. parasitus.

Fig. 25. Distribution of the Black Kite. 1970 - 1979.

203

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Yellowbilled Kite Milvus migrans parasitus

Distribution (Figs 26 and 27)

1700-1969: This distribution corresponds with that given by Snow (1978) for M. migrans.
1970-1979: This distribution is far more restricted than that given by McLachlan & Liversidge
(1978) sub nom. M. aegyptius.

Habitat

In the E. Cape M. m. parasitus is found in all habitats and breeds in forests and planta-
tions. In the S.W. Cape this taxon is found in areas of open macchia and cereal crops. In both
regions the taxon is frequently associated with habitation.

Fig. 26. Distribution of the Yellowbilled Kite, 1700 - 1969.

204

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A common breeding summer migrant which overwinters in equatorial Africa. The status
of M. m. parasitus remained unchanged in the Province between the two periods. This study
demonstrates the presence of two populations in the Cape Province.

Fig. 27. Distribution of the Yellowbilled Kite, 1970 - 1979.

205

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983

Blackshouidered Kite Elanus caeruleus

Distribution (Figs 28 and 29)

Although the species may occur anywhere in the Province, it is not as evenly distributed as
shown by McLachlan & Liversidge (1978). The distribution for both the pre-1970 and
1970- 1979 periods accords with that given by Snow (1978). Known to occur in the N.W.
Cape; not considered during a 1977 field survey in this region.

Habitat

Primarily grassland but found in all habitats except forest. Partial to Pinus and Eucalyptus
spp. for roosting and breeding. Often found in man-made habitats, more especially croplands
and town commonages.

Fig. 28. Distribution of the Blackshouldered Kite, 1700 - 1969.

206

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS ( AVES: FALCONIFORMES)
Status

A nomadic species which moves into suitable environments and remains there for as long
as conditions are favourable. E. caeruleus is resident in optimal areas, however, such areas
represent only a small portion of its range in the Province. There has been no apparent change
in status. E. caeruleus is the common and most frequently recorded raptor in the Province.

Fig. 29. Distribution of the Blackshouidered Kite, 1970 - 1979.

207

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Cuckoo Hawk Aviceda cuculoides

Distribution (Figs 30 and 31)

The distribution for both the pre-1970s and the 1970s accords with that given by Snow
(1978), whereas the 1970s distribution is somewhat more restricted than that given by McLach-
lan & Liversidge (1978).

Habitat

Forest and Eucalyptus plantations.

Fig. 30. Distribution of the Cuckoo Hawk, 1700 - 1969.

208

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS ( AVES: FALCONIFORMES)
Status

Resident. Although there are several records of sub-adult birds there is no definite evi-
dence that A. cuculoides breeds in the Province. The species is, and always has been, rare and
is assigned this status by Siegfried et al. (1976). A. cuculoides may have decreased in numbers
in the Province as there are fewer records for the 1970s than for the earlier period; there has
been no apparent change in range in the Province between the two periods.

Fig. 31. Distribution of the Cuckoo Hawk, 1970 - 1979.

209

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Honey Buzzard Pernis apivorus

Distribution (Fig. 32)

There are no records from the 1970s and this situation accords basically with that given by
McLachlan & Liversidge (1978).

Habitat

No information for the Cape Province. Elsewhere in open woodland.

Fig. 32. Distribution of the Honey Buzzard, 1700 - 1969.

210

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A non-breeding Palaearctic migrant which is present during the austral summer. Extreme-
ly rare in the Cape Province; probably restricted to the E. Cape region. It is noteworthy that
six of the eight records for P. apivorus are from specimens. The present authors interpret this
to suggest that the species may not have undergone a decrease but simply that few ornithol-
ogists are able to identify the birds in the field. This contention is supported somewhat by the
fact that Cramp & Simmons (1980) consider that P. apivorus has increased in numbers in its
western Palaearctic range.

211

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Black Eagle Aquila verreauxii

Distribution (Figs 33 and 34)

The 1970s distribution accords with that given by McLachlan & Liversidge (1978) and
Snow (1978).

Habitat

Associated with mountains and gorges where there are cliffs for eyries and abundant Pro-
cavia capensis prey. In the forests of the E. Cape where such conditions occur the Black Eagle
is replaced by the Crowned Eagle Stephanoaetus coronatus. Precluded from the N. Cape
through lack of suitable habitat.

Fig. 33. Distribution of the Black Eagle, 1700 - 1969.

212

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

Resident. Despite persecution by small-stock farmers A. verreauxii shows no change in
status between the two periods. This species is the most numerous and widespread of the large
eagles in the Cape Province ( vide Boshoff & Vernon 1980b).

Fig. 34. Distribution of the Black Eagle, 1970 - 1979.

213

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983

Tawny Eagle Aquila rapax

Distribution (Figs 35 and 36)

1700-1969: This distribution accords with that given by Snow (1978).

1970-1979: This distribution is more restricted than that given by McLachlan & Liversidge
(1978).

Habitat

Bushveld, karoo and grassland. In habitats devoid of indigenous trees the species breeds
in exotic trees, e.g. Pinus spp., and on powerline pylons.

Fig. 35. Distribution of the Tawny Eagle, 1700 - 1969.

214

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS ( AVES: FALCONIFORMES)
Status

Resident but scarce; the species has decreased in range in the Province. A. rapax is perse
cuted by small-stock farmers and outside the Kalahari Gemsbok National Park the total popu
lation may be relatively low (vide Boshoff & Vernon 1980b).

Fig. 36. Distribution of the Tawny Eagle, 1970 - 1979.

215

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983

Steppe Eagle Aquila nipalensis

Distribution (Figs 37 and 38)

McLachlan & Liversidge (1978) do not present a distribution map and only list one re-
cord, from about 1940, from Tarkastad in the E. Cape (vide Brooke 1974).

Habitat

Bushveld and karoo.

Fig. 37. Distribution of the Steppe Eagle, 1700 - 1969.

216

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A non-breeding Palaearctic migrant, which may occur in relatively large numbers in the
northern parts of the N. Cape during the austral summer. To the south of this area it is a rare
vagrant. The Cape Province is on the periphery of the range of this species. There has been no
apparent change in status in the Province between the two periods.

Fig. 38. Distribution of the Steppe Eagle, 1970 - 1979.

217

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Wahlberg’s Eagle Aquila wahlbergi

Distribution (Figs 39 and 40)

The 1970s distribution is similar to that given by Snow (1978) but the S. Cape record
mapped by Snow (op cit.) is not accepted by the present authors due to lack of substantiating
evidence and atypical habitat. The distribution map given by McLachlan & Liversidge (1978)
virtually excludes A. wahlbergi from the Cape Province.

Habitat

Bushveld in the Cape Province. Elsewhere in Africa A. wahlbergi is a woodland species.
Where it occurs in savannas it utilises large trees along water courses.

Fig. 39. Distribution of the Wahlberg’s Eagle, 1700 - 1969.

218

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

The status of A. wahlbergi in the Province is equivocal and at best can be considered a
rare vagrant. It is a breeding summer migrant in southern Africa and the N. Cape is at the
southern periphery of its range. There are 12 records of this species in the Province, most of
them from the Kalahari Gemsbok National Park. However, none of these are substantiated
and there is a possibility that misidentifications may have occurred.

Fig. 40. Distribution of the Wahlberg's Eagle, 1970 - 1979.

219

ANN. CAPEPROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Booted Eagle Hieraaetus pennatus

Distribution (Figs 41 and 42)

The maps given for the distribution of H. pennatus by both McLachlan & Liversidge (1978)
and Snow (1978) are inadequate.

Habitat

Karoo, macchia and grassland, especially at their ecotones. Specifically the species is
found in kloofs and gorges where it breeds on cliffs, however, birds do hunt in open country
well away from such sites.

Fig. 41. Distribution of the Booted Eagle, 1700 - 1969.

220

BOSHOFF ET AL : HISTORICAL ATLAS OF DIURNAL RAPTORS ( AVES: FALCONIFORMES)
Status

H. pennatus is a breeding summer migrant with nesting confined to the karoo — macchia
and karoo — grassland ecotone. Non-breeding Palaearctic migrants supplement the southern
African population (Brooke et al. 1980). Birds breeding in the southern parts of the Province
move to areas north of 29° S during winter. The species was long thought to be a rare migrant
and is listed as “rare and vulnerable” by Siegfried et al. (1976). However, recent studies have
shown H. pennatus to be a common eagle in the Province (Brooke et al. op cit.). There is con-
troversy as to whether the species has increased in recent years (Benson 1979), or has always
been present but was overlooked and misidentified (Brooke et al. op cit.).

Fig. 42. Distribution of the Booted Eagle, 1970 - 1979.

221

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Hawk Eagle Hieraaetus fasciatus

Distribution (Figs 43 and 44)

The distribution maps presented by McLachlan & Liversidge (1978) and Snow (1978) are
inadequate as none of the published records for the S.W. Cape, S. Cape and E. Cape is ac-
cepted by the present authors (see “Status”).

Habitat

No information for the Cape Province but apparently bushveld. Elsewhere in Africa in
woodland associated with hilly terrain.

Fig. 43. Distribution of the Hawk Eagle, 1700 - 1969.

222

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

With only three records for the entire 1700 - 1979 period, H. fasciatus is an extremely rare
vagrant in the northern Cape Province, which is on the southern periphery of its range. In
South Africa the furthest records to the south are from Greytown in Natal (approx. 29° S) (Cy-
rus & Robson 1980). There is no apparent change in the status in the Province between the
two periods.

The record of a former specimen of H. fasciatus from the Wynberg district near Cape
Town (Layard 1867) is not accepted here due to lack of substantiating data and atypical habi-
tat. Similarly the published records of this species from the E. Cape and S. Cape (one) are not
accepted due to incorrect labelling of specimens (two) and unconfirmed sight records (three).

Fig. 44. Distribution of the Hawk Eagle, 1970 - 1979.

223

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Ayres’ Eagle Hieraaetus ayresii

Distribution (Fig. 45)

1700-1969: This distribution accords with that given by McLachlan & Liversidge (1978). The
S.W. Cape and N.W. Cape records given by Snow (1978) sub nom. H. dubius are
rejected (Brooke & Vernon 1981).

1970-1979: No records.

Habitat

Forest and thornveld in the Cape Province. Woodland in normal range of the species.

Fig. 45. Distribution of Ayres’ Eagle, 1700 - 1969.

224

BOSHOFFET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)

Status

A rare vagrant, well south of its normal range in Mocambique and Zimbabwe. Five of the
six records from the 1700 - 1969 period are of immature birds. The earlier assumption that this
species was resident in the Cape Province was based on misidentifications (Brooke & Vernon
op cit.). Status trend in the Province indeterminate but may have ceased to occur. Siegfried et
al. (1976) list H. ayresii as “rare and vulnerable” in South Africa.

225

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983

Longcrested Eagle Lophaetus occipitalis

%

Distribution (Figs 46 and 47)

1700-1969: The Karoo record is of a specimen from Nieu Bethesda early in the last century.

This distribution accords with that given by Snow (1978) except for one S.W. Cape
record mapped by that author.

1970-1979: The map in McLachlan & Liversidge (1978) fails to indicate any records for the
Cape Province.

Habitat

Grassland, especially marshy areas or on the edge of forests or plantations.

Fig. 46. Distribution of the Longcrested Eagle, 1700 - 1969.

226

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCON IFORMES)
Status

L. occipitalis has decreased in range in the Cape Province. At present its status is uncer-
tain and at best there is only a small resident population; there are no breeding records from
the Province, individuals sighted seldom remain long in the same area. The decrease is attribu-
ted to the destruction and misutilisation of suitable habitat. The Nieu Bethesda record is con-
sidered feasible as wooded kloofs were present in the area in the past. L. occipitalis is listed as
“rare and vulnerable” in South Africa by Siegfried et al. (1976).

Fig. 47. Distribution of the Longcrested Eagle, 1970 - 1979.

227

ANN. CAPE PROV. MUS. (NAT. HIST. ) VOL. 14 PT 7, FEBRUARY 1983
Martial Eagle Polemaetus bellicosus

Distribution (Figs 48 and 49)

The distribution for the pre- 1970s and 1970s accords with that given by McLachlan &
Liversidge (1978) and Snow (1978).

Habitat

May be found in all habitats but more regularly in open country; seldom in forested and
mountain areas.

Fig. 48. Distribution of the Martial Eagle, 1700 - 1969.

228

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A scarce resident which shows no change in overall range in the Province between the two
periods. P. bellicosus is severely persecuted by small-stock farmers (Boshoff & Vernon 1980b)
but, although a decrease in numbers is considered to have taken place, there is little reliable
published evidence to that effect. This species has adapted to man’s activities and breeds in
exotic trees, windmills and powerline pylons. P. bellicosus is rightly listed as “vulnerable” by
Siegfried et al. (1976).

Fig. 49. Distribution of the Martial Eagle, 1970 - 1979.

229

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Crowned Eagle Stephanoaetus coronatus

Distribution (Figs 50 and 51)

The pre-1970 and 1970s distribution of the species accords with that given by McLachlan
& Liversidge (1978) and Snow (1978) but does not extend into the Karoo pace McLachlan &
Liversidge (op. cit.). The early record (breeding) from the S.W. Cape is questionable.

Habitat

Forest and also thornveld, even where such habitat is restricted to kloofs and gorges in
otherwise open country.

Fig. 50. Distribution of the Crowned Eagle, 1700 - 1969.

230

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A relatively scarce resident species. S. coronatus is directly persecuted by small-stock
farmers but still has a viable population in the Province ( vide Boshoff & Vernon 1980b). There
has been no major change in range in the Province between the two periods but there may
have been a disappearance from the fringes of the Karoo where the species inhabited densely
wooded kloofs and ravines. The kloof vegetation was largely destroyed through misutilisation
by man.

Fig. 51. Distribution of the Crowned Eagle, 1970 - 1979.

231

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Brown Snake Eagle Circaetus cinereus

Distribution (Figs 52 and 53)

The distribution for the pre-1970s and 1970s accords with that given by McLachlan &
Liversidge (1978) and Snow (1978).

Habitat

No information is available from the Cape Province. Elsewhere in Africa it is a woodland
species.

Fig. 52. Distribution of the Brown Snake Eagie, 1700 - 1969.

232

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

C. cinereus is a rare vagrant to the Province and is well south of its normal range in Bot-
swana, Transvaal, Zululand and further north. There has been no change in range in the Prov-
ince between the two periods. A pre-1900 record from the Berg River is not accepted here due
to lack of supporting evidence and atypical habitat.

Fig. 53. Distribution of the Brown Snake Eagle, 1970 - 1979.

233

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Blackbreasted Snake Eagle Circaetus gallicus

Distribution (Figs 54 and 55)

The distribution for the pre-1970s and 1970s accords with that given by McLachlan &
Liversidge (1978) and Snow (1978).

Habitat

All habitats, except forest and mountain areas but most often in open country.

Fig. 54. Distribution of the Blackbreasted Snake Eagle, 1700 - 1969.

234

BGSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A resident species which is, and has always been, rare. Although there are five breeding
records from the Province, the majority of the birds recorded may be vagrants from Botswana,
Namibia and the Transvaal. C. gallicus does not exhibit any change in range in the Province
between the two periods.

Fig. 55. Distribution of the Blackbreasted Snake Eagle, 1970 - 1979.

235

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Bateleur Terathopius ecaudatus

Distribution (Figs 56 and 57)

1700-1969: This distribution accords to some extent with that given by Snow (1978), except
that two records from the Karoo are given by this author.

1970-1979: This distribution is far more restricted than that given by McLachlan & Liversidge
(1978).

Habitat

Bushveld in the N. Cape and open country in the E. Cape. Elsewhere in its range in Afri-
ca the species is found in woodland and savanna.

Fig. 56. Distribution of the Bateleur. 1700 - 1969.

236

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

There is no satisfactory evidence that T. ecaudatus ever bred in the Province and so may
always have been a rare vagrant. A record of a nestling from Nieu Bethesda in the early part
of the last century is not accepted by the present authors. Reasons for this are given as atypical
breeding habitat and the fact that it is the only breeding record of T. ecaudatus from the entire
Province. These aspects raise suspicions of incorrect labelling or a transported captive bird.
The species may be resident in the Kalahari Gemsbok National Park where regular sightings
are made. Recent sight records from the S. Cape and S.W. Cape require confirmation due to
possible confusion with Buteo rufofuscus. T. ecaudatus exhibits a decreased range in the Prov-
ince between the two periods. The species is actively persecuted by small-stock farmers in the
N. Cape. The species should be included in a revised South African Red Data Book: Aves.

Fig. 57. Distribution of the Bateleur. 1970 - 1979.

237

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983

Fish Eagle Haliaeetus vocifer

Distribution (Figs 58 and 59)

The distribution for the pre- 1970s and 1970s accords with that given by Snow (1978) but is
not as widespread as that given by McLachlan & Liversidge (1978).

Habitat

Perennial rivers, lakes, dams and estuaries.

Fig. 58. Distribution of the Fish Eagle, 1700 - 1969.

238

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A common resident species which shows no change in status in the Province between the
two periods ( vide Boshoff & Vernon 1980b). Although listed as “vulnerable” in South Africa
by Siegfried et. al. (1976), due to loss of natural habitats, this has been largely balanced by new
habitats created by means of artificial water impoundments.

Fig. 59. Distribution of the Fish Eagle, 1970 - 1979.

239

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Steppe Buzzard Buteo buteo

Distribution (Figs 60 and 61)

The distribution for the 1970s is more restricted than that given by McLachlan & Liver-
sidge (1978).

Habitat

Grassland, macchia and to some extent thornveld; particularly croplands and road verges.

Fig. 60. Distribution of the Steppe Buzzard, 1700 - 1969.

240

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A common non-breeding Palaearctic migrant which is present during the austral summer.
The species exhibits an unchanged overall status in the Province between the two periods.
B. buteo may, however, have adjusted its distribution in the Province in accordance with
changes in farming activities.

Fig. 61. Distribution of the Steppe Buzzard. 1970 - 1979.

241

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Mountain Buzzard Buteo tachardus

Distribution (Figs 62 and 63)

The distribution for the pre- 1970s and 1970s accords with that given by Snow (1978) but is
more extensive (into the S.W. Cape) than that given by McLachlan & Liversidge (1978).

Habitat

Forest and plantations of exotic species, e.g. Pinus spp. Despite its common name
B. tachardus is rarely associated with mountains and, where this does occur, the birds are in or
near densely wooded ravines in the footslopes.

Fig. 62. Distribution of the Mountain Buzzard, 1700 - 1969.

242

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A relatively scarce resident species which shows no change in range in the Province be-
tween the two periods. In fact the species may have been overlooked in the past due to mis-
identification and so acquired rarity status. Although listed as “vulnerable” by Siegfried et al.
(1976) B. tachardus appears to have a viable, albeit small, population in the Province. The
species has adapted to plantations.

Fig. 63. Distribution of the Mountain Buzzard, 1970 - 1979.

243

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Jackal Buzzard Buteo rufofuscus

Distribution (Figs 64 and 65)

1700-1969: This distribution accords with that given by Snow (1978).

1970-1979: This distribution is more restricted than that given by McLachlan & Liversidge
(1978).

Habitat

Found in all habitats, except bushveld, where there are cliffs and gorges. Appears to pre-
fer open habitats e.g. macchia, grassland and karoo.

Fig. 64. Distribution of the Jackal Buzzard, 1700 - 1969.

244

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A common resident which is the third most frequently recorded raptor in the Province.
There has been no apparant change in status in the Province between the two periods.

Fig. 65. Distribution of the Jackal Buzzard, 1970 - 1979.

245

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Redbreasted Sparrowhawk Accipiter rufiventris

Distribution (Figs 66 and 67)

The distribution for the pre-1970s and 1970s accords with that given by McLachlan &
Liversidge (1978) and Snow (1978).

Habitat

Forests, plantations and groves of exotic trees in grasslands. In the E. Cape and S. Cape
the species is absent from coastal forests.

Fig. 66. Distribution of the Redbreasted Sparrowhawk, 1700 - 1969.

246

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A relatively scarce resident species, A. rufiventris extended its range into the grassland
(treeless areas) of the E. Cape as exotic trees were planted and is now found around home-
steads in country where it would not previously have occurred. This range increase is not ob-
vious from a comparison of the maps for the two periods.

Fig. 67. Distribution of the Redbreasted Sparrowhawk, 1970 - 1979.

247

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Ovambo Sparrowhawk Accipter ovampensis

Distribution (Figs 68 and 69)

McLachlan & Liversidge (1978) do not record this species from the Cape Province.

Habitat

Bushveld; usually Acacia trees along water courses.

Fig. 68. Distribution of the Ovambo Sparrowhawk, 1700 - 1969.

248

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

An extremely rare vagrant to the northern part of the N. Cape; the two records for the
Province are far south of the normal range in northern Namibia, northern Botswana, northern
Transvaal and Zimbabwe. Siegfried et al. (1976) consider that A. ovampensis might be includ-
ed in a revised South African Red Data Book: Aves.

Fig. 69. Distribution of the Ovambo Sparrowhawk, 1970 - 1979.

249

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Little Sparrowhawk Accipiter minullus

Distribution (Figs 70 and 71)

1700-1969: This distribution accords' with that given by Snow (1978).

1970-1979: This distribution accords with that given by McLachlan & Liversidge (1978) except
that the species does not extend into the Karoo as shown by these authors.

Habitat

Forest and occasionally groves of exotic trees.

Fig. 70. Distribution of the Little Sparrowhawk. 1700 - 1969.

250

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AYES: FALCONIFORMES)

Status

A scarce resident which shows no change in status in the Province between the two
periods.

Fig. 71. Distribution of the Little Sparrowhawk, 1970 - 1979.

251

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Black Sparrowhawk Accipiter melanoleucus

Distribution (Figs 72 and 73)

1700-1969: This distribution coincides with that given by Snow (1978).

1970-1979: The distributions map given by McLachlan & Liversidge (1978) does not record
the species in the S.W. Cape and includes the southern part of the Karoo in the
range.

Habitat

Forest and plantations.

Fig. 72. Distribution of the Black Sparrowhawk, 1700 - 1969.

252

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A scarce resident which because it is unobtrusive may be commoner than thought. There
has been no major change in the status of A. melanoleucus in the Province between the two
periods, except an apparent disappearance from the Karoo. This is no doubt linked to the de-
struction of wooded kloof habitat. The species is recorded as “threatened at least in part of its
range” in South Africa by Siegfried et al. (1976).

Fig. 73. Distribution of the Black Sparrowhawk, 1970 - 1979.

253

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Little Banded Goshawk Accipiter badius

Distribution (Figs 74 and 75)

The distribution for the pre- 1970s and 1970s is more restricted than that given by both Mc-
Lachlan & Liversidge (1978) and Snow (1978). There are reported sightings from the E. Cape
(Skead 1967) and Courtenay-Latimer (1964) lists the species as “resident” in the East London
area. There are two reported sightings from the E. Cape from the 1970s. However, no speci-
mens of A. badius have ever been taken from the E. Cape and the occurrence of the species
there needs confirmation.

Habitat

No information from the Cape Province but apparently bushveld. Elsewhere in Africa in
savanna woodland .

Fig. 74. Distribution of the Little Banded Goshawk, 1700 - 1969.

254

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A rare vagrant in the Province; the species normally occurs in Zululand, Transvaal, Nami-
bia and northwards. There has been no apparent change in status in the Province between the
two periods.

Fig. 75. Distribution of the Little Banded Goshawk, 1970 - 1979.

255

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
African Goshawk Accipiter tachiro

Distribution (Figs 76 and 77)

The distribution for the pre-1970s and 1970s accords with that given by McLachlan &
Liversidge (1978) and Snow (1978).

Habitat

Forest and wooded kloofs.

Fig. 76. Distribution of the African Goshawk, 1700 - 1969.

256

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A. tachiro , a resident species, is the commonest accipiter in the E. Cape and S. Cape.
However, this species has decreased in the southern and eastern Karoo due to destruction and
misutilisation of suitable habitat by man.

Fig. 77. Distribution ol the African Goshawk, 1970 - 1979.

257

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7. FEBRUARY 1983
Gabar Goshawk Micronisus gabar

Distribution (Figs 78 and 79)

The pre-1970s and 1970s distribution differs from that given by McLachian & Liversidge
(1978) and Snow (1978) in that the species is recorded from the S.W. Cape, albeit only one re-
cord from the early period.

Habitat

Bushveld and thornveld, more especially Acacia karroo trees along watercourses. The
species has adapted to thornveld in overgrazed areas and to trees around homesteads in arid
areas.

Fig. 78. Distribution of the Gabar Goshawk, 1700- 1969.

258

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A scarce resident in most of the Province. M. gabar is, however, more numerous in the
Kalahari Gemsbok National Park. There has been no apparent change in status in the Province
between the two periods and the lack of records from the Karoo in the pre- 1970s may be due
to inadequate sampling.

Fig. 79. Distribution of the Gabar Goshawk, 1970- 1979.

259

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983

Pale Chanting Goshawk Melierax canorus

Distribution (Figs 80 and 81)

The distribution for the pre-1970s and 1970s accords with that given by McLachlan &
Liversidge (1978) and Snow (1978). Known to occur over most of the N.W. Cape; not consid-
ered during a 1977 field survey in this region.

Habitat

Karoo, bushveld, and less frequently macchia, thornveld and grassland.

Fig. 80. Distribution of the Pale Chanting Goshawk, 1700 - 1969.

260

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS ( AVES: FALCONIFORMES)
Status

M. canorus is a common resident species in the karoo and bushveld but is subject to local
movements and so may not be permanently resident in any locality and may move temporarily
into the peripheral macchia, thornveld and grassland. The species shows no change in status in
the Province between the two periods.

Fig. 81. Distribution of the Pale Chanting Goshawk, 1970 - 1979.

261

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983

African Marsh Harrier Circus ranivorus

Distribution (Figs 82 and 83)

1700-1969: This distribution is similar to that given by Snow (1978) except that this author
also records the species in the N. Cape.

1970-1979: The increase in records from the N. Cape is due to improved mapping coverage.

This distribution is more restricted than that given by McLachlan & Liversidge
(1978).

Habitat

Marshy areas along rivers, around dams, lakes and estuaries and vleis.

Fig. 82. Distribution of the African Marsh Harrier, 1700 - 1969.

262

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A scarce resident which shows no overall change in range in the Province between the two
periods, although there may be some disappearance from the central Karoo. The large number
of records from the N. Cape from the 1970s is due to increased sampling in that region. Much
habitat has been lost, e.g. in the Karoo, due to modification for agricultural purposes and the
overall population has decreased in size. As C. ranivorus is vulnerable to extirpation in the
Province it should be included in a revised South African Red Data Book: Aves.

Fig. 83. Distribution of the African Marsh Harrier, 1970 - 1979.

263

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 P^ 7, FEBRUARY 1983
Montagu’s Harrier Circus pygargus

Distribution (Figs 84 and 85)

The 1970s distribution is more restricted than that given by McLachlan & Liversidge
(1978).

Habitat

No information for the Cape Province but elsewhere in grassland and open country.

Fig. 84. Distribution of the Montagu’s Harrier, 1700 - 1969.

264

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A rare non-breeding Palaearctic migrant which is present during the austral summer The
decrease in records and range in the Province may reflect a decrease in the breeding grounds
of this species in the Palaearctic (Cramp & Simmons 1980).

Fig. 85. Distribution of the Montagu’s Harrier, 1970 - 1979.

265

ANN. CAPE PROV. MUS. (NAT HIST.) VOL. 14 PT 7, FEBRUARY 1983

Pallid Harrier Circus macroiirus

Distribution (Figs 86 and 87)

Apart from the S.W. Cape record, the 1970s distribution is similar to that given by Mc-
Lachlan & Liversidge (1978).

Habitat

No information for the Cape Province but elsewhere in grassland and open country.

Fig. 86. Distribution of the Pallid Harrier, 1700 - 1969.

266

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A non-breeding Palaearctic migrant which is present during the austral summer. The
species is, and always has been, rare in the Province. There has been no change in range in the
Province between the two periods but infrequently recorded in the 1970s, particularly from the
Karoo. This is surprising in view of the population increase in the western Palaearctic (Cramp
& Simmons 1980).

Fig. 87. Distribution of the Pallid Harrier, 1970 - 1979.

267

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983

Black Harrier Circus maurus

Distribution (Figs 88 and 89)

1700-1969: This distribution accords with that given by Snow (1978).

1970-1979: This distribution is more restricted than that given by McLachlan & Liversidge
(1978).

Habitat

Grassland, karoo and macchia. In open country but not necessarily dose to water. The
species has adapted to cultivated lands.

Fig. 88. Distribution of the Black Harrier, 1700 - 1969.

268

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

C. maurus is a scarce resident species in the Cape Province. Breeding takes place primar-
ily in late winter and early spring in the Province south of 31° S. The birds undergo a norther-
ly migration during the non-breeding season (mainly June to July) to the northern parts of the
Province and adjacent provinces and territories (Van der Merwe 1981). Due to various pub-
lished statements there is controversy regarding the past and present status of C. maurus ; the
species was considered by some to have decreased to near extinction but subsequently numbers
were said to have increased again. Van der Merwe (op cit.) has assessed the situation and con-
tends that, on available evidence, the population in South Africa has remained largely un-
changed in recent times. C. maurus does not warrant inclusion in a revised South African Red
Data Book: Aves; the species is included in a supplementary list by Siegfried et al. (1976).

Fig. 89. Distribution of the Black Harrier, 1970 - 1979.

269

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Gymnogene Polyboroides typus

Distribution (Figs 90 and 91)

1700-1969: This distribution largely accords with that given by Snow (1978).

1970-1979: This distribution is somewhat more restricted than that given by McLachlan &
Liversidge (1978).

Habitat

Wooded kloofs and gorges.

Fig. 90. Distribution of the Gymnogene, 1700 - 1969.

270

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A relatively scarce resident. There has been no change in status in the Province between
the two periods.

Fig. 91. Distribution of the Gymnogene, 1970 - 1979.

271

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983

Osprey Pandion haliaetus

Distribution (Figs 92 and 93)

The 1970s distribution is more restricted than that given by McLachlan & Liversidge
(1978). The single record given by Snow (1978) is a breeding record from the S.W. Cape which
is now discounted.

Habitat

Mainly estuaries and coastal lakes but also large dams in the interior.

272

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A rare non-breeding Palaearctic migrant which occurs mainly during the austral summer;
a few birds are recorded during the winter (Boshoff & Palmer in prep). There has been no
change in status in the Province between the two periods.

Fig. 93. Distribution of the Osprey, 1970 - 1979.

273

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Peregrine Falco peregrinus

Distribution (Figs 94 and 95)

The distribution given by McLachlan & Liversidge (1978) is too generalised, whereas the
range given by Snow (1978) is limited mainly to the S.W. Cape and E. Cape.

Habitat

Mountainous and broken terrain where there are suitable cliffs for roosting and breeding.
The distribution maps suggest that the species may prefer the coastal areas and adjacent in-
terior.

Fig. 94. Distribution of the Peregrine, 1700 - 1969.

274

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS ( AVES: FALCONIFORMES)
Status

Two races of F. peregrinus occur in the Cape Province, viz. F.p. minor , which is resident,
and F.p. calidus , which is a non-breeding Palaearctic migrant. No distinction between the races
is made here. They overlap in range as the migrant race has been collected as far south as Port
Elizabeth. It is possible that the Karoo and N. Cape records refer to F.p. calidus and that the
resident race is confined to the coastal areas. F. peregrinus is rare in the Cape Province and ap-
parently has always been rare; there has been no apparent change in status in the Province be-
tween the two periods. The species is listed as “possibly threatened” by Siegfried et al. (1976).

Fig. 95. Distribution of the Peregrine, 1970 - 1979.

275

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Lanner Falco biarmicus

Distribution (Figs 96 and 97)

The distribution for the pre-1970s and 1970s accords with that given by Snow (1978) and
McLachlan & Liversidge (1978).

Habitat

Usually associated with cliffs suitable for breeding but may nest in trees. Birds range wide-
ly and can be found in all habitats except forest.

276

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A relatively scarce resident species. There has been no apparent change in status in the
Province between the two periods.

277

ANN. CAPE PROV. MUS. (NAT HIST. ) VOL. 14 PT 7, FEBRUARY 1983
Hobby Falco subbuteo

Distribution (Figs 98 and 99)

The 1970s distribution differs markedly from that given by McLachlan & Liversidge
(1978).

Habitat

No information for the Cape Province but apparently preferably thornveld and macchia.

278

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCON 1FORMES)
Status

A rare non-breeding Palaearctic migrant which is present during the austral summer.
There has been no apparent change in status in the Province between the two periods,
although not recorded from the interior of the Province during the 1970s.

279

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
African Hobby Falco cuvierii

Distribution (Figs 100 and 101)

1700-1969: This distribution accords with that given by Snow (1978).

1970-1979: This distribution is more restricted than that given by McLachlan & Liversidge
(1978).

Habitat

No information from the Cape Province but apparently thornveld.

Fig. 100. Distribution of the African Hobby, 1700 - 1969.

280

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

There are only five records for the entire 1700- 1979 period. F. cuvierii is considered a
rare vagrant in the Province; a breeding record from Idutywa (Transkei) in 1924 is considered
equivocal (W. R. J. Dean in litt. per A. C. Kemp). There has been no apparent change in
status in the Province between the two periods. F. cuvierii is listed as “rare and endangered” in
South Africa by Siegfried et al. (1976).

Fig. 101. Distribution of the African Hobby, 1970 - 1979.

281

ANN. CAPE PROV. MUS. (NAT. HIST.) VOT. 14 PT 7, FEBRUARY 1983
Rednecked Falcon Falco chicquera

Distribution (Figs 102 and 103)

The distribution for the pre- 1970s and 1970s largely accords with that given by McLachlan
& Liversidge (1978) and Snow (1978).

Habitat

Bushveld. Often associated with Borassus palms, e.g. in northern N.W. Cape.

Fig 102. Distribution of the Rednecked Falcon, 1700 - 1969.

282

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A rare species in the Cape Province, which is on the periphery of its range. Although
there are no breeding records from the Province F. chicquera may be resident as there is suit-
able habitat available. There has been no apparent change in status in the Province between
the two periods. The species is listed as “vulnerable” in South Africa by Siegfried et al. (1976).

Fig. 103. Distribution of the Rednecked Falcon, 1970 - 1979.

283

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Eastern Redfooted Kestrel Falco amurensis

Distribution (Figs 104 and 105)

The 1970s distribution is similar to that given by McLachlan & Liversidge (1978).

Habitat

No information from the Cape Province but apparently thornveld and possibly macchia.

Fig. 104. Distribution of the Eastern Redfooted Kestrel, 1700 - 1969.

284

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A rare non-breeding Palaearctic migrant which is present during the austral summer.
There has been no change in range in the Province between the two periods.

Fig. 105. Distribution of the Eastern Redfooted Kestrel, 1970 - 1979.

285

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983

Common Kestrel Falco tinnunculus

Distribution (Figs 106 and 107)

The distribution for the pre-1970s and 1970s accords with that given by McLachlan &
Liversidge (1978) and Snow (1978). Known to occur over most of the N.W. Cape; not consid-
ered during a 1977 field survey in that region.

Habitat

All habitats except forest. Breeds mainly on cliffs but is not confined to them and may be
found in flat open country.

Fig. 106. Distribution of the Common Kestrel, 1700 - 1969.

286

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A resident species which is, and appears to have always been, common i.e. there has been
no change in status in the Province between the two periods. F. tinnunculus may have ben-
efited from man’s activities as extensive use is made of telephone poles and powerline pylons
as hunting perches and nest sites. The species is subject to local movements so that birds may
be present in an area at one time and absent in another; these nomads appear to be the non-
breeding dispersing segment of the population.

Fig. 107. Distribution of the Common Kestrel, 1970 - 1979.

287

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Greater Kestrel Falco rupicoloides

Distribution (Figs 108 and 109)

1700-1969: This distribution accords with that given by Snow (1978).

1970-1979: The distribution map given by McLachlan & Liversidge (1978) excludes the S.W.
Cape from the species’ range but includes all other coastal areas.

Habitat

The arid and semi-arid interior, encompassing the karoo and bushveld and occasionally
the grassland, thornveld and macchia.

Fig. 108. Distribution of the Greater Kestrel, 1700 - 1969.

288

BOSHOFFET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A relatively common resident which is subject to nomadic movements so that it may be
temporarily absent from the interior and temporarily present in the peripheral areas. This no-
madism may be governed by climatic factors, specifically droughts, but the precise mechanisms
are not known. Thus the absence of F. rupicoloides from the moister coastal areas of the E.
Cape, S. Cape and S.W. Cape during the 1970s is probably a natural phenomenon. In the E.
Cape the species is considered a rare vagrant with only one breeding record (from the Bedford
district). There has been no apparent change in overall status in the Province between the two
periods. F. rupicoloides makes extensive use of telephone poles and powerline pylons for hunt-
ing perches and nest sites.

Fig. 109. Distribution of the Greater Kestrel, 1970 - 1979

289

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Lesser Kestrel Falco naumanni

Distribution (Figs 110 and 111)

The distribution for the 1970s is more restricted than that given by McLachlan & Liver-
sidge (1978).

Habitat

Grassland primarily and secondarily macchia, karoo and thornveld. The birds forage over
open country and roost communally in groves of trees; these groves are often exotic types such
as Eucalyptus spp. and Populus spp. which are planted around farmsteads and in towns.

Fig. 110. Distribution of the Lesser Kestrel, 1700 - 1969.

290

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A relatively common non-breeding Palaearctic migrant which is present during the austral
summer. The species may have increased its range in the eastern Karoo while undergoing a de-
crease in the southern regions. The increase may be due to the establishment of groves of
exotic trees in the Karoo.

Fig. 111. Distribution of the Lesser Kestrel, 1970 - 1979.

291

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983
Pygmy Falcon Polihierax semitorquatus

Distribution (Figs 112 and 113)

The distribution for the pre-1970s accords with that given by Snow (1978) but, as with that
for the 1970s, is more restricted than the distribution given by McLachlan & Liversidge (1978).

Habitat

Bushveld, especially where the camelthorn tree Acacia erioloba occurs; these trees are
used as nest trees by the Social Weaver Philetarius socius , a species to which the distribution of
Polihierax semitorquatus is closely linked since it usually nests in a chamber of a Social Weaver
nest.

Fig. 112. Distribution of the Pygmy Falcon, 1700 - 1969.

292

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)
Status

A resident species which is rare outside the Kalahari Gemsbok National Park. There is no
evidence of a change in range in the Province between the two periods. However, numbers
may have decreased due to a decrease in camelthorn trees, and consequently Social Weaver
nests, as a result of overfrequent veld burning by stock farmers. P. semitorquatus is listed as
“vulnerable” in South Africa by Siegfried et al. (1976).

Fig. 113. Distribution of the Pygmy Falcon, 1970 - 1979.

293

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983

DISCUSSION

The authors contend that the coverage obtained for the 1970s permits a relatively accurate
assessment of the distribution of the diurnal raptors of the Cape Province for that decade and
also permits a comparison with the pre-1970s period. Of the 55 taxa surveyed the majority
(72,6%) show no change in their range. However, almost a quarter (23,6% ) of the taxa ex-
hibit a decrease in range of varying degrees. In contrast only two taxa (3,6% ) show a small in-
crease. Of the 13 taxa showing a decrease eight are resident in the Province and another for-
merly bred in the Province.

Many of the distribution maps presented here modify to a greater or lesser extent the
maps given for certain taxa by McLachlan & Liversidge (1978) and Snow (1978). Particularly
in the case of the latter publication no distinction is made between old and modern records and
as a result the maps for certain taxa, e.g. Lappetfaced Vulture, Lammergeyer, Bateleur, are
totally inadequate.

It is perhaps relevant that all the taxa, shown by this survey to have decreased their range
in the Province, can be classed as “specialists” in that they occupy narrow ecological niches.
Such taxa are the first to disappear when habitats are degraded. For example five of the eight
taxa of vulture (including the Lammergeyer) recorded in the Province exhibit a decreased
range; the removal of the game herds and their replacement by well managed domestic stock
has largely contributed to this decrease. Similarly four of the eight true forest taxa have exhibi-
ted a decrease, having disappeared from the southern and/or eastern parts of the Karoo. The
taxa concerned, viz. Longcrested Eagle, Crowned Eagle, Black Sparrowhawk and African
Goshawk, used to inhabit thickly wooded kloofs and ravines in that area but man has now vir-
tually totally destroyed the habitat through injudicious burning and use of the trees for build-
ing and fencing material. On the other hand the “generalists” are able to persist, usually
through adaptation.

The data presented in this paper must be assessed in terms of the constraints imposed
upon them. The extraction of data from published sources actually presented less of a problem
than those from unpublished sources. This is due to the fact that, although often located in
obscure publications, the published data were at least available, whereas there is undoubtedly
much material still to be consulted but which is in the form of notes in personal, and thus un-
published, notebooks of many professional and non-professional ornithologists. It is hoped
that this paper will stimulate the release of additional unused data. Many valuable data, par-
ticularly from the pre-1900 period, were excluded due to the omission of such basic informa-
tion as date and locality. Furthermore, and this applies mainly to the early records, some of
the localities given were so vague as to render the record useless, e.g. “Cape of Good Hope”
or “Eastern Province”. It was not possible to substantiate all records on the file but for most
taxa the incorrect records are in the minority and they are masked by the majority of (correct)
records. There is little doubt that more records will come to light in the literature, particularly
from obscure sources. For the 1970s, and to a lesser extent the 1950s and 1960s, the authors
were able to use many unpublished records, whereas for the preceding period they had to rely
almost entirely on published material. The authors consider that 95% of the literature records
have been extracted.

There are many variables pertaining to the study of the distribution of raptor populations
and the type of data presented here does not lend itself to any form of statistical analysis. Con-
sequently the maps give the data a higher degree of mathematical uniformity than they actually
possess. The main factors precluding quantification of trends is the lack of any standardized
method of data collection. For example in both the pre- 1970s and 1970s, particularly the for-
mer period, there are many cases where only interesting or conspicuous taxa were recorded
from a locus. Further the “degree of effort” per locus was not controlled, i.e. the time spent
and the area covered within a locus shows large variation. Some loci were visited only once

294

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AVES: FALCONIFORMES)

whereas others were visited many times. Many of the records were from roadside counts and
the many biases involved in recording raptors from vehicles are well known. Certain taxa, e.g.
kestrels, kites and buzzards, are prone to the use of telephone poles or powerline pylons as
perches and are thus more conspicuous and therefore readily observed. Other taxa, by virtue
of their habitat requirements, are inconspicuous and are therefore rarely observed, e.g. forest
and mountain taxa.

Interpretation of the maps is complicated by various factors. An important aspect is that
of the number and distribution of observers throughout the entire period covered by the maps.
Probably only since 1940, the year in which the first edition of Birds of South Africa (Roberts
1940) was published, has there been a marked increase in interest in South African birds.
Through time there has also been an increase in the quantity (and quality) of observers. Fur-
thermore, some parts of the Cape Province, e.g. N. Cape and N.W. Cape, are even today very
isolated, with poor roads. Accordingly these areas have not been subjected to the same degree
of observation as in the case of areas close to or including the cities or more settled parts where
active birdwatching clubs may exist. Thus any gaps in distribution, in certain cases, could be at-
tributed to lack of observation rather than the absence of the birds. However, the maps and
graphs should be viewed only in terms of general temporal and spatial trends. For certain taxa
the two periods selected were adequate to indicate major range fluctuations. However, for
others several range fluctuations may have taken place during one or both periods. An ex-
ample of the latter case is provided by the Cape Vulture population which was widespread and
thriving at the time of the rinderpest epidemic in 1896 and the Anglo-Boer War at the turn of
the century. The population then decreased in range and numbers during the first part of the
present century. This was followed by an increase, particularly during the 1960s, and then an-
other decrease in the 1970s (Boshoff & Vernon 1980a). However, space and the general qua-
lity of the data precluded the presentation of one map per decade per taxon.

A relevant criticism of the maps presented here is the fact that no distinction is made be-
tween the types of record used, e.g. breeding records, records of regular occurrence and re-
cords of vagrants. Mundy (1980) discusses this aspect of distribution surveys in some detail. In
the present study it is purely a case of quality and quantity of data which precludes such a
breakdown. For virtually all taxa there are too few breeding records available to be able to
present a meaningful picture. Further, too little is known of nomadism and local migration to
separate the breeding and non-breeding ranges of resident taxa. For most taxa it is generally
accepted that the range given includes the breeding range, with the obvious exception of non-
breeding Palaearctic migrants. Extralimitals and stragglers can be expected on the periphery of
the range of a taxon. In addition there is a security risk attached to the pin-pointing of loci
where rare and endangered raptors breed.

A further problem in the interpretation of the survey data is that of the nomadism which
several raptor species exhibit, e.g. Blackshouldered Kite, Pale Chanting Goshawk. The
authors have little data to aid the understanding of this phenomenon but it appears to be re-
lated to changing conditions, which in turn are regulated by climatic events, e.g. an east-west
shift of populations according to droughts and floods in the E. Cape and Karoo. To further
complicate the issue the natural trend is masked by man’s modifications of various habitats.
Taxa undergoing local migrations, e.g. Black Harrier, Booted Eagle, present additional inter-
pretation problems.

The scope of this study and the type of data collected do not permit any statements on the
dynamic numerical status of the taxa listed. However, it must be emphasized that a static range
does not imply a static population in terms of abundance. The following resident taxa which
show no range change in the Cape Province are subjectively considered to have undergone a
decrease in numbers: Secretary Bird, Whitebacked Vulture, Cuckoo Hawk, Blackbreasted
Snake Eagle, African Marsh Harrier, Pygmy Falcon.

295

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 7, FEBRUARY 1983

The conservation value of this survey is seen mainly as the provision of a set of data which
can serve as a baseline for further monitoring of diurnal raptor populations, on a temporal and
spatial basis, in the future. In addition the survey has indicated which taxa are most in need of
urgent conservation attention, viz. the vultures, the large eagles and the forest species.

Button & Clancey (1972), Brooke (1978) and Liversidge et al. (1981) have published re-
ports of extremely rare taxa from the Cape Province and Transkei. However, as many of the
taxa reported need confirmation and none can be considered as part of the Cape Province rap-
tor avifauna they are not listed by this survey. Similarly there is only one acceptable record for
the Western Redfooted Kestrel, Falco vespertinus , that of a specimen from the Beaufort West
district collected in 1875. F. vespertinus is a non-breeding Palaearctic migrant which normally
occurs in northern and central Namibia and the single record referred to is probably of a rare
vagrant.

ACKNOWLEDGEMENTS

The Directors, and where applicable the resident ornithologists, of the following museums
are thanked for allowing access to their collections and for help in other ways: Albany Mu-
seum, Grahamstown; Alexander McGregor Museum, Kimberley; Durban Museum; East Lon-
don Museum; National Museum, Bloemfontein; Pietermaritzburg Museum; Port Elizabeth
Museum; South African Museum, Cape Town; Transvaal Museum, Pretoria.

The authors are obliged to the South African Ornithological Society for access to nest re-
cord cards housed at the Percy FitzPatrick Institute for African Ornithology at the University
of Cape Town and to P. Morant, of SAFRING at the time of the study, for making various
data available.

The project would not have been possible without the cooperation of the many profes-
sional and non-professional ornithologists who provided unpublished records. To all of them
the authors extend their grateful thanks, particularly to: P. Ashton, G. Bligh, M. H. Currie, J.
H. Grobler, J. Kieser, P. S. Lockhart, M. Lorraine-Grews, I. A. W. and S. Macdonald, R., J.
and E. Martin, P. J. Mundy, P. and M. Neatherway, N. G. Palmer, P. Steyn, C. T. Stuart, L.
and G. Underhill, C. J. Uys.

Messrs M. H. Currie and N. G. Palmer are thanked for assisting with the literature search
and the coding of records, as are Mrs H. Glover and Messrs P. Green and A. Bastiaans of the
Cape Provincial Administration Computer Centre for assistance with databanking and map
production.

During the course of this study R.K.B. was employed under the South African National
Programme for Environmental Sciences (C.S.I.R., Pretoria).

REFERENCES

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

Benson, C. W. 1979. Review of Snow, D. W. (ed.): An atlas of speciation in African non-passerine birds. Ibis 121:
529 - 531.

Boshoff, A. F. and Vernon, C. J. 1980a. The past and present distribution and status of the Cape Vulture in the Cape
Province. Ostrich 51: 230 - 250.

Boshoff, A. F. and Vernon, C. J. 1980b. The distribution and status of some eagles in the Cape Province. Ann. Cape
Prov. Mus. (nat. Hist.) 13 (9): 107 - 132.

Boshoff, A. F., Brooke, R. K. and Crowe, T. M. 1978. A computerized distribution mapping scheme for vertebrates
in southern Africa, illustrated by a range decrease in the Bearded Vulture Gypaetus barbatus (Linn.). S. Afr.
J. Wildl. Res. 8: 145 - 149.

Broekhuysen, G. J., Broekhuysen, M. R., Martin, J., Martin, E., Martin, R. and Morgan, H. K. 1968. Obser-
vations on the birdlife of the Kalahari Gemsbok National Park. Koedoe 11: 145 - 160.

Brooke. R. K. 1974. The African southern limits of the Steppe Eagle in winter. Bull. Brit. Orn. Club. 94: 62.

296

BOSHOFF ET AL.: HISTORICAL ATLAS OF DIURNAL RAPTORS (AYES: FALCONIFORMES)

Brooke, R. K. 1978. Sooty Falcon in Transkei. The Bee-eater 29: 1

Brooke, R. K. 1979. Predation on Ostrich eggs by tool-using crows and Egyptian Vultures. Ostrich 50: 257 - 258
Brooke, R. K. and Vernon, C. J. 1981. Early names and records of two small Hieraaetus eagles (Aves: Accipitridae) in
the Cape Province of South Africa. Ann. Cape Prov. Mus. (nat. Hist.) 13 (10): 133 - 139.

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.

Brown, L. H. 1977 The status, population structure and breeding dates of the African Lammergeyer Gypaetus barba-
tus meridionalis. Raptor Research 11: 49 - 58.

Button, E. L. and Clancey, P. A. 1972. Some interesting records of wintering Palaearctic raptors from the Northern
Cape. Durb. Mus. Novit. 9: 201 - 207.

Clancey, P. A. (ed.). 1980. S.A.O.S. Checklist of southern African birds. Johannesburg. Southern African Ornithologi-
cal Society.

Courtenay-Latimer, M. 1964. Check list of the birds of the East London area. S. Afr. Avif. Series. No. 20.

Cramp, S. and Simmons, K. E. L. 1980. The birds of the Western Palearctic. II. Oxford. Oxford University Press.

Cyrus, D. and Robson, N. 1980. Bird atlas of Natal. Pietermaritzburg. University of Natal Press.

Layard, E. L. 1867. The birds of South Africa. Cape Town. Juta.

Liversidge, R., Richardson, P. and Gubb, A. 1981. The Sooty Falcon Falco concolor in the southern Kalahari. Bull.
Brit. Orn. Club. 101: 268 - 270.

McLachlan, G. R. and Liversidge. R. 1978. Roberts Birds of South Africa. Cape Town. Trustees of the John Voelcker
Bird Book Fund.

Mills, M. G. L. 1976. A revised checklist of birds in Kalahari Gemsbok National Park. Koedoe 19: 49 - 62.

Mundy, P. J. 1978. The Egyptian Vulture ( Neophron percnopterus ) in southern Africa. Biol. Cons. 14: 307 - 315.
Mundy, P. J. 1980. The comparative biology of south African vultures. Unpub. Ph.D. thesis, University of Zimbabwe.
Salisbury.

Roberts, A. 1940 The Birds of South Africa. London. Witherby.

Siegfried. W. R.. Frost, P. G. H., Cooper. J. and Kemp, A. 1976. South African Red Data Book: Aves. (South African
National Scientific Programmes, Report No. 7.) Pretoria. CSIP.

Skead, C. J. 1967. Ecology of birds in the Eastern Cape Province. Ostrich Suppl. No. 7.

Snow, D. W. (ed.) 1978. An atlas of speciation in African non-Passerine birds. London Brit. Mus. (Nat. Hist.).

Van der Merwe, F. 1981. Review of the status and biology of the Black Harrier. Ostrich 52: 193 - 207.

Vernon, C. J. 1978. A review of the status of raptors in the Eastern Cape. In A symposium on African predatory birds.
Pretoria. Northern Transvaal Ornithological Society, pp. 87 - 95.

297

.3

i

.ANNALS

OF THE

CAPE PROVINCIAL

MUSEUMS

NATURAL HISTORY

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

24th JUNE 1983

PUBLISHED JOINTLY BY THE

CAPE PROVINCIAL MUSEUMS AT THE ALBANY MUSEUM, GRAHAMSTOWN

SOUTH AFRICA

Printed by CTP Book Printers, Cape

BD1289

On the Hydropsychidae (Trichoptera) of Southern Africa with keys to
African genera of images, larvae and pupae and species lists

by

K. M. F. SCOTT

Albany Museum, Grahamstown, South Africa
CONTENTS

Page

Abstract 300

Introduction 300

Scales and abbreviations used 302

Keys to genera of Afrotropical Hydropsychidae

(a) Imagos 303

(b) Larvae 305

(c) Pupae 308

Descriptions, discussion, biology and species lists

Subfamily Hydropsychinae 311

Hydropsyche 311

Hydropsyche longifurca Kimmins 311

African species of Hydropsyche 321

Cheumatopsyche 324

Cheumatopsyche thomasseti (Ulmer) 324

African species of Cheumatopsyche 336

Hydromanicus 341

Hydromanicus seychellensis Ulmer 341

African species of Hydromanicus 342

Subfamily Diplectroninae 342

Sciadorus 342

Sciadorus obtusus Barnard 343

African species of Sciadorus (endemic genus) 351

Diplectronella 351

Diplectronella medialis Marlier 352

African species of Diplectronella 358

Diplectrona 358

Diplectrona felix McLachlan 359

African species of Diplectrona 359

Subfamily Macronematinae 360

Tribe Macronematini

Macrostemum (replaces Macronema, see pp 360 & 368) (for change of name see

p. 368) 360

Macrostemum capense (Walker) 360

African species of Macrostemum 370

Amphipsyche 371

Amphipsyche scottae Kimmins 372

African species of Amphipsyche 380

299

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Protomacronema 381

Protomacronema pubescens Ulmer 382

African species of Protomacronema (endemic genus) 388

Leptonema 389

Leptonema natalense Mosely 389

African species of Leptonema 397

Tribe Polymorphanisini

Polymorphanisus 398

Polymorphanisus bipunctatus (Brauer) 399

African species of Polymorphanisus 406

Aethaloptera 407

Aethaloptera maxima Ulmer 408

African species of Aethaloptera 415

Histograms (Fig. 214) showing larval instars of Cheumatopsyche , Macrostemum,

Amphipsyche and Aethaloptera species 416

Acknowledgements 417

References 417

ABSTRACT

The different stages of one species of each of the genera of Hydropsychidae (Trichoptera)
found in Southern Africa (south of the Zambezi River), are described, figured and discussed as
far as available material allows, the main emphasis being on the immature stages. The genera
included are Hydropsyche , Cheumatopsyche, Sciadorus, Diplectronella, Macrostemum , Am-
phipsyche, Protomacronema , Leptonema, Polymorphanisus and Aethaloptera. Of these, Di-
plectronella, from Central Africa, has been included in case it should be found further south
and for comparative purposes. Notes on Hydromanicus, from the Seychelles, and Diplectrona,
from North Africa, have been included, and they have also been included in the keys. Illus-
trated keys to imagos, larvae and pupae of African genera are given (“Africa” here means the
Afrotropical Region, including Africa south of the Sahara and the Malagasy Subregion, to-
gether with North Africa). In the case of each genus a list of the species recorded from Africa
is appended, with references and distribution. Macrostemum natalense (Ulmer) is placed in the
synonomy of M. capense (Walker), and Sciadorus sinuatus Marlier transferred to the genus Di-
plectronella as D. sinuata (Marlier).

INTRODUCTION

This account of the Hydropsychidae of Southern Africa is based mainly on specimens in
the Albany Museum Collection of Freshwater Invertebrates, which derives largely from collec-
tions made by the staff of the National Institute for Water Research of the South African
Council for Scientific and Industrial Research, during hydrobiological surveys of various river
systems throughout South Africa. To these have been added several other collections, and sup-
plementary material has been made available to me by various museums and other workers in
the field (see Acknowledgements).

The aim in the present paper has been to describe and illustrate as fully as possible the im-
mature stages of the Southern African genera, including earlier larval instars wherever possible
but giving only a very brief description of the adults, together with keys and necessary dis-
cussion. References to the literature have been included, and generic diagnoses have been
quoted or adapted from it as they are seldom accessible to field workers, yet are invaluable for
checking identifications. In each genus the species recorded from Africa have been listed with

300

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

comments, noting the known stages, but it has not been possible to check the validity of doubt-
ful species.

The genera fall into three of the four formerly recognized subfamilies of the Hydropsychi-
dae, the Hydropsychinae, Diplectroninae and Macronematinae. The fourth is now usually
raised to family status, as the Arctopsychidae, and does not occur in Africa, apparently being
confined to the Northern Hemisphere.

The procedure followed is outlined below. The keys, to imagos, mature larvae and pupae,
are given first, each couplet being illustrated wherever necessary by reference to the text
figures. These keys include all genera recorded from Africa. The term ‘Africa’ is here used to
denote the Afrotropical Region (i.e. Africa south of the Sahara together with South-West Ara-
bia, which formed the old Ethiopian Region, together with the Malagasy Subregion), plus
North Africa. (See Crosskey & White, 1977, for the term ‘Afrotropical Region.) This seemed
desirable for completeness and for use by workers in other parts of Africa.

In point of fact, Hydromanicus Brauer, found in the Seychelles, Diplectrona Westwood, a
Palaearctic genus extending into North Africa, and Diplectronella Ulmer from Central Africa,
are probably the only valid genera included that do not occur in Southern Africa. Hydropsy-
chodes Ulmer has been placed in the synonymy of Cheumatopsyche Wallengren (Kimmins
1963), Plesiopsyche Navas and Symphitopsyche Ulmer in that of Hydropsyche Pictet (see Ross
& Unzicker 1977 and Schmid 1979); the genera Phanostoma Brauer and Amphipsyche
McLachlan have been combined, and Chloropsyche McLachlan placed in the synonymy of Ae-
thaloptera Brauer (Kimmins 1962a). Excepting in the cases of Protomacronema and Diplectro-
nella, Southern African specimens have been used in making the drawings.

The keys are followed by a descriptive section, forming the main body of the paper, in
which is given as complete a coverage of the developmental stages of each genus as is possible
from the material available, together with a brief account of the imagos, the latter mainly for
recognition purposes.

In writing these sections, and in building up the keys, reference was, of course, made to
the not inconsiderable literature on the subject, and the keys, particularly those to imagos and
pupae, owe much to the excellent papers produced by the late Dr Georg Ulmer (1951, 1957),
as well as to those of many other workers in the field. Following Ulmer’s practice, the couplets
have been expanded to enable further points to be checked and identifications made more cer-
tain, particularly in cases where parts of a specimen are damaged or missing. Much use has
also been made of the late Dr F. C. J. Fischer’s invaluable Trichopterorum Catalogus (volumes
IV, 1963, and XIII, 1972); his nomenclature has in most cases been followed.

A general treatment of African families of Trichoptera (Marlier’s “Genera des Tri-
chopteres de l’Afrique”) appeared in 1962, but because this admirable work covers the entire
Order, the descriptions are necessarily brief, not very fully illustrated, and include only final
instar larvae. It was therefore felt that a much fuller treatment of the family was needed for the
use of ecologists in Southern Africa, and for use by workers on any African material. The keys
were produced prior to the appearance of Marlier's, being based primarily on the older litera-
ture and on specimens in hand. They have, however, been checked against his and others and
tested in use for some time.

The genus Protomacronema , while mainly Central African, has also been recorded from
Zululand (Kwazulu), from the Zambezi River in Mozambique and Rhodesia (now Zimbabwe),
and from its upper reaches and tributaries in Angola. Mature larvae were first described by
Gibbs (1973), even then tentatively, from Ghanaian material, and the account in this paper is
based on specimens generously sent to me by him, and compared with his own description.
Statzner (in litt.) has confirmed this correlation. Marlier (1965/66) listed Protomacronema lar-
vae as occurring in Angola, basing his identification on their occurrence in the same area as

301

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Protomacronema imagos, together with their notable differences from known Hydropsychid
larvae. I borrowed those specimens for study from the Museo do Dundo, but it transpired
from Gibbs’ work that they were in fact Leptonema larvae, probably of L. normale Banks,
Leptonema evidently being a genus showing a wide variation in certain larval characters,
notably the procoxae, size of stridulatory files and shape of anal gills.

The genus Diplectronella , first described from Ceylon (Ulmer 1928), has only been record-
ed from Central Africa to date, but may yet be discovered in mountainous regions further
south. It has been described from specimens kindly made available by Dr Decelle of the
Musee Royale de l'Afrique Centrale at Tervuren, and compared with larvae determined for
Dr E. J. Young, collected in Kenya.

Comments on the material studied, particularly the larvae, are given under each genus,
only one species being described in each case. Where possible, however, several species have
been compared in an endeavour to ensure that characters used in the keys are of generic rather
than specific value. Further localities are given from records in the Collection catalogues to
supplement the known distribution of certain species. Except where otherwise specified, the
specimens studied are in the Albany Museum.

It should be noted that the keys which follow are not necessarily applicable in other parts
of the world.

Methods of study have been described by Barnard (1934), Ross (1944), Wiggins (1977)
and may others. Terms used in this paper are illustrated in the earlier figures (cf. Figs 1-6,
10-12, 17, 19, 24) and occasionally in later ones where required.

Most of the adult specimens studied have been in spirit; in these the setae, particularly of
wings, may have been lost, though their bases may still be distinguishable. Pinned (dry) speci-
mens are very fragile, and must be handled most carefully; they tend to lose antennae, legs or
wings, rather than setae. Drawings have been made either from spirit material under a stereo-
microscope (magnification from x8 to x216) or, in the case of genitalia, from material cleared
in KOH and drawn in clove oil.

In the case of larvae and pupae all were in spirit. Parts too small for convenient study
under the stereomicroscope were mounted in glycerine, gum chloral or Euparal for drawing
under a compound microscope (magnification used from Xl5 to x600). An Abbe Drawing
Apparatus was used to ensure correct proportions.

SCALES AND ABBREVIATIONS USED

SCALES

Scale lines represent 1,0 mm unless otherwise indicated.

ABBREVIATIONS
Wing notation:

C, costa; Sc, subcosta; RS, Radial sector; Ri, R2, Rs, R-t, Rs, branches of radius; Mi, M2, M3,
M4, branches of media, M; Cma, Cuib, Q12, branches of cubitus; 1A, 2A, 3A, anal veins; arc,
arculus; dc, discoidal cell; he, humeral crossvein; me, median cell; m-cu, crossvein between
media and cubitus; pt, pterostigma; r-m, crossvein between radius and media; sdc, subdiscoi-
dal cell; tc, thyridial cell; 1, 2, 3, 4, 5, apical forks; wf, wing fold; wh, wing-coupling hooks.
Note corneous points in fork 2 and thyridial cell.

Other abbreviations:

ac, anteclypeus; ava, anterior ventral apotome; ant, antenna; ats, apical tibial spurs; co, coxa;
ep, epimeron; fc, frontoclypeus; fca, frontoclypeal apotome; lp, labial palpus; mp, maxillary

302

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

palpus; pe, pre-episternum (1st pleural sclerite); p2, second pleural sclerite; psw, posterior se-
tose wart; pt, pterostigma; pts, pre-apical tibial spurs; sf, stridulatory file; st, stipes (plural =
stipites); ve, vertex.

KEYS TO GENERA OF AFROTROP1CAL HYDROPSYCHIDAE
(a) IMAGOS

This key in various forms has been in use for a long time. It owes much to keys and data
given by Ulmer (1951), also by Barnard (1934) and Ross (1944); many other authorities have
also been consulted, all have been included in the list of references.

1. in hind wing RS first forks at anastomosis (i.e. at level of main cross-veins, cf Figs 91
& 4), from a common base, forming a long subdiscoidal cell; Ri, generally with Sc,
joins RS distal to anastomosis (in Polymorphcmisus by a cross-vein only. Fig. 180);
discoidal cell open or absent, strongly curved macrotrichia on costal margin. In fore-
wing discoidal cell closed, open or absent. Antennae 2 to 4 times length of forewing
in c3, longer than forewing in 2, very slender, threadlike. Maxillary and labial palpi
present or absent. Vertex of head generally with anterior setose warts larger than
posterior ones (Figs 88, 116); smaller in Leptonema (Fig. 158) and 9 Protomacrone-

ma (Fig. 144) Subfamily MACRONEMATINAE Ulmer 3

— In hind wing RS forks before anastomosis (Figs 4, 46), subdiscoidal cell absent, Ri

and Sc do not joing RS apically; discoidal cell closed; strongly curved macrotrichia
absent. In forewing discoidal cell closed. Antennae not much longer than forewings,
often shorter, slender to fairly stout, not threadlike. Maxillary and labial palpi al-
ways present. Vertex with paired posterior warts larger than anterior pair (Figs 1,
20,43,66) 2

2. In hind wing Sc and Ri join apically before joining costa. 5th abdominal sternum

lacks paired filamentous appendages. Outer leg claws may be concealed by tuft of
black setae. African genera with tibial spurs 2.4.4 or 0.4.4. Antennae smooth. 2
genitalia with sternal plates of VIII partly or completely separated; mid-tibiae and
tarsi of 9 usually dilated. (See Figs 1-6 & 20-25)

Subfamily HYDROPSYCHINAE Curtis 8

— In hind wing Sc and Ri remain separate to margin, 5th abdominal sternum with a

pair of filamentous lateral appendages, each arising from an internal gland. Outer
leg claws not concealed by tuft of setae. African genera with tibial spurs 2.4.4. An-
tennae appear serratulate owing to setal arrangement and internal dilatation, 9 gen-
italia with sternal plates of VIII separated to base, mid-tibiae and tarsi of 2 not di-
lated. (See Figs 43-48) Subfamily D1PLECTRONINAE Ulmer 10

3. Palpi absent or rudimentary, medium-sized to very large species, wings usually

glassy, middle tibiae with 2 or 3 spurs Tribe POLYMORPHANISINI Lestage 4

— Palpi always present, usually well developed, medium-sized species, wings glassy or

hairy, middle tibiae with 4 spurs Tribe MACRONEMATINI Lestage 5

4. Tibial spurs 0.2.2 or 0.3.2. Forewing with discoidal cell present or absent, small false

discal cell enclosing the corneous spot, median cell very large, quadrangular. Medi-
um-sized (<3 forewing c. 10-18 mm, 2 smaller) (Figs 192-198)

Aethaloptera Brauer (Syn. Chloropsyche McLachlan)

— Tibial spurs 1.3.2, 1.3.3, or 2.3.3. Forewing with discoidal cell closed, median cell

subtriangular. Very large (6 forewing up to 25 mm). Some African species have a
pair of conspicuous dark spots on the mesoscutellum. (Figs (Figs 177-180)

Polymorphcmisus Walker

303

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

5. Forewing with discoidal cell open or absent; palpi large or small. Tibial spurs 0.4.4,

1.4.4, 1.4.3, 1.4.2, 0.4.3, 0.4.2, 0.3.2 or 0.2.2. Wings glassy or hairy, hind wings very

wide, triangular, much folded 6

— Forewing with discoidal cell closed; palpi well developed. Tibial spurs 0.4.4, 1.4.4 or

2.4.4. Wings usually hairy, hind wings normally shaped 7

6. Palpi present but poorly developed, forewing with false fork 5 between Cui & Cuz.

Tibial spurs 1.4.4, 1.4.3, 1.4.2, 0.4.3, 0.4.2, 0.3.2 or 0.2.2. No stalked processes on
5th abdominal sternum. Wings glassy, 3 forewing c. 11-18 mm, $ smaller. (Figs
116-121) Amphipsyche McLachlan

— Palpi well developed but slender, often inconspicuous in 3 , shorter in $ ; 3 fore-

wing with forks, 1, 2, 3, 4 (no false fork 5, Cui, Q12 & 1A all ending in an anastomo-
sis), 9 with forks 1, 2, 3, 4, 5. Tibial spurs 0.4.4. Pair of small knob-like processes,
stalked, on 5th abdominal sternum in 3 . Wings hairy, 3 forewing c. 10—13 mm, 9
smaller. (Figs 137-144) Protomacronema Ulmer

7. Vertex with anterior warts smaller than posterior. Hind wing with fork 5 longer than

fork 2. 2nd segment of maxillary palp longer than third. 5th sternum in both sexes
with pair of fenestrae. Spurs 2.4.4 or 1.4.4. Wings usually hairy, a few species with
clear wings. 3 forewing c. 10-17 mm. 9 usually larger. (Figs 158-163)

Leptonema Guerin-Meneville

— Vertex with anterior warts larger than posterior. Forewing with Sc joining Ri before

reaching wing margin; hind wing with fork 5 shorter than fork 2. 2nd segment of
maxillary palp shorter than 3rd. 5th sternum in 3 (not in 9) without fenestrae but
with a pair of lateral processes. Spurs 2.4.4 or 1.4.4. Wings usually glassy, some-
times hairy, often brightly patterned, 3 forewing c. 10-20 mm, 9 smaller. (Figs
88-93) Macrostemum Kolenati

8. Hind wing with fork 1 present (may be small), median cell closed or open; tibial spurs

2.4.4 9

— Hind wing usually with fork 1 absent (if present compare with couplet 9), median
cell open, veins M & Cui usually well separated, cross-vein large. Spurs 2.4.4, rarely
0.4.4. 3 copulatory organ without long forked ventral process. 9 with sternal plates
of VIII separated to base of segment; mid-tibiae and tarsi of 9 dilated. 3 forewing

6- 8 mm long, 9 slightly larger. (Figs 20-25)

Cheumatopsyche Wallengren (Syn. Hydropsychodes Ulmer)

9. Median cell in hind wing closed, veins M & Cui usually very close together, cross-

vein short. 3 copulatory organ with forked ventral process longer than the organ it-
self (Afrotropical species & certain N. American species). 9 with sternal plates of
VIII separated for only 2/3 distance to base of segment, mid-tibiae and tarsi strongly
widened. 3 forewing 9-11 mm, 9 slightly larger. (Figs 1-6)

Hydropsyche Pictet (Synonyms Plesiopsyche Navas & Symphitopsyche Ulmer)

— Median cell in hind wing open, veins M & Cm well separated, crossvein clear. 3

copulatory organ without long forked ventral process, in this species terminal seg-
ment of 3 claspers apically concave, blackened, with thick inner covering of black
setae. 3 mid-leg not widened. 3 forewing 7-8 mm, 9 8-10 mm. (See Ulmer 1910a:
47-49, figs 11-14 for wings & 3 genitalia)

Hydromanicus seychellensis Ulmer (Seychelles)

10. Fork 1 in hind wing present, obvious; hind wing with Sc & Ri sinuous, strongly bent
upwards to meet costa apically. Vertex of head with two pairs large warts. 3 with 2
pairs large, rounded reticulated internal organs in abdomen (number depends on
species, 2 pairs being most usual): processes on sternum of V very long. 3 forewing

7- 9 mm (in D. felix McLachlan,. the only known African species). (For figures see

304

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Mosely 1939, figs 411-415) Diplectrona Westwood (North Africa only)

— Fork 1 in hind wing absent (if present then minute) , other characters various 11

11. Eyes rather small, sparsely pubescent; hind wing with Sc & Ri not strongly bent up-
wards to meet costa, wing not deeply rounded posteriorly, fork 1 absent (occasional-
ly present but minute). Vertex of head with narrow anterolateral warts, posterior
warts very large. 8 abdomen with 2 pairs large, rounded reticulate internal organs,
processes on abdomen V (c?, $) short, not extending beyond segment V. 8 fore-
wing 5-9 mm. (Figs 43-48) Sciadorus Barnard

— Eyes fairly large, strongly pubescent; hind wing with Sc & Ri sinuous, bowed to-

wards costa basally & strongly bent up to meet costa apically; wing deeply rounded
posteriorly, fork 1 always absent. Vertex of head with sinuous beanshaped antero-
lateral warts, posterior warts fairly large (compare Figs 66 & 43). 8 abdomen with-
out such large internal organs, processes on abdomen V very long, reaching end of
abdomen in 8 , in $ to mid-VI or longer. 8 forewing 7-11. (Figs 66-72)

Diplectronella Ulmer

(b) LARVAE

All Flydropsychid larvae are campodeiform, having long, two-jointed anal prolegs project-
ing beyond and free from the membranous lobes of the tenth abdominal segment; all three
thoracic terga are sclerotised, and tufted tracheal gills are present on the ventral surface of tho-
rax and abdomen in all instars except the first and, in some Macronematine genera, the sec-
ond. In counting gills it must be borne in mind that, as the abdomen enlarges in later instars,
the double stems of certain gills may become quite widely separated. The larval head is oval to
oblong, with eyes (grouped ocelli) placed below several cuticular lenses; antennae minute, situ-
ated close to the bases of the large mandibles; maxillary palpi stout, sometimes very long, la-
bial palpi very small, not easily seen. The ventral side of the head usually bears stridulatory
files (except in the Polymorphanisini), the scrapers being situated on the forelegs. When pres-
ent these organs are presumably used by a larva as a warning to other larvae trespassing on its
“territory”, i.e. too close to its net (see Johnstone 1964), or actually invading it (Jansson &
Vuoristo 1979), when stridulation often accompanied by fighting between the larvae has been
observed. Abdominal terga I- VIII are without dorsal sclerites, though there may be one on IX;
setose sternal plates may be present on VIII & IX, or IX only or absent. The larvae lack trans-
portable cases but construct shelters and spin complex catching nets in which to trap food.
These structures differ according to genus, and sometimes within a genus. Most genera may
also sally forth at times to capture prey or, being very pugnacious, to repel invaders, or to es-
cape unfavourable conditions (see Sattler 1963, Kaiser 1965, Edington 1965, Wallace 1975 and
others).

The larvae live on and under rocks or stones or amongst roots of aquatic macrophytes in
running water or on wave-washed shores of large lakes. They often abound where food sup-
plies are plentiful and flow is rapid, as may be the case below impoundments. Biological notes
are given under each genus.

In this family there appear to be either five or six larval instars, the Macronematinae hav-
ing six, the other subfamilies five. The key which follows applies to mature larvae and cannot
be used for first or second instar larvae in which neither gills nor pre-episterna (fore trochantin
of Wiggins 1977) nor limb claws are as yet sufficiently developed. It must be used with caution
in the case of third instar larvae. Fourth and fifth instars (the latter where there are six) are
sufficiently like the mature larva for reasonably easy recognition, though the colour pattern
may not as yet be fully developed. Early instars can, however, sometimes be recognized from
characters mentioned in the text, particularly when collected together with later instars of the
same species.

AM— 2

305

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Synonymy, given in the key to imagos under the genera concerned, is not repeated in the

keys to larvae and pupae. Synonyms can, of course, also be found, with references, in the sec-
tions on the genera which follow the keys.

1. Pre-episternum at base of foreleg forked (all African genera in this group) (Fig. 17);

apex of submentum with deep median incision (Figs 12, 32); dorsal surface of head
without carina, ventral surface with stridulatory files (Figs 12, 32); anteclypeus not
sclerotised; left mandible with inner brush of setae (Fig. 14); branched tracheal gills
on abdominal segments II-VI in four rows, small conical dorsolateral gills (“Zipfel-
kiemen” of Ulmer) present ; setose ventral sclerites on VIII & IX

HYDROPSYCHINAE Curtis 3

— Pre-episternum at base of foreleg not forked but simple (Figs 61, 105a, 132); apex of

submentum entire (Figs 58, 131); dorsal surface of head with or without carina (Figs
100, 102, 57); stridulatory files present or absent, anteclypeus with or without scle-
rites; left mandible with or without inner brush; branched tracheal gills on ab-
dominal segments II-VI in two, four, six or eight rows, “Zipfelkiemen” present or
absent 2

2. Meso- and metanotal shields with transverse sutures along which they break at ec-

dysis (Fig. 59); head without carina; frontoclypeal apotome very broad, with deep
indentations opposite eyes, behind which it is strongly expanded (Fig. 57); anterior
and posterior ventral apotomes large (Figs 58, 81), stridulatory files present; left
mandible with inner brush; tracheal gills in four or six rows, “Zipfelkiemen” pres-
ent; setose ventral sclerites on segments VIII & IX DIPLECTRONINAE Ulmer 5

— Meso- and metanotal shields without transverse sutures, remaining whole at ecdysis

(Figs 111, 207); head with or without carina, frontoclypeal apotome not widely ex-
panded behind eyes (Fig. 100); anterior ventral apotome small, posterior ventral
apotome minute or absent, stridulatory files present or absent; left mandible with or
without inner brush; tracheal gills in two, six or eight rows, “Zipfelkiemen” absent;
setose ventral sclerites on IX only (except in Leptonema) or absent

MACRONEMATINAE Ulmer 7

3. Naked dorsal branch of pre-episternum about \\ times length of bristly branch and

as wide basally; head and thorax with small, weak setae, mainly tapered, a few blunt
or scale-like, glabrous in appearance; 3 metasternal gills, larvae up to 15 mm long.
(Marlier 1978c, Figs 12-13.) Hydromanicus Brauer (Seychelles)

— Naked dorsal branch of pre-episternum as long as or shorter than the bristly branch,

basally narrower (Fig. 17); head and thorax strongly setose, rough in appearance 4

4. Most setae on head short, blunt, peglike; right mandible without setulae on upper

apical tooth; anterior margin of frontoclypeal apotome smooth; large paired oval
sclerites present on intersegmental fold posterior to prosternal shield; 2 pairs meta-
sternal gills, four anal gills. Larva up to c. 15 mm long. (Figs 11-19, Table I)

Hydropsyche Pictet

— Setae on head tapered; right mandible with small tuft of setulae on upper apical

tooth; anterior margin of frontoclypeal apotome often notched or crenulate; sclerites
on intersegmental fold small, sometimes concealed or absent; 3 metasternal gills,
four or five anal gills. Larvae up to c. 8 mm in length. (Figs 30-40, Table III)

Cheumatopsyche Wallengren

5. Frontoclypeal apotome widest in centre, anterior and posterior ventral apotomes

very large, together much longer than mid-ventral ecdysial line; no line of setulae on
upper surface of mandibles; head & thoracic nota with many small sharply pointed
setae 6

306

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

— Frontoclypeal apotome at least as broad anteriorly as in middle or broader, anterior
and posterior ventral apotomes fairly large, but together shorter than mid-ventral
ecdysial line; a line of setulae on upper surface of each mandible; head and thoracic
nota with many large setae, some of which are blunt much as in Hydropsyche ; basal
seta on foreclaw slender, fairly long, on other claws stumpy; branched tracheal gills
in four rows, no fan on anal prolegs. Length of larva c. 15-18 mm. (Figs 79-87,

Table VII) Diplectronella Ulmer

6. Foreclaw and basal seta slender, pointed; branched tracheal gills in four rows; few

apical setae on anal prolegs (see Wiggins 1977, fig. 6.4)

Diplectrona Westwood (N. Africa)

— Basal seta of foreclaw short, stumpy, spinelike, as is claw; branched tracheal gills in

six rows; large fans on anal prolegs. Length of larva 13-14 mm. (Figs 56-65, Table
V) Sciadorus Barnard

7. Stridulatory files absent; mandibles without lines of setulae or inner brush; pre-epi-
sternum large, sharply pointed, either collarlike (Fig. 188) or dagger-shaped (Fig.

208); body glabrous, gills on abdominal segments II-VI in eight rows. (Early instars
of both genera have the frontoclypeal apotome with characteristic triangular mar-
gin.) Tribe POLYMORPHANISIN1 Lestage 8

— Stridulatory files present; mandibles with or without lines of setulae and inner brush;

pre-episternum small, conical (Fig. 132); body glabrous or rough, gills on abdominal
segments II-VI in two, six or eight rows Tribe MACRONEMAT1NI Lestage 9

8. Pre-episternum collar-like with sharp anterior points; head and thorax long, narrow,

smooth, shining, much narrower than abdomen; frontoclypeal apotome narrow, lab-
rum large, almost circular, lacking strong lateral brushes, maxillary palpi very long,
slender, retractile; forelegs without brushes; setose plates on sterna of VIII, IX, ab-
sent. Larvae large, up to 40 mm in length or more. (Figs 185-191, Table XVI)

Polymorphanisus Walker

— Pre-episternum narrow, dagger-shaped, as long as coxa; head relatively short &
broad; thoracic segments broader than long, not strikingly narrower than abdomen;
frontoclypeal apotome broadly triangular, broad labrum with strong lateral brushes,
maxillary palpi small; forelegs with weak brushes on tibia and tarsus; setose plates
present on sternum of IX. Larva medium-sized, 14-15 mm long. (Figs 203-213,

Table XVIII) Aethaloptera Brauer

9. Plead with U- or horseshoe-shaped carina on dorsal surface, enclosing a broad, part-

ly sunken area (Figs 100-102); mandibles without inner bristles or lines of setulae;
foreleg with or without brush, fore femur with thumblike basal process (Fig. 132a);
tracheal gills on abdominal segments II-VI in six or eight rows, body smooth 10

— Head without carina; inner brush on left mandible, line of setulae along upper tooth

in both mandibles; forelegs without brushes, anterior femur without basal process
but with stridulatory protuberance, fore coxae variously armed according to species;
tracheal gills on abdominal segments II-VI in two rows, body rough, bristly (thickly
clad with small scalelike setae). (Figs 168-176, Table XIV)

Leptonerna Guerin-Meneville

10. Carina cuts off a triangular piece from apex of frontoclypeal apotome, head (known
species) bright chestnut or orange-brown, smooth, with fringe of long, stiff setae
bordering carina; anteclypeus with two sclerotized plates; prosternal plate with
median sulcus; brushes on forelegs very strong; gills on abdominal segments II & III
in eight rows, on IV-VI in six rows. (Figs 99-109, Table IX) Macrostemum Kolenati

— Carina does not cut off a triangular piece from apex of frontoclypeal apotome, head
(known species) dark brown or yellow with orange-brown or yellowish patterning;

307

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

anteclypeus with four sclerotized plates; prosternal plate without median sulcus;
brushes on forelegs weak; gills on abdominal segments II-VI in either six or eight

rows 11

11. Carina runs through apex of frontoclypeal apotome exactly through point, head
brown or yellow, usually with two conspicuous pale marks on frontoclypeal apotome
and many setulae, no long curled setae on head; abdominal segments II-VI with
gills in eight rows. (Figs 127-136, Table XI) Amphipsyche McLachlan

— Carina runs along apex of frontoclypeal apotome outlining a triangle, head in only
known species brown patterned with orange-brown, long curling secondary setae on
frontoclypeal apotome and within border of carina; gills on abdominal segments II-

VI in six rows. (Figs 148-157, Table XIII) Protomacronema Ulmer

(correlated by Gibbs (1973), confirmed by Statzner, in litt. (6.ix.l980) & paper
(1981)

(c) PUPAE

Hydropsychid pupae are easily recognizable by their branched, tufted abdominal gills and
typical anal processes (Figs 7, 8, 49, 52 etc.). The ventral gills are very similar to those of the
larvae, however, the lateral gills may have enlarged pouch-like bases or resemble conical pock-
ets, several often being superimposed (Figs 49, 122). The paired anal processes are situated at
the end of the abdomen and may be wide or narrow, usually terminating in finger-like, plate-
like or bifurcated apices (Figs 95, 125, 75). In the male there are paired ventral pockets con-
taining the genitalia. Presegmental hook-bearing plates occur on abdominal terga II/III to VI/
VII/VIII, varying according to genus, and postsegmental hook-bearing plates on terga III, III
and IV or III, IV and V, at any rate in African species (Figs 7, 7a). The abdomen does not
bear a long fringe of lateral setae, though the venter may be hairy and the dorsum may show
lines or patches of setae or patches of shagreening (cuticular thickenings resembling those on
sharkskin).

The mid-legs, particularly of the female, are generally flattened, widened and fringed with
setae to form oars. The final instar larva constructs a stout case of sand grains or rock frag-
ments, sometimes with an admixutre of vegetable matter, in which to pupate. Within this case,
which is anchored to the rock or stone adjacent to it, is a silken cocoon enclosing the pupa.
The pupal case almost invariably has sieve plates at each end to allow free passage of water. At
the anterior end is a cap of sand grains which, on emergence, is removed by the pupa (pharate
imago) using its mandibles (Figs 27-29). The discarded larval sclerites are generally retained
within the case, tucked into the posterior end. When still present they can be used to assist in
identification of the species and in the correlation of larva, pupa and imago in the case of a ma-
ture male pupa with undamaged genitalia. The pupa is exarate, leaving the case as a pharate
adult, able to swim to the surface of the water and climb up any projecting object on which
eclosion takes place and from which the perfect insect flies away.

1. Antennae short, stout, not more than If times body length at most in 8, usually
little longer than body, often shorter, in 9 . Anal appendages with apices deeply
forked (Figs 53, 75) or ending in two points with a concavity between them (Figs 8,

26a), or truncate (Seychelles genus only); mandibles and palpi always present; lat-
eral gills conical, trachael gills branched 2

— Antennae very long, filiform, 2-4 times body length, coiled round or dorsal to pos-
terior end of body in <3, in $ shorter, not coiled but recurved; anal appendages not
apically forked or cup-shaped, but apices bluntly or sharply pointed (see Figs 95,

124, 145, 165, 184, 201); mandibles present or absent, if present with well-developed
apical & lateral teeth (Fig. 97), palpi present or absent; presegmental dorsal plates

308

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

on segments II/1II— VI/VI I/VIII , postsegmental on III or III & IV; lateral gills not
conical but may have expanded pouch-like bases; tibial spurs various

MACRONEMATINAE Ulmer 7

2. Anal appendages straight, narrow, apices deeply forked (Figs 53, 75); antennae as

long as or shorter than body; sternum of V with pair of ventrolateral filiform pro-
cesses; gills on segments II— VII, (Figs 49-79); presegmental dorsal plates on II— VII-
/VIII, postsegmental on III & IV or III, IV, V; tibial spurs 2.4.4

DIPLECTRONINAE Ulmer 3

— Anal appendages curved dorsad, apices truncate or ending in two points with a con-

cavity between them (Figs 8, 26a); antennae not more than If times body length,
usually less, in 2 shorter than body; sternum of V without filiform processes; gills on
segments II— IV/VII; presegmental dorsal plates on II/III— VIII, postsegmental on III

& IV; tibial spurs 2.4.4 or 0.4.4 HYDROPSYCHINAE Curtis 5

3. Anal appendages with both branches of apical forks very long, dagger-shaped; inner

surface of mandibles proximal to teeth smooth; presegmental dorsal plates on seg-
ments II-VII, postsegmental on III & IV. (For figs see Ulmer 1957 pi. 30 figs
607-610) Diplectrona Westwood (North Africa)

— Anal appendages with branches of apical fork shorter, not dagger-shaped (Figs 53,

75) 4

4. Anal appendages with both terminal points of apical fork subtriangular, pointed,
strongly denticulate, particularly within fork; inner surface of mandibles proximal to
teeth smooth; presegmental dorsal plates on segments II— VIII, postsegmental on III,

IV and V. (Figs 49-53, Table VI) Sciadorus Barnard

— Anal appendages with terminal points of apical fork broad, blunt; inner surface of

mandibles proximal to teeth denticulate; presegmental dorsal plates on II-VII, post-
segmental on III and IV. (Figs 73-77, Table VIII) Diplectronella Ulmer

5. Anal appendages with wide, blunt apices, finely crenate or serrated apically but not

concave; presegmental places on III— VIII; mandibles with large apical teeth; ventral
and lateral gills on segments II-VII; mid-legs of 6 & 2 not widened but tarsi
fringed; antennae (<?) wound once round apex of abdomen; pupal exuviae 13 mm
long. (See Marlier 1978b figs 24-26) Hydromanicus Brauer (Seychelles)

— Anal appendages with apices clearly concave between lateral points; antennae (6)

shorter, not wound round apex of abdomen but recurved (Fig. 7); mid-legs 6, 2,
widened, flattened, with swimming fringe 6

6. Presegmental plates on III— VIII (II— VIII in a few species), postsegmental plates on

III forming, long narrow strips, on IV smaller, inconspicuous; mandibles with fairly

large terminal teeth, inner margin smooth, anal appendages with concave apices,
long fine setae within these; pupae 8-14 mm long. (Figs 7-9, Table II)

Hydropsyche Pictet

— Presegmental plates on II— VIII. postsegmental plates on III widely oval, on IV
clearly visible; mandibles with small terminal and lateral teeth, separated by lateral
serrations; anal appendages with concave apices studded with papillae and setate tu-
bercles, no long fine setae within them. Pupae 5-7 mm long. (Figs 26-28, Table IV)

Cheumatopsyche Wallengren

7. Mid-leg with two or three tibial spurs; maxillary and labial palpi absent or rudimen-
tary, mandibles present or absent; postsegmental plates on segments III or III and

IV 8

— Mid-leg with four tibial spurs; maxillary and labial palpi clearly present, mandibles
present, large; postsegmental plates on segment III only except in Protomacronema 9

309

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

8. Medium-sized pupa (length c. 10 mm); anal appendages widened at base, apices
rounded; mandibles present, small, with inflated bases and large teeth; presegmental
plates present on II— VIII , prominent, only two denticles each, postsegmental plates
narrow strips of minute denticles on segment III only; tibial spurs 0.2.2 or 0.3.3.

(Figs 199-202, Table XIX) Aethaloptera Brauer

— Pupa much larger (length c. 22 mm); anal appendages with widened apices ending in
several large blunt points; mandibles absent; presegmental dorsal plates on III— VII,
two to five denticles on each, postsegmental plates on III or III and IV, oval with
many points on III, also a narrow strip with a few minute points on IV according to
Marlier, but not evident on our material; tibial spurs 1.3.3 or 2.3.3. (Figs 183, 184,

Table XVII) Polymorphanisus Walker

9. Tibial spurs 0.4.4; each anal appendage ends in a sharp sclerotized point (Figs 145,

146); presegmental plates on V-VIII (only part of exuviae available, probably II or
III— VI 1 1 ) , there appear to be two sets of postsegmental plates (3 plates on torn part
of exuviae), probably from III and IV; no gills present on exuviae. (Figs 145-147) . .

Protomacronema Ulmer

— Tibial spurs not 0.4.4 (may be 2.4.4, 1.4.4, 1.4.3, 1.4.2, 0.4.3, 0.4.2, 0.3.2 or 0.2.2);

anal appendages do not end in such sclerotized points; presegmental plates on
II/III-VIII 10

10. Anal appendages short, broad, contracted at ends to form terminal plates fringed
with setulae; presegmental plates on II— VIII, tufted gills on II— VII/VIII, most lateral
gills with pouchlike bases; palpi present but small, mandibles short, with inflated
bases and four to five large serrated teeth; tibial spurs 1.4.4, 1.4.3, 1.4.2, 0.4.3,

0.4.2, 0.3.2 or 0.2.2. (Figs 122-125, Table XII) Amphipsyche Wallengren

— Anal appendages long, terminating in finger-like processes; presegmental plates on

II/III-VIII; branchial gills on II— VII, some lateral gills either pouchlike or with
pouchlike bases; palpi large, well-developed, bases of mandibles not inflated though
they may be wide, apical teeth very large, several smaller teeth as well, margins ser-
ratulate; tibial spurs 2.4.4, 1.4.4 or 0.4.4 11

11. Anal appendages large with “bare” dorsal apical finger (is minutely spinose, appear-
ing bare except under high magnification), otherwise thickly fringed ventrally with
very long setae; presegmental plates on III— VIII ; tufted gills in six rows on II & III,
four rows on IV- VII; a pair of small processes on fifth sternum ( <3 only), shagreen-
ing on dorsal side strong, obvious. (Figs 94-98, Table X) . . .Macrostemum Kolenati

— Anal appendages long, cylindrical, contracted apically into a fingerlike process cov-

ered with minute setiferous papillae and spinules, ventrally a shorter projection out-
lined by long setae; presegmental plates on II— VIII ; ventral tufted gills in two rows
on II- VII, conical lateral gills in two rows on III— VII . No processes on fifth sternum,
but traces of paired fenestrae (present in imago); shagreening on dorsal side scarcely
visible. (Figs 164-167, Table XV) Leptonema Guerin-Meneville

310

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS
DESCRIPTIONS, DISCUSSION, BIOLOGY AND SPECIES LISTS
Subfamily HYDROPSYCHINAE Curtis 1835

Genus HYDROPSYCHE Pictet 1834

Hydropsyche Pictet 1834: 23, 199 (for entire family plus others)

Type species cinerea Pictet, synonym of instabilis (Curtis) (subsequent selection of Ross
1944). Switzerland.

Hydropsyche Pictet; Ulmer 1907a: 169, pi. 23 figs 208b-e. 6 9 .

(Key, generic diagnosis, list of known species, including the African H. propinqua.)

Hydropsyche Pictet; Barnard 1934: 395 (diagnosis for imago).

Hydropsyche Pictet; Lepneva 1964 (1970 translation): 547 (diagnosis of larval and pupal stages).

Hydropsyche Pictet; Wiggins 1977: 106, 107, figs 6.6A-G (larva).

Hydropsyche Pictet; Fischer 1963: 6-8; 1972: 99-100.

Generic diagnosis (derived from Ulmer 1907a, Barnard 1934, Mosely and Kimmins 1953
and other sources):

Imago: tibial spurs of (3, $, 2.4.4; anterior tarsus of 6 with one daw normal, the other
concealed by a tuft of dark setae. Maxillary palpi with first segment short, second longer,
third subtriangular, third and fourth subequal, fifth as long as rest together, articulated.
Labial palpi with first and second segments short, subtriangular, third slightly longer than
others together, articulated. Mandibles sclerotized, probably functional, meeting beneath
labrum (Crichton 1957: 75, 76, fig. 43). Head: vertex with two large, oval posterior
warts, anterior warts smaller; antennae moderately slender, in 6 slighty longer than fore-
wings, in 9 shorter, with fine dark ridge winding spirally round the joints, becoming in-
distinct apically. Eyes (oculi) of c? in some species larger than in 9 . Forewings with api-
cal forks 1, 2, 3, 4, 5 present, discoidal cell short, broad, crossveins m-cu and cu well
separated; hind wings with forks 1, 2, 3, 5 present, discoidal cell always closed, median
cell present, usually closed, M and Cui close together and crossvein very short. Mid-tib-
iae and tarsi of 9 flattened, dilated. Fifth abdominal sterna lack filamentous processes or
internal vesicles (though there may be small tufts on pleura of some abdominal seg-
ments). 6 genitalia with a strong sclerotized prolongation of 9th and 10th terga and a
pair of slender, two-jointed claspers; prominent copulatory organ (phallotheca plus en-
dotheca, the aedeagus and parameres having been lost by specialization, see Schmid
1968), which in Afrotropical species (and one group of North American species) has a
long apically forked ventral endothecal lobe. $ genitalia with sternal plates of 8th seg-
ment separated for two-thirds distance to base of segment; distal margin of XI with a pair
of cerci between two pairs of small lobes.

The genus is the largest in the family and is almost cosmopolitan, being found in all
regions with the exception of the Neotropical.

Hydropsyche longifurca Kimmins
(Figs 1-19, Tables I, II)

Hydropsyche longifurca Kimmins, 1957a: 40, 41, figs 11-13 (3 genitalia).

Portuguese East Africa, Chirinda, Busi River. <3.

Holotype S in British Museum (Nat. Hist.), 4 paratypes in B.M. (Nat. Hist.) and Coryn-
don Museum, Nairobi, Kenya.

Distribution: Mozambique (Portuguese East Africa); South Africa; Natal, Eastern Transvaal;
Zimbabwe (formerly Rhodesia), at Victoria Falls.

311

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

pt

Figs 1-9. HYDROPSYCHINAE: Hydropsyche longifurca, <3, 9 pupa and probable $ . 1. <3: dorsal view of head & tho-
racic nota, 2. 6 : lateral view of head & palpi, 3,4. 3 : fore & hind wings (right), 5. 6 : lateral view of genitalia, 6. prob-
able 2: lateral view of genitalia, 6a, same, clasper receptacle further enlarged, 7. 6 pupa: habitus, right appendages
only shown, 7a, same, dorsal plates of pupa further enlarged (right), 8. 6 pupa: anal appendages, dorsal, 8a, same, ter-
minal view of anal appendages, 9. 6 pupa: right mandible, dorsal.

312

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS
Description of imagos (Figs 1-6)

Male imago (in spirit): a medium-sized species, forewings c. 9 mm long, pale fuscous, with
thin pubescence, hind wings pale yellowish; head and thorax brown with paler warts, palpi and
legs yellowish. Antennae slender, scape short, rounded, basal part of flagellum with dark spiral
ridge which fades out after about six segments. Eyes large, black. Vertex with large pair of
posterior warts, two pairs of smaller anterolateral warts, and three still smaller anterior warts
near antennal bases. Face with wide lozenge-shaped median wart. Setae on head short, curled,
golden brown. Pronotum with pair of long, tapered median warts, very small outer pair. Meso-
scutum without warts, mesoscutellum with subtriangular wart with lateral bands of setae. Ab-
domen purplish brown with usual shrunken tufts on pleura. Genitalia as in fig. 5 (see also Kim-
mins 1957a figs 11-13).

Kimmins (1957a) described the pinned 6 as: “General appearance ochraceous, abdomen
fuscous. Forewing pale fuscous, with numerous ochraceous irrorations. Venation as in H. nam-
wa Mosely.”

Probable female imago (in spirit. Figs 6, 6a): three females available. Very similar to <3,
somewhat larger, with similar large eyes; length of forewings 12 mm. Sternal plates of VIII
subquadrangular in shape, with short setae, longer setae along distal margins, only partially
separated ventrally. 8th tergites finely pubescent, with setose posterolateral angles; X hood-
like, clasper receptacles rounded (Fig. 6a), clasper grooves unclear, sclerotized knob present
on each side near genital atrium, ventral plate mainly membranous. I do not know what the
knobs may be; Barnard described the $ of his H. ulmeri as having such “chitinous knobs”, and
figures them (1934: 359, fig. 36), so both these species, ulmeri and ? longifurca , would appear
to possess them.

Material of H. longifurca in the Albany Museum comprises 6 males, 3 probable females
and over 100 larvae, also 3 pupae, including a mature male and a fairly mature female, to-
gether with pupal cases and some larval sclerites. These would be sufficient to establish the re-
lationship between the different stages of this species, were there proof as to whether the scle-
rites pertained to the male or the female (both were in the same container, with remains of
pupal cases and a few larval sclerites). The larvae have a definite and very characteristic colour
pattern on their heads, which renders recognition even of larval sclerites easy. The fact that
both adult females and female pupae were found in association with males, male pupa and lar-
vae renders it very likely that all belong to the same species (the only one so far found in these
areas), but there is still the possibility that larva and/or female could be those of ulmeri , whose
females appear very similar from Barnard’s figures, though the ventral plates of the latter dif-
fer somewhat in shape, as ulmeri has been recorded from the same general area (Zululand,
Transvaal, Orange Free State and probably Rhodesia). All males in the Albany Museum col-
lections are, however, H. longifurca (identified by both D. E. Kimmins and myself), and I
have as yet seen no ulmeri. H. ulmeri and H. longifurca are the only species recorded from
Southern Africa to date, though we do have what appears to be the female of another species
collected from the Orange River in the Warrenton area; this has subtriangular sternal plates
and lacks the lateral sclerotized knobs.

Description of probable larva (Figs 11-19; Table I)

Mature larva (in spirit; Figs 1 1-19): length 12-15 mm. Larvae considerably larger than
those of the other Hydropsychine genus found in Southern Africa (Cheumatopsyche) , and with
one characteristic feature which facilitates recogniton, even in early instars. This is the distinc-
tive shape of the great majority of the setae on head and thoracic nota, which are short and
blunt-ended, either truncated or shaped like an Indian club (Fig. 11). These short, blunt setae
are very numerous in later instars, giving the head (apart from the frontoclypeal apotome) and
thoracic nota a bristly unshaven appearance. This feature greatly facilitates separation from

AM— 3

313

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Figs 10-19. HYDROPSYCHINAE: Hydropsyche longifurca , probable larva (mature). 10. habitus of mature larva, 10a,
same, mesocoxal gill & conical lateral gill further enlarged, 11, 12. dorsal & ventral views of head, colour pattern shown
on left gena & frontoclypeal apotome, setae on right; blunt setae further enlarged, 13. prosternal plate & intersegmental
sclerites, 14. left & right mandibles, 15. right anal proleg & claw, 16. ventral view of right half of abdominal segments
VIII & IX, showing setose sclerites, 17. right foreleg, showing pleural sclerites at base, arrow indicates position of scrap-
er of stridulator; plumose seta enlarged, 17a, same, margin of femur further enlarged to show scraper, 18. mid-leg, an-
terior view, 18a, margin of femur & single plumose seta further enlarged, 19. hind leg, anterior view.

314

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Cheumatopsyche larvae even in the early instars. Diplectronine larvae also have some club-
shaped or scale-like setae, but are easily separable from they Hydropsychinae on other charac-
ters (see key).

Head (Figs 11, 12): subquadrate, dorsally flattened but without carina, with strongly
marked dark chestnut brown and yellow pattern; dorsal and lateral surfaces of genae thickly
studded with short, dark, blunt setae, very few such setae on frontoclypeal apotome except
along the anterior margin; none on posterior part of head which can be withdrawn inside the
pronotum. The colour pattern as illustrated is that most commonly seen, and at first I thought
that a dark variant might be another species, but several samples show a clear gradation from
the usual pattern by a progressive darkening, in which the light hour-glass shape on the fronto-
clypeal apotome gradually closes until only two light patches remain; the rest of the head dark-
ens similarly. Eyes black, as usual set in paler areas, each below six small cuticular lenses. Stri-
dulatory hies large, elongate oval in shape, colour pattern continued across them. Anterior
ventral apotome widely subtriangular, light brownish yellow, with sutures on both sides; pos-
terior ventral apotome minute, black.

Mouthparts (Figs 11, 12, 14): labrum brown, retractile, setose, with small, pale brushes on
ventral side; anteclypeus pale, unsclerotized; mandibles strong, dark brown, left mandible with
inner brush of setae and five apical and subapical teeth, the proximal one having two cusps;
right mandible with four teeth, no inner brush; neither mandible bears setulae on upper apical
tooth; mandibles triangular in section, outer face set with strong setae between flanges. Maxil-
lary palpi short, stout, five-segmented; labium short, thick, palpi not visible; only mandibles
seen when mouthparts retracted, as is often the case in preserved specimens. Submentum
brown, darkening to a blackish apex with deep median incision. Stipites light brownish yellow,
conspicuous.

Thorax (Fig. 10): pro-, meso- and nretanota brown, pronotum darkest, all thickly studded
with short, blunt setae interspersed with small, fine setae; nota fringed anteriorly with blunt se-
tae and finer setae; meso- and metanota with darker brown diagonal stripes, the former with
median crescent-shaped black mark at posterior margin, the latter with a smaller similar black
mark there; all three nota with wide black lateral borders. Pre-episternum bifid, dorsal branch
smooth, blackish, ventral branch bearing several setae (Fig. 17). Prosternal plate (Fig. 13) a
transverse bar with heavily sclerotized anterior and posterior margins; posterior to this a pair
of oval sclerites on the intersegmental fold, often partly or completely covered by a fold of the
epidermis. One pair of branched tracheal gills on mesosternum, two pairs on metasternum.
Thoracic pleura and sterna covered with small, fine setae interspersed with many larger, blunt
brown ones, giving a rough, hairy appearance to the larva.

Legs (Figs 17-19): stout, foreleg with tarsal claw bearing a single long basal seta, scraper
of stridulator on anterior angle of femur (see Fig. 17a), a few large plumose setae on posterior
margin of trochanter and femur, many smaller ones on coxa, a clump of stout setae on inner
side of femur. Mid- and hind legs with tarsal claws bearing a single thick basal spine, femora
with a row of feathered setae near posterior margin, coxae with a few small plumose setae on
outer side; fewer plumose setae on mid- than on hind legs.

Abdomen (Figs 10, 15, 16): this, like thoracic pleura and sterna, is rough and hairy in ap-
pearance. Densely tufted tracheal gills present in two rows on segments I and VII, four rows
on segments II-VI (see Table I). There are also minute conical lateral gills, “Zipfelkiemen” of
Ulmer, on segments III— VII. Sterna of segments VIII and IX each bear a pair of subtriangular
sclerotized plates set with stout, mainly short, spinelike setae, the plates on IX being much
larger than those of VIII. Segment IX has in addition a pair of small, pale ventrolateral scle-
rites and a pair of larger dorsolateral sclerites, often partically concealed beneath VIII. Anal
prolegs long, well developed, each bearing an apical tuft of very long dark bristle-like setae;
anal claws strong, right-angled, plain. Four anal gills.

315

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Earlier instars: there are five larval instars in the Hydropsychinae, not six as found in
those Macronematinae for which we have adequate material. All five appear to be represented
in our collection, though not in sufficient numbers to give adequate ranges of head width or ad-
equate separation of instars on a histogram.

First instar: no tracheal gills: pre-episternum minute, as yet unbranched, very few setae
but those present are comparatively long; blunt setae on head and thorax (8 on head, 6 on pro-
thorax, 9 each on meso- and metanota); only head and pronotum moderately well sclerotized
(latter showing dark margin); head with very faint indication of pattern, stridulatory files ab-
sent, two small “teeth" on anterior margin of head (egg bursters); meso- and metanota very
pale, scarcely distinguishable. All tarsal claws with three basal spines ( cf . Fig. 41), a few small
plumose setae on legs. Recognizable as Hydropsyche by presence of blunt setae, which in this
instar are very long. Four anal gills.

Second instar: simple, unbranched gills present on meso- and metasterna and abdominal
segements I-VII, arranged as in final instar but none with two main stems as yet. Pre-epister-
num with future dorsal branch visible as a point on ventral branch. More blunt setae present,
but still scanty. Head pattern still faint but recognizable; stridulatory files present but un-
coloured, scrapers just visible on fore femora. Thoracic nota all well sclerotized, with dark
margins. All tarsal claws with one basal seta or spine only (as in subsequent instars); a few
plumose setae on trochanters and femora of fore and hind legs, none seen on mid-legs.

Third instar: ventral gills with 1-2 branches, anterior ventro-lateral gills with 2-4
branches, posterior ventro-lateral gills with one only; gills on 1 and VII have developed two
main stems. Pre-episternum clearly bifid, but dorsal branch not yet fully developed. Several
feathered setae on trochanter and femur of foreleg (as in subsequent instars) and a few on hind
femur. Head pattern clear. More blunt setae on thorax, a few on abdomen.

Fourth instar: all ventrolateral gills have two main stems; ventrolateral and ventral gills
have 4-12 branches. Pre-episternum with both processes fully developed. Many blunt setae
present on head, thoracic nota and abdomen. Stridulatory files coloured and clearly visible,
head pattern fully developed.

Fifth instar: this is the mature larva described above. Ventral and ventrolateral gills with
many branches (c. 12-25); minute conical lateral gills present on segments III— VI . Farvae
strongly setose, but in older larvae whose abdomen has become much enlarged many abdomi-
nal setae may have been lost.

Description of pupa (Figs 7-9, Table II)

There are only two mature pupae in the collection, one male and one probable female,
both in fairly good condition, with cases and some larval sclerites (see above for comments).
The 6 pupa was identified by Mr D. E. Kimmins as well as by myself. There is also one imma-
ture female pupa.

Male pupa (pharate adult) (in spirit. Figs 7-9): length 8,0 mm. Fully mature, light yellow-
ish to golden brown in colour. Antenae moderately slender, slightly longer than body; imago
clearly visible through pupal pelt, head warts as imago, setae on them long, silky, pressed for-
wards; short brown setae scattered on pupal skin covering vertex, also on thoracic terga and,
more thickly, on face. Labrum rounded, with a pair of rounded setose protuberances on anter-
ior margin, a smaller lateral pair. Mandibles smooth, dark red-brown, with four (right man-
dible) or five (left mandible) large teeth along inner edge at apex. No smaller teeth or serra-
tions. Bases are minutely pubescent along inner margin. Palpi long, as in imago. Legs much as
in imago, with tibial spurs 2.4.4, mid-leg widened, flattened and with swimming fringe. Fore-
legs reach to end of II, mid-legs to VI and hind legs well beyond genitalia. Wing cases brown-
ish, forewing reaching to middle of VI, hind wing to base of V. Abdomen mottled, mainly pur-
plish brown; segments II- VII with large flattened ventral tracheal gill tufts, small conical

316

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Table I:

Number and position of tracheal gills and other characters 1st to 5th larval instars of

Hydropsyche ? longifurca

Segment no.

1st instar

2nd instar

3rd instar

4th instar

5th instar

Mesosternal

—

1 pair

1 pair

1 pair

1 pair

gills

Metasternal

—

2 pairs

2 pairs

2 pairs

2 pairs

gills

Abdominal

VL V

VL V

VL V

VL V

VL V

gills

I

— —

1

2*

2

2

II

— —

1 1

1 1

2 1

2 1

III

— —

1 1

1 1

2 1

2 1

IV

— —

1 1

1 1

2 1

2 1

V

— —

1 1

1 1

2 1

2 1

VI

— —

I 1

1 1

2 1

2 1

VII

— —

1

2

2 —

2

Tarsal claws

3 basal spines

1 basal spine

1

1

1

Gills

absent

simple

1-4 branches

c. 4-12

c. 12-25

(unbranched)

branches

branches

Head width

c. 0,2mm

c. 0,33mm

c. 0,45-0,67

c. 0,78-0,98

c. 1,25-1,40

at eyes

mm

mm

mm

Pre-episternum

simple.

2nd point just

bifid but

bifid, fully

bifid, fully

minute

visible

unequal

developed

developed

* The figure 2 denotes that the abdominal gill in question has two main stems.

Head widths are from specimens available, but numbers are insufficient to give a good indication, hence the addition of
c. = cerca.

VL = ventrolateral gills.

V = ventral gills.

N.B. There are also very small conical lateral gills on segments III— VII , mature larva.

pocket-like lateral gills on II I— VI. The ventral tufted gills on II are separate, but on III— VI
have paired stems and numerous filaments. Paired dorsal plates on III— VIII (see Table II for
gill and dorsal plate arrangement); presegmental plates on III— VIII, teeth small to very small,
postsegmental plates on III and IV, the former being elongated double strips of minute for-
ward-projecting denticles, the latter smaller and in this specimen largely concealed beneath an
epidermal fold. Shagreening practically invisible (some on anal appendages but only seen at x
400); Transverse belts of long setae on IV and V (dorsal), venter hairy, particularly posterior
to gills. Anal appendages of medium length, stout, strongly curved dorsad ending in wide cup-
like sclerotized apeces which appear shortly bifid in dorsal view, mitt-shaped in terminal view
(Fig. 8a); there are long, fine setae within the concavity. Each anal appendage bears a thick
belt of long, stout lateral setae and two or three very stout, shorter setae projecting downwards
from near the “toe”. Male genitalia in pockets as usual.

317

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Pupa of probable female: very like that of male but lacking pockets for genitalia. Female
genitalia not fully mature, but appear to be the same as those of the female imago described
above. Anal appendages similar to those of the male. Length of $ pupa 9,0 mm.

Pupal cases much damaged, but remains indicate that they were firmly built of stone frag-
ments joined with silk, with the ends capped with stones between which two or three curved
slits allowed for the passage of water. Some larval sclerites within cases, but many evidently
lost during collection. The pupa lies within a silken cocoon in the case. These two cases had
evidently been attached to one another as well as to a rock or stone, the side adpressed to the
stone being closed by a silken membrane.

Table II:

Tracheal gills, dorsal plates etc. of 6 pupa of Hydropsyche longifurca Kimmins

Abdominal
segment no.

Tracheal gills
(right side)

Dorsal plates

Setae etc.

I

L

V

11

1

1

—

III

(1)

2*

Presegmental pair

IV

(2)

2

Postsegmental pair
Presegmental pair

Thick fringe of setae

V

(3)

2

Postsegmental pair
Presegmental pair

near posterior margin
thin fringe of setae

VI

(2)

2

Presegmental pair

near posterior margin

VII

VIII

I

Presegmental pair
Presegmental pair

IX

—

—

—

* The figure 2 denotes that the gill in question has 2 main stems

(1) Brackets indicate that the gill in question is unbranched and pocket-like, with one, tw-o or three pockets superim-
posed on one another.

L = lateral gills (pocket-like)

V = ventral gills — flattened, thickly tufted branchial gills.

Remarks

Hydropsyche being an almost cosmopolitan genus rich in species (less so in Africa than in
the Northern Hemisphere), there is much literature on it.

African species of Hydropsyche include 16 species from the Afrotropical mainland, of
which only two have been recorded from Southern Africa, and nine or ten from the Malagasy
Subregion. In addition there are about ten North African species, which form part of the Pal-
aearctic fauna. Several of the Afrotropical species may be invalid because they were described
from females only, which, when finally correlated, may prove to belong to species at present
known from the male sex only. Such lack of correlation between males and females in many
species is a clear warning against description of new species based on females.

318

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Navas’s genus Plesiopsyche has been placed in synonymy with Hydwpsyche by Ross &
Unzicker (1977), who state that the 9 concerned is a typical member of the H. propinqua
group.

The genus Symphitopsyche was originally created by Ulmer (1907b) for Hydropsyche mau-
ritiana McLachlan, but has been generally regarded as a probable synonym of Hydropsyche.
Ross & Unzicker (1977), however, resurrected the genus and applied it to many American
species of Hydropsyche as well as to all the Afrotropical species (the “propinqua group”).
Schmid (1979) pointed out that Ross & Unzicker’s “new genera” (including Symphitopsyche)
were merely species groups of Hydropsyche. The name Symphitopsyche thus falls into the syn-
onymy of Hydropsyche. This is borne out by the fact that I found the “Symphitopsyche” larvae
from the U.S. A. (very kindly sent to me by Dr G. A. Schuster) were virtually indistinguishable
from our Hydropsyche larvae, such minor differences as are present certainly being no more
than distinctions between groups of species. Thus Symphitopsyche mauritiana reverts to Hy-
dropsyche. S. plutonis Banks, however, is not a Hydropsyche and has been transferred to the
genus Cheumatopsyche (Mosely 1939b).

Larval and pupal stages of Hydropsyche have been described, often in great detail, from
other parts of the world, notably by Siltala (1907, two species) and Lepneva (1964, 1970 trans-
lation, several species). As far as Afrotropical species are concerned, Marlier (1962b) gave a
general description of larva and pupa; he also briefly described the pupa of H. bwambana
Mosely (Marlier 1943b).

The larvae and pupae ascribed tentatively to H. longifurca Kimmins and described above,
almost undoubtedly pertain to that species, the only doubt being the recorded presence of
H. ulmeri Barnard in at least part of the same area. All the males in the Albany Museum col-
lections are, however, longifurca. In addition to the longifurca-type larvae and pupae, there is
material of a different species from the Inyanga Highlands of Rhodesia (Zimbabwe) which in-
cludes larvae and immature male pupae, unfortunately too immature for identification to
species. These show definite differences from the ones described above. The larvae are very
dark brown in colour, younger instars uniformly so, as are some mature larvae, other mature
larvae show one or two small pale patches on the frontoclypeal apotome. There are also a few
plain brown larvae from other areas in Southern Africa, including places from which we have
many more of the patterned longifurca- type larvae. The Inyanga pupae, while clearly a species
of Hydropsyche, do show minor differences from those of longifurca, particularly in gill ar-
rangement, the ventral gills on the seventh abdominal segment having paired stems, and there
being lateral gills on it as well. Abdominal sterna II- VI each have a pair of round blackish
spots, placed towards the sides, not seen in longifurca. It is possible that these larvae and pu-
pae are of ulmeri, the only other species so far recorded from Southern Africa.

Biology

Young larvae are often found amongst vegetation or under stones, and feed on micro-
plankton (Kaiser 1965). Older larvae are predacious. They build themselves shelters of bits of
stone and vegetable matter fastened together with silk, within which they live. In front of
these, facing into the current, they weave coarse-meshed catching nets of silk to trap their
prey, small planktonic or benthic animals, and inevitably also scraps of plant material, algae
and detritus, most being eaten but unwanted matter thrown out. The nets are usually held
open by small sticks and silken guy ropes as well as by the pressure of the current. There is an
exit hole at the posterior end of the shelter. Where stones or sand are scarce. Hydropsyche lar-
vae utilize crevices in which the shelter may be made of silk only, without the usual reinforce-
ment.

319

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8. JUNE 1983

The larvae remain in their retreats, holding on by means of their anal claws, though they
may scramble half out to forage (Fremling 1960). Under unfavourable conditions (slow flow or
still water, as in the laboratory), larvae may leave their retreats, fail to build nets and fight one
another (Philipson 1969). Malicky (1973) mentions that larvae do not weave nets in very cold
northern winters, but in the less severe African climate nets always appear to be present in
suitable situations. Hydropsyche larvae may be found in fast-flowing, clean, cool, clear, silt-
free streams and rivers. They also occur in large lakes along stony wave-washed shores where
the water is well aerated.

The Inyanga (Zimbabwe, formerly Rhodesia) specimens (not longifurca but another
species) were collected from a small cascade just above the Pungwe Falls, from rock surfaces in
fairly deep water, flowing fast and strongly. At this site shelters and pupal cases were common
and conspicuous, being made of white quartz chips. At the upstream end of each shelter was a
wide net, bellied out by the current. In living larvae the head, thorax and legs were dark
brown, the abdomen bright green. The net was supported by bundles of twiglets bound to-
gether with silk and arranged round the opening of the shelter and also around the edge of the
net. Water temperature at the time of collection was 12-14° C, and pH was 6, 8-6, 9. Turnbull-
Kemp (1970) commented that the Hydropsyche and Cheumatopsyche species at Inyanga form
an important source of food for trout. They are known to be among the Hydropsychid pred-
ators of Simulium larvae in Uganda (Hynes & Williams 1962), and the same may well apply
here.

Hydropsyche species are, on the whole, poorly represented in Africa. In Central Africa
there are some 13 species. In South Africa only two have been recorded to date and these may
well be restricted to Natal and parts of the Transvaal and Orange Free State. The Albany Mu-
seum collections include larvae from the Orange River at Prieska and a single female from the
Storms River (South Cape), but no other specimens from the Cape Province.

Prior to pupation the larva constructs a strong dome-shaped shelter of large sand grains,
sometimes with scraps of vegetation, either by altering the larval shelter (Fremling 1960) or by
construction of a new one (Sattler 1958 and others). Inside this it enters the prepupal stage,
subsequently pupating within a silken cocoon. At each end of the case there are slits to permit
a current of water to flow through. At the anterior end these are situated in a cap or lid of
stones which can be cut loose by means of the pupal madibles when eclosion takes place.
Should the current speed or water level drop and the oxygen supply diminish, the pupa main-
tains rhythmic undulatory movements of the abdomen which keep the water circulating
through the case and round the gills. At eclosion the pupa, by now a pharate adult, cuts open
cocoon and top of case, swims to the surface by means of its oar-like mid-legs, climbs out onto
a stone or stick and there ecdyses, freeing itself from the pupal exuviae and flying away.

Fremling (1960) gave an interesting and detailed account of the life cycle of H. orris in the
Mississippi River below the dam at Keokuk, Iowa, where this species abounded (together with
other Trichoptera) during the summer months, creating an acute nuisance problem in the adja-
cent town where adults were attracted to the town lights in vast numbers. At dusk and shortly
before dawn adult males formed dense swarms, to which females were attracted, in the air
near the river banks. Mating couples dropped to the ground. Mating also occurred on lighted
windows and elsewhere without prior swarming. Females appeared to have scent glands along
the sides of the abdomen which attracted the males. On heavily overcast days they might still
swarm during daylight hours but on bright sunny ones sheltered under vegetation. The H. orris
females not having fully mature eggs on emergence laid them some time later. They were able
to survive for up to ten days if provided with water, longer if sugar solution was available. A
day or two after mating the females walked or swam down to a depth of as much as 12 ft
(c. 360 cm) to oviposit on submerged objects or floats, evidently preferring these to floats at
the surface, oviposition being most concentrated at depths of 3-4 ft (c. 90-120 cm). They then

320

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

emerged from the water but might re-enter it before dying. Larvae were found in large num-
bers in the river at depths of up to 40 ft (c. 1 200 cm) on buoy anchoring chains and in turbine
chambers, and appeared to pass through five larval stadia. Fremling considered that the species
was bivoltine, adults emerging throughout the warmer months, with peaks in early and late
summer. Recent work (Mackay 1978) shows that Hydropsyche species may be either univoltine
or bivoltine and that this may vary in different areas.

Fremling found that the H. orris larvae occurred in faster current than the other Hydro-
psychidae present, forming 100% of samples at a speed of 5 ft/sec. (c. 150 cm/sec.) and 90% at
3 ft/sec. (c. 90 cm/sec.) whereas Cheumatopsyche campyla larvae only formed 2% of samples at
the latter speed. He found that H. orris eggs took 8-11 days to hatch in the laboratory and that
the first instar larvae were able to swim well. Later instars were not observed swimming.

D. Scott (1958), in his study of the River Dean, observed that at surface current speeds of
over 40 cm/sec. Hydropsyche larvae occurred under stones, preying on the rich concentration
of smaller insects sheltering there. Chutter (1970) noted that in Natal Hydropsyche larvae were
found in stony runs in the eroding zone of rivers at all seasons of the year.

African Species of Hydropsyche

H. abyssinica Kimmins, 1963:140-142, figs 54-59 (wings, 3, 2 genitalia). 3 $ .

Distribution: Ethiopia.

H. alluaudina (Navas), 1931a: 127, fig. 64 (2 wing), as Plesiopsyche. 2 .

Ross & Unzicker 1977: 304-305, synonymize Plesiopsyche with Hydropsyche. In any case
the species may not be a valid one, having been described from the 2 only.
Distribution: Ruwenzori.

H. atlanta Botosaneanu, 1975: 272, 273, fig. 1 ( <5 genitalia). <3.

Synonym of H. obscura Navas (Botosaneanu 1980).

Distribution: Grand Atlas, North Africa.

H. angulata (Navas), 1934c: 73, fig. 14 (wing), as Symphitopsyche. Sex not stated, identity un-
clear.

Distribution: Madagascar.

H. bwambana Mosely, 1939b: 24-25, figs 72-74 (intromittent organ). 3 2 .

Marlier 1934b: 20, 21, fig. 27 (pupal mandibles and appendages). Description of 3 pupa.
Distribution: Ruwenzori; Belgian Congo.

H. contubernalis McLachlan, 1865: 129-130, pi. 13, fig. 12. 3. Europe.

Botosaneanu (1967, Limnofauna Europaea) lists this from Morocco.

Fischer, 1963: 41, places contubernalis in the synonymy of H. guttata Pictet.

Dakki, 1978: 113, refers the specimens concerned to H. resmineda Malicky.

Distribution: remains Europe only, Morocco is incorrect.

H. diminuta Walker, 1852, now placed in Cheumatopsyche.

H . fasciolata Navas, 1926: 109, 110, fig. 43 (3 genitalia). 3.

Distribution: Belgian Congo.

H.fezana Navas, 1932b: 111-112, fig. 4(2 forewing). 2. (Also given as fenzana.)

Dakki, 1978: 116, figs 4, 5, 6, describes Navas’ 2 from the Inst. Sci. Mus. Rabat. He con-
siders that material described as H. instabilis from North Africa is in fact H. fezana,
thus also extending the distribution to Algeria.

H. ouedorum Botosaneanu 1975 is a synonym of H. fezana (Dakki 1978).
Distribution: Morocco, Algeria.

H.fulvipes (Curtis), 1834: 213, as Philopotamus. 2.

Tobias, 1972: 262-264, figs 126-129, redescription of 3, 2 .

Dakki, 1978: 113; Navas’ specimens identified as fulvipes are in fact a 3 pellucidula and a

AM— 4

321

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

9 maroccana; the North African distribution thus falls away.

Distribution: Europe (Morocco incorrect).

H. furcata Jacquemart, 1963b: 1-2, fig. 1 (<? genitalia). 8.

Distribution: Mauritius.

H. guttata Pictet, 1834: 203-204, pi. 17f, figs 3a-d (also metam.) 8 9 .

Mosely 1939c: 186, 189, figs 397-399 (8 genitalia). 8 .

Tobias 1972: 248-249, figs 79-84 (8, 9 genitalia). 8 9.

Dakki 1978: doubts presence of this species in North Africa.

Distribution: Europe, ? Algeria.

H. hirta Jacquemart, S. & Statzner, B., 1981: 13, pis VI figs 10-13 & XV fig. 8 (8 gen., wing).
8.

Distribution: Zaire.

H. inflata Navas, 1917: 118, 119, fig. la-d (wing tips, 8 genitalia). 8.

Dakki 1978: 113; type appears to have been lost, but he thinks that Navas7 material was
probably of H. lobata, whose larvae have been found in this area.

Distribution:? Morocco, Algeria.

H. instabilis (Curtis), 1834: 213, as Philopotamus. 9.

Hildrew & Morgan, 1975: 217-219, figs 3, 7 (d genitalia), discuss this species, pointing out
that it has frequently been misidentified, the caddisfly in question being in fact siltalai
(Ibid., figs 1, 5). It is therefore uncertain whether the North African specimens are
instabilis or siltalai. Hildrew & Morgan gave a key to adult 8 8 and notes on larvae.
Dakki, 1978; considers North African material to be H. fezana Navas.

Distribution: Europe and possibly North Africa (Morocco, Algeria).

H. jeanneli Mosely, 1939a: 299-300, figs 16-19 (8 genitalia. 8. (9 collected but not de-
scribed.) (Note that Moseley’s figures have been reversed and the captions attached to the
wrong ones; figs 16-19 labelled H. jeanneli are to be found on p. 301 over the captions for
Chimarra elga.)

Distribution: Mount Elgon (Kenya).

H. lobata McLachlan, 1884: 43, pi. 5, figs 1-4. 8 .

Dakki, 1978; attributes Navas' H. inflata specimens to H. lobata, describes the 9 and has
collected larvae from Morocco.

Distribution: Europe, Morocco.

H. longifurca Kimmins, 1957a: 4(1-41, figs 11-13 (8 genitalia). 8 .

Jacquemart 1963a: 370, fig. 26a-d (wings, 8 genitalia).

Distribution: Portuguese East Africa; South Africa (Natal, Eastern Transvaal); Zim-
babwe.

H. longipalpis Banks, 1920: 357, pi. 2, fig. 22. 8 .

Distribution: Madagascar.

H. maniemensis Marlier, 1961: 159, 189-191, fig. 18 (wings, 8 genitalia). 8 9 .

Distribution: Kivu.

H. marlieri Jacquemart, S. & Statzner, B., 1981: 13, pis VI figs 7-9 & XV fig. 7 (8 gen., wing).
8.

Distribution: Zaire.

H. maroccana Navas, 1936b: 95, 96, fig. 16 (8 genitalia). 8 .

Dakki, 1978: 113, 114, places H. timha Mosely in synonymy and illustrates 9 genitalia.
Distribution: Morocco, Canary Islands.

H. maura Navas, 1932b: 110, 111, fig. 3 ( 9 hind wing). (Also as mauri). 9 .

Probably a doubtful species as 9 only and whereabouts of type unknown.

Distribution: Morocco.

322

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

H. mauritiana McLachlan, 1871: 133, pi. 4, fig. 17 {3 genitalia). 3 .

Ulmer, 1907b: 32, 33, fig. 48 (wings), places it in the genus Symphitopsyche, as did subse-
quent authors until Schmid (1979) indicated that its place is in the genus Hydro-
psyche. (See under Remarks below.)

Distribution: Mauritius, Madagascar.

H. mokaensis Jacquemart, 1960: 3, fig. 2a-d (wings, 3 genitalia). 3 .

Distribution: Mauritius.

H. moselyi Jacquemart, S. & Statzner, B., 1981: 12, 13, pis VI figs 4-6 & XV fig. 6 (3 gen.,
wing). 3 , 1, p.

Distribution: Zaire.

H. namwa Mosely, 1939b: 23, 24, figs 67-71 (wings genitalia, 3). 3 9 .

Distribution: Ruwenzori.

H. nubila Gauthier, 1928. A nomen nudum never later substantiated.

H. obscura Navas, 1928a: 189-190, fig. 4 (wingtip, 3 genitalia). 3 .

Botosaneanu, 1980: 191, places H. atlanta Botosaneanu in synonymy with it.

Distribution: Algeria, Morocco.

H. ornatula McLachlan, 1878: 363-364, pi. 39, figs 1-3 (3 genitalia). 3 9 .

Mosely, 1939a: 186, 188-189, figs 394-396 (wings, 3 genitalia).

Dakki, 1978: 113, considers that the Moroccan specimens were in fact H. resmineda Mc-
Lachlan.

Distribution: Europe, Morocco and Algeria falling away.

H. ouedorum Botosaneanu, 1975: 272-275, fig. 2 (d genitalia). 3 .

Dakki, 1978; places it in synonymy with H. fezana.

Distribution: Morocco.

H. palpalis Navas, 1936a: 108, 109, fig. 41 (wing). 3.

Distribution: Madagascar.

H. pellucidula (Curtis), 1834: 213, as Philopotamus. 9 .

Mosely, 1939a: 186-187, figs 388-391 {3,9 genitalia). 3 9 .

Dakki, 1978: 113, 3 identified by Navas (1935) as fulvipes was pellucidula, and the species
is widely distributed in Morocco.

Distribution: Europe, Morocco, Algeria.

H. plesia Navas, 1934a: 372, fig. 3 (hind wing). 9 .

Distribution: Kenya.

H. propinqua Ulmer, 1907b: 21-23, figs 32, 33 (3 genitalia). 3 .

Distribution: Cameroons; Togo; Tanganyika; Ethiopia; Congo; Ruanda; Madagascar.
South African material transferred to H. ulmeri n. sp. by K. H. Barnard (1934).

H. sexfasciata Ulmer, 1904b; now placed in Cheumatopsyche.

H. resmineda Malicky, 1977: 6, 7, pi. 3 (d genitalia). 3 . Morocco.

Dakki, 1978: 112, 113, 118, fig. 2 ( 9 genitalia), describes 9 & attributes the H. ornatula
McLachlan from Morocco to this species.

Distribution: Morocco.

H. striolata Navas, 1934b: 88, 89, fig. 54 ( 9 wing). 9 .

Distribution: Cameroons.

H. timha Mosely, 1938: 272-274, 277 , figs 1-3 (3 genitalia). <5.(9 not described although ma-
terial available).

Dakki, 1978: 114, places timha in synonymy with H. maroccana Navas.

Distribution: Morocco; Canary Islands.

H. ulmeri Barnard, 1934: 359, 360, fig. 36a-c (3 genitalia), d 9. Barnard here described the
Zululand material placed by Ulmer (1907b) in H. propinqua as a new species.

Distribution: South Africa (Zululand, Transvaal).

323

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

H. vulpina Navas, 1934b: 89-90, fig. 55 (9 hind wing). 9 .

Distribution: North Africa (Cyrenaica).

H. wamba Mosely, 1939b: 25-26, figs 75-79 (8 genitalia). <3,9.

Distribution: Ruwenzori.

H. wittei Marlier, 1943a: 8-9, fig. 5 (<? genitalia). 8.

Distribution: Belgian Congo.

Genus CHEUMATOPSYCHE Wallengren 1891

Cheumatopsyche Wallengren 1891: 138, 142.

Type species Hydropsyche lepida Pictet.

Cheumatopsyche Wallengren; Kimmins 1960a: 255-267, 76 figs, review of African species.
Cheumatopsyche Wallengren; Kimmins 1963: 130, 131, synonymy of Hydropsychodes Ulmer
with Cheumatopsyche and new diagnosis.

Cheumatopsyche Wallengren; Lepneva 1964 (1970 translation): 581-590 (larval and pupal
stages).

Cheumatopsyche Wallengren; Wiggins 1977: 100, 101, Figs 6.3A-F (larva).

Cheumatopsyche Wallengren; Fischer 1963: 106; 1972: 133.

Generic diagnosis (derived from Kimmins 1963, Mosely and Kimmins 1953 and other sources):
Imago: tibial spurs 2.4.4, sometimes 0.4.4 in male, forelegs with both tarsal claws clearly vis-
ible. Maxillary palpi with first segment short, second usually cylindrical, sometimes triangularly
dilated, third and fourth subequal, fifth about as long as first three together, articulated. Labial
palpi small, slender, three segmented, third articulated. Antennae may or may not have diag-
onal dark markings on basal segments. Head with large posterior warts, smaller anterior ones.
In forewing discoidal cell closed, forks 1, 2, 3, 4, 5 present, crossveins m-cu and cu situated
fairly close together. In hind wing forks 2, 3, 5 (occasionally 1, 2, 3, 5) present, discoidal cell
closed, median cell open, veins M and Cm widely separate and crossvein between them nor-
mal, obvious. Male genitalia of Afrotropical species show no special differences from those
elsewhere. Sternal plates of VIII in 9 separated to base of segment (cf. Figs 6 & 25), distal
margin of XI as in Hydropsyche , with one pair of cerci between two pairs of small lobes;
median leg of 9 dilated, fringed.

The genus is a very large one, and is represented in all regions except the Neotropical.
Cheumatopsyche thomasseti (Ulmer)

(Figs 20-42, Tables III, IV, also histogram, Fig. 214A)

Cheumatopsyche thomasseti (Ulmer), 1931: 18-19, Figs 16, 17, (8 genitalia) as Hydropsy-
chodes. Natal, Weenen. 8.

7 syntypes in British Museum (Nat. Hist.).

Distribution: South Africa: Natal, Cape Province, Transvaal, Orange Free State; South West
Africa; Angola; Zimbabwe (Rhodesia). Larvae were also found in a collection from Kenya.

Description of imagos (Figs 20-25)

Male imago (described from fresh specimens; pinned and spirit material also available): a
small-medium species, forewings 6-7 mm long, when alive silver-grey, wings densely pubes-
cent, silvery mottled with darker grey, abdomen greenish to olive. Colour may vary in differ-
ent areas as Ulmer described the type material (presumably pinned) as being light yellow to
golden, the forewings marbled with dark brown, fringed with grey-brown setae, hind wings iri-
descent, abdomen greenish to blackish olive. Spirit material yellowish brown. Antennae slight-
ly longer than wings, slender, very like those of Hydropsyche. Eyes large, black; head warts
shown in Fig. 20, the anterior warts being somewhat variable in shape. Genitalia discussed and
well figured by Kimmins (1960a: 263-265, Figs 28-31), who also gives a key to the then known

324

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Figs 20-29. HYDROPSYCHINAE: Cheumatopsyche thomasseti, <3 , 2 , d pupa and pupal case. 20. <3: dorsal view of
head and thoracic nota, 21. <5: lateral view of head and palpi, 22, 23, 3: fore and hind wings (right). 24. 3 : hind leg
(left), 24a, same, fifth tarsal joint, paired claws and pulvilli, further enlarged, 25. 2: lateral view of genitalia, 26. 3
pupa: apex of abdomen showing anal appendages and pockets for 3 genitalia, 26a, apex of anal appendage further en-
larged, 27. right mandible of pupa, ventral, 28. left mandible of pupa, dorsal, 29, pupal case, lateral, 29a, end view of

same, showing sieve plate.

325

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

species with comparative drawings of males and females of a number of them.

Female imago (in spirit): slightly larger, forewing 7-8 mm long, colour when alive as 3 ;
antennae shorter than wings; eyes, palpi etc. as in 3. Sternal plates of VIII subquadrangular in
shape with rounded apical angles, well separated right to base medially, setose and with long
posterior fringe, X hood-like, clasper receptacle small with long groove (see Fig. 25 and Kim-
mins 1960a: 265, Fig. 71, $, drawn from specimens bred out from Ch. thomasseti larvae col-
lected from the Great Berg River, Western Cape Province).

Material available in the Albany Museum Collection includes over 150 males, over 90 fe-
males (the latter not all with abdomen cleared, but Fig. 25 was drawn from cleared material
compared with a Berg River specimen identified by D. E. Kimmins and bred out from a
known larva). There are also hundreds of larvae and about ten 3 and 9 pupae, of which two
mature male pupae with all relevant sclerites in their cases establish the correlation between
adults, larvae and pupae of this species. The larvae, described below and illustrated, are very
distinctive and easily distinguished from those of other Cheumatopsyche species of which we
have correlated larvae.

Description of larva (Figs 30-42, Table III, histogram. Fig. 214A.)

Mature larva (in spirit Figs 30-40): length up to 8,0 mm, abdomen at widest point up to
1,3 mm, sclerotized parts light yellow to brown, head with strong brown and yellow pattern
(Figs 30-32), abdomen in life olive brown to ochraceous in colour. The larvae are strongly
setose, the setae on head and thoracic nota being short and sharply pointed, with none of the
blunt setae so characteristic of Hydropsyche larvae.

Head (Figs 30-32): short, almost square, dorsal side flat but without carina, with very
characteristic white or yellowish square on centre of frontoclypeal apotome, rest of head
brown dorsally and laterally, apart from yellowish areas round eyes and near posterior margin
as shown; ventral side yellowish. Dorsal side with many short setae except on frontoclypeal
apotome which is minutely pitted but bears few setae, though it may often be covered with a
woolly coat of “Aufwuchs”, very characteristic of this species. Anterior margin of fronto-
clypeal apotome black, crenate, straight or slightly concave. Eyes medium-sized, black, placed
well forward beneath small cuticular lenses. Stridulatory files present but not easily distin-
guishable as striae are fine and same colour as genae. Anterior ventral apotome light brown,
posterior ventral apotome minute, brownish black.

Mouthparts (Figs 32-34): labrum sclerotized, pale brown, setose, without lateral expan-
sion but with large ventral brushes of setae, usually partly or entirely retracted beneath it;
anteclypeus membranous. Mandibles short, broad, three-sided, the narrow outer side bearing
a few large setae. Left mandible with three terminal and three subterminal teeth, below these
an irregular shelf-like protuberance across which lies a strong inner brush of setae. Right man-
dible with two terminal and three subterminal teeth, below the latter a shelf-like projection, no
inner brush but a small tuft of setulae on upper apical tooth. Maxillary palpi short, stout, five-
segmented, labium also short, thick, palpi not seen. Submentum light brown with deeply cleft
apex, stipites paler brown.

Thorax (Figs 30, 35): thoracic nota ochraceous, pubescent, lateral margins blackish, meso-
and metanota with faint brownish diagonal stripes, the former with crescent-shaped black mark
in centre of posterior margin, the latter with a much smaller one. Anterior part of pronotum,
particularly front edge, with scattered branched setulae in addition to simple ones. Pre-epister-
num bifid, dorsal branch smooth, ventral branch setose. Prosternal plate with dark anterior
and posterior borders, no intersegmental sclerites posterior to it. Membranous parts of thorax
whitish, with many small inconspicuous setae. One pair of mesosternal gills, three metasternal
gills (one pair coxal gills, a single median gill).

326

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Figs 30-42. HYDROPSYCHINAE: Cheumatopsyche thomasseti, mature larva. 30. habitus of mature larva, 31, 32. dor-
sal and ventral views of head, colour pattern shown on left gena and frontoclypeal apotome, setae on right, with 2 setae
further enlarged, 33. labrum (dorsal), 34. left and right mandibles, dorsal view, 35. prosternal plate, 36. left foreleg (lat-
eral view) with pre-episternum, 37. left mid-leg (median view), 37a. the same, femoral margin & single seta further en-
larged, 38. left hind leg (median view), 39. ventral view of abdominal segments VIII & IX showing sclerites and setae,
40. right anal proleg and claw. Cheumatopsyche thomasseti , first instar larva. 41. claw of left mid-leg, 42. anal claw.

327

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Legs (Figs 36-38): yellowish with a few dark markings, claws red-brown, with single basal
spine. Foreleg relatively stout, particularly coxa and femur, claw large, strong, basal spine
long, no tarsal brush, branched and plumose setae on coxa, trochanter and femur as indicated.
Scraper of stridulator on femur much as in Hydropsyche. Mid-leg longer, hind leg longest,
bearing feathered setae as shown, basal spines short, stout.

Abdomen (Figs 30, 39, 40): ochraceous in life, white in spirit, with many inconspicuous se-
tuiae which tend to disappear in larger individuals except in folds and on the ventral side of the
body. Tufted tracheal gills present on abdomen in two to four rows (see Table III). Those ab-
dominal gills which have two main stems retain a common basal stalk even in the largest speci-
mens, and clearly arise as branches of a single gill. Single-stalked gills have 5-9 filaments,
double-stemmed ones 9-16 filaments. Conical lateral gills present on segments III— VII , very
small and inconspicuous. Sternum of segment VIII bears small pair of triangular sclerites set
with short, stout setae and bordered posteriorly with long, dark ones; sternum of IX with a
pair of much larger bifid sclerites similarly set with setae, also two pairs of smaller sclerites, lat-
eral and dorsolateral, each with a few setae. Anal prolegs long, well developed, each with well
sclerotized lateral sclerite and terminating in a fan of long, dark setae; anal claws right angled,
plain. Four anal gills.

Earlier instars: the large number of larvae present in many of the samples studied has
made it possible to work out the instars with reasonable accuracy, basing this primarily on the
width of the head capsules at the eyes. Heads of 166 larvae collected from the Vaal River near
Warrenton, North Cape, were measured, and the results are shown as a histogram (Fig.
214A), which indicates that there are five larval instars as is usual in the Hydropsychinae, the
fifth showing two peaks, presumably representing males and females (the latter being the
larger). Many specimens of each instar were present in the samples, and a brief description of
their characteristics follows. All instars were clearly recognizable as Cheumatopsyche , of which
only this species was encountered in the Warrenton area at the time.

First instar (Figs 41, 42): no tracheal gills; pre-episternum simple, head capsule and prono-
tum lightly sclerotized, the latter showing brown margins, no head pattern or stridulatory files
visible as yet; meso- and metanota scarcely sclerotized, the latter barely distinguishable, no
prosternal plate. Few setae, long and tapered, on head and limbs; legs with three basal spines
to each claw (Fig. 41), a few feathered setae on tibiae. Anal claws (Fig. 42), simple, curved but
not right-angled, no terminal fan.

Second instar: simple unbranched gills present on meso- and metathorax and abdomen.
Pre-episternum shows second point, but may appear undivided except under compound micro-
scope. Meso- and metanota clearly sclerotized. pubescent, head darker but pattern not yet vis-
ible. Stridulatory file distinguishable under compound microscope. Prosternal plate present.
Tarsal claws with a single basal spine as in later instars, forelegs with a few feathered setae on
femora. Anal prolegs without brushes, but with two long, strong setae next to each claw. Ven-
tral plates on abdominal segments VIII and IX just visible.

Third instar: ventrolateral gills on abdominal segments II- VI, with with two main stems
arising from a common base (ventral gills always remain single). All gills still unbranched (i.e.
without lateral filaments). Pre-episternum with second point well-developed, but only clearly
seen under compound miscroscope. Head darker, with heavy labrum, pattern not yet clear.
Pro-, meso- and metanota with many fairly long setae. Ventral plates on VIII, IX clear, but
stridulatory file still only visible under compound microscope. Foreleg with a few branched
(trident or quadrident) setae on coxa, several plumose setae on trochanter and femur; mid-and
hind legs with a few branched and plumose setae on femora; anal prolegs with thin brush of se-
tae, anal claws not yet right-angled.

Fourth instar: meso- and metasternal gills with three filaments each, abdominal gills with
one to three according to position. Head pattern fairly clear. Stridulatory file just visible under

328

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

stereomicroscope. Pre-episternum clearly bifid; forelegs with several long plumose setae on
femora, many trident setae on coxae. Mid- and hind legs with a row of branched and feathered
setae on femora and a few other feathered setae; a few branched setae on tarsi and coxae.
Brushes on anal prolegs not yet fully developed.

Fifth instar: this is the final instar described above. See Table III for details of gills etc.

Table III:

Number and position of tracheal gills and other characters 1st to 5th larval instars of

Cheumatopsyche thomasseti

Segment no.

1st instar

2nd instar

3rd instar

4th instar

5th instar

Thoracic gills

Mesosternal

—

1 pair

1 pair

1 pair

1 pair

Metasternal

—

3 gills

3 gills

3 gills

3 gills

Abdominal

VL V

VL V

VL V

VL V

VL V

gills

I

— —

1 1

1 1

1 1

1 1

II

— —

1 1

2* 1

2 1

2 1

III

— —

1 1

2 1

2 1

2 1

IV

— —

1 1

2 1

2 1

2 1

V

— —

1 1

2 1

2 1

2 1

VI

— —

1 1

2 1

2 1

2 1

VII

— —

1

1

1

1

Tarsal claws:

basal spines

3

1

1

1

1

Gill branching

no gills

simple gills

simple gills

1-3 branches

5-9 branches
per stem

Head width at

eyes in mm

0,12-0,20

0,23-0,32

0,37-0,45

0,55-0,65

0,78-0,81 3
0,83-0,91 2

Pre-episternum

simple.

simple with
lateral point

bifid

bifid

bifid

* The figure 2 denotes that the abdominal gill in question has two main stems arising from a common base (these orig-
inate as branches of a single gill and do not become completely separate basally). Gills of right side shown.

VL = ventrolateral gills.

V = ventral gills.

N.B. There are also very small conical lateral gills present on segments III— VII in the mature larva.

Description of pupa (Figs 26-29, Table IV)

There were some 10 pupae available for study, including two mature males, one almost
mature male and three females (one nearly mature), each in case with larval sclerites, also pre-
pupae in cases.

AM— 5

329

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983
Table IV:

Tracheal gills, dorsal plates etc. of <3 pupa of Cheumatopsyche thomasseti (Ulmer)

Abdominal
segment no.

Tracheal gills
(right side)

Dorsal plates

Setae and points
(shagreening)

I

L

V

—

—

—

—

II

2

Presegmental pair
(3-4 crochets)

fine shagreening; a few small
ant. setae & 3 pairs post,
setae

III

1

2

Presegmental pair
(c. 6 crochets)
Postsegmental pair
(large, oval, 17-20
crochets)

fine shagreening over whole
central area; some small
ant. setae & 4 pairs post,
setae

IV

2

2

Presegmental pair
(c. 6 crochets)
Postsegmental pair
(smaller oval, 6-7
crochets)

fine shagreening in centre,
post, band of long setae &
a few small ant. setae

V

3

2

Presegmental pair
(4 crochets)

fine shagreening down

centre; c. 36 pairs of post,
setae & some smaller ant.
setae

VI

3

2

Presegmental pair
(4 crochets)

fine shagreening over central
area; c. 3 pairs setae

VII

2

—

Presegmental pair
(4-5 crochets)

median shagreening & c. 6
pairs setae

VIII

—

—

Presegmental pair
(4 crochets)

row of c. 10 long setae

L = lateral diverticula (“conical gills”); where there are two or three the anterior one is smallest, the posterior one larg-
est.

V = ventrolateral tufted gills, the figure 2 denoting that the gill has two main stems.

The postsegmental dorsal plates have minute crochets or teeth, much smaller than those of the presegmental dorsal
plates, projecting forwards not backwards.

Male pupa (pharate adult) (in spirit. Figs 26-28): length c. 5,5 mm, head rounded, pale
yellow, with erect colourless setae on vertex, labrum and bases of antennae; imago visible
through pupal pelt. Eyes large, dark. Antennae slender, about 1J times body length, ends
forming a single coil. Labrum subtriangular, a tuft of setae on each side and a small median
tuft near apex. Mandibles brown, slender, sharply pointed, right mandible with four teeth, left
with five (two of them bicuspid), cutting edged serrated between teeth, each mandible with
two tufts of 4-5 long setae on ventral side, a single small seta on outer margin, inner basal area
with large patch of spinules. Maxillary and labial palpi as in adult. Forelegs fairly slender, mid-

330

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

and hind legs sturdier but not noticeably expanded, mid-legs with swimming fringe. Spurs 2.4.4.
Tufted gills, conical lateral gills and dorsal plates present on abdomen as indicated in Table IV.
Anal appendages fairly short, stout, almost straight, apices slightly curved dorsal, the tip of
each appendage forming two sharp points between which is a shallow saddle covered with
small protuberances; each appendage fringed with long, dark setae along outer margin, ventral
side with apical patch of shorter, colourless setae, and spinules along outer edge. Genitalia in
lobed pockets with light shagreening (raised points) on surface.

Female pupa (in spirit): length c. 6 mm; much as male, but antennae slightly shorter than
body, straight; mid-legs expanded, with swimming fringe. Anal appendages almost straight,
with fewer setae than in <?, apices less symmetrical, the outer points being shorter than the in-
ner. Genitalia not in special pockets.

Pupal case (Fig. 29): oval, about 8 mm long, rather roughly constructed of large sand
grains or rock fragments with small ones filling interstices, not usually very rigid, completely
lined with silk except for the rounded ends, which are partly closed with sand grains joined
within by bands of silk forming a course network, allowing water to circulate; sides basally in-
turned and affixed to rock or stone, basal opening closed with silk lining only. Larval sclerites
packed into posterior end of case.

Remarks

Males and females in the Albany Museum collection are typical of the species, and cleared
genitalia of both sexes have been compared with material identified by Mr D. E. Kimmins,
who himself identified some of these specimens. The larvae and pupae have not, to my knowl-
edge, previously been described, excepting that Barnard (1934: 63, fig. 38c) evidently had a
mixed series of at least three species among the larvae he drew, because his fig. 38c, labelled
“variations in the front margin of clypeus of Hydropsychodes lateralis ”, in fact shows what may
be a variation of Cheumatopsyche afra (top drawing), C. thomasseti (middle drawing) and
C. maculata ( = C. lateralis ), bottom drawing. The larva fully described (frontoclypeal apotome
fig. 38b) appears to have been that of C. afra.

Ulmer (1957) discussed the larvae of the then known Flydropsychinae, including inter alia
Cheumatopsyche and Hydropsychodes (now in synonymy), and providing a tentative key. Ul-
mer concluded from the evidence in his possession that Cheumatopsyche species had four
metasternal and five anal gills, whereas Hydropsychodes species had three metasternal and
four anal gills. As has been noted above, C. thomasseti has three metasternal gills and four
anal gills, as has C. afra (described by Barnard as C. lateralis in 1934), though Marlier (1961)
in describing afra larvae mentions five anal gills (see comment below). C. tenerrima on the
other hand has three metasternal gills but appears to have five anal gills (Marlier 1961). C.
copiosa , the larva of which was described by Hickin (1956) also has five anal gills (number of
metasternal gills not given) and C. maculata (material collected by the author) definitely has
three metasternal and five anal gills. Of the uncorrelated species of Cheumatopsyche described
by Marlier, one (Marlier 1943a) has three metasternal gills (anal gills not mentioned); the
other. Marker’s Larva B (1943b), is described as having three metasternal gills in the younger
instars, four in mature larvae (anal gills not given). As regards Larva B, Ulmer suggests, I
think with reason (there are also differences in the frontoclypeal apotome), that two different
species were included in Marker’s material. Indeed from a careful consideration of Marker’s
descriptions and figures, it appears to me probable that “Larva B’’ included two species of
Cheumatopsyche (the small and medium-sized larvae respectively) and one of Hydropsyche
(the large one); and that his “aged larva” seems to have been a specimen of the Cheumatop-
syche sp. larva described in his 1943b paper. His “Hydropsychinae sp. C” larva (Marlier
1943c), sounds very like the plain brown Hydropsyche larvae mentioned under H. longifurca

331

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8. JUNE 1983

above. The larva described by Jacquemart (1957) as C. cf. leloupi also has three metasternal
gills.

Of the Cheumatopsyche larvae so far described from Africa, C. tenerrima (Marlier 1961) is
the only species which might be confused with C. thomasseti as the larvae have the same al-
most straight crenate margin to the frontoclypeal apotome on which there is also a light patch.
In C. tenerrima this light central patch broadens posteriorly to form a pale transverse band
across the head behind the eyes whereas in C. thomasseti there is a dark transverse band in
that position. There are also differences in shape of mandibles and the ventral plates on IX;
these are clearly shown in the drawings. Gibbs (1973) describes two species (sp. A & sp. D,
fig. 109A) with similarly shaped frontoclypeal apotomes, but both lack the central light square
and show other differences from C. thomasseti.

It thus appears that in those Afrotropical species of Cheumatopsyche , the larvae of which
have been described so far, the larvae have three metasternal gills and either four or five anal
gills. Ulmer’s Sunda Island species of Hy dropsy chodes (Ulmer 1957) also had three metaster-
nal gills. His ascription of four to Cheumatopsyche was based on the description of C. lepida
(Pictet), originally described as Hydropsyche , in which there are four metasternal gills (Silfve-
nius 1905) and his retention of the genus Hydropsy chodes. As, however, Hydropsychodes has
since been placed in synonymy with Cheumatopsyche (Kimmins 1963) and C. lepida is the type
species of the genus and widespread in Europe and down into North Africa, it is advisable to
regard Cheumatopsyche larvae as possessing two mesosternal gills and either three or four me-
tasternal gills and either four or five anal gills. There would seem, nevertheless, to be stability
in this count within a particular species; certainly in C. thomasseti larvae there appears to be no
variation in thoracic gill numbers between the different instars, there being two mesosternal
and three metasternal gills from the time of their first appearance in the second instar to the
final instar.

The most useful characters for distinguishing between larvae of different Cheumatopsyche
species in the Afrotropical Region appear to be the shape of the anterior margin of the fronto-
clypeal apotome and the colour pattern of the head, both of which seem to be reasonably sta-
ble characters, though there may be minor variations in the intensity of the colouration and in
the shape of the margin. No doubt anal and metathoracic gill counts, setation, shape of man-
dibles and other characters can be used as additional criteria. Statzner (1981), in his study of
the Hydropsychidae of the N’Zi River (Ivory Coast), concluded that the most useful characters
for recognition of early instars of Cheumatopsyche species were the stridulatory ridges and sec-
ondary setae on the head; in later instars the size of the prosternal sclerites, the shape of the
frontoclypeal apotome margin, shape of submentum and pre-episternum and the secondary
setae on the head.

There are, in the Albany Museum collection, larvae representing at least a dozen species,
of which, to date, three only have been correlated with their imagos, namely C. thomasseti ,
C. afra and C. maculata (unpublished data). Gibbs’ “sp. A” is near South African C. afra lar-
vae (Gibbs 1973), as of course is Marlier’s C. afra larva from Kivu (Marlier 1961), which, how-
ever, differs from South African material in several respects. It is very possible that C. afra
may eventually prove to be a complex of species as Kimmins (1960b) suggested might be the
case.

In the same paper Marlier (1961) proposed the subdivision of the genus Cheumatopsyche
into four subgenera, Cheumatopsy chodes, Achirocentra, Abacarioides and Cheumatopsyche
sensu stricto. He gave some discussion as to their relationships and differences and listed the
species that he attributed to each. As I do not find the distinguishing characters very easy to
use, I am not attempting to key out the proposed subgenera here. In the list of African species
which follows, however, Marlier’s placement into subgenera is indicated, and reference should
be made to his paper if details are required.

332

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS
Biology

Cheumatopsyche larvae are considerably smaller than those of Hydropsyche , but are simi-
lar in habits, also being predacious omnivores and building shelters with attached catching nets
in which algae and smaller animals are trapped and from which they are collected and eaten.
The catch taken includes small crustaceans, such as Copepoda and Cladocera, aquatic insect
larvae, such as Simulium larvae and Ephemeroptera nymphs (Chutter 1968, 1971; Hynes &
Williams 1962, and personal observations), and a higher proportion of algae than in the case of
Hydropsyche. Cheumatopsyche larvae evidently prefer weaker currents to those favoured by
Hydropsyche and build larger, flimsier nets which collapse on removal from water.

Much is known regarding the life cycle and habits of Cheumatopsyche species from work
done in various parts of the world. Fremling (1960) in the study of Hydropsyche orris at Keo-
kuk quoted above, also gave a detailed account of the almost equally common C. campyla. He
noted that the larvae abounded in tailwaters and other areas of moderate current; that the
adults sheltered during the day and swarming was not seen, though mating couples were col-
lected; that females entered the water to oviposit, laying on submerged objects at depths of up
to 12 ft (c. 360 cm), often re-entering the water on several occasions. In the case of eggs laid in
the laboratory it took 51 days from hatching through five larval instars and pupation to the
emergence of the adults. The pupal stage lasted for six days. In the river the larvae were found
at depths of up to 24 ft (c. 720 cm), preferring deeper to shallower water. Only first instar lar-
vae appeared to swim. Older larvae seemed to be continually active, feeding, cleaning nets and
discarding inedible objects. New areas were populated by drifting of young larvae, this being
compensated for by upstream flight of females. C. campyla appeared to be bivoltine, with
peaks in early and late summer.

Cheumatopsyche thomasseti is a eurytopic species, widespread in Southern Africa, and is
found in many types of flowing water. This species shows a wide range of tolerance to climatic,
physical and chemical conditions, much more so than Hydropsyche and more so than some of
the other species of Cheumatopsyche found in Southern Africa, though the genus as a whole is
more tolerant of mild pollution of various kinds than are most other Trichoptera.

In the Berg River (West Cape) C. thomasseti larvae commonly occurred in the lower
reaches during the summer months. Another species, C. afra (Mosely), was characteristic of
the upper reaches, although in summer this species was present in small numbers, together
with thomasseti , in the lengthy lowest zone. The main habitat of thomasseti thus lay in a stretch
of river running through farmlands, where stony runs and stickles alternated with large, sandy,
shallow pools, and where turbidity, dissolved solids, temperature and pH were higher than in
the upper reaches. The whole river was subjected to periodic flooding during the winter rains.
A few comparative values are shown below, taking Station 3 (a fast-flowing foothill stream) as
being one at which afra was well represented, and Station 18 (on the lower Berg) for thomasse-
ti, the only station where it was abundant (figures extracted from Harrison & Elsworth 1958).

Station

Temperatures ° C

pH

TDS ppm

Turbidity ppm S1O2

3: C. afra

winter min.
9,0

summer max.
29,0

4,7-6, 8

19-78

0, 0-6,0

18: thomasseti

10,0

31,7

6,4-8, 2

45-584

0,0-13 000

In the East Cape the rivers may be broadly divided into two categories; highlying, often
acid mountain streams; more or less highly mineralized, silty lowland rivers, subject to alter-

333

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

nate droughts and floods. The Cheumatopsyche species in the area show a clear distribution
pattern, as might be anticipated, though perhaps not one that would have been expected.

In the upland streams on the Hogsback mountains (Amatola Range near Alice, altitude
1 200-1 500 m), the overwhelming preponderance of Cheumatopsyche larvae were a dark form
of C. afra, C. thomasseti larvae being very rarely found. In the lower altitude (800-900 m)
montane streams in the vicinity of Grahamstown, C. afra occurs, though rarely, unless two
other possibly new species found prove to be variants of it. At present 1 tend to the opinion
that they are different species as the larvae also appear to show consistent differences. The
commonest of the local species is, however, C. maculata (Mosely). All these streams are clear,
cold and clean, those on the Amatolas being permanent, whereas those near Grahamstown
may be reduced to a trickle or to subsurface moisture during severe droughts.

Temperatures on the Hogsback ranged from snow cover at times in winter to 15-17° C
(water temperature) in summer, and pH values ranged from 6,4 in summer to 7,5 in winter,
the change to slight alkalinity being due to run-off during the rainy season, the rock formations
being mainly basaltic. (Values from B. C. Wilmot, pers. com.) Highlying streams near Gra-
hamstown drain Witteberg quartzites and remain acid throughout the year, pH values ranging
from 6, 0-6,9.

Two lowland rivers were studied in the course of other work (Scott 1974, Scott, Allanson
& Chutter 1972), namely the Sundays River, which rises in the Sneeuberg mountains north of
Graaff-Reinet and flows to the sea 35 km NE of Port Elizabeth, and the Great Fish River,
which lies to the east of the Sundays, drains several mountainous areas and reaches the sea
some 130 km to the east of Port Elizabeth. Both rivers lie largely in areas of rain shadow and
flow over soft, easily eroded geological formations (mainly Beaufort and Ecca Series of the
Karroo System, though the lower Sundays flows over cretaceous rocks of marine origin). Both
rivers are liable to dry out during droughts and both are subject to catastrophic floods. Both
had small resident Trichoptera populations limited to those parts which did not dry out com-
pletely. In the Sundays River the Hydropsychidae were represented by C. afra and in the Fish
by C. thomasseti.

A brief summary of relevant water conditions encountered in the two is given below, ex-
tracted from Scott et al. (1972), including only values at which living, apparently healthy,
Cheumatopsyche larvae were found:

pH

TDS mg/1

Ca mg/1

Na mg/1

Fish River

C. thomasseti

8,2-9, 0

511-3 368

27,7-115

110-830

Sundays River

C. afra

8, 2-9, 3

998-4 349

56

1 480

Such temperatures as were taken varied from 11,5-27° C. However, these do not really
reflect the range obtaining, for air temperatures varied with the seasons from winter frosts to at
least 46° C in summer and, as the water is often shallow, it tends to be directly affected. In ex-
treme heat larvae could, however, seek shelter under stones and imagos under bridges or vega-
tation. The Hydropsychid larvae also had other factors with which to contend, such as slow
current, low water levels and heavy silt load during floods.

Calcium values have been included because the C. thomasseti larvae from most stations on
the Fish River were more or less heavily encrusted with a white coating of CaCCb, often with
head completely covered and gills matted with calcareous material. This could have been due
to a high load of calcium bicarbonate which can be precipitated as CaCCb on the bottom of the

334

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

river and evidently also on some of the animals. The larvae when caught, were alive, but must
have been seriously hampered in their activities (Scott 1976).

The difference in Cheumatopsyche species between the two rivers may have been due to
the very heavy silt load in the Fish River during floods, considerably heavier than in the Sun-
days, as it has been generally observed that C. afra usually occurs in larger numbers in the
eroding zone, where the water is clean and silt-free whereas thomasseti is more abundant in
lower, more turbid waters (Harrison & Elsworth 1958, Chutter 1969a, Scott et al. 1972). Since
the above study was completed conditions may have changed considerably with the inflow of
water from the Orange to the upper Fish River, via the Orange-Fish tunnel.

River studies in the Transvaal and Orange Free State (Chutter 1968, 1971) included obser-
vations on C. thomasseti , which was one of the commonest animals in stony runs in most parts
of the Vaal River and its tributaries, particularly in areas of mild pollution. For example, such
mild pollution, due to seepage of treated sewage effluent, occurred in the Vaal River below
Standerton, resulting in increased particulate organic matter and algae (mainly diatoms), large
numbers of Cladocera and Copepoda and, evidently feeding on all these, large numbers of
C. thomasseti and Amphipsyche scottae. Chutter’s work showed that C. thomasseti larvae were
present all the year round, though most numerous in spring and autumn, and that the imagos
were abundant in summer. Very few early instars were found in winter samples. Density of lar-
vae was significantly correlated with current speed in certain areas, increasing sharply at speeds
above 50 cm/sec.

Statzner’s studies of emerging insects from the River Kalengo (Zaire) (Statzner 1976) also
showed an all-year-round pattern but with two peaks for C. explicanda imagos and four smaller
peaks for C. boettgeri. Statzner concluded that the main controlling factor could be light inten-
sity. The Zairese climate is tropical.

Sharp (1976) made some laboratory observations on C. maculata larvae from hill streams
near Grahamstown where they inhabit small falls and stony runs. In the latter situation they
were usually under stones. Optimum conditions appeared to be a water temperature of about
10° C and a relatively high PO2 level, indicating a high oxygen tension. Undulatory (respirato-
ry) movements of the abdomen which increased both with rising temperature and with falling
PO2 levels did not occur at all in fast current when temperature and oxygen levels were favour-
able. If the temperature was increased too quickly or too much (25-30° C) or the PO2 level
dropped too low (less than 55 mg Hg) the larvae left their nets and attempted to escape, swim-
ming with spiralling movements of the abdomen. Intraspecific fighting did sometimes occur but
was not serious and did not result in injuries or death. Unlike Hydropsyche , C. maculata larvae
spun nets at very low current speeds, and indeed did so in the laboratory even in still water
provided that the substratum was suitable. In the small waterfalls they spun nets in the usual
way, on open rock faces in the stream, usually in small crevices in the rock where there was
some shelter (personal observations).

Gibbs (1973), in observations on Ghanaian species, noted that at times C. digitata larvae
occurred in vast numbers (as many as 250 000/m2 in a stream at Tafo), completely covering the
bedrock with a dense incrustation of retreats, inhabited by fourth and fifth instar larvae, while
younger larvae moved freely around them. Beneath these larvae retreats were pupal cases and
beneath those older broken down cases occupied by chironomid larvae and worms. Gibbs
noted that Cheumatopsyche larvae could be separated to species by their head markings and
differences in secondary setation.

Statzner’s study (1981) of hydropsychid larvae in the N’Zi River, a running water habitat
with intermittent flow, included three species of Cheumatopsyche. The survey period of 15
months enabled him to study densities, life cycles and their relationship to discharge patterns.
Parts of the N’Zi were treated with chlorphoxim to control Simuliidae, and Statzner concluded

335

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

inter alia that Cheumatopsyche larvae survived unfavourable periods in the larval stage and
that they increased in numbers when flow slackened. He considered that there were only two
to three generations per year.

African Species of Cheumatopsyche

C. afra (Mosely), 1935: 229-231, figs 17-20 (wings, 6 genitalia), as Hy dropsy chodes. 6 .

Mosely 1939b: 27, transfer to Cheumatopsyche.

Kimmins 1960a: 266-267 , figs 36-67 (6 genitalia), figs 72-76 (genitalia of probable 2 ).

C. lateralis (Barnard), 1934, partim (as Hydropsychodes ), is a synonym.

Marlier 1961: 162-164, figs 1-3 (<J genitalia, larva), placed it in his new subgenus
Cheumatopsychodes , and (1962a) proposed it as subgenus-type of this taxon.
Distribution: Sierra Leone, Tanganyika, Northern and Southern Rhodesia, South Africa,
Belgian Congo, Ethiopia, Angola and possibly Ghana. It is thus widespread in
Africa from the West Cape in the south to Ethiopia in the north-east and Sierra-
Leone in the north-west, but may prove to be a complex of closely related species
(Kimmins 1960a).

C. albomaculata (Ulmer), 1905a: 34, 35, fig. 22 (wings), as Hydropsychodes. 6 2 .

Marlier 1957: 289-292, figs 7, 8 (wings, head & thorax of 2 ), as Hydropsychodes. 2 .
Kimmins 1963: 132-134, figs 36-39 (wings, 6 genitalia), as Cheumatopsyche.

Distribution: Belgian Congo, Katanga, Lakes Edward & Albert, Angola, Ghana, Ethio-
pia.

C. alfierii (Navas), 1927: 212-213, as Hydropsychodes. 2 .

Distribution: Egypt.

C. amboinica (Navas), 1928b: 112, 113, fig. 9 (c? genitalia, labelled C. ambonica), as Hydro-
psychodes. 6 .

Distribution: British East Africa (Amboni).

C. anema Marlier, 1959: 111-114, figs 4, 5 (wings, 6 genitalia; 5c, not labelled as such,
appears to be 2 genitalia). 6 2.

Distribution: Sao Tome Island, Gulf of Guinea.

C. apicata (Navas), 1931a: 126, 127, as Hydropsychodes. 6 .

Distribution: Belgian Congo.

C. aterrima Marlier, 1961: 159, 172-174, figs 11A-E (wings, 6 genitalia, palps). 6. Marlier
places this in his new subgenus Achirocentra in the same paper.

Distribution: Kivu.

C. atlantis (Navas), 1930b: 142-143, fig. 48 (hind wing), as Hydropsychodes. 2 .

Distribution: N. Africa, Grand Atlas Mountains.

C. bifida Statzner, 1975: 382, 383, figs 13-17, 31 (<3 genitalia & wings), figs 21-23 (2
genitalia). 6 2.

Distribution: Zaire.

C. bimaculata (Ulmer), 1930: 491-493, fig. 15 (wings), as Hydropsychodes. 2 .

Mosely 1939b: 27, transfers it to Cheumatopsyche.

Kimmins 1963: 132, fig. 35 (2 genitalia).

Distribution: Abyssinia.

C. bimaculata (Jacquemart) 1966b: 45, fig. 10 (wings), as Hydropsychodes. 6 .

Distribution: Congo.

336

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Note: C. bimaculata Ulmer was described from the 9 only; if this is regarded as the 3 of
his species it should be attributed to Ulmer, if not, it would have to be renamed, as
Hydropsychodes has been placed in synonymy with Cheumatopsyche (Kimmins
1963). Jacquemart gives no drawings of the genitalia.

C. boettgeri Statzner, 1975: 382, 383, 388-397, figs 18-20, 32 (3 genitalia & wings), figs 24-26
(9 genitalia). 3 9 .

Distribution: Zaire.

C. brunnea Jacquemart, 1961d: 228. 3 .

Jacquemart 1961b: 23-25, fig. 14 (wing & 3 genitalia).

Distribution: Congo.

C. brunnea (Jacquemart), 1963a: 378-379, fig 33A-D (wings & 3 genitalia), as Hydropsy-
chodes. 3 .

Distribution: South Africa (West and East Cape).

Note: As Hydropsychodes is now in synonymy with Cheumatopsyche, this species should
be renamed.

C. burgeonia (Navas), 1931b: 138-139, fig. 74 ( 9 hind wing), as Hydropsychodes. 9 .
Distribution: Belgian Congo.

C. comorina (Navas), 1931a: 125-126, fig. 63 (hind wing), as Ulmeria. 3 9.

Distribution: Comoro Islands.

C. copiosa Kimmins, 1956: 196-197, figs 1-4 (d, 9 genitalia). 3 9 .

Hickin 1956b: 132-133, figs 1-7 (larva), description of larva.

Distribution: Uganda, Gold Coast, Dakar.

C. digitata (Mosely), 1935: 228-229, figs 14-16 (wing & 3 genitalia), as Hydropsychodes. 3 .
Gibbs 1973: 393, 397, fig. 109a (larval head cf. with other species). Brief description of
larval head and habitat.

Distribution: Belgian Congo (not Tanganyika, Mosely 1936), Uganda, Ghana, Nigeria.
Cape Verde Islands (Nybom 1960: 3, 3 9).

‘C. digitifera Mosely’, Jacquemart 1966b, is presumably a mis-spelling of C. digitata.

C. diminuta (Walker), 1852: 115, as Hydropsyche.

Ulmer 1931: 16-18, figs 13-15 (wing & 3 genitalia), as Hydropsychodes. 6 .

Betten & Mosely 1940: 187-188, fig. 92 (9 wings), as Cheumatopsyche. Original descrip-
tion plus further description.

Marlier 1961: 160, 161, proposed this as subgenus-type of his subgenus Cheumatopsy-
chodes, but later withdrew it and substituted C. afra (1962a).

Distribution: Sierra Leone, Congo, German East Africa, South Africa, Cameroons, Su-
dan, Egypt.

C. explicanda Statzner, 1975: 382-387, 394, figs 1-4, 9-13, 30 (d genitalia & wings), figs 5-8
(9 genitalia), fig 14 ( 6 & 9 genitalia in copula). 3 9.

Distribution: Zaire.

C.falcifera (Ulmer), 1930: 482-485, figs 4-7 (wings, 3 genitalia), as Hydropsychodes. 3 .
Barnard 1934: 360-362, figs 37h-j (3 genitalia), as Hydropsychodes zuluensis Barnard.
Mosely 1939b: 27, transfer to Cheumatopsyche.

Kimmins 1960a: 263, figs 24-27 ( 3 genitalia); synonymy of H. zuluensis Barnard with
C. falcifera (Ulmer).

AM— 6

337

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Kimmins 1963: 130-131, fig. 46 (2 genitalia), as Cheumatopsyche. 9. He notes that
C. falcifera var., Ulmer 1930, is actually C. afra (Mosely).

Corbet 1958: 208-210, fig. 3a-f (larval parts & pupal case).

Distribution: Abyssinia, South Africa, former Belgian Congo, Uganda, Ghana.

C. kimminsi Marlier, 1961: 159, 165-168, figs 5-7 (wings & d genitalia). 8 2 .

Marlier places it in his new subgenus Cheumatopsychodes in the same paper.

Distribution: Kivu.

C. kissi Marlier, 1961: 159, 170-172, figs 9-10 (wings, 8 genitalia & palp). 8 .

Placed in Marlier's new subgenus Cheumatopsychodes in the same paper.

Distribution: Kivu.

C. kitutuensis Marlier, 1962a: 3, fig. 1 (9 maxillary palps). 2.

Marlier uses it as subgenus-type of his new subgenus Hydrocheumatopsyche. Originally
described in Marlier 1961: 186-188, figs 16, 17A (wing & leg) as Cheumatopsyche sp.
2.

Distribution: Kivu.

C. lateralis (Barnard), 1934: 362-364, as Hydropsychodes. It proved to be a synonym of C. ma-
culata (Mosely): part of his material was, however, C. afra (Mosely).

C. leleupi Marlier, 1961: 159, 177-180, figs 13a-c (wings, 8 genitalia). 8.

Placed in Marlier’s subgenus Abacarioides in the same paper.

Distribution: Kivu.

C. leloupi Jacquemart, 1957: 94-95, figs 78-83 (d genitalia & wings), pp. 95-100, figs 61a, 63,
75 (parts of probable larva). 8 .

Marlier, 1961, suggests that leloupi may be synonymous with C. digitata (Mosely), to
which it is certainly very similar.

Distribution: Lakes Kivu, Edward and Albert (Belgian Congo).

C. lepida (Pictet), 1834: 207, pi. 18f, also larva & case, as Hydropsyche.

Fischer 1963: 114-122, as Cheumatopsyche.

Distribution: Europe, North Africa (Algeria).

C. lesnei (Mosely), 1932: 4, pi. 1 & figs 1-3 (wings, 8 genitalia), as Hydropsychodes. 8 2 .
Mosely 1939b: 27, transfer to Cheumatopsyche.

Kimmins 1957b: 11-12, fig. 7 (d & 2 genitalia), as C. uncata sp. n. in error; he discovered
it was synonymous with lesnei , appending a note to that effect in the same publica-
tion.

Distribution: Mozambique, Uganda, Dakar, Rhodesia, as well as probable material from
Natal.

C. lestoni Gibbs, 1973: 395, figs 99-101 (8 genitalia). 8 .

Distribution: Ghana.

C. lobata Marlier, 1943a: 10, 11, fig. 6 (d genitalia). 8 2 .

Distribution: Belgian Congo.

C. maculata (Mosely), 1934: 22-24, figs 11-13 (d genitalia), as Hydropsychodes. 8 .

Mosely 1939b: 27, transfer to Cheumatopsyche.

C. lateralis (Barnard) 1934: 362 falls into synonymy with maculata , his drawings were,
however, of afra (Kimmins 1960a).

Distribution: South Africa (West and East Cape).

338

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

C. madagassa (Navas), 1923: 27, fig. 38 (d hind wing), as Hydropsychodes. 3 .

Distribution: Madagascar.

C. marieni (Jacquemart), 1965: 1-3, figs 1, 2 (wings & d genitalia), as Hydropsychodes. 3 .

Distribution: South Africa (Natal).

C. meruana Navas, 1934d: 21-22, fig. 107 (hind wing). 9.

Kimmins 1960a, notes that the 3 is unknown and regards it as unidentifiable at present.
Distribution: Kenya.

C. nubila Kimmins, 1963: 137-138, figs 47-50 (3, 9 genitalia). 3 2.

Distribution: Ethiopia, South Africa (Natal).

C. obscurata (Ulmer), 1930: 485-488, tigs 8-10 (wings, 3 genitalia), as Hydropsychodes. 3 .
Mosely 1939b: 27, transfer to Cheumatopsyche.

Kimmins 1963: 130, fig. 45 (2 with description).

Distribution: Abyssinia.

C. obtusa (Jacquemart), 1963a: 380, fig. 35a-c (wings, 3 genitalia), as Hydropsychodes. 3 .

Distribution: South Africa (Orange River, Upington).

C. pallida (Banks), 1920: 357-358, pi. 7, fig. 89, as Hydropsychodes. 3 .

Distribution: Madagascar.

C. plutonis (Banks), 1913: 239, pi. 23, figs 2, 4, 5 (d & 9 genitalia, wings), as Symphitopsyche.

3 2.

Mosely 1939b: 27, transferred it to Cheumatopsyche.

Distribution: Abyssinia.

C. pulverulenta Gibbs, 1973: 393-395, figs 95-98 (d genitalia & forewing). 3 2 .

Distribution: Ghana.

C. punctata (Jacquemart), 1961d: 228, also 1961b: 25-26, fig. 15 (forewing & 3 genitalia), as
Hydropsychodes. 3.

Distribution: Congo.

C. rhodesiana (Jacquemart), 1963a: 375-378, figs 32A-D (wing & 3 genitalia), as Hydropsy-
chodes. 3

Distribution: South Africa (East Cape, Natal), Basutoland, Southern Rhodesia, Katanga.

C. roscida (Navas), 1934c: 71, 72, fig. 13 (forewing), as Hydropsychodes. 3 2 .

Distribution: Madagascar.

C. sessilis Marlier, 1961: 159, 168-170, figs 8A-E (wings, 3 & 2 genitalia), placed in his new
subgenus Cheumatopsychodes in this paper. 3 2.

Distribution: Kivu.

C. sexfasciata (Ulmer), 1904b: 421, figs 10, 11, 12 (wings, 3 genitalia), as Hydropsyche. 3 2 .
Kimmins 1963: 131, figs 30-34 (wings, 3 & 9 genitalia); transfer to Cheumatopsyche.
Marlier 1961: 159, 175-177, figs 12A, B (wings, 3 genitalia), places it in his new subgenus
Abacarioides , which he describes here.

Gibbs 1973: C. striata Jacquemart is doubtfully distinct.

Distribution: Cameroons, Ethiopia, Kivu, Belgian Congo, Angola, Ghana, Gabon.

C. simplex Kimmins, 1963: 134-135, figs 40-44 (wings, 3 & 2 genitalia). 3 2 .

Distribution: Ethiopia.

C. socia (Navas), 1927: 213, fig. 9 (d genitalia), as Hydropsychodes. 3 2 .

Ulmer 1963: 258, 260, figs 4, 5 (d genitalia), description and transfer to Cheumatopsyche.
Distribution: Egypt.

339

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

C. spinifera (Jacquemart), 1966b: 43, 44, figs 8, 9 (wings, 3 genitalia), as Hydropsychodes. 3 .
Distribution: Congo.

C. striata (Jacquemart), 1959: 125, figs 39-43 (wings & 3 genitalia), as Hydropsychodes. 3 .
Gibbs 1973: doubtfully distinct from C. sexfasciata (Ulmer).

Distribution: Lake Albert.

C. tectifera Marker, 1959: 115-117, figs 6-8 (wings, 3 & 2 genitalia). 3 9.

Distribution: Sao Tome Island, Gulf of Guinea.

C. tenerrima Marker, 1961: 159, 180-186, figs 14, 15 (wings, 3 genitalia, larval & pupal parts),
larva and pupa described. Marker left this species in Cheumatopsyche sensu stricto in this
paper, but in his 1962b paper placed it in his subgenus Ethiopsyche. 3 .

Distribution: Kivu.

C. thaba Mosely, 1948: 70-71, figs 9, 10 {3 genitalia). 3 .

Distribution: Yemen (South-west Arabia is regarded as a part of the Afrotropical Re-
gion).

C. thomasseti (Ulmer). 1931: 18, figs 16, 17 ( 3 genitalia), as Hydropsychodes. 3 9 .

Kimmins 1960a: 263-265, figs 28-31 (d genitalia), fig. 71 (9 genitalia); fig. 70, which he
gave as ? thomasseti , he later transferred to C. nubila Kimmins (1963).

Distribution: South Africa (all Provinces), Angola, Rhodesia, South-west Africa, prob-
ably Kenya (larvae).

C. triangularis (Ulmer), 1931: 19, figs 18, 19 (d genitalia), as Hydropsychodes. 3.

Kimmins 1960a: 275, figs 1-4 (3 genitalia), figs 68, 69 (Probable 9 genitalia).

Distribution: South Africa (Natal, Transvaal).

C. trifida (Mosely), 1935: 228, figs 11-13 (wing & 3 genitalia), as Hydropsychodes. 3 9 .

Mosely 1939b: 27, transferred it to Cheumatopsyche.

Kimmins 1960a: 257-259, figs 5, 6 (<3 genitalia).

Distribution: Belgian Congo, Southern Rhodesia, Uganda.

C. uncata Kimmins 1957b: synonym of C. lesnei (Mosely), q.v.

C. urema Mosely, 1936: 436-437, figs 15-17 (<3 wings & genitalia).

Kimmins 1960a: 257, 265, figs 32-35, stated that urema might be a synonym of thomasseti
(Ulmer).

Marker 1961: 159 et seq., 165, fig. 4 (3 genitalia) placed it in his newly erected subgenus
Cheumatopsychodes.

Distribution: Kenya, Uganda, Kivu.

C. villosa (Navas), 1934d: 22-23, fig. 108 ( 9 posterior wing), as Hydropsychodes. 9 .

Mosely 1935 commented that the 9 hind wing hardly suggested the genus Hydropsy-
chodes, with which I would agree.

Distribution: Africa.

C. zuluensis (Barnard), 1934: 360-362, fig. 37h— j . as Hydropsychodes. 3 9 .

Kimmins 1960a: 263, placed zuluensis in synonymy with C. falcifera (Ulmer).

340

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Genus HYDROMANICUS Brauer 1865

Hydromanicus Brauer 1865: 420.

Type species irroratus Brauer. 9 . Java.

Hydromanicus Brauer; Ulmer 1907a: 168 (key to genera), 172, 173 (characters & species list);
pi. 23, fig. 211 (wings).

Hydromanicus Brauer; Ulmer 1951: 226, 227, 275-281 (imagos).

Hydromanicus Brauer; Ulmer 1957: 388, 390 (larval & pupal keys).

Hydromanicus Brauer; Fischer 1963: 93, 94; 1972: 127.

Generic diagnosis (derived from Ulmer 1907a: 172 & 1951: 279):

In general this genus is very similar to Hydropsyche.

Imago: tibial spurs of 3 , 9 , 2.4.4, inner spurs longer than outer, tarsal claws usually smaller in
3 than in 2. Maxillary palpi with fourth segment clearly longer than third, fifth articulated.
Head with large median wart and pair of large lateral warts; antennae little longer than fore
wings in <3, shorter in 9 , without spiral black line. Wings broad; forewing with apical forks, 1,
2, 3, 4, 5, discoidal and median cells closed; hind wing with apical forks 1, 2, 3, 5 (1 may be
diminutive), discoidal cell closed, median cell open, M and Cm well separated, with a clear
crossvein between them. 3 with large superior appendages and apically cleft tenth tergite;
claspers (inferior appendages) two-jointed; copulatory organ prominent, large (lacking the
long ventral endothecal lobe characterising Afrotropical species of Hydropsyche). 2 mid-legs
not widened.

The genus is not a very large one, mainly represented in India, Indonesia, the Far East
and Borneo, with a single species in the Seychelles, hence its inclusion here.

Hydromanicus seychellensis Ulmer

Hydromanicus seychellensis Ulmer, 1910: 47-49, figs 11-14 (wings & 3 genitalia). Mahe,
Seychelles Islands. 3 2 .

Syntypes comprised 10 33, 5 2 9, divided between the Museums in Cambridge and
London and Ulmer’s private collection.

Hydromanicus seychellensis Ulmer; Marker 1978b: 34-38, pi. 2, figs 12-27 (larva & pupa).
Distribution: endemic to the Seychelles Islands.

Remarks

Ulmer’s description of this species accords with his generic diagnosis, except that he
neither draws nor mentions superior appendages in the 3 genitalia, and although M and Cui
are clearly separate in the hind wing, they do lie close together. The 3 genitalia are, however,
very easily recognizable on account of the distinctive apical joints of the claspers, sharply
pointed, blackened, and thickly clothed with dark setae. Length of 3 forewing 7-7,8 mm, 2
8-10 mm.

H. seychellensis appears to be the sole Hydropsychid found on the Seychelles Islands,
where Marker (1978b) found it to be widespread and the commonest Trichoptera species. He
found the larvae and pupae in all streams on the Island of Mahe, even slightly polluted ones,
and also on the Island of Praslin. Marker, who reported on the results of collections made both
by himself, in 1976, and by Professor F. Starmiihlner of the University of Vienna in 1974, re-
corded that the larvae were found under similar conditions to European species of Hydro-
psyche, also constructing wide-mouthed catching nets and building rigid pupal cases of pebbles
and sand.

341

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8. JUNE 1983

Distinguishing characters of imagos, larvae and pupae of this species as described in the
literature have been used to separate Hydromanicus from other genera in the keys, together
with generic characters used by Ulmer (1951). The keys are not, therefore, necessarily applic-
able to species of Hydromanicus from other parts of the world.

The larvae and pupae of Hydromanicus appear to be remarkably similar to those of Hy-
dropsyche■, though there are minor differences such as the proportions of the dorsal and ventral
branches of the pre-episternum, number of metasternal gills and type of setation in the larva.
The anal appendages of the pupa are simpler than those of Hydropsyche, lacking the two
points seen there; presegmental dorsal plates on III— VIII ; tracheal gills on II— VII, and c? an-
tennae wound once round apex of abdomen. Length of Hydromanicus larva up to 15 mm,
length of pupal exuviae 13 mm.

Reference should be made to Ulmer (1910) for description and figures of 8 imagos, and
to Marker (1978b) for illustrated accounts of larva and pupa.

African species of Hydromanicus

H. seychellensis Ulmer, 1910: 47-49, figs 11-14 (wings & 8 genitalia).

Mahe, Seychelles Islands. 8 2 .

Marlier 1978b: 34-38, pi. 2, figs 12-27 (larva & pupa).

Distribution: Seychelles Islands only.

Subfamily DIPLECTRONINAE Ulmer 1951

Genus SCIADORUS Barnard 1934

Sciadorus Barnard 1934: 358 (key), 364 (diagnosis) (also metam.)

Type species acutus Barnard (original designation).

South Africa (Cape Province). 8 9 larva, pupa.

Sciadorus Barnard; Scott 1975: 47, figs 8-13, 31, 32 (larva).

Sciadorus Barnard; Fischer 1963: 151, 1972: 154.

Generic diagnosis (amplified from Barnard 1934: 364):

Imago: tibial spurs of 8 , 2 , 2.4.4; anterior tarsal claws of 8 normal. Maxillary and labial palpi
relatively small, slender, the former with 1st segment short, 2nd, 3rd, 4th subequal, longer
than 1st, 5th longest, articulated, the latter 3-segmented, 3rd segment longest, articulated.
Head: vertex with very large paired posterior warts, anterior warts narrow strips, median wart
inconspicuous, mid-cranial sulcus present, may be incomplete; eyes noticeably small, anteriorly
placed, sparsely pubescent; antennae with flagellum slender, about length of forewings in 8,
shorter in 2, with the serrated appearance usual in this subfamily. Forewing with apical forks
1, 2, 3, 4, 5 present, discoidal and median cells closed, forks 1 & 3 with long footstalks, Cu:
and 1A meet at arculus; 8 forewing 5-9 mm, $ slightly larger. Hind wing narrower than in Di-
plectronella , forks 2, 3, 5 present (occasionally a minute fork 1), discoidal cell closed, median
cell open. Sc and Ri entirely separate, united by cross-vein, little upturned at ends; fork 2
shortly stalked or sessile. Abdomen: 5th sternum (8 & 9) bears a pair of short, tubular ven-
trolateral filaments, each arising from an internal gland; two pairs of large, rounded, reticula-
ted sacs present in 8 only, opening on posterior margins of VI and VII and extending forward
over two or more segments. Abdomen thicker than is usual in males, presumably due to pres-
ence of these organs. 8 genitalia with tenth tergum bifid, copulatory organ large, apically trun-
cate, claspers slender, two-segmented, apical segment small. Small lateral tufts sometimes
present on pleura of II-VII. 2 genitalia with sternal plates of VIII completely separate ventral-
ly, as in Cheumatopsyche, and the usual two pairs of small papillae separated by a pair of small
cerci; mid-legs not expanded, but fringed with setae in pupa.

342

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

This small genus is endemic to the Afrotropical Region, to date comprising two South
African species, one from the West Cape, the other from the South-west and South Cape. A
third, S. sinuatus, was described by Marlier (1943a) from the former Belgian Congo, but this
should be transferred to the genus Diplectronella (see under Remarks below).

Sciadorus obtusus Barnard
(Figs 43-65, Tables V, VI)

Sciadorus obtusus Barnard, 1934: 366, fig. 39a-i (wings, 6 genitalia, detail of abdomen,
antenna etc.). Zwartberg Range, South-west Cape. 6 .

Syntypes (d) in collection of South African Museum.

Distribution: South Africa: Cape Province (Swartberg Range, various streams, forest stream
on Robinson Pass, Vet River on Garcia’s Pass near Riversdale, Storm’s River tributary and
Blaauwkrantz River, all South or South-west Cape; also recorded by Jacquemart (1963a) from
Upington on the Orange River (North Cape) and the Swartberg Pass (but see under Re-
marks).

Description of imagos (Figs 43-48)

Male imago (in spirit. Figs 43-47): (colour description from Barnard 1934: 364, 365 as his
spirit material is old and faded); a medium-sized species, forewings 7-8,5 mm long, pale
brown, veins darker, when fresh hyaline marks visible on distal cross-vein of discoidal cell, r-m
cross-vein and Mi+: proximal to it, also on distal part of Cm. Head, thorax and abdomen main-
ly brown, antennae brown, legs ochraceous. Vertex: a line of setate spots present, close behind
each narrow anterior wart (not seen in S. acutus)', mid-cranial sulcus present, incomplete. Lat-
eral filaments on V no longer than segment, may be shorter, not easily seen. Easily separable
from the other Cape species, S. acutus , by the shape of the 9th and 10th terga, larger size,
darker colour and presence of lines of setate spots on vertex.

Female imago (in spirit. Fig. 48): Barnard had no females, larvae or pupae of this species,
but all three are represented in the Albany Museum collection, which includes one good 9
from a forest stream above the Storm’s River Pass (SU 68, 25. IX. 1959); it was together with
two larvae of S. obtusus , and its size, locality and association with the larvae all indicate that it
is the female of this species. There are also good 2 pupae, evidently of the same species, from
the Vet River, which were associated with an identifiable <5 pupa, prepupal larvae and larvae
(personally collected). The female imago is larger than the male (9 forewing 10,2 mm in
length), but is very similar in general appearance, colouring and wing venation. Head and tho-
rax mainly dark brown, warts cream; abdomen mid-brown with lighter marks on dorsal side,
pleura purplish, paired filaments on V not longer than segment, inconspicuous; legs ochra-
ceous; palpi almost colourless, smaller than in 6. Genitalia: IX almost entirely contracted
within VIII, scarcely visible, though XI can be seen, with the usual small cerci and papillae;
sternal plates of VIII rounded, brown, apices with blunter, shorter dorsolateral points than in
acutus , hairy, completely separated ventrally.

Material available comprises Barnard’s spirit specimens from the Swartberg Range (sever-
al males, mainly fragmented, including cleared abdomens showing genitalia and reticulated in-
ternal sacs), on loan from the South African Museum. Also 21 larvae, 9 prepupal larvae and 8
pupae (including 4 6 pupae, one good, fully mature, three immature, damaged, four good 2
pupae, two mature, two immature) from the Vet River, Garcia’s Pass, near Riversdale (MISC
306f, g); 22 larvae from the Ruiterbos Forest Station on Robinson Pass (MISC 282), all in the
Albany Museum Collection. In addition one larva from the Blaauwkrantz Pass, and one good
2 and two larvae from a forest stream with falls above the Storm’s River Pass, collected by Dr
H. Bertrand (also in the Albany Museum Collection).

343

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Figs 43-55. DIPLECTRONINAE: Sciadorus obtusus , 3,9,6 pupa. 43. 6 : dorsal view of head and thoracic nota, 44.
6: lateral view of head and palpi, 45, 46. 6: fore and hind wings (right), 47. 6: dorsal view of genitalia, 48. 9: lateral
view of genitalia, showing process on abdomen V, 49. 6 pupa: habitus; left appendages only shown, 49a. 9 pupa: part
of abdomen VI showing gill arrangement: gills spring from distal part of pleuron. Note that lateral branchial gills arise
from base of conical gills, 49b. same: mid-leg showing swimming fringe, 50, 51. 6 pupa: left and right mandibles, 52.
same: apex of abdomen, ventral, 53. same: apex of left anal appendage further enlarged, 54. part of egg string from 2
pupal case, 55. apex of 6 pupal case showing meandrine opening and smaller lateral openings; larger surrounding stones

omitted.

344

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Description of larva (Figs 56-65, Table V)

Mature larva (in spirit, Figs 56-65, Table V): length 13-14 mm, larvae large, similar to
Hydropsyche in size, but with the characteristic Diplectronine head and thoracic nota. Sclerites
dark brown in colour, almost black in life, with cream abdomen (in spirit abdomen turns pale
purplish and sclerites fade somewhat); rough and setose in appearance.

Head (Figs 56, 57): rounded, without carina, frontoclypeal apotome very broad, widest in
centre, with deep indentations opposite eyes, behind which it is strongly expanded; ventral side
with very large anterior and posterior apotomes, together much longer than the mid-ventral ec-
dysial line; stridulatory files present, large, broadly oval, inconspicuous because same colour as
genae, lines fine, even, close together. Dorsal side of head blackish to dark brown except for
two small light patches on frontoclypeal apotome and pale areas round eyes; thickly set with
short, dark, pointed setae. Eyes small, set well forward beneath clusters of cuticular lenses.
Ventral side of head brown, mainly dark.

Mouthparts (Figs 58, 60): labrum rounded, sclerotized, dark brown, setose, frequently
largely retracted, with thick lateral brushes of setae, anteclypeus membranous. Mandibles
blackish, heavily built, strong, triangular in section, the left one with 3 apical and 4 lateral
teeth, the proximal tooth bicuspid, and inner brush of bristles; the right mandible with 2 apical
and 3 lateral teeth, the proximal tooth shelf-like; no lines of setulae on either mandible. Outer
surface of both mandibles with a number of short setae. Maxillary palpi short, stout, five-seg-
mented; labium short, thick, palpi minute. Submentum with rounded apex, it and stipites dark
brown.

Thorax (Figs 56, 59): pro- and mesonota very dark brown, the latter paling slightly poste-
riorly, pronotum with thickened black posterior and lateral margins, thickly set with simple
dark setae, metanotum paler brown; meso- and metanota pubescent, with black lateral mar-
gins and transverse lines where nota break at ecdysis, common to the Diplectroninae; both
with sinuous posterior margins, each with a conspicuous median black mark. Pre-episternum
simple, it and pleural sclerites dark brown, setose. Prosternal plate angled, light brown with
darker marks; membranous parts of thorax cream with scattered dark setae; single mesosternal
tracheal gill (Fig. 56a), two pairs of metasternal gills, the median pair close together but not
joined.

Legs (Figs 61-63): forelegs stout, strong, dark to mid-brown, strongly setose, claws blunt
with short blunt basal spine; scraper of stridulator on femur as usual. Mid- and hind legs less
stout, light brown with darker markings, thickly fringed with strong dark setae, many of them
sword-like; most other setae paler, less conspicuous, claws with stout basal spine, less blunt
than foreclaws. Lightly feathered setae present on coxa, trochanter and tibia of foreleg and tib-
iae of mid- and hind legs.

Abdomen (Figs 56, 64, 65): cream in life, mauve-tinged in spirit; many short dark setae
scattered over whole abdomen, those on venter very small, inconspicuous. Tufted tracheal gills
present in four to six rows (see Table V). In this genus the gill filaments are few in number,
usually 3-4 filaments (branches) per gill, occasionally 2 or 5. Minute conical lateral gills (‘Zip-
felkiemen’ of Ulmer) on segments III (1), VI (2), V (2), VI (2). Sternum of segment VIII
bears small pair of triangular sclerites set with yellow spines and fringed posteriorly with long
dark setae, these sclerites are joined medially by a glabrous scierotised area. Segment IX bears
a pair of much larger triangular spine-clad sclerites on venter, a small pair of setose lateral scle-
rites and a larger pair of setose dorsolateral sclerites. Anal prolegs fairly long, with well-devel-
oped lateral sclerites and strong, plain claws, each leg terminating in a large fan of black setae.
Five anal gills.

Earlier instars: few immature larvae were present in the material available; one of these
appeared to be in the fourth instar, six in the third. These are briefly described below.

AM— 7

345

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Figs 56-65. DIPLECTRONINAE: Sciadorus obtusus, mature larva. 56. habitus of mature larva, 56a. mesothoracic gill,
further enlarged, 57, 58. dorsal and ventral views of head, 59. dorsal view of thoracic nota showing transverse ecdysial
lines, 60. left and right mandibles, dorsal, 61. right foreleg, with pre-episternum, episternum & epimeron, 61a. foreclaw
further enlarged, 62, 63. right mid- & hind legs, 64. right anal proleg, claw and brush, 65. ventral view of abdominal seg-
ments VIII & IX to show sclerites & setae.

346

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Third instar: three larvae in the Vet River sample, one from the Blaauwkrantz River and
two from the Storms River tributary. Length 5-7 mm, gill count as in mature larva, the gills on
the whole longer relative to size of larva than in final instar, nearly all gills with one or two fila-
ments per gill, occasionally three in ventral gills. Colour of head and nota varies from dark
brown (Vet River) to mid-brown (Blaauwkrantz and Storms Rivers), light spots on head vari-
able in intensity, not always easy to distinguish. Colour differences possibly attributable to alti-
tude differences. Small conical gills absent. Anal setal fans very thin.

Fourth instar: single larva in sample from Vet River, much as final (fifth) instar in appear-
ance; length 10 mm. All gills present, smaller than in final instar, with two to three branches.
Conical gills present, minute, on segments III— VI. Colour markings on head etc. as in final
instar. Anal fans thicker but not as dense as in final instar.

I would expect first instar larvae to lack tracheal gills entirely, second instar larvae to have
the full number of gills found in later instars, those gills being simple and all or almost all un-
branched.

Table V.

Number and position of tracheal gills in mature larvae of Sciadorus obtusus and S. acutus (right

side, abdominal)

Segment no.

Sciadorus obtusus

Sciadorus acutus

Thoracic gills

Mesosternal

Metasternal

single (median)
2 pairs

3 gills (1 pair & 1
median gill with 2 stems)

Abdominal gills

DL

VL

V

DL

VL

V

I

1 1

1

1

11

1

2*

1

1

2

1

III

1

2

1

1

2

1

IV

1

2

1

1

2

1

V

1

2

1

1

2

1

VI

1

2

1

1

2

1

VII

1 2

1

2

*The figure 2 denotes that the gill has 2 main stems, either joined basally or contiguous within the same basal circle.

DL = dorsolateral gills: 2-3 filaments per stem.

VL = ventrolateral gills: 3-4 filaments per stem.

V = ventral gills: 4 filaments per stem as a rule.

Note: There are also very small simple conical gills (lateral) on segments III— VI , dorsal to the lateral line area (no vis-
ible lateral line). (See Fig. 56)

Description of pupa (Figs 49-53, Table VI)

Specimens available for study comprised 8 pupae together with their cases. Of these four
were males (one good, mature, one fairly good, one very poor and one very immature), and
four were females, two of them very good. All came from the Vet River on Garcia’s Pass, and
were collected in February, 1976. Of these, the good, mature male pupa proved to be Sciado-

347

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

rus obtusus Barnard, thus providing the correlation between larvae, 6 and 9 pupae and ima-
gos. Although none of the Sciadorus pupal cases contained larval sclerites, which is very un-
usual in Hydropsychidae, there was evidently only a single species involved, hence the
possibility of correlation. The loss of larval sclerites from the pupal cases could well have been
due to the large size of the terminal openings (Fig. 55) combined with pupal movements and
fast flowing water.

Male pupa (pharate imago) (in spirit. Figs 49-53): length 7-7,5 mm, imago clearly seen
through pupal pelt, mainly dark brown with cream warts set with long, dark brown setae, flat-
tened beneath pelt; mesoscutum with two large patches of setae, mesoscutellum with long se-
tae, antennae appearing serratulate even within the pupal pelt. Face of pupa with scattered
dark setae and four clumps of setae just above labrum; labrum small, brown, oval, thickly set
with dark, shorter setae; mandibles brown, strong, the left one with five apical and subapical
teeth, the right with four. Maxillary and labial palpi as in imago, but pale, the labial palpi very
small and inconspicuous. Forelegs and mid-legs with tufts of long dark setae on the coxae, mid-
legs slightly flattened, with swimming setae; tibial spurs 2.4.4. Fore-, mid- and hind legs reach
to distal end of abdominal segments II and V and apex of abdomen respectively. Wings (unex-
panded in sheaths), dark golden brown, strongly setose, fore- and hind wings reaching to distal
end of segments VII and VI respectively. Tufted gills, conical lateral gills and dorsal plates
present on abdomen as shown in Table VI and Figs 49, 49a. Conical gills smaller than in Di-
plectronella, and ventrolateral branchial gills actually arise from proximal side of conical gills.
Anal appendages shortly bifurcated, both apices curving to a point with a saddle between, out-
er point slightly shorter than inner; appendages apically strongly denticulate, set with short,
heavy spines and long, strong, dark lateral setae. Both sexes with a small, inconspicuous pair
of filaments on abdomen V, towards ventral side of pleura. Usual pockets enclosing male geni-
talia.

Female pupa (in spirit): length 7-7,5 mm, cases as in males. As male pupa, but gills larger
and very clear and dorsal plates appear smaller though otherwise similar. One female had an
egg string tangled up inside its case, a small part is shown in Fig. 54.

Pupal case: oval, about 9-10 mm long, strongly built, usually very shaggy in appearance,
incorporating much vegetable matter (pieces of wood and charcoal) as well as sand grains
(white quartz) and small rock fragments. Both ends have large, somewhat meandrine ope-
nings with silk flaps. No larval sclerites found within cases.

Remarks

Comments on the genus Sciadorus Barnard and the two Cape species, 2S. acutus and
S. obtusus , appear above. It may be noted here, however, that S. obtusus imagos differ from
those of acutus not only in detail of genitalia and head warts, but also in size and colouring, be-
ing larger and much darker. The same applies to the larvae, which are larger, with almost
black sclerites, but the simplest distinction is the possession by obtusus larvae of a single well-
developed mesosternal gill (Fig. 56a) and two pairs of metasternal gills, whereas acutus larvae
have no mesosternal gills and three metasternal, of which the median gill has two branches
joined basally.

Jacquemart’s record (1963a: 383, 384, Fig. 38A-D) of S. obtusus from the Orange River,
Upington, has been omitted from the distribution of this species because his drawings appear
to indicate two different taxa. He also recorded S. obtusus from the Swartberg Pass, S. W.
Cape (altitude c. 1525 m), which is a known habitat of that species, and his Fig. 38D (genitalia,
laterial) could be of S. obtusus , as could Fig. 38A, B (wings). Fig. 38C (genitalia, ventrodorsal)
does not, however, appear to be obtusus and many well not be of a Sciadorus at all. This is a
possibility as Jacquemart worked entirely from specimens already prepared and mounted on

348

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Table VI.

Tracheal gills and dorsal plates of 8 pupa of Sciadorus obtusus Barnard

Abdominal segment no.

Tracheal gills (right side)

Dorsal plates

I

Lc

Lb

V

II

1

2*

Presegmental pair

III

(1)

1

2

Presegmental pair

IV

(2)

1

2

Postsegmental pair
Presegmental pair

V

(2)

1

2

Postsegmental pair
Presegmental pair

VI

(2)

1

2

Postsegmental pair
Presegmental pair

VII

(1)

1

2

Presegmental pair

VIII

—

—

—

Presegmental pair

*2 indicates double tracheal gills, either with a common base or contiguous bases.
(1) Brackets indicate conical gills (lateral diverticula). (See Fig. 49a)

Lc = lateral conical gills.

Lb = lateral branchial gills.

V = ventral branchial gills.

slides so that a lateral and a “ventrodorsal” view could not be drawn from the same specimen.
Sciadorus has small, slender claspers and simple genitalia lacking the long, spiny processes the
drawing shows. The lower Orange River at Upington is not a known habitat for Sciadorus ,
which is usually confined to the upper reaches of fast-flowing mountain streams (see also under
Biology).

S. sinuatus Marlier (1943a) was described as having longer lateral filaments on V than
either of the Cape species, as long as in Diplectronella, and there appeared to be other resem-
blances to that genus. Study of imagos kindly lent to me by Dr Decelle (Tervuren) revealed
that the male abdomen lacked the large paired reticulated organs found in Sciadorus and Di-
plectrona , but not in Diplectronella , and that the wings were almost identical with those of
D. medialis Marlier (also borrowed from Tervuren). A careful comparison was therefore
made, and it is clear that Sciadorus sinuatus must in fact be transferred to the genus Diplectro-
nella, becoming D. sinuata (Marlier) (Stat. nov.). D. sinuata is very similar to D. medialis Mar-
lier, differing in a few minor details such as the position of some of the cross-veins in the wings
and the shape of the apices of the tenth tergum in the 8 genitalia. On that account I have left
them as separate species for the present although it is possible they may prove to be synony-
mous should series of both species be studied.

The genus is said to resemble closely the genus Diplectrona apart from the usual absence
of fork 1 from the hind wings. Their hind wings, however, also differ in shape, and in Sciado-
rus the apices of Sc and Rt are not strongly upturned as in Diplectrona and Diplectronella , and
the lateral filaments on V are short, not long. The larvae and pupae also show consistent dif-
ferences. Males of both Diplectrona and Sciadorus possess in the abdomen the large reticulated

349

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

organs the function of which is still obscure. These organs also occur in one or two other Di-
plectronine genera not found in Africa.

Biology

Little is known about the biology of Sciadorus species. Being endemic to the Cape Province
and usually somewhat inaccessible they have not to my knowledge been reared in the laborato-
ry or used in any experimental work. Field observations indicate that Sciadorus appears to be a
high altitude genus, infrequently encountered in highlying, fast-flowing, clean cold streams,
where the species occur higher up than Cheumatopsyche although there may be a small over-
lap. It is not possible to compare their distribution directly with that of Hydropsyche species,
as that genus has not as yet been recorded from the same areas.

Barnard (1934) collected S. acutus from the West Cape (French Hoek and Hottentots
Holland Mountains) between altitudes of 2 000 and 4 000 ft (610—1 220 m), 5. obtusus from the
Zwartberg Range, South-West Cape, between 3 000 and 5 000 ft (915-1 525 m). Streams at
such levels are highlying and fast-flowing, often torrential, the mountains themselves rising
from the coastal plain. The Albany Museum material of S. obtusus was collected from the
South-West Cape mountains at altitudes of 500-600 m, in similar streams, where it was the
only Hydropsychid present. Bertrand’s specimens from the Blaauwkrantz River (at 180 m) and
Storms River tributary (no altitude given) were, however, found much lower down. This can
probably be explained by the fact that both are “drowned rivers”, their original estuaries and
lower reaches now lying submerged beneath the sea, as is the case with all the rivers in that
part of the South-West Cape. The Storms River specimens were taken together with a single
larva of Cheumatopsyche afra.

On the Robinson Pass S. obtusus larvae were taken from strongly flowing water in a forest
stream, from the outer face of a weir and from adjacent rock faces. There they had built thick-
walled, rather untidy shelters incorporating much vegetable matter such as twigs and charcoal
fragments as well as sand grains. Attached to each shelter, facing full into the current, was a
neat, coarse-meshed, wide-mouthed catching net, attached to the rock on one side, often by
means of a line of small stones, and held open and reinforced laterally and along the outer
edge by bundles of twiglets bound together by silk secretion. Within this shelter lay the sturdy,
powerfully built larva, usually facing upwards and into the current. The posterior exit was sur-
rounded by a double or treble ring of stones, cemented together and smoothed over with se-
cretion. Gut contents included many insect sclerites and some detritus, indicating that the lar-
vae are undoubtedly predators. The shelters were unusual in that some also contained a
walled-off compartment filled with detritus, possibly simply sealed off with silk when much ex-
traneous material had accumulated, perhaps at a time of low flow. Some shelters contained
half-eaten prey, such as Corixid bugs.

The Garcia’s Pass specimens came from a small fast flowing mountain stream near the
summit, on a rocky bush-covered hillside. The shelters and nets were of similar type to the
others, but incorporated a higher admixture of charcoal blocks, probably reflecting a recent
mountain fire.

The larvae pupate in stout cases of the same materials as the larval shelters, with a thick
soft lining, affixed to the rock and with rather large terminal openings, meandrine and guarded
by silken flaps (flaps closed in early prepupal case, later ones and pupal cases open). The ope-
nings are evidently large enough to allow the larval sclerites to escape with the water flowing
through during ventilation. Mature pupae were found in February (late summer), together
with mature larvae and prepupae, so emergence could well have continued into the autumn.
Bertrand’s female imago was taken in late September (spring), and Barnard’s males in Novem-
ber and February (late spring and summer), as were his S. acutus (possibly a function of col-
lecting dates rather than of life cycles). At any rate it would appear that these species probably

350

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

emerge during the spring, summer and autumn months, no doubt overwintering as young lar-
vae.

African species of Sciadorus

S. acutus Barnard, 1934: 364-366, fig. 39j-n (<3, 2 genitalia, larval head). 3 $ metam.
Distribution: South Africa (West Cape).

S. obtusus Barnard, 1934: 366, fig. 39a-i (3 wings, genitalia, antenna, ventrolateral process on
abdomen V, internal vesicles). 3 .

Distribution: South Africa (South-West Cape).

S. sinuatus Marlier, 1943a: 11, 12, fig. 7 (3 genitalia). 6 (2 not described). Distribution: Con-
go (Albert National Park).

Transferred to Diplectronella in this paper, as D. sinuata (Marlier).

Genus DIPLECTRONELLA Ulmer 1928

Diplectronella Ulmer 1928: 317.

Type species Hydropsyche taprobanes Hagen (selected by Fischer 1963: 150).

Diplectronella Ulmer; Mosely 1931: 195, 196 (discussion and redefinition of genus).

Diplectronella Ulmer; Marlier 1962b: 140-142.

Diplectronella Ulmer; Fischer 1963: 150, 151, 1972: 153.

Generic diagnosis (based on Mosely 1931, Marlier 1962b, Ulmer 1951):

Imago: tibial spurs of 3, 2, 2.4.4, claws normal. Maxillary palpi with 1st segment short, 2nd
longer, dilated, with tuft of setae on truncate apex (inner side), 3rd and 4th progressively
shorter, 5th about as long as rest together, articulated. Labial palpi 3-segmented, 3rd longest,
articulated. Vertex with fairly large bean-shaped anterolateral warts, posterior warts larger,
oval, median wart clear; eyes strongly hairy. Antennae faily short, thick, of usual Diplectro-
nine type, each segment with short spines or setae and slight dilatation, giving the antenna a
serrated appearance (Fig. 67). Forewing with apical forks 1, 2, 3, 4, 5, discoidal and median
cells closed, forks 1 and 3 stalked; wing apically broad, apex obliquely truncate; subcosta and
radius much thickened from base to cross-vein; length about 7-11 mm. Hind wing very broad,
lower margin almost semicircular, forks, 2, 3, 5 present, occasionally a small fork 1 (very small
cf. with that in Diplectrona)\ discoidal cell closed, median cell absent; subcosta elbows down-
wards nearly meeting radius near apex of wing, so may appear to form a false fork; cubitus of-
ten pigmented. Venation sometimes variable. Abdomen: 5th sternum (<3 & 2) bears a long
pair of ventrolateral filaments, reaching apex of abdomen in 3, to mid-VI in 2; each arises
from an internal gland. No large internal vesicles such as characterize 3 Diplectrona and Scia-
dorus. 3 genitalia with tergum of X forming a pair of large, apically pointed lobes, each bear-
ing a setose lateral boss; between them lies a smaller pair of slender lobes; copulatory organ di-
lated at apex to form a cup-like depression containing four stylets (cf. Fig. 78), lateral to which
is a pair of small erectile lobes; claspers long, two-segmented, basal part long, apically thick-
ened, apical part shorter (but much longer than in Sciadorus). 2 genitalia with sternal plates of
VIII separated to base and the usual small cerci and papillae; mid-legs not expanded.

351

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983
Diplectronella medialis Marlier
(Figs 66-87, Tables VII, VIII)

Diplectronella medialis Marlier, 1961: 159, 191-197, figs 19-20 (wings, larval parts). 6 2.

Kivu, Belgian Congo.

Type material in Musee Royal de l’Afrique Centrale, Tervuren, Belgium.

D. medialis Marlier; 1962b: 140, 142, figs 76, 79 (larval & pupal parts).

Distribution: Belgian Congo (Kivu); Zaire (River Kalengo).

NOTE: The material described below was very kindly sent to me on loan by Dr Decelle of the
Musee Royal de l’Afrique Centrale at Tervuren, and forms part of that originally described by
Dr Georges Marlier (1961, 1962b). It comprised 3 male, 2 female imagos in spirit, 6 parts on
slides, 6 larvae and a male pupa. In addition I had a few Diplectronella larvae collected by Dr
J. O. Young in Kenya, which were useful for comparison; they could well have been of D. afra
Mosely.

Description of imagos (Figs 66-73)

Male imago (in spirit. Figs 66-71 and cf. 78); (colour description from Marlier 1961); a
medium-sized species, forewings (both sexes) 7, 5-9, 5 mm, membrane brownish-yellow, vena-
tion distinct, hyaline mark on r-m cross-vein; insect generally brownish-yellow in colour, head
yellowish, warts paler. Vertex with clear median wart, bean-shaped anterolateral warts and
large, oval posterior warts, mid-cranial sulcus clear, complete; a pair of conical tubercles on
vertex. Lateral filaments on V long, reaching end of VIII. Marlier remarks that the genitalia
are very similar to those of D. afra Mosely, but does not illustrate them, separating the two
species on the basis of forewing venation, particularly the position of the r-m cross-vein in re-
lation to that closing the median cell, which in D. medialis is in fact very similar to that in
D. indica Mosely (1931) from India. The male genitalia are figured here (Figs 70, 71) and show
the pubescent median processes of X.

Female imago (in spirit, Fig. 73): as male in colour and general appearance. Lateral ap-
pendages of V shorter than in d, reaching to middle of VI. Genitalia: IX not retracted within
VIII, XI with usual small cerci and papillae; sternal plates subrectangular, subgenital plate
with prominent terminal lip, downturned in lateral view.

Description of larva (Figs 79-87, Table VII)

Mature larva (in spirit. Figs 79-87, Table VII): length 16-18 mm, larvae fairly large, of
typical Diplectronine type, with transverse ecdysial lines on meso- and metanota, and fronto-
clypeal apotome which widens behind eyes, though less so than in Sciadorus ; abdomen rough
in appearance due to presence of small setae and minute scales. Sclerites pale brown, abdomen
yellowish. Although the larva is quite strongly setose it appears much less so than Sciadorus as
the setae are pale brown in colour and inconspicuous. There are few anal setae, not forming a
fan.

Head (Figs 79-81): oblong, brownish with paler patches as indicated, without carina,
frontoclypeal apotome about as broad behind eyes as in front; anterior and posterior ventral
apotomes fairly large, but together shorter than ventral ecdysial line; head thickly set with both
blunt and pointed setae. Stridulatory files present, shortly oval. Eyes small, set well forward
under usual cuticular lenses.

Mouthparts (Figs 80-82): labrum semicircular, yellowish brown, finely setose with large
lateral brushes, anteclypeus membranous. Mandibles brown, triangular in section, setose on
outer side; left one with two large apical teeth, between these a smaller tooth, four lateral
teeth (one on dorsal side) and an inner brush of setae; right mandible with one apical and four

352

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Figs 66-78. DIPLECTRONINAE: Diplectronella medialis , 3,9,6 pupa. 66. 6 : dorsal view of head and thoracic nota,
67. 6: lateral view of head and palpi, 68, 69. 6 : fore and hind wings (right) (from 2 individuals, hence hind wing rela-
tively larger than forewing), 70, 71. 6: dorsal and lateral views of genitalia, showing apices of processes on abdomen V,
72. 9 : lateral view of genitalia, 73. 6 pupa: habitus (dorsal), left dorsal plates of segment II (presegmental). III (pre- &
postsegmental), VI (presegmental) shown further enlarged; apical parts of exuviae removed, 74. same: apical part of
exuviae (dorsal view), from slide, 75. same: apex of right anal appendage further enlarged (ventral view), 76, 77. same:
left and right mandibles; part of dentate margin shown further enlarged. Diplectronella sinuata, d : 78. Apex of copulato-
ry organ (right lateral view) showing cup-likc tip containing 2 pairs of stylets (of which 2 are visible), also small erectile

lateral lobe.

AM— «

353

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

lateral teeth, no inner brush; a line of setulae on the dorsal tooth or teeth in both mandibles.
Maxillary palpi short, thick, five-jointed, much as in Sciadorus. Submentum with rounded
apex, slightly concave anteriorly, pale brown.

Thorax: (Fig. 79): thoracic nota brownish with narrow black margins as indicated, set with
small brownish setae, again including both blunt and pointed types. Pre-episternum simple, se-
tose, pro-epimeron wraps round base of foreleg. Prosternal plate stout, brown anteriorly,
blackish posteriorly, angled as in Sciadorus. One pair of mesosternal gills, three metasternal
gills (coxal pair and single two-stemmed median gill).

Legs (Figs 83-85): foreleg stout, mid- and hind legs longer and more slender than in Scia-
dorus, brownish, densely setose, but setae pale and inapparent except under compound
microscope; foreleg with plumose setae on coxae, trochanter and femur; mid- and hind legs
with feathered setae on femora, tibiae and tarsi. A scraper is present on the fore-femur, as in
many other hydropsychids, but is not easily seen as it projects inwards; it is clear in ventral
view, forming a small triangular projection.

Abdomen (Figs 79, 86, 87): tufted tracheal gills present in four rows (see Table VII). Gills
clear, with more filaments than in Sciadorus, usually 7-12 per gill stem. Small conical lateral
gills (“Zipfelkiemen”) present on segments III (1), IV (3), V (3), VI (3) and VII (3), larger

Table VII:

Number and position of tracheal gills in mature larvae of Diplectronella medialis

(right side, abdominal)

Segment no.

Diplectronella medialis

Thoracic gills

Mesosternal

Metasternal

1 pair gills

1 pair coxal gills, single median gill

Abdominal gills

VL

V

I

1

1

II

2*

1

III

2

1

IV

2

1

V

2

1

VI

1

1

VII

—

—

2*: denotes that the gill has 2 main stems, joined basally. Tracheal gills bushy, with 7-12 filaments per stem, roughly in
whorls.

VL = ventrolateral gills.

V = ventral gills.

Dorsolateral gills are absent in this genus in contrast to Sciadorus.

Note: The conical lateral gills on the pleural area of Diplectronella larvae are conspicuous and much larger than in Scia-
dorus, occurring on segments II I— VII (see Fig. 79).

354

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Figs 79-87. DIPLECTRONINAE: Diplectronella medialis, mature larva. 79. habitus, lateral (right appendages only
shown), 79a. ventrolateral gill (double) of segment III, spread out & enlarged, 80, 81. dorsal & ventral views of head;
dorsal view shows colour pattern on right half, setation on left, types of setae further enlarged, 82. left & right man-
dibles, dorsal, 83, 84, 85. right foreleg, mid- and hind legs, foreleg with pleural sclerites; foreclaws and mid-claws fur-
ther enlarged, 86. ventral view of abdominal segments VIII & IX showing sclerites & setae, 87. right anal proleg, lat-
eral.

355

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

than in other genera so far described. Sternum of VIII with a pair of small sclerites each bear-
ing a few stout setae, apparently joined medially by a very lightly sclerotized area. Sternum of
IX has larger paired patches of spine-like setae, not confluent medially, and a pair of small lat-
eral setate sclerites, but no dorsal ones. Anal prolegs with pale brown lateral sclerites and a
few terminal setae, no fans; anal claws plain, strong; five anal gills.

Description of pupa (Figs 73-77, Table VIII)

A good male pupa was kindly made available to me for study by Dr Decelle; the following
description was based on it and compared with Marlier’s own account. A few larval sclerites
accompanied the pupa. The anal appendages and mandibles had been removed and mounted
on a slide.

Male pupa (pharate imago) (in spirit. Figs 73-77): length about 11 mm, imago clearly vis-
ible through pupal pelt. Antennae clearly serratulate even through pelt, short, thick, about 5/6
body length. Labrum rounded, with pair of strongly setose lateral protuberances and median
area; mandibles slightly bowed, each with several setae on outer side, inner margin appearing
finely serrated due to presence of blade-like spinules; left mandible with four apical and subap-
ical teeth, right with five. Maxillary and labial palpi large, clear. Wings (unexpanded) short,
just over half body length. Abdomen with anterior dorsal plates on segments II- VII , posterior
dorsal plates on III and IV; lateral gills very large and prominent, including ventral tufted gills
and conical, branched, pouch-like lateral gills (see Table VIII). Genitalia relatively small, clear
(Figs 70, 71). The anal appendages are forked, apices broadly squared (not sharply pointed);

Table VIII:

Tracheal gills and dorsal plates of d pupa of Diplectronella medialis Marlier

Abdominal

Tracheal gills

Dorsal plates

segment no.

(right side)

I

II

Lc

V

(1)

1

Presegmental pair

III

(3)

2*

Presegmental pair
Postsegmental pair

IV

(3)

2

presegmental pair
postsegmental pair

V

(3)

2

presegmental pair

VI

(3)

1

presegmental pair

VII

(3)

—

presegmental pair

VIII

—

—

—

*2 indicates double tracheal gills, either with a common base or contiguous bases. Tracheal gills are bushy, with about
20 long, fine filaments per double gill.

(3) Brackets indicate conical pocketlike gills (lateral diverticula); these are large, long and prominent.

Lc = lateral conical gills (1-3 pockets superimposed).

V = ventral branchial gills.

There are no lateral branchial gills as in Sciadorus.

356

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

apices densely covered (particularly on ventral side) with strong, closely set denticles, a few
long lateral setae present, also a pair of long, stout, shortly feathered setae on the inner side of
each appendage (Figs 74, 75).

Pupal case: an oval case made of small stones affixed to the substrate.

Remarks

Mosely (1931) states that the paired ventrolateral filaments on the abdomen arise from
segment IV, but this is an error, as they definitely arise from segment V (Marlier 1961 and per-
sonal observation).

Marlier (1962b) mentions “three species of Diplectronella in Africa”; I only know of two,
however, namely D. afra Mosely (1931) and his own D. medialis (Marlier 1961, 1962b). Mar-
ker’s sinuata of course makes a third, but that was regarded at the time of his writing as a
species of Sciadoms. He notes that the genus Diplectronella appears to occur only in small
localized populations, remarking that the resulting inbreeding may be the cause of the aberra-
tions in wing venation known to occur in this genus (1961 p. 193).

As indicated under “Remarks” on the genus Sciadorus , 5. sinuatus Marlier (1943a) is in
fact a species of Diplectronella , as is evidenced by the general appearance of the imago and
more particularly by the shape of the hind wings and their venation, the head warts, the 6 gen-
italia and the length of the lateral appendages on V. Most of these points (and others) will be
clearly seen on comparison of the figures for the two genera, as D. sinuata is very similar to D.
medialis. The acid test is, however, the possession of large paired reticulated internal organs by
the <3 of Sciadorus (and Diplectrona) and not by Diplectronella males.

Biology

Marlier (1962b) commented that Diplectronella larvae lived in cold, clear running water,
on stony bottoms in highlying streams or cascades, being found in small numbers in crevices
between stones where they wove their silken nets. Larvae collected had been feeding on veg-
etable matter (debris of higher plants) and on small crustaceans ( Darwinula sp., Ostracoda)
(Marlier 1961). Known African species are mountain dwellers, living almost entirely at high al-
titudes in Central Africa; the lowest found was at 950 m (over 3 000 ft).

Little else was known about their biology until the publication of Statzner’s work on the
emergence of Trichoptera imagos from a Central African stream, the Kalengo, which flows
from the highlands of Zaire into Lake Kivu (Statzner 1976). From his studies it became clear
that Diplectronella medialis imagos, male and female, emerged throughout the year (though
occasionally only one sex was taken). Numbers were always low, from one or two to a maxi-
mum of eighteen, and (at least from September to February) males were somewhat more nu-
merous than females. Statzner calculated the numbers collected twice monthly, the total num-
bers of both sexes and their dry weight, and gave the percentages of females to males and of
total dry weight of D. medialis to dry weight of the whole catch. These are listed below (ex-
tracted from Statzner 1976, using his figures for the period April 1972-March 1973 inclusive).
The numbers given had been extrapolated to cover days when no collecting was done.

Diplectronella medialis Marlier

Totaled Total?? ?% % Abundance Drywtc? Drywt? %, Total dry wt

184 85 31,6% 2,56% 372,97 mg 277,70 mg 14,08%

Statzner (1976) also contributed to a knowledge of their food, since he found that
D. medialis larval guts contained parts of Trichoptera, Diptera and water mites as well as plant

357

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

material. Life cycles appeared to take roughly four months, which would give three peaks of
emergence during the year in a tropical climate. The actual peaks on the graph (Fig. 21), oc-
curred in October and February of 1972 (both in the rainy season) and June, July of 1973, in
the dry season (the sampling was not continued beyond July). Statzner suggested that the fac-
tor controlling emergence was neither rainfall nor phases of the moon nor waterlevel (the lat-
ter did not decrease in the Kalengo during the dry season), but changes in light intensity,
which occurred at the beginning and end of the dry season. Emergence values as counted,
might have been affected by the fact that breeding and egg-laying could take place in the emer-
gence house, which was sited over a portion of the river.

African species of Diplectronella

D. afra Mosely, 1931: 202-205, figs 10-13 (d wings & genitalia). <3.

Uganda: Kampala.

Distribution: Uganda (Kampala, Ruwenzori), Ethiopia.

D. medialis Marlier, 1961: 159, 191-197, figs 19, 20 (wings & larval parts). 8 9 .

Belgian Congo: Kivu. Larva also described.

Marlier 1962b: 140-142, fig. 79 (pupal parts); further description of adults, larvae and pu-
pae.

Distribution: Congo (Kivu), Zaire (River Kalengo).

D. sinuata (Marlier), 1943a: 11-12, fig. 7 (8 genitalia), as Sciadorus. 8 2 .

Belgian Congo.

Transferred in the present paper to Diplectronella.

Distribution: Belgian Congo.

May prove to be synonymous with D. medialis Marlier; see discussion under “Remarks”
on Sciadorus.

Genus DIPLECTRONA Westwood 1840

Diplectrona Westwood 1840: 49

Type species: Aphelocheira flavomaculata Stephens 1836 nec Pictet 1834, syn. of felix
McLachlan (original designation). Europe, North Africa.

Diplectrona Westwood; Ulmer 1957: 329, 331, 414 (larva and pupa).

Diplectrona Westwood; Wiggins 1977: 102, 103, figs A-G (larva).

Generic diagnosis (based on Mosely 1939c, Mosely & Kimmins 1953, Ulmer 1951, Schmid
1980 and others):

Imago: tibial spurs of 8 , 9 , 2.4.4, anterior tarsal claws of 8 normal. Maxillary palpi with basal
segment short, second, third and fourth long (second being longest), fifth articulated, scarcely
as long as rest together. Head: warts on vertex of usual Diplectronine type, two large pairs
present, posterior ones largest; antennae slender, about length of forewings, distal segments
slightly inflated in middle, appearing serrated. Forewing short, broad, with apical forks 1, 2, 3,
4, 5; 1 and 3 having long footstalks, discoidal cell short, closed, median cell longer, closed, thy-
ridial long, narrow, Cu: and 1A meet, or almost meet, at arculus. Hind wings shorter, broad,
blunt, with forks 1, 2, 3, 5 (1 being large), discoidal cell closed, median open, Ri, and often Sc,
strongly curved upwards near apices to meet costa, in both fore and hind wings. Abdomen:
fifth sternum bears a pair of very long, slender ventrolateral tubular filaments, each arising
from an internal gland (these glands being larger in the 8 than in the 9); the filaments may
reach beyond the end of the abdomen. Within the abdomen in the 8 there are also two pairs
of large, rounded reticulate, internal sacs, opening on the posterior margins of VI and VII, and

358

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

frequently extending over two or more segments (usually four sacs, but number varies accord-
ing to species). 3 genitalia usually with dorsal plate of X divided into two or more lobes.
Claspers with long, stout basal segment, short, narrower apical segment. Phallic apparatus
large, simple. $ genitalia of usual Diplectronine type, sternal plates of VIII separated to base,
mid-legs not widened.

The genus occurs in North Africa, Diplectrona felix McLachlan having been recorded
from there, but nowhere in the Afrotropical Region. It does not occur in the Neotropical Re-
gion either, but is widespread in the Northern Hemisphere and Australian Region.

Diplectrona felix McLachlan

Diplectrona felix McLachlan 1878: 376, pi. 40, figs 1-9. 3 2 . Britain, France, Guernsey.
Diplectrona felix McLachlan; Gauthier 1928: 26. North Africa.

Diplextrona felix McLachlan; Mosely 1939c: 194, figs 411-415. 3 2. (Wings and genita-
lia).

This is a European species whose distribution extends into North Africa, which forms part
of the Palaearctic Region, not the Afrotropical. Should further information about it be requir-
ed, Fischer (1963 and 1972) gives many references, including ones to the larval and pupal
stages.

Remarks

Larvae of this genus are excellently illustrated and described in Wiggins (1977). Charac-
ters distinguishing them from Sciadorus and Diplectronella will be found in the keys at the be-
ginning of the present paper. The larvae are typically Diplectronine in shape of head, fronto-
clypeal and ventral apotomes and in method of ecdysis by rupture of transverse ecdysial lines
on the meso- and metanota. Meso- and metasternal gills are present, and the larvae have four
rows of tracheal gills on the abdomen, most ventrolateral ones being double, as have Diplec-
tronella larvae. Sciadorus has six, most ventrolateral ones also being double, and there are dor-
solateral gills in addition to the others. Older larvae in all three genera possess in addition the
smaller conical lateral gills termed by Ulmer “Zipfelkiemen”.

Diplectrona larvae inhabit small, cool swift-flowing streams, often in vegetation, construct-
ing shelters and catching nets in the usual way. Their normal life cycle appears to be one year,
and swarming adults have been observed in August (Edington & Hildrew 1981).

African species of Diplectrona

D. felix McLachlan, 1878: 376, pi. 40 figs 1-9. 3 $ . Britain, France, Guernsey.

Fischer 1963: 142-145, 1972: 148, 149.

Gauthier 1928: 26. North Africa (Algeria, Tunisia).

Mosely 1939c: 194, figs 411-415 (3 wings & genitalia). Britain.

Tobias 1972: 395, figs 158-161 (Genitalia). 3 9.

Lestage 1921: 544-546, fig. 183a-i (larval & pupal parts).

Distribution: Europe, North Africa.

359

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983
Subfamily MACRONEMATINAE Ulmer 195

Genus MACROSTEMUM Kolenati 1859

Macrostemum Kolenati, 1859: 168, 239.

Type species Macronema hyalinum (Pictet) (described as Hydropsyche), mentioned by
Ulmer 1957: 339 as one of Kolenati’s two Macrostemum species; selected by Fischer (1963:
176). (For description of M. hyalinum see Ulmer 1907d: 75, 76, fig. 78 & col. pi. Ill figs
15, 16 6 imagos.)

Macrostemum Kolenati; Barnard 1934: 372, key to South African species and varieties (as
Macronema ): larva (as Chloropsychel ), p. 368 & figs 40h-l.

Macrostemum Kolenati; Wiggins 1977: 1 10-111, figs 6.8A-E (larva), as Macronema.

Macrostemum Kolenati; Flint & Bueno-Soria 1982: 358-9, 367-369, resurrection of name
Macrostemum for their second group of Macronema species (for reasons see under Re-
marks below).

Generic diagnosis (derived from Flint & Bueno-Soria, 1982, and other sources and checked
against all other African species as far as descriptions and figures allow, as well as some from
other regions; diagnosis is applicable to African species but not necessarily to those from else-
where).

Image: tibial spurs of 8, 9, usually 2.4.4, sometimes 1.4.4, apical spurs on foreleg sometimes
small but always present. Antennae very long, threadlike, twice to thrice forewing length in 6 ,
shorter in 9. Maxillary palpi well developed, labial palpi relatively small. Head: vertex with
large pair of anterior warts, sometimes a median wart, small posterior pair. Forewing elongate,
rounded at apex, rarely pubescent (if so, setose not scaly), usually translucent, shining, mem-
brane often brightly patterned; discoidal and median cells present, closed, discoidal cell small
but not minute, median cell larger than discoidal, thyridial cell very long, costal cross-veins
may be present, Sc united to Ri shortly before reaching wing margin; forks 1-5 present (oc-
casionally only 1-4 in <5, or a false 5 present). Hind wing hyaline, sometimes iridescent, short-
er than forewing, broad, anal portion much widened in 6, less so in 9, folded, discoidal and
median cells absent, forks 1, 2, 3, 5 or 2, 3, 5 present; costal margin excised before apex, par-
tially bordered with hooked macrotrichia. Length of forewing 9-21 mm, in 9 less (9 usually
smaller than d). Fifth abdominal sternum bears a pair of short filamentous appendages in
males, each arising from an internal gland, not present in females. 6 claspers two-segmented,
tenth tergum elongate, usually simple, lacking additional lobes or processes, tip of copulatory
organ bulbous without special structures.

The genus Macrostemum as defined and delimited by Flint & Bueno-Soria (1982) still re-
mains a large and widespread one of almost world-wide occurrence; it is not, however, found
in Europe.

Macrostemum capense (Walker)

(Figs 88-113, Tables IX, X, histogram Fig. 214B)

Macrostemum capense (Walker), 1852: 77 (as Macronema capensis)

Macrostemum capense (Walker); Betten & Mosely 1940; 194-199, figs 96-99, as Macrone-
ma capense ; synonymy with many other African species and varieties and re-descrip-
tion of imagos (<3, 9) Holotype 9 in British Museum (Nat. Hist.). Port Natal.

Macrostemum capense (Walker); Barnard 1934: 368, 372, figs 42f-g, also larva, described
as? Chloropsyche , p. 368, figs 40h-l.

Macrostemum capense (Walker); Gibbs 1973: 386-389, figs 62, 72, 84—86, imago, larva &
pupa (as Macronema capensis).

Distribution: widely distributed over Southern and Central Africa; includes several colour var-

360

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Figs 88-98. MACRONEMATINAE: Macrostemum capense , 6,9,6 pupa. 88. 6 : dorsal view of head and thoracic
nota, 89. 6 : lateral view of head and palpi, 90, 91 .6: fore and hind wings (right), 92. 6 : dorsal view of genitalia (copu-
latory organ indicated mainly by dotted lines), 93. 9: lateral view of genitalia, 94. 6 pupa: habitus (lateral), 94a. pre-
and postsegmental dorsal plates of segment III, further enlarged, 94b. lateral gill from segment IV, further enlarged
(stem shaded), 95. 6 pupa: apical part of exuviae (dorsal), from slide (not all setae shown). 96. same: apex of right anal
appendage further enlarged, 97, 98. 6 pupa: right and left mandibles (dorsal).

361

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

ieties earlier described as separate species or subspecies. Colour of head and body ranges from
dark brown to yellow, of forewings from dark brown to fawn, yellow or orange, plain or vari-
ously patterned with brown, orange, yellow or white (see Ulmer 1907d, pi. V, for coloured
figures of several varieties); hind wings paler, plain, iridescent. Recorded from South Africa
(all four Provinces), through Zimbabwe, Zambia, Angola, Mozambique to West and East
Africa (Sierra Leone to Zanzibar). Material in the Albany Museum includes specimens from
South Africa (all Provinces), Malawi and Zimbabwe, mainly of typical capense but also includ-
ing the colour variation earlier termed Macronema inscripta Walker, then M. capense var. in-
scriptum, and later placed in the synonymy of M. capense , now Macrostemum capense (Walk-
er).

Description of imagos (Figs 88-93)

Imagos (in spirit. Figs 88-93): dark form, earlier described as Macronema capense var. ca-
pense, now of course Macrostemum capense. Males dark brown in general appearance, with
pubescent brown forewings obscurely marked with deeper brown, hind wings paler. Head and
thorax dark, shining chestnut brown, mesoscutum without warts, mesoscutellum with a pair of
pale warts. Antennae very long, slender, brown, minutely pubescent; maxillary palpi well de-
veloped, fuscous, finely pubescent; labial palpi comparatively small, pale. Legs yellowish
brown, tibial spurs 2.4.4. 8 abdomen with a pair of short filamentous processes on segment V;
genitalia simple, lacking cerci, claspers two-segmented, terminal segment long, strong; copula-
tory organ with large round apex and narrow stem. Females similar to males but paler, with
shorter antennae, and without filamentous abdominal processes on fifth sternum. 9 genitalia
of usual hydropsychid type; sternal plates widely separated medially. Specimens in our collec-
tion appear to be considerably darker in colour than those described in the literature, but
colour in this species is known to be highly variable, and as these are in fluid accurate colour
description is not possible. Gibbs (1973: 387) described M. capense imagos from Ghana (evi-
dently fresh material) as the dark form and as having an orange-brown head, orange-brown an-
tennae with thick grey pubescence, especially basally, dark brown thorax, yellow legs and pal-
pi, wings with pale brown membrane, forewings with short brown pubescence. (Wing length
(d forewing) 9-12 mm, 2 usually smaller. Fig. 110 depicts an imago of M. capense var. in-
scriptum with spread wings.

Description of Larva (Figs 99-109, 111-113, Table IX, Figs. 214B)

Mature larva (in spirit. Figs 99-102, 104—109, 111, Table IX): length up to 16 mm. Fairly
large larvae, brightly coloured, with typically macronematine head, flattened, with carina; pre-
episternum simple, ventral sclerites on IX but not on VIII; no setal fan on anal prolegs.

Head (Figs 99-102): bright chestnut, shortly rounded, with strong carina; frontoclypeal
apotome broadly triangular, with two pairs of slight lateral indentations, clypeal portion partly
sunk over a semicircular area anterior to the epistomal sulcus, frontal portion slightly raised
above level of carina, apex cut off by carina, which traverses it. Dorsal surface of head finely
punctate within the carina in certain areas only (indicated by stippling in Fig. 100), colour al-
most uniform bright red-brown, with a few small paler muscle marks. Anteclypeus with one
pair of sclerites. Each parietal sclerite bears a row of long, stiff setae dorsally, bordering the
carina. Eyes small, black, situated fairly far forward close to carina, within a paler area; no cu-
ticular lenses visible. Antennae minute, close to base of mandibles. Submentum almost semi-
circular, the anterior margin flattened and slightly indented, anterior ventral apotome showing
suture at one side only; posterior ventral apotome minute. Stridulatory files present, large,
curved, the distance between striae greater in the wider distal part than in the narrower proxi-
mal part.

362

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Figs 99-109. MACRONEMATINAE: Macrostemum capense , mature larva (except for Figs 99a & 103). 99. habitus, lat-
eral (right appendages only shown), 99a. ventral gill bearing seta, further enlarged, from 3rd instar larva, 100, 101, 102.
dorsal, ventral & lateral views of head, 103. labrum of 4th instar larva, 104. left and right mandibles (dorsal), 105. fore-
leg with claw further enlarged, also propleural sclerites including pre-episternum, on right, 106, 107. mid- and hind legs,
hind claw shown further enlarged, 108. prosternum, 109. ventral view of abdominal segment IX, with part of VIII and

anal prolegs (somewhat tilted).

363

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Mouthparts (Figs 101, 103, 104): labrum dark brown, retractile, sclerotized, with small
paired lateral membranous expansions each bearing a very large brush of setae, sometimes
tucked in beneath labrum, which is usually strongly retracted and mandibles extended in pre-
served specimens. Mandibles large, broadly three-cornered, heavily sclerotized, with strong,
blunt teeth, mainly in the apical half, the triangular outer side of each mandible set with small
setae below a strong ridge; no inner bristles.

Thorax (Figs 99-111): castaneous, pronotum with blackened lateral margins and posterior
division, meso- and metanota with lateral margins blackened, a medio-dorsal crescent-shaped
black spot near hind margin of former, a smaller oval spot on latter, diagonal brownish stripes
present but faint, no transverse ecdysial lines as in the Diplectroninae. Prosternal plate partly
divided by a sulcus. Mesosternum with one pair of branched tracheal gills, metasternum with
two pairs. Pre-episternum small, bluntly triangular.

Legs (Figs 105-107): strong, brownish yellow. Foreleg: coxa with strong basal process, fe-
mur with basal process on inner side, tibia and tarsus with dense brush of silky white setae,
claw with single slender basal spine. Mid- and hind limb claws apparently with two thick basal
spines, the proximal one small. Ulmer suggests that the proximal basal spine on each claw may
in fact merely be a basal extension of the claw itself (Ulmer 1957: 368-9). In the penultimate
instar it is, however, clearly socketed. The mid-tarsus bears a short, line brush. As Gibbs
(1973: 389) points out, there is a basket-like arrangement of long setae on the trochanter of the
foreleg; this is very clearly seen in our larvae too, and in one or two this contains a fluffy mass
which could have been detritus or algal material collected from the net.

Abdomen (Figs 99, 109): dorsal side covered with short, fine setae and sparsely scattered
longer ones; tufted gills present on abdominal segments II and III in eight rows, on IV-VI in
six rows (see Table XI). Segment IX with one pair of ventral and two pairs of lateral sclerites,
all bearing bristles or setae; the ventral ones are of the usual type, but only occur on IX, there
are none on VIII. Anal prolegs of medium length, stout, with strong lateral sclerite and a few
long setae, no anal fan. Five anal gills.

Earlier instars: The Vaal River material included third to fifth instars of M. capense in ad-
dition to the mature (sixth instar) larvae described above, and the Natal material included all
six instars, there being many specimens of all except the first, of which there were only three,
probably due to their passing through the meshes of the nets used. The instars were separated
as before, on the basis of head width, gill type etc. From the third instar on, the larvae showed
the characteristic head shape and colour seen in the mature larvae, although with slight differ-
ences in gill development (see Table IX) and other features. (See histogram Fig. 214B.)

First instar: head rounded, head capsule approximtely 0,17 mm in width at eyes; labrum
small, with a few setae but no brushes. No gills apart from the five anal gills. Only pronotum
clearly defined and complete, pre-episternum scarcely visible, mesonotum partially and very
faintly sclerotized, metanotum practically unsclerotized. Legs: foreleg with four large feathered
setae on tarsus, claw with three basal spines; mid- and hind claws with three basal spines and a
spinule. Ninth abdominal segment with two large feathered setae where ventral sclerites ap-
pear in later instars. Anal claw plain, curved, with single large seta. (See Figs 112, 113 a-c)

Second instar: head rounded, head capsule approximately 0,26 mm in width at eyes, la-
brum prominent, with strong lateral brushes; no stridulatory files visible. No abdominal or tho-
racic gills. All three thoracic nota well sclerotized; pre-episternum clearly visible. Legs: foreleg
with three feathered setae on tibia, two on tarsus, claw with two basal spines and two spinules
(all four graduated in size); claws of mid- and hind legs with two basal spines and several spin-
ules. No feathered setae or ventral sclerites on ninth abdominal segment, other setae long,
sparse.

364

SCOTT: HYDROPS YCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Table IX.

Number and position of tracheal gills in 3rd to 6th larval instars of Macrostemum capense

Segment

3rd instar

4th, 5th, 6th instars

Mesothorax

1 pair

1

pair

Metathorax

2 pairs

2

pairs

Abdomen

(right side)

Post Pre

VI

V

Post Pre

VI

V

I

1

1

1

—

1

1

1

11

1

1

1

1

2*

2

1

III

1

1

1

1

2

2

1

IV

— —

1

1

1

—

2

1

V

— —

1

1

1

—

2

1

VI

— —

1

1

1

—

1

1

VII

— —

1

—

1

—

1

—

VIII

* Abdominal gills marked 2 each have two main stems, either arising from a common base or separately.
Post = postsegmental lateral gill.

Pre = presegmental lateral gill.

VI = ventrolateral gill.

V = ventral gill.

Third instar: head flattened, carina visible; head width 0,37-0,50 mm at eyes; stridulatory
file and presumed scraper present on head and foreleg respectively. One pair mesothoracic and
two pairs metathoracic gills present, also abdominal gills as in Table IX; all abdominal gills
simple, unbranched, arising from one stem only, each ventral gill bears a long seta. No post-
segmental lateral gills present as yet. Forelimb claw with one slender basal spine; mid- and
hind claws with two thick basal spines, the proximal one very small; thin brush of setae on
forelegs, no feathered setae found; abdomen smooth, paired ventral sclerites now present on
IX. (See Fig. 99a and Table IX)

Fourth instar: head capsule 0,69-0,81 mm wide at eyes; abdominal gills as in mature larva,
postsegmental lateral gills present, gills that will develop double stems have done so, but gills
only bear 1-4 branches; ventral gills without setae. Fore claw with slender basal spine, mid-and
hind claws with two thick basal spines, the proximal one small, as in later instars. Abdomen
smooth. Brush on foreleg denser than in third instar. (See Fig. 103, labrum)

Fifth instar: head capsule 1,06-1,38 mm wide at eyes; abdominal gills as in mature larva,
usually with 4—8 branches. Brush on foreleg fairly well developed. Abdomen hairy in appear-
ance, as in final instar. As mentioned above, the proximal basal spine of each limb claw is
clearly socketed.

Sixth instar: this is the mature larva, described above. Head capsule 1,56-1,94 mm in
width at eyes; gills with 15-23 branches.

Description of pupa (Figs 94-98, Table X)

Material available included a number of mature male and female pupae with cases and lar-
val sclerites, thus establishing the correlation with M. capense imagos. Length of pupae 9-10
mm. Mature larvae were present in some instances, substantiating the correlation.

365

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Figs 110-115. MACRONEMATINAE: Macrostemum capense. 110. M. capense var. inscriptum: 6 imago with wings
spread (legs omitted). 111. M. capense'. pro-, meso- and metanotal shields of mature larva (larval sclerites from pupal
case), 112. M. capense : first instar larva, whole mount, 113. M. capense: first instar larva, claws (drawn freehand,
x 400): a. claw of foreleg, b. claw of mid-leg, c. anal claw. MACRONEMATINAE: Amphipsyche scottae. 114. meso-
thoracic claws of larvae: a. first instar, b. second instar, c. third instar., 115. dorsal view of pupal case, showing end
plates on left, further enlarged, to show respiratory holes.

366

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Male pupa (pharate imago — in spirit) (Figs 94-98, Table X): head rounded, brownish yel-
low, surmounted by two sparse flattened tufts of brown setae with bent tips, anterior ones pro-
jecting forwards, posterior ones backwards. Eyes large, dark; antennae long, slender, curled
round end of abdomen (two coils per antenna). Labrum semicircular, flattened; mandibles
small, each with an apical and three lateral slightly serrated teeth (proximal one small, distal
ones large, in left mandible second tooth from base has two cusps), strong setae along the out-
er margin. Maxillary and labial palpi well developed. All legs fairly stout, middle pair fringed
to form oars; spurs 2.4.4. Tufted gills and dorsal plates present as indicated in Table X. Lateral
gills on segments IV-VI each with a swollen pouch-like base from which arises a single, long,
much-branched gill (Fig. 94b). Anal appendages of medium length, strongly curved dorsad,
plump, with small “bare” finger-like apices (these lack long setae but are minutely spinose),
apart from the apices the appendages are heavily fringed with strong setae, some laterally but
most ventrally placed. Male genitalia enclosed in ventral pockets as usual.

Female pupa (pharate imago — in spirit): similar to male but antennae shorter, recurved
not coiled; middle swimming legs more strongly broadened; anal appendages similar but lack-
ing pockets for genitalia. Gill count as in male, but pouches at bases of dorsolateral gills less
pronounced; dorsal plates, setae etc. similar.

Pupal case: length 10-12 mm; a straight, rigid case of variously sized sand grains, mainly
medium to large in size, lower edges inturned where case rests on substratum, completely lined
with secretion except at ends. Secretion forms a loose inner lining, cocooning the pupa. Poste-

Table X.

Branchial gills, dorsal plates etc. of 6 pupa of Macrostemum capense

Abdominal
segment no.

Tufted gills
(right side)

Dorsal plates

Setae, sclerotised points
etc. on dorsal side of abdomen

I

D1

VI

V

2-3 pairs setae

II

1

2*

2

—

4 small pairs setae; scattered

III

(1)

2

2

1 ant. pair

points over whole
4 pairs setae; scattered points

IV

(1)

_

2

1 post, pair
1 ant. pair

over whole

Band of long setae along

V

(1)

2

1 ant pair

post. margin; scattered
points over whole
Scattered points over whole

VI

(1)

—

1

1 ant. pair

Scattered points over whole

VII

1

—

1

1 ant. pair

—

VIII

—

—

—

1 ant. pair

—

*The figure 2 denotes that the gill in question has 2 main stems.

(1) The brackets denote that the gill in question has a swollen pouch-like base, from which arises a single long, much
branched gill.

D1 = dorsolateral gills.

VI = ventrolateral gills.

V = ventral gills.

367

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

rior end of case rounded, of smaller grains connected by perforated sieve plates, anterior end
flatter, with lid of small sand grains between which are several holes guarded within by sieve
plates. Larval sclerites tucked into posterior end of case.

Remarks

Over the years there has been considerable confusion in regard to the nomenclature of
Macronema species in Africa, largely on account of the great variability in colour and small
amount of difference in other characters between the specimens being studied. Specimens have
frequently been made new species which have later proved to be colour variations of existing
species. Names have changed in rank from species to subspecies or variety and back and many
have later fallen into synonymy. The list of African species given below reflects this.

Although this background and position remain unaltered, there has been a recent change
in the generic name of all African species from Macronema Pictet to Macrostemum Kolenati
(Flint & Bueno-Soria 1982). As Flint and Bueno-Soria point out in their paper, it had long
been recognized that Macronema species fell into two groups (e.g. Ulmer 1907d, 1957 &
others). They had defined differences in an earlier paper (Flint & Bueno-Soria 1979) in which
they separated the Neotropical species of Macronema into two groups, the hyalinum group and
the percitans group, based on clear differences in adult structure, particularly in wing colour
and in shape and complexity of male genitalia. Whereas larvae of a number of species in the
hyalinum group were well known, no larvae had then been correlated with imagos of any
species of the percitans group. Thus the matter stood until Bueno-Soria was able to collect and
rear some Macronematine larvae of a type previously attributed to the genus Centromacrone-
ma by Marlier (1964), a very different larva from the well-known “ Macronema ” larva. These
larvae proved to be those of Macronema variipenne , a member of the percitans group, the lar-
vae of Centromacronema thus remaining unknown.

Flint & Bueno-Soria (1982) point out that variipenne larvae are so different from those of
previously described Macronema larvae (exemplified by their having been attributed to Centro-
macronema) that the two divisions of Macronema must be regarded not as species groups but as
different genera. The percitans group includes both M. variipenne and M. lineatum , the latter
being the type species of the genus Macronema. The percitans group (which is exclusively
South American) therefore retains the name Macronema. A new name has therefore had to be
found for the hyalinum group which comprises some South American species, all African
species and all species from the rest of the world. To this group Flint & Bueno-Soria (1982)
have allocated the name Macrostemum Kolenati 1859, of which the type species is Hydro-
psyche hyalina Pictet, selected by Fischer (1963: 176). Macrostemum had been listed as a syno-
nym of Macronema by Fischer (1963: 176) so was available for the hyalinum group.

Flint & Bueno-Soria (1982, table 1) give a useful comparison of the larvae and pupae of
Macronema and Macrostemum species. The food and probably the net of South African
Macrostemum larvae, however, evidently differ considerably from those of South American
ones (see under Biology).

Larvae and pupae of non- African species of Macronema (i.e. Macrostemum) have been
described by a number of authors, mainly from American and Asian material. Of African
species, the only valid descriptions to my knowledge are those of M. capense larvae and pupae
by Gibbs (1973) and of an unidentified species of Macronema by Barnard (1934) as Chloro-
psychel which was also M. capense. Descriptions of Macronema larvae were also given by Ul-
mer (1912) and Marlier (1943b) as Aethaloptera. Marlier (1965/6) described a pupa but gave no
figures.

Neither larvae nor pupae show clear interspecific differences, though it might well be poss-
ible to find reliable characters should sufficient material of several species be available for
study. Comparison of M. capense larvae with figures and descriptions in the literature indicates

368

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

that shape of mandibles differs from one species to another and that colour differences may
also exist. Statzner (1981) separates four species of Macronema (i.e. Macrostemum ) on the ba-
sis of shape of head and submentum and differences between stridulatory ridges and setae on
the ventral sclerites of IX. These differences are regrettably not illustrated.

Barnard (1934: 373) pointed out that M. natalense Ulmer (1931) only differed from
M. capense Walker in the length of the fifth segment of the maxillary palpi relative to that of
the other four segments, a difference which seems minimal in available material, if indeed
present. The recorded distribution of natalense coincides with part of that of capense and the
larvae appear to be identical (comparison of specimens in the Albany Museum collections,
considered on a geographical basis). M. natalense is a dark form; capense has dark forms, inter
alia. I therefore feel that natalense should be placed in the synonymy of capense (Syn. nov.).

Biology

In contrast to most of the other Macronematinae, much detailed and elegant work has
been carried out in various parts of the world on the biology and behaviour of Macronema ( =
Macrostemum ) species, particularly by the late Dr Werner Sattler on larvae living in forest
streams in the Amazon basin (Sattler 1963, 1965); see also Wallace (1975, 1976) and Wallace
and Sherberger (1974, 1975), working on North American material, and others.

Macrostemum larvae tend to inhabit large rivers and the larger streams, living in sandy or
gravelly bottoms, on or under rocks and large stones, or in and on the bark of submerged trees
or branches. There they make beautiful and complex retreats, varying somewhat in detail
according to species and site, but usually incorporating a U-shaped tube, comprising a longer
funnel-like tube, facing upstream, down which the water enters, and a second, shorter tube for
its exit. At the bend of the U a very fine-meshed net is placed (mesh openings 2-4 p. x 14 — 32/ll
in Amazonian species), in which incoming food particles are trapped, and just anterior to
which a third tube opens, which evidently houses the larva, enabling it to collect food from the
net. The net traps microseston (fine particulate organic matter, phytoplankton and bacteria) in
contrast to the much larger organisms commonly taken by most other Hydropsychidae in their
coarser nets. The tubes are made of silk secretion in which embedded sand grains act as stiffen-
ing. Dense brushes on forelegs and labrum and setose tracts within the oral region enable the
Macrostemum larvae to sweep the food particles collected by the net into the mouth. When not
feeding the larva can block the opening of the dwelling tube with its flat head.

When the tubes are situated in sandy bottoms they are vertically placed with the arms of
the U-tube projecting into the current and the dwelling tube penetrating into the substratum.
When under stones, as was the case with some South African specimens that I collected, the
tubes were horizontal, facing upstream, and usually situated between a large stone and the
substratum whether gravel or underlying rock or stone. Thus they did not need to be as com-
plete and the walls were only partly of sand or gravel and silk and partly formed by the shel-
tering stone and the substratum. Gibbs (1973) found similar retreats under stones in a river in
Ghana where the anterior tube was roughly made of rather large stones.

The South African stream-dwelling species, M. capense , evidently has a different feeding
pattern from that of South American species studied by Sattler. Chutter (1968), working on
the ecology of Simulium species in the Vaal River (South Africa), also studied their predators
which included several hydropsychid genera of which Macrostemum (as Macronema ) was one.
The Macrostemum larvae had a strongly toothed gizzard and large amounts of insect sclerites
were found in their guts. He concluded that large M. capense larvae were mainly carnivorous
and would probably readily eat Simulium larvae and, of course, other insect larvae.

M. capense imagos and pupae, in the Albany Museum material, were collected from Sep-
tember through to April (thus not in the winter months) and larvae all the year round.

369

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

African species of Macrostemum (all described as Macronema)

M. adpictum (Navas), 1934c: 70, 71, fig. 12. $ .

Distribution: Madagascar.

M. alienum (Ulmer), 1907d: 107-108, fig. 109. d 9.

Distribution: Sudan, Ghana.

M. bouvieri (Navas), (1922) 1923: 49, 50, fig. 16. d.

Distribution: Congo.

M. capense (Walker), 1852: 77. d 2 .

Synonyms: M. capense var. signatum Walker, 1852: 77.

M. capense var. inscription Walker, 1852: 77.

M. capense var. pulcherrimum Walker, 1852: 77.

M. sansibaricum Kolbe, 1897: 39.

M. sjostedti Ulmer, 1904b: 419, fig. 9.

M. capense var . rhodesianum Ulmer, 1931: 14, 15, fig. 12.

M. capense var. stigmatum Navas, 1930a: 331, fig. 46.

M. fulva Jacquemart, 1961b: 19.

M. natalense Ulmer, 1931: 12, fig. 11. Placed in the synonymy of M. capense in
this paper (Syn. nov.); see above under Remarks.

See Fischer 1963: 179-181 for full references. Most of the above varieties have been
treated as full species at times. All now fall into synonymy.

Distribution: widely distributed in Africa, from Sierra Leone and Ghana in the West to
Uganda and Zanzibar in the East, down to South Africa (all four Provinces), and in-
cluding the Comoros, Fernando Po and Sao Tome.

M. colini (Navas), 1932a: 289, fig. 86. $ . Belgian Congo.

Marlier (1960) states that this is undoubtedly a species of Dipseudopsis (Dipseudopsidae).

M. displicens (Navas), 1935: 73, 74, fig. 34. d.

Distribution: Madagascar.

Sykora (1964: 281, figs 13-16) redescribes the species and transfers it to the genus Lepto-
nema.

M. distinction (Ulmer), 1912: 100-102, fig. 22. d.

Distribution: Spanish Guinea, Ghana.

M.furcata (Jacquemart & Statzner), 1981: 15, pis X figs 1, 2, XV fig. 9. d.

The wings illustrated resemble those of a Protomacronema rather than those of a Macro-
stemum.

Distribution: Zaire.

M. graphicum (Navas), 1934c: 69, 70, fig. 11. d.

Distribution: Madagascar.

M. lacroixi (Navas), 1923: 25, 26, fig. 36. d .

Distribution: Madagascar.

M. madagascariense (Ulmer), 1905b: 31, fig. 18. d.

Distribution: Madagascar.

M. nigriceps (Navas), 1932a: 290, fig. 87. 9. Congo.

Marlier (1960) notes that this is a mistaken identification of Protodipseudopsis congolana
Navas (Dipseudopsidae).

370

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

M. obscurum (Banks), 1920: 355. (Sex not stated.)

Distribution: Madagascar.

M. placidum (Navas), 1935: 72, 73, fig. 33. 8.

Distribution: Madagascar.

M. pseudodistinctum (Marlier), 1965/66: 22, 43, 44, figs 14a, b. 8 .

Distribution: Angola.

M. scriptum (Rambur), 1842: 507. Sex not given but evidently 8 . Holotype in Coll. Selys.
Ulmer 1907d: 108, 109, fig. Ill (wings). 8.

Distribution: Madagascar.

M. subinscriptum (Jacquemart), 1966b: 42. Listed from the Congo.

I can find no reference to a description. If there is none, it is a Nomen Nudum.

M. trilineata (Jacquemart), 1961a: 1-3, figs 1,2. 2 .

Jacquemart 1966a: 8. 8 .

Distribution: Congo (9), Katanga (<?).

Macronema sp. Marlier, 1978a: 292-294, fig. 4 (2 wings & genitalia). 2 . (= Macrostemum)
Distribution: Mali.

Described as a new species but not named, awaiting 8 .

Genus AMPHIPSYCHE McLachlan 1872

Amphipsyche McLachlan, 1872: 68, pi. 2, figs 7-7e.

Type species Amphipsyche proluta McLachlan, 1872. Amur.

Phanostoma Brauer, 1875: 69, pi. 4, figs 5-5e.

Type species Phanostoma senegalense Brauer, 1875. Senegal.

Phanostoma Brauer; Barnard 1934: 369.

Amphipsyche McLachlan; Kimmins 1962a: 84-85, synonymy and discussion; key to then known
African species (males).

Amphipsyche McLachlan; Lepneva 1964 (1970): 604-617, description of larval and pupal stages.
Generic diagnosis (derived from McLachlan 1872, Ulmer 1907d & Kimmins 1962a, 1963)
Imago: Spur formula d, 2, 1.4.4, 1.4.3, 1.4.2, 0,4.3, 0.4.2, 0.3.2 or 0.2.2. Mid-tibia and first
three basal tarsal segments moderately flattened in 8 , very strongly so in 2 , fringed with se-
tae. Antennae very long, scape bulbous, flagellum threadlike, about 2\ times forewing length
in 8 , shorter in 2 ; eyes rather small, face somewhat inflated, vertex with paired anterior warts
very large. Maxillary palpi weak, rather short, five-segmented, labial palpi very small, three-
segmented, segments two and three subequal, not clearly separated. Forewing long, rounded
apically, excised in postcostal area proximal to arculus, particularly in 8 ; wing forks 1, 2, 3, 4,
5 present, several false costal cross-veins in 8, not in 9; discoidal cell absent, median cell
closed, short, a false discoidal cell present or absent. Hind wing subtriangular, very wide at
base in 8 , less so in 2 , with forks 2, 3, 5, discoidal and median cells absent, hooked macrotri-
chia present along costa; folded when at rest. 8 and 2 wings differ in shape and venation
(Kimmins 1962a gives figures of both sexes). Wings colourless to pale green or pale brown,
naked. Length of 8 forewing 11-18 mm, of 2 9-14 mm, in African species. 8 genitalia simple,
tenth tergite divided, copulatory organ with or without spiniform “parameres” (endothecal
spines); claspers slender. 2 smaller than 8 . 9 genitalia with sternal plates of VII widely separ-
ated medially, as in Macrostemum.

The genus occurs not only in Africa, but also in Russia, China and the Oriental Region as
far south-east as Borneo.

371

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Amphipsyche scottae Kimmins
(Figs 114-136, Tables XI, XII, histogram Fig. 214C)

Amphipsyche scottae Kimmins, 1962a: 93-94, figs 29-37 (3 palpi, 3 , $ wings and genita-
lia).

Holotype 3, Wilge River (tributary of Vaal), 5 miles downstream from Harrismith,
10. ii. 1959; Allotype 9, Vaal River below Barrage, 23.viii.1957, both in British Mu-
seum (Nat. Hist.). Paratypes in British Museum (Nat. Hist.), Albany Museum and
South African Museum collections.

Amphipsyche scottae Kimmins; Scott 1975: 48, figs 35, 36 (larva).

Distribution: this species appears to be endemic to South Africa, where it has been recorded
from the Transvaal, Orange Free State and Northern Cape, always from the Vaal River and its
tributaries; it has also been found in the Mooi River, Natal.

Kimmins' material all came from the National Institute for Water Research collections,
(now housed in the Albany Museum and forming part of their Collection of Freshwater Inver-
tebrates), and the rest of his specimens are now housed in the Albany Museum together with
further material from the same areas. Additional material is extensive, including male and fe-
male imagos, larvae and pupae; almost all are from the Vaal River System, a few from the
Mooi River, Natal.

Description of imagos (Figs 116-121)

Imagos (in spirit. Figs 116-121): large, cream to tawny yellow in colour with glassy wings;
hind wings very wide, males more brightly tinted than females. Antennae finely annulated with
red-brown at joints, very slender. Maxillary and labial palpi present, small. Legs with small
black spots on each trochanter and at apical end of fore and mid-tibiae. Mid-legs of male
somewhat flattened, of female strongly flattened and expanded, fringed. Tibial spurs 0.4.2.
Wings slightly clouded, iridescent, membrane covered with microtrichia and, particularly in <3,
with a sparse scattering of larger setae; setae also present along some wing veins. 3 forewing
length c. 18 mm, $ c. 14 mm. Vertex with pair of large anterior warts separated by a shield-
shaped raised median area, representing the dorsal triangle; posterior warts narrow. Mesoscu-
tum and mesoscutellum without warts. Females almost entirely glabrous, males with a fuzz of
pale hairs on parts of thorax and legs. Abdomen reddish purple. Male genitalia are simple, the
claspers slender and the copulatory organ has a pair of “parameres” (Kimmins and other
authors) inserted into its dorsal side. According to Schmid (1979 and other papers), these are
not true parameres but neoformations, endothecal spines.

Description of larva (Figs 126-136, Table XI, Fig. 214C)

Amphipsyche larvae are very characteristic in appearance, the larger instars always being
easily recognizable from the head shape and pattern. Even young larvae of A. scottae are often
separable from other Hydropsychidae by the comparatively large, broad, flat brown head with
paired whitish marks, the large humped labrum (occasionally retracted, in which case the man-
dibles are usually everted), and the relatively short, rapidly narrowing abdomen.

Mature larva (in spirit. Figs 126-136, Table XI): length up to 12 mm, width of abdomen
up to 3 mm.

Head (Figs 126-128): short, rounded, frontoclypeal apotome and surrounding area flat-
tened and bounded by a carina; top of head mainly dark brown in colour, dark colour extend-
ing slightly beyond carina except where indented by yellowish areas round eyes; paired yellow-
ish patches on frontoclypeal apotome. Head not uniformly dark brown, but spotted with
minute yellow pits from each of which arises a small seta. Carina runs exactly through apex of
frontoclypeal apotome, which is short and widely triangular; setae sparse, small, pale, apart

372

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Fig^ 116-125. MACRONEMATINAE: Amphipsyche scottae , <3, 2 pupa, 8 pupa. 116. <3 : dorsal view of head & tho-
racic nota, 117. 3 : lateral view of head and palpi, 118, 119. 3: fore and hind wings (right), 120. <3 : lateral view of geni-
talia, 121. 8 : copulatory organ (dorsal); “parameres”, shown by transparency, lie mainly within the copulatory organ on
dorsal side; the apical internal sclerites lie below & between them, 122. 2 pupa: habitus, lateral, 122a. pre-segmental
dorsal plate from V & postsegmental dorsal plate from III (part), 122b. ventrolateral and dorsolateral gills from III & V
respectively, 123. 2 pupa: left & right mandibles, dorsal, 124. 8 pupa: apical part of exuviae (ventral) from slide (not
all setae shown; long setae practically all on ventral side, 125. same: apex of right appendage further enlarged.

373

ANN. CAPE PROV. MUS. (NAT. HIST. ) VOL. 14 PT 8, JUNE 1983

from two pairs of longer, dark ones near anterior margin. Eyes (clusters of ocelli), black,
placed well forward, beneath a few small cuticular lenses, not easily seen. Antennae minute, at
base of mandibles. Parietal sclerites yellow laterally, brownish ventrally, patterned with paler
muscle spots; stridulatory files triangular, anterior margin of head capsule heavily sclerotized,
dark. Anterior ventral apotome represented by a small sclerite, with cleavage line on left side
only, posterior ventral apotome smaller.

Mouthparts (Figs 129-131): labrum brown, largely sclerotized, with membranous anterola-
teral lobes, anterior margin with long tufts of pale, silky setae, dorsal surface with small sec-
ondary setae in addition to the primary ones. Labrum retractile, when fully retracted scarcely
visible; membranous base of labrum (anteclypeus) bears four brown sclerotized plates. Man-
dibles broad, powerful; right mandible with four apical teeth, narrowing below, then widening
midway to the base where a blunt, three-cusped lateral tooth arises; left mandible with five api-
cal teeth and two lateral teeth, the latter with two or three cusps each, no inner brushes; man-
dibles triangular in cross-section, as is usually the case in hydropsychids, the outer surface se-
tose, enclosed by lateral flanges. Maxillo-labium: palpifers with strong tufts of setae, maxillary
palpi stout, five-jointed, galeae narrow, tipped with sensilla; submentum triangular, brown,
bearing curved setae; labium shorter than maxillae, labial palpi minute, not easily distin-
guished.

Thorax (Fig. 126) thoracic segments short, little narrower than abdomen; pronotum with
paired sclerites, brownish, with narrow darker anterior and posterior borders and black lateral
thickenings, a transverse sulcus near posterior margin; meso- and matanota undivided, brown-
ish with black lateral margins and darker diagonal stripes, a small central black mark near
posterior margin of each. Prosternal plate narrow, dark; meso- and metasterna membranous;
one pair of tufted gills present on mesosternum, two pairs on metasternum. Propleuron with
pre-episternum (first pleural sclerite) simple, shortly pointed, bearing a few bristles (see Fig.
132).

Legs (Figs 132-134): comparatively short, strong, yellowish to brownish in colour, with
darker markings as indicated in figures, anterior pair (particularly tibiae and tarsi) darker than
rest, claws brown. Legs with many secondary setae. Forelegs: coxa with stout, striated basal
projection which articulates with propleuron, femur with short, blunt projection on inner side
of proximal end, this appears to have no function unless it serves as scraper, as I could find no
other in the usual place along the anterior margin of the femur. Tibia and tarsus have a weak
brush of hairs, some finely plumose, appearing serrated, and a short fringe of stout setae along
the posterior margin. Limb claws each with a single short stout spine at base.

Abdomen (Fig. 126): not deeply segmented, dorsal surface largely covered with minute,
blunt microchaetae, occasional longer setae and a few very long ones, the microchaetae be-
come smaller laterally, disappearing at the pleura. They are usually more abundant in the in-
tersegmental folds, and may be lacking in patches dorsally, for example along transverse, in-
dented bands on segments I and II, which instead are set with bent spinules several times as
long as the microchaetae. Tufted tracheal gills are present in six rows on segments I and VII,
eight rows on segments 1 1- VI, and two rows on segment VIII, there are none on IX. These
gills are borne on the lateral, ventrolateral and ventral surface of the abdomen (see Table XI),
lateral gills have both presegmental and postsegmental tufts. Gill filaments arise from the sides
and apex of a common stalk (in thoracic gills along one side only). On segment IX there is a
pair of ventral sclerites bearing spines, also a small pair of lateral sclerites and a larger pair of
dorsolateral sclerites, both setose; there are no such sclerites on VIII. Anal prolegs short,
stout, with strong lateral sclerites, anal claws large, incurved; five anal blood gills.

Earlier instars: The differences between the larval instars were noted as in the other gene-
ra, and they appeared to fall naturally into six groups, which have been taken as representing

374

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Table XI:

Number and position of tracheal gills and other characters in 3rd to 6th larval instars of

Amphipsyche scottae

Segment No.

3rd instar

4th instar

5th instar

6th instar

Mesosternal

1 pair

1 pair

1 pair

1 pair

gills

Metasternal

gills

2 pairs

2 pairs

2 pairs

2 pairs

Abdominal

gills

Post Pre VI

V

Post Pre VI V

Post Pre VI

V

Post Pre VI

V

I

— — 1

1

— — 1 1

— — 1

1

1 1

1

11

1 1

1

1 2* 2 1

1 2 2

1

1

2 2

1

III

1 1

1

12 2 1

1 2 2

1

1

2 2

1

IV

— — 1

1

1 12 1

1 1 2

1

1

1 2

1

V

— — 1

1

1 12 1

1 1 2

1

1

1 2

1

VI

— — 1

1

1 12 1

1 1 2

1

1

1 2

1

VII

— — 1

—

1 2 —

1 1 2

—

1

1 2

—

VIII

— — 1

—

- - 2 -

— — 2

—

- — 2

Tarsal

claws

Gills

2 spines
(1 minute)

unbranched

2 spines
(1 minute)

c. 2-4 branches

1 spine

c. 6-10 branches

c.

1 spine

16-24 branches

Head width
at eyes

0,5mm

0,75-0, 80mm

1,1mm

1,5-1, 7mm

* The figure 2 in this section of Table I indicates that the abdominal gill in question has 2 main stems, either arising
from a common base or separately.

Post = postsegmental lateral gill; Pre = presegmental lateral gill; VI = ventrolateral gill; V = ventral gill.

six larval instars, as was the case in Macrostemum capense. Only 69 larvae were available for
study, so the histogram derived was not as good as that for Cheumatopsyche , nevertheless it
tallied with the other larval features observed, see Fig. 214C (histogram).

First instar: head rounded, neither flattened nor with carina, uniformly pale brown in
colour but with the characteristic extensile labrum; thoracic nota very lightly chitinized, scarce-
ly darkened marginally; tarsal claws of all legs with three large basal spines, several large feath-
ered setae on tibiae, tarsi and femora (Fig. 114a); no tracheal gills on thorax or abdomen. Av-
erage head width of available specimens at eyes 0,17 mm.

Second instar: similar to first, but head darker, sometimes faintly patterned; thoracic nota
more heavily sclerotized, pronotum strongly darkened marginally; tarsal claws of all legs with

375

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Figs 126-136. MACRONEMATINAE: Amphipsyche scottae, mature larva. 126. habitus, lateral (right appendages only
shown), 127, 128. dorsal & ventral views of head; small portion of surface shown further enlarged, 129. left & right
mandibles, ventral view, 129a. right mandible, lateral view, 130. labrum and anteclypeus, dorsal, 131, labium, ventral,
132, 133, 134. foreleg, mid- & hind legs (left); foreleg with pleural sclerites attached & serrate seta further enlarged,
132a. fore-coxa & part of femur, median view; process of femur further enlarged, 135. ventral view of abdominal seg-
ment IX, including anal prolegs, 136. right anal proleg, lateral.

376

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

two large basal spines (Fig. 114b), large feathered setae on tibia, tarsus and femur of foreleg,
less strongly feathered setae on tibiae and tarsi of mid- and hind legs; files of stridulators not
yet visible on head, but basal processes present on anterior femora and coxae; no tracheal gills.
Average head width 0,33 mm.

Third instar: carina sometimes recognisable on head, but dorsal surface still convex,
characteristic colour pattern faint but discernible; stridulatory files present; tarsal claws of all
legs with one large basal spine and one minute one (Fig. 114c); normal setation of the type
found in the final instar developing, strongly plumose setae absent, but some setae on tibia and
tarsus with very fine, short feathering (“serrated”), as in later instars. Simple tracheal gills
present on both thorax and abdomen (nearly all unbranched), no postsegmental gills present
and few presegmental gills, so that gill count is less than in later instars (see Table XI): all gills
arise from one stem only, and each ventral gill bears a subterminal seta, as in Macrostemum.
Average head width 0,5 mm.

Fourth instar: typical flattened frontoclypeal apotome, colour pattern strong (whole area
bounded by carina is not completely flattened even in mature larvae); full number of gills pres-
ent except for presegmental gill on segment I and postsegmental gill on VII, but gills usually
only have one to four branches, although nearly all those that will develop double stems have
already done so; forelegs more strongly setose, but as yet no proper brush; tarsal claws as in
third instar. Average head width 0,75-0,80 mm.

Fifth instar: full number of gills present except for presegmental gill on segment I; gills
more fully branched, usually having from six to ten branches; a thin brush present on forelegs,
tarsal claws with one basal spine only, as in mature larva. Average head width 1,1 mm.

Sixth (final) instar, fully described above: all gills present and fully branched (usually
16-24 branches per gill). While the actual number of gills per segment appears to be constant,
the number with two main stems seems to be slightly variable, and an occasional larva may
have from one to three more gills with double stems than is indicated in the table. Forelegs
with full brush; this is not, however, as thick as in, for example, Macrostemum or Aethalopte-
ra. Average head width 1,5-1, 7 mm.

Description of pupa (Figs 115, 122-125, Table XII)

Female pupa ( in spirit; Figs 122, 123): length 9,5 mm. Head somewhat flattened, vertex
dark brown surmounted by four flattened tufts of dark setae with bent tips, of which the two
smaller tufts project forwards, the two larger backwards; each seta arises from a white circle,
giving the head a spotted appearance. Eyes large, dark; antennae very slender, about H times
length of pupa, not curled round body but loosely recurved; labrum short, rounded, with three
setae at each side; mandibles strong, with large serrated teeth (left mandible with five teeth, in-
cluding apical teeth, right with four), tufts of setae on outer side near base and a few setae
near outer edge of dorsal face; maxillary and labial palpi small but present. Forelegs fairly
slender, mid-legs much broadened, flattened and fringed with setae to form swimming oars;
hind legs very slender. Tibial spurs 0.4.2. Tufted gills and dorsal plates present on abdomen as
indicated in Fig. 122 and Table XII. Gill branches are slightly variable in size and number, and
branches are easily broken off (gill numbers are a more reliable character in larvae than in pu-
pae). While the dorsolateral and ventrolateral gills are clearly separate on segments II and III,
where the ventrolateral gills are double, from segment IV caudad they draw closer and closer
together until on segment VII both could be branches of the same gill. On VIII the dorsolate-
ral gills are definitely two-branched and there are no others. Those dorsolateral gills on seg-
ments III-VI arise from broad pouch-like bases, from which may arise one, two or three differ-
ent stems bearing gill filaments. Anal appendages fairly short, bluntly rounded, directed
somewhat dorsad, bearing strong, dark setae on the ventral side.

377

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Male pupa (pharate imago, in spirit. Figs 124, 125): as female, but mid-leg narrower, less
expanded, though still fringed and oar-like; antennae much longer than in female, and more
slender, ends forming a coil on the posterior half of the dorsal side of the pupa, four turns of
each antenna in the coil; dorsal plates and gills as in female. Anal appendages less rounded
(Fig. 124), with pockets for male genitalia, more heavily setate, ending posteriorly in small ver-
tical plates fringed with spinules.

Pupal case (Fig. 115): oval, constructed of variously sized sand grains (mainly large), lined
with silk; sides basally inturned to rest on substrate, case closed ventrally with silk only; head
end of case (cover) flattened, edged with closely set sand grains which surround a silken lid in
which are set a number of sand grains, between which are several small slits for the passage of
water; posterior end of case rounded, with about six slits between larger, more closely set sand
grains. Case rigid, sand grains very firmly cemented together, completely lined with silk except
at ends. Larval sclerites tucked into posterior end of case as is usual.

Table XII:

Tracheal gills, dorsal plates etc. of 9 pupa of Amphipsyche scottae

Abdominal
segment no.

Tufted gills
(right side)

Dorsal plates

Setae, points (shagreening)

I

D1

VI

V

irregular band of setae

11

1

2*

2

—

broad anterior band of scattered points;

III

(1)

2

2

1 anterior pair

narrow posterior band of scattered
points

scattered points anteriorly; band of points

IV

(1)

1

2

1 posterior pair
(strips)

1 anterior pair

posteriorly

anterior patch and posterior band of long

V

(1)

1

2

1 anterior pair

setae

VI

(1)

1

2

1 anterior pair

VII

(1)

1

2 or

1 anterior pair

VIII

2

2

2

1 anterior pair

* The figure 2 denotes that the gill in question usually has 2 main branches or stems.
(1) Brackets indicate that the gill in question has a broad, pouch-like base.

D1 = dorsolateral gill; VI = ventrolateral gill; V = ventral gill.

Remarks

Ulmer distinguished between the larvae of Amphipsyche and Phanostoma on the basis of
length of tarsal claws compared with tarsus (Ulmer 1963). In view of the combination of the
two genera by Martynov (1935) and Kimmins (1962a) this difference would be specific rather
than generic.

378

SCOTT': HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

A comparison of the larvae of Amphipsyche scottae with the descriptions of the larvae of
senegalensis (Brauer) as given by Hickin (1955) and Jacquemart (1957) and with those of the
presumed larvae of curvinerve Navas given by Ulmer (1963), of meridiana Ulmer (Ulmer 1957)
and of proluta McLachlan (Lepneva 1964 (1970) ) indicates that whereas they are very similar
one to another, as might be expected, there are noticeable though often small differences in
such characters as the mandibular teeth and the colour pattern and basic colour of the head.
The basic colour of the head may be brown, as in scottae and senegalensis , or yellow. In curvi-
nerve the head is yellow and lacks the typical pair of pale spots; proluta also lacks these and
has a paler transverse stripe. The setose ventral plates on the last abdominal segment may dif-
fer somewhat in shape from species to species. Similarly the labrum may differ in shape and
extent of sclerotization. There also appear to be slight differences in gill counts (where these
are available) and in the suture or sutures shown as bounding the anterior ventral apotome lat-
erally (though this may reflect inaccurate observation in some cases).

Only the central part of the transverse bar in senegalensis , tentatively called the gular scle-
rite by Hickin (1955), represents the anterior ventral apotome. This is small and triangular
with a suture on the left side only (Fig. 128). Marlier (1943b) described as Aethaloptera sp. an
Amphipsyche larva, from the Parc National Albert, which could be senegalensis .

Kimmins (1962a) suggested that Ulmer’s curvinerve , described from females only, might
be synonymous with senegalensis , for though their wing venation differed from that of the male
senegalensis , normal in this genus, it was similar to that of the females assigned to senegalensis
on a locality basis. Ulmer’s description of the probable larva (Ulmer 1963) is not of senegalen-
sis but might well prove to be that of ulmeri not curvinerve larvae. In the Albany Museum col-
lection there are two larvae (and an immature pupa), very kindly sent to me by Mr Gibbs from
his Ivory Coast material, which may be of ulmeri. They are fourth or fifth instar and have some
gills missing. They correspond to some extent to Ulmer’s colour description of curvinerve as
the heads are yellow to brownish yellow, however, they are not entirely unpatterned, as Ul-
mer’s appear to have been, but, as in proluta , show the two paler marks faintly on the fronto-
clypeal apotome and a pale stripe close to the anterior margin. The gill count is different from
the other known species. These larvae are evidently neither senegalensis nor scottae and could
thus well be those of ulmeri as Gibbs considered likely (Gibbs 1973). To establish the facts
with certainty it would be necessary to correlate the larvae of both kinds (ulmeri and curvi-
nerve) with their male and female imagos (and incidentally establish the status of curvinerve).

The presumed larvae of senegalensis in the Albany Museum collection are more like the
scottae larvae than 1 would have expected from the published descriptions. However, they do
differ in having narrower setose plates on the ninth abdominal sternum, in the shape of the
mandibles, and in that the brown coloration of the head and nota lacks the slight reddish tinge
seen in scottae. Neither account of the larva of senegalensis gives the gill count in mature speci-
mens.

Statzner (1981) distinguishes between two species of Amphipsyche larvae from Ivory
Coast by using stridulatory ridges, submentum, and shape of sclerites on IX as a basis for com-
parison, but as far as I know no general scheme for distinguishing all known African species
has as yet been worked out.

The pupa of scottae can be compared with those of senegalensis (Marlier 1978a, Gibbs
1973) and ulmeri (immature pupa from Ivory Coast); descriptions of the non- African species
meridiana (Ulmer 1957), proluta (Lepneva 1964 (1970)) and indica (Hafiz 1937) are also avail-
able. Pupae of the different species are very similar in general appearance, but do show minor
differences in gill counts, in the shape of the anal appendages and their setation, particularly in
males, and in the shape of the pupal cases, which may be rounded apically or urn-shaped, with
lip, as in most species other than scottae.

379

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

In imagos of A. scottae fork 5 has been shown as present in both 8 and 9 forewings,
though whether the 8 fork 5 is a true or false one is a moot point. In drawings of A. senegalen-
sis and berneri males (Kimmins 1962a Figs 4 & 24) it could be a true fork 5 as Cm does not join
Cui directly but ends in a cross-vein connecting Cm and 1A with the anal margin. In other
species, however, including scottae , the posterior branch of “fork 5” might represent Cm, or
Cm + Cm or Cm alone. In females, however, there is a true fork 5 in the forewing as Cm
clearly ends in a cross-vein to the margin as does 1 A, and Cm has two branches.

Biology

Little is known of the biology of Amphipsyche larvae. Hickin (1955) described the larvae
of senegalensis as inhabiting tough, whitish-grey silk-like tubes underneath stones and boulders
a few metres into Lake Victoria, where the water was moving rapidly and they were extremely
common, and on the banks of the Nile a couple of miles downstream from the Owen Falls.
Their identity was confirmed by breeding some of the larvae to male imagos. Hickin suggested
that they might be important predators of Simulium larvae. This was confirmed by Chutter
(1968) in his search for predators of Simulium larvae. He found that A. scottae larvae pos-
sessed a moderately strongly toothed gizzard (though not as strongly so as Macrostemum lar-
vae) and that their gut contents comprised mainly fragments of insect exoskeleton with some
algae. They are thus predatory and omnivorous, possibly eating more insect than plant food,
and are in all probability one of the hydropsychid predators of Simulium.

Amphipsyche larvae are also abundant in certain other African lakes particularly where
the water is in constant motion, as is the case near the entry or exit of rivers (Marlier 1962b).

A. scottae may be numerous, with other hydropsychid larvae (notably Macrostemum, Hy-
dropsyche and Cheumatopsyche species, with a few Aethaloptera), below dam outlets and bar-
rages in rivers where they exploit the plentiful supply of food in the outflow.

In South Africa the imagos fly from August to April, most commonly in the summer
months, and the larvae appear to be present throughout the year, there being a preponderance
of younger instars in spring.

According to Corbet (1966), Amphipsyche senegalensis appears to be able to reproduce
parthenogenetically, perhaps accounting partly for its wide distribution.

African species of Amphipsyche

A. africana Ulmer, 1905c: 48-50, pi. 1, figs 36-39. <3. West Africa.

This species was removed to the genus Protomacronema by Kimmins (1962b: 99-101).

A. berneri Kimmins, 1962a: 91-92, figs 24-28. 8 .

Distribution: Gold Coast (Volta River).

A. corbeti Kimmins, 1962a: 89-91, figs 19-23. 8.

Distribution: Uganda (Victoria Nile).

A. curvinerve Navas, 1927: 214, fig. 10. $ .

Ulmer 1963: 257, discussion of curvinerve and description of larvae which he ascribes to it.
8 unknown.

Kimmins (1962a) suggested that Navas’s curvinerve, described from the 9 only, may
prove to be the female of senegalensis (and see 'Remarks’).

Distribution: Egypt.

A.fuscata Kimmins, 1962a: 128-129, figs 24-29. 8.

Distribution: Ethiopia.

A. instabilis Kimmins, 1962a: 126-128, figs 15-23. 8 9 .

Distribution: Ethiopia, N. Rhodesia (Zambezi River).

380

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

A. plicata (Jacquemart), 1963a: 363-364, fig. 22A-F. 8 ; as Phanostoma.

Distribution: S. Rhodesia (Victoria Falls).

A. scottae Kimmins, 1962a: 93-94, figs 29-37. 8 2 .

Scott 1975: 49-51, figs 35, 36, larva.

Distribution: South Africa: Transvaal, O.F.S., N. Cape, Natal.

A. senegalensis (Brauer), 1875: 71, as Phanostoma. 8 .

Kimmins 1962a: 85-89, figs 4-13, descriptions and figures of 8 , $ .

Hickin 1955: 155-163, figs 10-14, larva.

Gibbs 1973: figs 89-91, pupa.

Marlier 1978a: 299-301, fig. 7, description of 8 pupa.

Distribution: Senegal, Congo, S. & N. Rhodesia, Nyasaland, Uganda, Ethiopia, Sudan,
Mozambique, Ghana, Togo, Tchad, also Madagascar.

A. ulmeri Kimmins, 1962a: 89, figs 14-18. 8.

Distribution: Sudan (Sennar).

Genus PROTOMACRONEMA Ulmer 1904

Protomacronema Ulmer 1904b: 416-417.

Type species Protomacronema pubescens Ulmer 1904b: 417-419, figs 4-8. 8 . Kamerun.

Protomacronema Ulmer; Ulmer 1907d: 34 (and see col. pi. 1 figs 1, 3, for imagos with spread
wings). 8 9 .

Generic diagnosis (based on Ulmer 1904b & 1907d):

Imago: tibial spurs 0.4.4, 8 9.8 mid- and hind femora with a fuzz of long, white hair; tibiae
of 8 mid-legs weakly widened, in 9 mid-tibiae and first tarsal segments very strongly so,
fringed. Antennae with short bulbous scape, flagellum very long, slender; eyes not very large,
dark. Flead reminiscent of that of Amphipsyche but palpi well developed. Vertex with large,
paired anterior warts, posterior warts not as large. Maxillary palpi long, sparsely setulose, ba-
sal segment very short, 2nd longer, 3rd longer than 2nd, 4th shorter than 2nd, 5th very slender,
about 1| times as long as rest together, clearly ringed. Labial palpi with 1st and 2nd segments
short, 3rd longer than both together, ringed. Wings very similar in shape to those of Amphi-
psyche, but show little or no indentation at arculus (except in P. splendens)\ venation also re-
sembles that in Amphipsyche. Forewing with brownish membrane, densely hairy, with hyaline
mark at anastomosis: forks 1, 2, 3, 4 in 8 , 1, 2, 3, 4, 5 in 2 discoidal cell open or absent, SC
and R end in costa, a few indistinct costal veins present; in 8 Cui, Cujand 1A do not join mar-
gin directly but end in a curved cross-vein, hence the absence of fork 5. (Marlier, 1962b, re-
gards this cross-vein as representing the two branches of Cui, in which case, of course, a rather
peculiar fork 5 would be present.) Hind wing with forks 2, 3, 5 or 1, 2, 3, 5; hooked macrotri-
chia along costa, no discoidal cell; Sc and Ri join apically just before joing RS; clear false cos-
tal veins present; wing much folded when at rest. 8 forewing in known species 8-15 mm in
length, 2 smaller. 8 genitalia simple, 10th tergum divided into a pair of lobes, copulatory or-
gan may have spiniform processes of the endotheca (“parameres”); claspers slender, hairy,
with long apical segment. Fifth abdominal sternum in 8 bears a pair of small, knob-like or-
gans. 2 genitalia of usual hydropsychid type but I can find no description of one (see that
given below).

The genus is endemic to the Afrotropical Region, particularly to tropical West Africa,
from which seven species have been described; one of these has also been recorded from
Madagascar, home, too, of an eighth species.

381

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Protomacronema pubescens Ulmer
(Figs 137-156, Table XIII)

Protomacronema pubescens Ulmer 1904b: 417-419, figs 4-8 (<3 palpi, wings & genitalia).
Kamerun. <3 .

Lectotype 6 and paralectotype 6 in the Stockholm Museum, designated by Kimmins
(1962b).

Protomacronema pubescens, Ulmer 1907d: 36 ( partim , Kamerun only, nec figs 28-30).

Protomacronema pubescens, Ulmer 1912: 98, 99 (Kamerun only).

? Protomacronema pubescens, Ulmer 1913: 191 (Zululand).

Protomacronema pubescens Ulmer; Lestage 1919a: 303 ( partim , Cameroon).

? Protomacronema pubescens Ulmer; Barnard 1934: 370, figs 41a-e (Zululand).

Protomacronema pubescens Ulmer; Kimmins 1962b: 99-101, figs 1, 2-4 (wing & 6 genita-
lia), from a re-examination of the material gives the above synonymy and separates
off P. africana (Ulmer). He gives the known distribution as Cameroons, Sierra Leo-
ne.

Protomacronema pubescens Ulmer; Gibbs 1973: 389, 391, figs 73-76, 87, 88 (larva &
pupa). Ghana.

Protomacronema pubescens Ulmer: Scott 1975: figs 43, 44 (larva). Ghana.

Distribution: Cameroons, Sierra Leone, Ghana, Ivory Coast.

Material available comprises 1 6 from Ivory Coast, presented by Mr F. M. Gibon of
O.R.S.T.O.M., also two larvae and part of a pupal skin from Ghana, given by Mr D. G.
Gibbs, one <S and two larvae given by Dr B. Statzner and one 2 on loan from him, and a fur-
ther 6 and 2 received from Mr Gibon.

Description of imagos (Figs 137-145)

Imago (c?, in spirit. Figs 137-143): a fairly large, pale yellow caddisfly, vertex, thoracic
nota and genitalia yellowish brown, warts pale, setae on anterior head warts fine, stiffly up-
standing, on posterior head warts, pronotum and scapulae long, coarse, curly; anterior warts
separated by an almost quadrangular raised “dorsal triangle”. Antennae yellowish with narrow
brown annulations, at least 3x length of forewings. Maxillary and labial palpi noticeably long
and slender, mouthparts very clear (see Fig. 138): madibles of usual hydropsychid type, meet-
ing beneath labrum, laciniae with small denticles, stipites rounded, strongly setose, haustellum
(in specimen drawn) expanded, small labial lobes visible below it (cf. Crichton 1957: 75, 76,
fig. 43). Legs slender, pale yellow, with black spot at bases of femora. Wings as in generic de-
scription, forks 1, 2, 3, 4 in forewings (length 12,0 mm), 1, 2, 3, 5 in hind wings, which are no-
ticeably short in the genus and much folded when at rest. Abdomen mauve-white dorsally, yel-
lowish ventrally, genitalia brown, much retracted into VIII in this specimen.

Ulmer (1904b) described his specimens (dry material) as having red-brown head and tho-
rax with yellowish-red setae, face and mouthparts reddish, apical segments of labial palpi dark
brown. Legs reddish, hind legs yellowish red. Abdomen dorsally reddish with white shimmer,
pleura clothed with white pubescence, ventral side whitish. Forewings pubescent, a golden yel-
low strip over the hyaline mark at the anastomosis, otherwise copper-red to copper-brown, two
dark patches over the pterostigma. Hind wings hyaline, iridescent, with scattered brown setae
in the apical cells. Antennae yellowish red ringed with brown at the joints.

Male genitalia with dorsal margin of 8th segment entire, 9th with slight median excision
and setal fringe, 10th divided to form two lateral digitate lobes, upturned apically, bearing

382

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Figs 137-147. MACRONEMATINAE: Protomacronema pubescens, 6,9,6 pupa. 137. 6 : dorsal view of head & tho-
racic nota, 138. 6: face showing maxillary & labial palpi and mouthparts, comprising: median labrum, lateral to that
mandibles (black), paired lacinia (maxillary lobes), papillate or denticulate, median haustellum, below it minute median
labial lobes (ligula)., 139, 140. 6 : fore & hind wings (right), wing folds indicated by broken lines, 141. <3 : lateral view of
genitalia (abdominal segments VII, VIII shown partly by light broken lines, setae on IX drawn as if VIII omitted), 142.
6 : copulatory organ (lateral) showing paired spines ("parameres”). 143. $: genitalia (ventral), 144. 9: dorsal view of
head, 145. 9: right mid-leg, 145a, same, portion of serrated inner margin of tibial spur, further enlarged, 146. 6 pupa:
apical part of exuviae (ventral), slightly twisted, setal fringe not marginal on right side, thus small dorsal spine visible;
position of left dorsal spine indicated by inset circle, 147: <3 pupa: apex of left appendage further enlarged, not all setae
shown & some shortened; 3 denticles (shagreening) greatly enlarged, 147a, b, anterior & posterior dorsal sclerites from

segment III, enlarged.

383

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

some long, erect setae; copulatory organ with slender stem and clavate apex, within which is a
pair of claw-shaped spiniform processes (“parameres”). Claspers slender, to very slender, se-
tose, with long apical segment.

Female imago (in spirit; Figs 143-145): similar in general appearance to male, though
smaller, antennae shorter, and palpi smaller, very slender and inconspicuous. Forewings have
forks 1, 2, 3, 4, 5. A notable difference is that the anterior head warts are smaller than the
posterior in both available females instead of the reverse, and the head narrower from side to
side, which makes the eyes appear relatively larger. Mid-legs much widened, fringed, tibial
spurs 0.4.4. Abdominal sternum V without paired lateral knobs or slits, paired sternal plates of
VIII separated for entire length, with apical fringe of setae, lateral margins more heavily scle-
rotized than rest; sterna of apical segments represented by median plate, XI with usual paired
papilli and cerci, dorsal hood fringed with long curved setae.

Description of larva (Figs 148-157, Table XIII)

Protomacronema larvae are strongly reminiscent of Amphipsyche larvae in appearance,
though smaller, with less conspicuous gills and with characteristic long, curly setae on their
heads, in later instars of the only known species at least. There are also certain resemblances
to Macrostemum larvae, such as shape of submentum, stridulatory files and mandibles.

Mature larva (in spirit. Figs 148-157, Table XIII): length about 7,0 mm. One of the speci-
mens is commencing to pupate; the other is about the same size and evidently a mature larva,
both come within the same head width range (head width at eyes 4,3 and 4,9 grid squares re-
spectively), and are therefore presumably in the same instar, in spite of slight size and colour
differences.

Head (Figs 148-150): rounded, with flat top and carina, frontoclypeal apotome longer
than wide, a pair of small knobs on the anterior margin, apex acute, dividing carina, which
runs alongside it almost to the posterior margin; colour described by Gibbs (1973) as follows:
area within carina dark brown with orange-brown areas in centre and adjacent to eyes, head
otherwise dark brown with orange-brown stridulatory files. Head within carina with long, curl-
ing setae arising from clear pits, anchor-shaped central area smooth. Eyes dark, placed well
forward. Stridulatory files long, narrow, reaching to occipital foramen, a pair of small depres-
sions lateral to them. Anterior ventral apotome brown, represented by a small sclerite with
cleavage line on left side only; posterior ventral apotome minute, black. Head in available
specimens retracted within pronotum.

Mouthparts (Figs 149-151): labrum brown, largely sclerotized, with membranous lateral
lobes bearing tufts of silky setae, anterior margin with smaller tufts on ventral side, antecly-
peus with four brown, sclerotized plates, of which the median ones are large, the lateral ones
small and partly concealed by the anterior corners of the frontoclypeal apotome. Mandibles
large, strong, subtriangular in section, with a number of setae on outer side, which is partially
demarcated by a short, oblong lateral flange in both cases; left mandible with seven teeth, one
flattened; right mandible with five teeth, no inner brushes. Submentum oval with setose mar-
gin, flanked by stipites, all brownish.

Thorax (Fig. 148): thoracic nota dark brown, with few, small setae; pronotum has paired
sclerites with indented anterolateral corners and posterolateral black marginal thickenings.
Meso- and metanota undivided, with slight black lateral thickening and large posteromedian
black spots, lateral diagonal stripes only faintly indicated. Prosternal plate narrow, brown,
without median sulcus. Meso- and metasterna membranous, mesosternum with one pair of
coxal gills; these are very long and distinctive because the basal part is contracted and sclero-
tized, widening to form a soft stalk tipped with a single pair of leaf-like apical branches. Meta-
sternum with two pairs of tufted gills. Propleuron also unusual, with pre-episternum set into

384

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Figs 148-157. MACRONEMATINAE: Protomacronema pubescens , mature larva. 148. habitus, lateral (right append-
ages only shown) mesosternal gill shown further enlarged (same scale as legs), 149, 150. dorsal & ventral views of head,
151. a. left mandible, dorsal, b and c. right mandible; a. dorsal, b. ventral., 152, 153, 154. foreleg, mid- & hind legs
(right), foreleg with pleural sclerites, 155. a. palmate setae from coxa, b, c, d. claws of foreleg, mid- & hind legs, all fur-
ther enlarged, 156. ventral view of abdominal segment IX, with anal prolegs, 157. right anal proleg, lateral.

385

ANN. CAPE PROV. MUS. (NAT. HIST. ) VOL. 14 PT 8, JUNE 1983

side of large 2nd pleural sclerite, epimeron narrow, bearing a row of small spine-like setae (see
Fig. 152).

Legs (Figs 152-155): short, strong, subequal, yellowish brown to brown, forelegs as usual
darker than rest, claws brown, legs with many secondary setae, particularly on and near poste-
rior margin. Forelegs: coxa with high basal process which articulates with propleuron just be-
low pre-episternum, a row of short spine-like setae along anterior margin and many small pal-
mate setae as indicated; femur with small thumb-like basal projection on inner side, as in
Macrostemum-, tibia and tarsus with row of ventral spine-like setae and very weak brush of
long setae; claws blunt with thick blunt basal spine. Mid- and hind legs also with palmate or
feathered setae on coxae, both with stout spines at apices of tibiae and tarsi, claws larger but
still blunt, with blunt basal spines.

Abdomen (Figs 148, 156, 157): smooth, not deeply segmented, few setae except along the
transverse indented bands on segments I and II, which are set with minute hooked spinules as
in Macrostemum and Amphipsyche. Tufted tracheal gills (rather small and inconspicuous, with
fine branches), present in six rows on segments I- VI, four on VII and two on VIII. Gill fila-
ments tend to form tufts towards apices of gill stalks. Segment IX bears a pair of fairly large
but inconspicuous setose ventral plates, also smaller ventrolateral plates; paired dorsolateral
plates present but almost concealed under VIII, each bears a few short setae. There are no
such sclerites on VIII. Anal prolegs short, stout, with somewhat indistinct lateral sclerites,
large anal claws and lateral fringe of long setae, no fan. Anal gills not seen.

Table XIII.

Number and position of tracheal gills in mature larva of Protomacronema pubescens

Segment

Mature larva

Mesothorax

Metathorax

1 pair

2 pairs

Abdomen

(right side)

DL

VL

V

1

1

1

1

II

2*

2

1

III

2

2

1

IV

2

2

1

V

2

2

1

VI

1

2

1

VII

1

2

—

VIII

—

1

—

IX

—

—

—

‘Abdominal gills marked 2 have two main stems whose bases are contiguous but not joined.

DL = dorsolateral gill, these are not separated into pre- and postsegmental gills, but are, however, all presegmental in
position whereas the ventrolateral gills are postsegmental. Gill branches range in number from 10-14 per gill stem
on segment II to 4-6 per stem on V-VIII, and are always small and slender.

VL = ventrolateral gill.

V = ventral gill.

386

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS
Description of pupa (Figs 144-147)

The only pupal material available is the middle and posterior (apical) part of the pupal
skin of a d pupa (see Figs 144-147). It has been mounted on a slide and from it the following
facts could be ascertained:

Anal appendages short and stout compared with other genera so far described, each tip-
ped with a large, sharp, sclerotized point, and fringed laterally (largely ventrally) with long,
stout setae, some with curled tips, arising from prominent bases; between them, towards the
dorsal side, there is a pair of smaller sclerotized points. The appendages are heavily sha-
greened (i.e. covered with scattered triangular points as in shark skin). Usual pockets for 8
genitalia, smaller than in Amphipsyche or Macrostemum. Dorsal plates large, with long stout
teeth; anterior plates apparently on segments III— VI 1 1 , posterior on III only. There were no
gills on the material available, and as I have been unable to obtain any further pupal material
it is not possible to provide a table showing gills or dorsal plates (of which only some are pres-
ent).

Remarks

Ulmer (1904b) made Protomacronema pubescens Ulmer the type species of his new genus,
Protomacronema , from Kamerun. Shortly afterwards he described Amphipsyche africana from
West Africa (Ulmer 1905c), but later decided that he had been mistaken, placing it in synony-
my with his Protomacronema pubescens (Ulmer 1907d). This led to subsequent confusion in
the identification of specimens and consequently in the literature. The position was later clar-
ified by Kimmins (1962b) in a revision of these two species. He concluded that both were cer-
tainly Protomacronema but that africana was in fact a valid species, clearly different from pu-
bescens, and gave new figures and descriptions of both species with their known distributions
and synonymy. The parts relevant to P. pubescens are given above. Ulmer’s record from Zulu-
land (Ulmer 1913) is given with a query as Kimmins felt that from the description the speci-
mens appeared to be intermediate between pubescens and africana and should be re-examined.
Material from Rhodesia (Zambezi River) examined by Kimmins proved to be of africana. Jac-
quemart (1963a) figured a male from the Victoria Falls (Zambezi) as pubescens , however, his
drawing clearly shows the U-shaped excision of africana , to which species it must be referred.

As will be seen from the list which follows, eight species have been described. Of these
probably only four or five represent valid species, since three were described, with very poor
figures, by Navas from females only.

Statzner (1981) had two kinds of Protomacronema larva from Ivory Coast, which he separ-
ated by differences in the submentum and the dorsal sclerites on abdominal segment IX.

Biology

Little is known of the biology of Protomacronema larvae. Gibbs’ larvae and pupae were
taken from the Black Volta River at New Bator, in an area where fast-flowing rapids linked by
slow-flowing deeper parts occurred (Gibbs 1973; Petr 1970). The Protomacronema larvae were
found where the current velocity exceeded 20 cm/sec, increasing in numbers with increased
current speed, and making a contribution to the standing crop of, for example, 2 534,9 mg/sq.
m out of a total of 42 957,9 mg/sq. m where the current speed was 140 cm/sec. The total Tri-
choptera contribution at that point was 30 835,2 mg/sq. m, of which the great majority were
Cheumatopsyche species.

The substratum in the rapids consisted mainly of stones, often with deep crevices between
them, the depth of water during the dry season being under 30 cm. During floods the current
speed increased to 300 cm/sec at times.

387

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983
African species of Protomacronema (Endemic Genus)

P. africana (Ulmer) 1905c: 48-50, pi. 1, figs 36-39, as Amphipsyche. 8 . West Africa.

Kimmins 1962b: 99-101, figs 5-9 {8 genitalia), in a revision of the two species concerned,
transferred africana from Amphipsyche to the genus Protomacronema as a species sui
generis , removing it from the synonymy of P. pubescens (Ulmer) in which Ulmer had
himself placed it (Ulmer 1907d), and giving the following synonymy:

Protomacronema pubescens Ulmer ( partim ), 1907d: 36, figs 28-30, Congo examples.
Protomacronema pubescens Ulmer, 1912: 98-99 (Congo examples).

Protomacronema pubescens Ulmer, Lestage, 1919: 303, {partim , West Africa).

? Protomacronema pubescens Ulmer, Navas 1930a: 329, Congo.

? Protomacronema divisum Navas 1930a: 329, Congo. $ .

? Protomacronema pubescens Ulmer, Navas 1931c: 277 (Congo); 1931b: 141 (Congo).

? Protomacronema pubescens Ulmer, Mosely 1932: 5 (Zambezi).

Protomacronema pubescens Ulmer, Jacquemart, 1961b: 19, fig. 12 (Congo).

Distribution: West Africa, Congo Rhodesia (Victoria Falls), ? Portuguese East Africa
(Zambezi).

P. divisum Navas 1930a: 329-330, fig. 44. $ . Belgian Congo.

Kimmins 1962b: 101 places this in probable synonymy with P. africana.

Distribution: Belgian Congo.

P. hyalinum Ulmer 1904a: 354-355, 359, fig. 1. 8. French Congo.

Ulmer 1905a: 27, figs 13, 14. $ . Congo.

Distribution: French Congo, Belgian Congo, Madagascar.

P. pellucidum Navas 1923: 26, fig. 37. ? . Madagascar.

Distribution: Madagascar.

P. pubescens Ulmer 1904b: 417-419, figs 4-8 ( 8 palpi, wings & genitalia).

Ulmer 1907d: 36 {partim , Kamerun only, nee figs 28-30).

Ulmer 1912: 98, 99 (Kamerun only).

? Ulmer 1913: 191 (Zululand).

Lestage 1919: 303 {partim , Cameroon).

? Barnard 1934: 370, figs 41a-e (Zululand).

Distribution: Cameroons, Sierra Leone, Ghana, Ivory Coast.

P. splendens Ulmer 1905c: 65-67, pi. 3 fig. 106. 8 . Gabun.

Ulmer 1907d: 34, 37, 38, fig. 31, col. pi. 1, fig. 3 (whole insect).

Distribution: Gabun.

P. tanganikae Jacquemart 1961c: 69-71, figs 7-10. 8 . Lake Tanganyika, Albertville.
Distribution: Belgian Congo (Lake Tanganyika).

P. testaceum Navas 1934b: 91-92, fig. 57. $ . French Guinea.

Distribution: French Guinea.

388

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Genus LEPTONEMA Guerin-Meneville 1843

Leptonema Guerin 1843: 165, 396.

Type species pallida (recte pallidum) Guerin 1843: 396. Brazil.

Location of type unknown; Type species currently unrecognizable, see Mosely 1933: 7-8.

Leptonema Guerin; Mosely 1933: 7-12.

Revision of the genus, with key to known species.

Leptonema Guerin; Ulmer 1957: 346-348, 352, 354, larval and pupal stages.

Leptonema Guerin-Meneville; Fischer 1947: 313-315.

Leptonema Wiggins 1977: 108, 109, figs 6. 7A-E.

Generic diagnosis (derived mainly from Mosely 1933: 9-10)

Imago: tibial spurs of 8, 2, 2.4.4 or 1.4.4, inner spurs longer than outer; mid-leg of 2 not
usually dilated; antennae very long, twice to three times length of forewing, scape short and
thick, flagellum slender; maxillary palpi with basal segment short, second long, third shorter,
fourth shorter still, fifth as long as or longer than the rest put together, ringed. Forewing
usually with short, broad discoidal cell, forks 1, 2, 3, 4, 5 present, fork 1 with footstalk; hind
wings broader and shorter than forewings, forks 1, 2, 3, 5 present. Wings semi-transparent or
streaked with black pubescence or irrorated in American species, generally coloured or pat-
terned in African and Madagascan species. Forewing length in 8 8-28 mm, in 2 10-19 mm
(not many females known); where known 2 usually larger than 8. Male genitalia with tenth
tergum divided, bearing simple or complex cerci; copulatory organ with or without spiniform
processes. 2 with sternal plates of VIII widely separated. 8 , 2 both with paired fenestrae on
fifth sternum.

To date over 40 species have been described from the Neotropical Region, extending as
far north as Mexico and the Texan border, fourteen from Central and Southern Africa and
Madagascar, and one each from the Mediterranean and India.

Leptonema natalense Mosely
(Figs 158-176, Tables XIV, XV)

Leptonema natalense Mosely, 1933: 24-25, figs 43-48 (forewing, 8 genitalia). Natal,
Kloof, 1 500'. 8.

Holotype 8 in British Museum (Natural History).

Leptonema natalense Mosely; Barnard 1934: 370, fig. 4 If— h (wings, 8 genitalia), as L. oc-
cidental Ulmer; a mistaken identification which he later corrected (Barnard 1940).

Leptonema natalense Mosely; Marlier 1961: 159, 202-207, figs 23, 24. 8, 2, larva, pupa
(2 not described). Kivu.

Distribution: South Africa: Natal, Eastern Transvaal, Pondoland; Lesotho, Congo (Kivu).
Description of imagos (Figs 158-163)

Male imago (in spirit): a fairly large species, medium to dark brown in colour, forewings
11-13 mm long, pubescent, fuscous, hind wings hyaline, often iridescent, wide at base (folded
at rest); legs yellowish, not noticeably widened, spurs 1.4.4 (2.4.4 in some Natal specimens,
second spur on foretibia very small); eyes medium-sized, dark; antennae very long (c 35 mm),
dark, scape enlarged, flagellum very slender; palpi well developed, finely pubescent; face
rounded, with few setae; a single small, flat, median wart near antennal bases; vertex with two
small warts at bases of antennae, a large protuberant median wart between them, all brown,
posterior warts larger, cream-coloured, setae long. Thus in this genus the anterior warts on the

389

158

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

160

Figs 158-167. MACRONEMATINAE: Leptonema natalense, 6, 2, pupae. 158. 6 : dorsal view of head & thoracic
nota, 159. 6: head, lateral, with maxillary & labial palpi, 160, 161. 6 : fore & hind wings (right), 162, 6: genitalia, right
lateral: a. apex of left clasper, ventral, b. copulatory organ, right lateral., 163. 2: genitalia, ventral view, 164. 6 pupa:
apical part of exuviae, ventral, 165. 6 pupa: apex of left appendage, lateral, 166. 167. 2 pupa: right & left mandibles,

ventral.

390

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

vertex are, most unusually for Macronematinae, considerably smaller than the posterior ones.
Pronotum with median division and two pairs of oval cream warts; mesonotum without warts,
but mesoscutellum with large pair of poorly defined cream warts. Abdomen brown, with
paired fenestrae on 5th sternum, genitalia brown, 9th tergum with margin rounded, entire,
tenth divided to form a pair of triangular plates with basal setose processes, copulatory organ
with small truncate apex flanked by wing like expansions; claspers with basal segment apically
widened, apical segment short, broad, truncate in dorsal view, in lateral view bluntly pointed
with short blunt spines and fine setae.

Mosely (1933) described the pinned male as having a black head, dark fuscous antennae
with yellowish annulations, ochraceous palpi, black thorax, brown forewings with light brown
pubescence and fuscous hind wings.

Female imago (in spirit): much same size as male (forewing 12,4 mm), antennae shorter,
mid- and hind legs somewhat flattened, scarcely widened, not fringed; spurs 1.4.4 (only two fe-
males available). Pale fenestrae on V obvious; paired sternal plates of VIII widely separated,
apical margins thickened, pale, setose, lateral margins darkened, sterna of apical segments rep-
resented by scelerotized median plates, usual apical papillae and cerci, dorsal hood fringed.

Material in the Albany Museum collection includes 20 males, 2 females, 32 larvae and 9
pupae, about half the specimens being in good condition. Amongst the pupae are mature
males disclosing genitalia, also one good female bred out in the laboratory by Dr Chutter. In
each case these are accompanied by the pupal case, larval sclerites and pupal exuviae, thus de-
finitely linking both sexes of the species with their larval and pupal stages.

Description of larva (Figs 168-176, Table XIV)

Mature larva (in spirit; Figs 168-176): length 22-23 mm, greatest width (abdomen) c. 3
mm. Large larvae, sclerotized parts amber to pale orange or brown in colour, abdomen cream
to pale mauve, gills white. Mean head width of mature larvae examined, at eyes, 1,34 mm
(mean of 18 mature larvae, range 1,21-1,53 mm). As these larvae are somewhat smaller than
those described by Marlier (1961), I checked the size of the frontoclypeal apotome against that
of the frontoclypeal sclerites found in the pupal cases, and found that they fell into the same
range. 1 am therefore assuming that these are mature larvae.

Head (Figs 169, 170): almost oblong, widest near occipital foramen, rounded, without Car-
ina, nearer in shape to those of Hydropsychinae than Macronematine larvae; frontoclypeal
apotome with slight median indentation in rounded anterior margin, head bright amber yellow
to chestnut with a few obscure muscle marks, thickly set with short, fine setae and a few longer
ones, ventral side paler. Eyes black, set in lighter areas, each beneath six small cuticular
lenses. Antennae minute, at base of mandibles. Stridulatory files large, long, inconspicuous un-
less darkened. Anterior ventral apotome a shallow triangle, with sutures on both sides, poste-
rior ventral apotome dark, minute.

Mouthparts (Figs 169-171): labrum large, retractile, setose, brownish, almost semicircu-
lar, with silky lateral brushes; anteclypeus pale, with shiny areas which may represent sclerites,
but if so are scarcely distinguishable. Mandibles strong, red-brown, left one with two large api-
cal teeth, in mature larvae usually worn down to appear a single broad, llat-topped curved
tooth, dorsal to which is a smaller tooth bearing a row of white setulae, five smaller, less
prominent teeth along cutting edge, mandible hollowed with inner brush of setae. Right man-
dible with two apical teeth, a smaller dorsal tooth with row of setulae, three irregularly shaped
teeth on cutting edge, mandibles triangular in section, outer faces setose. Maxillary palpi short,
thick, 5-segmented, longer than maxillary lobes; labium short, thick, palpi minute. Submentum
a truncated triangle, brown, as are stipites.

Thorax (Fig. 168): terga bright amber yellow to chestnut, unpatterned, with many small
setae and a few longer ones, most margins narrowly black or brown; pronotal sclerites paired

391

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Figs 168-176. MACRONEMATINAE: Leptonema nulalense mature larva. 168. habitus, lateral (right appendages only
shown), paired gills from abdominal segment II (left) further enlarged, 169, 170. dorsal & ventral views of head, 171.
right & left mandibles, dorsal, 172, 173, 174. foreleg, mid- & hind legs (left); foreleg shown with pleural sclerites, mid-
and hind legs with attached coxal gill; in each case tarsal claw & adjacent setae shown further enlarged, 175. ventral
view of abdominal segments IX & part of VIII, showing ventral sclerites, 176. left anal proleg, lateral.

392

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

with narrow posterior division, mesonotum with a conspicuous black M-shaped mark, centrally
placed near posterior margin, metanotum with a much smaller crescent. Prosternum with
large, deep, flanged sclerite, attached laterally to the pronotum by muscles & epidermis; a pair
of smaller sclerites lie behind it, along the posterior border of the mesonotum, sometimes par-
tially obscured by an epidermal fold. Pre-episternum unbranched, conical, set with stout setae
as is the epimeron; episternum glabrous. One pair of branched mesosternal gills present, two
pairs of branched metasternal gills (see Figs 173, 174).

Legs (Figs 172-174): slender, brownish yellow to yellow, forelegs as usual darkest, thickly
set with setae, mainly small, inconspicuous, except marginally. Foreleg with strong thumb-like
process on upper distal margin of coxa and line of strong setae, trochanter divided, femur with
usual protuberance (scraper of stridulator) on anterior margin, tibia and tarsus with ventral
band of multifid or plumose setae, the latter also mingled with strong spine-like setae on the
other joints; claw blunt with slender basal spine. Mid- and hind legs subequal, tibiae and tarsi
with marginal belt of strong spine-like setae, in hind leg mingled with long pinnate setae, which
also occur on femur, tibia and tarsus of both legs; claws small, stout, blunt, with single short
thick spine near base.

Abdomen (Figs 168, 175, 176): thickly covered with small, narrow scale-like setae, pale
and hard to see unless larva comes from turbid water. Abdominal tracheal gills: two pairs on
segment I, one pair on II— VII (see Table XIV), gills on segments II— VII each with two stems
and 12-20 filaments per stem on the more anterior gills, fewer on the posterior ones. On eighth
sternum a small pair of triangular setose sclerites, on ninth a large pair; no dorsal or lateral
sclerites on VIII or IX, but four small clumps of lateral setae on IX. Anal appendages fairly
long, setose, with strong yellow lateral sclerites, claws stout, no anal brush. Anal gills present,
only three seen but according to Ulmer (1957) there are five.

Earlier instars:

Only one identifiable younger larva was found in the material available, and this was ap-
parently very recently ecdysed (many gills still in cuticle of earlier instar). Head width at eyes
0,9 mm; gill count as in mature larva, all double gills already present, but filaments only 1-4 in
number per gill stem. Posterior prosternal sclerites present but very pale and inconspicuous,
otherwise as mature larva in general appearance, only paler and smaller. Assuming that Lepto-
nema has six larval instars as in the other Macronematinae studied, this one would appear to
be in the fourth instar.

There were in addition five probable early instar larvae, of which brief descriptions follow,
but it was not possible to be certain of their identity, though they were clearly separable from
early Hydropsychine larvae present in the same samples, and do appear to be of Leptonema.

Probable second instar larvae: large head relative to body, width at eyes 0,23-0,26 mm
(three larvae), head setae few, long, no stridulatory files as yet; mandibles strong, labrum
large, setose. Gills absent. All three thoracic nota well sclerotized with fine brown lines demar-
cating most edges, a few long setae; pre-episternum present, simple, with one seta. Legs: fore-
leg with two long feathered setae on tarsus, claw long, stout, with two long basal setae, one
shorter one and a small spine. Long plumose setae also present on coxa and tibia; coxa with
anterior row of about three long setae but no thumb-like process as yet. Abdomen smooth
with a row of very long setae, IX with two long ventral plumose setae but no ventral sclerites
as in older larvae. Anal prolegs long, claw strong, curved, a few long setae.

Probable third instar: head width at eyes 0,35 and 0,36 mm (two larvae); setae over whole
larva shorter and much more abundant, stridulatory files present on ventral side of head. Lat-
eral margins of thoracic nota darker, one pair of simple gills on mesothorax, two pairs on me-
tathorax, pre-episterna with four setae. Forelegs: coxae still lack thumb-like process, femora
with scrapers; all limb claws with a single long basal spine-like seta and a small basal spine.
Abdomen hairy, now with many small scale-like setae; patches of long setae on eighth and

393

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

ninth sterna. One pair of simple gills, each with a single filament, on segments II— VII, no
double stems as yet. No gills bear a seta, as in third instars of Macrostemum and Amphipsyche,
but this is to be expected as Leptonema does not have ventral gills.

Table XIV.

Number and position of tracheal gills in mature larva of Leptonema natalense

Mesothorax

Metathorax

1 pair

2 pairs

Abdomen
(right side)

L VL

V

I

1

1

II

— 2*

—

III

— 2

—

IV

— 2

—

V

— 2

—

VI

— 2

—

VII

— 2

—

VIII

— —

—

*Gill with 2 main stems arising from a common base.

L = lateral gill.

VL = ventrolateral gill.

V = ventral gill.

Description of pupa (Figs 164-167, Table XV)

Nine pupae were available, including one good, mature male pupa with case and larval
sclerites from Lesotho, and one good male from Natal (Inkomba stream), also six male and fe-
male pupae in very poor condition from the Ingeli Forest and from the Inkomba stream. Natal.
In each case they were associated with larvae and identifiable as Leptonema natalense. There
was also a female imago from the Inkomba stream which had been correlated in the laborato-
ry, together with its associated case, pupal pelt and larval sclerites.

Male pupa (in spirit; Figs 164, 165): pupa still within pupal pelt, length 12 mm; head simi-
lar to that of imago, eyes, head and thoracic warts and palpi clearly visible. Labrum small,
rounded, a few setae along upper margin and near antennal bases; above labrum paired
clumps of long, hooked brown setae. Mandibles as in $ (Figs 166, 167), with inflated bases,
outer faces set with setae; left mandible with apical and four lateral teeth, right with apical and
three lateral teeth, inner margins between apical and lateral teeth serrated. Palpi long, well de-
veloped. Antennae very long, slender, each terminating in 2\ coils round apex of abdomen.
Forewing cases brownish, hind wing cases cream, reaching anterior end of segments VI and V
respectively. Mesoscutum shows wide, paired median bands of pubescence through pelt. Legs
cream with small brown spots at apices of trochanters and tibiae, spurs 1.4.4. Abdomen: seg-
ment III with a pair of tapering unbranched lateral gills, IV- VII with two such gills each side;
II- VII also with paired ventral tufted gills, each with two main stems and numerous filaments.
Paired dorsal plates as in Table XV; presegmental plates on segments II— VIII, postsegmental

394

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

plates (wide strips with many teeth) on III only. Shagreening practically absent; a belt of long
setae on IV and a number of hairs on V. Anal appendages of medium length, plump, strongly
curved dorsad, contracting abruptly to form short finger-like apices densely fringed with small
spine-like setae. Proximal to the apex the appendage is fringed with eight to ten long stout se-
tae with curved tips. Male genitalia enclosed in pockets as usual.

Female pupa (in spirit; Figs 166, 167): general appearance as male, but larger, antennae
shorter, abdomen stouter; female genitalia not in pockets; anal appendages much as in male,
but somewhat stouter.

Pupal case: about 15 mm in length, almost completely covered with large stones or rock
fragments, interstices filled with small sand grains. Pupa within a complete sac-like silken lin-
ing, closed at both ends, which are formed from a fine meshwork of strands, allowing for pass-
age of water. Ventral side of case adpressed to rock, of silk only; larval sclerites packed into
posterior end of case as usual.

Table XV.

Number and position of branchial gills, dorsal plates etc. of 6 pupa of Leptonema natalense

Mosely.

Abdominal

segment

Tracheal gills
(right side)

Dorsal plates
(paired)

Setae

I

L.

V.

—

—

—

II

—

2

Presegmental

III

(1)

2

Presegmental

Postsegmental

IV

(2)

2

Presegmental

Belt of hairs

V

(2)

2

Presegmental

Scattered hairs

VI

(2)

2

Presegmental

VII

VIII

(2)

2

Presegmental

Presegmental

L. = lateral gills; these are conical pouch-like gills, either single (1), or double, lying one above the other (2).
V. = ventral gills; these are branched gills, each with 2 main stems arising from a common base.

Remarks

The larvae and pupae described above are definitely those of L. natalense Mosely, correla-
tion with the imagos having been reliably established.

Marlier (1961) described larvae of L. natalense from Kivu. These show small differences in
patterning from the larvae described above, due perhaps to a certain amount of geographical
variation. There are also slight differences in mandibular teeth. Marlier described the gills as
“two pairs per segment”, but presumably meant one pair of double gills per segment, except
on segment I, as is the case in all Afrotropical specimens that I have seen.

Gibbs (1973) described the larvae of three species of Leptonema from Ghana: normale
Banks, guineense Gibbs, and a third uncorrelated species near to natalense but showing some

395

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

differences in head pattern and structural details. He stated that the larvae of all three species
had one pair of two-branched gills with many filaments on segments II— VI, however, his dia-
gram shows in addition two pairs of single gills on segment I and a pair of two-branched gills
on VII as in natalense.

It is somewhat more difficult to reconcile the gill position in natalense (and other known
African species) with that in the American Leptonema sp. larva from Mexico, beautifully de-
lineated in Wiggins’ book (1977, Fig. 6.7A). In the latter species the pairs of double gills are
present though slightly separated basally but there are in addition single gills in a presegmental
position on segments II-VI (presumably also paired), so that the gill count would be two pairs
on segments I and VII and three pairs on II-VI. However, Dr O. S. Flint, Jr. (in litt.
4. viii . 1982) has very kindly informed me that there are two groups of Leptonema species rep-
resented in the Americas, one of which being that which includes the African species and the
other being entirely American. All Flint’s associated larvae belong to species in the second
group and all have the additional ventral gills, as figured by Wiggins (1977). When next in the
Neotropics Dr Flint will endeavour to collect larvae of the first group. Should these resemble
known African larvae in gill arrangement, it may prove necessary to reconsider the generic
status of the two groups.

There is a surprisingly wide variation in the coxal armature in Leptonema species, in Afri-
can species at any rate. In L. normale the fore-coxa (“anterior femur” of Gibbs) has its ante-
rior margin produced into a flat flange-like plate fringed with heavy sword-like setae, in gui-
neense the coxa is bordered with a row of long stiff spine-like setae, and in natalense and
Gibbs’ unknown species it bears an anterior thumb-like process and a few stiff spine-like setae.
Wiggins’ Mexican species also possesses such an apical coxal process (Wiggins 1977, Fig.
6.7B). In species known to me the easiest characters to use in separation are the coxal arma-
ture and the shape of the stridulatory files. Statzner (1981) in addition uses the abdominal se-
tae in separating his three species.

The Angolan larvae that were kindly lent to me as Protomacronema sp. by Drs de Barros
Machado and Carvalho of the Museo do Dundo (see Marlier 1965/66) proved to be either Lep-
tonema normale Banks, as described by Gibbs (1973), or a similar species. They differed from
natalense not only in coxal structure but also in having narrower stridulatory files and two thin
stiff anal gills as well as three normal ones (thus making up the usual five).

The most interesting features of Leptonema larvae, however, lie in their many similarities
to Hydropsychine and Diplectronine larvae, particularly those of Cheumatopsyche and Diplec-
tronella , for which they may easily be mistaken unless closely examined. Resemblances include
the rounded head, mandibular structure, hairy appearance and possession of setose ventral
plates on both VIII and IX. Nonetheless, Leptonema larvae have the usual main Macronema-
tine characters such as simple pre-episterna, entire labium, frontoclypeal apotome not widely
expanded behind eyes, minute posterior ventral apotome and meso- and metanota without
transverse sutures (Scott 1975). It is possible that these wide differences and similarities in lar-
val structure may be due to differences and similarities in larval biology, currently not well
known.

Biology

Leptonema larvae inhabit flowing water, usually living beneath stones or on rocks in cur-
rent or in cascades. There they construct typical hydropsychid retreats covered with sand
grains, rock fragments or plant material, rather more rigidly constructed than is the case in Hy-
dropsyche and provided with a simple net for catching food, which appears to include insects
and plant material. In Kivu Leptonema larvae were common in high altitude streams
(2 000-3 000 m) with acid or neutral water, often slightly peaty. In Ghana they were found

396

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

both in highlying forest streams and also in streams in cultivated lowlands. (See Marlier 1961,
Gibbs 1973, Wiggins 1977.)

South African specimens in the Albany Museum collection (all L. natalense) came from
streams. The larvae were collected from stones in current and the imagos were either bred out
in the laboratory or taken in the field on riverine vegetation or at light. They appeared to be
present throughout the year.

African Species of Leptonema
L. affine Ulmer, 1905a: 27-28. 3 9 .

Ulmer, 1907d: 57, 58, fig. 60 (d wings) & pi. II fig. 8 (col. fig. of imago).

Mosely 1933: 27, figs 54-56 (6 genitalia).

Distribution: Madagascar.

L. alatum Marlier, 1961: 159, 199-200, fig. 22 (wings). 3 2 .

Distribution: Kivu.

L. displicens (Navas), 1935: 73-74, fig. 34. 3 , as Macronema.

Sykora, 1964: 281, figs 13-16 (d genitalia), as Leptonema.

Distribution: Madagascar.

L. guineense Gibbs, 1973: 385-386, figs 56-59 (3 genitalia), 66-68 (larva), 80-82 (pupa). 3 2
larva pupa.

Distribution: Ghana.

L. latipenne Marlier, 1947: 31-32, figs 1, 2 (genitalia). 3 9 .

Distribution: Ivory Coast.

L. machadoi Marlier, 1965/66: 22, 44-45, fig. 15 (wings). 3 .

Distribution: Angola.

L. madagascariense Ulmer, 1905b: 81, fig. 50 (wings). 3 .

Ulmer 1907d: 58, figs 61, 62 (wings, d genitalia).

Mosely 1933: 28, figs 57-59 (d genitalia).

Distribution: Madagascar.

L. milae Sykora, 1964: 279-291, figs 5-9 (d genitalia), d.

Distribution: Madagascar.

L. natalense Mosely, 1933: 24, 43-48 (d genitalia & forewing), d.

Barnard, 1934: 340, figs 41f-i (wings, d genitalia), as occidentale.

Marlier, 1961: 159, 202-207, figs 23, 24 (larval & pupal parts).

Distribution: South Africa: Natal (type locality), Pondoland, East Transvaal; Belgian
Congo (Lake Kivu); Basutoland.

L. normale Banks, 1920: 357, t. 7, fig. 108, as L. normalis. Sex not stated in text, but figure
given is of d .

Mosely, 1933: 64, regarded normale as an unrecognizable species.

Gibbs, 1973: 383-385, figs 53-55 (d genitalia), 63-65 (larvae), 83 (pupal appendages), d
2 larva pupa. Ghana.

Distribution: Cameroons, Ghana. Larvae are similar to those described by Marlier
(1965/66) from Angola as Protomacronema , so distribution could extend south to An-
gola.

397

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

L. occidentale Ulmer, 1907c: 57, 58, fig. 87 (wings). 3 $ (9 undescribed). Cameroons.

Ulmer, 1913: 191, Barberton, Transvaal. I have examined this material, which proved to
be L. natalense Mosely.

Mosely, 1933: 23, figs 38-42 (d genitalia). Gives probable synonymy of L. vanderysti
Navas with this species.

Marlier, 1961: 159, 201, 202. Description of 3 pupa. Urundi.

Distribution: Cameroons, Urundi, Zaire.

L. tholloni Navas, (1922) 1923: 48, 49. Gabun. 3.

Mosely, 1933: 30, figs 65-68 (<3 genitalia). 3 .

Distribution: French Gabun.

L. vanderysti Navas, 1930a: 330-331, fig. 45 (wing fragments). 2 .

Mosely, 1933: 66 states that this is possibly the 2 of occidentale .

Distribution: Belgian Congo.

L. zahradniki Sykora, 1964: 277-279, figs 1-4 (3 genitalia). 6 .

Distribution: Madagascar.

Note:

Leptonema tridens Mosely, 1933: 10, 17-18, figs 17-21. 3 . Brazil.

Ross, 1967: 195, stated that tridens also occurred in Africa.

Flint (in litt. 14. ix. 1979), however, corrected this as being a mistaken reference, L. agra-
phum Kolenati being the Brazilian species in Ross’s mind, and being a species closely
related to L. tholloni from Gabun. L. tridens is thus confined to South America, and
should not be placed on the African list.

Genus POLYMORPHANISUS Walker 1852

Polymorphanisas Walker 1852: 78 (under Leptoceridae).

Type species Polymorphanisus nigricornis Walker 1852: 79. North India.

Oestropsis Brauer 1868: 263, 407.

Type species Oestropsis semperi Brauer 1868: 264-265, pi. 2A f, 1— Id, 6 2. Philippines. Syno-
nymy established by Ulmer 1905a: 24, 25.

Polymorphanisus Walker; Ulmer 1907b: 14, 19, 20, figs 4—9.

Polymorphanisus Walker; K. H. Barnard 1934: 369.

Polymorphanisus Walker; P. C. Barnard 1980: 78, 79, generic diagnosis, comment and key to
known species.

Generic diagnosis (based on P. C. Barnard 1980, K. H. Barnard 1934 & Ulmer 1907b, checked
against specimens in Albany Museum collection):

Imago: tibial spurs of 3, 2, 1.3.2, 1.3.3 or 2.3.3, spurs, particularly of foretibia in some
species, may be very small; tibiae and tarsi of mid-legs of 3 slightly broadened, of 2 very
strongly so, forming swimming legs with fringe of stiff setae. Antennae (3 2 ) up to twice fore-
wing length, scape rounded, less than half head length. Vertex with one pair large anterior
warts, behind these a transverse ridge interrupted by the mid-cranial sulcus; posterior head
warts inconspicuous or absent. Palpi vestigial. Face glabrous, subquadrangular. Wings prui-
nose, appearing naked, sometimes with several lines of setae in apical cells of forewings. These
are the largest African Hydropsychidae, forewing length 13-25 mm. Forewings long, narrow,
widest near arculus, apical border in 6 may be sinuous and more strongly produced than in 2 ;
discoidal cell closed, short, broad, median cell larger, additional costal cross-veins present;
wing forks 1, 2, 3, 4, 5 (3 2). Hind wing much broader and shorter than forewing, subtriangu-

398

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

lar, with forks 1, 2, 3, 5 or 2, 3, 5; no true discoidal or median cells, anterior margin with two
rows curved macrotrichia for wing coupling. Insects dark green to yellowish brown or whitish,
wings colourless to pale green, in African species unmarked except in guttatus and similis ,
which have brown patches; some species with a conspicuous pair of dark spots on mesoscutel-
lum; fifth sternite (3 2) with paired oval apertures.

P. C. Barnard (1980) has revised the genus, giving key, full descriptions and figures of the
19 known species, and dividing it into two species-groups, the nigricornis group, with small
eyes and often with dark spots on the mesoscutellum, and the ocularis group, with large-eyed
males and lacking dark spots on thorax. Barnard suggests that these two groups may prove to
warrant separate generic status, but is leaving them as species-groups for the present. Compo-
sition of the two groups in Africa is indicated in the species list pertaining to this genus.

The genus Polymorphanisus has been recorded from the Afrotropical Region, including
Madagascar, from the Oriental Region from India through South-east Asia to Indonesia, and
from the Philippines and Borneo. Of the 19 known species, seven are Afrotropical, one
( P . guttatus) being endemic to Madagascar, while the rest are all from the Afrotropical main-
land.

Polymorphanisus bipunctatus (Brauer)

(Figs 177-191, Tables XVI, XVII)

Oestropsis bipunctata Brauer, 1875: 73. 9 . Blue Nile.

Lectotype 2, designated by P. C. Barnard (1980:81), in Naturhistorisches Museum,
Vienna.

Polymorphanisus bipunctatus (Brauer); Ulmer 1907b: 20-22, figs 4, 6-9. <3 2 . Natal.

Polymorphanisus bipunctatus (Brauer); P. C. Barnard, 1980:81, figs 47-53 (2 wings,
head, thorax, genitalia).

Distribution: Sudan (type locality Beni Sciangul, Blue Nile); widely spread in the Afrotropical
Region from the Sudan through Ethiopia and Kenya in the north, to Angola, Zimbabwe and
South Africa (Natal, Transvaal) in the south.

Description of imagos (Figs 177-182)

P. bipunctatus appears to be the largest of the Afrotropical species, body length about 22 mm,
forewing 22-26 mm in 3 , up to 28 mm in 2 ; description below taken from pinned specimens in
this collection, checked against others in spirit; colour from dry specimens.

Male imagos (see Figs 181, 182): cream-coloured insects with black eyes and a pair of
large, rounded black spots on mesoscutellum. Antennae very long (up to 40 mm), slender,
scape and pedicel yellowish-brown, flagellum brown with darker annulations at joints. Head:
face smooth, shining, without warts; vertex with usual pair of large anterior setose warts, be-
tween these a prominent raised dorsal triangle, posterior to them a transverse ridge; posterior
warts represented by a small pair of setose patches, clearly demarcated in some specimens,
scarcely visible in others. Palpi rudimentary. Pronotum high, narrow, collarlike, with deep
median division, a pair of long, narrow median warts and a smaller pair of lateral warts, all
strongly hairy. Meso- and metanota without warts, largely smooth, shining, but with some
long, pale anterior and posterior setae, particularly on metanotum; mesoscutellum with pair of
large black spots and fringe of pale setae. Wings apparently naked, pruinose, covered with
microtrichia, in life pale green, in dry specimens cream to pale brown. Legs yellowish, darken-
ing distally to orange-yellow, a small dark spot at apices of tibial and tarsal joints; forelegs
slender, mid-legs somewhat widened, fringed, oarlike, hind legs not widened, but tibiae and
tarsi with short fringe of stiff setae. Tibial spurs 1.3.3 or 2.3.3. Abdomen cream with purplish

399

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Figs 177-184. MACRONEMATINAE: Polymorphanisus bipunctatus, 2, <3 pupa (mature). 177. 2: dorsal view of
head, pro- & mesonota, 178. 2: head, lateral, note vestigial palpi, 179, 180. 2 fore & hind wings (left), 181. 6 pupa:
genitalia, dorsal, 182. 6 pupa: mid-leg, pupal exuviae removed, 183. 6 pupa: abdomen, dorsal (genitalia and apical part
of exuviae removed), 184. 6 pupa: apical part of exuviae, ventral.

400

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

marks, darkening in dry specimens, a pair of inconspicuous oval apertures on sternum of V,
small lateral segmental processes present, possibly remains of pupal gills. Genitalia brown; ter-
minal segment of claspers short, wide, ninth segment relatively broad ventrally, rounded dor-
sally.

Females resemble males in colouring and general appearance, excepting that mid-legs are
very strongly widened and antennae have scape and pedicel cream above, black beneath, 3rd
to 6th segments blackish, rest of flagellum yellowish to brownish cream with brown annula-
tions, paling distally. $ genitalia have sternal plates of VIII widely separated to base, inner
margins reflected and strongly setose; tenth segment forms an apically setose dorsal hood and
has a median ventral sclerite; XI bears the usual pair of cerci and two pairs of small lobes. (See
Figs 177-180, which also apply largely to males, though the wings differ slightly in details and
the antennae in colour.)

The large size of the imagos, and the conspicuous black spots on the mesoscutellum, make
recognition of this species easy.

Description of larva (Figs 185-191, Table XVI)

As Polymorphanisus bipunctatus is the only species that has been recorded from Southern
Africa, the larvae described here have been taken as belonging to it, particularly as all the lar-
vae examined, both from the Eastern Transvaal and from Natal, evidently represent a single
species. The only available pupa, though bred out in the laboratory, was not associated with
any larval sclerites.

Mature larva (in spirit; Figs 188-191, Table XVI; Figs 185-187, of a fifth instar larva, may
also be used for comparison): a very large, powerful larva, length of present specimens up to
42 mm, greatest width (abdomen) 5 mm. Sclerotized parts smooth, shining, rich chestnut
brown to mahogany in colour, head and thoracic nota strongly patterned with paler brown to
yellowish markings; abdomen pinkish buff to mauve (blue-green in life).

Head: ( cf . Figs 185-187): long, slender, without carina, frontoclypeal apotome narrow,
widening anteriorly, anteclypeus membranous; eyes medium-sized, black, placed well forward;
antennae minute, at base of mandibles; anterior ventral apotome triangular, with suture on left
side only; posterior ventral apotome and stridulatory files absent.

Mouthparts (Fig. 190 & cf. Figs 185-187): labrum large, rounded, without lateral expan-
sions but with numerous small setae on dorsum and forming cushions under lateral margins;
mandibles strong, blackish, with narrow apices, each with two large, blunt apical teeth and sev-
eral small lateral teeth, the proximal ones poorly defined, no brushes or rows of setulae, outer
side rounded, not keeled, bearing two or three long setae; maxillary palpi five-jointed, exten-
sile, very long when extended, labial palpi minute; submentum subquadrangular.

Thorax (cf. Figs 185, 186, 188a): segments very long and slender compared with abdomen,
patterned as indicated; prosternum with narrow transverse sclerotized plate, anteriorly flanged,
articulated laterally with pronotum, behind it a second median sclerite concealed beneath the
integument; pre-episternum collarlike, with anterior point, row of flattened setae and inner
strengthening ridge, second pleural sclerite and epimeron large, conspicuous, glabrous. Meso-
sternum with one pair of tracheal gills, metasternum with two.

Legs (Figs 188, 189): foreleg with stout coxa, trochanter divided, femur strongly bent,
with dense band of setae along inner side and ventral margin, this is continued narrowly along
tibia and tarsus and on distal part of trochanter; claw stout, with long, slender basal bristle.
Mid- and hind legs more slender, with thin marginal fringe of setae, claws smaller, with stout
basal spine and small point. Legs subequal in length, foreleg slightly shortest.

Abdomen (Fig. 191 & cf. Figs 185, 186): tough, leathery, eight rows of tufted gills present
on segments I- VII, four on segment VIII (see Table XVI). Gills have straight main stem and

401

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Figs 185-191. MACRONEMATINAE: Polymorphanisus bipunctatus, larvae. 185. 5th (penultimate) instar larva: ha-
bitus, lateral (right appendages only shown), note absence of ventral sclerites on IX, 186. dorsal view of head, thorax &
abdomen I, II of same, 187. ventral view of head of same, 188, 189. mature larva: foreleg & hind leg (left, median
view); hind claw further enlarged. 188a. outer view of pleural sclerites of right foreleg (stippled areas indicate sclero-
tized structures shown by transparency), 190. mature larva: left & right mandibles, dorsal, 191. Gill from abdominal seg-
ment IV (mature larva).

402

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

many slender filiform branches (40-50 per gill stem anteriorly, fewer posteriorly), but appear
small in comparison with large size of larva; most lateral and ventrolateral gills have two main
stems set so far apart that they almost appear to represent separate gills. No ventral sclerites
on VIII, but several long setae; ventral sclerites on IX represented by a patch of pubescence
and a few longer setae; two minute lateral seta-bearing sclerites present. Anal appendages fair-
ly long, slender, with stout lateral sclerites and simple, strongly curved anal claw with few se-
tae, no anal brush. Anal gills not visible on available specimens, but probably five in number
since Aethaloptera, the other genus in this Tribe, has five.

Earlier instars: the 17 larvae available were measured as before (width of head capsules at
eyes) and other features studied. On this basis they were tentatively sorted as follows, number
of specimens in each case given in brackets (see also Table XVI). The instars have been taken
as being six in number as in the other Macronematinae studied, but this may of course prove to
have been an erroneous assumption should a large amount of material become available.

Third instar (3 larvae): head width 0,38-0,43 mm. Gills as in Table XVI, all gills simple,
unbranched, each gill ending in a single filament; ventral gills each bear a seta ( cf . Fig. 99a);
no gills with double stems as yet. Body setae very long; colour pattern not yet distinguishable,
but frontoclypeal apotome and anterior margin of thoracic nota darkened.

Fourth instar (3 larvae): head width 0,63-0,64 mm. Gill numbers and arrangement as in
the mature larva (Table XVI), but each stem with only 3-6 filaments; all gills which will have
two main stems now have them, arising from a common base. Colour pattern j ust distinguish-
able.

Table XVI:

Number and position of tracheal gills in 3rd to 6th larval instars of
Polymorphanisus bipunctatus

Segment

3rd instar

4th, 5th, 6th instars

Mesothorax

1 pair

1

pair

Metathorax

2 pairs

2

pairs

Abdomen (right side)

L

Post.

Pre.

V

L

Post.

Pre.

V

1

1

1

1

1

2*

1

1

1

II

1

1

1

1

2

2

2

1

III

1

1

1

1

2

2

2

1

IV

1

1

1

1

2

2

2

1

V

1

1

1

1

2

2

2

1

VI

1

1

1

1

2

2

2

1

VII

1

1

1

—

2

2

2

1

VIII

—

1

1

—

—

2

2

—

* Abdominal gills marked “2” each have two main stems, in instars 4-6 usually separated more or less widely at base.

On segment I the post- and pre-segmental gills are alongside one another and could be branches of the same gill.

L = lateral gill.

Post = postsegmental ventrolateral gill.

Pre = presegmental ventrolateral gill.

V = ventral gill.

403

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Fifth instar (6 larvae) (Figs 185-187): head width 1,0-1, 2 mm. Gill stems with 15-22 fila-
ments; where gill stems are double the bases may already be widely separated. Colour pattern
reasonably clear, but not as strong as in mature larva. No ventral setose plates on VIII or IX,
only small patches of long setae. The paired lateral setae on IX arise from very small sclerites.
(Figs 185-187 were made from a fifth instar larva owing to the difficulties posed by the size of
the mature larva.)

Sixth instar (mature larva, 5 specimens) (Figs 188-191): head width 1,5-1, 9 mm. Gills very
bushy, each stem with 40-50 filiform branches. Ventral setae on VIII and IX much as in fifth
instar, also minute paired setose lateral sclerites on IX; no ventral plates. Colour pattern fully
developed (as in Fig. 185 but darker).

Description of pupa (Figs 183, 184, Table XVII)

A single good male pupa from Natal was available in the collection. This is described be-
low.

Male pupa (in spirit; Figs 183, 184, Table XVII): very large, length c. 23 mm, head round-
ed, two pairs of small setae on vertex; antennae more than twice body length; labrum small,
transversely oblong, with thick fringe of marginal setae; mandibles lacking, maxillary palpi ves-
tigial, labial palpi absent. Paired dorsal plates present as follows: presegmental plates on seg-
ments III- VII, each with a single very large hook bearing 1-3 very small ones, hooks increas-
ing in size aborally. Postsegmental plates on segment III large, oval, with many small hooks,
no postsegmental plates on IV as recorded by Marlier (1962) for specimens from Lake Tanga-
nyika (later identified as probably P. elizabethae by P. C. Barnard. 1980). Small paired patches
of setae occur on segments II— VII , on III situated between the posterior plates. Paired tracheal

Table XVII:

Tracheal gills and dorsal plates of 6 pupa of Polymorphanisus bipunctatus

Abdominal segment

Tracheal gills

Dorsal plates (paired)

DL

VL

V

I

—

2

—

II

2*

2

2

III

2

2

2

Presegmental plates

Postsegmental plates

IV

2

2

2

Presegmental plates

V

2

2

2

Presegmental plates

VI

2

2

2

Presegmental plates

VII

2

2

2

Presegmental plates

VIII

—

—

—

—

* Gills marked “2" are each represented by two main stems, in a few cases arising from a common base, otherwise well
separated. The dorsalateral and ventrolateral gills are so placed that they are, respectively, presegmental and postseg-
mental in position. All are tufted branchial gills and none has an inflated base.

DL = dorsalateral gill.

VL = ventrolateral gill.

V = ventral gill.

404

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

gills occur on segments I-VII, in three rows on each side except on segment I (see Table
XVII). The dorsolateral and ventrolateral gills usually lie close together, in post- and preseg-
mental positions. Gills have paired stems, each forming a long, feathery branch which is easily
broken off. None of them have pouch-like bases. Anal appendages large, apically widened,
each terminating in several strong blunt points; on ventrolateral surface many setiferous tuber-
cles, increasing in size towards apex of appendage. Usual ventral pockets enclosing male geni-
talia. Tibial spurs 1.3.3. Mid-legs widened and fringed.

Pupal case: no case has to my knowledge been described, but see comments under Biol-
ogy below.

Remarks

P. C. Barnard in his ‘Revision of the Old World Polymorphanisini’ (1980) has sorted out
the nomenclatural problems of the species of Polymorphanisus and reference should be made
to his paper when African specimens have to be identified. As mentioned earlier on, he has
divided the genus into two species-groups, the nigricornis- group and the ocularis- group, the
composition of which as far as African species are concerned is reflected in the species list
given below.

All material collected to date in South Africa falls into bipunctatus, a very widely distrib-
uted species belonging to the nigricornis-group. Barnard retains bipunctatus as a valid species
separate from the Oriental species nigricornis to which it is closely allied. Of the other three
African members of this group, elizabethae includes specimens with a wide variety of black
mesoscutellar markings whereas two new species, hargreavesi and marlieri, lack such spots.
Those two species known only in the female sex differ from one another in antennal colour.

African members of the ocularis- group include angustipennis, similis (of which bisignatus
is a synonym) and guttatus , all of which lack dark markings on the mesoscutellum.

The three colour varieties of P. bipunctatus, described by Marlier (1965/66) a Forma A,
Forma B and Forma incolor, are referred by P. C. Barnard to P. elizabethae (as is also P. pu-
pillatus Navas). Those specimens of Forma incolor which entirely lack the typical black spots
would, however, belong to one of the species lacking mesoscutellar spots but as all are females
it is not clear which. Marlier and Botosaneanu’s (1968) ‘ Polymorphanisus cf. similis' has
proved to be P. angustipennis (Barnard 1980).

A larva (as “Macronematinae Larva C’) and pupa of Polymorphanisus were first described
by Marlier (1943c: 85-88, fig. 13) who later (1961) identified the larva as P. bipunctatus , com-
mented briefly on both stages and gave some biological notes. Barnard (1980), however, com-
mented that the larvae were probably of elizabethae as the adults collected at the same time
were elizabethae. This explains the small discrepancies between the South African larvae and
pupa of bipunctatus and those described by Marlier, mentioned below, which would thus indi-
cate species differences.

Marlier's pupa from Tanganyika had a single gill stem in three places where the South
African one has two; he also found small postsegmental dorsal plates on segment IV in ad-
dition to those on III. Apart from those differences the pupae seem to be very similar. Aetha-
loptera, the other African genus in the Polymorphanisini, also has postsegmental plates on III
only.

Biology

Marlier (1958, 1961) has published observations on the biology of larvae found in Lakes
Tumba and Kivu. Very young larvae were often found in the floating sudd where they mined
the older grass stems. Mature larvae inhabited galleries in rotting submerged wood, often using
pre-existing crevices, and spun fine longitudinal nets within the gallery in which micro-organ-
isms (bacteria and algae) were trapped for food. Marlier was able to watch their activities in

405

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

glass tubes, closed at one end, in the laboratory aquaria. He noted that the net was fastened to
the sides of the tube but was incomplete basally so that the larva retained space to pass it. The
larva usually remained curled in a U-shape with head and thorax to one side of the net and ab-
domen and anal prolegs to the other. Constant undulations of the abdomen kept the water cir-
culating assisting both with respiration and in food capture. Marlier could not see whether the
larvae collected food from the net or ate the net at intervals but did note that their guts were
filled with microscopic green algae. The feeding habits of Polymorphanisus larvae are thus
closer to those of some species of Macrostemum than to those of the other Macronematinae.

Marlier found Polymorphanisus larvae in lakes, large rivers and torrents. He considered
that in its mining habits the genus was comparable with the mayfly Povilla adusta. The larvae
in the Albany Museum Collection all came from stones in current in small rivers. Their habits
there are unknown although the blue-green colour of the abdomen may well indicate that
microscopic algae form their main food supply. Imagos in the Albany Museum collection were
taken in early summer (October — December) and larvae in various months, probably being
present all the year round.

It seems probable that the larvae may pupate in the larval dwelling tubes simply plugging
the opening loosely with debris. Wiggins (1960) in a study of Phryganeidae, the pupae of which
lack mandibles, concluded that such pupae have no need for mandibles as they do not have to
cut their way out of the pupal cases but merely have to push out a loosely inserted plug. Poly-
morphanisus pupae are included in Wiggins' discussion.

African species of Polymorphanisus

A. P. nigricornis- group (eyes small, well separated ventrally)

For other characters and comments see P. C. Barnard, 1980: 78, 79; he also gives keys,
descriptions of species and drawings.

P. bipunctatus (Brauer), 1875: 73, as Oestropsis. 2. Ethiopia.

P. C. Barnard, 1980: 81, figs 47-53 (d, 2 wings, 9 head, thorax, 6, 9 genitalia), map
(fig. 72). He places P. bipunctatus pupillatus Navas (1931b: 139, fig. 75), in synonymy
with P. elizabethae Navas, and considers Marlier's material of “bipunctatus" (1961:
208; 1965/1966: 40) to be in fact P. elizabethae.

Distribution: Sudan, Ethiopia, Kenya, Angola, Zimbabwe, South Africa (Natal,
Transvaal).

P. elizabethae Navas, 1931b: 140, fig. 76. 6 .

P. C. Barnard, 1980: 81, 84, figs 54-71 (d, 2 wings, genitalia, variations in mesoscutellar
spots), also map (fig. 72). Barnard synonymizes P. bipunctatus pupillatus Navas
(1931b: 139, d, Zaire) with elizabethae; this synonymy naturally includes P. pupilla-
tus Navas, erected to generic status by Marlier ( 1965/66: 22, 77).

Distribution: Sierra Leone, Ghana, Nigeria, Cameroun, Congo, Zaire, Uganda, Zambia,
Zimbabwe, Angola.

P. hargreavesi P. C. Barnard, 1980: 87, figs 74-76 ( 9 wings, head, genitalia), map (fig. 73). 2 .
Distribution: Sierra Leone, Zaire, Zambia, Zimbabwe.

P. marlieri P. C. Barnard, 1980: 64, 87, figs 77, 78 ( 9 wings, 8th sternite), 2 .

Distribution: Zaire.

P. nigricornis Walker, 1852: 79. d. Not African: see note below.

P. pupillatus Navas, 1931b: 139, fig. 75. d.

Marlier (1965/66: 22, 77) raised this to generic status and described 2 .

P. C. Barnard (1980: 81) places it in synonymy with P. elizabethae Navas.

Distribution: former Belgian Congo, Angola.

406

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Note:

P. nigricornis Walker, 1852: 79. 6 . North India.

Brauer 1875 recorded this species from Central Africa and India.

Betten & Mosely (1940) suggested that P. nigricornis and P. bipnnctatus were almost cer-
tainly synonymous.

P. C. Barnard, 1980: 90, figs 91-93 (d thorax & genitalia), corrects earlier erroneous de-
scriptions and misidentifications of these two species, concluding that nigricornis is
much rarer and more restricted than thought earlier, and that although it is very simi-
lar to bipunctatus, there are differences in the d and 2 genitalia which, together with
the widely separated distributions, suggest that the two species should be kept dis-
tinct.

Distribution: India, Vietnam, Sumatra, Java.

B. P. ocularis- group (eyes of male large, almost meeting ventrally)

For other characters and comments see P. C. Barnard, 1980: 78, 79, 98.

P. angustipennis Ulmer, 1912: 97, fig. 21. d.

P. C. Barnard, 1980: 98, figs 120-124 (d wings, head, genitalia), figs 125-128 (2 wings,
head, 8th sternite).

Distribution: Ghana, Nigeria, Cameroun, Sudan, Uganda.

P. bisignatus Navas, 1931c: 276, fig. 61. 2. Belgian Congo.

P. C. Barnard (1980) places this in synonymy with P. similis Ulmer.

P. guttatus Navas, 1935: 71, 72, fig. 32. 9 .

P. C. Barnard, 1980: 98, 99, figs 133, 134 ( 2 wings & 8th sternite).

Distribution: Madagascar.

P. similis Ulmer, 1912: 96-97 , figs 19, 20. <3 .

Marlier (1961: 159, 211, 212, fig. 26) description of 2.

P. C. Barnard, 1980: 103, figs 129-132 (d genitalia, 2 wing & 8th sternite). Barnard
places P. bisignatus Navas in synonymy with similis , and points out that Marlier &
Botosaneanu’s (1968) “ Polymorphanisus cf similis” is actually P. angustipennis.
Distribution: Sierra Leone, Nigeria, Cameroun, Zaire, Uganda.

Genus AETHALOPTERA Brauer 1875

Aethaloptera Brauer, 1875: 71-72.

Type-species Aethalopera dispar Brauer, 1875: 72. d. Senegal.

Chloropsyche McLachlan, 1880: lxix, pi. Lvii, 5 figs.

Type-species Chloropsyche evanescens McLachlan, 1880: 69. 6 .

Synonymized by Kimmins, 1962a: 96.

Primerenca Navas, 1915: 181, fig. 6. 6 .

Type-species Primerenca maesi Navas, 1915: 182, figs 6a-c. 6 .

Synonymized by Lestage, 1919a: 293.

Aethaloptera Brauer; Kimmins 1962a: 94, figs 43-63, synonymy and discussion.

Paraethaloptera Martynov, 1935: 193. 9 only.

Type species Paraethaloptera gracilis Martynov, 1935.

Synonymized by P. C. Barnard, 1980: 66.

Aethaloptera Brauer; P. C. Barnard, 1980: 66, 68, generic description, remarks and key to
species.

Aethaloptera Brauer; Lepneva 1964 (1970): 617-627, description of larva of A. rossica Marty-
nov.

407

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Generic diagnosis (derived from P. C. Barnard, 1980, with additions from Ulmer 1907d, Mar-
lier 1962b, and study of available material):

Imago: tibial spurs of 3 , 9 , 0.2.2 or 0.3.2. Sub-apical spur of mid-tibia near base at outer
side; spurs small, unequal. Legs slender in 3, mid-tibiae slightly dilated; in 2 tibiae and tarsi
of mid-legs greatly dilated, fringed. Antennae of 6 three to four times as long as forewing, of
2 little longer than wing; scape almost as long as head, flagellum very long, slender. Eyes
small, blackish, not meeting ventrally, slightly larger in 2 . Vertex with two pairs of setose
warts, the anterior larger than the posterior, latter weakly developed in 2. Palpi absent or ru-
dimentary. 6 forewings long, basal half narrow, apical half widened along postcostal border;
several false cross-veins between C and Sc, discoidal cell absent, but a false discal cell enclos-
ing the corneous spot, formed by R.4 and R5 rejoining shortly after separating; apical forks 1-5
present, though 5 may be a false fork between Cuia and Cu?; median cell large, bent. Hind
wing subtriangular, very wide basally, with forks 1, 2, 3, 5; discoidal cell absent. 9 wings short-
er, forewing wider, hind narrower than in 6 . Wings glabrous, unpatterned, greenish to brown-
ish or colourless. 6 forewing 10-18 mm, of 2 8-10 mm.

The genus occurs throughout most of the Afrotropical Region, Egypt, India, Sri Lanka,
South-east Asia and Indonesia, to Northern Australia, with one species in the U.S.S.R.

Aethaloptera maxima Ulmer
(Figs 192-293, Tables XVIII, XIX, Fig. 214D)

Aethaloptera maxima Ulmer, 1906: 62, fig. 66 6 . Orange Free State.

Holotype 3 was in Hamburg Museum; later destroyed (Weidner 1964: 66). Type
locality Bothaville, Orange Free State, South Africa, 25. III. 1899.

Neotype 3, South Africa: Waterval River, National Road between Standerton and
Greylingstad, 12. I. 1959; in Zoologisches Museum, Hamburg; designated by
P. C. Barnard, 1980: 72.

Chloropsyche maxima (Ulmer), 1907d: 17, fig. 2. 3 . Orange Free State.

Chloropsyche maxima (Ulmer); K. H. Barnard 1934: 368, figs 40 b-g, also suggested larva
from Natal (fig. 40 h-1); this however proved to be of Macrostemum capense.

Aethaloptera maxima Ulmer; Kimmins 1962a: 100, figs 40, 41, 45, 47, 56-59, 62, 63 (d, 2
wings, mid-legs, genitalia).

Aethaloptera maxima Ulmer; P. C. Barnard 1980: 72, figs 29-34 (3 , 9 wings & genitalia),
also distribution map (fig. 28).

Distribution: South Africa (Orange Free State, Transvaal, Northern Cape, Natal), South West
Africa, Mozambique, Zimbabwe, Zambia. There is much material in the Albany Museum, in-
cluding males, females, larvae and pupae; this is mainly from South Africa, but five imagos
come from Zimbabwe.

Description of Imagos (Figs 192-198)

Imagos (in spirit): fairly large, tawny yellow insects with cream to purplish abdomen, fore-
wings almost hairless, slightly milky, membrane covered with microtrichia, a few scattered se-
tae present, mainly along veins, several dark marks on most cross-veins (see Fig. 195); hind
wing iridescent, with fork 2 stalked (sessile in A. dispar). Legs yellowish with dark apical spot
on mid-tibia, tibial spurs 0.2.2. Males have plain brownish antennae, and a fuzz of soft, fine,
white hairs on sternum, proximal end of mesoscutum and upper parts of legs. Females differ
somewhat in appearance, having strongly striped antennae (segments proximally white, distally
brown), and lacking the soft white hairs. Both sexes are heavy-bodied, with stout thorax and
fairly stout abdomen; genitalia are small, those of male with slender claspers and large-ended

408

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Figs 192-202. MACRONEMATINAE: Aethaloptera maxima, 6,9, mature 6 & 9 pupae. 192. 6 : dorsal view of head,
pro- and mesonota, 193, 194. 6: frontal & lateral views of head, note vestigial palpi, 195, 196. 6: fore & hind wings
(right), 197. 6 genitalia, dorsal (VIII removed to show IX, retracted within it), 198. 9 genitalia, ventral, 199. 9 pupa:
head, left lateral, 200. 9 pupa: apical part of abdomen with pupal appendages, ventral, 201. 6 pupa: pupal appendages
from exuviae, ventral, 202. d pupa: left & right mandibles from exuviae, dorsal.

409

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

copulatory organ, those of female with sternal plates of IX apparently not separated medially
(see Fig. 198 and also figs 60-63, Kimmins 1962a). Length of male forewings 14-18 mm, of fe-
male 9-13 mm.

Description of Larva (Figs 203-213, Table XVI11, Fig. 214D)

The 89 available larvae include at least four, possibly five or six, instars. Instars 3-6 were
clearly Aethaloptera and easily recognizable as such. Of the very young larvae, there were
three that looked as though they might belong to Aethaloptera , but as they were colourless and
transparent it was difficult to be certain. They have, however, been included with a query.

Mature larva (Figs 203-213, Table XVIII): length up to 15 mm, width of abdomen up to
3 mm, all sclerotized parts pale yellow to brownish, largely transparent (muscles etc. show
through).

Head (Figs 203-205), broadly oval, dorsally rounded, without carina; frontoclypeal apo-
tome triangular with two pairs of very slight lateral indentations, anterior margin strongly con-
vex, with paired lateral emarginations; anteclypeus membranous; eyes black, medium-sized,
situated well forward within a paler area; antennae minute, near base of mandibles. Anterior
ventral apotome small, triangular, clearly demarcated on left side only; posterior ventral apo-
tome minute; no stridulatory files.

Mouthparts (Figs 204—206): labrum large, retractile, lightly sclerotized, with partial medi-
an sulcus, and mebranous lateral expansions fringed anteriorly and laterally with long, silky se-
tae, dorsal surface with two long brown setae and a number of shorter colourless ones; maxil-
lary palpi five-segmented, galeae small, palpifers with large tufts of lateral setae, labium stout,
labial palpi minute, submentum a truncated oval, distal end concave, without setae, lateral
margins setose. Mandibles (Fig. 206), large, strong, left one with five terminal teeth, right with
four, each mandible with one lateral tooth situated midway along inner edge; no inner brushes
or rows of setulae, scattered setae along outer rounded margin.

Thorax (Figs 203, 207): thoracic segments short, wide, pale yellowish brown, anterior and
lateral margins darkened, the latter with black thickenings, pronotum with transverse sulcus
near posterior margin; mesonotum with faint brown diagonal stripes extending half way across
shield and central black spot on posterior border; metanotum with darker brown diagonal
stripes extending right across shield, no median black spot. Pre-episternum long, slender, dag-
ger-shaped, as long as coxa. Prosternal plate sclerotized, yellowish, with wide anterior and nar-
row posterior parts.

Legs (Figs 208-210): sub-equal in length, pale yellowish brown with slight darkening at
joints. Foreleg with distal spine on anterior surface of coxa; trochanter divided; long setae
fringing trochanter and femur, fairly thick brush on tibia and tarsus. Mid- and hind legs with
relatively few setae apart from a thin fringe along posterior border of tibia and tarsus, which
also have larger bladelike apical bristles. Claws small, shown enlarged, of unusual structure;
anterior tarsal claw with two sharp basal spines, those of mid- and hind tarsi each with three
basal spines and a “cushion” edged with spinules (the third spine and “cushion” are only clear-
ly visible under a compound microscope, see Figs 208a, 209a, b, 210a, b).

Abdomen (Figs 203, 211-213): smooth, bare, except for two transverse areas on segments
1 and II respectively, which are grooved and covered with minute hooked spinules as in Am-
phipsyche. Tufted tracheal gills (15-20 filaments per stem) present on mesosternum, metaster-
num and abdominal segments as shown in Table XVIII and Fig. 203. Thoracic gills have fila-
ments at tip and along one side only; abdominal gills have a more ring-like arrangement but
filaments tend to bunch. Certain gills have two main stems; in these the bases may be contigu-
ous, but as the abdomen of a mature larva increases in size they usually become quite widely
separated. Abdominal gill count as follows: six rows on segments I and VII, eight rows on seg-
ments II to VI, none on segment VIII. Sternum of IX with two sub-ovate sclerites bearing

410

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Figs 203-213. MACRONEMATINAE: Aethaloptera maxima, mature larva. 203. habitus, left lateral (right appendages
only shown), pair of abdominal gills shown further enlarged, 204, 205. dorsal & ventral views of head, 206. left & right
mandibles, dorsal, 207. pro-, meso & metanota, dorsal, 208. left foreleg with pleural sclerites, 208a. claw of same, fur-
ther enlarged, 209. left mid-leg, 209a, b. claw of same, inner & outer views, further enlarged, 210. left hind leg, 210a, b.
claw of same, further enlarged, inner & outer views, 211. abdominal segments VIII & IX and anal prolegs, ventral, 212.
left anal proleg, further enlarged, lateral, 213. ventral plate from IX, further enlarged.

411

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

bladelike setae, also paired lateral patches of setae (no dorsal or dorsolateral sclerites, and no
ventral sclerites on VIII). Anal prolegs fairly long and stout with long setae on lateral sclerites
and inner side; lateral sclerites very lightly sclerotized, scarcely visible); anal hooks smooth,
plain, sharply curved; five anal gills present.

Earlier instars: comments on earlier instars follow, assuming that there are six altogether
as in Macrostemum and Amphipsyche , as will be the case if instars 1 and 2 definitely prove to
belong to this genus (see Fig. 214D).

Probable first instar (2 larvae): head width 0,15 mm; head large, rounded, labrum large,
mandibles with several small teeth along margin; pre-episternum present, comparatively large,
which is unusual in first instar larvae, sub-ovate, with downturned blunt apical point; tarsal
claw of foreleg with three basal spines, graduated in length; mid- and hind claws appear to be
similar, though longer basal spines are not in same plane. Gills absent. Anal claw large, clear.
Setae long, sparse, two to three long ventral setae on VIII and IX.

Probable second instar (1 larva): head width 0,22 mm; head large, rounded, labrum large,
frontoclypeal apotome with rounded anterior margin, mandibles large, solid, with about five
apical and lateral teeth and a median lateral projection; pre-episternum longer, with thick base
tapering to pointed apex; fore claw with two basal spines, mid- and hind claws unclear. Simple
gills present. Setae shorter, more numerous.

Third instar (4 larvae): head width 0,35-0,37 mm; pale but easily recognizable as Aetha-
loptera, head shape and mandibles typical, also dagger-shaped pre-episternum, now two-thirds
length of fore-coxa, with lateral seta, fore-coxa with distal spine present, apparently basally ar-
ticulated with fore-coxa. Gills now bear a single filament (branch), the ventral ones carrying a
seta (cf. Macrostemum , Fig. 99a); same number of gills as in later instars, but each has only
one main stem. Foreclaw with two short basal spines, mid- and hind claws each have a single
stout basal spine and ring of smaller spinules.

Table XVIII:

Tracheal gills of mature larva of Aethaloptera maxima

Mesothorax

Metathorax

1 pair

2 pairs

Abdomen (right side)

Post

Pre

VI

V

I

—

—

2*

1

II

1

2

2

1

III

1

2

2

1

IV

1

2

2

1

V

1

2

1

1

VI

1

2

1

1

VII

1

2

1

—

VIII

—

—

—

—

* The figure 2 denotes that the gill in question has two main stems, either arising from a common base or separately.
Post = postegmental lateral gill.

Pre = presegmental lateral gill.

VL = ventrolateral gill.

V = ventral gill. *

412

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Fourth instar (2 larvae): head width 0,55-0, 65 mm; in foreleg coxal spine is short, fused
to coxa, and carries a stout basal seta. Most gills bear two to three filaments, some ventral gills
still have a terminal seta and some gills now have double stems. Foreleg with thin brush, fore-
claw with two basal spines, mid- and hind claws with two stout basal spines (one larger than
the other) and small “cushion” surrounded by spinules (i.e. basically the same as in the mature
larva). Claw structure can only be clearly seen under compound microscope x 400.

Fifth instar (17 larvae): head width 0,87-1,0 mm; these resemble final instar larvae but
have fewer gill filaments (usually 6-12), none of which bears a seta.

Mature larvae, i.e. sixth instar (many larvae): head width 1,27-1,54 mm. These have been
described above.

Description of pupa (Figs 199-202, Table XIX)

Pupae available included mature males and females, in a few cases complete with pupal
cases and larval sclerites, thus affording correlation between larvae and imagos.

Male pupa — pharate imago — (in spirit. Figs 201, 202, & cf. Figs 199, 200, 2, Table XIX):
3 specimens, length c. 9 mm. Head yellowish, vertex rounded with four flattened groups of pale
setae, each of the two anterior groups situated above a small corneous projection, setae with
bent tips. Eyes medium-sized, dark; antennae very long, slender, brownish, forming a coil on
posterodorsal side of abdomen (four turns of each antenna in coil). Labrum somewhat bul-
bous, a few small, colourless setae and several long ones with bent tips centrally placed, rows
of setate tubercles giving a transversely striated appearance to labrum. Mandibles small, nar-
row, very similar, each with bulbous base and an apical and three lateral teeth, maxillary and

Table XIX:

Branchial gills, dorsal plates etc., of ? pupa of Aethaloptera maxima

Abdominal
segment no.

Tufted gills
(right side)

Dorsal

plates

Setae (shagreening
scarcely visible)

I

D1

VI

V

1 pair setae

II

1

2*

2

1 ant. pair

few short setae

III

(1)

2

2

1 ant. pair

few short setae

IV

(1)

2

2

1 post pair
(strips)

1 ant. pair

2 scattered areas

V

(1)

2

1

1 ant. pair

of long setae
few short setae

VI

(1)

2

1

1 ant. pair

few short setae

VII

2

1

1

1 ant. pair

few short setae

VIII

—

—

—

1 ant. pair

few short setae

* The figure 2 denotes that the gill in question has 2 main stems.

D1 = dorsalateral gills; brackets indicate that the gill in question has a broad pouch-like base.

VI = ventralateral gills; these have been bracketed with the dorsolateral gills when they arise at virtually the same place.
V = ventral gills.

413

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

labial palpi rudimentary. Fore and hind leg small, slender; mid-legs longest, broader, flattened,
with swimming fringes; tibial spurs 0.2.2. Tufted gills and dorsal plates present on abdomen as
shown in Table XIX. All anterior dorsal plates with two points only; posterior plates long nar-
row strips with many minute points; few setae on abdominal tergites, shagreening almost invis-
ible. Anal appendages of medium length, blunt-ended, curved dorsad, bearing strong ventrola-
teral and apical setae set on papillae; pockets for male genitalia present as usual.

Female pupa (in spirit, Figs 199, 200, Table XIX): very similar to male, but mid-legs more
expanded, antennae shorter, equally slender, not coiled but curled round; gill count as in male,
gills very clear: dorsal plates as in 6 , anal appendages very similar to those of 6 but smaller
and without pockets for genitalia. Two specimens, length c. 10 mm.

Pupal case: oval, examples incomplete, but evidently made from variously sized sand
grains, mainly either very large or very small, completely lined with silk except at ends; head
end with neat cap of stones overlying a silken meshwork, these stones being attached to the
meshwork round rim of lid only; lid partially surrounded by a raised flange of stones (probably
originally forming a complete ring); larval sclerites packed into posterior part of inner lining as
usual.

Remarks

Kimmins (1962a) discussed the genus Aethaloptera and concluded that there were only
two doubtfully distinct species in Africa, A. dispar Brauer (1875) and A. maxima Ulmer
(1906), the former having a more northerly distribution than the latter. P. C. Barnard (1980)
regards dispar as a widespread and variable species and considers maxima to be a valid species,
separable from dispar by hind wing venation and tibial spur formula. He further considers that
the two species meet at the Victoria Falls, as shown on his distribution map (1980 fig. 28). In
the Albany Museum collection there are adults from the Victoria Falls of which several are
maxima but one, which differs from them in colour and certain other features, proved to be
dispar (P. C. Barnard det.). All South African imagos are of maxima.

Various larvae have been ascribed in the literature to Aethaloptera (see Ulmer 1957),
however, of these the only valid ones appear to be those of A. rossica Martynov, described by
Lepneva (1964 (1970): 617-627, figs 762-773), and A. dispar , described by Gibbs (1973: 391,
figs 77-79). The larva figured by K. H. Barnard (1934: 368, figs 40h-l) as ? Chloropsyche is in
fact a Macrostemum. The larval and pupal stages of A. maxima have not, to my knowledge,
been described previously.

The maxima larvae are on the whole very similar to those of rossica and dispar but there
are distinct differences in mandibular teeth, anterior margin of frontoclypeal apotome, sub-
mentum and setation. The limb claws have a similar though not identical arrangement of basal
spines (spinelike setae) and they and the long, slender pre-episternum probably represent valid
generic characters, as does the type of mandible. Statzner (1981) in his comparison of dispar
with another species also used the margin of the frontoclypeal apotome, submentum and seta-
tion, together with comparison of the ventral sclerites on IX.

Gibbs (1973; Figs 92-94) figured the pupa of dispar (evidently 2), but without description.
I can find no description of Aethaloptera pupae in the literature. The anal appendages as
figured by Gibbs are similar to those of maxima , however, the mandibular teeth do show dif-
ferences. The absence of palpi is characteristic of the Polymorphanisini, to which Tribe Aetha-
loptera belongs.

Biology

Statzner (1981), who observed A. dispar larvae in streams in Ivory Coast, noted that they
built tubes which protruded into the current, somewhat altering the character of the substra-
tum. They were not able to survive dry periods without flow and were killed by the application

414

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

of the insecticide Chlorphoxim. They took some months to repopulate, presumably by imagos
from the nearest water, some 80 km distant. Instar distribution suggested a long larval life,
similar to that shown by the Cheumatopsyche species observed.

Aethaloptera maxima , like Amphipsyche scottae, is found in the Vaal River and its tributa-
ries, always amongst stones in current, and in similar habitats in Natal, South West Africa and
Zimbabwe as far north as the Zambezi River. It particularly favours localities below dam out-
lets or barrages where there is a good food supply. The retreats have not been seen, but may
well be similar to those of dispar. Larvae have been recorded from October to April, most
commonly in summer, and imagos over much the same period. Dissection showed gut content
to be very finely divided material, possibly indicating a diet of microseston as in some species
of Macrostemum.

African Species of Aethaloptera

A. dispar Brauer, 1875: 72, pi. iv, figs 4a-c, 6 9 .

Lectotype 6 , Senegal, near Taoue, xi.1869; in Naturhistorisches Museum, Vienna. Desig-
nated by P. C. Barnard, 1980: 68.

Primerenca maesi Navas, 1915: 182. Holotype 9, Zaire. Synonymy by Lestage, 1919: 293,
294.

Primerenca maerina Navas, 1916: 242 (unjustified emendation).

K. H. Barnard, 1934: 366-368, fig. 40a (<3 wings).

Kimmins, 1962a: 98-100, figs 38, 39, 43, 46, 48-55, 60, 61, <3 9 .

P. C. Barnard, 1980: 68-72, figs 19-27 & fig. 28 (distribution map); revision of genus and
redescription of species with synonymy.

Marlier, 1965/66: 22, figs 11a, b, recorded two new forms of dispar , from Angola, namely
Forma A and Forma B, both from females, of which Forma A is in fact A. maxima ,
Forma B A. dispar (Barnard 1980).

Distribution: Senegal, Sierra Leone, Ivory Coast, Ghana, Nigeria, Zaire, Sudan, Uganda,
Tanzania, Angola, Zambia, Malawi, Zimbabwe. P. C. Barnard considers that dispar
and maxima meet only at the Victoria Falls, so that the Transvaal and Madagascar re-
cords need confirmation.

A. maxima Ulmer, 1906: 62, fig. 66. 6 .

Neotype <3, South Africa, Waterval River, 12. i. 1959, in Zoologisches Museum, Hamburg,
designated by P. C. Barnard, 1980: 72.

Marlier, 1965/66: 40, A. dispar Forma A, misidentification of maxima.

Scott, 1975: 49, fig. 37, larva. (For other references see above under A. maxima.)

Distribution: South Africa (Orange Free State, Transvaal, Natal, Northern Cape), South
West Africa, Zimbabwe, Zambia, Angola.

A. sexpunctata (Kolenati), 1859: 226, pi. III. Not African, see Note below.

Note:

Ulmer, 1912: 95, recorded A. sexpunctata from the Cameroons with a query, since it is an
East Indian species. Navas (1932a: 288 & 1934b: 90), probably following Ulmer, re-
corded sexpunctata from the Belgian Congo and the Zambezi River. Kimmins (1962a:
100), suggested that the above records probably refer to dispar , possibly to maxima ,
but not to sexpunctata.

415

214

A

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Fig. 214 A-D. Larval instars of certain Hydropsychidae shown as block histograms, based on head width at eyes. A:
Cheumatopsyche thomasseti, B: Macrostemum capense , C: Amphipsyche scottae, D: Aethaloptera maxima , Y axes: num-
ber of individuals, X axes: a = head width at eyes in microcrometer unit scale (6,0 units = 1 mm) (actual head widths in

mm given in text), b = instar number.

416

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

ACKNOWLEDGEMENTS

The studies leading to the publication of this paper have been spread over many years,
and I am greatly indebted to many people and institutions for assistance, whether in the form
of information or specimens.

First of all, I would like to thank the Director, former Director and professional staff (par-
ticularly the Head of the Limnology Division) of the National Institute for Water Research of
the C.S.J.R., for which I worked in various capacities for some 26 years and from which
stemmed much of the material used in the present study. This material is now part of the Nat-
ional Collection of Freshwater Invertebrates of the Albany Museum.

Grateful thanks are also due to the Director and former Director of the Albany Museum,
in which this piece of work has been completed, for accommodation and assistance, and to the
Editor of Annals of the Cape Provincial Museum, Dr F. W. Gess, for his encouragement and
editorial work.

Many Trichopterist and ecologist colleagues have also been of great help, always willing to
provide information or to lend or give specimens when needed. In this context 1 must thank Dr
P. C. Barnard and, earlier, Mr D. E. Kimmins of the British Museum (Natural History), par-
ticularly for checking identifications; Dr O. S. Flint, Jr. (Smithsonian Institution, Washington)
for informatiqn about American species of Leptonema, Macronema and Macrostemum ; Dr
F. Schmid (Biosystematics Research Institute, Ottawa) for comments on Hydropsyche, Sym-
phitopsyche and Diplectrona\ Dr G. A. Schuster (Eastern Kentucky University) for material of
“ Symphitopsyche Mr D. G. Gibbs (U.K.) for specimens of Protomacronema , Leptonema and
Amphipsyche from Ghana; M. F.-M. Gibon (O.R.S.T.O.M.) for Protomacronema imagos
from Ivory Coast; Dr B. Statzner (Karlsruhe University T.H.) for imagos and larvae of Proto-
macronema, also from Ivory Coast; Dr G. Marker (Institut Royal des Sciences naturelles de
Belgique) and Dr J. Decelle (Musee Royal de l’Afrique Centrale, Tervuren) for the loan of
specimens of Diplectronella medialis and Sciadorus sinuatus\ also, not least, to the South Afri-
can Museum for Dr K. H. Barnard’s spirit collection of Trichoptera, on long loan to me, of
which much use was made.

I must also express my gratitude to Dr F. M. Chutter (Head of Limnology, National Insti-
tute for Water Research) for his interest in this project over the years and for sending me
much material stemming from his meticulous field work, including correlated series of several
of the species studied, as well as for the fruits of his ecological observations on them, together
with supplementary information for incorporation in the histograms.

REFERENCES

Banks, N. 1913. Synopses and descriptions of exotic Neuroptera. Trans. Am. ent. Soc. 39: 201-242.

Banks, N. 1920. Neuroptera, Panorpata and Trichoptera collected by the American Museum Congo Expedition, with
lists of the species known from the Belgian Congo. Bull. Am. Mus. nat. Hist. 43: 21-33.

Barnard, K. H. 1934. South African Caddis-flies (Trichoptera). Trans, roy. Soc. S. Afr. 21: 291-394.

Barnard, K. H. 1940. Additional records, and descriptions of new species, of South African Alder-flies (Megaloptera),
May-flies (Ephemeroptera), Caddis-flies (Trichoptera), Stone-flies (Perlaria), and Dragon-flies (Odonata).
Ann. S. Afr. Mus. 32: 609-661.

Barnard, P. C. 1980. A revision of the Old World Polymorphanisini (Trichoptera: Hydropsychidae). Bull. Br. Mus.
nat. Hist. Ent. 41 (2): 59-106.

Betten, C. 1934. The Caddis Flies or Trichoptera of New York State. N. Y. St. Mus. Bull. No. 292: 500pp.

Betten, C. & Mosely, M. E. 1940. The Francis Walker types of Trichoptera in the British Museum , London: Trustees of
the British Museum.

Botosaneanu, L. 1967. Trichoptera. In: lilies, J., ed., Limnofauna europaea. eine Zusammenstellung alter die
europdischen Binnengewasser bewohnenden mehrzelligen Tierarten mit Angaben iiber ihre Verbreitung und
Okologie Stuttgart: Gustav Fischer Verlag. Trichoptera pp. 285-309.

Botosaneanu, L. 1975. Trichopteres recuillis au Maroc par M. J. Miron. Folia ent. hung. (2) 28: 269-276.

Botosaneanu, L. 1980. Etude de quelques types de Trichopteres Ouest-Palaearctiques, deposes au Museum de Paris
(Insecta, Trichoptera). Bui. zool Mus. Univ. Amsterdam 7 (19): 189-196.

417

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Brauer, F. 1865. Zweiter Bericht iiber auf der Weltfahrt der kais. Fregatte Nevara gesammelten Neuroptcren. Verh.
zool.- bot. Ges. Wien 15: 415-422.

Brauer, F. 1868. Neue von Herrn Dr G. Semper gesammelte Neuropteren. Verh. zool.- bot. Ges. Wien 18: 263-268.

Brauer, F. 1875. Beschreibung neuer und ungeniigend bekannter Phryganiden und Oestriden. Verh. zool.- bot. Ges.
Wien 25: 69-74.

Brindle, A. 1960. The larva of Hydropsyche angustipennis Curtis (Trichoptera, Hydropsychidae). Entomologist's Rec.
72: 267-270.

Chutter, F. M. 1963. Hydrobiological studies on the Vaal River in the Vereeniging Area. Part 1. Introductory water
chemistry and biological studies on the fauna of habitats other than muddy bottom sediments. Hydrobiologia
21: 1-65.

Chutter, F. M. 1967. Hydrobiological studies on the Vaal River and some of its tributaries including an introduction to
the ecology o/Simulium in its lower reaches. Ph D. Thesis, Rhodes University, South Africa. 2 vols. pp. iv +
216, 27 figs in vol. 1, tables & 13 pis in vol. 2.

Chutter, F. M. 1968. On the ecology of the fauna of stones in the current in a large South African river supporting a
very large Simulium (Diptera) population. J. appl. Ecol. 5: 531-561.

Chutter, F. M. 1969a. The effects of silt and sand on the invertebrate fauna of streams and rivers. Hydrobiologia 34:
57-76.

Chutter, F. M. 1969b. The distribution of some stream invertebrates in relation to current speed. Int. Revue ges. Hy-
drobiol. 54: 413-422.

Chutter, F. M. 1970. Hydrobiological studies in the catchment of Vaal Dam, South Africa. Part 1. River zonation and
the benthic fauna. Int. Revue ges. Hydrobiol. 55: 445-494.

Chutter, F. M. 1971. Hydrobiological studies in the catchment of Vaal Dam, South Africa. Part 2. The effects of
stream contamination on the fauna of slones-in-current and marginal vegetation biotopes. Int. Revue ges. Hy-
drobiol. 56 (2): 227-240.

Corbet, P. S. 1958. Larvae of certain East African Trichoptera. Revue Zool. Bot. afr. 58: 203-213.

Corbet, P. S. 1966. Parthenogenesis in Caddisflies (Trichoptera). Can. J. Zool. 44: 981-982.

Crichton, M. I. 1957. The structure and function of the mouthparts of adult caddisflies (Trichoptera). Phil. Trans. R.
Soc. Ser. B. No. 677, 241: 45-91.

Crosskey, R. W. & White, G. B. 1977. The Afrotropical Region, a recommended term in zoogeography. J. nat. Hist.
11:541-544.

Curtis, J. 1834. Phil. Mag. 4: 213.

Curtis, J. 1835. British Entomology. Vol. 3. London, [pi. 544 (text), pi. 601 (text).]

Dakki, M. 1978. Le genre Hydropsyche au Maroc (Trichoptera, Hydropsychidae). Bull. Inst, scient. Rabat 3: 111-120.

Edington, J. M. 1965. The effect of water flow on populations of net-spinning Trichoptera. Mitt. int. Ver. Limnol. 13:
40-48.

Edington, J. M. & Hildrew, A. G. 1981. A key to the caseless caddis larvae of the British Isles with their ecology.
Scient. Pubis Freshwat. biol. Am. 43, pp. 92.

Fischer. F. C. J. 1963. Trichopterorum Catalogus IV. Hydropsychidae, Arctopsychidae. Amsterdam: Nederl. ent. Ver-
een. vi + 225 + (1) pp.

Fischer, F. C. J. 1972. Trichopterorum Catalogus XIII. Supplement to Vols III & IV. Amsterdam: Nederl. ent. Vereen.
vii + 172 pp.

Flint, O. S., Jr. & Bueno-Soria, J. 1979. Studies of Neotropical Caddisflies XXIV. The genus Macronema in Meso-
America (Trichoptera: Hydropsychidae). Proc. ent. Soc. Wash. 81: 522-535.

Flint, O. S., Jr. & Bueno-Soria, J. 1982. Studies of Neotropical Caddisflies XXXII. The immature stages of Macro-
nema variipettne Flint & Bueno, with the division of Macronema by the resurrection of Macrostemum (Tricho-
ptera: Hydropsychidae). Proc. biol. Soc. Wash. 95: (2): 358-370.

Fremling, C. R. 1960. Biology and possible control of nuisance caddisflies of the Upper Mississippi River. Res. Bull.
Iowa State Univ. 483: 856-879.

Gauthier, H. 1928. Rech. Faune Eaux contin. Algerie, Tunisie: 26.

Gibbs. D. G. 1973. The Trichoptera of Ghana. Dt. ent. Z. N.F. 20: 363-424.

Guerin, F. E. 1843. Insecta. In: Cuvier, Inconogr. Regn. Anim. : 395-397.

Hafiz, H. A. 1937. Rec. Ind. Mus. 39: 113-116.

Harrison, A. D. & Agnew, J. D. 1962. The distribution of invertebrates endemic to acid streams in the western and
southern Cape Province. Ann. Cape Prov. Mus. 2: 273-291.

Harrison, A. D. & Elsworth, J. F. 1958. Hydrobiological studies on the Great Berg River, Western Cape Province.

Part 1. General description, chemical studies and main features of the flora and fauna. Trans. R. Soc. S. Afr.
35 (3): 125-226.

Hickin, N. E. 1955. Larvae of some East African Trichoptera. Proc. R. ent. Soc. Lond. (A) 30: 155-163.

Hickin, N. E. 1956. The larva of an East African caddis fly, Cheumatopsyche copiosa Kimmins (Hydropsychidae). Proc.
R. ent. Soc. Lond. (A) 31: 132-133.

Hildrew, A. G. & Morgan, J. C. 1975. The taxonomy of the British Hydropsychidae (Trichoptera). J. Ent. (B) 43:
217-229.

418

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Hinton, H. E. 1963. The ventral ecdysial lines of the head of the endopterygote larvae. Trans. R. ent. Soc. Lond. 115:
39-61.

Hynes, H. B. N. & Williams, T. R. 1962. The effect of DDT on the fauna of a Central African stream. Ann. trop.
Med. Parasit. 56 (1): 78-91.

Jacquemart, S. 1957. Trichoptera des lacs Kivu et Edouard. Explor. hydrobiol. lacs Kivu, Edouard et Albert (1952-54)
3(2): 65-129.

Jacquemart, S. 1959. Trichoptera du lac Albert. Explor. hydrobiol. lacs Kivu, Edouard et Albert ( 1952-1954 ) 3 ( 3):
107-136.

Jacquemart, S. 1960. Trois Trichopteres nouveaux de Pile Maurice. Bull. Inst. r. Sci. nat. Belg. 36 (57): 1-6.

Jacquemart, S. 1961a. Un Trichoptere nouveau d’Afrique centrale: Macronema trilineata sp. n. Bull. Inst. r. Sci. nat.

Belg. 37 (13): 1-3.

Jacquemart, S. 1961b. Trichoptera. Bruxelles. Explor. Parc natn. Upemba Miss. G. F. de Witte (1946-1949): 1-46.

Jacquemart, S. 1961c. Contribution a I’etude de la faune Trichopterologique africaine. Trois Trichopteres nouveaux du
lax Tanganika. Explor. hydrobiol lacs Kivu, Edouard et Albert (1952-1954) 3 (4): 65-75.

Jacquemart, S. 1961d. Trichopteres nouveaux du Parc National de l’Upemba. Revue Zool. Bot. afr. 63 (3-4): 227-230.

Jacquemart, S. 1963a. Trichoptera. In Hanstrom et al., eds, S. African Animal Life 9: 335-415.

Jacquemart, S. 1963b. Trichopteres nouveaux des lies Maurice et de la Reunion. Bull. Inst. r. Sci. nat. Belg. 39 (39):

1-7.

Jacquemart, S. 1965. Contribution a l’etude des Trichopteres de l'Afrique. Un Trichoptere africain nouveau: Hydro-
psychodes marieni sp. n. Bull. Inst. r. Sci. nat. Belg. 41 (30): 1-3.

Jacquemart, S. 1966a. A propos de quelques Trichopteres du Katanga. Bull. Inst. r. Sci. nat Belg. 42 (11): 1-10.

Jacquemart, S. 1966b. Trichoptera. Explor. Parc natn Garamba Miss. H. de Saeger (1949-1952) 51 (2): 35-56.

Jacquemart, S. & Statzner, B. 1981. Trichopteres nouveaux du Zaire. Bull. Inst. r. Sci. nat. Belg. 53 (21): 1-25.

Jansson, A. & Vuoristo, T. 1979. Significance of stridulation in larval Hydropsychidae (Trichoptera). Behaviour 71:
167-186.

Johnstone, G. W. Stridulation by larval Hydropsychidae (Trichoptera). Proc. R. ent. Soc. Lond. (A) 39: 146-150.

Kaiser, P. 1965. Uber Netzbau und Stromungssinn bei den Larven der Gattung Hydropsyche Piet. (Ins.. Trichoptera).
Int. Revue ges. Hydrobiol. 50: 169-224.

Kimmins, D. E. 1956. A new species of Cheumatopsyche (Trichoptera, Hydropsychidae), from Central Africa. Entomol-
ogist 89: 196-197.

Kimmins, D. E. 1957a. New species of East African Neuroptera, Mecoptera and Trichoptera. Ann. Mag. nat. Hist. (12)
10: 36-41.

Kimmins, D. E. 1957b. New and little-known species of African Trichoptera. Bull. Br. Mus. nat. Hist. Ent. 6 (1): 1-37.

Kimmins, D. E. 1960a. A review of the African species of the genus Cheumatopsyche (Trichoptera, Hydropsychidae),
with special reference to those of Southern Africa. Bull. Br. Mus. nat. Hist. Ent. 9 (4): 253-267 .

Kimmins, D. E. 1960b. Proposed use of plenary powers to suppress the generic name Macronema Stephens 1829 so as to
preserve the generic name Macronema Pictet 1836 (Class Insecta, Order Trichoptera). Z. N. (S). 706. Bull.
Zool. Nomencl. 17 (9-11): 313, 314.

Kimmins, D. E. 1962a. New African Caddis-flies (Order Trichoptera). Bull. Br. Mus. nat. Hist. Ent. 12 (2): 81-121.

Kimmins, D. E. 1962b. A note on Protomacronema pubescens Ulmer (Trichoptera, Hydropsychidae). Entomologist's
mon. Mag. 98: 99-101.

Kimmins, D. E. 1963. On the Trichoptera of Ethiopia. Bull. Br. Mus. nat. Hist. Ent. 13 (5): 119-170.

Kolbe, H. J. 1897. Neuropteriden, Netzfliigler Deutsch-Ost-Afrikas. In: Deutsch-Ost-Afrika 4. Trichoptera: 37-41.

Kolenati, F. A. 1859. Genera et species Trichopterorum. Gen. Spec. Trich. 2. Moscow 1859: 143-296.

Lepneva, S. G. 1964 (1970). Fauna of the U.S.S.R. Trichoptera. Vol. II (1). Larvae and pupae of Annulipalpia. pp. 638.

Akademyia Nauk SSSR. 1964. (Translated from Russian and published for the Smithsonian Institution and the
National Science Foundation, Washington, D C., by the Israel Program for Scientific Translations, 1970.)

Lestage, J. A. 1919a. Les Trichopteres d’Afrique. Catalogue synonymique et systematique des especes connues. Revue
zool. afr. 6: 251-336.

Lestage, J. A. 1919. Notes sur quelques Nevropteres (Ephemeres et Trichopteres) du Conge beige. Revue zool. afr. 6
(2): 193-204.

Mackay, R. J. 1978. Larval identification and instar association in some species of Hydropsyche and Cheumatopsyche
(Trichoptera: Hydropsychidae). Ann. ent. Soc. Am. 71 (4): 499-509.

Malicky, H. 1973. Trichoptera (Kocherfliegen). In Handbuch der Zoologie, Berlin, Vol. IV Arthropoda, (2) Insecta,
Division 2 (ed. by M. Beier, Vienna), No. 29, 114 pp. W. de Gruyter, Berlin.

Malicky, H. 1977. Ein Beitrag zur Kenntnis der Hydropsyche g;<m»a-Gruppe (Trichoptera, Hydropsychidae). Z. Arb-
Gem. ost. Ent. 29 (1/2): 1-28.

Marlier, G. 1943a. Trichoptera. Explor. Parc natn. Albert Miss. G. F. de Witte (1933-1935) 44: 1-17.

Marlier, G. 1943b. Trichoptera. Explor. Parc natn. Albert Miss. H. Damas ( 1935-36 ) 11: 1-34.

Marlier, G. 1943c. Trichopteres du Conge beige. Revue Zool. Bot. afr. 37 (1): 64-88.

Marlier, G. 1947. Trichopteres de la Cote d'Ivoire. Revue fr. Ent. 14 (1): 31-33.

419

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Marlif.r, G. 1957. Sur une petite collection de Trichopteres provenant du Katanga. Bull. Annls. Soc. r. ent. Belg. 93
(9-10): 283-292.

Marlier, G. 1958. Recherches hydrobiologiques au lac Tumba. Hydrobiologiu 10: 352-385.

Marlier, G. 1959. Trichopteres de Sao Tome. Revue fr. Ent. 26 (2): 108-120.

Marlier, G. 1960. Deux synonymies nouvelles dans la Sous-famille des Macronematinae (Trichoptera, Hydropsychi-
dae). Bull. Annls Soc. r. ent. Belg. 96 (11-12): 255.

Marlier, G. 1961. Hydropsychidae du Kivu (Trichoptera). Revue Zool. Bot. afr. 63 (1-2): 158-212.

Marlier, G. 1962a. Notes sur les Trichopteres. Bull. Inst. r. Sci. nat. Belg. 38 (27): 1-9.

Marlier, G. 1962b. Genera des Trichopteres de l’Afrique. Annls Mus. r. Afr. cent. Ser. 8vo 109: 1-263.

Marlier, G. 1964. Trichopteres de l’Amazonie recuillis par le Professeur H. Sioli. Mem. Inst. r. Sci. nat. Belg. Ser. 2
(76): 1-167.

Marlier, G. 1965/66. Las Trichopteres du Musee de Dundo. Publqoes cult. Co. Diam. Angola 72: 14-79.

Marlier, G. 1978a. Sur une collection de Trichopteres de l’Afrique occidentale. Rev. zool. afr. 92 (2) 283-302.

Marlier. G. 1978b. Les larves et nymphes des Trichopteres des Seychelles. Proc. 2nd int. Symp. Trich. : 31-54.

Marlier, G. & Botosaneanu, L. 1968. Trichopteres du Ghana et de la Cote d'Ivoire. Bull. Inst. r. Sci. nat. Belg. 44
(16): 1-16.

Martynov, A. B. 1935. On a collection of Trichoptera from the Indian Museum. Part I, Annulipalpia. Rec. Ind. Mus.
37: 93-209.

McLachlan, R. 1865. Trichoptera Britannica: a monograph of the British species of Caddis-Flies. Trans, ent. Soc.
Lond. (3) 5: 1-184.

McLachlan, R. 1871. On new forms etc., of extra-European Trichopterous insects. J. Linn. Soc. (Zool.) London 11:
98-141.

McLachlan, R. 1872. Materiaux pour une Faune nevrepterologique de l’Asie Septentrionale. Ann. Soc. ent. Belg. 15:
60-71.

McLachlan, R. 1878. A monographic revision and synopsis of the Trichoptera of the European fauna (with Supplement
and First Additional Supplement, iv+ 523+ xii + 103 + 76pp & Vol. of 66 plates, 1874-1880. 1884). London:
John van Voorst.

McLachlan. R. 1880. Ibid. Suppl. II: xiii — lxxxiv. London: John van Voorst.

McLachlan, R. 1884. Ibid. 1st Additional Suppl. 76pp. London: John van Voorst.

Mosely, M. E. 1931. The genus Diplectronella Ulmer (Insecta: Trichoptera). Ann. Mag. nat. Hist. (10) 8: 195-205.

Mosley, M. E. 1932. Contributions a l’etude de la faune du Mozambique. Voyage de M. P. Lesne (1928-1929). 12e
note — Trichoptera. Mems Estud. Mus. zool. Univ. Coimbra Ser. I. No. 64: 1-8.

Mosely, M. E. 1933. A revision of the genus Leptonema (Trichoptera). London: British Museum (Nat. Hist.). 69pp.

Mosely, M. E. 1934. New Fijian and African Trichoptera. Eos. Madr. 9: 17-27.

Mosely, M. E. 1935. New African Trichoptera. Ann. Mag. nat. Hist. (10). 15: 221-232.

Mosely, M. E. 1936. New African Trichoptera — I. Ann. Mag. nat. Hist. (10), 17: 429-451.

Mosely, M. E. 1938. Trichoptera collected in Morocco by Messrs K. H. Chapman & G. A. Bisset. Ann. Mag. nat. Hist.
(11), 1: 271-277.

Mosely, M. E. 1939a. Mission scientifique de l’Omo. Mem. Mus. natn. Hist. nat. Paris. 9: 293-301.

Mosely, M. E. 1939b. Trichoptera. Ruwenzori Exped. 1934-5, 3 (1): 1-40.

Mosely, M. E. 1939c. British Caddis Flies. London: George Routledge & Sons, xiii -F 320pp.

Mosely, M. E. 1948. Trichoptera. Expedition to South-west Arabia 1937-8, 1 (9): 67-85.

Mosely, M. E. Kimmins, D. E. 1953. The Trichoptera (Caddis-flies) of Australia and New Zealand. London: Trustees of
the British Museum. iv+ 550pp.

Navas, L. 1915. Notes sur quelques Nevropteres du Congo beige. III. Revue zool. afr. 4: 172-182.

Navas, L. 1916. Neuropteros nuevos o poco conocidos. 7a serie. Mems R. Acad. Cienc. Artes Barcelona 12 (13): 1-27.

Navas, L. 1917. Deux Trichopteres nouveaux de l’Afrique du Nord. Bull. Soc. Hist. nat. Afr. N.: 118-120.

Navas, L. (1922) 1923. Algunos insectos del Museo de Paris. Revta Acad. Cienc. exact, fls.-quim. nat. Zaragoza 7:
15-51.

Navas, L. 1923. Insecta nova. Ser. 9 & 10. Memorie Accad. pont. Nuovi Lincei (2) 6: 1-27.

Navas, L. 1926. Insecta nova. Serie II. Memorie Accad. pont. Nuovi Lincei (2) 9: 101-110.

Navas, L. 1927. Nevropteres d’Egypte et de la Palestine. 3me Partie (1). Bull. Soc. r. ent. Egypte 19 (1926): 192-216.

Navas, L. 1928a. Insectes nevropteres et voisins de Barbarie. 6e Serie. Bull. Soc. Hist. nat. Afr. N. 19: 183-191.

Navas, L. 1928b. Comunicaciones entomologicas. 10. Mis excursiones cientificas en 1927. Revta Acad. Cienc. fis.-quim.

nat. Zaragoza. 11: 79-137.

Navas, L. 1930a. Insectes du Congo beige, Serie 4. Revue Zool. Bot. afr. 19: 305-336.

Navas, L. 1930b. Insectos del Museo de Paris. 6a Serie. Broteria. Ser. Zool. 26: 120-144.

Navas, L. 1931a. Insectos del Museo de Paris, 8a Serie. Broteria, Ser. Zool. 27: 101-113, 114-136.

Navas, L. 1931b. Insectes du Congo beige. Serie 6. Revue Zool. Bot. afr. 21: 123-144.

Navas, L. 1931c. Insectes du Congo beige. Serie 5. Revue Zool. Bot. afr. 20: 257-279.

NavAs, L. 1932a. Insectes du Congo beige. Serie 7. Revue Zool. Bot. afr. 22: 269-290.

Navas, L. 1932b. Insectos de Berberia. Serie 10. Neuropteros. Boln. Soc. iber. Cienc. nat. 31: 106-122.

420

SCOTT: HYDROPSYCHIDAE (TRICHOPTERA) OF SOUTHERN AFRICA WITH KEYS

Navas, L. 1934a. Tricopters nous. Butll. Inst, catal. Hist. nat. 33: 369-372.

Navas, L. 1934b. Tricopteros nuevos o interessantes. Serie 10 a. Broteria Ser. trimest. Cienc. nat. 3: 81-95.

Navas, L. 1934c. Insectos de Madagascar. Revta. Acad. Cienc. exact, fis.-quim. nat. Zaragoza, 17: 49-76.

Navas, L. 1934d. Decadas de insectos nuevos. Broteria Ser. trimest. Cienc. nat. 3 (l): 20-24.

Navas, L. 1935. Insectos de Madagascar. Revta Acad. Cienc. exact, fis.-quim. nat. Zaragoza. 18: 42-74.

Navas, L. 1936a. Comunicaciones entmologicas. 19. Insectos de Madagascar. 3a serie. Revta Acad. Cienc. exact, fis.-
quim. nat. Zaragoza 19: 100-110.

Navas, L. 1936b. Insectos de Berberia. Serie 12. Boln. Soc. ent. Esp. 18: 77-100.

Nielsen, A. 1942. Entwicklung und Biologie der Trichopteren mit besonderer Beriicksichtigung der Quelltrichopteren
Himmerlands. Arch. Hydrobiol. Suppl. 17: 255-631.

Nybom, O. 1960. Trichoptera of the Cape Verde Islands. Commentat. biol. Helsinki 22 (8): 1-3.

Petr, T. 1970. The bottom fauna of the rapids of the Black Volta River in Ghana. Hydrobiologia 36: 399-418.

Philipson, G. N. 1953. The larva and pupa of Hydropsyche instabilis Curtis (Trichoptera, Hydropsychidae). Proc. R.
ent. Soc. Lond. (A) 28: 17-23.

Philipson, G. N. 1969. Some factors affecting the net-spinning of the caddis fly Hydropsyche instabilis Curtis (Tricho-
ptera, Hydropsychidae). Hydrobiologia 34: 369-377.

Pictet, F. J. 1834. Recherches pour servir a l’histoire et l'anatomie des Phryganides. Rech. Phryg. Geneva, 235pp, (for
entire family + others).

Pictet, F. J. 1836. Description de quelques nouvelles especes de Nevropteres du Musee de Geneve. Mem. Soc. Phys.
Hist. nat. Geneve 7: 396-403.

Rambur, J. P. 1842. Histoire naturelle des Insectes Nevropteres. Hist. nat. Nevr. Paris 17 + 534pp.

Ross, H. H. 1944. The Caddis Flies or Trichoptera of Illinois. Bull. III. nat. Hist. Surv. 23 (1): 326pp.

Ross, H. H. 1967. The evolution and past dispersal of the Trichoptera. A. Rev. Ent. 12: 169-206.

Ross, H. H. & Unzicker, J. D. 1977. The relationships of the genera of American Hydropsychinae as indicated by
phallic structures (Trichoptera, Hydropsychidae). J. Georgia ent. Soc. 12 (4): 298-312.

Sattler, W. 1958. Beitrage zur Kenntnis von Lebensweise und Korperbau der Larve und Puppe von Hydropsyche Piet.
(Trichoptera), mit besonderer Beriicksichtigung des Netzbaues. Z. Morph. Okol. Tiere 47 : 115-192.

Sattler, W. Uber den Korperbau, die Okologie und Ethologie der Larve und Puppe von Macronema Piet. (Hydro-
psychidae), ein als Larve sich von “Mikro-Drift” ernahrendes Trichopter aus dem Amazonasgebiet. Arch. Hy-
drobiol. 59 (1): 26-60.

Sattler, W. 1965. Uber Messvorgange bei der Bautatigkeit von Trichopteren-Larven (Trichoptera). Zool. Anz. 175:
378-386.

Schmid, F. 1968. La Famille des Arctopsychidae (Trichoptera). Mem. Soc. ent. Quebec No 1: 1-84.

Schmid, F. 1979. On some new trends in Trichopterology. Bull. ent. Soc. Can. 11 (3): 48-57.

Schmid, F. 1980. Genera des Trichopteres du Canada et des Etats adjacents. In Les Insectes et Arachnides du Canada
(7): 1-296. Agriculture, Canada.

Scott, D. 1958. Ecological studies on the Trichoptera of the River Dean, Cheshire. Archiv. Hydrobiol. 54: 340-392.

Scott, K. M. F. 1974. The Trichoptera of the Sundays and Fish Rivers, Eastern Cape Province, South Africa. Ann.
Cape Prov. Mus. {nat. Hist.) 9: 223-235.

Scott, K. M. F. 1975. The value of larval stages in systematic studies of the Trichoptera, with particular reference to the
Hydropsychidae from Africa south of the Sahara. Proc. I Congr. ent. Soc. sth. Afr. : 41-52.

Scott, K. M. F. 1976. Caddisfly studies in relation to ecological work. Water Report no. 3, June 1976: 2, 3.

Scott, K. M. F., Allanson, B. R. & Chutter, F. M. 1972. Research Report: Orange River Project. Working Group
for ORP Hydrobiology of the Fish and Sundays Rivers. CS1R Research Report 306: 1-6.

Sharp, B. 1976. Observations on ventilatory and other behavioural activities of larvae of Cheumatopsyche maculata
(Trichoptera, Hydropsychidae). Experimental Project for Entomology III, Rhodes University, South Africa.

Silfvenius, A. J. 1905. Beitrage zur Metamorphose der Trichopteren. Helsingfors. Acta Soc. Fauna Flora Fenn. 27,
No. 6: 1-168.

Siltala, A. J. 1907. (Formerly Silfvenius). Trichopterologische Untersuchungen No. 2. Uber die postembryonale Ent-
wicklung der Trichopteren-Larven. Zool. Jb. Suppl. 9: 309-626.

Statzner, B. 1975. Funktionsmorphologische Studien am Genitalapparat von drei neuen Cheumatopsyche- Arten
(Trichoptera-Hydropsychidae). Zool. Anz. 193 (5/6): 382-398.

Statzner, B. 1976. Die Kocherfliegen-Emergenz (Trichoptera, Insecta) aus dem zentralafrikanischen Bergbach Kalen-
go. Arch. Hydrobiol. 78: 102-137.

Statzner, B. 1981. A progress report on Hydropsychidae from the Ivory Coast: Characters for the specific identification
of larvae and population dynamics of four abundant species. Proc. 3rd int. Symp. Trichoptera: 329-335.

SVkora, J. 1964. New species of Leptonema Guerin from Madagascar (Trichoptera). Tijdschr. Ent. 107 (5): 277-282.

Tobias, W. 1972. Zur Kenntnis europaischer Hydropsychidae (Insecta: Trichoptera). II. Senckenbergiana biol. 53 (3/4):
245-268.

Turnbull-Kemp, P. St. J. 1970. Trout in Rhodesia. Rhodesia agric. J. 67 (6) (reprinted as Bulletin 2564: 1-6).

Ulmer, G. 1904a. Uber westafrikanische Trichopteren. Zool. Anz. 28: 353-359.

421

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 14 PT 8, JUNE 1983

Ulmer, G. 1904b. Uber die von Herrn Prof. Yngve Sjostedt in Kamerun gesammelten Trichopteren. Ark. Zool. 1:
411-423.

Ulmer, G. 1905a. Neue und wenig bekannte Trichopteren der Museen zu Brussel und Paris. Ann. Soc. ent. Belg. 49:
17-42.

Ulmer, G. 1905b. Neue und wenig bekannte aussereuropaische Trichopteren. hauptsachlich aus dem Wiener Museum.
Annin naturh. Mus. Wien. 20: 59-98.

Ulmer, G. 1905c. Zur Kenntnis aussereuropaischer Trichopteren (Neue Trichopteren des Hamburger und Stettiner
Museums und des Zoologischen Instituts in Halle, nebst Beschreibung einiger Typen Kolenati’s und Burmei-
ster’s). Stettin, ent. Ztg. 66: 1-119.

Ulmer, G. 1906. Neuer Beitrag zur Kenntnis aussereuropaischer Trichopteren. Notes Leyden Mus. 28: 1-116.

Ulmer, G. 1907a. Trichoptera. In: Wytsman, P., ed.. Genera Insect. 60: 1-259.

Ulmer, G. 1907b. Neue Trichopteren. I. Exotisches Material. Notes Leyden Mus. 29: 1-53.

Ulmer, G. 1907c. Trichopteren. Collections Zoologiques Baron Edm. de Selys Longchamps. Catalogue systematique et
descriptif. Bruxelles, Hayez. Fasc. 6 (1): 102 pp.

Ulmer, G. 1907d. Trichopteren. Ibid. 6 (2): 1-121.

Ulmer, G. 1909. Trichopteren von Madagascar und den Comoren. In: Voeltzkow, A. ed.. Reise in Ostafrika in den Jah-
ren 1903-1905 , 2 Stuttgart. E. Schweitzbart, pp. 357-363.

Ulmer, G. 1910. Trichoptera. In The Percy Sladen Trust Expedition to the Indian Ocean in 1905, 3. Trans. Linn. Soc.
Lond. (Ser. 2, Zool.) 14: 41-54.

Ulmer, G. 1912. Trichopteren von Aquatorial-Afrika. Wiss. Ergebn. dt. ZentAfr.-Exped. 4, Zool. 11:81-125.

Ulmer, G. 1913. South African Trichoptera. Ann. S. Afr. Mus. 10: 189-191.

Ulmer, G. 1924. Trichopteren und Ephemeropteren. In: Wissenschaftliche Ergebnisse der mit Unterstiitzung der Aka-
demie der Wissenschaften in Wien aus der Erbschaft Treitl von F. Werner unternommenen zoologischen Ex-
pedition nach dem Anglo- Agyptischen Sudan (Kordofan) 1914. 12. Denkschr. Akad. Wiss., Wien 99: 1-9.

Ulmer, G. 1928. Ueber die Trichopterengattung Hydromanicus und Verwandte. Konowia 1: 314-318.

Ulmer, G. 1930. Trichoptera und Ephemeroptera. In: Scott. H. & Omer-Cooper. J., Entomological Expedition to
Abyssinia 1926-1927. Ann. Mag. nat. Hist. (10), 6: 479-511.

Ulmer, G. 1931. Trichopteren von Afrika (hauptsachlich aus dem Britischen Museum). Dt. ent. Z. (1): 1-29.

Ulmer. G. 1951. Kocherfliegen (Trichopteren) von den Sunda-Inseln (Teil 1). Arch. Hydrobiol. ISuppl. 19: 1-528.

Ulmer, G. 1957. Ibid. (Teil 3). Arch. Hydrobiol. ISuppl. 23: 109-470.

Ulmer, G. 1963. Trichopteren (Kocherfliegen) aus Agypten. Arch. Hydrobiol. 59: 257-271.

Walker, F. 1852. Catalogue of the specimens of Neuropterous Insects in the Collection of the British Museum. Part 1:
1-135. London.

Wallace, J. B. 1975. Food-partitioning in net-spinning Trichoptera larvae: Hydropsyche venularis, Cheumatopsyche
etrona and Macronema zebratum (Hydropsychidae). Ann. ent. Soc. Am. 68 (3): 463-472.

Wallace, J. B. 1976. A progress report on the North American Macronema larvae, their retreats, food and feeding nets
(Trichoptera, Hydropsychidae). Proc. 1st int. Symp. Trich.: 145-146.

Wallace, J. B. & Sherberger, F. F. 1974. The larval retreat and feeding net of Macronema Carolina Banks (Tricho-
ptera, Hydropsychidae). Hydrobiologia 45: 177-184.

Wallace, J. B. & Sherberger, F. F. 1975. The larval dwelling and feeding structure of Macronema transversum (Walk-
er) (Trichoptera: Hydropsychidae). Anim. Behav. 23 (3): 592-596.

Wallengren, H. D. J. 1891. Skandinaviens Neuroptera. Andra afdelningen. Svenska Ak. Hand., 24(10): 173 pp.

Walker, F. 1852. Catalogue of the specimens of Neuropterous Insects in the collections of the British Museum , Part 1:
1-135. London.

Weidner. H. 1964. Die entomologischen Sammlungen des Zoologischen Staatsinstituts und Zoologischen Museums
Hamburg. Mitt. hamb. zool. Mus. Inst. 62: 55-100.

Westwood, J. O. 1840. An introduction to the modern classification of Insects. Vol. 2. London: Longman.

Wiggins, G. B. 1960. The unusual pupal mandibles in the Caddisfly Family Phryganeidae (Trichoptera). Can. Ent. 92
(6): 449-457.

Wiggins, G. B. 1977. Larvae of the North American Caddisfly Genera (Trichoptera). Toronto: University of Toronto
Press, pp. xi + 401.

422

_j Z _J “J Z _j

liSIMI NVIN0SH1IWS S3!HVaan LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSH1IW!

z

00

g) >

30

m w 00 n^£uu^x m :z m

C/> V, CO — CO — to

R1ES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH11WS S3iHVd8n LIBRARIES SMITHSONIAI

CO z Sj- (S> Z _ CO z

2 .x^Tx < JR’ • 2 < s

^ Z

> ^pr- 2 ‘^\ > X^msSX

_ 00 z <0 z co v z

I1SNI NVINOSH1IIAIS S3ldVdan LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSH1IW

co 3^ co ~ co

q o dcx — -•^r" O dsz _ O

R I ES^SMITHSOmAN~*INSTITUTION NOliniliSNl^NyiNOSHlI WS S3ldVd8n LIBRARI ES^SMITHSONIA
Z r* z r^»z r* Z

o ~ O z^5Tin7>v _ xs. O x35firS>v _ xo’qa/Sx O

pn xtjyo^ix " x^vas>^x m >■ ^ pri

| IS N 1 NVINOSHIIINS^S 3 I HVd an“LI BRAR I ES SMITHSONIAN'lNSTITUTION^NOliniilSNI NVIN0SH1I/

Z CO z s CO Z V to z

xCTTnx 2 aA:. *£ 'x s! /C5^Tlr<o5x *£

<

w JWJ?* £ 0*S& o life Jl £ oSfc £ Mr ffi' x

s >• 5 2 >- 2 N^ofSgiy

R I ESWSMITHSONIAN2!NSTITUTION°0NOIinillSNI_NVINOSHlHAlSC/>S3 I BVaa Il^LI BRAR I ES^SMITHSONIA

00 = CO -7> \ CO — CO

to

co /o^wcs^gX — /^isw&y-t\ _ 1 —

C£ '

<

_ _ V —1 Qt

/ * ij 'vv5?4WJK/ m ^ CO c US r

c_> x^yos hjx

iisni-nvinoshims ^saiavaan ""libraries Smithsonian institution NoiinmsNi nvinoshiiw

z r* __ z r*

UJ

m w v^uu-rx m x^n^x — m x; x^uus^x m

j, \ x co — co £: co

R I ES^SMITHSONIAN- INSTITUTION NOlinillSNI NVINOSHIIWS S3IUVd8n LIBRARIES SMITHSONI/
co z ,v. co z ^ w z -,•• co

< *&N\. 2 ..< S - <

x

CO

Vf€AkP?/ - " ^

^iu.y*x ^ > :'#’5sr 2

_ ^5 v z co z co -e.

LiisNi nvinoshiiws saiavaan libraries smithsonian^institution NoiiniiiSNi nvinoshiii/

— CO 5 ^ co 2 C^ “

cci XCvSwTXv /A ^ -CO y'S'iJSi'C^X 111 ZiV

‘ O x^A),pc.x — O _ ^

sR I ESZSMITHSOmAN“‘lNSTITUTION NOlinillSNI “‘NVINOSHimS S 3 I d V H 8 13 LIBRARIES SMITHSONI)

-yr J Z S j* Z I Z

O xCTvrmT^v ~ O xSswij?x ~ xoa'v7?x O

m -■ w z; x^vAsq>- pn

_ co — co

LIISNI _ NVIN0SH1IIAIS SSIHVHai! LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIII
2: 00 2 ^2 - z: v co

o

M Z "" _l 2 _

aan libraries Smithsonian institution NoiiniusNi nvinoshiiws saiavaan librarie

~~y

2 .-m /$S§S*!§K 2 xSSSTSx rn 2 ^

rcTHaAKpy t- \i*4LA\Wg/ _ >' ’nkj, VoXsMjsxy — KcJUMU^ss-j

m N^ot^ ” • m 'S2^£S>' 2 rn ~ Xguiicsx m

co ± co — to \ ± co

UTION NOliniliSN! NVINOSHillNS S3iaVa8IT LIBRARIES SMITHSONIAN INSTITUTION NOIlHlIlSi

CO Z CO 2 E? 2 .v CO

k

- t ^ I

s 'ciEii/ > 2 > x^ongx s >

* z ■ to 2 CO * 2 CO * 2

yell LIBRARIES SMITHSONIAN INSTITUTION NOliniilSNI NVIN0SH1IWS S3 I ava a I1_ LI 6 RAR I E
— <05 _ to — to

2 \ «.)) ><visc WX SjJ 2iwi>v ^

CO

V\ q:

i <>,

cc

O “ W 5 xo^dc^x “ o ^ " Xoo^r^x o

2 . Z — 1 z *t i 2

JTION NOlirUliSNI NVIN0SH1IIAIS S3iaVHen LIBRARIES SMITHSONIAN INSTITUTION NOlifUliS

z r~ Z r* z [_ 21

o x£?™<ox “ ™ 9 m zS'SSsX 2

CD

SJ /Si

3“* W'i

jo m rn ^ * m

aan_LIBRARi ESC/5SMITHSONlAN~INSTlTUTIONC°NOIinili$Ni“NVINOSHilWS SBIbVdail LIBRARIE

z <0 z « co z co z

< >k g ^ g

18k ^ 4a 9 z 9 ^

&.. 1 1 #4 1 'W

CO ($(£ to jflwj&’f' £

> ’’ 5 > ' 2E

to ' " Z CO ' Z CO 2 CO

JTION NOliniilSNI NVINQSHillNS S3iavaaiT LIBRARIES SMITHSONIAN INSTITUTION NOlinillS

CO 5 \ <0 „ _ =£ to — _ ^ 9 ^-rr^

0

w W — — "

_ 2 »j 2 ~i

aan libraries Smithsonian institution NoiiniusNi nvinoshiiws saiavaan ubrarii

r- 2 f z r- _ z “

m — " CO ± " CO V £ CO

JTION NOliniilSNI NVIN0SH1IIA1S S3 1 ava a 11 L I B R AR I ES SMITHSONIAN INSTITUTION NOlifUI-LS

2 ~ 2 .1 .xK., 1 i

vV\

\

31 to

t/sl o yw//v%F£

v?m &&/ 2 t xx' 'N^ 2 \i*CALjZ£/ t ^ xxx 2

2 PCX J,, ’yypr 2 X > Xopi^X 2 >

yan^LIBRARIES^SMITHSONIAN INSTITUTION NOliniilSNI NVIN0SH11 WS^SS I a V a 3 H LJBRARM

M ' 8 — ^ — co — </> —

x LU
X —

a 0=

_ ^ 4 <

01 -1 vc'tt

q XQm dcJX “ X^uTIs^X q NX " X^iiTIs^X O

ution ^ NoiiniusNi^NviNOSHiiws^sa 1 ava a n librari es^smithsonian^institution Noiinms

Z l~ , z f— Z t- 2

rn X^vosvEX 2 m \CNOS>tV

aan~ LIBRARI ES</>SMITHSONIAN-|NSTITUTIONWNOlinillSNI~NVINOSHllWS S3IHVaan~LIBRARII

Z CO Z CO Z to 2